JOURNAL OF SHELLFISH RESEARCH VOLUME 18, NUMBER 1 JUNE 1999 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. Long Island University Southampton, NY 11968 Dr. Standish K. Allen, Jr. (2000) School of Marine Science Virginia Institute of Marine Science Gloucester Point. V A 23062- 1 1 346 Dr. Peter Beninger (1999) Laboratoire de Biologie Marine Faculte des Sciences Universite de Nantes BP 92208 44322 Nantes Cedex 3 France Dr. Andrew Boghen (1999) Department of Biology University of Moncton Moncton, New Brunswick Canada El A 3E9 Dr. Neil Bourne (1999) Fisheries and Oceans Pacific Biological Station Nanaimo, British Columbia Canada V9R 5K6 Dr. Andrew Brand (1999) University of Liverpool Marine Biological Station Port Erin, Isle of Man Dr. Eugene Burreson (1999) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Peter Cook (2000) Department of Zoology University of Cape Town Rondebosch 7700 Cape Town, South Africa EDITORIAL BOARD Dr. Simon Cragg (2000) Institute of Marine Sciences University of Portsmouth Ferry Road Portsmouth P04 9LY United Kingdom Dr. Leroy Creswell (1999) Harbor Branch Oceanographic Institute US Highway 1 North Fort Pierce. Florida 34946 Dr. Lou D'Abramo (2000) Mississippi State University Dept of Wildlife and Fisheries Box 9690 Mississippi State, Mississippi 39762 Dr. Ralph Elston (1999) Battelle Northwest Marine Sciences Laboratory 439 West Sequim Bay Road Sequim, Washington 98382 Dr. Susan Ford (2000) Rutgers University Haskin Laboratory for Shellfish Research P.O. Box 687 Port Norris, New Jersey 08349 Dr. Raymond Grizzle (1999) Randall Environmental Studies Center Taylor University Upland, Indiana 46989 Dr. Mark Luckenbach (1999) Virginia Institute of Marine Science Wachapreague, Virginia 23480 Dr. Bruce MacDonald (2000) Department of Biology University of New Brunswick P.O. Box 5050 Saint John, New Brunswick Canada E2L 4L5 Dr. Roger Mann (2000) Virginia Institute of Marine Science Gloucester Point. Virginia 23062 Dr. Islay D. Marsden (2000) Department of Zoology Canterbury University Christchurch, New Zealand Dr. Tom Soniat (2000) Biology Department Nicholls State University Thibodaux, Louisiana 70310 Dr. J. Evan Ward (2001) Dept. of Marine Sciences University of Connecticut Groton, CT 06340-6097 Dr. Gary Wikfors (2000) NOAA/NMFS Rogers Avenue Milford, Connecticut 06460 Journal of Shellfish Research Volume 18, Number 1 ISSN: 0077571 1 June 1999 Journal of Shellfish Research, Vol. 18. No. 1. 1-3. 1999. MBLW/HOl Library .; DEC 0 6 2007 WOODS HOLE Massachusetts 02543 IN MEMORIAM DR. L. EUGENE CRONIN 1917-1998 Dr. L. Eugene Cronin. NSA member and NSA President from 1959 to 1961, died at his home in Annapolis, MD on December 18, 1998 at the age of 81. He was known nationally and internationally for his efforts on behalf of estuarine research, resource management, and habitat restoration. A native of Aberdeen. MD. Gene Cronin received an AB in Chemistry from Western Maryland College ( 1938). and MS ( 1943l and PhD ( 1946) degrees in Zoology (blue crab biology) from the University of Maryland. He was a high school Biology teacher from 1938 to 1940 in Bel Air. MD. From 1943 to 1950. he was a biologist at Chesapeake Biological Laboratory (CBL) in Solomons. MD with a focus on biology and management of blue crabs, oysters, and fish. He subsequently spent five years with the University of Delaware's Department of Biological Sciences. There he established a marine laboratory in Lewes. DE. beginning in rented space in the local high school. In 1952, he and his staff of three moved into a former restaurant while making plans for a permanent research building; a converted fishing-party boat served as the University's first research vessel. Acartia. By the time the new research building was dedicated in 1955, Gene had laid the foundation for a program in research and education that has since grown into the University of Delaware Graduate College of Marine Studies. Gene returned to Maryland in 1955 as Director of the Maryland Department of Research and Education, with headquarters at CBL. In 1961. the Department became the Natural Resources Institute (NRI). a 4-laboratory entity within the University of Maryland. In 1975. NR1 merged with the newly-established Horn Point Environmental Laboratories (now Horn Point Laboratory) to form The University of Maryland Center for Environmental and Estuarine Studies (now The University of Maryland Center for Environmental Science). In 1977. Gene became Director of the Chesapeake Research Consortium that coordinates Chesapeake Bay-related research among the area's major research universities and institutions. In 1984, he "retired" to become a consultant, while remaining active in conservation organizations such as The Chesapeake Bay Foundation and The Alliance for the Chesapeake Bay. In Memoriam: L. Eugene Cronin ,J During the course of his career. Gene was involved with the governance of a number of scientific societies. In addition to being President of The National Shellfisheries Association as noted above, he was the first President ( 1949) of the Atlantic Estuarine Research Society (AERS). His continued interactions with AERS and estuarine scientists led to his involvement in the founding of the Estuarine Research Federation (ERF), of which he was Founding President from 1971 to 1973. He was a Fellow Emeritus of the American Institute of Fishery Research Biologists. Within Maryland, he was Director of the Maryland Biology Teachers Association in his early career and a Trustee of The Natural History Society of Maryland. Gene made numerous contributions towards understanding and conserving natural resources by serving on many State and Federal committees and commissions. He was an advisor to the Atlantic States Marine Fisheries Commission from 1955 to 1969, served on a variety of committees of the National Academy of Sciences and of the US Department of State, and sat on the Marine Board of the National Research Council. Through these activities, he played important roles in influencing fisheries policies in the Chesapeake Bay region. He had a key role in involving the US Environmental Protection Agency in a 5-year study of Chesapeake Bay and in the subsequent establishment of the Chesapeake Bay Program, a state-federal initiative. His efforts brought him many awards. In 1994 he became the second recipient of the Mathias Medal. This biennial award "In Recognition of Scientific Excellence" is sponsored by the Sea Grant Programs of Maryland and Virginia and the Chesapeake Research Consortium to honor those who work to enhance understanding of the Chesapeake Bay ecosystem and to encourage application of this knowledge to solving environmental problems in the Bay. In 1997, ERF created an award in his name to recognize outstanding contributions to estuarine research by a young scientist, a fitting recognition of Gene's contributions to ERF. Other awards included recognition by the Oyster Institute of North America (1967), Chesapeake Bay Seafood Industries Association (1968), and the Isaac Walton League of America ( 1990). The Governor of Maryland appointed him as "Admiral of the Chesapeake Bay" in 1987. Gene remained active professionally to within a few weeks of his death. He was one of seven distinguished scientists who participated in a workshop in November 1998 for university students interested in learning how research, policy, and management have been applied to Chesapeake Bay problems in the past. He was also involved in co-editing a volume on the biology of the blue crab. Gene is survived by his wife, Alice, herself a chemist and founding member of AERS, three sons, and four grandchildren. Victor F. Kennedy Horn Point Environmental Laboratory PO Box 775 Cambridge, MD 21613 PUBLICATIONS Cronin. L.E. 1947. Anatomy and histology of the male reproductive system of Callinectes sapidus Rathbun. J. Morphol. 81:209-239. Cronin, L.E. 1949. Comparison of methods of tagging the blue crab. Ecology 30:390-394. Cronin, L.E.. J.C Daiber & E.M. Hulbert. 1961. Quantitative seasonal aspects of zooplankton in the Delaware River estuary. Chesapeake Set. 3:63-93. Cronin. L.E. 1967a. The role of man in estuarine processes. Pages 667-689 in G.H. Lauff (ed.). Estuaries. Publication 83. American Association for the Advancement of Science. Washington DC. In Memoriam: L. Eugene Cronin 3 Cronin. L.E. 1967b. The condition of the Chesapeake Bay. Transactions North American Wildlife and Natural Resources Conference 32:137-150. Cronin. L.E. & D.A. Flemer. 1%7. Energy transfer and pollution. Pages 171-183 in T.A. Olson and F.J. Burgess (eds.). Pollution and Marine Ecology. Enter/science Publications. New York. Cronin. L.E. 1971. IV. Prevention and monitoring. Pollution prevention. Proc. Roy. Soc. Loud. B. 177:439-450. Cronin, L.E. & A. J. Mansueti. 1971. The biology of the estuary. Pages 14—39 in A Symposium on the Biological Significance of Estuaries. Sport Fishing Institute. Washington DC. Mihursky. J. A. & L.E. Cronin. 1973. Balancing needs of fisheries and energy production. Transactions North American Wildlife and Natural Resources Conference 38: 459-476. Wiley. M.L.. T.S.Y. Koo & L.E. Cronin. 1973. Finfish productivity in coastal marshes and estuaries. Pages 139-150 in R.H. Chabreck (ed.). Proceedings of the Coastal Marsh and Estuary Management Symposium. Louisiana State University. Baton Rouge LA. Cronin. L.E.. D.W. Pritchard. T.S.Y. Koo & V. Lotrich. 1976. Effects of enlargement of the Chesapeake and Delaware Canal. Pages 18-32 in M.L. Wiley (ed.). Estuarine Processes, Volume II. Academic Press, New York. Tsai, C.-F., J. Welch. K.-Y. Chang. J. Schaeffer & L.E. Cronin. 1979. Bioassay of Baltimore Harbor sediments. Estuaries 2:141-153. Cronin. L.E. 1981. The Chesapeake Bay. Transactions of the North American Wildlife and Natural Resources Research Conference 46:223-229. Cronin. L.E. 1982. Pollution in the Chesapeake Bay: A case history and assessment. Pages 17— 16 in T.W. Duke (ed.). Impact of Man on the Coastal Environment. US Environmental Protection Agency Publication EPA-600/8-82-021. Washington DC. Roberts, M.H., Jr., J.E. Warriner. C.-F. Tsai, D. Wright & L.E. Cronin. 1982. Comparison of estuarine species sensitivities to three toxicants. Arch. Environm. Contain. Toxicol. 11:681-692. Cronin. L.E. & R.B. Biggs. 1981. Special characteristics of estuaries. Pages 3-23 in B.J. Neilsen & L.E. Cronin (eds.). Estuaries and Nutrients. Humana Press, Clifton NJ. Officer. C.B., L.E. Cronin. R.B. Biggs & J.H. Ryther. 1981. A perspective on estuarine and coastal research funding. Env. Sci. Tech. 15:1282-1285. Officer. C.B.. R.B. Biggs. J.L. Taft, L.E. Cronin, M.A. Tyler & W.R. Boynton. 1984. Chesapeake Bay anoxia: Origin, development and significance. Science 223:22-27. Cronin, L.E. 1985. Chesapeake Bay: Productive? Polluted? Planned? Pages 339-358 in N.L. Chao & W. Kirby-Smith (eds.). Proceedings of the International Symposium on the Utilization of Coastal Ecosystems: Planning. Pollution and Productivity. Fundacao Universidade do Rio Grande. Brasil. Cronin. L.E. 1986. Chesapeake fisheries and resource stress in the 19th Century. J. Wash. Acad. Sci. 76:188-198. Cronin. L.E. 1987. Actions needed to reduce contamination problems impairing Chesapeake Bay fisheries. Pages 555-566 in S.K. Majundar. L.W. Hall & H.M. Austin (eds.). Contamination Problems and Management of Living Chesapeake Bay Resources. Pennsylvania Academy of Science. Easton PA. SELECTED TECHNICAL REPORTS Cronin. L.E. 1942. A histological study of the development of the ovary and accessory organs of the blue crab. Callinectes sapidus Rathbun. MS thesis. University of Maryland, College Park, MD. 37 p. Cronin, L. E. 1947. Anatomy and histology of the male reproductive system of Callinectes sapidus Rathbun. PhD dissertation. University of Maryland, College Park. MD. 71 p. Cronin, L. E. 1950. The Maryland crab industry. 1949. Chesapeake Biological Laboratory Pub. 84. 41 p. Pyle. R. & L.E. Cronin. 1950. The general anatomy of the blue crab. Chesapeake Biological Laboratory Pub. 87. 40 p. Cargo. D. G. & L. E. Cronin. 1951. The Maryland crab industry. 1950. Chesapeake Biological Laboratory Pub. 92. 23 p. Cronin. L. E., W. A. Van Engel. D. G. Cargo & F. J. Wojcik. 1957. A Partial Bibliography of the Genus Callinectes. Virginia Fisheries Laboratory Spec. Sci. Rep. 8: Maryland Dept. Research and Education Ref. 57-26. 21 p. Cronin. L.E.. M.G. Gross, M.P. Lynch & K.J. Sullivan. 1977. The condition of the Chesapeake Bay — A consensus. Pages 37-61 in the Proceedings of the Bi-State Conference on the Chesapeake Bay. Chesapeake Research Consortium Pub. 61. Shady Side MD. Cronin, L. E. 1988. Report of the Chesapeake Bay Blue Crab Management Workshop. November 9-10, 1997. Waldorf. MD. 68 p. EDITED WORKS Cronin. L.E. (ed.). 1975. Estuarine Research. Volumes I and II. Academic Press. NY. 738 and 587 p. Neilson. B.J. & L.E. Cronin (eds.). 1981. Estuaries and Nutrients. Humana Press. Clifton NJ. 643 p. Cronin, L.E. (ed.). 1982. Chlorine - Bane or Benefit? Proceedings of a Conference on the Uses of Chlorine in Estuaries. Chesapeake Research Consortium. Shady Side MD. 212 p. Cronin, L.E. (ed.). 1983. Ten Critical Questions for Chesapeake Bay in Research and Related Matters. Chesapeake Research Consortium Pub. 1 13. Shady Side, MD. 156 p. + 1 Appendix Journal oj Shellfish Research. Vol. 18. No. 1. 5-7. ITO IN MEMORIAM TERRANCE HENRY BUTLER 1923-1998 Terrance (Terry) Henry Butler, a recognized world authority in the field of crustacean biology and crustacean fisheries, passed away on March 10, 1998 in Nanaimo, British Columbia, Canada. He was 74-years-old. Terry was born in Nelson, British Columbia, where his father worked as an engineer on the Canadian Pacific ferries. He received his elementary and secondary education in the interior of British Columbia (B.C.) and, prior to World War II, entered the engineering department of Victoria College in Victoria, B.C. In 1942 he joined the Canadian army, enlisting in the engineer corps. He transferred to the Canadian Airborne Regiment, with whom he served overseas, and was stationed in the Netherlands by the end of the War. He was discharged from the army in January 1946. Upon returning to Canada, he enrolled in the University of British Columbia where he completed a BA degree in Honors Biology in 1949. He received an MA degree from the same University in 1953; the subject of his thesis was a study of Dungeness crab biology and fisheries. During the summer of 1948. Terry had been hired as a summer student at the pacific Biological Station in Nanaimo. B.C.. where he served as a groundfish technician conducting biological sampling of commercial groundfish catches. Thus began an association with the Biological Station that lasted for the remainder of his working career. In May 1949, he accepted a permanent position at the Pacific Biological Station as a research scientist in charge of the Crab and Shrimp Investigation. He continued in this capacity, and also served for a period as head of the Shellfish Investigation, until he retired from the Pacific Biological Station in March 1984. During his working career, Terry undertook research studies on a wide range of subjects that included basic biology, population dynamics and stock assessment of some British Columbia crustacean species, exploratory fishing for harvestable stocks, and improve- ments in fishing methods. He became a recognised authority on crustacean fisheries of the West Coast of North America. As a result of his work, several crustacean fisheries were started in British Columbia. Throughout his working career, Terry maintained a close working relationship with the industry and was frequently consulted by industry for his advice. Terry published widely in the field of crustacean biology and fisheries, having over 55 publications including his "Shrimps of the Pacific Coast of Canada." which has remained a classic reference work. He received the Queen's Silver Jubilee Medal in 1977 in recognition of his career. 6 In Memoriam: Terrance Henry Butler Terry also had an active interest in crustacean fisheries in developing countries. From January 1957 to November 1958, he served with the Food and Agriculture Organization of the United Nations in Indonesia, assisting in the development and management of local shrimp fisheries. In 1988. he served with the Canadian Executive Services Organization to assist in the development of a southern king crab fishery in Chile. After his retirement in 1 984, Terry continued to work several days a week at the Pacific Biological Station, even when his health was failing, to continue with his studies and publications on British Columbia crustaceans. At the time of his passing, he was writing a major work, "The Crab Fisheries of British Columbia." Terry was highly respected for his research accomplishments and was an inspiration to younger scientists at the Pacific Biological Station and elsewhere. He was a kind and patient man who always had time to talk with younger staff members, to encourage them in their work, and to give them the benefit of his long years of experience. Terry had many outside interests. He was keenly interested in plants, shrubs, and trees, and always claimed that he was a frustrated botanist. Wherever he travelled, he was very interested in the local fauna and how it compared to that in British Columbia. He had a great appreciation of classical music and he was an enthusiastic and much respected golfer and a regular member of a Saturday morning golfing group from the Pacific Biological Station. Terry married D. Joan Abel in 1947; they celebrated their golden wedding anniversary in 1997. Together, they had six children, nine grandchildren, and a great granddaughter. Terry was a warm-hearted, kind and generous companion. Those who knew him appreciated his understanding, his guidance and his example and these qualities served to inspire his associates and to assist them in their research and in their daily lives. Terry will be sorely missed by his family and by all that knew him. James Boutillier and Neil Bourne Fisheries and Oceans Pacific Biological Station Nanaimo. British Columbia CANADA V9R 5K6 PARTIAL PUBLICATION LIST FOR T.H. BUTLER Butler. T. H. 1949. The status of the pink shrimp (Prawn). Pandalus bo- realis Kroyer, in the commercial shrimp fishery of English Bay British Columbia. B.A. Thesis. Univ. of British Columbia, 37 p. Butler. T. H. 1950. The commercial shrimps of British Columbia. Fish. Res. Bd. Canada. Pacific Coast Sta., Prog. Rep. 83:30-34. Butler. T. H. 1950. Two records of shrimps from English Bay. B.C. Can. Field Nat. 64(5): 188 p. Butler. T. H. 1951. The 1949 and 1950 tagging experiments in the Graham Is. Crab fishery. Prog. Rep. Pac. Coast Stations, Fish. Res. Bd. Can. 89:88-87. Butler. T. H. 1953. The appearance of a new commercial shrimp in a newly developed shrimp fishery. Fish. Res. Board Can. Progr. Rep. Pac. Coast Stn. 94:30-31. Butler. T. H. 1953. A shrimp survey by the investigator No. 1". April. 1953. Fish. Res. Bd. Canada. Nanaimo Biol. Sta. Circular 55. 4 p. Butler. T. H. 1953. The Life of the Commercial Crab. Western Fisheries. January, 1953. Butler. T. H. 1954. Food of the commercial crab in the Queen Charlotte Islands Region. Canada. Fish. Res. Bd.. Pac. Prog. Rep. No. 99. pp 3-5. Butler. T. H. 1955. Re-discovery of the parasitic cirripede. Mycetomorpha vancouverensis Potts. In British Columbia waters. J. Parasit. 41(3): 321. Butler. T. H. 1956. The distribution and abundance of early post-larval stages of British Columbia commercial crab. Fish. Res. Bd. Canada. Pacific Prog. Rep., No. 107. pp. 22-23. Butler. T. H. 1956. A first British Columbia record of a cragonid shrimp. Can. Field Nat. 70(30): 142. Butler. T. H. 1957. The tagging of the commercial crab in the Queen Charlotte Islands region. Fish. Bd. Canada. Pacific Prog. Rep.. No. 109, pp. 16-19. Butler, T. H. 1959. Results of shrimp trawling by investigator No. 1', June 1959. Fish. Res. Bd. Canada. Nanaimo Biological Station Circular, No. 55. 7 p. Butler. T. H. 1960. Maturity and breeding of the Pacific edible crab. Can- cer magister Dana. J. Fish. Res. Bd. Canada. 1 7(51:641-646. Butler. T. H. 1961. Records of decapod Crustacea from British Columbia. Can. J. Zool. 39:59-62. Butler, T. H. 1961 . Growth and age determination of the Pacific edible crab Cancer magister Dana J. Fish. Res. Board Can. 18:873-891. Butler. T H. 1963. An improved prawn trap. Fish. Res. Bd. Canada. Nan- aimo Biol. Station Circular. No. 67. 7 p. Butler. T. H. 1963. The prawn Pandalus platyceros Brandt 1851. F. A. O. Fish. Rep.. 57(4). Butler. T. H. 1964a. Records of shrimps (Order Decapoda) from British Columbia. J. Fish. Res. Board Can. 21:419-121. Butler. T. H. 1964b. Redescription of the parasitic isopod Holophrysxus alaskenis Richardson And a note on it synonymy. J. Fish. Res. Board Can. 21:971-976. Butler. T. H. 1964c. Growth, reproduction, and distribution of pandalid shrimps in British Columbia. J. Fish. Res. Board Can. 21:1403-1452. Butler. T. H. 1967. A bibliography of the Dungeness crab. Cancer magister Dana. Fish. Res. Bd. Canada. Tech. Rep. 1. 12 p. Butler. T. H. 1967. Shrimp exploration and fishing in the Gulf of Alaska and Bering Sea. Fish. Res. Rd. Canada, Tech. Rep.. 18. 49 p. Butler. T. H. 1968. The shrimp fishery of British Columbia. FAO Fish. Rep. 57(21:521-526. Butler. T. H. 1969. Catch and effort statistics on the Canadian shrimp fishery on the Pacific Coast in 1967 Fish. Res. Bd. Canada. Manuscr. Rept. Ser. No. 1031.4 p. Butler. T. H. 1970. Synopsis of biological data on the prawn Pandalus platyceros Brandt, 1851. FAO Fish. Rep. 57(4): 1289-1315. Butler. T. H. 1971a. A review of the biology of the pink shrimp, Pandalus borealis Kroyer 1838. Can, Fish. Rep. 17:17-24. Butler. T. H. 1971b. Eualus berkeleyorum n.sp., and records of other cari- dan shrimps (Order Decapoda) from British Columbia. J. Fish. Res. Board Can. 28:1615-1620. Butler. T H. 1980. Catch and effort statistics of the Canadian shrimp fishery on the Pacific Coast in 1979. Can. Data Rep. Fish. Aquat. Sci. 224. 5 p. Butler. T. H. 1980. Shrimps of the Pacific coast of Canada. Can. Bull. Fish. Aquat. Sci. 202. 280 p. Butler, T. H. 1981. Dungeness Crab: A Primer. Western Fisheries, vol 103. no 3, pp 26-27. Butler, T. H. & G. V. Dubokovic. 1955. Shrimp prospecting in the offshore In Memoriam: Terrance Henry Butler Region of the British Columbia coast. June to August. 1953. Fish. Res. Bd. Canada. Nanaimo Biol. Sta. Circular 39. 33 p. Butler. T. H. & G. V. Dubokovic. 1955. Shrimp and prawn prospecting on the British Columbia coast. June to December. 1954. Fish. Res. Can. Nanaimo Biol. Stn. Core. (Gen. Ser.) 35: 92 p. Butler. T. H. & D. G. Hankin. 1992. Comment on mortality rates of Dunge- ness crabs Cancer Magister. Can. J. Fish. Aquat. Sci. 49: 1518-1521 p. Butler. T. H. & H. E. J. Legare. 1954. Shrimp prospecting in regions of the British Columbia coast. November 1953 to March 1954. Fish. Res. Board Can. Gen. Ser. Circ. 31. 42 p. Butler. T. H.. J. G. Lindsay & C. B. Chic. 1975. Prawn trap exploration British Columbia Central coast November 1974 to February 1975. Fish. Res. Board Can. MS Rep. 1357, 1 19 p. Butler. T. H. & M. S. Smith. 1968. Shrimp sampling and Temperature Data obtained during Exploratory fishing off British Columbia. 1966 and 1967. Fish. Res. Bd. Canada, Tech. Rep.. 61. 92 p. Butler. T. H. & M. S. Smith. 1968. Shrimp sampling and Temperature Data obtained during Exploratory fishing off British Columbia. 1966 and 1967. Fish. Res. Bd. Canada. Tech. Rep.. 61. 92 p. Butler. T. H. & M. Stocker. 1990. Surplus production model analysis of crab (Cancer magister Dana) fisheries of British Columbia. Canada. Fish. Res. 9:231-254. Butler. T. H.. A. N. Yates & C. C. Wood. 1973. G. B. Reed shrimp cruise 73-S-l. May 7-23. Fish. Res. Board Can. MS Rep. 1255. 46 p. Butler. T. H.. A. N. Yates & D. C. Miller. 1974. G. B. Reed shrimp cruise in Queen Charlotte Sound, April 1974, Fish. Res. Board Can. Nanaimo Biol. Stn. Circ. (Gen. Ser.) 98. 44 p. Allen. J. A. & T. H. Butler. 1994. The Caridea (Decapoda) collected by the Mid-Pacific Mountains expedition, 1968. Pac. Sci. 48(4):410-445. Boutillier. J. A.. T. H. Butler, el al. 1998. Assessment of the 'Area A' Crab (Cancer magister) Fishery in British Columbia. Can. Stock Assessm. Seer. Res. Doc. 98/86. 39 p. Boutillier. J. A.. J. R. Carmichael & T. H. Butler. 1978. Shrimp population survey, west coast of Vancouver Island. May 1978. Fish. Mar. Serv. Data Rep. 100. 84 p. ( 1 ) Boutillier. J. A., A. N. Yates & T. H. Butler. 1976. B. B. Reed Shrimp Cruise 76-S-l. May 3-19. 1976. Fish. Mar. Serv. Dat Rep. 13. 45p. Boutillier. J. A., A.N. Yates & T. H. Butler. 1977. B. B. Reed Shrimp Cruise 77-5-1. May 3-14. Fish. Mar. Serv. Data Rep. 37. 42 p. Hankin, D. G. and T. H. Butler. 1997. Does intense fishing on males impair success of female Dungeness crabs? Can. J. Fish. Aquat. Sci. 54(3): 655-669. Jamieson. G. S.. C. K. Robinson & T. H. Butler. 1986. King and Tanner Crabs in northern British Columbia mainland inlets. Can. Manuscr. Rep. Fish. Aquat. Sci.. no. 1880. 130 pp. Ketchen. K. S.. N. Bourne & T. H. Butler. 1983. History and present status of fisheries for marine fishes and invertebrates in the Strait of Georgia. British Columbia. Can. J. Fish. Aquat. Sci. 40:1095-1119. Margohs. L. & T. H. Butler. 1954. An unusual and heavy infection of a prawn. Pandalus borealis Kroyer. by a nematode. Contracaecum sp. J. Patasitol. 40(6):649-655. Schrivener. J. C. & T. H. Butler. 1971. A bibliography of shrimps of the family Pandalidae. Emphasizing economically important species of the genus Pandalus Fish. Res. Board Can. Tech. Rep. 241. 42 p. Wicksten. M. K. & T. H. Butler. 1983. Description oiEualus lineatus new species, with a redescription of Heptacarpus herdmani (Walker). (Car- idea: Hippolytidae). Proc. Biol. Soc. Washington. Washington D.C.. v 96. no 1, pp 1-6. Journal of Shellfish Research Vol. 18. No. I. 4-17. 1999. OBSERVATIONS ON THE BIOLOGY OF THE VEINED RAPA WHELK, RAPANA VENOSA (VALENCIENNES, 1846) IN THE CHESAPEAKE BAY JULIANA M. HARDING AND ROGER MANN Department of Fisheries Science Virginia Institue of Marine Science College of William and Mary Gloucester Point, Virginia 23062 ABSTRACT The recent discovery of the Veined Rapa whelk {Rapana venosa. Valenciennes. 1846) in the lower Chesapeake Bay provides an opportunity to observe the initial biological and ecological consequences of a novel bioinvasion. These large predatory gastropods occur in subtidal. hard bottom habitats in the lower Bay and are capable of feeding, mating, and moving while completely burrowed. Hard clams (Mercenaria mercenaria) are consumed preferentially in the laboratory when offered concurrently with oysters (Crassostrea virginica), soft clams {Mya arenaria), and mussels (Mytilus edulis). Chesapeake Bay R. venosa readily open and consume large hard clams (30 to 85 mm SH) leaving no visible signs of either drilling or boring behavior. Shell morphology and thickness may provide an inherent size-selective predation refuge for Rapa whelks in the Bay. These same shell characteristics may change the dynamics of shell selection by local hermit crabs, particularly the striped hermit crab. Clibanarius vittatus. Recent collections of striped hermit crabs from the Hampton Roads area indicate that very large striped hermit crabs are using empty Rapana shells as shelters. KEY WORDS: Rapana venosa. Veined Rapa whelk. Muncidae. Thaididae. ballast water, bioinvasion, Chesapeake Bay. Clibanarius vittatus, Mercenaria mercenaria INTRODUCTION The Veined Rapa whelk, Rapana venosa, (Valenciennes, 1846) is a large, predatory gastropod that has recently been found in the lower portion of the Chesapeake Bay. As with other representa- tives of the Thaididae family [Earlier classifications of the Neo- gastropods place Rapana sp. in the family Muricidae. Recent taxo- nomic revisions include Rapana in the Thaididae (R. Germon, Smithsonian Institution, Washington, D.C., pers. comm.)], this animal is a carnivore whose principal prey items include many commercially valuable bivalves. Rapana venosa is one of several modern Rapana species including R. bezoar and R. rapiformis. Although R. thomasiana was originally described by Crosse in 1861 as a separate species (Thomas's Rapa whelk), it is currently recognized as a synonym fori?, venosa (R. Germon. Smithsonian Institution. Washington, DC. pers. comm.). Rapana venosa is native to the Sea of Japan, the Yellow Sea, the East China Sea, and the Gulf of Bohai (Tsi et al. 1983, Chung et al. 1993. Zolotarev 1996. Chung and Kim 1998). Three species of Rapana occur sympatrically in Chinese waters: R. venosa. R. bezoar, and R. rapiformis (Tsi et al. 1983). All three species are found in coastal subtidal habitats and are commercially harvested (Hwang etal. 1991. Chung et al. 1993, Morton 1994). Rapa whelks were discovered in the Black Sea in 1947 (Drapkin 1963) and have subsequently spread throughout the Black Sea and into the Sea of Azov as well as the Aegean (Koutsoubas and Voultsiadou- Koukoura 1990, Zolotarev 1996) and Adriatic (Bombace et al. 1994) Seas. R. venosa from Korean waters described by Chung et al. (1993) ranged from 32.5 to 168.5 mm shell length (the maxi- mum distance from the tip of the spire to the bottom of the col- umella. SL). Rapana venosa is easily distinguished from native gastropods of the Chesapeake Bay. It has a short spired, heavy shell with a large inflated body whorl and a deep umbilicus (Fig. I ). The slightly concave columella is broad and smooth. Small, elongate teeth are present along the edge of the large, ovate aperture's outer lip. External shell ornamentation includes smooth spiral ribs that end in regular blunt knobs at both the shoulder and the periphery of the body whorl. In addition, fine spiral ridges are crossed by low vertical riblets. Older specimens can be eroded, but the color is variable from gray to orange-brown (one specimen is atypically blonde), with darker brown dashes on the spiral ribs. The aperture and columbella vary from deep orange-red to yellow or off-white. Spiral, vein-like coloration, ranging from black to dark blue, oc- casionally occurs internally, originating at the individual teeth at the outer lip of the aperture. The first collection of Rapana venosa in the Chesapeake Bay was made in the summer of 1998 during a routine trawl collection by the Virginia Institute of Marine Science (VIMS) trawl survey in the vicinity of the Monitor-Merrimac Tunnel (Fig. 2). This speci- men was positively identified as Rapana venosa by Drs. Jerry Harasewych (Smithsonia Institution. Washginton. DC) and Yuri Kantor (Russian Academy of Science, Moscow). A subsequent sampling trip specifically for Rapana venosa in the same vicinity on August 24, 1998 yielded two masses of R. venosa egg cases (Fig. 3; a total of 50+ egg cases) but no live animals. The egg cases were returned to VIMS and maintained at ambient temperature and salinity conditions on a 14 h light: 10 h dark regime. Within a week postcollection, individual egg cases began hatching with the last egg case hatching on September 21, 1998. Larvae were cultured and used in salinity tolerance experiments (Mann and Harding, in review). Given the size of the specimens collected to date from the lower Bay (68 to 165 mm SL) and the presence of viable egg cases, it seems reasonable to assume that the local Rapa whelk population is sexually mature and actively breeding. As in the eastern Mediterranean and Black Seas (Zolotarev 1996), ballast water from commercial and/or military ship traffic is the probable source of introduction into the Chesapeake Bay. R. venosa larvae are planktonic for 14 to 17 days (Chung et al. 1993. Mann and Harding in review). Normal transit time to the Hampton Roads/Norfolk area from the Baltic, Black, Adriatic, or Aegean Seas is approximately 10 to 24 days (G. Ruiz. Smithsonian Envi- ronmental Research Center, pers. comm.). This time interval is well within the temporal window for survival of viable planktonic R. venosa larvae. At certain times during the year (e.g.. May 10 Harding and Mann Figure 1. Picture of an adult Rapana venosa 115(1 mm SL) from the Chesapeake Bay. The arrows highlight the hroad columella, opercular teeth, and bright orange aperture. through October) temperature and salinity regimes on both ends of the trip are similar (see Mann and Harding, in review for a detailed discussion). The Hampton Roads/Norfolk area is a major foci of container, coal transport, and military ship activity. The area ranks third among U.S. ports in terms of volume of ballast discharged on an annual basis (G. Ruiz. Smithsonian Environmental Research Center, pers. comm.). Given the sheer volume of ballast water arriving in Chesapeake Bay annually from ports with active Rapa whelk populations (15 million metric tons; G. Ruiz. Smithsonian Environmental Research Center, pers. comm.). the possibility of obtaining sufficient numbers of Rapa whelk larvae needed to even- tually establish a breeding population in the Chesapeake Bay may be quite high. International traffic aside, the Hampton Roads/ Norfolk area is also a major hub for coastal shipping along the eastern seaboard of North America (G. Ruiz, Smithsonian Envi- ronmental Research Center, pers. comm.). If a local population of Rapa whelks becomes established in the Bay, it is likely that the Chesapeake would eventually become a source population for other coastal ports with similar habitat conditions. This scenario places ports throughout the Middle Atlantic Bight (e.g.. New York, Boston) as well as the South Atlantic Bight (e.g., Charleston) at higher risk for introduction of the species in that they would he receiving both international and local inoculations. Since the discovery of Rapana venosa in the Chesapeake Bay. live Rapa whelks have been under observation in wet laboratory tanks at VIMS. To date, 412 animals have been donated to VIMS (these numbers include live animals, dead animals with shells, or shells only), mostly by commercial watermen and seafood pro- cessing companies, indicating the presence of an established popu- lation of Rapana venosa in the lower portion of the Chesapeake Bay. Observations to date on the basic biology and ecology of Rapana venosa in the Chesapeake Bay are described herein and placed in the context of potential trophic interactions of this animal in the lower Chesapeake Bay. Current Distribution The current distribution of Rapana venosa in the Chesapeake Bay extends from the Chesapeake Bay Bridge Tunnel northward along the western shore line in a continuous swath across Little Creek. Ocean View, Fort Monroe, and Buckroe Beach (Figs. 2 and 4). Several unconfirmed reports from the Poquoson flats area are punctuated by two confirmed discoveries of Rapana at Tue Marshes Light in the York River (Fig. 2). The northernmost report of a Rapa whelk in the Bay is from Butler's Hole, a small oyster rock near the mouth of the Rappahannock River; this 130 mm SL Figure 2. Map showing known Rapana venosa distribution as of March 1999 in the Chesapeake Bay proper (A.) and the Ocean View/Hampton Roads/James River region (B.). The black circles (A.) indicate the Rappahannock and York River collection sites. The black zone (B.) shows the known distribution within the lower Bay/Hampton Roads/James River. The first collection location is indicated with an asterisk (B.). Biology of Rapana venosa ll Figure 3. Rapana venosa egg cases collected from Hampton Roads, VA in August 1998. The yellow egg case cluster was attached basally to a hydroid mat (A.). Note the broad phyllopodus egg case tops and egg pores shown in the top view (B.I. individual was collected by the authors during an annual oyster stock assessment dredge survey. The majority of Rapa whelks have been collected by either commercial clammers or crab dredgers working in the lower Bay. In early September 1998, VIMS established an ongoing Rapana bounty system with the help of the Virginia Saltwater Commerical Fishing Develoment Fund and, as of January 1999, the Virginia Sea Grant program. A bounty is paid for each snail turned in to VIMS personnel, provided that collection information (i.e.. loca- tion, gear, depth, and bottom type) are reported at the time of donation. The bounty program yielded an average of 8 to 10 ani- mals per week through the end of November 1998 donated pri- marily by clammers working off Ocean View and Buckroe Beach (Fig. 2). Clammers in the lower Chesapeake fish for hard clams or quahogs {Mercenaria mercenaria) with patent tongs. Quahogs > 50 mm shell height are abundant (approximately 1 to 1 1 animals m2) in portions of the lower bay (Roegner and Mann 1991). and the commercial hard clam fishery in the reigon is economically important, annually landing 1.1 million pounds with a dockside value of approximately $6 million (Kirkley 1997). The lower Bay also supports a winter crab dredge fishery tar- geting blue crabs (Callinectes sapidus) that burrow into the sand/ mud bottom to overwinter. When crab dredge season opened in the lower Bay on December 1 . 1998. Bay water temperatures were still 8 to 12°C. During the first 2 weeks of December 1998. over 30 Rapa whelks/day were donated to the VIMS colleciton by crab Chesapeake Bay James Rivet >^kk Jk * 4 V4 -*<*> dp t 1 N 10 0 km 10 Figure 4. Distribution map of Rapana venosa from the lower Chesapeake Bay showing collection zones: 1.) Above the SR 258 James River Bridge, 2.) Between the James River Bridge and the Monitor-Merrimac Bridge Tunnel, 3.1 Between the Monitor-Merrimac Bridge Tunnel and the Hampton Roads Bridge Tunnel, including the Hampton Bar area, 4.1 the Lafayette River, and 5.) the Buckroe Beach. Fort Monroe, Ocean View, and Little Creek areas in the bay proper. 12 Harding and Mann TABLE 1. Summary of Rapana venosa collections through March 1, 1999 from the lower Chesapeake Bay and its tributaries. Collection Location Average Shell Length (mm) Standard Error (SE. mm) Lower Bay: Little Creek/Ocean View/Buckroe Beach James River: Hampton Bar James River: between Monitor-Merrimac Bridge Tunnel and the James River Bridge James River: above the James River Bridge Lafayette River Nansemond River York River: Tue Marshes Light Rappahannock River: Butler's Hole 185 7 198 II 7 1 2 1 141.8 138.8 132.7 131.3 102.7 135.0 149.0 1 30.0 0.93 5.23 0.84 3.63 4.9 9.00 "n" refers to the number of animals collected from each location. Locations are shown in relation to the mainstem Chesapeake Bay and each other in Figures 2 and 4. dredgers and seafood processing companies. The arrival of a cold front just before Christmas 1998 caused water temperatures to fall below 5°C and coincided with a reduction in both fishing activity and R. venosa donations. Throughout January and February 1999. crab dredgers working the lower Bay have reported few R. venosa. Presumably, the sustained colder temperatures have driven them either into deeper waters as reported in their home range (Wu 1988) or deeper into the sediment below the zone of dredging activity. Although donations from crab dredgers in the lower Bay es- sentially stopped in January 1999. Rapa whelk donations from clammers working in the James Rivet continued until the closing of the area to commercial fishing in mid-March at an average rate of 6 animals/day"1. As of this writing, there have been no R. venosa reported by commercial oystermen working on extant oys- ter beds in the James River upstream of the Route 258/17 James River bridge (Haven and Whitcomb 1983). A majority of the ani- mals collected to date from all sources have been collected from regions with hard sand bottom in depths ranging from 10 to 60 m at salinities of 18 to 28 ppt. The collection data from commercial sources do not lend them- selves to an accurate Rapa whelk stock assessment, because it is impossible to separate the effects of fishing effort in a particular location from potential gear biases in that both crab dredges and patent tongs selectively catch larger snails (>100 mm SL) given standard ring size for both (0.06 m). However, an examination of R. venosa length-frequency distributions for sites in the lower Bay and Hampton Roads area yields an interesting pattern. The shell lengths (SL. mm) of animals from the five different regions with >5 confirmed Rapa whelk reports (Table I ) were compared with an ANOVA followed by Fisher's test for multiple comparisons (per Zar 1996). Data satisfied assumptions of both homogeneity of variance and normality without transformation. Animals collected from the Ocean View/Buckroe Beach/Little Creek area (Fig. 4) or from regions outside the Hampton Roads Bridge tunnel are sig- nificantly larger than animals collected from either the Lafayette River, above the James River Bridge, or between the Route 258/17 James River bridge (hereafter JRB) and the Monitor-Merrimac bridge tunnel (Figs. 4 and 5; ANOVA. p < .05; Fisher's test, p < .05). Animals collected from the Lafayette River are significantly smaller than Rapa whelks collected from any other site (Figs. 4 and 5: ANOVA. p < .05; Fisher's test, p < .05). It is interesting to note that the Little Creek/Ocean View area is immediately adjacent to 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' i ' i 1 ' 1 ' 1 ' • 50 Zone 2 40 2 30 20 r ■ 10 0 r —■■■-'-■-'-■-■ ' ' ^^ ■ ■■la I'IM'I' Zone 3 i.i.i.i nil L Zone 4 H H II i . i . i ■ i . i . i . i . i . i 50 _ Zone 5 1 i ! . | ■ | i | i | i | i , i 2 40 - 30 j 20 : 10 ■ Jail n --■-■■■-•■ ...'■■II !■■■■■_ Shell length (mm) Figure 5. Length-frequency distribution of Rapana venosa collected from the Chesapeake Bay. Zones 1 through 5 correspond to the zones shown in Figure 4; i.e., 1.) Above the SR 258/17 James River Bridge (JRB. n = ID, 2.) Between the JRB and the Monitor-Merrimac Bridge Tunnel tn = 194), 3.) Between the Monitor-Merrimac Bridge Tunnel and the Hampton Roads Bridge Tunnel, including the Hampton Bar area (n = 7), 4.) the Lafayette Ri\er (n = 7), and 5.) the Buckroe Beach, Fort Monroe, Ocean View, and Little Creek areas in the Bay proper (n= 181). Biology of Rapana venosa 13 both the anchorage for commercial and military ships awaiting pilots and clearance to enter the port and the Thimble Shoals shipping channel (Figs. 2 and 4). The area between the JRB and the Monitor-Merrimac bridge tunnel includes the Newport News coal container terminal, a major site of deballasting for interna- tional ships awaiting coal. Age Estimates In the absence of age and growth estimates for Chesapeake Bay Rapana venosa, age and growth estimates for a Knobbed whelk (Busycon carica) population from Virginia's Eastern Shore may offer a conservative estimate of potential Rapa whelk growth rates. Kraeuter et al. ( 1989) and Castagna and Kraeuter 1 1994) provide growth and length-at-age estimates for B. carica from both labo- ratory and field studies extending over a 14-year period. Growth rates for B. carcia were greatest during the first year (approxi- mately 32 mm/y-1 ) and then subsequently decreased to 14.4 mm/ yr~' for the first 10 years followed by growth rates of 6.5 to 9.5 mm/y-1 for animals older than 10 y and/or greater than 160 mm SL (Fig. 6; Kraeuter eta 1. 1989. Castagna and Kraeuter 1994). In the absence of any data on Rapana growth rates in the Chesapeake Bay. it is reasonable to consider growth rates of such sympatric species as B. carica for initial estimates of Rapa whelk age. Ninety-five percent of all R. venosa collected thus far in the Chesa- peake Bay are between 110 and 160 mm SL. If this range of Rapana shell lengths is overlaid onto the B. carica growth curve presented by Kraeuter et al. (1989) and Castagna and Kraeuter (1994). the resulting age distribution extends from approximately 7.5 to 13 years (Fig. 6). These are conservative growth estimates when considered in relation to the growth rates for Black Sea Rapana reported by Chukhchin (1984). Rapa whelk length-at-age relationships have been described by * • 12 24 36 48 60 72 84 96 108 120 132 144 156 168 Age (months) Figure 6. Plot at Busycon carica length-at-age relationship from labo- ratory and field observations of an Eastern Shore, VA B. carica popu- lation after Kraeuter et al. (1989) and Castagna and Kraeuter (1994). The shaded zone indicates the size range (SL, mm I and corresponding age estimate for 95% of the Rapana venosa collected in the Chesapeake Bay thus far. Chukhchin (1984) for animals from the Black Sea. Chukhchin ( 1984) estimates reports growth rates for individuals in Sevastopol Bay of 20 to 40 mm during year 1. with mean shell length (SL) values of 64.6 mm, 79.4 mm. 87.5 mm, and 92.1 mm in years 2 through 6. respectively. This terminal size is smaller than the maximum SL of 120.1 mm reported by Smagowicz (1989) for a specimen in a collection from Bulgaria and Georgia, whose exact collection location was not reported. Chukhchin (1984) correlates shell thickening with spawning events and notes that the first spawning occurs in the second year at sizes ranging from 35 to 78 mm SL with a mean value of 58 mm SL. Habitat Preferences Both field collections and laboratory observations confirm that Rapana venosa prefers hard sand bottom habitats. These animals are avid burrowers and remain completely burrowed for more than 95% of the time in the laboratory. A 150 mm SL R. venosa can burrow into a sand bottom so that its shell is completely covered in less than 1 h. The only visible sign of a burrowed Rapa whelk is the maroon U-shaped siphon that is usually extended 1 to 3 cm above the surface of the sand. Rapa whelk siphons are sensitive to both light and motion and are retracted immediately at the slightest disturbance. Siphonal sensitivity combined with the animal's bur- rowing speed and low visibility conditions in the Bay may make conventional benthic survey methods that relay on direct observa- tion of the animal (diver transects, video surveys) difficult as non- invasive stock assessment techniques. Bombace et al. (1994) ob- served an apparent increase in R. venosa biomass after artificial reef deployment in the Adriatic Sea. It is possible that there were burrowed Rapana at the sites at the time of reef deployment and that increases in Rapana sightings after reef construction are at- tributable to the emergence of local snails to feed on the reefs not the arrival of snails from other areas. Laboratory observations indicate that Rapa whelks are capable of both feeding and mating while burrowed. They move reason- ably quickly while burrowed (approximately 1 body length per minute). Hard sand bottom habitat is relatively common in the lower Chesapeake Bay (Fig. 7) and is not likely to be a limiting factor for potential range expansion of the animal in the bay. Prey preferences Rapana bezoar was described by Morton (1994) as "a gener- alist predator of subtidal molluscs." This description is certainly apt for R. venosa in the Chesapeake Bay. In laboratory feeding studies. Chesapeake Bay R. venosa prefer hard clams to oysters ( Crassostrea virginica), soft clams (Mya arenaria), or local mus- sels {Mxtilus edulis). although they will eat these other bivalves when hard clams are rare or unavailable (Fig. 8). A 140 mm SL Rapa whelk is capable of consuming a 75 to 80 mm hard clam in less than 1 h. Previous reports on the feeding behavior of R. thomasiana (now recognized as R. venosa) from the Black Sea place R. venosa among the gastropods that drill their prey (Gomoiu 1972. Carriker 1981) or use paralytic toxins during feeding (Chukhchin 1984). Morton ( 1994) describes feeding behavior of R. bezoar in terms of boring or crude rasping usually on the posterioventral shell margin. Similar rasping behavior has been observed for Chesapeake Bay Rapana venosa feeding on small hard clams [<30 mm shell height. SH (distance from hinge to the opposite shell margin)]. Small chips or rasp marks are visible on the posterioventral shell margin 14 50 Okm 50 100~ Harding and Mann B. A x0 Atlantic Ocean so Okm 50 100 A .0 Atlantic Ocean Figure 7. Maps of the lower Bay showing sand bottom habitat (A.) and hard clam populations (B.l in black per Roegner and Mann ( 1991 1 Ocean View/Hampton Roads/James River region is indicated by a square in both maps. The of some small clams attacked and eaten by large R. venosa. How- ever, Chesapeake Bay R. venosa readily open and consume large hard clams (30 to 85 mm SH) leaving no visible signs of either drilling or boring behavior. R. venosa grasps its prey along the shell margin and covers the clam with its foot until the clam gapes slightly (Fig. 8). When the clam gapes, the Rapana inserts its proboscis between the clam valves and begins feeding. The entire clam is consumed leaving clean, empty, articulated valves with no visible predation signature as the end product. Food is not likely to be a limiting factor for Rapana venosa in the Chesapeake Bay. Rapa whelks seem to share habitat preferences with their favored food item: the preferred habitat for both hard clams and Rapa whelks is sand bottom. The known Rapana venosa distribution overlaps regions of moderate to high hard clam densities in the lower bay (Fig. 7). The absence of a predation signature on large hard clams con- sumed by Rapa whelks is troubling in light of recent conversations with commercial clammers working in the Ocean View and Hamp- ton Roads area (Fig. 4). The clammers report an increase in the number of empty shell valves caught within the last 1 to 2 years and attribute the increase in empty valves to a corresponding in- crease in natural clam mortality. Given the number and size of Rapana venosa reported from these same areas during 1998 and the absence of a predation signature on large hard clams consumed A. B. Figure 8. Adult Rapana venosa consuming a hard clam (A.) and an oyster (B.t. Biology of Rapana venosa 15 in the laboratory, it is possible that the recent increase in empty, articulated shell valves observed by local watermen is attributable to Rapa whelk predation and not natural mortality. Rapa whelks have also been described as scavengers consum- ing carrion (Chukhchin 1984. Morton 1994). Laboratory observa- tions indicate that Rapa whelks prefer to capture and kill their own food; they will not feed on carrion in the presence of live prey. However. Chesapeake Bay Rapana have been caught incidentally by recreational fishermen that were using fresh squid as bait. Potential Predators: Rapa Whelks Rapana venosa are prey for native octopods in their native waters. Few of the habitats that Rapana have invaded include resident octopods as upper-level predators enabling Rapana popu- lations to grow quickly and inflict considerable damage on local shellfish resources; for example, the decimation of the Black Sea oyster population as described by Chukhchin (1984). Within the Chesapeake Bay. the only upper-level or apex predators that might be capable of using Rapa whelks as a food resource are those that currently eat the local whelk species; that is. Channeled whelks (Busycotypus canaliculatus) and Knobbed whelks (Busycon carica). Crabs and other gastropod species are potential predators for very small Rapana. Sea turtles may be capable of eating Rapa whelks Os O* Os Os Os OS Year Year Year Figure 1. Production trends in China (MAC 1985-97). A, aquaculture versus capture fishery; B, total mariculture production versus molluscan aquaculture; and C, aquaculture production of major molluscan groups from 1986 to 1996. Productions are in wet. whole body weight. lasted for 4 weeks and covered much of the central coast of China, from Qingdao in Shandong province to Wenzhou in Zhejiang province (Fig. 2). It consisted of on-site visits to aquaculture fa- cilities and discussions with local scientists, managers, and farm- ers. Major aquaculture sites we visited included those at and around Qingdao. Rizhao, Ganyu (Jiangsu). Lianyungang, Dafeng (Jiangsu), Qidong (Jiangsu). Wenzhou, and Yueqingjiang. In Oc- tober 1997. the first author visited the north and south coasts of China, which were not covered by our 1996 visit. The 1997 visit was part of a study on the Chinese molluscan aquaculture industry Figure 2. A map of China's coastal provinces show ing major cities and areas visited during this study. conducted by the first and third authors under sponsorship of the US-China Living Marine Resources Exchange Program. The 1997 visit covered major aquaculture facilities and research institutions in Liaoning and Shandong provinces in the north, and Fujian. Guangdong and Guangxi provinces in the south. Cities and areas covered by the 1997 visit included Dalian. Yantai. Rongcheng. Qingdao. Wenzhou, Xiamen. Guangzhou. Shenzhen (bordering Hong Kong), and Beihai. Production figures cited in this paper are mostly from official statistics published by the Ministry of Agriculture of China (MAC 1986-1997). Collecting accurate production data for molluscan aquaculture is always challenging, which is particularly true in China for several reasons. First, production of most molluscan species in China is estimated by multiplying the total culture area by an average yield per unit area. This method alone contributes a large degree of uncertainty. Second, the culture of a large number of species in various forms and across culturally diverse regions is a source of confusion and error. For example, the word "clam" (Ge in Chinese) may include different species in different areas. Oyster production from south China is traditionally reported in meat weights and that from the north in whole weight (with shells). Certain types of culture practices may be considered as aquacul- ture in one area, but not in another. Finally, production may be over- or underestimated by local officials for management and/or political reasons. Despite all these factors, the official figures are still the best or only available statistics. Some Chinese scientists and managers believe that the official statistics may overestimate the overall production by 20 to 30%. For some species, such as blood cockles and razor clams, the official statistics are very close to estimates from local scientists. The decline of the shrimp aqua- culture industry because of diseases is reflected in the official statistics, which corresponds well to expert estimates. (We should see a decline in scallop production, because of mortalities, in 1997 to 1998 if the official statistics are accurate.) Starting in 1996, a new reporting system, or standard, was implemented, which cor- rected some problems. For example, oyster and some clam pro- duction before 1996 were reported as meat weight, and the new statistics converted all molluscan production to whole body weights. For this report, oyster production data prior to 1996 were converted to whole body weight using a factor of 6.1 1, as recom- mended by the Ministry of Agriculture of China (MAC 1997). Scientific names are presented for all species discussed in this paper. The English common names, if available and generally accepted, are also used. For species with no common English Moi.luscan Aquaculture in China 21 names or conflicting ones, the Chinese common name is given in standard pinyin to avoid confusion with different translations. OVERVIEW OF PRODUCTION Mollusks are cultured all along China's 18,000-km coastline. The scope of molluscan aquaculture is reflected by the large num- ber of species cultured. A quick survey indicates that at least 32 species of marine mollusks are cultured commercially in China (Table 1 ), and the list is probably incomplete and growing. The list contains five gastropods and 27 bivalves. The bivalve species in- clude three oysters, four scallops, five mussels, one pearl oyster, and 14 clams. The list in Table 1 is arbitrarily divided into "major" and "minor" species. Most of the 14 major species are well-known mollusks that support large aquaculture industries in China. The minor species are also cultured at commercial scales, but with less significant production. In 1996, China produced 6.4 million metric tons of mollusks (whole body wet weight) from aquaculture, which is about eight times that of 1986 (Table 2). Oysters topped the species list, with production of 2.3 million tons, or more than one-third of the total. Clam production ranked second, with 1.6 million tons. The actual clam figure may be much higher, because some clam species from certain areas are reported under "others mollusks." Most of the 1 .2 million tons of "other" mollusks listed in Table 2 are probably TABLE 2. Aquaculture production I "hole weight in metric tons) of mollusks in China; 1986 vs. 1996/ Species Group 1986 1996 Oyster Clam Scallop Mussel Abalone Pearl Oyster Others Total 336.013 191,951 23,686 210,657 41.592 803,899 2,284,663 1,568.325 999.573 366,251 700 10.000" 1.187.083 6,416,595 'From MAC (1987, 1997) Oyster production in 1986 was converted to whole weight using a factor of 6. 1 1 . Clam statistics include razor clam and cockle species. h Estimated from the production of 20 tons of pearls. miscellaneous clams, and the total clam production from aquacul- ture could be anywhere between 1.6 to 2.5 million tons. Scallops ranked third, with a production of I million tons. Abalone produc- tion was negligible in weight but significant in value. An estimated 20 tons of marine pearls was produced, corresponding to about 200 million pearl oysters or 10,000 tons in whole weight. TABLE 1. A list of marine mollusks cultured in China. English Name Scientific Name Major Culture Areas, Notes Major species Zhe oyster Suminoe oyster Pacific oyster Manila clam Colorful clam Meretrix clam Mud cockle Razor clam Zhikong scallop Bay scallop Wrinkled abalone Colorful abalone Pearl oyster Blue mussel Minor species Huagui scallop Japanese scallop Thick-shell mussel Green-jade mussel Senhouse mussel Penshell Xishishe surfclam Square surfclam Chinese surfclam Cyclina clam Chinese glaucomya River corbula Morella clam Hairy cockle Giant cockle Mud snail Red conch Sea hare Crassostrea plicatula Gmelin Crassostrea rivularis Gould Crassostrea gigas Thunberg Ruditapes philippinarum Adams and Reeve Ruditapes variegata Sowerby Meretrix meretrix Linnaeus Tegillarca granosa Linnaeus Sinonovacula constricta Lamarck Chlamys farreri Jones and Preston Argopecten irradians Lamarck Haliotis discus hamuli [no Haliotis diversicolor Reeve Pinclada martensii Dunker Mytilus edttlis Linnaeus Chlamys nobilis Reeve Patinopecten yessoensis Jay Mytilus comscus Gold Pema viridis Linnaeus Musculus senhousei Benson Pinna pectinatu Linnaeus Mactrn antiquata Spengler Mactra veneriformis Reeve Mactra chinensis Philippi Cyclina sinensis Gmelin Glaucomya chinensis Gray Potamocorhula rubromuscula Zhuang and Cai Moerella iridescens Benson Scapharca subcrenata Lischke Scapharca broughtonii Schrenck Bullacla exarala Philippi Rapana venosa Valenciennes Notarchus leachii cirrosus Stimpson Fujian. most important oyster, wild seeds Guangdong. Fujian. low salinity, wild seeds Liaoning. Shandong, hatchery seeds, longlines All coast, most common clam, mostly wild seeds Fujian. Guangxi, some hatchery seeds Jiangsu. Shandong, extensive culture Zhejiang to Guangdong, high-value, some hatchery Zhejiang, Fujian, wild seeds, intertidal flats Shandong, Liaoning, wild seeds, cage culture Shandong, Liaoning, Fujian, hatchery seeds, cage Liaoning, Shandong, hatchery seeds, cage, raceway Guangdong, Fujian, hatchery seeds, cage culture Guangdong. Guangxi, hatchery seeds, cage culture Shandong, Liaoning. wild seeds, longline culture Guangdong. Fujian Liaoning Zhejiang. Fujian Fujian. Guandong Fujian, Guangdong Zhejiang. Fujian Shandong. Jiangsu, Fujian Jiangsu. Liaoning Jiangsu Jiangsu. salt Guangdong. Fujian Jiangsu Zhejiang Hebei, Liaoning, Shandong Hehei. Liaoning. Shandong Zhejiang Liaoning, Shandong Fujian 22 GUO ET AL. With the exception of mussels, production in all species groups increased several fold over the 10-year period between 1986 and 1996 (Fig. 1C). Production statistics before 1986 did not separate species groups, so it is impossible to make comparisons with ear- lier years. Economic reform and government promotion are prob- ably the dominant factors behind the rapid growth. Mussel pro- duction has been declining since 1992 probably because of com- petition from high-value species such as scallops, oysters, and abalone. Molluscan aquaculture is widely distributed among the Chinese coastal provinces (Table 3). Shandong province in the north, with a production of 2. 1 million tons, or one-third of the national total, is the largest producer of cultured mollusks and leads other prov- inces in scallop, clam, and mussel culture. Yantai. Rongcheng, and Rizhao are the major mariculture areas in Shandong. Liaoning is the leader in abalone aquaculture and also contributes significantly to scallop and mussel production. Jiangsu province cultures a va- riety of clam species and little else. Zhejiang province leads the nation in cockle and razor clam culture. Fujian. Guangdong, and Guangxi are major oyster culture provinces. Fujian is also a major producer of clams, especially razor clams. Guangdong and Guangxi are the dominant producers of pearl oysters. Hebei and Hainan are relatively poor in coastal resources and contribute little to molluscan production. As a generalized pattern, the Chinese molluscan aquaculture can be characterized, with some exceptions, by the scallop, Manila clam, and abalone culture in the north, oyster and pearl oyster culture in the south, and a variety of clams in the middle. We could not find any statistics on the trading of cultured mollusks. Most of the abalone facilities we visited ship a large portion of their product to markets in Hong Kong and Japan. A significant portion of cultured scallops is also exported. Some Meretrix and Manila clams are exported to Japan, and some oys- ters are sold to Hong Kong. Most of the cultured mussel, razor clams. Ruditapes clams and blood cockles, oysters, and scallops are consumed in China. China also imports some mollusks, such as geoduck and squid from North America, green mussel from New Zealand, and blood cockle from Korea. A hotel in Wenzhou, where we stayed in 1996. was selling geoducks from Canada at US$33/ kg. We were told by sources that China has become a net seafood importer in recent years, which may be partly because of the large quantities of fish and fishmeal imported for processing. OYSTERS China produced over 2.3 million tons of cultured oysters in 1996. By weight, oysters are the largest molluscan group cultured in China, and most of the output comes from three species. The most important species is the zhe oyster, Crassostrea plicatula Gmelin. Official statistics do not distinguish individual oyster spe- cies, but experts estimate that the zhe oyster accounts for 50-60% of the total oyster production. The second largest production comes from the Suminoe oyster. C. rivularis Gould (or C. aria- kensis Fujita). which accounts for 20-30% of the total. The other major species is the Pacific oyster. Crassostrea gigas Thunberg. which may account for 10-20% of the national production. China has about 20 recorded species of oysters occurring along its coast, and the classification is sometimes problematic. There are three points of uncertainty concerning aquaculture species. First, oyster farmers in southern China recognize two forms of oysters traditionally regarded as C. rivularis. One is called the "white" oyster, and the other is called the "red" oyster (referring to the meat color). Some experts now believe that the "white" oyster is actually true C. rivularis, and the "red" oyster may be C. iredalei (Li et al. 1988). The two oysters are usually present in the same area at variable proportions. Another uncertainty is about the spe- cies status of the Dalianwan oyster. C. talienwhanensis Crosse, which is cultured in the Dalian area. Some people believe that the Dalianwan oyster is actually a variety of C. gigas. Also, people disagree on whether the monk-hat oyster. C. cucullata, is synony- mous with the zhe oyster (C. plicatula Gmelin). and both names are used in the literature. Zhe Oyster The zhe oyster is found along the entire coast of China. It is small as compared with the Suminoe and Pacific oysters, and thin-shelled. The left shell is deeply cupped, similar to that of Kumomoto oyster (C. sikamea). but even more pronounced (Fig. 3A). The right shell is flat and covered with radial plates. The zhe oyster grows rapidly during the first year, after which shell growth usually stops. Zhe oysters are cultured primarily in Fujian province and other parts of the southern coast. In Fujian alone. 23.000 hectares are used for oyster culture, 80% of which are for zhe oysters. Tradi- tionally, the zhe oyster is cultured on stone pilings, vertical stone TABLE 3. Aquaculture production of major molluscan groups (whole weight in metric tons) from nine coastal provinces of China/ Province Total Oyster Clamb Scallop Mussel Razor Cockle Other Liaoning 871.657 95.969 224,787 223,286 104.888 7.012 1.119 214.596 Hebei 51.568 — 31.591 8.882 1.900 — 9,195 8.493 Shandong 2,144.166 338.209 481,974 753.902 140,983 39,997 31.356 357.745 Jiangsu 122.012 — 105.820 — — 3,338 4.361 8,493 Zhejiang 357,861 62,459 19,121 877 11.826 189.521 56.288 17,769 Fujian 1.138.226 804.845 106.605 12.325 64.281 102.651 3.363 44,156 Guangdong 1,204,212 558.950 39,534 130 40.352 — 20.222 545.024 Guangxi 5 1 1 .885 422.555 82,765 15 2.02 1 — 4,529 Hainan 5,008 1,676 1.751 156 — — 1 .425 — Total 6,406,595 2.284,663 1,093.948 999.573 366.251 342.519 131.858 1.187,783 "From MAC (1997). ' Clam here refers to Ruditapes clams for some provinces and may include meretrix and other clams (excluding razor clams and blood cockles provinces. for other MOLLUSCAN AQUACULTURE IN CHINA 23 Figure 3. Oyster culture. A, The deeply cupped left shells of the zhe oyster cultured in Fujian and Guangdong. B. intertidal racks and stakes and suspended longlines used for the culture of zhe oyster in Xiamen. C, Suminoe oysters cultured on cement stakes in Beihai area. D, a hatchery in Wenzhou area where Pacific oyster spat are heing set on shell-strings. The large concrete tanks (20-5(1 nr'l are typical for all hatcheries and are used for larval culture of a variety of species. E, bamboo rafts used to culture Pacific oysters on shell-strings in Vueqingjiang, Wenzhou. F. Pacific oysters cultured in Dalian reach 8-1(1 cm in 6 months. strips (over 1-m tall), and bamboo or wooden stakes. These ma- terials are arranged in a variety of configurations and formations in intertidal areas (Cai and Li 1990). and oyster-growing areas filled with these structures cover miles of coastline (Fig. 3B). Culture of zhe oysters depends entirely on natural seeds. Large stone strips are permanent structures used for both seed collection and grow- out. Often, seeds are collected on stones, shells, bamboo, and cement blocks in one area and moved to another area for culture. Recently, suspended longlines (Fig. 3B) and raft culture of shell strings has gained popularity with farmers, because it results in better growth and allows for utilization of open-water areas. Stone and bamboo racks in traditional oyster fields have been abandoned in some areas because of their lower productivity. Seeds are usu- ally collected in May and September. Seeds collected in May are usually harvested starting in December. Production peaks in Feb- ruary, around the Chinese New Year, and ends in March. The total culture time is under 1 year. For the fall seeds. 14 to 17 months are needed to reach a market size of 6-7 cm. Suminoe Oyster The Suminoe oyster occurs naturally along most of the Chinese coast, where it is called the Jinjiang ("close-to-river") oyster in Chinese. It tolerates a wide range of salinity, but prefers low sa- linity estuaries and riverbeds, especially for settlement. Local oys- ter farmers recognize two forms of the oyster: the white and red oysters. The former is preferred for its flavor and valued more than the red oyster. The Suminoe oyster is cultured primarily in two southern prov- inces, Guangdong and Guangxi. Culture methods are similar to those used for the zhe oyster. At the Guangxi site we visited. concrete stakes, 50-cm long and 6 x 6 cm at the cross section, are used for both spat collection and growout (Fig. 3C). The stakes are transported to upper river sites for spat collection in the spring. After the spat set. the stakes are planted along the lower river beds for growout. A dozen oysters per stake are considered optimal. About 30.000 stakes are planted per hectare, which produce about 1 2,000 pounds of oyster meat at harvest. At low tide, the river beds are covered with a forest of oyster-carrying stakes. Suminoe oys- ters are also cultured on shell strings hanging on rafts and long- lines. Unlike the zhe oyster, which usually stops shell growth after the first year, the Suminoe oyster maintains rapid growth through out the first 3 years. Oysters are usually harvested in 2 to 3 years at a size of 10-15 cm. In some areas, oysters are moved to pro- ductive areas for fattening before being harvested. Pacific Oyster Pacific oysters occur naturally along the Chinese coast. How- ever, most of the Pacific oysters being cultured in China were originally introduced from Japan. This species is cultured in all parts of the coast, but the major producers are Liaoning and Shan- dong provinces in the north, and Guangdong in the south. Pacific oyster culture depends exclusively on hatchery- produced seeds (Fig. 3D). As with most other bivalves, larvae are cultured in large concrete tanks (10-100 m3). Vitamin supplements and antibiotics are often used during larvae culture to maximize yields. Spat are set on strings of scallop or oyster shells. The recommended density is about 20 to 30 spat per shell, but. in practice, the density is often two to three times that. Farmers may break the shell in half for growout if the density is high. Spatted shells sell for about US$0.0 1-0.02 per piece, depending upon sea- 24 GUO ET AL. son. They are inserted into nylon ropes and cultured on suspended longlines (Fig. 3B) or rafts (Fig. 3E). Bottom culture is also prac- ticed in certain areas. Pacific oysters grow rapidly. Oysters at most sites we visited, including Dalian in the north, reach 8 to 10 cm after the first growing season (Fig. 3F). One factor is that the seeds are produced early in the spring so that they have a full season to grow. Another factor may be the high productivity of the culture sites. Depending upon demand, oysters may be harvested within the first year. Oys- ters cultured in the intertidal areas may need 2 to 2.5 years to reach market size. Triploid Pacific oysters, which undergo little or no gametogenesis. are used for production in Shandong and Liaoning provinces because of their superior growth and improved sur\i\al against "summer mortality." a syndrome linked to reproduction (Perdue et al. 1981). Some oysters are sold live on local markets or to restaurants. Vendors in the market would shuck the oysters if requested. Oys- ters are cooked in a variety of recipes, but rarely consumed raw in China. In southern China, most oysters are cooked and dried for storage, and the broth from processing is condensed into oyster sauce, a popular seasoning in Chinese cooking. CLAMS At least 14 species of clams are cultured in China (Table 1 ). The major species include the Manila clam (Ruditapes philippi- narum Adams and Reeve), the colorful clam (R. variegata Sow- erby). the Meretrix clam (Meretrix meretrix Linnaeus), the mud cockle (Tegillarca granosa Linnaeus), and the constricted razor clam (Sinonovacula constricta Lamarck). In addition to the five major species, at least nine others are commercially cultured in China (see Table 1 ). The official figure for clam production in 1996 is 1.6 million tons, but the actual figure may be higher, because some clam species may not be listed as clams in official figures. On the other hand, clam production listed as from aqua- culture may not be cultured in a strict sense. Culture of most clams, such as razor clams, cockles, and most of the Ruditapes clams, involves seed collection or hatchery production, nursery, and planting, and can be clearly classified as aquaculture. In some areas, however, culture of Meretrix and some other clams may simply mean protective enhancement of natural resources. The culture fields are protected from predators and poachers, but no seed collection and planting are involved. In some cases, a mixture of culture and wild harvest is practiced. Seeds of one species are planted, but multiple species are harvested from the same planting ground. All considered. China's clam production from aquaculture may be anywhere between 1.6 and 2.5 million tons. Ruditapes Clams Ruditapes clams are the most widely cultured clams in China and account for probably 60-70% of the total clam production. Two species predominate: the Manila clam. R. philippinarum Ad- ams and Reeve and the colorful clam R. variegata Sowerby (Fig. 4A). The two are often not distinguished in China, and both are referred to as "Gezi" in Chinese. The Manila clam is cultured along most of the coast and accounts for most of the production. The colorful clam is cultured in the southern provinces. Most seeds are collected from the wild, with Shandong and Fujian provinces being the major producers. Seed collection in- volves selection and building of seed-collection beds, eradication of predators, and routine maintenance. In southern China, clam Figure 4. Clam culture. A, Manila (upper) cultured in Jiangsu and the colorful Ruditapes clam (lower) cultured in Fujian. B, the cyclina (upper) and meretrix (lower) clams cultured in Daxing, Jiangsu: the dark coloration is characteristic of clams cultured in shrimp ponds. C, the square surfclam (upper) and hairy cockle (lower) cultured in Ganyu, Jiangsu. D, salt ponds (upper) in Qidong used for the production of cyclina clam seeds (lower), which usually reach a commercial size of 1 cm in a year. E, unprotected clam beds in Qidong used to culture meretrix, cyclina clams, square, and xishishe surfclams. F, net-fenced clam plots (upper) in Ganyu. Jiangsu. used to culture various clam species; large shrimp ponds (lower) in Wenzhou area, used for polyculture of shrimp, mud cockle and other clams. G, beds for the collection of the constricted razor clam seeds in Wenzhou area; the beds should be completely drainable during collecting season. MOLLUSCAN AQUACULTURE IN CHINA 25 seeds are also produced from hatchery production. Clam seeds, 5 to 10-mm in size, are planted at a density of about 35 million per hectare, but density may vary, depending upon the size of seeds and sediment type. In most cases, clam beds are not protected with nets. Clams are harvested at a size of 30 mm or larger, which is usually reached in 10 to 16 months. Manila clams are one of the most common seafoods in coastal regions of China. Most clams are sold live in local markets, for about USS0.5/kg. They are either stir-fried or used in soup with shells on. Ruditapes clams are usually not depurated before reach- ing the local market. "De-sanding" by placing clams in saltwater for an hour or two is usually the first step in the kitchen. Now there is a frozen product that uses depurated clams, vacuum-packed and frozen in a microwave oven-ready plastic bag. Some frozen clams and clam meat are sold to Japan. Meretrix and Other Clams in Jiangsu The meretrix clam (Fig. 4B) is found in most parts of China's coast. It resembles the hard clam (Mercenaria mercenaria) in bi- ology. It prefers sandy substrates and Jiangsu province, which has a long sandy coast, is the major producer of meretrix clams. Sev- eral other minor clam species are also cultured in Jiangsu using the same culture system as for meretrix clams, including the cyclina clam (Cyclina .sinensis Gmelin) (Fig. 4B). the square surfclam (Mactra venerifonnis Reeve) (Fig. 4C). and the xishishe surfclam (M. antiquum Spengler). Seeds of meretrix and most other clams are collected from the wild. Most cyclina clam seeds are artificially produced in earth ponds (Fig. 4D). Earth pond seed production is a low-tech and low-cost substitute for hatcheries and is effective for certain mol- lusks, including the Ruditapes, the cyclina clam, razor clam, and oysters. Large earth ponds, usually 1-3 hectares in area and 1-m deep, are dried and treated with bleach or herbal poisons before filling with filtered seawater (to 0.1 mm). Brood stocks are in- duced to spawn either in the pond or inside the hatchery. A density of i_4 D-stage larvae/mL is desired. Ponds are fertilized to boost algae growth. No water is discharged before larvae are completely settled. Seeds that settle in the ponds may be thinned if the density is too high. About 3,000 seed clams/nr. at a commercial size of 10 mm, can be expected from earth pond production by the end of the first year. The seed clams are mostly planted in unprotected intertidal flats (Fig. 4E). although planting areas are usually treated with herbal pesticides to remove predator species before planting. After the seed stage is passed, predation is not a major problem. Meretrix and surfclams live close to the surface and are capable of moving and relocating, particularly during storms; therefore, some clam beds are protected with net fences to prevent escape (Fig. 4F). In Jiangsu. the same bed may be used for meretrix and other clams. The clam beds are continuously harvested by manually picking out clams that have reached market size. Another major form of cul- ture for meretrix and other clams is in shrimp ponds (Fig. 4F). Because of the shrimp disease problem, most shrimp ponds are now used for other species or polyculture with other species, es- pecially clams. In the polyculture ponds, shrimp are stocked at a low density and harvested at a small size before mortality starts. One or two clam species are selected and planted in the shrimp ponds. One Jiangsu farm we visited grows both meretrix and cy- clina clams in their shrimp-clam polyculture system. Each shrimp pond is about 3.3 hectares and produces about 10 tons each of meretrix clams and shrimp. Because the harvested shrimp are small, clams constitute a significant proportion of the farm's in- come. Abandoned salt ponds are also used for clam culture. Mor- tality has been reported for meretrix clams cultured in ponds, which coincides with spawning of the clams in August when the water temperature reaches 30-3 1°C. The mortality is usually 30- 40%. but may reach 80-90%. Growers believe the problem is related to high temperature and poor water quality (e.g.. low oxy- gen and, possibly, high bacterial loads) adding to the "stress" of spawning. Mortality in other clam species is rare. Blood Cockles Three species of blood cockles are cultured in China: the mud cockle (Tegillarca granosa Linnaeus), the hairy cockle (Scapharca subcrenata Lischke), and the giant cockle (Scapharca broughtonii Schrenck). The mud cockle is the most desired and widely cultured of the three. Total cockle production from aquaculture was about 131,858 tons in 1996. mostly from the mud cockle. The mud cockle is found in muddy sand beaches of the Shandong peninsula and south. Larval settlement is found to be best on substrates that are more on the sandy side, but adult mud cockles usually grow faster on muddy sediments. The mud cockle is a small and slow growing species, usually taking 2 to 3 years to reach a market size of 2.5 cm. Both wild and hatchery-produced seeds are used in cockle cul- ture. Considerable experience in seed collection and nursery rear- ing has been accumulated throughout the history of cockle culture (1.700+ years). Cockle seeds, called "cockle-sand" (about 0.5 mm), are usually collected in September and October, using cloth bags or nytex screens. Cockle sand sells for about US$l,200/kg. but prices vary greatly from year to year. Seeds of the cockle-sand size are usually cultured in nursery beds for a year before reaching the next stage— "cockle-beans" (about 2.000 cockles/kg). Nursery beds are elevated so water can drain off completely at low tide. Predator species are eradicated before planting by applying herbal poisons before spat fall. A layer of fine sand is added to overly muddy substrates. Growout planting density varies greatly among different areas and depending upon seed size, ranges from 150 to 1 .500 seeds/nr (Wang et al. 1993). Hatchery techniques have been developed to meet the increasing demand for cockle seeds. The Zhejiang Mariculture Institute in Wenzhou we visited is a leader in cockle hatchery technology. Larval culture is the same as for other bivalves, but a unique practice in cockle hatcheries is that mud is added as a substrate to induce settlement. Mud cockle is cultured in intertidal plots and ponds. The yield from cockle ponds is about 7 to 10 tons/hectare. The mud cockle is a delicacy in the Shanghai and Zhejiang areas. It is sold for about US$1 0/kg in Zhejiang area, but imports from South Korea drove the price down by about 30% in 1997. Cockles are prepared by briefly dipping them in boiling water, after which they are consumed semiraw. Because of incomplete cooking and poor sanitation in some areas, cockle consumption, particularly with hairy cockle, has occasionally been associated with Hepatitis infection. Constricted Razor Clam in Zhejiang and Fujian The constricted razor clam (Sinonovacula constricta Lamarck) is found in most parts of the Chinese coast. It tolerates wide temperature and salinity ranges and prefers substrates with a muddy top layer and fine sand bottom. This species is primarily 26 GUO ET AL. cultured in Zhejiang and Fujian provinces and is probably the most important mollusk for Zhejiang province, where its production is greater than all other mollusks combined. Zhejiang produced 189.521 tons of razor clams in 1996. which represents 55% of the national total. Zhejiang and Fujian together account for 85% of all razor clams produced in China. Most razor clam seeds are collected from the wild. Hatchery and earth pond production are used when wild seeds are insuffi- cient. Wild seed collection has a long history and well-established protocols. Collection beds are elevated plots with drainage chan- nels so that the beds are completely exposed at low tide (Fig. 4G). Newly settled razor clam spat would move away from areas that are submerged at low tide. Larval settlement occurs in the fall, and seeds (1-2 cm) are harvested 3 month later and planted at a density of 1.000-1,500 kg/hectare for growout. The razor clams typically grow to about 5 cm in 5-8 months after planting ( 1 year of age). Large clams are harvested at 1 year of age. and small ones are cultured for a second year. Razor clams have several predators, including moon snails, crustaceans, and several fish, especially eels. Herbal pesticides are used to control predators. The razor clam is the first intermediate host of a parasitic worm ( Vesisocoe- liiini solenophagum Tong) and suffers frequent mortalities as a result (DFC 1979; Wang et al. 1993). Several finfish species are the final host of this parasite. SCALLOPS China produced 1 million metric tons of scallops from mari- culture in 1996. Most of the production was from two major spe- cies: the native zhikong scallop (Chlamys farreri Jones and Pres- ton) and the introduced bay scallop (Argopecten irradians La- marck) (Fig. 5A). The native zhikong scallop accounted for about three-quarters to four-fifths of the total, and the bay scallop ac- counted for one-fifth of the total or about 200.000 tons annually. There are two other species, Patinopecten yessoensis Jay (Fig. 5B ) and Chlamys nobilis Reeve, being cultured along the coast with little output. P. yessoensis was introduced from Japan. Its life history resembles that of the sea scallop. Placopecten magellani- cus, of the North Atlantic Ocean. It is a low-temperature species and is cultured only in the northern provinces. Liaoning and Shan- dong. It is larger in size and commands a higher market price than the zhikong and bay scallops, although its yield is low. C. nobilis occurs naturally in the South China Sea and southern Japan It is cultured in southern China on a limited scale. Zhikong Scallop The zhikong scallop is naturally found in north China. Korea, and Japan. It can survive at -1.5°C but does poorly when the temperature exceeds 25°C (DFC 1979. Wang et al. 1993). Most of the zhikong scallop culture is in Shandong and Liaoning provinces. Rizhao. in southern Shandong, probably represents the southern- most extent of its range. Shandong province is the leading pro- ducer, accounting for more than 80% of the national total. Zhikong scallop culture was first developed, between 1973 and 1983. using hatchery seed. Large-scale culture has led to the es- tablishment of breeding populations in many areas of Shandong. Now zhikong scallop culture uses almost entirely natural seed. Breeding grounds on Shandong peninsula currently provide suffi- cient seed for the scallop culture industry. One of the most pro- ductive bays in Shandong produced about 130 billion scallop seeds in 1996. The zhikong scallop has two natural spawning seasons in most areas, one in the early summer and one in the fall. Protocols for seed collection have been well established through years of Figure 5. Scallop culture. A. the introduced ba> scallop (upper) and the native zhikong scallop (lower) cultured in Shandong (photo by H. Yang). B, the Japanese scallop cultured in Dalian. C, ropes made from palm tree fibers used for the collection of scallop spat and seaweed seedlings in hatcheries. D. suspended longlines used for the culture of scallops, ahalone. Pacific oyster, and seaweeds in Rongcheng, covering most of the bay. E, lantern nets used for scallop culture. K, zhikong scallops are harvested at 1.5 years of age (photo by H. Vang). MOLLUSCAN AQUACULTURE IN CHINA 27 research and experience (DFC 1979, Wang et al. 1993). Successful collection involves site selection, preparation of collection mate- rial, and forecasting collection dates by obtaining data on gonadal development, larval stages, and density at the collection site. Seeds are usually collected using bags (30 x 40 cm) stuffed with nylon screens. Each bag may collect 100 to 1,000 spat, depending upon location, season, and year. Scallop culture primarily uses summer seeds, which set be- tween late May and early July. Commercial seeds, about 1-cm in size, are harvested and sold in October. They are then put in lantern nets and reared in a nursery area until the following March, when they reach about 3 cm and are ready for growout. Lantern nets on suspended longlines are used for scallop culture (Fig. 5D). The lantern nets are about 35-cm in diameter with 8-10 layers (Fig. 5E). In the nursery, about 200-300 seed scallops are stocked in one layer, or 2.000-3,000 per cage. The growout density is usually 30 to 50 scallops per layer, although higher densities are often used by farmers. One of the scallop farms we visited in Rizhao produces about 2.500 tons of whole zhikong scallops per year. Total annual production in the Rizhao area is about 80,000 tons. Typically, 1-cm scallop seeds are bought from northern Shandong in October. The seeds are maintained in a nursery area at five times growout density until the following March, when the scallops have reached 2.5-3.0 cm. Young scallops are thinned to 50-80 per layer or 400-500 per cage for growout. They usually reach market size. 6-7 cm, by December (Fig. 5F). The lantern nets are hung on longlines, which are usually 80 to 100-m long and supported by rubber floats. The water depth at this site is about 20 m. At the time of our visit (September), the lantern nets were heavily fouled by a variety of organisms. Scallop culture in Shandong is experiencing a mortality prob- lem. Mortality generally begins in early August as the water tem- perature reaches and exceeds 28°C. It lasts for about 20 days and ends when the temperature begins to decrease. Mortalities at the Rizhao site are in the 20 to 30% range each year, but they may reach 80% farther north, in sites nearer to where the seeds origi- nate. To the north, the timing of the deaths is similar to that in Rizhao, and the temperature is about the same or perhaps slightly (~1°C) warmer. The mortalities were first observed in 1994 at the Rizhao site and have continued since then. Both 1994 and 1995 were warmer (by 2 and 1°C. respectively) than normal. Tempera- tures in 1996 were more typical, and the death rate lessened. Mor- tality worsened in 1997 to 1998, reaching 80% or more in many areas throughout Shandong. There are several suspected causes for the scallop mortality, although they have not been studied extensively. Most scientists in China believe that the mortalities are caused by a combination ot overcrowding, high summer temperature, and deteriorating water quality. Scallop farmers often culture scallop at 2 to 3 times the density (30 to 50/layer) recommended by local scientists. The number of longlines and culture plots (not just for scallop culture) has been increasing rapidly in recent years and may have exceeded the carrying capacity of many coastal areas. Overcrowding at both cage and bay level may have added considerable stress to the culture environment. A haplosporidan parasite of the type respon- sible for extensive mortalities in oysters in the United States was identified in bay scallops in China, but there was no evidence that it was causing mortalities in that species or that it had been trans- ferred to zhikong scallops (Chu et al. 1996). Finally, there is also suspicion that the scallop stock is deteriorating. Although all seeds are collected from the wild, they are collected from a restricted area where the wild population is believed to have originated from hatchery production during the late 1970s and early 1980s. Bay Scallop The bay scallop is not native to China. It was introduced from the United States in 1982 by the third author and colleagues. Of the original shipment, a total of 26 bay scallops survived and were spawned in January 1983. producing the first generation of bay scallops in China (Zhang et al. 1986). The juveniles reached an average 6.9 mm in May and were transferred to culture sites in Shandong and Fujian provinces. The scallops grew to 50 mm by September and 59 mm by December 1983. Market size. 50 to 60 mm. can therefore be reached within a year. This is a major ad- vantage over the zhikong scallops, which usually take 1 .5 to 2.0 years to reach market size. The shorter turn-around time of bay scallops is partly attributable to its faster growth, and partly to the fact that they are spawned in the early spring (or late winter) so that they catch a full growing season. Because zhikong scallop seeds are collected in the fall, they miss most of the first growing season. Because of the short growout time, bay scallops quickly gained acceptance by scallop farmers, and aquaculture expanded rapidly after 1984. By early 1990, the annual production of bay scallops had reached about 200,000 tons. Bay scallop culture pro- duction has declined somewhat in the past few years. Because of the recent summer mortality problem in zhikong scallops, bay scallop culture may increase again. Bay scallop seeds are produced exclusively in hatcheries, where thousands of mature adults, which are hermaphroditic, are placed in lantern nets and induced to spawn in large concrete tanks, ranging from 10 to 100 cubic meters. When the desired egg density is reached (about 50/mL), the adults are moved to the next tank to continue spawning. The first water change is made as soon as the larvae reach D-stage, which is usually 24 h after spawning. Larvae reach the eyed stage in 10 to 14 days at a size of 170-190 p.m. at which time spat collectors are placed in the tanks. Two common types of spat collectors are used. One is a rope curtain made from natural palm tree fiber (Fig. 5C). The other is polyethylene or nylon nets/screens. Spat, which attach to the collectors by byssal threads, are cultured in the hatchery until they reach 500-600 p.m, after which they are transferred to shrimp ponds or nursery areas. Commercial seeds are sold at a size of 0.5 to 1 .0 cm. Lantern nets on suspended longlines is the predominant form of culture for all scallops. The culture density of bay scallops is about the same as the zhikong scallops. 30 to 50 scallops per layer, 250 to 400 per net. Hatchery production of larvae usually occurs between March and May, and scallops are harvested between November and De- cember. Most of the bay scallop production continues to be from the 26 scallops first introduced in 1982. and there have been signs of inbreeding, such as larval and juvenile mortality. Several new broodstock introductions have been made to expand the gene pool, but offspring of the recently introduced scallops have not yet en- tered the mainstream production. Most of the scallops are processed upon harvest. Adductor muscles are either individually frozen, or cooked and then dried. A small fraction of scallops is sold live to local restaurants. MUSSELS Mussel culture is relatively new in China. It started with blue mussel in the 1950s as a byproduct of seaweed {Laminaria 28 GUO ET AL. japonica) culture. Five species of mussel are cultured commer- cially. The most widely cultured is the blue mussel. Mytilus edulis Linnaeus (Fig. 6A). which is produced chiefly in Shandong and Liaoning provinces. Four other species, the thick-shell mussel {Mytilus coruscus Gould), the Senhouse mussel [Musculus sen- housei Benson), the green-jade mussel {Perna vividis Linnaeus) (Fig. 6B), and the penshell (Pinna pectinata Linnaeus) (Fig. 6C). are also cultured in Guangdong and other southern provinces on limited scales. The Senhouse mussel is primarily used for shrimp, fish, and chicken feed. All five species are native to China. For blue mussel culture, hatchery-produced seeds were used to supplement wild set during the 1970s. Now wild mussel seeds are abundant, and hatcheries are used for other molluscan species. Seeds are usually collected in May and June, occasionally also in the fall. When the juveniles reach about 10 mm, they are thinned and reattached to ropes for growout. Culture on suspended long- lines is the predominant form of growout for the blue mussel. Bottom culture is also used for the green and thick-shell mussels in the south. A portion of the juveniles that set in May and June are har- vested between October and December, at a size of 60 to 70 mm. Smaller mussels are returned and cultured for longer periods, but most mussels are harvested within a year. Mussels with full gonads are considered most desirable on the market. Some mussels are sold fresh on local markets, but most are steamed and dried into a traditional product called "Dan-cai." Mussel meat that is dried without cooking is called "butterfly meat." Some mussels are cooked to produce "oyster sauce." Mussels are also used as feed for cultured shrimp and Rapana snails. Mussel culture has been in decline in recent years probably because of competition from other high-value species, such as scallops, oysters, and abalone. ABALONE AND OTHER GASTROPODS Abalone culture is new in China. Large-scale culture started in the late 1980s, and it has been developing rapidly in the past decade. Now it is one of the largest components of the molluscan culture industry by value. It is probably also the most sophisticated industry in terms of production technology. Abalone is not indi- vidually listed in official statistics, and production estimates vary greatly. The annual production is probably between 500 to 900 tons. Two species of abalone are cultured: the wrinkled abalone, Haliotis discus hannai Ino (Fig. 7E) and the colorful abalone, Haliotis diversicolor Reeve. The colorful abalone is a southern species and is cultured mainly in Guangdong and Fujian. The wrinkled abalone is the major species and accounts for about three- quarters of the total production. It is a northern species and is cultured mainly in Shandong and Liaoning provinces. Liaoning's Dalian area is the leader in abalone aquaculture. and much of the technology has been developed there. There are three major aba- lone culture companies in Dalian: the Bilong Seafood Co.. Pacific Seafood Co., and Xinda Products Co. We visited all three facili- ties. The Bilong Seafood Co. has been the leader in abalone pro- duction for some time. The Pacific Seafood Co., which was es- tablished in 1993 with an investment of US$8.5 million, is posed to become the largest abalone culture company in China. It is designed to produce 600 tons of abalone per year, although it had not produced the first crop when we visited in 1997. Xinda has successfully used hybrid abalone to combat disease problems. Abalone culture in Shandong province is catching up fast. Over 40 large abalone facilities with investments of more than US$1 mil- lion each, have been built in Shandong's Rongcheng area. Most of the new facilities have not reached their full production capacity, and many are running at a loss because of disease problems. On the other hand, one facility in Rizhao that we visited has recovered all its investment in 3 years. The Rizhao facility is primarily designed for hatchery production of seed and has little growout production. All abalone seeds are hatchery produced with well-developed technology. Production starts in early spring with broodstock con- ditioning at elevated temperatures. After about 1.000 degree-days of conditioning (at about 18-20°C), abalone are ready to spawn. To induce spawning, they are left without water for 1 hour and then exposed to UV-treated seawater (600 u.W/h/nr1). Males usu- ally begin to spawn 1 hour after being in UV-treated water, and females within 30 to 40 min after the males. Fertilized eggs are incubated at density of 15-20 mL. Eyed larvae are set on corru- gated plastic plates, which are precoated with a layer of diatoms to induce settlement (Fig. 7A). Abalone spat remain on the settlement plates until they reach 3 mm. After that, they are separated from the settlement plates and transferred to large punctured plastic plates for nursery culture (Fig. 7B). The holes allow for better water circulation and for the young abalone to move from side to side. The plates are supported in net-pans and placed in raceways, typically 0.5-m deep. 1 to 2-m wide, and 10 to 20-m long (Fig. 7C). Commercial diets (formulated) are used during the nursery phase. Juvenile abalones are cultured to a size of 1 to 2 cm in the nursery before growout. Abalone seeds sell for about $0.25 each. Three major forms of growout are used in abalone culture. The N& &JP&T V* A--*" Figure 6. Mussel culture. A, the blue mussel harvested as a byproduct from seaweed and oyster longlines in Dalian. B, green-jade mussel cultured in Guangdong and Fujian. C, the penshell cultured in Zhejiang. MOLLUSCAN AQUACULTURE IN CHINA 29 Figure 7. Abalone culture. A, settlement plates are coated with diatoms before use in a Dalian hatchery. B. juvenile abalones are cultured on plates in indoor raceways in a Dalian hatchery. C, typical indoor raceways used for abalone larval culture and nursery. D, large and specially designed cages for abalone culture on suspended longlines, Dalian. E, close to market size abalone cultured in raceways inside an abandoned air-raid bunker in Lianyungang. F. sea urchin and sea cucumber are alternative species cultured in abalone hatcheries. first, and most widely used, is culture in cages on suspended long- lines. Most abalone cages are much larger and more sophisticated than the lantern nets used for scallops (Fig. 7D). They are about 70-cm in diameter and l-m tall with 4—5 layers. One type of cage is made from large plastic tubes (35-cm in diameter and 60-cm long), with screens on each end and 1-cm holes on the side. Some farmers also use scallop lantern nets for abalone culture. The sec- ond form of culture is on plastic plates placed in indoor concrete raceways (Fig. 7E). Abandoned air-raid bunkers are now popular places for indoor abalone culture. The third form, which is not widely used, is in intertidal ponds. Intertidal net fences are also used in some areas for abalone culture. Fresh kelp, mostly Lcuni- naria japonica, is the primary food during growout, and artificial diets are used when algae are not available. Abalone is considered to be one of the best and most valuable seafoods in Chinese culture and other parts of Asia. Commercial size abalone (7 to 9 cm) is priced at S3 to $4 per animal, and a large proportion of the abalone produced in China is sold live to markets in Hong Kong and Japan. Some abalone are sold to local restau- rants. Abalone is also valuable as an ingredient in Chinese medi- cine. Abalone culture was highly profitable in the late 1980s and early 1990s, but recent disease problems have been plaguing aba- lone culture in the north. Many abalone facilities in Shandong and Liaoning have reduced or stopped abalone culture and started growing sea urchins and sea cucumbers (Fig. 7F), which have similar culture requirements. One of the diseases is known as the pustule disease and is caused by Vibrio fluvialis-U (Li et al. 1998). Another major dis- ease affecting the wrinkled abalone has a distinct syndrome. It includes a long incubation period and slow disease progression and is characterized by a shrinking of the meat within the shell. The condition is reminiscent of "Withering Syndrome", which affects wild black abalone (Haliotis cracherodii Leach) along the Cali- fornia coast. This disease is transmissible and is strongly associ- ated with a rickettsial infection (Gardner et al. 1995, Friedman et al. 1997). A similar syndrome of Nordotis discus discus in Japan is associated with a virus (Otsu and Sasaki 1997). Whether these syndromes are related is presently unknown. Several other gastropods are also cultured in China, including the red conch (Rapana venosa Valenciennes), the mud snail iBul- lacta exarata Philippi) and the sea hare (Notarchus leachii cirro- sus Stimpson). For the red conch, juveniles are collected from the wild and cultured in cages or lantern nets. They are fed with blue mussels. The mud snail is a small gastropod, belonging to Order Cephalaspidea (bubble shells), with an adult size of 2 to 3 cm. It has a large foot that covers much of its thin shell. The culture of mud snails involves the selection of muddy flats that have heavy larval settlement. The flat is treated with pesticides to remove predators and fertilized to stimulate diatom growth just before settlement occurs. No other management is required. Production at harvest usually ranges between 35-75 metric tons per hectare. The mud snail, pickled in liquor, is a delicacy in the Zhejiang and Shanghai areas and is priced at about $l/lb. Mud snail culture is a significant industry in Zhejiang and is highly profitable. The sea hare has been cultured for hundreds of years in southern China, where juveniles are collected from the wild and cultured in ponds. 30 GUO ET AL. Sea hares are not cultured for their meat, but for their egg cases, which when processed, constitute a valuable remedy in Chinese medicine, referred to as "Sea Powder." PEARL OYSTER Pearls are produced from both freshwater and marine bivalves. China has a long history of culturing freshwater pearls and cur- rently produces about 800 tons annually, mostly from the fresh- water mussel Hyriopsis cumingii Lea. Freshwater pearls are not only marketed as jewelry, but also used as ingredients in Chinese medicine and cosmetic creams. Marine pearls have a higher market value than freshwater pearls. Wild marine pearls have been har- vested for several thousand years in China, but artificial culture is less than 50 years old. China now produces about 20 tons of marine pearls annually, second to Japan's production of about 40 tons. Marine pearls are produced by species of Pteriidae. In China, almost all cultured marine pearls are from the Martensii pearl oyster, Pinctada martensii Dunker (Fig. 8A). This species is also the major species cultured in Japan and accounts for over 95% of worldwide marine pearl production by weight. Other species such as Pinctada maxima Jameson and Pinctada margaritifera Lin- naeus are also cultured experimentally in China. Pearls from the latter two species are more valuable because of their larger size, unique coloration, or both. Marine pearls are primarily cultured in provinces on the South China Sea. Guangdong and Guangxi provinces produce over 90% of China's total. We visited Beihai, the pearl city of China, where the famous Hepu pearls are produced. This area has about 360 pearl oyster hatcheries and 2,000 farms, and produces about 40- 50% of the national total. All seeds are hatchery produced using methods similar to those used for bay scallops. As in other mol- luscan hatcheries, simple technologies are often practiced. One example is the culture of age in plastic bags (Fig. 8B). Hatchery production usually starts in April. When spat reach 1 mm, they are put into fine-mesh bags and moved to nursery areas in the sea. Seeds are transferred to growout cages at a size of 5-8 mm and a density of about 5,000 per cage. Unlike the multiple-layer lantern nets used for scallops, cages used for pearl oyster culture are small (25 x 25 x 10 cm), single compartment units (Fig. 8C). They consist of metal frames with nylon net sides. Some cages are made with a metal ring, about 30-cm in diameter (Fig. 8D|. These cages are hung on suspended rafts, longlines. or intertidal longlines that are supported by wood stakes about 60-cm tall. Juveniles are thinned five to seven times during a 2 to 3-year period to a final density of 30-50 adults per cage. Pearl oysters are cultured for 2 to 3 years to a size of 50-70 mm before being used for pearl production. Marine pearls are produced by inserting a nucleus (5-8 mm) attached with a small piece of mantle (2-3 mm) from a donor oyster into the mantle of the re- cipient pearl oyster. The attached mantle tissue will grow, encap- sulate the nucleus, and produce nacre. Freshwater pearls are usu- ally produced by inserting a piece of mantle only. Nucleus pearls have recently been introduced to freshwater pearl production and have become strong competitors with marine pearls. Nuclei are usually made from molluscan shells or synthetic materials. The number of nuclei inserted per oyster depends upon the size of the nucleus and the oyster, but on average, about two nuclei per oyster are inserted. Nucleus insertion is performed in spring or fall. Sum- mer is avoided because of the additional stresses of high tempera- ture and reproduction. When nuclei are inserted between February and April, pearls are harvested in November and December of the Figure 8. Pearl oyster culture. A, the outside (upper) and inside (lower) appearance of the Martensii pearl oyster used for the production of marine pearls in southern China. B, plastic bags are used for algae culture (second stage) in many molluscan hatchery; the last stage culture is usually done in shallow concrete tanks. C and D, single compartment cages used to culture pearl oysters on intertidal longlines in Beihai. E, pearls are harvested by sacrificing oysters; one pearl per oyster is expected, on average. F, pearl necklaces on display at markets in Beihai. the pearl city of China. M( 1LLUSCAN AQUACULTURE IN CHINA 31 same year. The recovery rate is about 50% or one pearl per oyster. It takes about 10 million pearl oysters at harvest to produce one ton of pearls. At harvest, pearl oysters are sacrificed to collect the pearls (Fig. 8E). The meat is used for chicken and duck feed. Raw pearls come in various shapes and colors. They are sorted and processed in a series of chemical treatments before reaching the jewelry market (Fig. 8F). They are classified into four size categories: large (>8 mm), medium (6-8 mm), small (5-6 mm), and fine (<5 mm). Large pearls are worth considerably more than smaller ones. Nucleus-free pearls are produced by inserting only donor tissue and are intended for use in Chinese medicine. Pearl powder from pearls and shells is used in toothpaste and facial creams. One problem in pearl production is sexual maturation. When pearl oysters are full of gametes, nucleus insertion is difficult, and the survival and pearl recovery rates are low. Some farmers inhibit maturation in the fall by placing oysters at high density and in deep water. Mature oysters are sometimes induced to spawn beforehand to improve their condition for nucleus insertion. Sterile triploids are being tested for pearl production at the South China Sea In- stitute of Oceanology in Guangzhou and Guangxi Institute of Oceanology in Beihai. and preliminary results are encouraging. PERSPECTIVES Molluscan aquaculture in China is impressive in both scope and magnitude. It is practiced in almost all inhabited parts of the coast and covers all major molluscan species. A wide range of produc- tion technology is used, ranging from sophisticated intensive cul- ture of abalone, scallop, and pearl oysters, to primitive extensive culture of certain clams and snails. Some practices are distinctive. One example is the "semiartificial" collection of clam seed, which involves site selection, bed construction, substrate modification, larval forecast, and predator control. Another example is seed pro- duction from earth ponds, which seems to be an effective approach under low-tech conditions. Polyculture of mollusks, shrimp, and/or fish in ponds is widely practiced. The past decade represents the fastest growing period for mol- luscan aquaculture in China. This period coincides with rapid growth in the overall Chinese economy and is primarily influenced by China's economic reforms. The replacement of central planning with a market economy is probably the leading force responsible for the rapid growth. Marine mollusks are among the best-loved seafood in China. As the Chinese economy grows and people's income rises, demand for mollusks will continue to increase and prompt further growth of the aquaculture industry. The rapid development of molluscan aquaculture has brought with it some problems, the most pressing of which is the deterio- ration of the culture environment. The expansion of mariculture during the past decade has put considerable stress on the marine environment, and the carrying capacity of the coastal water may be exceeded in many areas. Longlines and rafts often cover much of a bay (Fig. 3B. E; Fig. 5D). Large shrimp ponds are densely situated (Fig. 4F). Scallops are often cultured at excessive densities at both cage and baywide levels. Excessive feeding from shrimp and abalone culture leads to accumulations of tremendous waste. Red tides have become more frequent along China's coast. Over- crowding and poor water quality, coupled with high water tem- perature, are believed to be the leading causes for the massive scallop mortalities in 1997 to 1998. Abalone culture has been seriously affected by diseases. There is a great need for the devel- opment of new management strategies to maintain yields while minimizing conditions that degrade the environment, cause dis- ease, or both. Future growth of the molluscan aquaculture industry may largely depend upon technological advances that make mol- luscan aquaculture more efficient and environmentally friendly, instead of crude expansion in scale. ACKNOWLEDGMENTS We thank all our hosts for their hospitality and assistance dur- ing our visits. Many Chinese scientists and aquaculturists contrib- uted personal knowledge to this paper. We particularly thank Pro- fessors Rucai Wang, Zhaoping Wang, Huiping Yang, Guofan Zhang. Zhihua Lin, Aimin Wang, and Zhinan Zeng for discussion and comments. This study and our visits were supported partly by NOAA's US-China Joint Program in Marine Living Resources. China's State Bureau of Foreign Experts, Institute of Oceanology Chinese Academy of Science, and Rutgers University. This is publication No. 99-14, IMCS/NJAES. LITERATURE CITED Cai, Y. & X. Li. 1990. Oyster culture in the People's Republic of China. World Aquacult. 2 1 :67-72. Chu. F.-L. E., E. M. Burreson, F. Zhang & K. K. Chew. 1996. An uniden- tified haplosporidian parasite of bay scallop Argopecien irradians cul- tured in the Shandong and Liaoning provinces of China. Di.s. Ai/imr. Org. 25:155-158. DFC (Dalian Fishery College). 1979. Molluscan aquaculture. Agriculture Press. China (in Chinese). FAO. 1997. Review of the state of world aquaculture. FAO Fisheries Circular No. 886. Rev. 1. Rome. Friedman. C. S.. M. Thomson, C. Chun, P. L. Haaker & R. P. Hedrick. 1997. Withering syndrome of the black abalone, Haliotis cracherodii (Leach): water temperature, food availability, and parasites as possible causes. J. Shellfish Res. 16:403^1 1 Gardner. G. R.. J. C. Harshbarger, J. L. Lake. T. K. Sawyer. K. L. Price, M. D. Stephenson, P. L. Haarker & H. A. Togstad. 1995. Association of prokaryotes with symptomatic appearance of withering syndrome in black abalone Haliotis cracherodii. J. Invertebrate Pathol. 66:111- 120. Li, G, Y. Hu & N. Qing. 1988. Population gene pools of big-size cultivated oysters (Crassoslrea) along the Guangdong and Fujian coast of China. Proceedings of Marine Biology of the South China Sea. 51-70. Li. T., M. Ding, J. Zhang. J. Xiang & R. Liu. 1998. Studies on the pustule disease of abalone (Hatliotis discus hanni Ino) on the Dalian Coast. J. Shellfish Res. 17:707-711. MAC (Ministry of Agriculture of China). Bureau of Aquatic Products. 1986-1997. China Fishery Annual Statistics. Beijing, China (in Chi- nese). Otsu, R. & K. Sasaki. 1997. Virus-like particles detected from juvenile abalone (Nordotis discus discus) reared with an epizootic fetal wasting disease. J. Invert. Pathol. 70:167-168. Perdue, J. A., J. H. Beattie & K. K. Chew. 1981. Some relationships be- tween gametogenic cycle and summer mortality phenomenon in the Pacific oyster (Crassoslrea gigas) in Washington state. J. Shellfish Res. 1:9-16. Wang. R., Z. Wang & J. Zhang. 1993. Marine molluscan culture. Qingdao Ocean University Press, Qingdao, China (in Chinese). Zhang. F.. Y. He. X. Liu. J. Ma. S. Li & L. Qi. 1986. The introduction, hatchery rearing, and culture of bay scallops. Oceanol. Limnol. Sinica 17:367-374 (in Chinese with English Abstract). Journal oj Shellfish Research, Vol. 18, No. 1. 33-39. 1999. SETTLEMENT OF THE BLUE MUSSEL MYTILUS GALLOPROVINCIALIS LAMARCK ON ARTIFICIAL SUBSTRATES IN BAHIA DE TODOS SANTOS B.C., MEXICO SERGIO CURIEL RAMIREZ AND JORGE CACERES-MARTINEZ Centra de Investigation Cientifica y de Education Superior de Ensenada Departamento de Acuicultura Apartado. Postal 2732, 2800 Ensenada, Baja California, Mexico ABSTRACT The culture of the blue mussel Mytilus galloprovincialis in Bahi'a de Todos Santos is a growing economic activity. This culture depends upon mussel seed collection from artificial collectors; however, there are no studies on time and duration of the mussel settlement season, collector material, and settlement pattern. Between December 1994 and November 1995, pieces of about 163 cm2 of nylon ropes, polypropylene ropes, ropes made with polypropylene and cotton, pads of synthetic fibrous material, and dried Luffa, sp. were used as collectors and were deployed at 2- and 5-m depths. Mussel settlement occurred during all the period of study, and its fluctuation was similar for all collectors tested. Major settlement occurred in December and January for all collectors and depths studied. The observed settlement pattern indicates that direct settlement of competent pediveligers from the plankton onto the substrates is the main source of recruitment in the area {929c ). There was a trend of greater settlement on pads of synthetic material than on the other collectors. This material seems to be appropriate for scientific studies; whereas, for commercial activity, any filamentous rope collectors are recommended. KEY WORDS: settlement. Mytilus galloprovincialis, mussel seed collectors, dispersion and culture INTRODUCTION The culture of the blue mussel Mytilus galloprovincialis using submerged longlines in Bahi'a de Todos Santos, Baja California (Fig. 1) started in 1991 (Caceres-Martfnez 1997). At present, the annual production is around 150 metric tons and it is marketed in Mexico and in the United States. Submerged longlines. 200-m long, are suspended from 200-L plastic floating barrels and are anchored with 0.8 or 1.2 ton concrete anchors. The main line is placed at a 5-m depth, from which culture ropes, 7-m long, are suspended. Mussel seed is obtained from nature on artificial col- lectors placed from late November to December. The collectors consist of a polypropylene rope placed inside a thin plastic net and suspended from surface longlines (Caceres-Martfnez 1997). This practice and type of collectors were established on the basis of the experience of mussel growers. On the other hand, it is known that colonization on natural and artificial substrates by 116- «'\ 1 >0- 05' "31" 50' ^^Ensenada Babia de Todos \ sanlos V Pacific Ocean A Mussel cultured ^ I) area 1 J V lslas Todos ,— ^?L/ Xi "aJa Santos ^s_^ California " 31* «' \ r.lK.r.1. !^_ \Js \ Figure 1. Map showing the study area, filled oval indicates the culture area were collectors were deployed. mussels, Mytilus sp. may occur by settlement of competent pedi- veliger larvae and/or by settlement of drifting postlarvae (Davies 1974, King et al. 1989, Caceres-Martmez et al. 1993, Caceres- Martfnez et al. 1994). This settlement pattern could have practical importance for the mussel grower, because its knowledge in a given area may improve the chance of collecting mussel seed from nature (Caceres-Martfnez and Figueras 1998). However, there are no scientific studies in the area to corroborate or improve mussel seed collection practices. The aim of the present study is to determine the time and duration of the mussel settlement season and the relative percent- age of competent pediveligers to postlarvae during settlement on different artificial collectors deployed at two depths in Bahi'a de Todos Santos. MATERIALS AND METHODS The study was performed in the mussel culture area of Bahfa de Todos Santos, which is approximately 18-km long and 14-km wide and has a surface area of 252 km2; it has a sandy bottom, and it is partially separated from the ocean by the two small islands. Islas de Todos Santos. The study was carried out from December 1994 to November 1995. Pieces of 25-cm long and 2-cm diameter (163 cm2) of nylon ropes, material similar to that used by mussel growers (FN); polypropylene ropes (FP); ropes made with poly- propylene and cotton (FPC); pads of 25-cm long and 6.5-cm wide (163 cm2) of synthetic fibrous material (Commercial Scotch Brite™) (SF) and similar pieces of dried fibrous Luffa sp. (Cucur- bitacea) (L) were used as collectors. Ropes were unraveled by passing them through a grinding machine to increase their fila- mentous nature. A comparison among surface area of the sub- strates was relative because of the difficulty for determining their exact surface area. The five collectors were attached to a PVC tube and hung at 2- and at 5-m depths, covering part of the depth at which mussel growers place their collecting ropes (1 to 7 m-depth). Collectors 33 34 Ramirez and Caceres-Martinez III.. I I, I ill. ..Ill 3 "5 to 5 2 E o ■ S3 I 1 c en o _l DJ F MAMJJ ASO N FCP I hitafc ft H nini DJ FMAMJJA SON I J, ■ II M. U DJFMAMJJASON DJFMAMJ JASON I DJFMAM JJASON Months Figure 2. Logarithm of the mean number of mussels settled (+ standard error) on different collectors placed at 2- (open bars) and 5-m (filled bars) depth, during a study period in Bahia de Todos Santos, Mexico. This graphic representation allows better visualization of results during low settlement periods. were replaced after periods of 30 ± 5 days. Each collector was taken to the laboratory in one plastic bag, then the collectors were immersed individually in a 10% solution of commercial sodium hypochlorite (Na CIO) for 5 min. (to dissolve organic material and to facilitate detachment of mussels) and were rinsed with running water directly onto a 0.09-mm sieve. After that, the seed was dried in an oven at 70°C for 24 h and passed through a series of sieves TABLE 1. Expected size of mussels during time intervals studied. Theorical Theorical Date of Time for Expected Expected Collector Temperature Colonizing Size (mm) Size (mm) Replacement <°C) (days) 2-m depth 5-m depth 15.12.94 35 1.212 1.210 08.02.95 54 1.687 1.685 27.02.95 19 0.812 0.810 29.03.95 17 30 1.087 1.085 25.04.95 18 27 1.012 1.010 18.05.95 17.6 23 0.912 0.910 16.06.95 20 29 1.062 1.060 13.07.95 20.7 27 1.012 1.010 24.08.95 21.4 42 1.387 1.385 28.09.95 18.6 35 1.212 1.210 30.10.95 18.6 32 1.137 1.135 22.11.95 17 23 0.912 0.910 of 0.09 to 0.7-mm mesh to facilitate mussel separation by size. Three subsamples of mussels were obtained from each sieve and counted. For graphic representation, data for Fig. 2 were log trans- formed. Thirty individuals per fraction were measured with an ocular micrometer under a stereoscopic microscope or with an electronic caliper if the mussel size was >5 mm to determine size distribution. Competent pediveliger larvae were separated considering mus- sels with shell lengths from 0.250 to 0.470 mm (mean 0.360 mm) according to the minimum and maximum size values recorded for this stage of Mytilus sp. (Rees 1954, Bayne 1965, Widdows 1991 ). Moreover, the growth rate of recently settled Mytilus galloprovin- cialis [ca. 25 u.md_1 at 15 to 17°C (Aguirre 1979)] was considered to estimate the probable increase in shell length of mussels during the sampling period, mussels greater than expected will confirm the attachment on collectors of drifting postlarvae. Water tempera- ture was recorded during samplings. The number of spat from different collectors and depths studied were compared using an analysis of variance (ANOVA). RESULTS Settlement fluctuations throughout the study period on different collectors are shown in Figure 2. In general, in all the collectors, a similar fluctuation was detected at the two depths studied: the settlement began to rise in December, with a peak in January (in these months, the 78% of the total spat obtained during all the Ramirez and Caceres-Martinez 35 study was recorded), and it remained low throughout the other months, except September, when a light increase was detected. This general fluctuation was best recorded on collector SF where the greater quantity of spat was detected. Particular differences are observed among the obtained spat on the studied collectors and depths: settlement seems to be more abundant at 2- rather than 5-m depth. This occurs during December and January in FP, January in FN, December in FPC, December and January in SF. and Decem- ber in L. The loss of FPC and L in January prevent a comparison in these collectors. During the months of low settlement, this dif- ference is not clear. The weakness of L favored their frequent loss during the study period. Statistical comparison of the spat among collectors showed that differences were neither significant (F = 2.18. p = .7) during the study period, nor between depths (F = 0.92. p = .34). Expected sizes of spat during permanence time on collectors and the temperatures recorded are shown in Table 1. Size dis- tributions of the spat in different collectors are shown in Figures 3 to 7. In general, a similar distribution was recorded in all collec- tors and depths. The maximum size recorded in December was 1.59 mm in all the collectors and at both depths; whereas, in January, it was 5.99 mm in FP at 2-m depth. 3.74 and 3.70 in FN and L. respectively, at 2-m depth, and <3.59 in the other collectors and depths. During December and January, about 7% of mussels in all the collectors and depths reached a size greater than ex- pected, lately this percentage accounts for \c/c. During the low settlement period, the maximum sizes recorded were 12.49 mm during April in SF at 2-m depth, 10.32 mm during February in FP at 5-m depth, and 5.33 during August in FPC at 2-m depth. In the other collectors and depths, the maximum sizes were <3.59 mm. Sizes of <0.470 mm were recorded during all the year and in all collectors. DISCUSSION The presence of mussels <0.470 mm throughout the year and their abundance in December and January reflects the presence of spawning mussels throughout the year and the occurrence of a major spawning period during late autumn. The reproductive cycle of Mytilus galloprovincialis in Bahfa de Todos Santos has not been studied; however, in the west coast of North America, the major spawning season of M. edulis (after the works of Harger ( 1972) in Santa Barbara, California, the Mytilus edulis-like in southern Cali- fornia was identified as M. galloprovincialis (see Mc Donald and Koehn 1988. Koehn 1991)) takes place in winter (see Suchanek 1981). It is known that in populations of M. edulis and M. gallo- December n 2=2 56 n5=159 Hk March n2=43 n5=46 100 January n2=1079 n5=458 80 60 40 _^ — , — 20 TlT~L 0 _j)M_JB_j»=b 100 April n5=9 80 60 40 . f 20 0 -II 100 July n2=17 n5=8 100 February n2=41 n5=18 80 60 40 • | 20 0 100 May n2=12 n5=11 IL. 100 80 August n2=46 D □ September n2=60 n5=69 [L 0.2S 0.471 0.9 1.6 2.6 >3.6 0.47 0.899 1.599 2.699 3.699 100 October n2=22 n5=33 100 November n2=2 n5=6 80 80 60 60 40 r| ■ 40 "I 1 . 20 0 3 0.26 0.47 I 1 2°o j 1 I ■ 2.6 3.599 0.471 0.9 1.6 0.899 1.699 2.699 2.6 3.699 >3.6 0.26 0.47 0.471 0.9 1.6 0.899 1.699 2.599 >3. Size classes of mussels Figure 3. Size distribution in percentage of mussels settled on collectors FP placed at 2- (open bars) and 5-m (filled bars) depth, during a study period in Bahia de Todos Santos, Mexico. 36 Ramirez and Caceres-Martinez i 40 December January n2=151 n5=206 too so 60 40 n2=739 n5=638 L 20 JO S 60 in 3 E ■8 20 £. ° s c O 100 a. March n2=26 n5=41 [K June n2=8 nS=8 Jl □ September n2=52 n5=51 [L 0.25 0.471 0.9 1.6 2.S >3.6 0.47 0.899 1.599 2.599 3.599 April n2=3 n5=6 .■I October n2=26 n5=13 0.25 0.471 0.9 1.6 2.6 >3.6 0.47 0.899 1.599 2.599 3.599 February n2=24 n5=14 II 100 80 May n2=12 n5=2 CD August n2=33 dD □ D November n2=39 n5=13 0.25 0.471 0.9 1.6 2.6 >3.6 0.47 0.899 1.599 2.599 3.599 Size classes of mussels Figure 4. Size distribution in percentage of mussels settled on collectors FN placed at 2- (open bars) and 5-m (filled bars) depth, during a study period in Bahia de Todns Santos, Mexico. provincialis from different areas of the world, after a main spawn- ing season, minor spawning during the year may occur (Seed 1976. Ferran 1991. Vfflalba 1995. Caceres-Martinez and Figueras 1998). Differences between reproductive cycles from mussels located in different environmental conditions could be explained by tempera- ture, salinity, photoperiod. food, nutrient reserves, hormonal cycle, and genotype differences (Seed 1976. Devauchelle and Mingant 1991, Robinson 1992. Seed and Suchanek 1992. Couturier 1994). It is important to note that M. califomianus is present in the exposed rocky shores of the ocean side of the Bahia de Todos Santos, where M. galloprovincialis is found only rarely. To the contrary, M. galloprovincialis is clearly dominant in the protected areas of the bay and obviously in culture area, where M. califor- nianus is rarely seen. However, the occasional presence of M. califomianus in the culture area, suggests that some larvae of this species may reach and survive culturing conditions, then it is rea- sonable to assume that some settled larvae may belong to this species. However, morphological identification of mussels <3 mm is practically impossible (Rees 1950. Loosanof et al. 1966. Hines 1979, personal observation). To resolve this point, it is necessary to do detailed studies on identification throughout the molecular genetics of mussel larvae and postlarvae. In this study, we assume that the major recorded spat corresponds to M. galloprovincialis, taking into account the prevalence of M. galloprovincialis and that its reproductive cycle presents one major spawning season in the west coast of North America with respect to the reproductive cycle of M. califomianus without a major spawning season (Young 1946. Suchanek 1981. Hines 1979, Curiel-Ramirez and Caceres- Martinez unpublished data). The similarities found in the settlement of mussels in different collectors and depths suggest a relatively uniform presence of competent pediveligers and postlarvae in the study area. A trend similar to a major settlement in surface collectors was found by Fuentes and Molares (1994) and Molares and Fuentes (1995) in Ria de Arosa, Spain, but a contrary tendency was found by Cac- eres-Marti'nez and Figueras (19981 in Ria de Vigo. Spain. These results have been explained by the presence of the thermocline and by the behavior of larvae and postlarvae settlement (Fuentes and Molares 1994, Caceres-Martinez and Figueras 1998). The prefer- ence of mussels to settle in substrate has been widely studied (de Blok and Geelen 1958. Bohle 1971, Davies 1974. Dare et al. 1983. Eyster and Pechenik 1987. King et al. 1990. Caceres-Martinez et al. 1994). and it has been found that rugose and filamentous sub- strates are the best for settlement, because thev are related to the Ramirez and Caceres-Marti'nez 37 December n2=279 n5=63 100 80 60 40 20 0 January n5=531 III. February ioo n2=18 n5=171 J March n2=33 n5=22 June n2=14 n5=14 April n2=9 n5=9 .ll I July n2=5 n5=3 May n2=14 n5=4 August n2=38 □ D D □ □ September n2=77 D 0.26 0.471 0.9 1.6 2.6 0.47 0.899 1.S99 2.599 3.599 October n2=27 n5=30 100 n2=27 n5=2 so 60 I ■Qhu 0.25 0.471 0.9 1.6 2.6 >3.6 0.47 0.899 1.599 2.599 3.599 Size classes of mussels November n2=11 n5=29 .1 0.26 0.471 0.9 1.6 2.6 >3.6 0.47 0.899 1.599 2.599 3.599 Figure 5. Size distribution in percentage of mussels settled on collectors FPC placed at 2- (open bars) and 5-m (filled bars) depth, during a study period in Bahia de Todos Santos. Mexico. use of long contact mucous threads of competent pediveliger and postlarval stages that are easily jammed between filaments and rugosities (de Blok and Tan Mass 1977, Caceres-Marti'nez et al. 1994). However, comparison among efficiencies of filamentous substrates is difficult to do, not only because of the enormous surface variability of filamentous artificial substrates, but also be- cause these substrates are colonized immediately in the water by filamentous algae, hydroids, and debris that modify their surface. In laboratory studies where small pieces of substrates are used and environmental conditions are under control, image analysis has proved to be a useful tool to calculate the substrates surface area (Caceres-Martinez et al. 1994). However, this is hardly applicable to large collectors used in field studies. In these circumstances, a useful comparison among filamentous substrates is related to du- rability of collectors, handling, cost, nature of the study, and com- mercial use. Despite the fact that statistical analyses indicate that there were no differences among the spat recorded in the collectors studied, there was a clear trend of major settlement on commercial pads (SF). The observed tendency allows us to suggest the use of SF for scientific studies in the field or laboratory; however, their handling for commercial purposes is limited because of the diffi- culty of extracting the seed, in comparison with the use of fila- mentous ropes. Therefore. FP. FN. and FPC could be useful for commercial purposes. The use of L is limited because of its weak- ness for handling and its low durability in the sea. Although 92% of total spat was considered as competent pe- diveliger that arrived and grew on the collectors during the per- manence of the substrates underwater between the sample collec- tion, the presence of mussels larger than expected confirm the recruitment of at least a small proportion of postlarvae by disper- sion. However, it is important to mention than temperatures over 17°C and other environmental conditions may have a direct effect on mussel growth rates. Offshore scarcity of drifting mussels agrees with the fact that extension of postlarvae dispersion is more or less limited to the vicinity of mussel beds and high current areas (Newell et al. 1991, Caceres-Martinez et al. 1994, Caceres- Marti'nez and Figueras 1997, Caceres-Martinez and Figueras 1998). The presence of very large postlarvae ( 12.49 mm) supports the notion that the higher limit in the size of drifting postlarvae is around 10 mm (Beukema and Vlas 1989, Caceres-Martinez and Figueras 1997). ACKNOWLEDGMENTS The authors thank Oc. Sergio Guevara, from the Company Acuacultura Oceanica for allowing us to perform this study in their 38 December n2=478 n5=180 Ju 0> 60 CO » 40 E 20 **- O o 0) O) a) 100 u fe 8° EL March n2=41 n5=73 I. June n2=34 n5=24 XI September n2=90 n5=147 [L 0.26 0471 0.9 1.6 2.6 >3.6 0.47 0 899 1.699 2.699 3.699 Ramirez and Caceres-Marti'nez January n 2=1 581 nS=966 April n2=11 nS=2 nS=20 LL October n2=115 n5=91 0 26 0.471 0.9 1.6 2.6 >3 6 0.47 0.899 1.699 2 599 3.699 100 80 February n2=171 n6=123 May n2=1S nS=13 August n2=52 Dd November n2=74 n5=12 .i . 0.26 0.471 0.9 1.6 2.6 >3.6 0.47 0 899 1699 2.699 3.699 Size classes of mussels Figure 6. Size distribution in percentage of mussels settled on collectors SF placed at 2- (open barsl and 5-m (filled bars) depth, during a study period in Bahia de Todos Santos, Mexico. n2=913 n5=270 1L 0) 60 10 » 40 E 20 O 0 o a> ra *- c 0> o c 0) 0L March n2=128 n6=43 January n6=936 III. 100 April n2=14 80 60 - I 1 40 20 100 , July n2=73 80 60 40 20 P P r- >3.6 0.26 0.471 0.9 1.6 2.6 O.t 7 0.899 1.5 99 2.699 3.599 February n2=66 n6=37 May n2=28 n5= 1, Size classes of mussels Figure 7. Size distribution in percentage of mussels settled on collectors L placed at 2- (open barsl and 5-m (filled bars) depth, during a study period in Bahia de Todos Santos, Mexico. Rami'rez and Caceres-Marti'nez 39 facilities, also Raul Silva. Hilario Cardona Sepulveda, and Victor Molina Armenta from the same company for their help and logistic support during field samplings. We also thank Rebeca Vasquez- Yeomans for her assistance in processing samples and M. C. Ig- nacio Mendez for statistical advice. This work was supported by the CICESE Project 623106. LITERATURE CITED Aguirre. M. P. 1979. Biologia del mejillon (M. eduli.s) de cultivo de la Ria de Vigo. Bol. Insr. Esp. Oceanogr. 5:107-160. Bayne. B. L. 1965. Growth and delay of metamorphosis of the larvae of Mytilus edulis (L). Ophelia 2:1-47. Beukema, J. J. & J. de Vlas. 1989. Tidal-current transport of thread-drifting postlarval juveniles of the bivalve Macoma balthica from the Wadden Sea to the North Sea. Mar. Ecol. Prog. Sen 52:193-200. B0hle. B. 1971. Settlement of mussel larvae Mytilus edulis on suspended collectors in Norwegian waters, pp. 53-69. In: D. J. Crisp led.). Pro- ceedings of the 4th European Marine Biological Symposium. Bangor. Cambridge University Press, London. Caceres-Marti'nez, J. 1997. Mussel fishery and culture in Baja California. Mexico: history, present status, and future. NOAA Tech. Rep. NMFS 128.41-55 pp. Caceres-Martines, J., J. A. F. Robledo & A. Figueras. 1993. Settlement of mussel Mytilus galloprovincialis on an exposed rocky shore in Ria de Vigo, NW Spain. Mar. Ecol. Prog. Ser. 93:195-198. Caceres-Martines, J., J. A. F. Robledo & A. Figueras. 1994. Settlement and postlarvae behavior of Mytilus galloprovincialis: field and laboratory experiments. Mar. Ecol. Prog. Ser. 112:107-1 17. Caceres-Marti'nez, J. & A. Figueras. 1997. Mussel (Mytilus galloprovin- cialis Lamarck) settlement in the Ria de Vigo (NW Spain) during a tidal cycle. J. Shellfish Res. 16:83-85. Caceres-Martines, J. & A. Figueras. 1998. Mussel (Mytilus galloprovin- cialis Lamarck) colonization on artificial substrates in the Ria de Vigo of NW Spain. J. Shellfish Res. 17:153-157. Couturier, C. 1994. Spawning in sea scallops. Placopecten magellanicus. Canadian Tech. Rep. on Fisheries and Aquatic Sciences 1994: Pro- ceedings of the 9th international pectinid workshop. Nanaimo. British Columbia. Canada. April 22-27. 1993. pp. 138-146. Curiel-Ramirez. S. & J. Caceres-Marti'nez. Reproductive cycle of coexist- ing mussel species, Mytilus galloprovincialis. Mytilus californianus. and Septifer bifurcatus in Baja California. NW Mexico (in preparation). Dare, P. J.. D. B. Edwards & G. Davies. 1983. Experimental collection and handling of spat mussels (Mytilus edulis L.) on ropes for intertidal cultivation. MAFF (Lowestoft) Fisheries Research Tech. Rept. 74:1- 23. Davies. G. 1974. A method for monitoring the spatfall of mussel (Mytilus edulis L.). ./. Cons. Int. Explor. Met: 36:27-34. de Blok, J. W. & H. J. Geelen. 1958. The substratum required for the settling of mussels (Mytilus edulis L.). Arch Neerl. Zool. Vol. Juhilaire 13:446-460. de Blok. J. W. & Tan Maas. 1977. Funtion of byssus threads in young postlarva Mytilus. Nature 267:558. Devauchelle. N. & Ch. Mingant. 1991. The conditioning of scallop spawn- ers: practical aspects. Aquaculture and the environment. Special Publ. Eur. Aquacult. Soc 14.. p. 89. Eyster. L. S. & A. J. Pechenik. 1987. Attachment of Mytilus edulis L. larvae on algal and byssal filaments is enhanced by water agitation. J. Exp. Mar. Biol. Ecol. 114:99-110. Ferran, A. E. 1991. Ciclo gonadal y del tejido de reserva del mejillon de las Rias de Galicia Mytilus galloprovincialis Lmk. Doctoral thesis, Uni- versidad de Santiago de Compostela, Spain. 229 pp. Fuentes, J. & J. Molares. 1994. Settlement of the mussel Mytilus gallo- provincialis on collectors suspended from raft in the Ria de Arousa (NW Spain); annual pattern and spatial variability. Aquaculture 122: 55-62. Harger. J. R. 1972. Variation and relative "niche" size in the sea mussel Mytilus edulis in association with Mytilus californianus. Veliger 14: 275-283. Hines. H. A. 1979. Effects of a thermal discharge on reproductive cycle in Mytilus edulis and Mytilus californianus (Mollusca. Bivalvia). Fish. Bull. 77:499-503. King, P. A., D. McGrath & E. M. Gosling. 1989. Reproduction and settle- ment of Mytilus edulis on an exposed rocky shore in Galway Bay. West Coast of Ireland. / Mar. Biol Ass. U.K. 69:355-365. King, P. A., D. McGrath & W. Britton. 1990. The use of artificial sub- strates in monitoring mussel (Mytilus edulis L.). Settlement on an ex- posed rocky shore in the west of Ireland. J. Mar. Biol. Ass. U.K. 70:371-380. Koehn. R. K. 1991. The genetics and taxonomy of species in the genus Mytilus. Aquaculture 94:125-146. Loosanof, V.L., C. H. Davies & P. E. Chanely. 1966. Dimensions and shapes of larvae of some marine bivalve mollusks. Malacologia 25. McDonald. J. H. & R. K. Koehn. 1988. The mussels Mytilus galloprovin- cialis and Mytilus trossulus on the Pacific coast of North America. Mar. Biol. 99:111-118. Molares. J. & J. Fuentes. 1995. Recruitment of the mussel Mytilus gallo- provincialis on collectors situated on the intertidal zone in the Ria de Arousa (NW Spain). Aquaculture 138:131-137. Newell. C. R., H. Hidu, B.J. McAlice. P. Podniesinski, F. Short & L. Kindblom. 1991. Recruitment and commercial seed procurement of the blue mussel Mytilus edulis in Maine. J. World Aquae. Soc. 22:134- 152. Rees, C. B. 1950. The identification and classification of lamellibrach lar- vae. Hull. Bull. 19:73-104. Rees, C. B. 1954. Continuous plankton records: the distribution of lamel- libranch larvae in the North Sea, 1950-51. Bull. Mar. Ecol. 4:21-46. Robinson, A. 1992. Gonadal cycle of Crassostrea gigas Kumamoto (Thun- berg) in Yaquina Bay. Oregon and optimum conditions for broodstock oysters and larval culture. Aquaculture 106:89-97. Seed, R. 1976. Ecology, pp. 13-65. In: B. L. Bayne (ed.). Marine Mussels: Their Ecology and Physiology. Cambridge University Press, Cam- bridge, UK. Seed, R. & T. H. Suchanek. 1992. Population and community ecology of Mytilus. pp. 87-157. In: E. Gosling (ed.). The Mussel Mytilus: Ecol- ogy, Physiology, Genetics, and Culture. Elsevier. Amsterdam. Suchanek. T. H. 1981. The role of disturbance in the evolution of life history strategies in the intertidal mussel Mytilus edulis and Mytilus californianus. Oecologia (Berl) 50:143-152. Villalba, A. 1995. Gametogenic cycle of cultured mussel, Mytilus gallo- provincialis. in the bays of Galicia (NW Spain). Aquaculture 130:269- 277. Widdows. J. 1991. Physiological ecology of mussel larvae. Aquaculture 94:147-163. Young, R. T. 1946. Spawning and setting season of the mussel, Mytilus californianus. Ecology 27:354-363. Journal of Shellfish Research, Vol. IS. No. 1, 41-46. 1999. INDUCTION OF SETTLEMENT AND METAMORPHOSIS OF THE SCALLOP ARGOPECTEN PURPURATUS LAMARCK BY EXCESS K+ AND EPINEPHRINE: ENERGETIC COSTS G. MARTINEZ,1 C. AGUILERA,1 AND E. O. CAMPOS2 1 Facultad de Ciencias del Mar, Universidad Catolica del Norte, Coquimbo, Chile Vnidad de Neurobiologia Molecular, Facultad de Ciencias Biologicas, Pontificia Universidad Catolica de Chile, Coquimbo, Chile ABSTRACT Settlement and metamorphosis of marine invertebrate larvae is known to be triggered by specific environmental cues. Neuroactive compounds, particularly some monoamines, have been implicated in this process, and depolarization of receptor cell membranes has been suggested to occur as a response to them. An increase of extracellular K+ in seawater has been used as an effective inducer of these processes for some species. This study describes work designed to assay effects of epinephrine and excess K* as inducers of settlement and metamorphosis of larvae of the scallop Argopecten purpuratus. Epinephrine and excess K* increased the percentages of settlement, metamorphosis, and survival of these larvae. Responses were dose-dependent, with a maxima under 10"5 M (epinephrine) and 10 mM (K+). In the case of epinephrine, the responses did not vary significantly with the time of exposure. An analysis of size and energy content of larvae induced to metamorphosis by the different methods showed that larvae induced with epinephrine produced postlarvae that were significantly smaller in size and energetically weaker than postlarvae produced using excess K* or no added exogenous inducer. KEY WORDS: Argopecten purpuratus larvae, metamorphosis, scallops, settlement INTRODUCTION The scallop Argopecten purpuratus, as do many benthic marine invertebrates, produces pelagic larvae that spend days or weeks in the plankton before settling and metamorphosing into adult forms. Important physiological, morphological, and biochemical changes occur during the transition from pelagic to benthic existence. Settlement and metamorphosis processes are triggered by larval sensory recognition of, and responsiveness to, exogenous chemical and other environmental stimuli (Morse 1990). Various types of settlement-inducing cues have been described, including: (1) physical [e.g., illumination, physical texture (Hadfield and Pen- nington 1990), vibration (Rittschof et al. 1998)]; (2) biological [e.g., conspecific individuals, microbial films, prey species (re- viewed by Rodriguez et al. 1993)]; and (3) chemical [e.g., cues of natural or artificial origin (Yool et al. 1986)). Many of these chemical cues are compounds whose roles as neurotransmitters are broadly known (Rodriguez et al. 1993) although other chemically similar substances and some fatty acids have also been described to induce settlement and metamorphosis (Pawlik 1988, Kitamura et al. 1993). The ability of epinephrine to induce settlement and metamorphosis has been reported by Coon et al. ( 1985) (Crasso- strea gigas); Beiras and Widdows ( 1 995 ) ( C. gigas). Kingzett et al. (1990) (Patinopecten yessoensis), Tan and Wong (1995) (C. belcheri), Chevolot et al. 1991. and Nicolas et al. (1996) (Pecten maximus). It has been assumed that the inducers act on external cellular receptors somewhere on the larva (Hadfield and Pennington 1990). The rapidity and cascade of events in settlement and metamoiphic induction suggest the existence of a preformed larval nervous sys- tem capable of detecting a specific signal (Hadfield and Penning- ton 1990, Fenteany and Morse 1993). Although precompetent lar- vae may have receptors, these may be not sufficient to induce settlement behavior; attainment of competency may be determined by the accumulation of a threshold number of receptors (Barlow 1990). It is known that the response of receptor cells to an appropriate stimulus is the depolarization of specialized cells. On this basis, Baloun and Morse (1984) have suggested that the perception of inductive cues by larvae may rely on the stimulus-mediated depo- larization of cells in a sensory-inductive pathway. Given the knowledge that an increase of extracellular K+ induces membrane depolarization, several authors have successfully assayed the ef- fects of an excess of this ion in seawater as an inducer of larval settlement and metamorphosis (Baloun and Morse 1984. Yool et al. 1986; Inestrosa et al. 1993a, Campos et al. 1994: Pechenick et al. 1995). The molecular events mediating the inducer-receptor interaction and the hypothesized membrane depolarization are not understood, but it has been suggested that a possible second mes- senger such as cAMP and IP, may participate (Leitz and Muller 1987, Morse 1990; Baxter and Morse 1992; Inestrosa et al. 1993b. Clare et al. 1995). These messengers might regulate ion channel activity (Wickman and Clapham 1995). Besides regulating mem- brane permeability to ions, these messengers are known to increase the activity of several catabolic enzymes through phosphorylation (Krebs 1985), a fact that must be considered when a compound is chosen for experimental induction of metamorphosis. Metamorphosis is an energetically costly process (Holland and Spencer 1973, Lucas et al. 1979). Storage reserves of energy-rich organic substrates are usually observed to increase before meta- morphosis (Lucas et al. 1986). Studies to this point have focused on variations in the composition of energy reserves and measure- ments of oxygen consumption during this process (Holland and Spencer 1973, Lucas et al. 1979, Rodriguez et al. 1990. Shilling et al. 1996). The present study evaluated the effects of excess K+ and epi- nephrine as inducers of settlement and metamorphosis in the pec- tinid Argopecten purpuratus. Because epinephrine is a powerful catabolic stimulant, it was important to evaluate the comparative energy costs between individuals stimulated to metamorphosis by epinephrine and those stimulated by excess K+ (plus nontreated 41 42 Martinez et al. controls). A null hypothesis was tested that there was no difference in energy cost to the larvae or postlarvae between treatments. The practical importance of this work is that there has been interest on the part of commercial scallop hatchery managers in obtaining reagents for the massive induction of metamorphosis in cultured scallop larvae destined for aquaculture growout. If a safe and reliable chemical inducer were available, it would significantly reduce production and labor costs and add reliability to this newly evolving industry. MATERIAL AND METHODS Assay of Larval Settlement, Metamorphosis, and Survival Larvae of A. purpuratus were obtained from a commercial scallop hatchery in Tongoy Bay, Chile (30°S). Larvae were mass cultured using methodology similar to that of DiSalvo et al. (1984) and transferred to our laboratory for experimentation at a stage when they were entering competence for metamorphosis (length near 200 u,m, with presence of eyespot and foot). Larvae were maintained in 0.45 p.m filtered seawater at 20°C and fed daily with mixture of the microalgae Isochrysis galbana (T-iso), Pavlova lutheri, Chaetoceros calcitrans, and Chaetoceros gracilis. Experimental configuration included the use of six 1-L plastic containers, each fitted with rippled plastic plates (ca. 150 cm2) to increase the surface area for larval settlement. Each replica con- tained 800 mL of test water containing 1,500-1,600 larvae. Test solutions included 0, 5, 10, 20, and 30 mM K+ ion. above the normal concentration in local seawater (9 mM); epinephrine con- centrations were 10", 10", and 10^* M, produced by dilution of a stock solution of the amine diluted in 0.0005 N HCI. Larvae were exposed to test solutions for 48 hours, after which, the free larvae were suspended into fresh, filtered seawater; settled and metamor- phosed larvae were quantified in three of the replicates. The ex- periment was then continued for 96 hours with the remaining replicates, with daily changes of filtered seawater and quantifica- tion of settled and metamorphosed larvae at required time inter- vals. Unattached, living larvae were also included in counts to account for all surviving individuals. Larvae were considered settled when swimming ceased, they had settled on container and plate surfaces and could not be detached by gentle washing with fresh seawater. Larvae were considered metamorphosed if the ve- lum had been resorbed and if they showed development of the ctenidium and deposition of dissoconch. Effect of Exposure Time A second set of experiments was carried out to determine how the time of exposure to excess K+ or epinephrine could affect the response to these signals. Using the same experimental configu- ration described above, and the optimal inducer concentration then obtained, groups of competent larvae were exposed for periods of up to 96 h (excess K+) or 48 h (epinephrine). At each time interval (see Fig. 3), larvae were rinsed and placed in fresh, filtered sea- water. After 144 h. metamorphosed, attached, and unattached or- ganisms that remained alive were counted to calculate final per- centages of survival and metamorphosis. Comparative Energetic Cost Between Inducers Energy depletion during metamorphosis induced by the two different treatments was determined by direct calorimetry of the test organisms. This measurement was carried out in triplicate for each treatment. For each replicate. 80.000 larvae were placed with filtered seawater in 10-L plastic containers, where they were ex- posed to either 10 mM K+ or to 10" M epinephrine over a period of 24 h. After this, the medium was changed to filtered seawater and maintained with daily changes of water for 6 days, at which time most larvae seemed to have passed metamorphosis. Postlar- vae were then removed from the containers with a small paint- brush, and collected on 250-|xm mesh plastic screen. These organ- isms were washed with isotonic ammonium formate, dried to con- stant weight, made into pellets, and ignited in an OSK calorimeter. Sub samples were used to count the postlarvae and measure their lengths. Results were analyzed using one-way analysis of variance (ANOVA) and a Tukey test was applied to evaluate probable differences (p < .05) between treatments. In the case of percentage values, these were subjected to an arc-sin transformation for cal- culations. RESULTS Effects of A* Concentration After 48-h exposure to increased external K+, the percentage of larvae settled with 10 mM was statistically higher (Tukey, p < .001) than that of larvae incubated under normal K+ (Fig. l.A); however, very few of the larvae had passed metamorphosis (Fig. LB), with no statistical difference between percentages of the different experimental groups at this time. At 144 h (96 h after removal of larvae from inductor solutions), almost half of the larvae from the 10 mM treatment had passed metamorphosis, with values for larvae from other concentrations either near or below those obtained with no treatment (Fig. 1 .B) The survival of larvae submitted to 10 mM excess K* was significantly higher than that of the other experimental groups (Fig. l.C). Effects of Epinephrine After 48 h in different concentrations of epinephrine, it was shown that the highest percentages of larval settlement were ob- tained with 10" and 10" M (Fig. 2. A). Although few larvae had metamorphosed after 48 h with epi- nephrine, significantly higher values were observed at 10" and 10" M (Fig. 2.B). On the following days, the number of larvae that metamorphosed increased, and at 144 h (48 h after removal of larvae from the inductor solutions), nearly all the settled larvae had passed metamorphosis, with the most effective concentrations hav- ing been 10" and 10" M (Fig. 2.B). Survival of larvae submitted to 10" and 10" M treatments was significantly higher than that of the other experimental groups (Fig. 2.C). Effect of Exposure Time When 10 mM excess K+ was used to induce metamorphosis, the percentage of larvae that metamorphosed increased as the time of exposure increased, from 6 to 48 hours. No greater increase was detected when larvae were under this treatment for 96 hours (Fig. 3. A). In the case of larvae exposed to 10" M epinephrine, from 6 to 48 hours, the time of exposure to this amine did not affect either the percentage of mortality or that of metamorphosis of larvae (Fig. 3.B). Argopecten Purpuratus Metamorphosis by K+ and Epinephrine 43 60 SO , . c p 40 d) *s CO 30 c 8 20 in a. 10 0 60 r"' CO SO CD a. (A 1 x: e- £ c 8 60 SO 40 30 20 10 0 60 50 40 -I, I- 90 c • m I > 70 y^^. e / X. S ^r ^< at t v^ ^S. c 50 — "^-^"^ x ^S. 8 """"""""^T ^T £ 40 3 CO 30 c 8 20 2L 10 0 10"6 10* 10-4 epinephrine concentration (M) ---• 48h 144 h Figure 2. Percentages of A. purpuratus larvae that settled (A), meta- morphosed (B), and survived (C) in different concentrations of epi- nephrine. Values represent the means ± SE of groups of larvae from three replicate treatments; (*) significantly different from the other values (Tukey's test, p < .05). Energetic Analysis of iMrvae Induced to Metamorphosis As it is seen in the Table 1, the energy content of A. purpuratus post larvae decreased to less than half the values they had previous to metamorphosis. In the case of larvae induced to metamorphosis by excess K+. the decrease in energy content was similar to that of the control group; when epinephrine was used as an inducer, the decrease was considerably greater. Although postlarvae showed considerably increased size over the larvae, no significant differ- ence was found between the groups of larvae induced to metamor- phosis by any of the three different treatments. DISCUSSION Raised levels of the K+ ion have not previously been demon- strated as an inducer of metamorphosis in A. purpuratus larvae. The present results now add this species to a group of other mol- lusks that have been shown to be induced to settle and metamor- phose by this ion. An early study was that of Baloun and Morse (19S4). who showed this effect on larvae of Haliotis rufescens. They showed that the larval response to 7-aminobutyric acid (GABA) may be inhibited by a decrease in external K+. Yool et al. (1986) showed that an increase in the concentration of this ion induced settlement and metamorphosis in larvae of the mollusks 44 Martinez et al. 80 c 0) u 40 di Q. 20 A r" " c C ..-I b y /■-[■■ t i a _ 12 24 48 hours with excess K 96 80 « 60 c 8 V a 40 I-- I- 20 6 12 24 48 hours with epinephrine — ' — metamorphosis mortality Figure 3. Effect of exposure time to 10 mm excess K+ (A) and to 10"5 M epinephrine (B) on metamorphosis and survival of competent larvae of A. purpuratus. Values represent means ± SE of groups of larvae from three replicate treatments. Means with same letter are not sig- nificantly different (Tukey's test, p < .05). Phestilla sibogae, Haliotis rufescens, and Astraea undosa and in larvae of the marine annelid Phragmatopoma califomica. Positive results were also shown for the mollusk Concholepas concholepas by Inestrosa et al. (1993a). In each case (as now with A. Purpu- ratus), the effect of potassium has been dose dependent, although optimal inductive dose may vary among species. Present results also show that, as time of exposure to excess K+ increases, up to 48 hours, the percentage of A. Purpuratus larvae that metamorphose also increases. In this respect, our results agree with those of Baloun and Morse (1984), Yool et al. (1986), and Inestrosa et al. ( 1993a). These authors showed that, not only dif- ferent doses of excess K+, but also cumulative time of exposure, produced a progressive increase in larval metamorphosis in the species cited above. The present results have shown that 10"6 and 10"5 M epineph- rine were the optimal concentrations for significantly increasing the percentage settlement and metamorphosis of A. purpuratus larvae; a higher dose (10~4 M) did not show any significant effect. TABLE 1. Size and energy content of Argopecten purpuratus larvae before and after metamorphosis induced by 10"' M epinephrine or by 10 mM excess potassium. Energy Content (mJoule/Larva) Larvae Length (urn) Before metamorphosis Competent larvae After metamorphosis Control Induced by epinephrine Induced by excess K+ 4.0113 ±0.2665 1.6983 ±0.0516 0.5168 ±0.0368* 1.6477 ±0.108 216.47 ± 10.57 406.00 ± 33.68 364.17 ±26.51 399.55 ± 25.69 Values are mean ± SD (n = 3). * Significantly different from control (p < 0.001 ). These values do not agree with those described for other bivalve species, which are also induced by this monoamine. Kingzett et al. (1990) assayed three concentrations of epinephrine OCT6, 10_s, and 1CT4 M) on larvae of the scallop Patinopecten yessoensis and showed similar increases in percentage metamorphosis with each dose. In the case of the scallop Pecten maximus, epinephrine ex- hibited an optimal action between 0.5 to 2.5 x 10~4 M (Chevolot et al. 1991, Nicolas et al. 1996). Coon et al. 1985. Coon et al. 1986. and Beiras and Widdows (1995) have described maximum induc- tive activity in the oysters Crassostrea virginica and C. gigas with 10"4M epinephrine. Tan and Wong ( 1 995 ) described 10"5 M as the best concentration of epinephrine to induce metamorphosis in the oyster. C. belcheri. Larvae of the polychaete Phragmatopoma lapi- dosa californica did not respond to epinephrine assayed at any subtoxic concentration. Thus, sensitivity to inducer compounds may vary from one species to another. Moreover, in the assays described for oysters, most of the metamorphosed individuals in- duced by epinephrine were "unattached" spat (Coon et al. 1985, Coon et al. 1986. Beiras and Widdows 1995). a fact that was not observed for pectinids. Pawlik (1990) has suggested that different routes to metamorphic activation may be involved in the responses to different cues by the species. It is apparent in the present data that metamorphosis is more rapidly triggered by the monoamine than by excess K+. When epinephrine was used as an inducer cue on A. purpuratus, a sig- nificant number of individuals metamorphosed within the first 48 hours. On the contrary, in that period, no larvae had metamor- phosed using excess K+. Pechenik and Gee ( 1993) and Pechenik et al. (1995) have shown that larvae of the gastropods Crepidula fornicata and Phestilla sibogae become responsive to excess K+ at a different time than to "natural cues." suggesting possible differ- ent sites of action of the inducers. In the case of P. sibogae, larvae become responsive first to the natural cue before they do to excess K+; the authors suggested that this ion acted internally rather than directly on surface receptors, showing increasing accessibility to those internal sites as larvae age. The assays of exposure time of A. purpuratus larvae to the different cues, agree with the preceding concept. It was shown that, as the time with excess K+ increases, more larvae pass metamorphosis. Meanwhile, using epinephrine, the percentage of metamorphosis did not increase with exposure time and even was less than the maximum obtained using ex- cess K+. The use of exogenous inducers for metamorphosis has pres- Argopecten Purpuratus Metamorphosis by K+ and Epinephrine 45 enlly been shown to increase the survival of A. purpuratus indi- viduals after metamorphosis. This may be of practical consequence for use in commercial production of hatchery "seed." However, in the case of individuals that had been induced to metamorphose by epinephrine, the energy content of postlarvae was significantly lower than that of individuals that had been treated with excess K+ or had metamorphosed without any added inducer. Depolarization of externally accessible, excitable cells has been suggested to be a mechanism common to the induction of settle- ment and metamorphosis for a number of species (Baloun and Morse 1984. Yool et al. 1986). It is known that the mechanism of action of neurotransmitters, many of which have been reported to induce settlement and metamorphosis of larvae (Rodriguez et al. 1993), is through changing the membrane permeability, and then (Wickman and Clapham 1995) changing the degree of polarization of the cell. The mechanism by which some of these neurotrans- mitters change this permeability is unclear. These compounds may, through G proteins, increase the concentration of any second messenger (cAMP, IP,, Ca+2). triggering the phosphorylation of some key proteins (Krebs 1985. Ho 1994). Some of these proteins may be intrinsic to membranes playing a part or a total role as an ion channel, and therefore, its phosphorylation may be changing the corresponding ion conductance of the membrane (Wickman and Clapham 1995). In addition to membrane proteins, there are some enzymes that are phosphorylated, changing their activity ( Krebs 1985). Many of the enzymes whose activity is enhanced by phosphorylation are catabolic (e.g., glycogen phosphorylase, li- pases). Such a cascade of events, could explain how, when neu- rotransmitters are used to induce the metamorphosis of inverte- brate larvae, the normally high energy cost of the process may be further incremented, perhaps to the point of stress. Existence of a common mechanism for the inductive effects o\ potassium and neurotransmitters action is not probable. The effects of K+ may be explained in terms of their ability to change mem- brane potential. It is well known that potassium gradients through cell membrane determine the degree of cellular polarization. This type of effect should not mobilize energy reserves of individuals over the degree expected in normal (noninduced) metamorphosis. In conclusion, we reject the hypothesis that there was no difference between treatments, having observed significantly greater energy depletion in organisms treated with epinephrine. In light of possible anomalous effects of membrane depolar- ization (K+). or possible energy stress associated with neurotrans- mitters, it is not possible at this time to recommend the practical use of chemical inducers in scallop hatcheries, because the simple measurement of percentage of metamorphosis is not indicative of several other factors that may affect viability of the larvae. ACKNOWLEDGMENTS This research was supported by FONDECYT (project # 1960058) and by a Program "FONDAP de Oceanografia y Biolo- gfa Marina." We are grateful to Alejandro Abarca and to Pesquera San Jose Scallop Hatchery, from Tongoy, for their help and supply of larvae. We thank Dr. Louis DiSalvo for his help with the En- glish lansuaee. LITERATURE CITED Baloun, A.J. & D. E. Morse. 1984. Ionic control of metamorphosis in larval Haliotis rufescens (Gastropoda). Biol. Bull. 167:124-138. Barlow. L. A. 1990. Electrophysiological and behavioral responses of lar- vae of the red ahalone {Haliotis rufescens) to settlement-inducing sub- stances. Bull. Mar. Sci. 46:537-554. Baxter. G. T. & D. E. Morse. 1992. Cilia from abalone larvae contain a receptor-dependent G protein transduction system similar to that in mammals. Biol. Bull. 183:147-154. Beiras, R & J. Widdows. 1995. Induction of metamorphosis in larvae of the oyster Crassostrea gigas using neuroactive compounds. Mar. Biol. 123:327-334. Campos, E. O.. A. Pinto, A. Bustos, S. R. Rodriguez & N. C. Inestrosa. 1994. Metamorphosis of laboratory-reared larvae of Concholepas con- cholepas (Mollusca; Gastropoda). Aquaculture 126:299-303. Chevolot. L„ J. C. Cochard & J. C. Yvin. 1991. Chemical induction of larval metamorphosis of Pecten maximus with a note on the nature of naturally occurring triggering substances. Mar. Ecol. Prog. Ser. 74:83- 89. Clare, A. S.. R. F. Thomas & D. Rittschof. 1995. Evidence for the involve- ment of cyclic AMP in the pheromonal modulation of barnacle settle- ment. J. Exp. Biol. 148:655-664. Coon. S. L.. D. B. Bonar & R. M. Weiner. 1985. Induction of settlement and metamorphosis of the Pacific oyster, Crassostrea gigas (Thunberg) by L-DOPA and catecholamines. J. Exp. Mar. Biol. Ecol. 94:21 1-221. Coon, S. L.. D. B. Bonar & R. M. Weiner. 1986. Chemical production of cultchless oyster spat using epinephrine and norepinephrine. Aquacul- ture 58:255-262. DiSalvo, L. H., E. Alarcon. E. Martinez & E. Uribe. 1984. Progress in mass culture of Chlamys (Argopecten) purpurata Lamarck (1819) with notes on its natural history. Rev. Chil. Hist. Nat. 57:35-45. Fenteany. G. & D. E. Morse. 1993. Specific inhibitors of protein synthesis do not block RNA synthesis or settlement in larvae of a marine gas- tropod mollusk (Haliotis rufescens). Biol. Bull 184:6-14. Hadfield. M. G & J. T. Pennington. 1990. Nature of the metamorphic signal and its internal transduction in larvae of the nudibranch Pheslilla sibogae. Bull. Mar. Sci. 46:455—164. Ho Ren-Jye. 1994. Transmembrane signaling and animal evolution. Zoolog. Studies 33:1-28. Holland, D. L. & B. E. Spencer. 1973. Biochemical changes in ted and starved oysters, Ostrea edulis (L.) during larval development, meta- morphosis and early spat growth. J. Mar. Biol. Assoc. U.K. 53:287- 298. Inestrosa, N. C. M. Gonzalez & E. O. Campos. 1993a. Metamorphosis of Concholepas concholepas (Bruguiere 1789) induced by excess potas- sium. J. Shell. Res. 12:337-341. Inestrosa, N. C, M. Gonzalez & E. O. Campos. 1993b. Molecular changes induced by metamorphosis in larvae of the prosobranch Concholepas concholepas Bruguiere (Mollusca: Gastropoda; Muricidae). J. Exp. Mar. Biol. Ecol. 168:205-215. Kingzett. B. C. N. Bourne & K. Leask. 1990. Induction of metamorphosis of the Japanese scallop Patinopecten yessoensis Jay. J. Shellfish Res. 9:119-124. Kitamura. H.. S. Kitahara & H. B. Koh. 1993. The induction of larval settlement and metamorphosis of two sea urchins. Pseudocentrolus depresus and Anlhocidaris crassispina, by free fatty acids extracted from the coralline red alga Corallina pilulifera. Mar. Biol. 115:387- 392. Krebs, G. E. 1985. The phosphorylation of proteins: a major mechanism for biological regulation. Biochem. Soc. Trans. 13:813-820. Leitz. T. & W. A. Midler. 1987. Evidence of the involvement of Pl- signaling and diacylglicerol second messengers in the initiation of 46 Martinez et al. metamorphosis in the hydroid Hydractima echinata Fleming. Dew Biol. 121:82-89. Lucas, M. I.. G. Walker, D. L. Holland & D. J. Crisp. 1979. An energy budget for the free-swimming and metamorphosing larvae of Balanus balanoides (Crustacea:Cirripedia). Mar. Biol. 55:221-229. Lucas, A.. L. Chebab-Chalabi & P. Beninger. 1986. Variation of relative organic matter in Mytilus edulis L. larvae and postlarvae. J. Exp. Mar. Biol. Ecol. 95:99-103. Morse, D. E. 1990. Recent progress in larval settlement and metamorpho- sis: closing the gaps between molecular biology and ecology. Bull. Mar. Sci. 46:465^83. Nicolas. L., R. Robert & L. Chevolot. 1996. Effect of epinephrine and seawater turbulence on the metamorphosis of the great scallop. Aqua- cull. Int. 4:293-297. Pawlik, J. R. 1988. Larval settlement and metamorphosis of sabellanid polychaetes, with special reference to Phragmatopoma lapidosa, a reef-building species, and Sabellaria floridensis. a nongregarious spe- cies. Bull. Mar. Sci. 43:41-60. Pawlik. J. R. 1990. Natural and artificial induction of metamorphosis of Phragmatopoma lapidosa californica (Polychaeta: Sabellariidae). with a critical look at the effects of bioactive compounds on marine inver- tebrate larvae. Bull. Mar. Sci. 46:512-536. Pechemk. J. A. & C. C. Gee. 1993. Onset of metamorphic competence in larvae of the gastropod Crepidula fomicata (L), judged by a natural and an artificial cue. J. Exp. Mar. Biol. Ecol. 167:59-72. Pechemk. J. A.. M. G. Hadfield & L. S. Eyster. 1995. Assessing whether larvae of the opisthobranch gastropod Phestilla sibogae Bergh become responsive to three chemical cues at the same age. /. Exp. Mar. Biol. Ecol. 191:1-17. Rittschof, D„ R. B. Forward, Jr. G. Cannon. J. M. Welch. M. McClary. Jr. E. R. Holm. A. S. Clare. S. Conova. L. M. McKelvey, P. Bryan & C. L. Van Dover. 1998. Cues and content: larval responses to physical and chemical cues. Biofoulirtg 12:31-44. Rodriguez, J. L., F. J. Sedano, L. O. Garcia-Martin. A. Perez-Camacho & J. L. Sanchez. 1990. Energy metabolism of newly settled Ostrea edulis spat during metamorphosis. Mar. Biol. 106:109-1 1 1. Rodriguez, S. R.. F. P. Ojeda & N. C. Inestrosa. 1993. Settlement of benthic marine invertebrates. Mar. Ecol. Prog. Ser. 97:193-207. Shilling, F.M., O.Hoegh-Guldberg & D. T. Manahan. 1996. Sources of energy for increased metabolic demand during metamorphosis of the abalone Haliotis rufescens (Mollusca). Biol. Bull. 191:402-412. Tan. S. H. & Wong, T. M. 1995. Induction of settlement and metamor- phosis in the tropical oyster. Crassostrea belcheri (Sowerby). by neu- roactive compounds. J. Shellfish Res. 14:435^138. Wickman. K & D. E. Clapham. 1995. Ion channel regulation by G proteins. Physiolog. Rev. 75:865-885. Yool, A. J., S. M. Grau, M. G. Hadfield, R. A. Jensen. D. A. Markell & D. E. Morse. 1986. Excess potassium induces larval metamorphosis in four marine invertebrate species. Biol. Bull. 170:255-266. Journal of Shellfish Research. Vol. 18, No. 1, 47-58, 1999. EVIDENCE FOR FALL SPAWNING OF NORTHERN BAY SCALLOPS ARGOPECTEN IRRADIANS IRRADIANS (LAMARCK 1819) IN NEW YORK STEPHEN T. TETTELBACH,1 CHRISTOPHER F. SMITH,2 ROXANNA SMOLOWITZ/ KIM TETRAULT,2 AND SANDRA DUMAIS2 ^Natural Science Division Southampton College Long Island University Southampton, New York 11968 'Marine Program Cornell Cooperative Extension Riverhead, New York 11901 Marine Biological Laboratory Woods Hole, Massachusetts 02543 ABSTRACT Spawning of Argopecten irradians irradians is generally believed to occur between late May to August; however, some literature reports and anecdotal observations have suggested that npe individuals may be present well into the fall. This paper reports on evidence for fall spawning of bay scallops that we sampled from different populations in Long Island. New York, waters, in different years. At two sites, a spawning peak in September followed a discrete spawning peak in early summer (late June/early July). Scallops at one or more of four different sites were conclusively shown to spawn well into the fall (late September, October, or early November) during 3 different years: one in which a brown tide (Aureococcus anophagefferens) algal bloom occurred (1995) and during nonbrown tide years (1993, 1994). Our work, coupled with reports of other researchers, suggests that fall spawning of A. i. irradians in NW Atlantic waters does not seem to be uncommon and may be important in some populations during particular years. KEY WORDS: Argopecten, spawning. New York, reproduction, scallop, seasonality, brown tide INTRODUCTION Much of the basic biological knowledge of the northern bay scallop, Argopecten irradians irradians (Lamarck 1819), is based on the early work of Risser (1901) and Belding (1910). These authors determined that the typical lifespan of this subspecies is 18-22 months; however, some individuals may reach the age of 3 years (Marshall 1960. S. Tettelbach, pers. obs.). Bay scallops are hermaphroditic and generally are regarded as semelparous, al- though Belding ( 1910) estimated that 10-20% of a given year class may survive to spawn in 2 successive years if not removed by the fishery. Bricelj et al. (1987a) reported that perhaps 30% of caged scallops at one site in Long Island, NY survived to spawn during their second year. Spawning of Argopecten irradians irradians over its natural geographic range is generally believed to occur between late May to midAugust (see Table 1 ), as water temperatures are rising (Sas- try 1963). Spawning of northern bay scallops has been described by some authors as a single, catastrophic event; whereas, others report spawning over much of the summer (see Table 1), some- times with a distinct secondary or tertiary peak later in the period (MacFarlane 1991 ). A few authors have reported early September spawning in Massachusetts (Kelley and Sisson 1981, Taylor and Capuzzo 1983, Hampson and Capuzzo 1984). Kelley and Sisson (1981) surmised, on the basis of scallop seed sizes, that spawning occurred after September in Nantucket Island. MA. MacFarlane (1991) concluded, on the basis of gross observation, that scallops in Orleans, MA spawned into October during 1980. The belief is still widely held, however, that spawning does not occur past early September; hence, sampling in many studies has been terminated prior to the fall. This paper reports on evidence for fall spawning of Argopecten irradians irradians that we obtained through the analysis of tem- poral patterns of bay scallop reproduction in different populations in Long Island. NY waters, during different years. This work was prompted by anecdotal observations made by baymen and our personal observations of visually ripe scallops in the Peconic Bay system during fall months between 1992 to 1996. Histological analyses on archived bay scallop gonadal samples were performed to determine if there was any concrete evidence of spawning be- tween September to November during 1993 to 1995. The 1995 data also provided the opportunity to determine whether bay scal- lops were able to spawn during a bloom of brown tide, Aureococ- cus anophagefferens (Sieburth et al. 1988). MATERIALS AND METHODS Analyses of reproductive activity primarily focused on three groups of scallops transplanted from natural populations in eastern Long Island, NY during 1994 and 1995. During 1994. scallops were dredged from Northwest Harbor (NWH) on 25 April and 1 May and free-planted directly on the bottom (density = 9.6/m") to the south of Red Cedar Point (RCP) in Flanders Bay (Fig. 1 ). In 1995, scallops were dredged from Sag Harbor and either deployed in rafts (density = 80.7/m2). on 2 May. in East Creek (EC). South Jamesport. NY or free-planted (density = 5/nr) on 4 May to the south of RCP (Fig. 1 ). These three groups of scallops were moni- tored as part of broader studies (Smith and Tettelbach 1996. Smith 47 4S Tettelbach et al. TABLE 1. Summary of previous studies assessing temporal patterns of reproduction in populations of northern bay scallops {Argopccten irradians irradians) in waters of the northeastern United States, in order of decreasing latitude. Peak Water Spawning Spawning Depth Temp. Means of Location Period Period! s) 16.5 <1910 GO. MO Belding 1910 Woods Hole, MA Mid-July to late Aug August 1965 Gl. H Sastry 1970 Narragansett Bay. Rl June to July Mid-June <1901 GO Risser 1901 Nantucket Hbr, MA Mid-June to early Sept Late June to early July 1980 GO, S. L. H Kelley & Sisson 1981 Niantic R.. CT Mid-June to late July Mid-June to July 19-28 1955 GO. S Marshall I960 Pocjuonock R.. CT Mid-June to August Mid-June to early July 0.3-0.8 17-24 1983,84 GO. S Tettelbach 1991 Lake Montauk. NY Early June to August Mid-June to early July 0.6-2.6 -17-23 1974 Gl. L Hickey 1977 Accabonac Hbr. NY 'late May to August Early June 0.6-2.3 -17-23 1974 GI. L Hickey 1977 Three Mile Hbr. NY Early June to July Mid-June 0.6-3.3 -17-23 1974 Gl, L Hickey 1977 Three Mile Hbr. NY Early June to Aug Early June 2 > 16-26 1984 Gl, GW Bncelj et al. 1987b Northwest Hbr. NY Early June to Aug Early to mid-June 2 £16-25 1984 GI. GW Bricelj et al. 1987b Northwest Hbr. NY Early June to Aug Early to late June 3.5 a 1 5-24 1984 GI. GW Bncelj et al. 1987b Sag Hbr. NY Late May to Aug Late May to mid-June 3.5 > 14.5-27 1984 GI. GW Bricelj et al. 1987b Flax Pond, NY Late June to mid-July Late June to mid-July 2 19-23 1985 GI, H Eppetal. 1988; Epp 1989 All analyses are based on samples taken from natural populations, except for Flax Pond. At the latter site, scallops were held in cages on the bottom, after being dredged from Northwest Harbor, New York. Means of Assessment: GO = gross observation; MO = microscopic observation; GW = gonad weight; GI = gonad index; H = histology, with oocyte measurement (Oo); L = larval abundance; S = spat abundance (partially adapted from Barber and Blake 1991). and Tettelbach 1997) intended to evaluate different reseeding tech- niques (see Tettelbach and Wenczel). Temporal changes in reproductive condition were evaluated by means of analyses of gonad dry weights (GDW) and gonad in- dexes (Gil (Barber and Blake 1991). as well as by histological examination (see below). For each individual, shell height was measured to the nearest mm, and then the gonad was dissected proximal to the foot (so the foot remained attached to the excised gonad). For the GDW and GI analyses, the gonad and remaining tissues were weighed separately for each individual after they were dried to a constant weight (>48 h) at ~82°C. A GI value was calculated for each individual, as follows: GI = Gonad Wt (g) x 100/[Gonad Wt + Remaining Tissue Wt (g)] (Barber and Blake 1991 1. Temporal changes in mean GDW and GI values were ex- amined to determine if spawning had occurred, as inferred from a significant drop in GDW or GI. At the times when scallops were LONG ISLAND SOUND Figure 1. Map of the Peconic Bay system in eastern Long Island, New York, USA, showing the location of bay scallop and Aureococcus sampling sites. Fall Spawning of Argopecten in New York 49 East Creek -1995 B 100°/c 57.0 57.3 53.8 (3.6) (3.2) (3.5) 57.9 51.4 55.2 (2.9) (3.3) (33) 53.2 67.3 70.8 (29) (4.6) (2.7) June 9 June 21 July 7 July 20 Aug 4 Aug 18 Sept 5 Oct 3 Nov 6 es Recovering |H Early Maturation m Mid-maturation ■ Ripe ■ Early-spawn g Mid-spawn 3 Late-spawn Very late-spawn Spent Figure 2. Temporal change in reproductive condition of bay scallops (male. A, and female. B, gonadal portions) deployed in rafts in East Creek, South Jamesport, New York during early May 1995. Percentage of individuals in a given gonadal stage were determined by histological analysis (n = 1 1-12 scallops per date) and scored using a modification of Naidu's ( 1970) method: stage 4 = recovering from spawning: stage 5 = early maturation of eggs and sperm; stage 6 = midmaturation of eggs and sperm: stage 7 = mature eggs and sperm (ready to spawn); stage 8 = spawning, where: 8e = early-spawn, (<25% of eggs or sperm were spawned); 8m = midspawn. (25-85% of eggs or sperm were spawned); 81 = late spawn, (85-95% of eggs or sperm were spawned); 8vl = very late spawn, (between 96-100% of eggs or sperm were spawned); stage 9 = spent (immediately postspawn). Bay scallop shell height data [mean, (SD)] are given at the top of the figure. transplanted from Northwest Harbor to RCP (29 April 1994) and from Sag Harbor to EC and RCP (2 May 1995). GI values (13.59 and 15.25. respectively) were marginally different (t = 2.14. p = .04, n = 20 and 25. respectively, for 1994 and 1995). During 1995, GDW and GI data (n = 23-25 scallops per sample) were collected biweekly between late May to midAugust (RCP) or early September (EC); histological sampling (n = 10-15 scallops per sample) started in early June, then followed the same schedule. Additional histological samples were collected monthly from early September through early October (at RCP) or Novem- ber (at EC). (On 21 July 1995, only, it was also necessary to include in the histological analysis scallops that were transplanted to RCP on 25 May.) During 1994. initial sampling of GDW and GI was done in late April; then GDW, GI, and histology samples were taken biweekly from around late May through late September. GDW/GI sample size was 13-20 scallops, except on 25 May 1994 (n = 8). For the 1994 histological analyses, sample size ranged from two (25 May only) to nine scallops, but was usually six. Eight scallop gonads sampled in 1993 from natural populations in NWH were also examined histologically to provide additional information on the possible occurrence of fall spawning. These samples were taken opportunistically (i.e.. tissues were only ar- chived from scallops that visually appeared to be very ripe (ovar- ian portion of gonad was bright orange with evident veins)) during October and November off Barcelona Neck and south of Alewife Creek, respectively (Fig. 1 ) and preserved in 70% ethanol. The methods employed for the fixation and preservation of scallop gonadal tissues for histological analyses (total n = 276) followed procedures described by Humason ( 1979). After fixation, gonadal tissues from each scallop were processed in paraffin. Six micron (6 p,m) sections were cut and stained with hematoxylin and eosin using standard histologic methods (Humason 1979). These cross sections (along the dorsoventral axis) were taken -1/3 of the way from the proximal and distal end of each gonad. The proximal and distal end sections contained predominantly male and female tubules, respectively. Gonadal developmental stages (see Figs. 2—1) were scored via 50 Tettelbach et al. Red Cedar Point - 1995 B 55.8 (4.1) 100% 52.9 (3.8) 51.3 53.9 455 47 6 47.8 60.3 (6.1) (6.1) (6.0) (5.1) (6.7) (3.6) JunelO June23 July5 July21 Aug3 Aug"! 7 Sept5 Oct3 Recovering Ripe Late-spawn Early Maturation | Early-spawn Very late-spawn Mid-maturation Mid-spawn Spent Figure 3. Temporal change in reproductive condition of bay scallops (male, A. and female, B. gonadal portions) free-planted to the south of Red Cedar Point in Flanders Bay, New York during early May 1995. Gonadal stages were determined by histological analysis In = 10-15 scallops per date) and scored using a modification of Naidu's (197(1) method: gonadal stages, scallop shell size data as given in Figure 2. traditional, subjective tissue evaluation methods performed hy a trained pathologist (RS), using a modification of Naidu's (1970) method. In the present study, stage 8 of Naidu (1970) was divided into four substages: 8e (early spawn) in which fewer than 25% of the eggs or sperm were spawned, 8m (midspawn) in which 25 to 85% of the eggs or sperm were spawned, 81 (late spawn) in which 85 to 95% of the eggs or sperm were spawned, and 8vl (very late spawn) in which between 96 to 100% of the eggs or sperm were spawned (see Fig. 5. A to D). The percentage of eggs or sperm spawned was subjectively determined by visual examination of the stained tissue section and was based on the amount of empty space within the mature tubules resulting from the loss of spawned ga- metes [in stage 7 ( = ripe), eggs and sperm are tightly packed within the gonadal tubules]. In cases where more than one stage was clearly evident in the male or female portion of the gonad from a given individual, partial designations were made (e.g.. 81/4). The degree of reproductive synchrony was assessed histologi- cally at two levels in this study: among scallops from the same sample group ( = interanimal) and within the same individuals ( = intra-animal). Synchronous interanimal maturation was identified when the male and female gonads of the animals examined varied little from the most common stage within that sample group (Fig. 2: 9 June female and male gonadal portions). Asynchronous inter- animal maturation was identified when numerous developmental stages were present in animals sampled at one given time (Fig. 2: 6 November female and male gonadal portions). Synchronous in- tra-animal maturation was most obvious in tubules of female go- nads that were in stages 6, 7, and 8. For example, normal synchro- nous maturation in stage 7 (Fig. 6A) was characterized by centrally located mature or submature eggs occupying >85% of the tubule lumen surrounded by lesser numbers of mid- to small-sized eggs lining the tubules (Naidu 1970). Asynchronous intra-animal matu- ration in stages 6. 7. and 8e female gonads was characterized by variable numbers of mature (70-85%) and submature eggs cen- trally located in tubules associated with increased numbers of im- mature eggs ranging from mid- to small sizes in outer layers of the tubules (Fig. 6B). Asynchronous intra-animal developmental maturation in male gonads was defined by alternating areas, within the same tubule or different tubules, of tufts at primarily later stages of sperm development (e.g., stage 8) interspersed with tufts at an earlier stage of development (e.g.. stage 6). Water temperature was monitored through the course of the field sampling during 1993 to 1995. A continuous temperature recorder (Ryan RTM2000). deployed at the bottom [-3 m at mean low water (MLW)|. was employed at the RCP site from May to Fall Spawning of Argopecten in New York Red Cedar Point - 1994 51 56.0 52.8 53.7 (1.4) (3.3) (1.2) 58.3 56.6 57.3 59.5 58.4 65.8 (3.2) (3.0) (2.1) (4.5) (2.8) (1.3) B R»™+^ri+jmmiL, ™ | I 1 I l H May 25 June 9 June 23 July 8 July 21 Aug 3 Aug 26 Sept 16 Sept 30 Recovering I Ripe Late-spawn | Early Maturation | | Early-spawn Very late-spawn Mid-maturation | Mid-spawn Spent Figure 4. Temporal change in reproductive condition of bay scallops (male, A. and female, B. gonadal portions) free-planted to the south of Red Cedar Point in Flanders Bay, New York during late April 1994. Gonadal stages were determined by histological analysis (n = 2 scallops on 25 May and 4-9 individuals on other dates) and scored using a modification of Naidu's ( 1970) method: gonadal stages, scallop shell size data as given in Figure 2. September 1994. In 1995, continuous temperature recorders (On- set Stowaway® Model #1405) were attached to midwater nets (depth = 1-2 m at MLW) at EC and RCP (May-September) and to one of the rafts (depth = 0.15 m at MLW) at EC (May- October). Readings were taken every 0.5 h. Continuous tempera- ture recorders were calibrated against a NITS standardized ther- mometer to within 0.1°C. Water temperature readings at other times and sites were taken in situ with a hand-held thermometer. Salinity samples were taken periodically during 1994 and 1995 and analyzed in the laboratory with a Beckman induction salinom- eter. Between 12 April and 6 October 1994. surface salinity ranged from 24.06-28.79 ppt at a site in central Flanders Bay [Sta. #170 of the Suffolk County Dept. of Health Services (SCDHS)] (see Fig. 1). -2200 m from our RCP site. During 1995. surface salinity at our EC site ranged from 28.21-29.99 ppt between 9 April and 5 September; surface salinity at our RCP site ranged from 27.20- 29.26 ppt between 10 June and 5 September. Cell concentrations of Aureococcus anophagefferens (Siehurth et al. 1988) were monitored throughout 1993 to 1995 by SCDHS. Sampling was done on a biweekly basis at East Creek (Sta. #101 ), s 100 m from our EC site, and weekly at Sta. #170 (see Fig. 1 ). For Flanders Bay, parameters of the Tetra Tech ( 1997) hydrodynamie model of the Peconic Bay system were used to calculate the av- erage transit time for tidal movement of surface waters from SCDHS Sta. #170 to our RCP site. At an averaged residual veloc- ity of 1.14 cm s_I, this transit time was calculated to be 69.1 h. (Tettelbach, unpubl. data). Using a maximum doubling time of 0.8 day-1 ( = every 30 h) for Aureococcus anophagefferens (based on laboratory studies by Cosper et al. 1989. Gobler 1995), cell con- centrations in Sta. #170 and our RCP site would be expected to vary by a factor of no more than 2.3. For 1993, biweekly Aureo- coccus cell counts were available for SCDHS Sta. #1 18. -1.5 km from our sampling sites in Northwest Harbor. RESULTS Histological analyses revealed that initial spawning of the RCP scallops in 1994 and of the RCP and EC scallops in 1995 occurred between late June and early July (Figs. 2^4); most scallops were spent (immediate postspawn) by the time of the late (20 and 21) July sampling dates. In 1994, the RCP free-plant group had just begun to spawn by 8 July, at which time 17 and 50% of the male and female gonads, respectively, were in early spawning condition (stage 8e). In 1995, by contrast, 100% of the RCP free-plants were 52 Tettelbach et al. wmM^ Figure 5. Photomicrographs of four spawning substages of female gonads from the bay scallop {Argopecten irradians irradians ) ( 25x ); A. early spawning (8e). only a few eggs are missing from the central lumens of the tubules: B. midspawning (8m), large numbers of eggs are missing from the tubules; C. late spawning (81), most eggs are missing from the tubular lumens; D. very late spawning (8vl), only rare eggs remain within the tubular lumens. past early spawning condition (stages 8m-9) on 5 July. Fifty-nine and 839}- of the male and female gonads, respectively, from EC raft scallops were in mid to late spawning condition (stages 8m-81) by 7 July 1995 (Figs. 2-4). Spawning continued after July in all three groups of scallops, but additional peaks of spawning activity occurred at different times. In both the 1994 RCP free-plants and the 1995 EC raft scallops, histological analyses revealed that gonadal maturation proceeded steadily from late July until mid to late August, and then a second spawning peak occurred between late August through late September to early October (Figs. 2. 4). In the 1995 RCP free- planted scallops, however, a second period of spawning was al- ready in progress by early August (Fig. 3). Spawning of these animals continued through early October, although a dramatic- spawning peak was not obvious in this group. The times of spawning suggested by the GI/GDW analyses, where they overlapped with histological analyses, agreed with the above results, except for the respective periods of spawning ini- tiation in the 1994 and 1995 RCP free-planted scallops. In the 1994 RCP group. GI/GDW trends (see Fig. 14) suggested that spawning commenced between late May and early June, rather than between late June to early July. In the 1995 RCP group, GI/GDW trends (see Fig. 13) showed a more gradual decline from peak values in late May. An analysis of variance (ANOVA) of differences in GI of RCP scallops between late May to early July 1995 (F = 104.65. 96 df, p < .0001 ) and subsequent Bonferroni multiple comparisons showed that there was no significant difference between GI values on 27 May and 10 June, but these GI values were both significantly greater than those on 23 June (p < .001). which were, in turn, significantly greater than those on 5 July (p < .001). These GI analyses suggest that spawning of the RCP free-planted scallops in 1995 commenced between 10 and 23 June rather than between 23 June and 5 July, as suggested by the histological analyses. Possible reasons for these apparent disagreements are discussed below. For the most part, inter- and intra-animal gonadal development stages were fairly comparable among the male and female portions of scallops sampled from a given site at a given time. Develop- mental stages of male and female gonadal portions from the same animal were almost always within one developmental stage of one another. The interanimal range of designated histological stages never exceeded two to three adjacent developmental stages for a given sample group at a given sampling period. Interanimal asynchrony was evident in the female and male gonads of scallops at EC and RCP following the end of the first major spawning in late July, as compared to the period between late May to early June. Interanimal asynchrony was more pro- nounced in 1995 than in 1994, particularly at RCP, where it was first noted in free-planted scallops sampled on 3 August 1995. In East Creek rafts, interanimal asynchrony (first noted on 7 July 1995) was not as severe. Interanimal asynchrony became common in both groups by August and September of 1995 (Figs. 2, 3), then began to decline. Interanimal asynchrony also was observed in the 1994 RCP samples, but was mild as compared to that seen in 1995 samples. Intra-animal asynchrony was also evident in female gonads sampled from the three groups described above. In 1995, it tended to become more apparent during midsummer to early fall (7 July-5 September) at both sites, but was most pronounced in the East Creek raft scallops. Intra-animal developmental asynchrony was noted in the 1994 samples during the second spawn, but was mild in comparison to that seen in both groups sampled in 1995. Spawn- ing of animals with marked intra-animal developmental asyn- chrony showed evidence of retention of several undeveloped eggs in the tubules at late (stage 81) spawning (Fig. 7). Interestingly, in Fall Spawning of Argopecten in New York 53 Figure 6. Photomicrograph of a bay scallop"s female gonadal tubules (50x). A. In this stage (7|. gonad tubules are filled with mature eggs. Rarely, eggs of smaller sizes are noted at the edges of the tubules. B. These female gonadal tubules are in stage 7/8e. but. in addition to the mature eggs, the tubules also contain many immature eggs (arrows). samples taken at both sites in September to October 1995. female gonads in stage 81 (late spawning condition) or 8vl (very late spawning condition) also often showed early proliferation of the germinal epithelium with small developing eggs (stage 4) (Fig. 8). Male gonadal tubules did not exhibit detectable intra-animal asynchrony. Male gonadal tubules from animals sampled in late spring rarely were observed to empty completely, as described by Naidu (1970). However, male gonads of scallops sampled in late summer, which were staged as 81 and 8vl. often showed active, mature, sperm-producing germinal epithelium that was sometimes only a few cells thick. Rarely, intra-animal developmental asyn- chrony was identified in these animals by the appearance of tufts of less differentiated spermatogenic cells interspersed with sper- matocytes and spermatids within the more mature (although greatly thinned) epithelium (Fig. 9). Postspawn or very late spawn intratubular invasion by hemocytes of the female and male gonadal tubules was rarely noted and was minimal when seen. Intratubular inflammation usu- ally consisted of a few hemocytes (usually between two and 20 cells) associated with degenerating retained eggs (Fig. 10). In cross sections of a single male gonad, approximately five closely asso- ciated tubules contained numerous hemocytes. resulting in signifi- cant inflammation that filled and slightly extended the walls of the tubules. The cause of inflammation in this male gonad was not apparent. Interestinglv. in one male and one female sonad from two different animals, rare foci consisting of tumorous proliferations formed cell mounds that projected into the tubular lumen from the germinal epithelium (Fig. 11). No tumorous cells were noted in- vading through the basement membranes. In both cases, the cells were undifferentiated. -10 pjn in diameter with a high nuclear/ cytoplasmic ratio and mitoses of 1/higher power field. Water temperatures recorded at the EC rafts were slightly higher than those recorded at the lantern nets at RCP; however, temporal trends were very similar in 1995 (see Figs. 12, 13). At RCP. peak temperatures for the 2 years of study were 29.4°C (on 3 August 1995) and 28.6°C (on 9 July 1994). More significantly, perhaps, there was a sharp drop in water temperature just before spawning seemed to commence during both years (from -26 to 20.5°C between 18-21 June 1995 and from -27 to 23°C between 19-22 June 1994) and then a more or less steady rise in tempera- ture after that (to ~27.4°C by 14 July 1995 and to ~28.6°C on 9 July 1994). Concentrations of Aureococcus anophagefferens were 2-105 cells mL"1 from 20 June through at least 18 July 1995 at East Creek and near Red Cedar Point (Figs. 12. 13). with respective recorded peaks at these two sites of 9 x 105 and 1.2 x 106 cells mL"1 on 3 July 1995 (SCDHS 1995). Thus, commencement of spawning at both sites coincided with the time of rising Aureo- coccus concentrations, before peak bloom conditions. The second spaw ning peak exhibited by EC raft scallops, which started in early September 1995. also coincided with rising Aureococcus concen- trations before the second brown tide peak (3.8 x 105 cells mL-1) on 12 September 1995. In 1994. Aureococcus concentrations at these sites did not exceed 1.5 x I03 cells mL-1 at any time (SCDHS 1994). At the time the last histological samples were taken in late September to early October or early November, all three of the above groups of scallops included individuals that were still in spawning condition (Figs. 2—1). This was not a rare phenom- enon, because about one-third of the EC raft scallops sampled on 6 November 1995 were still in spawning condition (either wholly or in part), while other individuals were ripe or in midmaturation (stage 6). Direct evidence of spawning was obtained on the after- noon of 3 October 1995, when East Creek raft scallops were seen to be spawning in situ. Water temperature on this day ranged from 17.0°C (at -0530 h) to I9.9°C (at -1300 h). Gametes collected from a few individuals yielded viable trocophore larvae by the next day. Additional evidence of fall spawning was provided by means of histological analyses of scallops that had been sampled from two different natural populations in NWH during 1993; GL/GDW data were not collected at that time. Of the six individuals sampled off Barcelona Neck on 14 October, all of the female gonadal portions and five of six male portions were in early to late spawning con- dition (stages 8e-81). The sixth male gonad was ripe (stage 7). One of two individuals sampled from south of Alewife Creek on 7 November was in midspawning condition (stage 8m). with the female gonadal portion also showing some sections that were re- covering from spawning (stage 4). The second individual was ripe (stage 7). but had not begun spawning. Only these eight individu- als had been selectively archived out of larger samples of scallops, because they visually appeared to be very ripe (ovarian portion of gonad was bright orange with evident veins). Thus, we can con- clude that the minimum proportions of scallops in the process of spawning at NWH and Barcelona Neck on 14 October and 7 November 1993 were 6/49 (12.2%) and 1/88 (1.1%). respectively. 54 Tettelbach et al. 9W W$Mit mm Figure 7. Photomicrographs of a bay scallop's female gonad in stage 8m/l sampled in the late summer. Most mature eggs have been spawned; however, numerous immature eggs of various sizes remain within the tubules (arrows) (25x) (5 September 1995. East Creek rafts). Figure 8. Photomicrographs of a bay scallop's female gonad in stage Svl. Although only few mature eggs remain in the tubules, early regeneration of the tubule epithelium has already begun (arrows) (25x) (3 October 1995, East Creek rafts). Figure 9. Photomicrograph of a bay scallop's male gonad in stage 8vl. (1) Foci of maturing spermatids alternate with (2) some foci of spermatocyte proliferation (25x) (7 July 1995, East Creek rafts). Figure 10. Photomicrograph of a bay scallop's female gonad in stage 81. Hemocytes surround and engulf portions of eggs remaining within the tubular lumens (1: hemocytes; 2: eggs) (lOOx) (7 July 1995, East Creek rafts). Figure 11. Photomicrographs of a bay scallop's female gonad. A. Nodules of tumor cells project into the tubular lumens ( lOOx). B. Mitotic figures are identified within the mass (arrow) (250x) (14 October 1993, Barcelona Neck). Water temperature on the latter date was 10.0°C. During 1993, Aureococcus coneentrations never exceeded 1.1 x 103 cells mL_1 in Northwest Harbor (SCDHS 1993). DISCUSSION Temporal patterns of bay scallop reproductive condition re- vealed by our histological analyses and GI/GDW monitoring agreed in many, but not all, instances in this study. Histological analyses may simply have missed the late May to early June 1994 spawning that was suggested by the Gl/GDW analyses of the RCP free-plants because of the small sample size (n = 2) on 25 May. However, we might then have expected that some evidence of postspawning recovery would show up in the next group of samples (n = 6) on 9 June 1994. It did not. Further work is needed to elucidate the apparent disagreement between the two methods: however, because histology is considered the most definitive way to assess reproductive condition in scallops (Barber and Blake 1991). we base the ensuing discussion on these results. Two clear spawning peaks (in late June to July and late August to September) were evident in the 1994 RCP free-plants and the 1995 EC raft scallops; whereas, spawning of the 1995 RCP free- plants showed one distinct peak (late June to early July) followed by a prolonged and less dramatic period of spawning. The latter group also showed the least synchronous pattern of reproductive development following the end of the first spawn in late July. The mean shell size of the 1995 RCP free-planted scallops was con- siderably smaller, several weeks after deployment, than the 1994 Fall Spawning of Arcopecten in New York 55 EAST CREEK -1995 RED CEDAR POINT -1995 J 1200- '■= 1000- u> 800 - O 600- • • g 400- 8 200- 5 i: 0- — ■«- • I ■ • (*- • • — I P-* — I — -*-* — 70 60 .s> X 150 V) 40 ' (|fHHn . Gonad Index Gonad Wt Total Body Wt Figure 12. Temporal changes in water quality parameters iAureococ- cus cell concentrations, water temperature), and size (shell height) and reproductive condition (gonad index, gonad dry weight, total body weight) of bay scallops deployed in floating rafts at East Creek, South Jamesport, New York in early May 1995. Initial values were obtained from a sample of scallops collected from Sag Harbor at the time when transplants were initiated. Datapoints for scallops are mean values ± 1 SD; n = 24-25 scallops per sampling date. RCP free-plants and the 1995 EC raft scallops (see Figs. 2-4, 12-14) (this may have been a sampling artifact, but the reason for the apparent decline in size of the surviving 1995 RCP scallops is unclear). Dry tissue weights are highly correlated with shell size in A. i. irradians (Epp et al. 1988). however, the exhibited magnitude of size differences for this group is not expected to have affected temporal spawning patterns: bay scallops of three groups ("large" natural, "small" natural, and hatchery-reared animals with respec- tive mean sizes of 53.4, 37.6, and 32.6 mm at the time of deploy- ment in pearl nets at the same site in Hallock Bay. NY during 60 50 40 Hi — i — i — >— ( — >- -•- — i — ^ 1 — i — i — i — i — i — -m- Gonad Index Gonad Wt Total Body Wt Figure li. Temporal changes in water quality parameters {Aureococ- cus cell concentrations, water temperature), and size (shell height) and reproductive condition (gonad index, gonad dry weight, total body weight) of bay scallops free-planted south of Red Cedar Point in Flanders Bay, New York in early May 1995. Initial values were ob- tained from a sample of scallops collected from Sag Harbor at the time when transplants were initiated. Datapoints for scallops are mean val- ues ± 1 SD; n = 23—25 scallops per sampling date. spring 1994) showed nearly identical temporal patterns of repro- duction as shown by changes in GI and GDW (Smith and Tettel- bach 1996). Several authors have suggested that when environmental con- ditions are less than favorable for synchronous spawning of scal- lops, "dribble" spawning may help to ensure that some larvae are able to survive (Langton et al. 1987, Paulet et al. 1988). The different patterns of inter- and intra-animal developmental asyn- chrony within scallop groups we observed in this study may sim- ply fall within the range of normal variability between different populations (see Bricelj et al. 1987b) and different years. However. 56 Tettelbach et al. RED CEDAR POINT -1994 -m- Gonad Index Gonad Wl Total Body Wt Figure 14. Temporal changes in water temperature, and size (shell height) and reproductive condition (gonad index, gonad dry weight, total body weight) of bay scallops free-planted south of Red Cedar Point in Flanders Bay, New York in late April 1994. Initial values were obtained from a sample of scallops collected from Northwest Harbor at the time when transplants were initiated. Datapoints are mean val- ues ± 1 SD; n = 8 scallops on 25 May and 13-2(1 individuals on other sampling dates. the differences in reproductive patterns exhibited by the 1995 RCP free-planted scallops warrant further examination of the possible effects of the brown tide (vs. nonbrown tide years) and depth in the water column (i.e.. surface rafts vs. on-bottom). Our finding that scallops at EC and RCP first spawned during the period of rising Aureococcus concentrations, before the peak of the brown tide bloom in 1995, is of particular interest because: ( 1 ) it demonstrates, for the first time, that bay scallops definitely spawned during a brown tide algal bloom (this seems to be coin- cidental rather than to have been a causative factor in spawning); and (2) it helps to answer the question posed by Bricelj et al. (1987b) and Bricelj and Kuenstner (1989) as to whether a temporal decline in scallop gonad weights or gonad indexes during a brown tide bloom represents actual spawning or gamete resorption. If gamete resorption had occurred, it is expected that extensive in- filtration of the gonadal tubules by hemocytes and the presence of numerous degenerative eggs within the tubules would have been apparent. However, only mild inflammation was noted in tubules and then only in association with a few degenerative eggs. This low level of inflammation was noted in scallops sampled during both 1994 and 1995 and thus probably represents a low level of atresia of unripened, but normal, eggs, as occurs in most other types of animals (Coe and Turner 1938, Van der Kraak et al. 1998). The occurrence of tumorous proliferations of cells of gonadal epithelial origin in gonads from animals that have actively pro- duced gametes for 3 to 4 months may parallel the phenomena of hormonally stimulated tumors as seen in other animals (Jubb et al. 1985) and at least partially result from repetitive gonadal cycling. Laboratory studies have shown that 3- to 10-day old bay scal- lop larvae experienced reduced growth and elevated mortality when exposed to Aureococcus concentrations of > 1.8 x 105 cells mL"1 (Gallager et al. 1989); thus, it seems unlikely that spawning of scallops during brown tide bloom conditions in 1995 resulted in successful recruitment of scallop larvae. This was borne out by the virtual absence of scallop seed in any part of the Peconic Bays during the fall of that year: the commercial harvest of adult (1+ v) bay scallops in 1996 was among the poorest in New York over the last 50 years. Histological analyses conducted in the present study conclu- sively demonstrated that spawning occurred at least through late September to early October in the 1994 and 1995 RCP free-plants and into November during 1993 and 1995 for the NWH and EC scallops, respectively. These are the latest spawning dates yet re- ported for Argopecten irradians irradians at this latitude. Based on the data we have presented here and prior reports from other locations (see Kelley and Sisson 1981, MacFarlane 1991) we believe that fall spawning of northern bay scallops is not an unusual phenomenon. Fall spawning probably has been missed in other studies, because reproductive sampling is usually termi- nated before the end of the summer, although Bricelj et al. (1987b) did conduct GI analyses through October 1984 in Northwest Har- bor, NY and found no evidence of a fall spawn. Research on other pectinid species has demonstrated the occurrence of "late" spawn- ing; for example, Placopecten magellanicus (Mac Donald and Thompson 1988) and Argopecten irradians concentricus (Bologna 1998). In the latter species, reproduction may occur throughout the year in St. Joseph Bay, FL (Bologna 1998). We do not yet know the full significance of fall spawning in Argopecten irradians irradians, especially given the reduced fer- tilization success and recruitment that may result when broodstock densities are low and/or when spawning individuals are separated by some critical distance (Levitan et al. 1992, Peterson and Sum- merson 1992). Virtually nothing is known about the latter phe- nomena in Argopecten irradians irradians. The problem of ensur- ing successful fertilization later in the fall is also likely to be exacerbated because of a reduced proportion of spawning indi- viduals in the population. Even if bay scallop spawning occurred throughout the Peconic Bay system after the brown tide subsided in 1995, the very low adult population size relative to 1994 and other years (based on reported commercial bay scallop landings by the NY State Dept. of Environmental Conservation) suggests that potential recruitment resulting from the 1995 fall spawn was prob- ably unimportant. Nevertheless, the occurrence of unusually small Argopecten irradians irradians seed at the end of some fall growing seasons or of adults with growth rings very close to the hinge (Kelley and Sisson 1981, MacFarlane 1991. Tettelbach et al. 1994) suggests that "late" spawning may be important to some bay scallop popu- lations in certain years. Tettelbach et al. (1994) found that during the winter of 1990 to 1991. following a nonbrown tide year, the Fall Spawning of Argopecten in New York 57 percentage of "small" (S20 mm) seed ranged from 0-9% in eight different bay scallop populations in eastern Long Island. In 1992, 1 year after a brown tide bloom, 100% (n = 268) of the adult scallops sampled at one of the same sites (south of Alewife Creek in NWH) had growth rings that were 2-7 mm from the hinge, indicating that the adults had only reached this size, as seed, at the end of their first growing season. These small seed may have resulted from late spawning, an extended larval period and/or slow growth following larval settlement, but our present findings lend further credence to the possibility that these small seed resulted from a fall spawn. In that case, fall spawning clearly may be essential to the persistence of certain populations during some years. Larvae of A. i. irradians have been shown to exhibit >90% survival, 8 days after fertilization in the laboratory, at a tempera- ture of 10°C and salinity of 25-30 ppt (Tettelbach and Rhodes 1981 ); thus, larval survival is probable well into November in local waters. However, although larval growth also occurred in these laboratory conditions, it was an order of magnitude slower than at 25°C (Tettelbach and Rhodes 1981 ). Our histological results suggest that eggs were not routinely resorbed during the spring to fall months when samples were taken, but continued to mature until they were spawned. The en- vironmental stimuli that induce spawning of Argopecten irradians irradians in the fall, however, are presently unknown. Barber and Blake (1991) have suggested that there does not seem to be any critical temperature, per se, at which scallop reproduction occurs and. furthermore, that a rapid temperature change (AT) is probably a more important spawning trigger than an absolute temperature or the direction of chan«e. Direct disturbance of the raft during the process of sampling on 3 October 1995 may possibly have trig- gered spawning, but the fact that water temperature at the EC rafts did spike to over 22°C at the end of September 1 995 and rose from 17 to 19.9'C on 3 October, just before the time when spawning was observed in situ in East Creek rafts suggests that these con- ditions may have provided the appropriate stimuli for spawning. Further work is necessary to elucidate the mechanisms that trigger fall spawning of northern bay scallops and the importance of this phenomenon in bay scallop populations. ACKNOWLEDGMENTS Many thanks to Ed Decort of Southampton College and Chris Pickerell, Gregg Rivara, and Mark Cappellino of Cornell Coop- erative Extension for assistance with field and lab work, and to Southampton College for a Faculty Research Release Time award (to STT). We thank the personnel of the Bureau of Shellfisheries, New York State Department of Environmental Conservation for their cooperation and assistance, and Dr. Bob Nuzzi, Vito Minei. and Mac Waters of the Suffolk County Department of Health Services. Office of Ecology, for Aureococcus and salinity data. We also thank the Town of Riverhead Shellfish Program for use of rafts at East Creek. We gratefully acknowledge funding for this work by the National Marine Fisheries Service, the Environmental Protection Agency's Near Coastal Waters Program, and the New York Sea Grant Institute. In particular, we thank Cornelia Schlenk of NYSGI, Jon Gorin, Rick Balla, and Felix LoCicero of USEPA, and Vito Minei and Walt Dywidiak of the Peconic Estuary Pro- gram Office. A special thanks to bayman Peter Wenczel for the many ways in which he inspired and assisted in this study. LITERATURE CITED Barber. B. J. & N. J. Blake. 1991. Reproductive physiology, pp. 377-428. In: S. E. Shumway (ed.). Scallops: Biology. Ecology, and Aquaculture. Elsevier. New York. Belding. D. L. 1910. The scallop fishery of Massachusetts. Marine Fish- eries Service — No. 3, Division of Fisheries and Game. Department of Conservation, Commonwealth of Massachusetts, Boston. MA. 51 pp. Bologna, P. A. X. 1998. Growth, production, and reproduction in bay scal- lops Argopecten irradians concentricus (Sayl from the northern Gulf of Mexico. J. Shellfish Res. 17:91 1-917. Bricelj, V. M., J. Epp & R. E. Malouf. 1987a. Comparative physiology of young and old cohorts of bay scallop Argopecten irradians irradians (Lamarck): mortality, growth, and oxygen consumption. J. Exp. Mar. Biol. Ecol. 112:73-91. Bricelj. V. M.. J. Epp & R. E. Malouf. 1987b. Intraspecific variation m reproductive and somatic growth cycles of bay scallops Argopecten irradians. Mar. Ecol. Prog. Ser. 36:123-137. Bricelj, V. M. & S. H. Kuenstner. 1989. Effects of the "brown tide" on the feeding physiology and growth of bay scallops and mussels, pp. 491- 509. In: E. M. Cosper. V. M. Bricelj, and E. J. Carpenter (eds.). Novel Phytoplankton Blooms: Causes and Impacts of Recurrent Brown Tides and Other Unusual Blooms. Coastal and Estuanne Studies 35. Springer. New York. Coe. W. R. & H. J. Turner. Jr. 1938. Development of the gonads and gametes in the softshell clam (Mya arenaria). J. Morphol. 62:91-1 I I. Cosper. E. M., W. Dennison. A. Milligan. E. J. Carpenter. C. Lee. J. Holza- pfel & L. Milanese. 1989. An examination of the environmental factors important to initiating and sustaining "brown tide" blooms, pp. 317- 340. In: E. M. Cosper. V. M. Bricelj. and E. J. Carpenter (eds.). Novel Phytoplankton Blooms: Causes and Impacts of Recurrent Brown Tides and Other Unusual Blooms. Coastal and Estuarine Studies 35. Springer. New York. Epp. J. 1989. Energy storage and utilization in the bay scallop, Argopecten irradians irradians (Lamarck). M.S. thesis. SUNY-Stony Brook. Stony Brook. New York. 83 pp. Epp, J., V. M. Bricelj & R. E. Malouf. 1988. Seasonal partitioning and utilization of energy reserves in two age classes of the bay scallop Argopecten irradians irradians (Lamarck). J. Exp. Mar. Biol. Ecol. 121:113-136. Gallager, S. M.. D. K. Stoecker & V. M. Bricelj. 1989. Effects of the brown tide alga on growth, feeding physiology, and locomotory be- havior of scallop larvae (Argopecten irradians). pp. 51 1-541. In: E. M. Cosper. V. M. Bricelj. and E. J. Carpenter (eds.). Novel Phytoplankton Blooms: Causes and Impacts of Recurrent Brown Tides and Other Unusual Blooms. Coastal and Estuanne Studies 35. Springer, New York. Gobler. C. J. 1995. The role of iron in the occurrence of Aureococcus anophagefferens blooms. M.S. thesis, SUNY-Stony Brook. New York. 127 pp. Hampson, G & J. M. Capuzzo. 1984. Growth and reproduction of bay scallops in shallow and deepwater embayments. WHOI Tech. Rept. WHOI-84-38:4-5. Hickey. M. T. 1977. Age. growth, reproduction, and distribution of the bay scallop. Aequipecten irradians irradians (Lamarck), in three embay- ments in eastern Long Island. New York, as related to the fishery. M.S. thesis, C. W. Post College, Long Island University, Brookville. New York. 101 pp. Humason. G. L. 1979. Animal Tissue Techniques. W. H. Freeman & Co.. San Francisco. Jubb. K. V. F.. P. C. Kennedy & N. Palmer. 1985. Pathology of Domestic Animals, vol. 3: 3rd ed.. Academic Press. Orlando. FL. Kelley, K. M. & J. D. Sisson. 1981. Seed sizes and their use in determining spawning and setting times of bay scallops on Nantucket, pp. 43^49. 58 Tettelbach et al. In: K. M. Kelley (ed.). The Nantucket Bay Scallop Fishery: The Re- source and Its Management. Shellfish and Marine Department. Nan- tucket. MA. Langton. R. W., W. E. Robinson & D. Schick. 1987. Fecundity and repro- ductive effort of sea scallops Placopecten magellanicus from the Gulf of Maine. Mar. Ecol. Prog. Ser. 37:19-25. Levitan, D. R.. M. A. Sewell & F. S. Chia. 1992. How distribution and abundance influence fertilization success in the sea urchin Strongylo- centrotus franciscanus. Ecology 73:248-254. MacDonald. B. A. & R. J. Thompson. 1988. Intraspecific variation in growth and reproduction in latitudinally differentiated populations of the giant scallop Placopecten magellanicus (Gmelin). Biol. Bull. 175: 361-371. MacFarlane, S. L. 1991. Managing scallops Argopecten irradians irradi- ans (Lamarck. 1819) in Pleasant Bay. Massachusetts: large is not al- ways legal, pp. 264-272. In: S. E. Shumway and P. A. Sandifer (eds, I An International Compendium of Scallop Biology and Culture. World Aquaculture Society. Baton Rouge. LA. Marshall, N. 1960. Studies of the Niantic River. Connecticut with special reference to the bay scallop. Aequipeclen irradians. Limnol. Oceanogr. 5:86-105. Naidu. K. S. 1970. Reproduction and breeding cycle of the giant scallop Placopecten magellanicus (Gmelin) in Port au Port Bay. Newfound- land. Can. J. Zool. 48:1003-1012. Paulet, Y. M., A. Lucas & A. Gerard. 1988. Reproduction and larval de- velopment in two Pecten maximus (L.) populations from Brittany. J. Exp. Mar. Biol. Ecol. 119:145-156. Peterson. C. H. & H. C. Summerson. 1992. Basin-scale coherence of popu- lation dynamics of an exploited marine invertebrate, the bay scallop: implications of recruitment limitation. Mar. Ecol. Prog. Ser. 90:257- 272. Risser, J. 1901. Habits and life-history of the scallop (Pecten irradians). pp. 47-55. In: 3 1st Annual Report to the Rhode Island Commissioners of Inland Fisheries. (Cited in: Barber. B. J. & N. J. Blake. 1991. Re- productive physiology, pp. 377-428. In: S. E. Shumway (ed.). Scal- lops: Biology, Ecology, and Aquaculture. Elsevier, New York). Sastry, A. N. 1963. Reproduction of the bay scallop Aequipecten irradians Lamarck, influence of temperature on maturation and spawning. Biol. Bull. 125:146-153. Sastry, A. N. 1970. Reproductive physiological variation in latitudinally separated populations of the bay scallop. Aequipecten irradians Lama- rck. Biol. Bull. 138:56-65. SCDHS. 1993-1995. Brown Tide Cell Counts— 1993-1995. Monitoring Reports. Suffolk County Department of Health Services, Office of Ecology, Riverhead, NY. Sieburth, J. M.. P. W. Johnson & P. E. Hargraves. 1988. Ultrastructure and ecology of Aureococcus anophagefferens gen. et sp. nov. (Chrysophyceae): the dominant picoplankter during a bloom in Nar- ragansett Bay, Rhode Island. Summer 1985. J. Phycol. 24:416-425. Smith, C. F. & S. T. Tettelbach. 1996. Bay Scallop Restoration: Western Peconic Bay. Final report submitted to the Environmental Protection Agency. 44 pp. Smith. C. F. & S. T. Tettelbach. 1997. Restocking Bay Scallops. Final Report to the National Marine Fisheries Service. 76 pp. Taylor. R. E. & J. M. Capuzzo. 1983. The reproductive cycle of the bay scallop, Argopecten irradians irradians (Lamarck), in a small coastal embayment on Cape Cod, Massachusetts. Estuaries 6:431^t35. Tetra Tech, Inc. 1997. Surface Water Quality Modeling of the Peconic Estuary — Calibration of EFDC Hydrodynamic Model. Interim Rept. 2. submitted to Peconic Estuary Program, Suffolk County Department of Health Services. Suffolk County, New York. Tettelbach. S. T. 1991. Seasonal changes in a population of northern bay scallops. Argopecten irradians irradians (Lamarck, 1819). pp. 164- 175. In: S. E. Shumway and P. A. Sandifer (eds). An International Compendium of Scallop Biology and Culture. World Aquaculture So- ciety. Baton Rouge, LA. Tettelbach, S. T. & E. W. Rhodes. 1981. Combined effects of temperature and salinity on embryos and larvae of the northern bay scallop, Ar- gopecten irradians irradians. Mar. Biol. 63:249-256. Tettelbach. S. T. & P. Wenczel. 1993. Reseeding efforts and the status of bay scallop Argopecten irradians (Lamarck, 1819) populations in New- York following the occurrence of "brown tide" algal blooms. /. Shell- fish Res. 12:423-431. Tettelbach, S. T.. P. Wenczel & S. W. T. Hughes. 1994. Size variability of juvenile (0+ yr) bay scallops in Long Island, New York populations. /. Shellfish Res. 13:284. Van der Kraak, G.. J. P. Chang & D. M. Janz. 1998. Reproduction, pp. 480-483. In: D. H. Evans (ed.). The Physiology of Fishes, 2nd ed. CRC Press. Boca Raton. FL. Journal of Shellfish Research, Vol. 18, No. 1. 59-66, 1999. SOME METHODS FOR QUANTIFYING QUALITY IN THE SCALLOP PECTEN MAXIMUS (L.) JULIE A. MAGUIRE, PIERRE G. FLEURY,1 AND GAVIN M. BURNELL Aquaculture Development Centre Department of Zoology and Animal Ecology Lee Makings, Prospect Row University College Cork Cork. Ireland ABSTRACT Because biological systems do not work in isolation, behavioral, biochemical, and physiological tests can give an overview of an individual's vital processes and reaction to stress. Two stress gradients were applied in this study, a short acute desiccation stress and a long-term density stress. These stress gradients were used to assess the usefulness of various techniques for quality assessment; namely, a standard salinity stress test, condition index, recessing speed of the scallop, adenylic energetic charge (AEC). and percentage carbohydrate content of the striated muscle. The results showed that AEC could be used effectively to measure the effect of a short-term stress. In the striated muscle. AEC levels were useful in discriminating between good and poor quality scallops. The total carbohydrate content in the striated adductor muscle and condition index were useful in assessing the effect of long-term stress on scallop quality. The most promising results arose from the recessing trials, because this nondestructive test successfully discriminated the different qualities of scallops arising from both long- and short-term stress. KEY WORDS: Pecten maximus, stress, quality, desiccation, density INTRODUCTION Juvenile scallops are either collected by natural settlement onto artificial collectors or produced in a hatchery. Intermediate culture of spat then takes place in suspended culture or cages on the sea bottom until the scallops reach a size (35-50 mm) that offers some protection from predation. Final outgrowth can take place in sus- pended cage culture or by ranching them on the seabed. Large variations in the survival and performance of spat and juveniles during transport, nursery, and outgrowth have demonstrated the need for research into the effect of stress on the quality of the scallop Pecten maximus (Maguire 1998). Stress has been defined as "the effect of any environmental alteration or force that extends homeostatic or stabilizing processes beyond their normal limits at any level of organization." (Esch and Hazen 1978). Chronic sublethal stress, such as pollution from heavy metals or stocking at high densities, can cause an even or negative scope for growth (Thompson and MacDonald, 1991) and can occur over months or even years. Short acute stresses can occur over hours or days for example, desiccation, thermal shock, and salinity, but both types of stress can eventually result in mortality. The stress effect of various husbandry practices on the physiology of bivalve mollusks is virtually unknown but is believed to be significant. Dhert et al. (1992a). Dhert et al. ( 1992b) considered stress tests to be invaluable in testing the nutritional requirements of aquacul- ture species at various stages of their development and established a standard stress test to determine the quality of shrimp and fish fry. in which they used elevated salinity as a stressor. Duran- Gomez et al. ( 1991 ) developed a test to be performed on postlarval prawns Penaeus japonicus (Bate) using salinity and pH shocks as stressors. Likewise, Ashraf et al. (1992) employed a standard sa- linity stress test to detect differences in nutritional studies when no 'Direction des Resources Vivantes. IFREMER. Centre de Brest. BP 70, 29280 Plouzane, France. Corresponding Author: Gavin M. Burnell. Tel-(353)21 904192. Fax-(353) 21 270562. email: g.burnell@ucc.ie. differences existed in survival and growth using larval striped bass Morone saxatilis (Walbaum) and the silverside Menidia beryllina (Cope) as the experimental organisms. Because biological systems do not work in isolation, a combi- nation of physiological, biochemical, and behavioral tests can give a more complete picture of an individual organism's reaction to stress. Examples of some techniques used for assessing quality in bivalve molluscs are listed in Table 1. Scallops have some unique behavioral traits among bivalves in that they have the ability to swim relatively long distances in an oriented way. They can also recess into the sediment, first de- scribed by Baird and Gibson (1956). Therefore, potential behav- ioral tests could include recessing and righting behavior (turnover after being placed flat side down), which would affect their ability to withstand predation. Recessing requires a large energetic cost, and scallops that are already weakened by the stress of handling or exposure to air during transport would be less able to escape from predators by recessing or swimming when returned to the sea. Fleury et al. (1997) completed a study of the recessing behavioral of three sizes of ranched scallops during three seasons and three sizes and used adenylic energetic charge as an index. They dis- covered that the best seeding time was in the spring and summer and that within this period, medium sized scallops (30 mm) re- cessed more effectively than the small (15 mm) or larger (42 mm) sized scallops. In our study, recessing speed was used as a method for stress assessment. The effect of a short-term stress on the biochemistry of the animal can be measured by its level of adenylic energetic charge (AEC). AEC is defined by the ratio: AEC = (ATP + 0.5 ADP) -^ (ATP + ADP + AMP) where (ATP = adenosine triphosphate, ADP = adenosine diphosphate, AMP = adenosine monophos- phate). The triphosphate bond of the ATP molecule has maximum energy, the diphosphate bond of ADP is half as rich, and the monophosphate bond (AMP) lacks energy. The AEC ratio ranges from 0 to 1; that is, (when 0, all nucleotides are AMP, and when 1. all nucleotides are ATP). Therefore, the relative level of these bonds can be used as a measure of the energy directly available to 59 60 Maguire et al. TABLE 1. A review of techniques used for quality assessment. General Category Technique Used Species Stress Reference Standard stress test "Aerial exposure Mytilus edulis (L.) Chronic Veldhuizentsoerkan et al. (1991) Acute Viarengo et al. ( 1995) Biometrics ""Condition index Flesh condition Oslrea edulis (L. ) M. edulis Pinctada fucata martensii (Dunker) Crassostrea virginica (Gmelin) Ruditapes philippinarum (Adams and Reeve) Argopecten ir radians i /radians (Lamark) M. edulis Chronic Rogan et al. ( 1991 ) Lundebye et al. (1997) Numaguchi (19951 Fisher et al. (1996) Isonoet al. (1998) Rheault and Rice (1996) Agirregoikoa et al. ( 1991 ) Behavior "Recessing Pectin maximus Fleury et al. (1997) Biochemical "Adenylic energetic charge "Carbohydrate content Lipid content Total oxyradical scavenging capacity RNA:DNA Bivalve mollusks C. gigas (Thunberg) Dreissena polymorpha (Pall.) M. edulis Eiinila ziczac (L.) P. maximus Placopecten magellanicus (Gmelin) Acute Moal et al. (1989a) Chronic Kaufmann et al. ( 1994) Short Sprung and Borcherding (1991) Chronic Regoli et al. (1998) Lodeiros et al. (1996) Robbins et al. (1990) Kenchington (1994) Cytochemical Lysosomal membrane fragility M. edulis Mya arenaria (L.) Digestive tubule and vesicular C. virginica connective tissue condition Chrome Pelletier et al. (1991) Tremblay et al. (1997) Fisher et al. (1996) Physiological Scope for growth Oxygen consumption: ammonia excretion Lipofusin accumulation O. edulis Pcrua viridis < L. ) M. edulis Amhlema plicata (Say) P. viridis Siinena scripla (L) Chronic Hutchinson and Hawkins ( 1992) and acute Chronic Cheung and Cheung ( 1995) Hatcher et al. (1997) Acute Barker and Horbach (1997) Mathew and Damodaran (1997) ' Techniques used to measure stress in this study. the cells at that particular time. For example, empirical studies have shown that a very weak, stressed scallop would have an AEC level (measured from the striated muscle) of 0.3 to 0.5 (Fleury. pers. comm.). Such a scallop would have a negative scope for growth and would have a poor chance of recovery. A scallop recording a level of 0.5 to 0.7 would have reduced growth, would not reproduce but could recover to its original quality. A healthy scallop would have an AEC level of 0.8 to 1. Adenylic energetic charge was first proposed as a stress index by Atkinson (1968). who suggested that modulations in the levels of adenylphosphate reflected variations of enzyme activity at key points in metablic pathways that yield energy in the form of high energy adenine- phosphate bonds. These variations are a result of external stress. In other words, the more stressed an animal becomes, the more en- ergy it uses to counteract the stress, thus lowering its AEC level. Many studies have been carried out using AEC as a stress index or in nutritional studies on different marine animals; for example, the marine isopod Cirolana boreulis (Lijborg) (Skjoldal and Bakke 1978). the European sea bass Dicentrarchus labrax (L.) (Reali et al. 1987), the oyster Crassostrea gigas (Moal et al. 1989b). the spider crab Hyas araneus (L.) (Harms 1992). the sturgeon Aci- penser beari (Brandt) (Salin 1992). the oyster C. angulata (Lamark) (Madureira et al. 1993) and the scallop P. maximus (Fleury et al. 1997). In juvenile scallops, the level of AEC varies between tissues. Le Coz (1989), in a comparative study of different tissues in the juvenile scallop P. maximus. found the highest AEC ratios in the adductor muscle. Within the muscle, the highest level was found in the striated part (mean = 0.93). which is concerned with the fast repetitive opening and closing of the valves of the scallop, thus enabling the animal to swim, to escape from predators, and to recess. In the smooth part of the muscle, the AEC results were Quantifying Scallop Quality 61 more variable. The smooth muscle has slower contractions and is capable of keeping the scallop shell closed for long periods, with little energy expenditure (Chantler 1991). Energy is transported from the muscle to the various organs via the haemolymph. The haemolymph of bivalves is also concerned with a variety of physiological functions; that is, transport of nu- trients and wastes, gas exchange, osmoregulation, and defence (Benniger and Le Pennec 1991). Therefore, in this study, we looked at the effect of a desiccation stress on AEC levels in the smooth and striated part of the adductor muscle and in the hae- molymph of P. maximus juveniles. The effect of a long-term stress on the biochemistry of an animal can be measured by the carbohydrate content of the smooth and striated adductor muscle, respectively. The adductor muscle is the main storage area for energy reserves. Many studies have concentrated on the seasonal partitioning of energy reserves in bivalves: for example. Epp et al. ( 1 988) studied energy partitioning of the bay scallop A. irradians. Walne ( 1970) assessed the seasonal variation of the glycogen content of seven populations of the oys- ter O. edulis. De Zwaan and Zandee (1972) studied the utilization of glycogen and accumulation of some intermediates during anaerobiosis in M. edulis. In this study, the effect of high stocking density on the carbohydrate content of cultured scallop spat was assessed. The criteria for a useful "stress detector" are that it should be reliable and significant; that is with little individual variation within the populations and significant differences between popu- lations. Quality in this study was defined by the degree of acute (emmersion) or chronic (density) stress endured by the scallops during these trials. Therefore, the objectives of this study were divided into two parts. First, to create different juvenile scallop qualities using a desiccation stress gradient and to use these ref- erence animals to compare different laboratory techniques, (stan- dard salinity stress test, recessing behavior and level of adeny lie- energy charge) for quality assessment. Second, to use the same laboratory techniques, (including total carbohydrate content in- stead of level of AEC) to measure quality in a case study where scallops were cultured at three different densities. MATERIALS AND METHODS The scallop spat (30 mm) used in this experiment were col- lected from the Rade de Brest, France. Shell length, height, depth, and total wet weight measurements were taken from a subsample of 100 spat used in each experiment, and a condition index was compiled: condition index = [Weight/! Height x Length x Depth)] x 10.000 (Fleury, pers. comm.). The scallops were acclimated in tanks and were maintained at a temperature of 15°C and a salinity of 35%e and fed an equal mixture (1 x 107 mL~') of batch cultured algae Pavlova lutheri (Droop). Isochrysis galbana, and Chaetoceros calcitrans (Paulsen) in volumes sufficient to give a tank concentration of 30-50 cells p-L.-1. The scallop were scrubbed clean of epibiota and used in experiments within 2 weeks. Creation of a Gradient of Scallop Qualities Using a Desiccation Stress (Short-Term Stress) Four batches (A-D) containing three replicates each of healthy scallop spat (n = 30) were used for each of three experiments (Expt. 1-3). The spat were individually weighed, labeled with a permanent marker, and placed out of water in a constant tempera- ture room for 0, 3, 6. and 12 h ( = A. B, C and D, respectively). The air temperature used to stress the scallops was 19°C for the first experiment, 15°C for the second, and 1 7°C for the third. The stress detector tests (standard stress test, recessing ability, and level of adenylic energetic charge) were carried out on each batch (A-D) to determine whether the tests could discriminate among the batches. Determination of Various Scallop Qualities L'sing Scallop Spat Cultured at Different Densities (Long-Term Stress) The scallop postlarvae (2 mm), were taken from Tinduff Hatch- ery/Nursery in April 1995. They were transferred to the Bay of St. Brieuc. Three months later (July 5). the scallops were removed from the cages and graded by shell size (mean ± SD height 12 ± 2 mm). They were placed in new cages (0.75 m2) with a larger mesh size (5x5 mm). A range of stocking densities from 700 to 900 to 1,250 scallops per tier was set up and was referred to as density 1, 2. and 3. Nine replicates of each experimental density were used. After a 3-month period (October 5). the scallops were retrieved by SCUBA diving from the cages at each density. During transport (4 h), the spat were wrapped in towels soaked with seawater. The juveniles were then stored in aerated seawater tanks at 16°C over- night. Over the next 2 weeks, various stress tests were carried out. These were a standard salinity stress test (2-wk duration), recessing ability (2-wk duration), and total carbohydrate content fixed im- mediately. A description of these tests follows. Standard Stress Test A useful stress test will pick up differences induced by a stress gradient. The ultimate reaction to stress is mortality, so this was used as a standard assessment. Shell height, length, depth, and various wet weight measurements were taken before and after the standard stress was completed to enable condition indices to be computed. The standard stress tests were performed in a cubic recirculat- ing tank (1.5 x 1.5 m). The experimental salinity was 25%c tem- perature 15 ± 1°C for experiments 1 and 2. this was made up using seawater and distilled water. This acted as a semi-severe stress to the already stressed spat to hasten mortality. The experiments took 2 weeks to complete. In experiment 3, the salinity stress test was carried out using freshwater (temperature 14 ± 1°C) to achieve a quicker result. The spat were given food daily at the same rate with the same species of algae used during their acclimation period. However, even in experiments 1 and 2 (257(c) the scallops were so stressed that they did not seem to feed. Survival was monitored twice per day over a 2-wk period in experiments 1 and 2 and every 15 min over a 2 h period for the freshwater test (experiment 3). The criterion for death was open valves with a lack of valve contraction when touched by a glass rod. All scallops were then reweighed and the shell length, height, and depth were recorded. Recessing Behavior Twenty scallops each from the different groups of spat were quickly measured for shell length, height, depth, and total wet weight. The spat were color labeled and placed in a tank (length 2m. width 0.5 m) with recirculating seawater (salinity 35% 5 mm particle size. 58% 2 to 5-mm. 35% 1 to 2-mm, and 3% < I -mm particle size. The juveniles were fed a mixture of batch-cultured algae, at the same volume used during their acclimation period. Recessing time was monitored every 4 h. and scallops were recorded as recessed (completely covered by substrate), semi-recessed (half covered by sediment), or not recessed. Extraction and Analysis of Nucleotides Scallop parameters (shell length, height, depth, and total wet weight) were quickly measured for each batch of spat. The scallop was rapidly dissected and the striated and smooth muscle sepa- rately removed and frozen in liquid nitrogen. There it was stored until analysis (within a few days). Moal et al. (1989a) found that a better nucleotide extraction was obtained when the required tis- sue, rather than the whole animal, was frozen. At the time of the analysis, the striated and the smooth part of the muscle were withdrawn from the liquid nitrogen. One mL of 0.5M ice-cold TCA was then added immediately to each sample, as better recovery of ATP was observed using TCA as compared to other acids; for example perchloric acid (PCA) (Moal et al. 1989a). Preliminary crushing of the extracts increases the stability of the neutralized extracts. The tissue (still frozen) was instanta- neously homogenized at 25.000 rpm for 10 s. The homogenate was centrifuged for 10 min at 4.500 rpm. and the supernatant was neutralized with 0.5 m fresh amine freon solution. The neutralized sample was either stored at -18°C or immediately analyzed by high-performance liquid chromatography (HPLC). Analysis The HPLC apparatus was composed of a pump (Waters model 510). an automatic injector (Kontron 460), and a spectrophotom- eter (Merck L4250). The separation took place in a C18 column of length 150-mm. diameter 4.6-mm (model SFCC/Shandon Spher- isorb 3u-OD52). and ultraviolet light (254 nm) was used for the detection of the nucleotides. An isocratic NaH2P04 (0.15 m) buffer (pH 6) containing an ion-pairing agent (0.005 M tetrabuty- lammonium) and 5% methanol was used to elute the nucleotides. All chemicals were of analytical grade and supplied by Sigma. Separation took approximately 30 min at a flow rate of 1 mL/min. Carbohydrate Content Biometric measurements were taken for each scallop from the different spat groups. The animals were rapidly dissected, and the striated muscle was removed and immediately placed in liquid nitrogen. At the time of analysis, the samples were withdrawn and freeze dried using a HETOSICC CD 53-1 freeze dryer. The car- bohydrate content was analyzed using a miniaturization of the Dubois et al. (1956) method. Twenty (xg of the muscle sample were crushed and resuspended in 1 mL of distilled water. Fifty p.L of the mixture was placed in an epindorff tube. 50 yiL of 5% phenol was added, and the resultant solution was allowed to stand for 20 min at 15°C. Five hundred u.L of 98% H2S04 was added, and the tube was placed on ice. After centrifugation. the absor- bance of the supernatant was read at 492 and 620 mu, using a spectrophotometer model SLT Spectra. A glucose standard was used at concentrations of 0. 50. 100. 150. and 200 p,g of glucose per mL of distilled water, and blanks were made using distilled water. Statistical Analyses Nonparametric data were normalized by log transformation or arcsine square root transformation for percentage data. One-way analyses of variance (ANOVAs) were used to test significant dif- ferences among treatments, and a posteriori Tukey test was used to contrast treatments. The level of significance was set at 0.05. RESULTS Standard Stress Test Figure I shows the mean survival times (over 2-week test pe- riod) of each population for each test (desiccation temperature 19° and 15°C). It showed that the degree of desiccation endured (0-12 h) by each group was directly proportional to the mortality rate of each group. However, the desiccation temperature of 1 9°C was too high, because all the spat from group D (12-h emmersion) died either during the last hour of desiccation or immediately after reimmersion. Despite this, a significant difference was found be- tween the spat groups created by using the higher desiccation temperature. The data for test 2 (desiccation temperature 15°C) showed a significant difference in the mean survival times between groups A/B, C, and D (0, 3, 6. and 12-h desiccation) with similar mortalities occurring between groups A and B (0- and 3-h desic- cation). Figure 2 shows the survival of the four populations (A-D) in test 3. using a freshwater standard stress (temperature 14°C). The data showed no significant difference between the populations (/> > 0.05). The stress used in this test was too severe to pick up the subtle differences in quality between the populations. The standard salinity stress test (water temperature 15°C, sa- linity 25%) was carried out on the groups 1-3 of the spat density experiment, and no significant difference was found between the survival of the different density treatments. Only 10% mortality was recorded in the test. Recessing Behavior Table 2 shows the recessing time of the four scallop groups in the desiccation experiment and the three groups in the density experiment. Recessing speed was directly proportional to the des- iccation endured (0-12 h) by the spat and the density (700-1,250 spat per tray). A significant difference was found among the treat- I mmersion Time (hours) Figure 1. The mean survival times over a 2-week period of four dif- ferent qualities (desiccation times: A = 0h, B = 3h, C = 6h, and D = 12 h) of juvenile scallops to a standard salinity stress of reduced sa- linity (S = 25%r, T = 15 C and 19°C). Quantifying Scallop Quality 63 45 60 75 Time (minutes) Figure 2. The = 0 h, B = 3 h. salinity stress survival of four different qualities (desiccation times: A C = 6 h, and D = 12 h( of juvenile scallops to a standard of freshwater ( 14°C). merits in the desiccation experiment (F76 3 = 74.2. p < 0.01) and the density experiment (F1172 = 13.67. p < 0.01 ). Adenylic Energetic Charge Table 3 shows the relationship between the striated and smooth adductor muscle of the four populations of spat. The highest levels of AEC for all groups was found in the striated adductor muscle. The AEC level in the striated muscle was significantly higher than the AEC level in the smooth muscle for each population (group A t34 = 19.56. p< 0.01; group B t36 = 12.12, p< 0.01; group C t3S = 3.34, p < 0.01; and group D t34 = 7.32, p < 0.01 ). The AEC levels in the striated adductor muscle clearly showed two significant groups (F72 3 = 24.15. p < 0.01). Scallops from group A/B had higher AEC levels (> 0.85) than the scallops from group C/D (< 0.75). In the smooth muscle, the highest levels of AEC were found in population B, and again levels significantly decreased from this in group C and D (F65 3 = 4.53, p < 0.01). Hemolymph was also extracted, but the AEC results were deemed to be unreliable because of the difficulty of extracting the hemolymph. Carbohydrate Content Table 4 shows percentage carbohydrate in dry weight of the striated adductor muscle and the condition index of scallops cul- TABLE 2. Mean ± SD recessing time of different spat qualities in the short- (desiccation) and long-term (density) experiments. Spat Group Experiment 1 Average Recessing Time (Days) A (0-h desiccation) B (3-h desiccation) C (6-h desiccation) D ( 12-h desiccation) Spat Group Experiment 2 1.6±0.4J 2.4 ± 0.6b 3.5±0.T 5.7 ± 1.6d Average Recessing Time (Days) Group 1 (density 700 per tray) Group 2 (density 900 per tray) Group 3 (density 1250 per tray) 1.73 ±0.93" 2.28 ± 1.24" 3.13 ± 1.38h Any two means sharing a common letter in each column are not signifi- cantly different at p < 0.05 (Tukey test). TABLE 3. Levels of AEC (mean ± SD) in the adductor muscle of four different groups of scallop spat. Group Striated Muscle Smooth Muscle A (0-h desiccation) B (3-h desiccation) C (6-h desiccation) D (12-h desiccation) 0.89 ±0.04'' 0.87 + 0.1 I" 0.71 ±0.1b 0.72 + 0.08b 0.55 ± 0.07ab 0.65 + 0.11- 0.61 ±0.08"b 0.53 + 0.08b Any two means sharing a common letter in each column are not signifi- cantly different at p < 0.05 (Tukey test). tured at three different densities. The carbohydrate content in- creased significantly from spat cultured at a density of 700 and 900 per tray to those cultured at 1,250 per tray (F56 2 = 5.25, p < 0.01). Biometrics Shell length, height, depth, and total wet weight measurements were taken from all spat held at each stocking density, and a condition index was calculated: condition index = [Weight/ (Height x Length x Depth)] x 10,000. Table 4 represents the average value calculated per scallop at each density. Spat cultured at a density of 700 and 900 per tray had a similar condition index. The condition index decreased signifi- cantly for those cultured at a density of 1,250 per tray (Fn72 = 4.188, p< 0.05). DISCUSSION Standard Stress Test In our study, salinity was reduced to 25 and 0%o, respectively, with the aim of inducing stress and, hence, mortality, which could be used to quantify the quality of the spat. Quality in this study was defined by the degree of acute (emmersion) or chronic (high- density) stress endured by the scallops during these trials. Simi- larly Viarengo et al. (1995) reported that a simple secondary stress response in mussels showed a sensitivity in the same range as other commonly used general stress indices at the cellular level. The results showed that short-term exposure of mussels to sublethal concentrations of pollutants significantly reduced mussel survival in air. Dredge (1997) suggested that saucer-shaped scallops Amu- sium japonicum balloti (Bernardi) can withstand exposure to air for up to 2 h before suffering significant mortality. TABLE 4. Mean ± SD percentage carbohydrate content of dry weight in the striated adductor muscle and the mean condition index of scallops cultured at three different densities. Scallop Density per Tray % Carbohydrate Content Condition Index Value 700 900 1,250 8.9 ± 2.4° 8.8 ± 2.5a 11.5±3.8b 5.81+0.42" 5.81 ±0.46a 5.55 ± 0.48" Any two means sharing a common letter in each column are not signifi- cantly different at p < 0.05 (Tukey test). 64 Maguire et al. Although different spat qualities were obtained when the spat were removed from water for 0. 3. 6, and 12 h (groups A-D) at a desiccation temperature of 19°C. this temperature was considered too high, particularly for the group D scallops, because some of the spat from this group died during the 12-h desiccation period. Therefore, an air temperature of 15°C is recommended to give a wider range of spat quality. Similarly. Hutchinson and Hawkins (1992) measured stress in the oyster O. edulis using scope for growth as an index. A severe reduction in scope for growth was observed when oysters were placed in conditions where high tem- peratures were combined with low salinity. The third test using freshwater as the stress test was found to be too severe, and no difference was found among the treatments because of the rapid mortality (within 1 20 min) of all groups. This is contrary to a study by Dhert et al. (1992b), who worked on the use of stress evaluation as a tool for the quality control of hatchery-produced shrimp and fish fry. In their experiments, the best results were achieved with stress tests performed within a 60 to 90 min period containing 15 to 30 evaluation points. A standard salinity stress test (water temperature 15°C, salinity 25%o) was carried out on the groups 1 to 3 of the density experi- ment, and no difference was found among the populations. Very few mortalities were recorded in the test. Dhert et al. (1992b) emphasized the importance of using the appropriate salinity level for each species and for each larval stage. Apparently, in our test, the salinity level was not severe enough to differentiate the differ- ent densities, or there was no difference in the quality of spat. Recessing Behavior It is not surprising that the best quality scallop recessed into the sediment quickest (Table 2|. Dao et al. (1985) found that when seeding scallop spat on the seabed, success seemed to depend upon three factors; namely, the quality, the size of the scallop, and the time of year that seeding takes place. By removing seasonal and size variables, we were able to demonstrate a relationship between quality and behavior in juvenile scallops. This is beleived to be the first time this has been demonstrated experimentally. Tyurin (1991) worked on the behavioral reactions of the scallop Mizu- hopecten yessoensis (Jay) to reduced salinity and oxygen exposure to synthetic detergents. Under unfavorable conditions, the test scallops were stressed, could not recess, and elicited an escape response instead. In this study, the recessing speed of scallops deteriorated significantly as the stocking density increased. The recessing test is, therefore, not only sensitive to subtle changes in the spat quality, but is also a very quick and simple test to perform. Adenylic Energetic Charge In general, the results indicated that as the stress level in- creased, the AEC level decreased in the striated muscle, to a cer- tain point where the AEC level did not decrease any more. This seems to be the threshold level for this test. Similar results have been shown by Madureira et al. (1993). who looked at the effect of polychorinated biphenyl (PCB) on adenylic energetic charge in the oyster C. angulata, which was fed a PCB-contaminated algal cock- tail. They found that the level of PCB increased with time within the animal and that this sublethal stress resulted in the reduction of AEC levels as PCB concentration increased. In our study, the striated muscle was found to be the best tissue to use when measuring AEC levels, because there was little indi- vidual variation within the groups and a large difference between stressed (group C and D) and unstressed (group A and B) treat- ments. Similarly, Le Coz (1989) reported that highest levels of AEC were found in the striated adductor muscle of P. maximus and that AEC results were more variable in the smooth muscle. In prior studies, (Fleury et al. 1997) the level of AEC in the striated muscle seemed to be a better measure of stress and quality, than the smooth muscle, because the decrease of AEC in this part of the muscle attributed to stress was more pronounced. In our study, as well, a similar AEC decrease was found in the striated muscle. The hemolymph of bivalves is concerned with a variety of physiologi- cal functions, but also the transport of ATP from the striated muscle to various organs. We expected similar results as those found in the muscle. Our results, however, were unreliable because of the difficulty of the hemolymph extraction procedure. The results of our study were consistent with those found by Moal (1989a. 1989b. 1991 ). who showed that the effect of short- term desiccation on the oyster C. gigas was dependent upon sea- son. AEC levels remained high after 3 h of desiccation in January, but decreased after 3 h of desiccation in May and July. Therefore, there is a negative correlation between AEC and season. Our study was only carried out in May, and the results showed a decrease in AEC levels after a 3-h desiccation period. Further experiments would have to be carried out to determine whether AEC levels could be used to quantify stress in P. maximus at other times of the year. Overall, the recessing test was just as reliable as the level of AEC in measuring the effect of stress on scallops. The recessing test is nondestructive: therefore it can be used for continuous monitoring of the same scallop and is more cost effective than the biochemical test. However, to monitor stress, the testing of the shellfish must take place immediately after sampling so that the condition of the scallop will not be altered by handling. Therefore, because a sample can easily be frozen for biochemical analysis later, it may be more convenient to use AEC rather than have to set up a recessing trial immediately after sampling. Carbohydrate Content The main energy reserve in scallops is glycogen, which is stored in the adductor muscle. It is mobilized and converted into usuable energy (ATP) when needed. In general, scallops contain relatively low levels of glycogen in the adductor muscle attaining maximum levels of up to 24% in P. maximus (Ansell 1978). 23- 25% of dry muscle weight in A. irradians (Epp et al. 1988) and 18% in Chlamys islandica (Muller) (Vahl 1981): whereas, the mussle A/, edulis attains glycogen levels of 42-53% in the mantle (Gabbott 1983). The percentage carbohydrate content in this study measured in October was quite low. ranging from 8.8-11% dry weight of the adductor muscle. This is the period when maximum levels of carbohydrate should be found in the adductor muscle after the summer period. Ansell (1978) suggested that the carbo- hydrate content varies among sites and among years in the maxi- mum levels found but generally varies from a minimum in March (2.5%) to a maximum in September (24%). Table 4 showed a surprising result, with the highest level of carbohydrate found in scallops cultured at the highest density ( 1 ,250 juveniles per tray). However, the microanalytical technique used for carbohydrate analysis was precise, and the coefficient of variation between subsamples of a single sample was 2.0. This result was contrary to a study by Kaufmann et al. (1994). who reported that the glycogen content of the Pacific oyster C. gigas Quantifying Scallop Quality 65 decreased by 90% after 5 weeks during a growth trial in Maderia Island. This decrease was attributed to a combination of stress factors. Biometrics The use of condition indices are the traditional method for measuring quality. In this study, the condition index used was sensitive enough to pick up a significant difference in quality between the scallops held at the lower density treatments (700 and 900 scallops per tray) and the high-density treatment (1,250 scal- lops per tray). Similarly, Rheault and Rice ( 1996) found that dou- bling the stocking density of the eastern oyster C. virginica re- sulted in a 20% reduction in the condition of the bivalve. CONCLUSIONS Scallop spat of significantly different quality were obtained and were used as a reference to test techniques for quality assess- ment. It was possible to detect a significant decrease in the qual- ity of scallops with increasing stress conditions in both experi- ments. Both AEC and recessing speed detected acute differences in spat quality in the desiccation experiment. Recessing speed, carbohydrate content, and condition index detected chronic dif- ferences in spat quality brought about by varied stocking densi- ties in the density experiment. These tests can now be reliably used to measure quality or the effect of a chronic or acute stress on scallops. LITERATURE CITED Agirregoikoa. M. G.. M. A. Perez, J. A. Marigomez & E. Angulo. 1991. Relationships between quantitative indexes of individual and digestive cell conditions in the common mussel, Mytilus edulis L. from the Biscay coast. Acta Hydrochim. Hydrobiol. 19:1:29-37. Ansell, A. D. 1978. Storage and utilization of reserves in Pectinid bivalves with particular reference to the adductor muscle. Proceedings of the Scallop Workshop. Brest. France 8-13 May 1978. 17 pp. Ashraf. M. K. L. Simpson, & D. A. Bengston. 1992. Development of sa- linity stress tests for larval striped bass. Morone saxatilis, and inland silversides, Menidia beryllina, used in nutritonal studies. Proceedings of Aquaculture '92, Growing Toward the 21st Century. 31 pp. Atkinson. D. E. 1968. The energy charge of the adenylate pool as a regu- latory parameter. Interaction with feedback modifiers. Biochemistry 7:4030-4034. Baird, T. H. & F. A. Gibson. 1956. Underwater observations on escallop {Pecten maximus L.) beds. J. Mar. Biol. Assoc. U.K. 35:555-562. Baker, S. M. & D. J. Hornbach. 1997. Acute physiological effects of zebra mussel (Dreissena polymorpha) infestation on two unionid mussels. Actinonaias ligamenta and Amblema plicata. Can. J. Fisheries Aquat. Sci. 54:3:512-519. Benninger, P. G. & M. Le Pennec. 1991. Functional anatomy of scallops, scallops: biology, ecology, and aquaculture. pp. 133-223. In: Shum- way S. (ed.). Developments in Aquaculture and Fisheries Science. Elsevier, New York. Chantler, P. D. 1991. The structure and function of scallop adductor muscles, scallops: biology, ecology, and aquaculture. In: Shuinway S. (ed.). Developments in Aquaculture and Fisheries Science. Elsevier, New York. Cheung, S. G. & R. Y. H. Cheung. 1995. Effects of heavy metals on oxy- gen consumption and ammonia excretion in green-lipped mussels (Perna viridis). Mar. Poll Bull. 31:4-12:381-386. Dao, J.C., D. Buestel. A. Gerard. C. Halary & J. C. Cochard. 1985. Le programme de replement de coquilles Saint-Jacques {Pecten maximus L.) en France: finable, resultats et prespectives. Colloque Franco- Japonais D'oceanographie: Marseille France, September 1985. De Zwaan, A. & D. I. Zandee. 1972. The utilization of glycogen and the accumulation of some intermediates during anaerobiosis in Mytilus edulis L. Comp. Biochem. Physiol. 43B:47-54. Dhert, P., P. Lavens & P. Sorgoloos. 1992a. A simple test for quality evaluation of cultured fry of marine fish. Me d. Fac. Landbouww. Univ. Gent 57/4b. 8 pp. Dhert. P., P. Lavens & P. Sorgoloos. 1992b. Stress evaluation: a tool for quality control of hatchery-produced shrimp and fish fry. Aquacul. Eur. 17:2:6-10. Dredge, M. C. L. 1997. Survival of saucer scallops, Amusium japonicum balloti. as a function of exposure time. J. Shellfish Res. 16:1:63-66. Dubois. M., K. A. Gilles, J. K. Hamilton, P. A. Rebers & F. Smith. 1956. Colorimetric method for determination of sugars and related sub- stances. Anal. Chem. 28:350-356. Duran-Gomez, R.. J. M. Rodriguez & J. Morales. 1991. Stress tests: a practical tool to control larval shrimp quality. In: P. Lavens, P. Sorgeloos. E. Jaspers and F. Ollevier (eds.l. Larvi '91, No. 15. Epp. J.. V. M. Bricelj & R. E. Malouf. 1988. Seasonal partitioning and utilization of energy reserves in two age classes of the bay scallop Argopecten irradians irradians (Lamark). ./. Exp. Mar. Biol. Ecol. 121:113-136. Esch, G. & T. C. Hazen. 1978. Thermal ecology and stress: a case history for red-sore disease in largemouth bass. pp. 331-363. In: J. D. Thorpe and J. W. Gibbons (eds.). Energy and Environment Stress in Aquatic Ecosystems. Technical Information Center. U.S. Department of En- ergy. CONF-771114. Fisher, W. S., J. T. Winstead. L. M. Oliver. H. L. Edmiston & G. O Bailey. 1996. Physiologic variability of eastern oysters from Apalachi- cola bay. Florida. J. Shellfish Res. 15:3:543-553. Fleury. P. G.. C. Mingant & A. Castillo. 1997. A preliminary study of the behavior of reseeded juvenile great scallops of three sizes in three seasons. Aquacul. Int. 4:325-337. Gabbott, P. A. 1983. Developmental and seasonal metabolic activities in marine molluscs, pp. 165-219. In: Hochachka P. W. (ed.). The Mol- lusca, 2, Environmental Biochemistry and Physiology. Academic Press, New York. Harms, J. 1992. Effects of nutrition (herbivore vs. carnivore) on energy charge and nucleotide composition in Hyas araneus larvae. Helgol Meeresunters. 46:1:29^14. Hatcher. A., J. Grant & B. Schofield. 1997. Seasonal changes in the me- tabolism of cultured mussels (Mytilus edulis L.) from a Nova Scotian inlet: the effects of winter ice cover and nutritive stress. J. Exp. Mar. Biol. Ecol. 217:1:63-78. Hutchinson. S. & L. E. Hawkins. 1992. Quantification of the physiological responses of the European flat oyster Ostrea edulis L. to temperature and salinity. J. Moll. Stud. 58:215-216. Isono, R. S.. J. Kita & C. Kishida. 1998. Upper temperature effect on rates of growth and oxygen consumption of the Japanese littleneck clam. Ruditapes philippinarum. Nippon Suisan Gakkaishi 64:3:373-376. Kaufmann, M. J., M. N. L. Seaman, C. Andrade & F. Buchholz. 1994. Survival, growth, and glycogen content of pacific oyster. Crassostrea gigas (Thunherg 1793), at Madeira Island (subtropical Atlantic). J. Shellfish Res. 13:2:503-505. Kenchington, E. L. R. 1994. Spatial and temporal variation in adductor muscle RNA:DNA ratio in sea scallops (Placopecten magellanicus) in the Bay of Fundy. Canada. J. Shellfish Res. 13:1:19-24. Le Coz. J. R. 1989. La charge energetique adenylique: mise au point, application a trois mollusques bivalves, synthese et perspectives. Mem- oire IFREMER, Centre de Brest, Direction des Ressources Vivantes. Lodeiros, C. J. M.. R. I. Fernandez. A. Boumati. J. H. Himmelman & K. S. Chung. 1996. Relation of RNA/DNA ratios to growth for the scallop Euvola [Pecten) ziczac in suspended culture. Mar. Biol. 126:2:245- 251. Lundbye, A. K.. W. J. Langston & M. H. Depledge. 1997. Stress proteins 66 Maguire et al. and condition index as biomarkers of tributyltm exposure and effect in mussels. Ecotoxicology 6:3:127-136. Madureira. M. J.. A. M. Picado. A. M. Ferreira. F. Mendonca & Y. Le-Gal. 1993. PCB contamination in the oyster Crassostrea angulata: effects on lipids and adenylic energetic charge. In: W. Sloof and H. de-Kruijf (eds.). Proceedings of the Second European Conference on Ecotoxi- cology 1993, vol. suppl. pts. 1-2. Maguire, J. A. 1998. Aspects of the biology of cultured scallops (Pecten maximus L.) with particular reference to stress. Ph.D. thesis. National University of Ireland. 152 pp. Mathew, S. & R. Damodaran R. 1997. Effect of ambient oxygen concen- tration on lipofuscin accumulation in a clam Sunelta scripla and a mussel Perna viridis. Indian. Mar. Sci. 26:1:57-63. Moal. J.. J. R. Le Coz, J. F. Samain. & J, Y. Daniel. 1989a. Nucleotides in bivalves: extraction and analysis by high-performance liquid chroma- tography (HPLC). Comp. Biochem. Physiol. 93B:307-316. Moal. J.. J. R. Le Coz, J. F. Samain. & J. Y. Daniel. 1989b. Responses and adaptations of adenylate energy charge and digestive enzyme activities to tidal emmersion of Crassostrea gigas population in Marennes. Ole- ron Bay. Sci. Mar. Bare. 53:2-3. 699-704. Moal. J.. J. R. Le Coz. J. F. Samain, & J. Y. Daniel. 1991. Oyster adenylate energy charge (AEC) and its natural variability: implications for envi- ronmental monitoring. Oceanis. Doc. Ocenogr. 17:3:279-280. Numaguchi. K. 1995. Effects of water temperature on catabolic losses of meat and condition index of unfed pearl oyster Pinciada fucata mar- tensii. Fisher. Sci. 61:5:735-738. Pelletier. E.. S. Ouellet & M. Paquet. 1991. Long-term chemical and cy- tochemical assessment of oil contamination in estuarine intertidal sedi- ments. Mar. Poll. Bull. 22:6:273-281. Reali, N.. A. Casti. G. Orlandini & R. Viviani. 1987. Effects of temperature on muscle adenylic nucleotides of European sea bass {Dicenlrarchus labrax L.). Ital. J. Biochem. 36:2: 136A-138A. Regoli. F., G. W. Winston, V. Mastrangelo. G. Principato & S. Bompadre. 1998. Total oxyradical scavenging capacity in mussel Mytilus sp. as a new index of biological resistance to oxidative stress. Chemosphere 37:14-15.2773-2783. Rheault, R. B. & M. A. Rice. 1996. Food limited growth and condition index in the eastern oyster, Crassotrea virginica (Gmelin 1791), and the bay scallop. Argopeclen irradians irradians (Lamark 1819). J Shellfish Res. 15:2:271-283. Robbins. I., P. Lubet & J. Y. Besnar. 1990. Seasonal variation in the nucleic acid content and RNA:DNA ratio of the gonad of the scallop Pecten maximus. Mar. Biol. 105:191-195. Rogan, E.. S. C. Culloty, T. 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. Aquaculture 97:311-315. Salin, D. 1992. The ammonia toxicity for sturgeon Acipenser haeri: mor- phological, physiological, and metabolic effects of an exposure to sub- lethal and lethal doses. Thesis. Bordeaux University, France. 176 pp. Skjoldal. H. R. & T. Bakke. 1978. Relationship between ATP and energy charge during lethal metabolic stress of the marine isopod Cirolana borealis. J. Biol. Chem. 253:3355-3356. Sprung, M. & J. Borcherding. 1991. Physiological and morphometric changes in Dreissena polymorphs (Mollusca, Bivalvia) during a star- vation period. Malacologia 33:1-2,179-191. Sunila. I. 1991. Respiration of sarcoma cells from the soft shell clam Mya arenaria L under various conditions. J. Exp. Mar. Biol. Ecol. 150:19-29. Thompson. R. J. & B. A. MacDonald. 1991. Physiological interactions and energy partitioning, scallops: biology, ecology, and aquaculture. In: Shumway S. (ed.l. Developments in Aquaculture and Fisheries Science 21:347-376. Tremblay, R. & J. Pellerin-Massicotte. 1997. Effect of tidal cycle in lyso- somal membrane stability in the digestive gland of Mya arenaria and Mytilus edulis L. Comp. Biochem. Physiol. 117:199-104. Tyurin, A. N. 1991. Behavioral reactions of the scallop. Mizuhopecten yessoensis and the mussel. Crenomytilus gray anas, to reduced salinity and oxygen and exposure to synthetic detergents. /. Hydro, biol. 24: 13-19. Vahl. O. 1981. Energy transformations by the Iceland scallop. Chlamys islandica (O. F. Muller) from 70°N.I. the age-specific energy budget and net growth efficiency. J. Exp. Mar. Biol. Ecol. 53:281-296. Veldhuizentsoerkan, M. B., D. A. Holwerda. A. M. T. Debon. A. C. Smaal & D. I. Zandee. 1991. A field study on stress indexes in the sea mussel. Mytilus edulis — application of the stress approach in biomonitoring. Arch. Environ. Contamination Toxicol. 21:497-504. Viarengo. A., L. Canesi. M. Pertica. G. Mancmelli, R. Accomando. A. C. Smaal & M. Orunesu. 1995. Stress on stress response — a simple moni- toring tool in assessment of a general stress syndrome in mussels. Mar. Environ. Res. 39:1-4.245-248. Walne. P. R. 1970. The seasonal variation of meat and glycogen content of seven populations of the oyster Ostrea edulis L. and a review of the literature. Series Invest. Series II XXVLT-33. Journal of Shellfish Research, Vol. IS. No. I. 67-70. 1999. CLONING AND CHARACTERIZATION OF TROPOMYOSIN cDNAs FROM THE SEA SCALLOP PLACOPECTEN MAGELLANICUS (GMELIN, 1791) MOHSIN U. PATWARY,1 MICHAEL REITH,2 AND ELLEN L. KENCHINGTON3 Department of Biology Medgar Evers College of The City University of New York Brooklyn, New York 11225 'Institute for Marine Biosciences, National Research Council of Canada Halifax, Nova Scotia, Canada B3H 3Z1 ^Science Branch. Bedford Institute of Oceanography Department of Fisheries and Oceans Dartmouth. Nova Scotia. Canada B2Y 4A2 ABSTRACT Two different complimentary DNAs (cDNAs) encoding tropomyosin have been characterized from adductor muscle of the sea scallop Placopecten magellanicus. These cDNAs fall into two size classes of approximately 2,540 and 2.030 base pairs with the larger clones containing a longer 3' untranslated region. This difference apparently arises from the utilization of two different polyadenylation signals. All clones are identical in both coding and noncoding regions, indicating that they represent the same gene. Northern analysis indicates that this gene is expressed highly in adductor muscle and at a much lower level in several other tissues. Southern blots indicate a small (1—3) number of tropomyosin genes in the sea scallop, and population studies detect a high degree of individual polymorphism at this locus. KEY WORDS: Placopecten magellanicus, sea scallop. cDNA, tropomyosin INTRODUCTION Tropomyosins are highly conserved, actin-binding proteins present in virtually all eukaryotic cells (see Lees-Miller and Helf- man 1991 for review). Different tropomyosin isoforms are ex- pressed in developmental and tissue-specific patterns and are broadly catagorized into three major classes: nonmuscle (cytoplas- mic), smooth muscle, and striated muscle specific. Tropomyosin mediates Ca2+-dependent actomyosin contraction through its in- teraction with troponins in striated muscle or caldesmon in smooth muscle and nonmuscle cells. In addition to this importance as an essential structural and functional component of the actin mi- crofilament system of the cell, tropomyosins have also been iden- tified as the major protein causing allergic reactions to shrimp (Shanti et al. 1993. Daul et al. 1994, Leung et al. 1994. Witteman ct al. 1994). Like many cytoskeletal proteins, the diversity of tropomyosin isoforms is generated from a few genes through alternative RNA processing or expression from alternate promoters rather than through individual genes for each isoform (Pittenger et al. 1994). In the rat, at least 16 different tropomyosin isoforms have been identified that are encoded by only four genes (Balvay and Fisz- man 1994). The four genes are the a-gene, p-gene, TM-4 gene, and hTMnm gene, each named after a protein they encode (striated muscle a- and B-TM, fibroblast TM-4, and human nonmuscular TM-30. respectively). The a-gene encodes at least nine different isoforms that are generated from two promoters (which results in the use of two different initial exons) and alternative splicing of exons 2. 6, and 9 (two alternate exons are encoded for exons 2 and 6, and four different exons are available for exon 9). The alterna- tive exons have been shown to encode tropomyosin sequences essential for critical interactions with other proteins. For example, the 9a exon of the a-gene, which is only expressed in striated muscle, is required for troponin to mediate high-affinity actin bind- ing (Hammell and Hitchcock-DeGregori 1996). The use of alter- nate promoters and splicing to generate multiple tropomyosin iso- forms has been found in all vertebrates investigated as well as Drosophila (Hanke and Storti 1988). In this communication, we describe the isolation and charac- terization of cDNA clones encoding tropomyosin from sea scallop adductor muscle. This paper contributes to a better understanding of the stucture-function relationship of the tropomyosin gene in bivalves, because there have been no detailed studies previously. We demonstrate that the region surrounding the tropomyosin gene is highly polymorphic in individual scallops and may prove to be an excellent marker for genetic studies of sea scallop. The mo- lecular characterization of tropomyosin cDNA will also be useful to future studies defining the physiological and molecular basis of allergic sensitivity. MATERIALS AND METHODS Sea scallops were obtained from commercial beds near Yarmouth and Sable Island. Nova Scotia and from St. Pierre Bank near Newfoundland, Canada. DNA extraction. cDNA library con- struction and screening, probe preparation for southern blot hy- bridization, and the preparation of genomic blots were as described previously (Patwary et al. 1996). Using a commercial RNA isolation kit (Stratagene). total RNA was extracted from pooled sea scallop adductor muscle, gill, go- nad, heart, liver, and mantle tissues from several individuals that had been snap-frozen in liquid nitrogen immediately after collec- tion and stored at -70°C. All RNAs were further extracted twice with phenolchloroform and once with chloroformisoamyl alcohol (24:1). precipitated with 7.5 m NH4C1 (DEPC-treated) and 2.5 volume ethanol and dissolved in DEPC-treated water. For northern blots, 15 jig of each RNA were electrophoresed on a 0.8% aga- rose-formaldehyde gel according to standard methods (Sambrook 67 68 Patwary et al. 1 ccctttcgagtctctgggagcccggtgtgtgttaggaataaggcagaagtcaggaggctctcgtctgcagttgttcagca 80 81 tttcccttctgcttcacacttcttcttcatctttctatttagataccgttaattctcaaacaacaaa AGT GAT GCT 156 1 M D A 3 15 7 ATC AAG AAG AAG ATG CAG GCC ATG AAG GTC GAC AGG GAG AAT GCC CAG GAC ATG GCC GAA 216 4 IKKKMQAMKVDRENAQDMAE 23 217 CAG ATG GAG CAG AAA TTG AAG GAC ACC GAG ACA GCC AAG GCA AAG TTG GAG GAA GAT TTC 276 24 QMEQKLKDTETAKAKLEEDF 43 2 77 AAC GAA CTC CAG AAG AAG CTC GGC ACC ACC GAA AAC AAC TTT GAT ATA GCC AAC GAA CAA 336 44 NELQKKLGTTENNFDIANEQ 63 33 7 TTG CAG GAA GCT AAT ACC AAG CTC GAA AAC TCA GAC AAA CAG ATC ACC CAG CTA GAA AGT 396 64 LQEANTKLENSDKQITQLES 83 397 GAT GTT GCT GGA CTC CAG AGG AGG CTC CAA CTG CTG GAA GAC GAT TAT GAG CGA TCT GAA 45 6 84 DVAGLQRRLQLLEDDYERSE 103 457 GAG AAG CTT AAC ACA ACA GCA GAG AAA TTG GAA GAG GCA TCC AAA GCT GCA GAT GAG AGT 516 104 EKLNTTAEKLEEASKAADES 123 517 GAG AGA AAT CGC AAG GTG TAT GAA GGC AGG AGT AAC ACT TGT GAG GAG AGG ATT GAT GAG 576 124 ERNRKVYEGRSNTCEERIDE 143 577 CTA GAA AAA CAG TTG GAT ACT GCT AAA ACC ATT GCA ACA GAT GCT GAC TCT AAG TTT GAT 636 144 LEKQLDTAKTIATDADSKFD 163 637 GAG GCC GCC CGT AAG CTT GCT ATT ACA GAA GTG GAC CTT GAG CGC GCC GAG ACT AGG CTG 696 164 EAARKLAITEVDLERAETRL 183 697 GAG GCC GCT GAC GCC AAA GTA CAC GAA CTC GAA GAA GAG CTC ACT GTT GTT GGT TCA AAT 756 184 EAADAKVHELEEELTVVGSN 203 75 7 ATC AAA ACC CTT CAG GTG CAA AAC GAT CAG GCA TCA CAG AGA GAG GAT AGC TAC GAG GAA 816 204 IKTLQVQNDQASQREDSYEE 223 817 ACC ATT AGA GAT CTC ACC AAA AGC CTG AAG GAT GCT GAG AAC AGG GCC ACA GAA GCT GAT 876 224 TIRDLTKSLKDAENRATEAE 243 87 7 AGA CAA GTA GTC AAA CTC CAG AAA GAG GTG GAC AGA CTC GAA GAT GAG CTG CTT GCG GAG 936 244 RQVVKLQKEVDRLEDELLAE 263 937 AAG GAA AGA TAC AAG GCA ATC AGT GAC GAA CTG GAC CAG ACC TTT GCC GAG ATT GCT GGT 996 264 KERYKAISDELDQTFAEIAG 283 997 TAC TAA tgtgtccagcaaggacatattcccctcaccaaatttatatcataaattacgaggaatgacagacaaagaaaa 1074 284 Y * 285 1075 gactgtcaattggaaaaataatattacatcagctttgtacagctatacaaatcgtcgtgcaagaattcgaaacagagaac 1154 1155 ctgccataaaggatccaacatttctttctcaatgtgtgtaatggtacagacgatgaaattccaaatttataaattattat 12 34 1235 taaqaaatqctaatctctatqtqaccttqccqcqatattqcqacccaqcqcqqqaqtqaqaqactaqaqqqcaaqqacqq 1314 1315 qtqqqqtcqaccqqqtccqctttttaacctctctactcttcqtqtatttqtatqtatqttacttttqtqaacqqtttctt 1394 1395 ttcqaacacctqtcaqacttctqcaaaqctactacttctqqqqqqtaaaataatqttcatttctatatttatatacaatq 1474 1475 tatatataactqatacatacatqqattcatatttttcqttattttattatatcatgttatqacatcqqaacacqacacaq 1554 1555 aaqattqtqttatccatqcctqqcqcattctqtacttqaqqctcqgqaaqqcaatqctaqcqqcacatqqttaccattta 1634 1635 aq-taqtaqctacqcaqaqq-tttqqaccaqqcqtacaaaaaccaatcqqqqqtaatattqcqaqatacqacaqtqq-tqqaa 1714 1715 atttqacaqctttaqacatqtaqaqtcqtttataqatacaaqttaqttqtaqqaqtqtaqaqaqactqtaaqtaqtacat 1794 1795 cctqtaqtcctaqaqaqqcqqtcacatctqcccttacattcaaaaqcqacqaccaaqaaattccaqcattcat-tttaacc 1874 1875 accacatqactatattttttttctatttttgttttatataaaaaqacttaatqqcaatatccaqacaacqccattqtqat 1954 1955 qattttttttcaaaqaaaaaactaaaaqctttaaatt-tccacqqcttctqq-tcctgcccctatttaaataaaqaaqqtql: 2034 * 20 35 atqtactatgtctttggaatgatttattttgcatgttgtttgtgtatagaagaatgtgttgtatggactacaaacaaaac 2114 2115 gtagctggctatattattaattgaaaaacaaatttatacattttccttcacagattatttaccttatatattatactttt 2194 2195 gattgagaattgatttttctcatctataaaatatccttattatggtacgaaatttttatcactatacatatatgtgtaga 22 74 2275 cacagataaagagaacttttgtaggtgtactaatttcgtggacattgttcatgttacatttgccatgtgaccaagactag 2354 2 355 ttagctgtacaggagggaaaagcgagaactattttgtcatgtgacaactgtagacggaagactctagtgtttgtcatgtg 24 34 24 35 actgtcatgtgacaactgtagacggaagactcctagtattcctcacaattcagcttccattgcatttccctggatattgc 2514 * 2515 caatttgttttaaccaaca a t a aacttgtattgcttacaaaaaaaaaaaaaaaaaaaaaaaaaaaaa 25 80 Figure 1. Nucleotide sequence and the derived amino acid sequence for sea scallop tropomyosin (nucleotides and amino acids follow standard abbreviations). The nucleotide residues are numbered from the 5' end of clone PmC 128. The amino acid residues are numbered from first in-frame methionine (M). The polyadenylation signals are in bold. The stars indicate the end of transcripts attributable to polvadenylation at different sites. Sequence for the 3' UTR probe is underlined. Tropomyosin cdna in Sea Scallops 69 Kb 4.40 2.37 . 1.35 abed e f g Kb M A B « Figure 2. Northern analysis of sea scallop tropomyosin cDNA. Fifteen ug of total RNA in each lane was elect rophoresed on a formaldehyde agarose gel, blotted to a nylon membrane, and probed with a ,2P- labeled 3'UTR probe. Source of RNAs are from adductor muscle (lanes a & b), gonad (lane c), heart (lane d). liver (lane e), mantle (lane f), and gill (lane g). The membrane was exposed for one hour at -70 C with an intensifying screen. 2.02- I Figure 4. Polymorphisms in the region surrounding the sea scallop tropomyosin gene. Each lane contains Hae Ill-digested DNA (10 ug) from a different animal hybridized with a tropomyosin cDNA coding region probe. DNAs in panel A lanes are from Yarmouth, Nova Scotia, panel B lanes are from Sable Island, Nova Scotia, and panel C lanes are from Newfoundland, Canada. M is digoxigenin-laheled DNA mo- lecular-weight marker (Boehringer Mannheim). el ul. 1989) and transferred to a nylon membrane (Boehringer Mannheim) with a Pharmacia Vacu-Gene XL unit following the manufacturer's protocol No 4. Tropomyosin cDNA probes were amplified from a plasmid clone by the polymerase chain reaction (PCR) using two nested primers. Amplification products were visualized on agarose gels, and the PCR products were excised and purified using a QIAquick Spin PCR Purification Kit (Qiagen Inc). PCR products (25-50 ng) were labeled by random priming with 32P-dCTP(3000 Ci/mmol) using a Ready-to-go labeling kit (Pharmacia Biotech), and the kb abode 2.69- 1.4 0.42- Figure 3. Southern hybridization of the sea scallop tropomyosin cod- ing region to genomic DNA. The blot contains DNA from a single animal digested with EcoR I (lane a), EcoR V (lane b). Hind III (lane c), Sal I (lane d), and Xba I (lane e). unincorporated nucleotides were removed using Nick columns (Pharmacia Biotech). Prehybridization and hybridization of northern blots were car- ried out in a hybridization oven at either 55 or 65°C in 15 mL of hybridization buffer (0.25 m Na,HP04, pH 7.2. 7% SDS, 50 mg/ mL sheared, denatured salmon sperm DNA). Hybridization was for 24-36 h with 1-2 x 106 cpm denatured probe. The membranes were then washed twice for 40-50 min each in 20 mm Na,HP04. pH 7.2, 5% SDS and twice for 35-45 min each in 20 mm Na2HP04. pH 7.2, 1% SDS at the same temperature used for hybridization. To characterize the major transcripts of scallop adductor muscle, 130 plaques from an adductor muscle cDNA library were randomly selected and sequenced on each end. About 4% of these clones were identified as encoding tropomyosin. Inserts from five clones (PmC 60, PmC 92. PmC 104, PmC 118. PmC 128) were subcloned and sequenced completely in both directions. RESULTS AND DISCUSSION The five DNAs characterized by sequencing (PmC 60, PmC 92, PmC 104, PmC 1 18, PmC 128) can be divided into two groups on the basis of size. PmC 60. PmC 104, and PmC 1 18 are approxi- mately 2.030 base pairs (bp) in length (excluding the polyA tail), and PmC 90 and PmC 128 are 2,531 and 2,546 bp. respectively. The slight differences in length within each group are caused by incomplete first strand cDNA synthesis; whereas, the difference between the two groups is caused by different lengths of the 3' untranslated region (3'-UTR). The two larger clones have a 1,550 nucleotide 3'-UTR. and that of the shorter clones is 1.036 nucle- otides. The difference in the 3'-UTR region seems to arise from polyadenylation at two different sites, with the shorter clones re- sulting from the recognition of a polyadenylation signal (AATAAA) at position 2020, whereas, the longer clones result from utilization of the polyadenylation signal at 2533 (Fig. 1 ). The use of alternative polyadenylation sites has been found previously in tropomyosins from a wide range of organisms (Balvay and Fiszman 1994). All five cDNA clones represent transcripts from the same tro- 70 Patwary et al. pomyosin gene, because they are identical in both their coding and noncoding nucleotide sequences. The cDNAs encode an open reading frame of 284 amino acid residues, with a predicted mo- lecular mass of 30.280 d. Sea scallop tropomyosin is approxi- mately 70% identical to other molluscan tropomyosins. 60% iden- tical to those of flukes. 55% identical to those of insects and worms, and 52% identical to vertebrate tropomyosins: the tro- pomyosin cDNA described herein may be useful as a heterologous probe. The observation that all five cDNA clones encode the same protein suggests that in adductor tissue, which contains both stri- ated and smooth muscle (Chantler 1991), both muscle types ex- press this tropomyosin isoform. Northern hybridization with a 3'-noncoding region probe to total RNA from several sea scallop tissues revealed intense signals only in the adductor muscle lanes (Fig. 2). Two bands consistent with the sizes of the two cDNAs that were isolated were seen, with the smaller, approximately 2.1 kilobase (kb) band present in greater abundance. Upon longer exposure, faint signals were also detected in lanes from other tissues (results not shown), indicating that the gene represented by this cDNA is also expressed in many scallop tissues. The cDNA we report here seems to represent the principal tropomyosin gene expressed in sea scallop adductor muscle. The number of genes encoding tropomyosin in sea scallop was estimated by southern hybridization with a probe covering most of the coding region (Fig. 3). The results indicate that there are only a few (1-3) tropomyosin genes in sea scallop, and similar results were obtained with a shorter probe from the 5' end of the coding region (not shown). As a part of an effort to develop DNA-based genetic markers to conduct population genetic studies on sea scallop (Patwary et al. 1994a. Patwary et al. 1994b. Patwary et al. 1996). we examined the utility of tropomyosin cDNA as a probe to reveal polymorphisms. A genomic blot containing Hae Ill-digested DNAs from 12 sea scallops from three distant locations was probed with a tropomyo- sin coding region probe. The probe revealed a highly polymorphic locus with a total of six alleles (Fig. 4). Although heterozygote deficiency has been reported to be common in sea scallop popu- lation (Foltz and Zouros 1984. Beaumont and Zouros 1991). this locus is highly heterozygous. Although the polymorphisms re- vealed here are limited by a small sample size, the tropomyosin probe seems to be a useful marker for various genetic studies in sea scallop. ACKNOWLEDGMENTS This project was supported in part by funds from the Depart- ment of Fisheries and Oceans. Canada through a contract to the NRC Institute for Marine Biosciences and in part by an NSERC Strategic Grant to E.K. and Prof. E. Zouros. Dalhousie University, Canada. LITERATURE CITED Balvay. L & M. Y. Fiszman. 1994. Analyse de la diversite des isoformes de tropomyosine. C. R. Soc. Biol. 18:527-540. Beaumont. A. R. & E. Zouros. 1991. Genetics of sea scallops. In: S. E. Shumway (ed.). Scallops: Biology. Ecology, and Aquaculture. Elsevier, Amsterdam. Chantler, P. D. 1991. The structure and function of scallop adductor muscles. In: S. E. Shumway (ed.). Scallops: Biology. Ecology, and Aquaculture. Elsevier. Amsterdam. Daul, C. B.. M. Slattery. G. Reese & S. B. Lehrer. 1994. Identification of the major brown shrimp (Penaeus aztecus) allergen as the muscle pro- tein tropomyosin. Int. Arch Allergy Immunol. 105:49-55. Foltz. D. W. & E. Zouros. 1984. Enzyme heterozygosity in the scallop Placopecten magellanicus (Gmelin) in relation to age and size. Mar. Biol. Lett. 5:255-263. Hammell. R. L. & S. E. Hitchcock-DeGregori. 1996. Mapping the func- tional domains within the carboxyl terminus of a-tropomyosin encoded by the alternatively spliced ninth exon. J. Biol. Chem. 271:4236—1242. Hanke, P. D. & R. V. Storti. 1988. The Drosphila melanogaster tropomyo- sin II gene produces multiple proteins by use of alternative tissue- specific promoters and alternative splicing. Mol. Cell. Biol. 8:3591- 3602. Lees-Miller, J. P. & D. M. Helfman. 1991. The molecular basis for tro- pomyosin isoform diversity. BioEssays 13:429—437. Leung. P. S. C. K. H. Chu, W. K. Chow. A. Ansari. C. 1. Bandea. H. S. Kwan. S. M. Nagy & M. E. Gershwin. 1994. Cloning, expression, and primary structure of Metapenaeus ensis tropomyosin, the major heat- stable shrimp allergen. J. Allergy Clin. Immunol. 94:882-890. Patwary. M. U.. E. L. Kenchington, C. J. Bird & E. Zouros. 1994a. The use of random amplified polymorphic DNA markers in genetic studies of the sea scallop Placopecten magellanicus (Gmelin, 1791). J. Shellfish Res. 13:547-553. Patwary. M. U.. R. M. Ball. C. J. Bird. B. Gjetvaj. S. Sperker, E. L. Kench- ington & E. Zouros. 1994b. Genetic markers in sea scallop and their application in aquaculture. Bull. Aquacult. Assoc. Can. 2:18-20. Patwary, M. U.. M. Reith & E. L. Kenchington. 1996. Isolation and char- acterization of a cDNA encoding an actin gene from sea scallop (Pla- copecten magellanicus). J. Shellfish Res. 15:265-270. Pittenger. M. F.. J. A. Kazzaz & D. M. Helfman. 1994. Functional prop- erties of nonmuscle tropomyosin isoforms. Curr. Opinion Cell Biol. 6:96-104. Sambrook. J.. E. F. Fritsch & T. Maniatis. 1989. Molecular cloning — a laboratory manual. Cold Spring Harbor Laboratory Press. New York. Shanti, K. N., B. M. Marin. S. Nagpal. D. D. Metcalfe & P. V. S. Rao. 1993. Identification of tropomyosin as the major shrimp allergen and characterization of its IgE-binding epitopes. J. Immunol. 151:5354- 5363. Witteman. A. M., J. H. Akkerdaas. J. V. Leeuwen. J. S. van der Zee & R. C. Aalberse. 1994. Identification of a cross-reactive allergen (pre- sumably tropomyosin) in shrimp, mite, and insects. Int. Arch Allergy Immunol. 105:56-61. Journal of Shellfish Research, Vol. IS. No. 1. 71-76, 1999. GROWTH CHARACTERISTICS OF CHLAMYS FARRER1 AND ITS RELATION WITH ENVIRONMENTAL FACTORS IN INTENSIVE RAFT-CULTURE AREAS OF SISHILIWAN BAY, YANTAI HONGSHENG YANG, TAO ZHANG, JIAN WANG, PING WANG, YICHAO HE, AND FUSUI ZHANG Institute of Oceanology Chinese Academy of Sciences Qingdao 266071 Peoples Republic of China ABSTRACT The growth characteristics of the scallop Chlamys farreri under intensive raft-culture and its relationships with major environmental factors were studied. The experiment was conducted from May 1997 to April 1998 in four farming areas of Sishiliwan Bay, Yantai, China. The instantaneous growth rates of shell height, wet weight, fresh and dried soft tissue of scallops were measured and calculated during the course of this study. The results showed obvious seasonal variation in the growth in Sishiliwan Bay. The main factors affecting the growth rate of the scallops were water temperature and food supply in the farming areas. The growth rates of the scallops cultured in Jinggouwan and Yueliangwan areas were faster than that in Kongtongdao and Qingqiangzhai areas. The rela- tionships between the instantaneous growth rates in dried tissue weight of C. farreri with water temperature, and particulate organic matter (POM) in Jinggouwan area and Yueliangwan area were simulated. Growth rate declined when water temperature was below 5°C. Between 5 and 23°C, growth rate increased with the increasing of water temperature. Growth rate sharply declined when water temperature was above 23°C. The scallops stopped growing when POM was less than 0.90 mg/L and grew rapidly with increasing POM. When POM was above 3.67 mg/L. the growth rate of the scallops decreased again. KEY WORDS: Sishiliwan Bay, C. farreri. instantaneous growth rate, intensive raft-culture, temperature, particulate organic matter, scallop, manculture INTRODUCTION The scallop C. farreri is the main cultured species over the coastal farming areas in the northern China Sea. In the recent decade, the growth rate of C. farreri sharply declined because of high culture density and exhaustion of food supply. Furthermore, spats used for farming were mostly collected from the recruitment reproduced by the cultured and possibly inbred stock in recent years. Mass mortality of this species occurred in most of the farm- ing areas in the northern China Sea, especially during the summer and the autumn 1997, and the mortality rate was above 60.0%. The exact cause for the mortality remains unknown. Therefore, it is necessary to study the growth characteristics of cultured scallops systematically among the different farming areas, and its relation- ship with the major environmental factors, such as water tempera- ture and availability of natural food supply. The main factors influencing growth are water temperature and the amount of food ingested. From feeding experiments carried out by Winter and Langton (1976) with Mytilus edulis, it is ob- vious that with increasing amounts of food available, there is an increase in growth rate. Since 1970s, the growth and its relation- ship with the main environmental factors have been studied in detail, in many pectinids, including Amusium japonicum balloti (Williams and Dredge 1981), Argopecten irradians (Briceli et al. 1987, Cahalan et al. 1989, Duggan 1972. Kirby-Smith and Barber 1974. Rhodes and Widman 1984, Zhang et al. 1987, Zhang et al. 1991a, Zhang et al. 1991b). Chlamys islandica (Vahl 1980), Chlamys opercularis (Broom and Mason 1978. Taylor and Venn 1978), Chlamys varia (Conan and Shafee 1978, Shafee 1980), Pecten alba (Gwyther and Mcshane 1988), Patinopecten caurinus (Haynes and Hitz 1971), Placopecten magellanicus (Macdonald 1986, Shumway et al. 1987). and Pecten maximus (Mason 1970). MATERIALS AND METHODS Sishiliwan Bay, Yantai, and its near sea areas, located on near Yantai city ( 121°20'-40'E, 37°25'-40'N) including Zhifuwan Bay, Jinggouwan Bay, and Sishiliwan Bay, is 26 km in width, and 13 km farthest from the shore. It is ear-shaped and half enclosed. The mouth of the bay faces eastward and is divided into two parts by Kongtondao Island. The smaller one is between Zhifudao Island and Kongtondao Island, and the larger one between Kongtongdao Island and Yangmadao Island (See Fig. 1). The bay has a muddy and sandy bottom. The total area is about 13,000 ha. and the depth is about 9-15 meters. Sishiliwan Bay is one of the earliest farming areas, where the kelp Laminaria japonica raft-culture was developed in 1949. At the end of 1960s, researchers from the Institute of Oceanology, Chinese Academy of Sciences, and other institutions had studied the artificial collection of mussel M. edulis spats and tested raft- culture techniques. The mariculture farms of Yantai had also col- lected C. farreri spats in Jinggouwan area in the middle of 1970s. Between the end of 1970s and the early 1980s, hatchery production of C. farreri seed was successfully developed and used, and the scallop was cultured in large scale. The northern bay scallop Ar- gopecten irradians has been one of the main species cultured in the bay since 1986. Now, M. edulis. C. farreri. A. irradians. and the kelp Laminaria japonica are the primary species under raft-culture in the bay. The culture areas include 830 ha ( 1 ha = 6.000 cages, the same as for mussels) for scallops, 460 ha for mussels, 250 ha ( 1 ha = 6000 strings) for kelps. Mussel and scallop seeds are mainly collected from the wild, and the spats of the bay scallop and the seedling of the kelp are hatchery-produced. Set-Up of Research Stations Four farming areas, Jinggouwan farming area, Yueliangwan farming area. Kongtongdao farming area, and Quingquanzhai 71 72 Yang et al. 1.4 Figure 1. Main farming areas in the Sishiiwan Bay. Yantai. A: Jing- gouwan, B: Yueliangwan, C: Kongtongdao, and I): Qingquanzhai; 1: Zhifudao Island, 2: Kongtongdao Island, 3: Yangmadao Island. farming area in Sishiliwan Bay were chosen in this study. In each farming area, three stations were set, totaling 12 stations. Sampling The specimens sampled in May to September 1997. were 1-year scallops, which were collected in the spring of 1996, and those sampled in October 1997 to April, 1998 were 1-year scallops collected in the spring of 1997. Fifty scallops were collected monthly at each station. After being taken back to the laboratory, the specimens were cleaned (the epibionts were cut off) and boiled, the adductor muscle and the viscera were divided, and the shell height, wet weight, shell weight, wet and dried weight of soft tissue (65°C. for 48 h) were measured. Water temperature, salinity, the biomass of seston and particulate organic matter (POM) were determined at the same time. The seston was filtered by GF/C and dried at 65°C for 48 h and weighed, and then washed at 450°C for 6 h and weighed again. The biomass of POM equals the difference between the weight of dried sestons minus that of ash. Calculation The instantaneous growth rates of the shell height, wet weight, the fresh and dried weight of the soft tissues were calculated using the following equation IGR = ((Ins, - Ins, )/t) x 100. IGR stands for instantaneous growth rate, s, stands for initial shell height, wet weight, the fresh or dried weight of the soft tissues measured first 1.2 1 ia> u 0.8 u sz 06 o a. o 0.4 -D — Jinggouwan area -• — Yueliangwan area :v*: > • S-0 N-D F-M A-M Month J-J A-S Figure 3. Annual variations of instantaneous growth rate in shell height of ('. farreri in Jinggouwan area and Yueliangwan area. time, s2 for ending shell height, wet weight, the fresh or dried weight of the soft tissues, and t for the interval days between initial and ending measurement. RESULTS AND ANALYSIS Annual Variation of the Growth of C. farreri in Relation to Water Temperature The surface water temperature and salinity variations of Sishi- liwan Bay. Yantai are illustrated in Figure 2. The lowest and highest water temperatures are observed in February and August, respectively. The annual fluctuation of salinity is small, around 29.95 ± 0.51. The annual variations of instantaneous growth rates of shell height, wet weight with shell, fresh and dried weight of soft tissue in Jinggouwan and Yueliangwan farming area are shown in Figures 3. 4, 5. and 6. It is obvious that the growth of C. farreri in two areas varies with the season. The growth rate of the scallops is fastest from May to July. The relationships between the instantaneous growth rate of soft tissue and water temperature are similar at these two areas (Fig. 7). The regression equations are as follows. Jinggouwan farming area IGR = - 0.003 IT1 + 0.091 IT2 - 0.4568T + 0.6218, r = 0.9788 Yueliangwan fanning area IGR = - 0.0027T3 + 0.0780T2 - 0.3801T + 0.4635. r : 0.9916 J F M A M J J Month A S O N D Z 3.5 3 2.5 2 1.5 I 0.5 0 S-0 Jinggouwan area Yueliangwan area N-D F-M A- Month ■M J-J A-S Figure 2. Annual variation of water temperature and salinity in the Figure 4. Annual variations of instantaneous growth rate in wet Sishiiwan Bay, Yantai. weight of C. farreri in Jinggouwan area and Yueliangwan area. Growth Characteristics of C. farreri 73 s-o N-D F^Vl A~M Month J-J A-S Figure 5. Annual variations of instantaneous growth rate in fresh soft weight of C. farreri in Jinggouwan area and Yueliangwan area. ay//' ^^x / ^\ / ° \ D Jinggouwan area /a 0 Yitliangwanarea o\ 8^ a/ ) 5 10 15 20 25 Temperature(°C) Figure 7. The relationship between water temperature and the instan- taneous growth rate of dried soft tissue of C. farreri in Jinggouwan and Yueliangwan areas. IGR stands for the instantaneous growth rate of soft tissue and T for water temperature (°C) These equations clearly show that there is a strong correlation between the growth of C. farreri and water temperature. The growth rate is slow when water temperature is below 5°C, and then increases sharply with increasing temperature. The fastest growth rate is at 16-18°C. The growth rate sharply declined when water temperature was over 23°C. Variation Among Different Culture Areas Results from the main growth period. May to September, show that there are differences in instantaneous growth rate of shell height, wet weight, fresh and dried weight of soft tissue in the four culture areas (Figs. 8, 9, 10, 1 1). especially from May to August. The dried weight of soft tissue in Yueliangwan and Jinggouwan areas increases fastest from May to July, and the growth rate in these two areas is faster than those in Kontongdao and Qingquan- zhai areas. The growth rate of C. farreri in Kontongdao and Qingquanzhai areas is slow, and their instantaneous growth rate varies little. That is quite different from the scallops cultured in Yueliangwan and Jinggouwan areas, where the growth rate is rela- tively faster in May to July, and gradually declines after that pe- riod. It is necessary to note that scallops in these four farming areas mostly died by September 18. 1997, up to 80.0%. The instanta- neous growth rate of the survivors decreased considerably, espe- cially the instantaneous growth rate of dried soft tissue. Relationship Between the Growth and POM Biomass The measurements of the biomass of seston and POM are listed in Table 1. The differences in the biomass of seston and POM are obvious among different culture areas. The relationship between the instantaneous growth rate of dried soft tissue and the biomass of POM can be formulated as IGR = -0.5695[POM]: + 4.1780[POM] - 3.4031. r = 0.9338 (Fig. 12). It shows that the instantaneous growth rate of dried soft tissue of C. farreri tends to be zero when the biomass of POM is less than 0.90 mg/L and increases with the increasing POM biomass. The scallop growth rate declines when the POM is more than 3.67mg/ L. It is clear that the growth of C. farreri is limited in some degree by the abundance of natural foods in the farming area. DISCUSSION Pectinids can be partitioned into four broad groups according to their patterns of life history (Orensanz et al. 1991 ): ( 1 ) long-lived, iteroparous species; (2) short-lived, iteroparous species: (3) short- lived, semelparous or quasi-semelparous species; and (4) small- sized, presumably short-lived, brooding species. The first group o o oi O 3.5 3 2.5 2 Jinggouwan area Yueliangwan area _J N-D F-M A- Month M J-J A-S OB J3 O 1.6 1.2 0.8 0.4 M-J ---A -B a -D „ ■ "~~"~~4 J-J Month J-A A-S Figure 8. The instantaneous growth rate in shell height of C. farreri in Figure 6. Annual variations of instantaneous growth rate in dried soft the culture areas. A: Jinggouwan, B: Yueliangwan, C: Kongtongdao, weight of C. farreri in Jinggouwan area and Yueliangwan area. and D: Qingquanzhai. 74 Yang et al. m-j J-J J-A A-S Month Figure 9. The instantaneous growth rate in wet weight of C. farreri in the culture areas. A: Jinggouwan, B: Yueliangwan. C: Kongtongdao, and I): Qingquanzhai. can be further divided into two types: large-sized (above 100 mm), relatively long-lived (maximum longevity usually above 12 years) species, and medium-sized (60-100 mm) species with maximum longevity usually less than 10 years. The scallop C. farreri belongs to the second type of the first groups. There are a lot of environ- mental factors influencing the growth of scallops, mainly being water temperature, water current, the biomass of natural food, culture density, and the amount of other filter-feeding animals within the farming area (Zhang et al. 1987. Zhang et al. 1991a, Zhang et al. 1991b). Water temperature and the biomass of natural food in culture area might be the most important factors affecting the growth of C. farreri. In most previous studies (Lou 1991. Wang et al. 1993. Zhang et al. 1956). the shell height of C. farreri was used to measure the growth of the scallops. Their results reflected that. C. farreri grow fast in the months in which water temperature is high, and vice versa in normal culture conditions. In the winter. C. farreri totally stop growing. From March in each year, the growth of C. farreri increases gradually with the increasing water temperature and reaches its peak in July. When water temperature is above 25°C, the growth obviously decreases. From January to March, the water o et Z Month Figure 11. The instantaneous growth rate in dried soft weight of C. farreri in the culture areas. A: Jinggouwan, B: Vueliangwan, C: Kong- tongdao, and D: Qingquanzhai. temperature is lower than 5°C, the growth of shell is nearly zero. In this paper, the instantaneous growth rate was used for the first time to describe the growth of this species, and the resulting model with the instantaneous growth rate and water temperature support the viewpoints above. A combination of many environmental fac- tors might have strongly affected the growth of C. farreri, leading to the death of most C. farreri in this area in 1997. The results of this study differ from previous studies on the influence of high temperature on the growth of C. farreri: when temperature is over 23°C, the instantaneous growth rate of dried weight of soft tissue of C. farreri sharply declines. Our findings indicate that the growth of C. farreri varies dif- ferently with different areas in the same season. The environmental differences of farming areas include the difference in water cur- rent, concentration of nutrients, primary production, and stocking density. The variation in water quality and food supply, especially the food supply, has an obvious influence on the growth of C. farreri. The biomass of POM is higher in the Yueliangwan and Jinggouwan areas than that in the Kongtondao and Qingquanzhai fanning areas. The Yueliangwan area lies upstream to the Sishi- liwan Bay (following the direction of water current), and it is the at o Month Figure 10. The instantaneous growth rate in fresh soft weight of C. farreri in the culture areas. A: Jinggouwan. B: Vueliangwan, C: Kong- tongdao, and D: Qingquanzhai. 4.5 * 3.5 3 2.5 Pi O 2 - 1.5 - 1 < 0.5 0 0 J 0<5> O 0 0.5 1 1.5 4.5 5 2 2.5 3 3.5 POMbioinass(mg/l) Figure 12. The relationship between the instantaneous growth rate in dried soft weight of C. farreri and the biomass of POM. Growth Characteristics of C. farreri 75 TABLE 1. Biomass of seston (S, mg/L) and particulate organic matter (POM, mg/L) in the farming areas of Sishiliwan Bay. Jinggouwan Yueliangw an Kongtongdao Qingqianzhai Sampling Date S POM S POM S POM S POM May 15 11.28 2.70 8.55 4.72 4.47 0.99 13.80 2.28 June 18 3.66 2.17 6.26 4.69 3.11 1.52 3.03 1.70 July 15 2.07 1.15 3.46 2.19 2.54 1.29 2.30 1 .03 Aug. 20 2.12 1.24 2.84 1.57 1.97 1.11 2.18 1.02 Sept. 18 3.48 1 .05 3.64 1 .53 4.37 2.03 3.93 1.07 front part of the whole culture area (Yantai Port is above it). The Jinggouwan farming area is downstream of it. and the Qingquan- zhai farming area is at the bottom of the bay. and water current in the Kongtongdao farming area is slow, and the culture density is highest. The modeling results of this study suggest that, when the bio- mass of POM is lower than 0.90 mg/L, the instantaneous growth rate of dried soft tissue tends to be zero; the growth rate increases with the increasing biomass of POM, and then the growth rate declines as the biomass of POM is above 3.67 mg/L. Roland and Brown (1990) established the relationships between POM and the growth of the Crassostrea gigas. which is similar to ours, except that the upper limit of POM that restricts the growth of C. farreri is 5.00 mg/L. The metabolism of C. farreri increases with the increasing temperature from 10°C to 23°C (Yang et al. 1998). The depiction of natural foods and the food selection of C. farreri (Wang et al. 1989) make the energy consumed by the scallops for metabolism, rather than for growth. The overabundance of natural foods limits the growth of filtering-food bivalves, because of the increased mucus excretion and the pseudofeces production of scal- lops, which consume large amounts of energy and lead to the discharge of carbon and nitrogen of the scallops. ACKNOWLEDGMENTS This work was supported by the National Commission of Sci- ence and Technology of China, Grant No. 96-922-02-04. and by Chinese Academy of sciences. Grant No. KZ951-A 1-102-02. Con- tribution No. 3501 from the Institute of Oceanology, Chinese Academy of Sciences. LITERATURE CITED Broom. M. J. & J. Mason. 1978. Growth and spawning in the pectinid Chlamys opercularis in relation to temperature and phytoplankton con- centration. Mar. Biol. 47:277-285. Bricelj, V. M. & S. E. Shumway. 1991. Physiology: energy acquisition and utilization, pp. 305-346. In: S. E. Shumway (ed.). Scallops: Biology. Ecology, and Aquaculture. Elsevier. Amsterdam. Bricelj, V. M. J. Epp & R. E. Malouf. 1987. Intraspecific variation in reproductive and somatic growth cycles of bay scallops Argopecten irradians irradians (Lamarck): mortality, growth, and oxygen con- sumption. J. Exp. Mar. Biol. Ecol. 112:373-397. Cahalan J. A., S. E. Siddall & M. W. Luchenbach. 1989. Effects of flow velocity, food concentration, and particle flux on growth rates of ju- venile bay scallops Argopecten irradians. J. Exp. Mar. Biol. Ecol. 129:45-60. Conan, G. & M. S. Shafee. 1978. Growth and biannual recruitment of the black scallop Chlamys varia (L.) in Lanveoc area. Bay of Brest. J. Exp. Mar. Biol. Ecol. 35:59-71. Duggan. W. P. 1972. Growth and survival of the bay scallop. Argopecten irradians, at various locations in the water column and at various densities. Proc. Natl. Shellfish Assoc. 63:68-71. Gwyther. G & P. E. Mcshane. 1988. Growth rate and natural mortality of the scallops Pecten alba Tate in Port Phillip Bay, Australia, and evi- dence for changes in growth rate after 20-year period. Fish. Res. 6: 113-131. Haynes, E. B. & C. R. Hitz. 1971. Age and growth of the giant Pacific sea scallop. Patinopecten caurinus. from the Strait of Georgia and outer Washington coast. J. Fish. Res. Bd. Can. 28:1335-1341. Kirby-Smith. W. W. & R. T. Barber. 1974. Suspension feeding aquaculture systems: effects of phytoplankton concentration and temperature on growth of the bay scallop. Aquaculture 3:135-145. Lou, Y. 1991. China, pp. 809-824. In: S. E. Shumway (ed.). Scallops: Biology. Ecology, and Aquaculture. Elsevier. Amsterdam. MacDonald. B. A. 1986. Production and resource partitioning in the giant scallop Placopecten magellanicus grown on the bottom and in sus- pended culture. Mar. Ecol. Prog. Ser. 34:79-86. Mason, J. 1970. The age and growth of the scallop. Pecten maximus L„ in Manx waters. J. Mar. Biol. Assoc. U.K. 37:653-671. Orensanz, J. M., A. M. Parma & O. O. Iribarne. 1991. Population dynamics and management of natural stocks, pp. 625-713. In: S. E. Shunway (ed.). Scallops: Biology. Ecology, and Aquaculture. Elsevier. Amster- dam. Rhodes. E. W. & J. C. Widman. 1984. Density-dependent growth of the bay scallop, Argopecten irradians Indians, in suspension culture. ICES CM 1984/k:18. 12 pp. Roland. W. G & J. R. Brown. 1990. Production model for suspended culture of the Pacific oyster, Crassostrea gigas. Aquaculture 87:35-52. Shafee. M. S. 1980. Application of some growth models to the black scal- lop. Chlamys varia (L.) from Lanveoc, Bay of Brest. J. Exp. Mar. Biol. Ecol. 43:237-250. Shumway S. E.. R. Selvin & D. E. Schick. 1987. Food resources related lo habitat in the scallop. Placopecten magellanicus (Gmelin. 1791): a qualitative study. J. Shellfish Res. 6:89-95. Taylor, A. C. & T. J. Venn. 1978. Growth of the queen scallop. Chlamys opercularis, from de Clyde Sea area. J. Mar. Biol. Soc. 54:9-28. Vahl, O. 1980. Seasonal variation in seston and in the growth rate of the Iceland scallop. Chlamys islandica (O. F. Muller) from Balsfjord. 70°N. J. Exp. Mar. Bio. Ecol. 48:195-204. Wang. R.. X. Lan & L. Liu. 1989. Analysis of the scallop Chlamys farreri food. J. Ocean Univ. Qingduo 19:12-18. Wang, R., Z. Wang & J. Zhang. 1993. Molluscan mariculture. Ocean University of Qingdao Press. Qingdao. China, pp. 155-204. 76 Yang et al. Winter. J. E. & R. W. Langton. 1976. Feeding experiments with Mytitm editlis L. at small laboratory scale. I. The influence of the total amount of food ingested and food concentration on growth, pp. 565-581. In: G. Persoone and E. Jaspers (eds. ). Proceedings of the 10lh European Sym- posium on Marine Biology Ostend. Belgium. Sept. 17-23, 1975. Uni- versa Press. Wetteren. Williams, M. J. & M. C. L. Dredge. 1981. Growth of the saucer scallop Amusium japonicum balloti Bernardi in central eastern Queensland. Aust. J. Mar. Freshwater Res. 32:657-666. Zhang. X., Z. Qi & J. Li. 1956. Observation on the reproduction and growth of scallop. Chlamys farreri. J. Zool. China 8:235-253. Zhang. F.. Y. He & J. Ma. 1987. A repon on the experimental cultivation of bay scallop and kelp by turns. Marine Sci. 6:1-6. Zhang. F.. Y. He & X. Liu. 1991a. Growth and mortality of bay scallop Argopecten irradians cultured at various water layers in Jiaozhou Bay. J. Fisheries China 15:42^17. Zhang. F., J. Ma & Y. He 1991b. Growth and mortality of bay scallop Argopecten irradians cultured in different vessels and different density in Jiaozhou Bay. Mar. Sci. 2:1-2. Yang, H.. T. Zhang, Wang P. et al. 1998. Effects of temperature on oxygen consumption and ammonia-N excretion of Chlamys farreri. Chinese J. Oceanol. Limnol. 16:167-172. Journal of Shellfish Research. Vol. IS, No. 1. 77-83. 1999. CULTURE OF MERCENAR1A MERCENARIA (LINNAEUS): EFFECTS OF DENSITY, PREDATOR EXCLUSION DEVICE, AND BAG INVERSION EVA M. FERNANDEZ,1 JUNDA UN,1 AND JOHN SCARPA2 'Florida Institute of Technology Melbourne, Florida 32901 'Harbor Branch Oceanographic Institution, Inc. Fort Pierce, Florida 34946 ABSTRACT Growth, survival, and condition index (CI) of the northern quahog, Mercenaria mercemiria (Linnaeus, 1758), cultured in nylon mesh bags ( 1.2 x 1.2 m) were assessed against density and predator exclusion device (PED: Vexar net with 2.5-cm openings) in the northern Indian River Lagoon at Oak Hill. Florida. Nursery seed [mean ± SD: 6.0 ± 0.8 mm shell length (SL)] were stocked in February 1997 at densities of 7.500 (5.210). 10.000 (6.944), and 12,500 (8,680) clams/bag (clams/nr) (n = 4) and monitored until the end of May 1997. Two replicates of each treatment were inverted 5 weeks before harvesting to smother fouling organisms and examine their influence on growth. Growout seed (mean ± SD: 21.1 ± 1.7 mm SL) were stocked in October 1996 at densities of 750 (521 ), 1.000 (694). and 1.250 (868) clams/bag (clams/irr) (n = 4) and monitored until early June 1997. At the end of the nursery seed experiment, the average final SL of clams was significantly different among the density treatments (p = .03) and not significantly different between the PED (p = .31) treatments. Nursery seed in the inverted bags were significantly larger (p = .03). and a higher percentage of them reached growout seed size ( 12 mm in SL). Density (p = .60) did not have a significant effect on survival; whereas, the bags with PED had significantly (p = .005) lower survivorship than that of the bags without PED. Density (p = .15) and PED (p = .79) did not significantly affect mean CI at the end of the study, but inversion significantly (p = .002) increased CI. At the end of the growout seed experiment, SL was not significantly different among the treatments (density, p = .40; PED. p = .17). There was a significant (p = .04) effect of density on percentage of the seed that reached legal harvest size (16 mm in shell thickness). In general, percentage of seed that reached harvest size decreased with increasing density. The effects of density (p = .04) and PED (p = .0009) on survival were significant, but there was no apparent pattern. Density (p = .29) and PED (p = .88) did not affect mean final CI. Chlorophyll a concentration and water current speed measured in April and May. 1997 indicated that food was not a limiting factor on growth of the northern quahog at the study site. Our recommendations for northern quahog culture in the Oak Hill area are: ( 1 1 use a planting density of 7.500 clams/bag for nursery seed and 750 clams/bag for growout seed; (2) could use PED to reduce fouling on the culture bags, although PED may not improve clam survivorship; and (3) invert culture bags periodically. KEY WORDS: Mercenaria mercenaria, density, Florida, fouling INTRODUCTION The northern quahog. Mercenaria mercenaria (Linnaeus, 1758). is a commercially important shallow-water species that has been harvested by subsistence fishermen since pre-Columbian times and today supports an important commercial and recre- ational fishery (Crawford 1992). Since the early 1950s, the feasi- bility of artificially enhancing the commercial crop of cultured northern quahog has been studied intensively (MacKenzie 1979). Some of the factors affecting the growth and survival of commer- cial northern quahog are planting density and predation (Flagg and Malouf 1983). Most studies dealing with densities found that high planting density resulted in slower growth of clams (Eldridge et al. 1976, Eldridge et al. 1979. Hadley and Manzi 1984, Walker 1984). although others indicated no significant difference in final size (Godwin 1968. Summerson et al. 1995). A major inhibition to commercially viable northern quahog culture has been predation (Carriker 1959. Menzel and Sims 1964. Haven and Loesch 1973. Menzel et al. 1976. MacKenzie 1977. Whetstone and Eversole 1978, Steinberg 1980. Castagna and Kraeuter 1981. McHugh 1981. Peterson et al. 1995). with crabs representing the most serious problem from Massachusetts to Texas (MacKenzie 1977. Virnstein 1977, Boulding and Hay 1984. Jory et al. 1984, Hines et al. 1990, Eggleston et al. 1992, Sum- merson et al. 1995. Kraeuter et al. 1998). Several techniques have been developed to minimize predation, including rafts, trays, cages, nets (Manzi et al. 1980). addition of gravel to inhibit for- aging effectiveness (Castagna and Kraeuter 1977). introduction of organisms that consume clam predators (Castagna and Kraeuter 1981, Castagna 1984, Jory et al. 1984, Bisker and Castagna 1989) and combinations of the various methods (Kraeuter and Castagna 1985). Despite the partial effectiveness of these control measures, predation remains a critical factor in northern quahog aquaculture. The field culture of bivalve mollusks is also dependent on the production and supply of phytoplankton and other food sources. Seston depletion is a major influence on cultured suspension feed- ers whose growth can be limited by both food quality and quantity (Fegley et al. 1992). It has proved most economical to grow quahogs in the natural environment at controlled densities, because space and food re- quirements increase exponentially as clams grow (Castagna and Kraeuter 1977). Although field nursery and growout offer a low- cost production system for the shellfish, one of the disadvantages is fouling. Fouling can affect production in various ways. The most obvious is a reduction of water flow through the enclosure, which, in turn, decreases food availability (Paul and Davies 1986, Wildish and Kristmanson 1984). In addition, fouling organisms are often themselves filter feeders, so they compete with the cultured species for food resources. Finally, fouling may reduce oxygen supply (Wallace and Reinsnes 1985). Several solutions have been used to remove fouling: addition of animals that prey upon the biofoulers (Flimlin and Mathis 1993). cleaning and changing structures often (Claereboudt et al. 1994), and inversion of bags (Mojica and Nel- son 1993). In Florida, growth of northern quahogs is more rapid than that observed for northern populations (Jones et al. 1990, Arnold et al. 1 99 1 ) because of the wanner temperature and longer growing 77 78 Fernandez et al. season. The best growth occurs in fall and spring (Eldridge et al. 1976. Eldridge et al. 1979. Menzel 1961). Fouling of culture bags and heavy predation force modifications in culture techniques. such as different mesh size of bags and predator excluding devices (Vaughan et al. 1988), as well as maintenance and cleaning of the system. In Florida, cultured clams reach legal harvest size ( 16 mm in ST) in approximately 9 to 10 months, and the commercial "littleneck" size (25.4 mm in ST, legal harvest size of wild clams) in approximately 12 to 18 months (Vaughan et al. 1988). Clam farming is an important industry in the Indian River Lagoon (IRL) of east central Florida. During the early to middle 1980s, a successful fishery for naturally occurring northern qua- hogs developed in the IRL. Landings from this fishery peaked in 1985 at more than 1.5 million pounds with an estimated value of more than US $8 million (Florida Department of Natural Re- sources 1986). Landings decreased substantially since then. The magnitude and success of this fishery influenced the development of a culture-based fishery in the IRL. The IRL possesses the nec- essary components of a suitable growout site for northern quahog aquaculture (Arnold et al. 1990): area of the lagoon is consider- able, water quality is generally good, and extensive shoreline is available from which to monitor and maintain aquaculture opera- tions. The present study was designed to determine how clam growth, survival, and CI was affected by planting density, predator exclusion device (in addition to the culture bag), and bag inversion at Oak Hill, Florida. MATERIALS AND METHODS Experimental Design This study took place in the IRL at Oak Hill. Florida (Fig. I ). In the IRL. water depth generally does not exceed 1-1.5 m, except near the Atlantic Intracoastal Waterway, and tidal range does not exceed 0.5 m (Sheng et al. 1990). Hatchery-reared northern quahog seed used in this study were produced by Harbor Branch Oceanographic Institution, Inc. in Fort Pierce. Florida. The seed were stocked in bags made of a flexible nylon mesh material. The experiment consisted of two different growing periods. Nursery seed clams were planted on February 27. 1997 and monitored until May 29, 1997 (13 weeks), and growout seed clams were planted on October 10, 1996 and monitored until June 6. 1997 (34 weeks). The experimental densities utilized for the nursery seed were 7.500/bag (5.210/nr), 10,000/bag (6.944/ nr ). and 12.500/bag (8.680/nr). with four replicates for each den- sity. Growout seed densities were 750/bag (521/nr), 1,000/bag (694/nr). and 1,250/bag (868/nr). with four replicates for each density. For both nursery and growout seed experiments, an addi- tional four replicates were planted and covered with PED for each density treatment. The PED was a 1 .6 x 1 .5 m Vexar cover net of 2.5-cm mesh size laid over the bags to exclude predators. Water temperature (to the nearest 0.LC with a thermometer), salinity (to the nearest 1 ppt with a hand-held temperature-compensated re- fractometer (Atago S/Mill), dissolved oxygen (to the nearest 0.01 ppm with a temperature-compensated dissolved oxygen meter (YSI Model 57), current speed (to the nearest 0.01 cm/second) and direction (with a mechanical flowmeter (Model 2030R, General Oceanics Inc.), and Secchi disc depth (to the nearest 1 cm with a 15-cm diameter Secchi disc) were measured weekly. Nursery Seed Nursery seed [mean ± SD shell length (SL): 6.0 ± 0.8 mm; n = 100] were stocked in 24 3-mm mesh size bags (1.2 x 1.2 m) on :a*3Q OAK HILL TrniSVIU-E.:: :3°oom- atlant;c qcsan SS3AS71AN CHEH!Cp\o, VEHO 3EACW.V 27*-Q'- FT. PIEHCc-rA -E.NSEN 3EAC STUAffS^ llu 10"30' 10" Figure 1. Map of the Indian River Lagoon, including the study area (Oak Hill). February 27. 1997. The number of seed placed in each bag was determined volumetrically. The bags and cover nets were kept in place on the bottom with metal stakes. PVC pipes were placed underneath the PED at the corners to maintain tension and prevent predators from entering the bags. Growth was assessed by measuring 100 clams per bag for SL, shell height (SH). and ST to the nearest 0.01 mm with Vernier calipers every 4 weeks. Only SL was used in further analysis because of the high correlations between the measurements (r = 0.97 for SL and SH: r = 0.80 for SL and ST). The measured clams were then returned to the bags. At the beginning of the study. 100 clams were sacrificed and dried in an oven at 65°C for 48 hours to determine shell and soft tissue dry weight (Walne and Mann 1975). At the end of the 13-week trial, another sample of 100 animals per bag was similarly sacrificed and measured. During weekly monitoring, any dead clams found were removed and re- corded, but not replaced. Bags and cover nets were inspected weekly and cleaned biweekly to assure proper water flow. Clean- ing consisted of manually removing fouling organisms that grew on the bags. Two replicates from each density and PED treatment combination were inverted on April 27. 1997. 5 weeks before harvesting to smother fouling organisms further and to examine their influence on growth. The 24 bags containing nursery seed clams were harvested on May 29. 1997. Surviving clams were counted, and the percentage of seed reaching the growout size was determined by sieving (10 to 11-mm mesh screen) clams from each bag. Clams that were retained on the screen were large enough for the growout phase, whereas, seed that passed through Culture of Northern Quahog 79 the sieve were not. The percentage of growout seed from each bag was calculated based on the total number of seed harvested. Growout Seed Growout seed (mean ± SD SL: 21.1 ± 1.7 mm: n = 100) were stocked in 10.5-mm mesh size bags (1.2 x 1.2 m, n = 24) and planted on October 10. 1996 following the same method described for nursery seed. One hundred clams per bag were sampled every 4 to 5 weeks. Growth was assessed by measuring SL, SH, and ST. but only analysis on SL was conducted, because that shell mea- surement was highly correlated (r = 0.94 for SL and SH; r2 = 0.83 for SL and ST). One hundred clams were sacrificed to mea- sure shell and tissue dry weight, as described earlier, at the begin- ning and again at the end of the study. Inversion of bags to smother fouling organisms was not performed on the growout seed. The 24 bags containing growout seed clams were harvested on June 5, 1997. Surviving clams were counted and the percentage of seed reaching legal harvest size was determined by grading (16 mm width on bar grader) clams from each bag. Clams that were re- tained on the bar grader had reached legal harvest size for cultured clams in Florida. The percentage of harvestable clams from each bag was calculated based on the total number of clams harvested. Food A vailability Chlorophyll a concentration and current speed were estimated at five locations in the study site to assess food availability along the prevailing flow gradient. Over a period of several days, float- ing objects were placed in the water during incoming and outgoing tides and followed to establish the prevailing current pattern in the area. Once the prevailing current pattern was established, five locations (two before, one inside, and two after the clam bed) were chosen to estimate food availability. Three 1-Liter water samples were taken at each location, on 3 days during incoming tide (April 10, May 8, and May 22, 1997) and 3 days during outgoing tide (April 18, May 15, and May 29, 1997). Water samples were taken to the laboratory and kept cool and dark until analyzed within a few hours. An appropriate volume (500 to 1000 mL) of seawater was vacuum filtered onto a synthetic filter (Millipore AA 47-cm diameter). Chlorophyll a was measured by spectrophotometries analysis (Strickland and Parsons 1976). Pigments were extracted from the filter with 90% acetone, and pigment absorbance was estimated spectrophotometrically at 750. 660, 647. and 630 nm wavelengths. A standard concentration curve was produced using a commercial chlorophyll a extract (SIGMA Chemical Company, St. Louis, MO). Dalit Analysis All data were examined for variance heteroscedascity using F max test (Sokal and Rohlf 1995), and no data transformation was necessary. Final SL of the clams was analyzed by a two-way (density and PED) ANOVA for growout seed and a three-way ANOVA (density. PED. and inversion) for nursery seed. Bonfer- roni's multiple comparison test was used to compare the means if there was a significant difference in the ANOVA (Sokal and Rohlf 1995). Survival of nursery seed was analyzed by a three-way ANOVA (density. PED. and inversion), and survival of growout seed was analyzed by a two-way (density and PED) ANOVA. Condition index was calculated using the formula: dry soft tissue wt. (g)* 1000/ dry shell wt. (g) (Walne and Mann 1975) and was analyzed by a three-way ANOVA (density, PED, and inversion) for the nursery seed and a two-way (density and PED) ANOVA for the growout seed. Percentage of clams that reached growout size was analyzed by a three-way (density. PED, and inversion) ANOVA. and percentage of clams that reached legal harvest size was analyzed by a two-way (density and PED) ANOVA. Pearson product moment correlation was used to correlate growth with temperature, salinity, and Secchi disc depth (Sokal and Rohlf 1995). The significance level (a) for all statistical tests was 0.05. RESULTS Environmental Parameters Water temperature ranged from 8.8 to 28.8°C during the study period, with a mean (± SD) of 21.7 (±4.1 )'C (n = 32). Mean (± SD) salinity was 32 (±1.7) ppt (n = 32) with a range of 29 to 34 ppt. Mean (± SD) dissolved oxygen concentration (D.O.) was 1 1.6 (±4.7) ppm (n = 32) at the surface (range: 4.8-20.0 ppm) and 1 1.3 (±4.8) ppm (n = 32) at the bottom (range: 5.0-19.0 ppm). Water depth was between 1 and 1.5 m. Mean (± SD) Secchi disc depth was 0.86 (±0.14) m (range: 0.57-1.00 m. n = 32). Mean (± SD) current speed was 6.8 (±4.2, n = 32) cm/s at the surface (range: 2.0-15.0 cm/s. n = 32) and 6.6 (±3.3. n = 32) cm/s at the bottom (range: 3.0-13.8 cm/s). Water temperature, salinity, D.O.. Secchi disc depth, and current speed did not show significant (p > .05) correlation with SL. Nursery Seed Nursery seed clams of all treatments grew at almost perfect linear rates over time, from an average initial SL of 6.0 mm to a mean final SL of 14.6 mm in the 13-week study. Density (p = .03) and inversion (p = .03) had a significant effect on the final SL; whereas, PED did not (p = .31 ). Low density clams tended to be larger than those of medium and high density treatments; and clams from inverted bags tended to be larger than those from the noninverted bags (Table 1 ). Mean percentage of nursery seed that reached growout size ranged from 25.5 to 91.8% (Table 1 ). There was a significant effect of density (p = .02) and inversion (p < .001) on percentage of clams attaining growout size. A higher percentage of clams in the low-density treatment reached growout size than that of clams in the medium- and high-density treatments (Table 1 ); inversion resulted in a higher percentage of nursery seed that reached the growout size (Table 1 ). Survival at the end of the study ranged from 59.0 to 94.5% (Table 1 ). PED (p = .005), and inversion (p = .002) effects were significant, but density effect was not (p = .60). Surprisingly, survival of clams in bags with PED was lower than that of clams in bags without PED (p < ,05, Bonferroni's test) (Table 1); inver- sion resulted in higher survivorship (p < .05, Bonferroni's test) (Table 1). Mean (± SD) initial condition index (CI) of nursery seed was 49.6(±28.3)(n = 100), and it changed little after the 13-week trial period (Table 1). The density (p = .15) or PED (p = .79) had no significant effect: whereas, the inversion (p = .002) did. The CI of the clams in the inverted bags was larger than that of the clams in the noninverted bags (p < .05. Bonferroni's test) (Table 1). Growout Seed Growout seed clams grew from an initial SL of 21.1 mm to a mean final SL of 33.3 mm. There was no significant difference (density: p = .40, PED: p = .17) in the final SL among the 80 Fernandez et al. TABLE 1. Mean )± SD) of shell length, percentage reached growout size, survival, and CI of nursery seed grown for 13 weeks. % Reached Inversion Densitv PED Shell Length Growout Survival Condition (I. NI) (Clams/Bag) (N, C) (mm) Size (%) Index I I 1 I 1 I NI NI NI NI NI NI 7.500 N 17.1 ±0.0a 91.8 + 9.8" 94.5 ± 6.4a 55.2±0.1abc 10,000 N 14.1 ±0.5ab 62.5 ± 13.4abc 76.0 ± 11.3abc 65.3 ± 3.2a 1 2,500 N 14.2±0.8ab 45.5 ± 6.4cd 81.0±5.7ab 52.9±6.1abc 7.500 C 15.1 ±0.2ab 67.5 ± 7.8ab 82.0±5.7ab 56.5 ± 2.6ah 10.000 C 14.4±0.3ab 56.0 + 7. 1KJ 83.5 ± 7.8ab 52.3±4.1abc 12.500 c 13.7±0.7b 57.5 + 0.7abcd 75.5 ± 3.5abc 42.9 ± 2.0bc 7,500 N 14.2±2.0ab 40.5 ± 3.5bcd 78.0 ± 2.8abc 45.9 + 9.0K 10.000 N 13.1 ± 1.0" 29.5 ± 6.4td 77.0 ± l.4abc 36.5 ± 8.0C 12.500 N 14.3+ l.lah 37.0 ± 1.4ud 81.0 + 1.4ab 43.2 ± 4.5* 7,500 C 14.1 ± 1.4ab 35.5 ± 28.9bcd 59.0±8.5C 52.8 + ICO60 10,000 c 13.3 ±0.8" 29.5 ± 10.6"1 69.5 ± 9.2C 43.5 ± 5.7bc 12.500 c I3.9±0.6b 25.5 ± 4.9d 62.5 ± 6.4C 47. 1 ± 4.9abc Values within a column with different superscripts were significantly different. Under inversion. "I" means inverted. "NI" means noninverted: under PED. "N" means not cover, and "C" means cover. treatments (Table 2, Fig. 2). Mean percentage of growout seed that reached 16 mm in ST (legal harvest size for cultured northern quahog) at the end of the study ranged from 30.1 to 66.8% (Table 2). The density effect was significant (p = .04); whereas, the PED effect was not (p = .25). In general, the percentage decreased with increasing density (Table 2). Survival at the end of the study ranged from 75.0 to 87.0% (Table 2). The effects of density (p = .04) and PED (p = .0009) were significant, but there was no apparent pattern (Table 2). Mean (± SD) initial CI of growout northern quahog seed was 65.4 (+28.3). The CI decreased after the 34-week trial period to an average of 36.3, with no significant difference among the treat- ments (Table 2). Food Availability Generally, chlorophyll a concentration was similar among the stations and between incoming and outgoing tides at a given date. Average chlorophyll a concentration in the April 1997 samples was 0.0642 u.g/L. Mean surface current during the April sample days was 1 1.3 cra/s, and mean bottom current was 9.5 cm/s. May 8 and May 15 samples showed an order of magnitude increase in chlorophyll a concentration to 0.96 p-g/L. Average late-May mea- surements of chlorophyll a was 0.81 fxg/L. In May. mean surface current speed was 6.7 cm/s and mean bottom current speed was 6.1 cm/s. DISCUSSION Ansell (1968) reviewed the growth of northern quahog in vari- ous locations along the eastern coast of the United States and concluded that the optimum temperature for growth was approxi- mately 20°C and that shell growth ceased below 9°C or above 31°C. In the present study, the mean water temperature over the 34-week period ranged from 15.1 to 26.3°C. Small changes in salinity do not have a major influence on growth rates, unless the salinity goes below 20 ppt (Castagna and Kraeuter 1981). The optimal salinity for the growth of northern quahog is reported to be about 26 and 27 ppt (Rice and Pechenik 1992). In the present study, salinity ranged from 29 to 36 ppt. In the Oak Hill area, D.O. was high (mean = 13 ppm) during the 34-week period. High D.O. has been found to be common in the southern IRL as well (Arnold et al. 1990, Dierberg et al. 1986). Manzi et al. ( 1981) recommended that intensive field culture is best initiated with seed size larger than 10 mm in SL. However, larger seed are more expensive, and their cost may be >60% of the total cost of producing the final product in northern quahog aqua- culture (Adams et al. 1991). In a study conducted in New Jersey TABLE 2. Mean (+ SD) of shell length, percentage reached legal harvest size, survival and CI of growout seed grown for 34 weeks. Shell "fc Reached Densitv PED Length Legal Harvest Survival Condition (Clams/Bag) (N, C) (mm) Size (%) Index 750 1,000 1,250 750 1.000 1.250 N N N C C C 33.5 ±2.1 32.9 ±2.1 32.8 ± 2.5 34.0 ± 2.2 33.0 ± 2.2 33.8 ± 2.5 53.7 ± 4.3" 49.1 ± U.2"b 30.1 ±4.8b 66.8 ± 8.2J 38.9 ± 10.0" 49.5 ± 3.8ab 75.0 ± 3.4" 87.0 ± 1 .3" 75.0 ± 3.1b 81.0±3.4al 81.0 + 2.3al 82.0±1.8al 37.2 ±6.0 34.8 + 4.9 35.5 ± 5.7 35.3 + 5.8 34.1 ±5.5 38.6 ± 6.4 Values within a column with different superscripts were significantly different. Under PED. "N" means not cover and "C" means cover. Culture of Northern Quahog Time (week) Figure 2. Mean (n = 4) shell length of growout seed over time at the different density and PEI) treatment combinations. (I,n: low-density bags without PEI); m.n: medium density without PKI); h,n: high density without PED; l.c: low with PED; m,c: medium with PED; h,c: high with PED). (Kraeuter et al. 1998) and the present study, successful culture was achieved with nursery seed of 5 and 6 mm SL. respectively. Our nursery seed grew on average 2.76 mm/month in SL. similar to that found by Kraeuter et al. ( 1998) in New Jersey in the summer and by Sturmer et al. (1995) on the west coast of Florida, and higher than that found by Summerson et al. (1995) in North Caro- lina. The growth rates of nursery seed were similar among the months. The winter months had much smaller effects in reducing grow rates on the nursery seed, as it did on the growout seed (Fig. 2). Growout clams in high density bags showed retarded growth during the winter months (Fig. 2). when environmental conditions were not optimal for clam growth. As soon as environmental con- ditions became optimal, clams in the high-density bags grew rap- idly to reach similar SL as the clams of the other treatments (Fig. 2). However, a lower percentage of clams in high-density bags reached the legal harvest size (Table 2). Menzel et al. (1976) and Kraeuter et al. ( 1998) suggested that survival of planted clams should be more than 40 to 50% for the commercial culture to be profitable. In the present study, average survival of nursery clams ranged from 59.0 to 94.5%. This survival is similar to the 87% found by Summerson et al. ( 1995) for the nursery seed grown in raceways in North Carolina and to the 58 to 88% found by Sturmer et al. ( 1995) in a 3-month field study on the west coast of Florida. Densely packed seed clams without PED in southeastern states have demonstrated massive losses to predation in very short peri- ods of time (e.g., Menzel et al. 1976. Gibbons and Castagna 1985. Peterson 1990, Kraeuter et al. 1998). Predators did not seem to be a significant problem in the present study. Although crabs and sheepshead were observed in the area, and sometimes small crabs were found inside the bags. Also, some of the dead clams found showed signs of crab predation: chipped margins or crushed shells (Vaughan et al. 1988). The PED used in the present study did not improve clam survival, and the maintenance of the PED was time consuming. However, biofouling on the culture bags w ithout the PED was heavier than those with the PED. Some PEDs were found float- ing or folded over the bags during the study and were re-installed. The use of PED may affect such biological processes as growth (Virnstein 1977, Dayton and Oliver 1980, Riese 1985). In the present study, survival was slightly higher in bags with PED in the growout period, and clams in bags with PED were found to grow slightly, but not significantly, faster than those in the bags without PED. Water flow in bags without PED seemed to be retarded because of fouling (by drift algae and sea squirts) of the bags. Fouling was found to be higher in bags without PED than in bags with PED where fouling was mainly observed on the PED itself. Fouling organisms diminish the water flow that passes through a bag. preventing clams from getting food necessary for optimal growth (Flimlin and Mathis 1993. Mojica and Nelson 1993). Drift algae (Gracilaria sp.) and tunicates were the most abundant foul- ing organisms in the present study. They grew rapidly if left un- checked. Inversion of the bags to smother fouling organisms was found to increase the growth of clams in the area. Inversion in- creased the percentage of nursery clams that reached the growout size and resulted in higher CI. Inversion also increased survival of the clams. A large increase in chlorophyll a concentration was observed from late April to late May. This increase in food supply in con- junction with increasing temperature may explain the rapid growth of the clams during this period. Since the pioneering work of Kellog (1903), it has been recognized that current speed has a major effect on the growth of northern quahog (Kerswill 1949, Haskins 1952. Hadley and Manzi 1984, Manzi et al. 1986). Grizzle and Morin ( 1989) and Grizzle and Lutz (1989) suggest that north- em quahog growth is primarily determined by horizontal seston flux past the animals and that intermediate seston flux rates pro- duce the highest growth rates in Mercenaria mercenaria in sandy sediments. In the present study, mean current velocity was 6.7 cm/s for the 34-week period, and mean chlorophyll a concentration was 0.61 u,g/L in April and May. Cahalan et al. (1989) indicated that growth rate of scallops peaked at 6.5 cm/s at 6.000 algal cells/mL. In conclusion, a planting density of 7,500 and 750 clams/bag for nursery seed and growout seed, respectively, should be used in the Oak Hill. Florida area, because the highest percentage of seed that reached growout seed or legal harvest size, respectively, was 82 Fernandez et al. found at these low densities. The PED used in this study did not improve survival, and its maintenance is time consuming. How- ever, it could be used to reduce biofouling on the culture bags. It is much easier to clean the PED (with larger mesh size and without clams and sediment inside) than to clean the culture bags. We rec- ommend periodic inversion of bags to smother biofouling organisms. ACKNOWLEDGMENTS Harbor Branch Oceanographic Institution. Inc. supplied the clam seed. We thank Sean Reif for his help in the field. An anony- mous reviewer provided valuable comments to an earlier version of the manuscript. LITERATURE CITED Adams, C. J. Cato. J. E. Easley. Jr.. S. Kemp. W. Mahan, J. J. Manzi, M. Oesterling, R. Pomeroy. E. Thunberg, D. Vaughan & R. Walker. 1991. Investing in commercial hard clam culture: a comprehensive guide to the south Atlantic states. Florida Sea Grant College. Rep. 104. Gaines- ville. Florida. 128 pp. Ansell, A. D. 1968. The rate of growth of the hard clam Mercenaria mercenaria throughout the geographical range. /. Cons. Int. Expl. Mar. 31:364-409. Arnold, W. S„ D. C. Marelli & C. Lund. 1990. Suitabdity of the Southern Indian River Lagoon for hard clam {Mercenaria spp.) culture. Aqua- culture report series. Final reports for the Aquaculture Market Devel- opment Aid Program 1989-1990 and 1990-1991. vol. III. Florida De- partment of Agriculture and Consumer Services, Tallahassee. Florida. Arnold. W. S„ D. C. Marelli. T. M. Bert. D. S. Jones & I. R. Quitmyer. 1991. Habitat-specific growth of hard clams Mercenaria mercenaria from the Indian River. Florida. J. Exp. Mar. Biol. Ecol. 147:245-265. Bisker. R. & M. Castagna. 1989. Biological control of crab predation on hard clams Mercenaria mercenaria by the toadfish Opsanus tau 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. Aquat. Sci. 41:521-525. Cahalan. J. A., S. E. Siddall & M. W. Luckenbach. 1989. Effects of flow velocity, food concentration, and panicle flux on growth rates of ju- venile scallops Argopeclen irradians. J. Exp. Mar. Biol. Ecol. 129:45- 60. Camker. M. R. 1959. The role of physical and biological factors in the culture of Crassostrea virginica and Mercenaria mercenaria in a salt water pond. Ecol. Monogr. 29:219-266. Castagna, M. 1984. Methods for growing Mercenaria mercenaria from postlarval to preferred size seed for field planting. Aquaculture 39:355- 359. Castagna, M. A. & J. N. Kraeuter. 1977. Mercenaria culture using stone aggregate for predator protection. Proc. Nat. Shellfish. Assoc. 67:1-6. Castagna, M. & J. N. Kraeuter. 1981. Manual for growing the hard clam Mercenaria. VIMS Special Rept. in Applied Marine Science and Ocean Engineering, No. 249. 110 pp. Claereboudt, M. R., D. Bureau. J. Cote & J. H. Himmelman. 1994. Fouling development and its effect on the growth of juvenile giant scallops in suspended culture. Aquaculture 121:327-342. Crawford. B. 1992. Oysters and clams. Aquaculture Species Resource Series. Final Repts. for the Aquaculture Market Development Aid Pro- gram 1989-1990 and 1990-1991, vol. III. Florida Department of Ag- riculture and Consumer Services, Tallahassee, Florida. Dayton, P. K. & J. S. Oliver. 1980. An evaluation of experimental analysis of population and community patterns in benthic marine environments, pp. 93-120. In; K. R. Tenore and B. C. Coull (eds.). Marine Benthic Dynamics. University of South Carolina. Columbia. South Carolina. Dierberg. F., J. Ryther. C. Traintafyllidis. R. Creswell. T. Debusk & M. Schilling. 1986. An inventory of bivalves and their food supply in the Indian River. Brevard County, Florida. Brevard County Water Re- sources Department. 61 pp. Eggleston, D. B., R. M. Lipcius & A. H. Hines. 1992. Density-dependent predation by blue crabs on infaunal clam species with contrasting dis- tribution and abundance patterns. Mar. Ecol. Prog. Ser. 85:55-68. Eldridge, P. J.. W. Waltz. R. C. Gracy & H. H. Hunt. 1976. Growth and mortality rates of hatchery seed clams Mercenaria mercenaria in pro- tected trays in waters of South Carolina. Proc. Natl. Shellfish. Assoc. 66: 1 3-20. Eldridge, P. J.. A. G. Eversole & J. M. Whetstone. 1979. Comparative survival and growth rates of hard clams Mercenaria mercenaria planted in trays subtidally and intertidally at varying densities in a South Carolina estuary. Proc. Natl. Shellfish Assoc. 69:30-39. Fegley, S. R . B. A. MacDonald & T. R. Jacobsen. 1992. Short-term varia- tion in the quantity and quality of seston available to benthic suspen- sion feeders. Est Coast. Shelf. Sci. 34:393— J 12. Flagg. P. J. & R. E. Malouf. 1983. Experimental plantings of juveniles of the hard clams Mercenaria mercenaria in the waters of Long Island, New York. J. Shellfish Res. 3:19-27. Flimlin. G. E. & G. W. Mathis. 1993. Biological biofouling control in a field-based nursery for the hard clam. World Aquaculture 24:47—48. Florida Department of Natural Resources (FDNR). 1986. Summary of Florida commercial landings, 1985. FDNR. Tallahassee, Florida. Gibbons. M. C. & M. A. Castagna. 1985. Biological control of predation by crabs in bottom cultures of hard clams using a combination of crushed stone aggregate, toadfish. and cages. Aquaculture 47: 101-104. Godwin. W. F. 1968. The growth and survival of planted clams. Merce- naria mercenaria. on the Georgia coast. Georgia Game Fish Coram. Contr. Series. No. 9. Brunswick, Georgia, pp. 1-16. Grizzle. R. E. & P. J. Morin. 1989. Effect of tidal currents, seston. and bottom sediments on growth of Mercenaria mercenaria: results of a field experiment. Mar. Biol. 102:85-93. Grizzle. R. E. & R. A. Lutz. 1989. A statistical model relating horizontal seston fluxes and bottom sediment characteristics to growth of Merce- naria mercenaria. Mar. Biol. 102:95-105. Hadley. N. H. & J.J. Manzi. 1984. Growth of seed clams Mercenaria mercenaria at various densities in a commercial scale nursery system. Aquaculture 36:369-378. Haskins. H. H. 1952. Further growth studies on the quahog Mercenaria mercenaria. Proc. Natl. Shellfish. Assoc. 42:181-187. Haven. D. S. & J. G. Loesch. 1973. An investigation into commercial aspects of the hard-clam fishery and development of commercial gear for the harvest of mollusks. Virginia Institute of Marine Science, An- nual Contract Rept. 3-124R. pp. 1-92. Hines, A. H.. A. M. Haddon & L. A. Weichert. 1990. Guild structure and foraging impact of blue crabs and epibenthic fish in a subestuary of the Chesapeake Bay. Mar. Ecol. Prog. Ser. 67:105-126. Jones. D. S.. I. R. Quitmyer. W. S. Arnold & D. C. Marelli. 1990. Annual shell banding, age. and growth rate of hard clams {Mercenaria spp.) from Florida. J. Shellfish Res. 9:215-225. Jory. D. E.. M. R. Carriker & E. S. Iverson. 1984. Preventing predation in molluscan mariculture: an overview. /. World Maricult. Soc. 15:421- 432. Kellog. J. I. 1903. Feeding habits and growth of Venus mercenaria. N.Y. State Mus. Bull. 71. Zool. 10:1-28. Kerswill. C. J. 1949. Effects of water circulation on the growth of quahogs and oysters. J. Fish. Res. Bd. Can. 7:545-551. Kraeuter. J. N. & M. Castagna. 1985. The effects of seed size, shell bags, crab traps, and netting on the survival of the northern hard clam Mer- cenaria mercenaria. J. Shellfish Res. 5:69-72. Kraeuter. J. N.. S. Fegley, G. E. Flimlin, Jr., & G. Mathis. 1998. The use of mesh bags for rearing northern quahog (hard clam). Mercenaria mercenaria. seed. J. Shellfish Res. 17:205-209. Culture of Northern Quahog 83 MacKenzie. C. L. 1977. Predation on hard clam Mercenaria mercenaria populations. Trans. Amer. Fish. Soc. 106:530-537. MacKenzie. C. L. 1979. Management for increasing clam abundance. Mar. Fish. Rev. 41:10-22. Manzi, J. J.. V. G. Burrell. Jr. & W. Z. Carson. 1980. A mariculture dem- onstration project for and alternative hard clam fishery in South Caro- lina: preliminary results. Proc. World. Maricult. Soc. 11:79-89. Manzi. J. J.. V. G. Burrell & H. Q. M. Clawson. 1981. Commercialization of hard clam. Mercenaria mercenaria, mariculture in South Carolina: preliminary report. J. World. Maricult. Soc. 12:181-195. Manzi. J. J.. N. H. Hadley & M. B. Maddox. 1986. Seed clam Mercenaria mercenaria culture in an experimental scale upflow nursery system. Aquaculture 54:301-31 I. McHugh, J. L. 1981. Recent advances in hard clam mariculture. J. Shellfish Res. 1:51-55. Menzel. R. W. 1961. Seasonal growth of the northern quahog Mercenaria mercenaria and the southern quahog M. campechinensis in Alligator Harbor. Florida. Proc. Nat. Shellfish. Assoc. 52:37-46. Menzel. R. W. & H. W. Sims. 1964. Experimental farming of hard clams. Mercenaria mercenaria in Florida. Proc. Nat. Shellfish. Assoc. 53: 103- 109. Menzel. R. W.. E. W. Cake. M. L. Haines. R. E. Martin & L. A. Olson. 1976. Clam mariculture in northwest Florida: a field study of predation. Proc. Nat. Shellfish. Assoc. 65:59-62. Mojica. R. Jr. & W. G. Nelson. 1993. Environmental factors of a hard clam aquaculture site in the Indian River Lagoon. Florida. Aquaculture 1 13: 313-329. Paul. J. D. & I. M. Davies. 1986. Effects of copper- and tin-based anti- fouling compounds on the growth of scallops and oysters. Aquaculture 54:191-203. Peterson. C. H. 1990. On the role of ecological experimentation in resource management: managing fisheries through mechanistic understanding of predator feeding behavior. In: R. N. Hughes (ed.). Behavioral Mecha- nisms of Food Selection. Spnnger-Verlag. Berlin, pp. 821-846. 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. Rice, M. A. & J. A. Pechenik. 1992. A review of the factors influencing the growth of the northern quahog Mercenaria mercenaria. J. Shellfish Res. 11:279-287. Riese. K. 1985. Predator control in marine tidal sediments. In: P. E. Gibbs (ed.). Proceedings of the 19th European Marine Biology Symposium. Plymouth. Cambridge University Press. Cambridge, pp. 311-321. Sheng. Y. P.. S. Peene & Y. M. Liu. 1990. Numerical modeling of tidal hydrodynamics and salinity transport in the Indian River Lagoon. Florida Scientist 53:147-168. Sokal, R. R. & F. J. Rohlf. 1995. Biometry. 3rd ed. W. H. Freeman and Co.. New York. 850 pp. Steinberg. M. N. 1980. Preliminary system dynamic model of the effec- tiveness of shellfish hatcheries on increasing harvestable yields. Pro- ceedings of the International. Conference on Cybernetics and Society. Institute of Electrical and Electronic Engineering New York. pp. 895- 900. Strickland, J. D. H. & T. R. Parsons. 1976. A practical handbook of sea- water analysis. Fish. Res. Bd. Can. Bull. 167:31 1. Sturmer. L. N„ E. Quesenberry. J. Scarpa & D. E. Vaughan. 1995. Devel- opment of a shellfish aquaculture industry on the west coast of Florida: seed production, growout monitoring, and species diversification. Ab- stracts from Aquaculture '95, San Diego. California, p. 279. Summerson. H. C. C. H. Peterson & M. Hooper. 1995. Aquacultural pro- duction of northern quahog, Mercenaria mercenaria (Linnaeus. 1758): high water temperatures in the nursery and growth penalties of predator control by gravel. J. Shellfish Res. 14:25-31. Vaughan. D.. L. Creswell & M. Pardee. 1988. A manual for farming the hard-shell clam in Florida. Aquaculture Report Series. Final Repts. Aquaculture Market Development Program 1989-1990 and 1990- 1991. vol. II. Florida Department of Agriculture and Consumer Ser- vices. Tallahassee, Florida. Virnstein, R. W. 1977. The importance of predation by crabs and fishes on benthic infauna in Chesapeake Bay. Ecology 58:1199-1217. 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. Nautilus 98: 1 14-1 19. Wallace. J. C. & T. G. Reinsnes. 1985. The significance of various envi- ronmental parameters for growth of the iceland scallop [Chlamys is- landica) m hanging culture. Aquaculture 44:229-242. Walne, P. R. & R. Mann. 1975. Growth and chemical composition in Ostrea edulis and Crassostrea gigas. pp. 587-607. In: Proceedings of the 9th European Marine Biology Symposium. Aberdeen University Press. Aberdeen. Scotland. Whetstone, J. M. & A. G. Eversole. 1978. Predation on hard clam Merce- naria mercenaria by mud crabs. Proc. Nat. Shellfish Assoc. 58:42-48. Wildish. D. J. & D. D. Kristmanson. 1984. Importance to mussels of the benthic boundary layer. Can. J. Fish. Aquat. Sci. 41:1618-1627. Journal of Shellfish Research. Vol. 18. No. 1. 85-89, 1999. RELATIONSHIP BETWEEN THE BURROWING WORM POLYDORA SP. AND THE BLACK CLAM CHIONE FLUCT1FRAGA SHOWERBY JORGE CACERES-MARTINEZ,1 GISSEL DALILA TINOCO,1 MARCO LINNE UNZUETA BUSTAMANTE,2 AND IGNACIO MENDEZ GOMEZ-HUMARAN3 Laboratorio de Patologia de Mohtscos del Departamento de Acuicultura. Centro de Investigation Cientifica y de Education Superior de Ensenada Apdo. Postal 2732, 2800 Ensenada Baja California, Mexico ~Centro de Investigaciones Biologicas del Noroeste, S.C. Unidad Guaymas Apdo. Postal 349, 85469 Guaymas, Sonora, Mexico ' Departamento de Estadistica del Colegio de la Frontera Norte. Zona Rio, Tijuana, B. C. 22320, Mexico ABSTRACT The black clam Chione fluctifraga is collected for human consumption in both coasts of the peninsula of Baja California. Mexico. An epibionts survey on the black clam from Bahfa Falsa, B.C. and Bahi'a de Guasimas, Sonora revealed an association between the clam and the burrowing worm Polydora sp. Approximately 94% of the worms were located around the siphon area in both valves of the clams taken from Bahia Falsa and 54.5% in clams from Bahi'a de Guasimas. The number of worms per host varied from 1 to 48 in Bahia Falsa and 2 to 15 in Bahia de Guasimas. There was a trend of increased intensity of worm infestation with increased clam size. After a period of 10 months under aquarium conditions, mean percentage of occupation of the siphon area by the worm varied from 94.2% at the beginning of the observation period to 88.3% at the end in clams from Bahia Falsa, and from 54.5% at the beginning of the observation period to 43.4% at the end in clams from Bahia de Guasimas. There was an increase in the mean number of worms on the clams after the observation period, from 9 to 15.7 worms in organisms from Bahi'a Falsa and from 5.2 to 9.5 worms in clams from Bahia de Guasimas. Worms may survive on the shell after the host is dead. Temperature during observation period varied from 20 to 25.5°C. The U-shape channels of the worm result in a very porous and brittle host shell. In heavily infested clams the shell is broken and this impinges on the clams ability to close its valves. This is the first record of burrowing worms associated with the siphon aperture area of the shell of C. fluctifraga. KEY WORDS: Polydora. Chione fluctifraga, burrowing worm, infestation INTRODUCTION worm can also render oyster shells brittle and easily broken during shucking, packaging, and transport (Korringa 1951). The named "polydorid" complex of the family Spionidae com- The black clam. Chione fluctifraga is a highly regarded food prises a number of highly diverse but closely related species, all and supports an extensive sport and commercial fishery in south- characterized by a modified fourth or fifth setiger (Light 1978). ern California (Haderlie and Abbott 1980). This species is also Among them, the genera Polydora and Boccardia contain a large gathered for human consumption on the Pacific coast of Baja number of species able to bore into calcareous substrates including California and the Gulf of California, Mexico (Martinez-Cordova shells of such commercially important bivalves as mussels, oys- 1988, Martinez-Cordova 1996). A survey of clams from Bahia ters. cockles, and scallops (Read 1975, Sato-Okoshi et al. 1990. Falsa on the Pacific coast of Baja California, and Bahia de Guasi- Blake 1996, Handley and Bergquist 1997). These species live in a mas, Sonora. Mexico revealed the presence of polychaetes on tube inside the hole bored in the shell of the host with two exterior shells. The aims of this study were to determine the identity of the apertures. The anterior end of the worm emerges from the tube and polychaete and to document some aspects of the relationship be- feeds from particles taken from the sediment surface or from the tween this worm and the black clam, overlying water column with the aid of its palps (Daro and Polk 1973, Blake 1996). Its boring activity may reach the inner surface MATERIALS AND METHODS of the mollusk's shell and induce the host to secrete calcite and conchiolin layers, forming a blister to isolate the worm (Kent In April 1997. a sample of 48 live Chione fluctifraga from 1979, Lauckner 1983). As a consequence, these worm species are Bahia Falsa. Baja California was collected and subsequently, in also named "mudworms" or "blisterworms" (Lauckner 1983). A June, a sample of 71 dead clams (empty paired shells) were ob- deviation of host energy for growth and reproduction to build a tained. Finally, in July 1997. a sample of 125 live black clams from blister has been suggested by some authors (Williams 1968. Kent Bahia de Guasimas, Sonora was also collected (Fig. 1). Bahia 1979). In edible oysters, the blisters affect the half-shell market. Falsa has a muddy bottom, and Bahia de Guasimas has a sandy and because the blisters can be punctured and release anaerobic me- muddy bottom. After washing the clams in running seawater, each tabolites. including hydrogen sulphide (Handley 1995). A blister- live clam was measured (length from the umbo to the posterior 85 86 Caceres-Martinez et al. 1 1 \ 116t,00W \ _-/•? Bahia \\ ^} 30°25'N- __J_^ MEXICO r1^ 110'35-W >l Bahia X , de Guasimas^--£n] -27°50'N fl-, f\r^ Figure 1. Map showing Bahia Falsa in Baja California and Bahia de Guasimas, Sonora. Mexico. Black dots indicate sampling areas. margin of the shell I. The surface of the shell was delimited for examination in three zones (approximately 30% of the total surface area each one): zone 1 (Zl), around the siphon area; zone 2 (Z2) in the middle area, and zone 3 (Z3) opposite to the siphon area (Fig. 2). Clams were placed individually in Petri dishes filled with seawater. and the number of worms emerging from the shell were counted by zone under the dissecting microsope. Two holes from the same "U"-shaped channel were considered as one worm. Sub- sequently, clams were opened with a knife, and the meat was discarded. Then, the inner sides of the right and left valves were checked for worms by zone, and Polydora infestations visible in the inner shell were enumerated by zone. The total infestation intensity was determined by comparing the number of worms ob- served on both sides of the shell. Dead clams were cleaned under running tap water, and worm blisters and holes related to burrow- ing worms were counted by zone. Prevalence was considered as the percentage of infested clams in the sample. To determine the degree of damage of the burrowing worm on the shell of the clam. X-ray radiographs were taken from clams with different infestation intensities. The number of worms was related to the size of the clams. Fifteen worms were extracted from the shell of six clams col- lected in Bahia Falsa, and seven were extracted from six clams from Bahia de Guasimas by crushing them with nippers around the edge of the shell, where the worms were located. The worms were removed from the shell fragments with dissection tweezers and fixed in 70% ethanol for identification. To assess worm behavior. 33 infested live clams from Bahia Falsa and nine infested live clams from Bahia de Guasimas were labeled and placed separately in aquaria without sand for 10 months. The water was changed every 3 or 4 days, and the tem- perature was recorded with a manual thermometer. The clams were fed once daily with hochrisis galvana and Chaetoceros sp. The water was checked for release of burrowing worm larvae and the Bahia Falsa Bahia de Guasimas MeanZl =94.2% N = 33 11 = 262 Mean = 9 SE = 1.3 Range = 1-25 N=23 n = 303 Mean = 15.7 SE = 1.8 Range = 6-37 MeanZl =54.5% N = 9 n = 47 Mean = 5.2 SE= 1 6 Range = 2-15 MeanZl =43 4 Figure 2. Zones delimited on the valves of Chione fluclifraga to determine distribution of Polydora sp. Percentage of Polydora sp. infesting different zones (Zl, Z2, Z}) of the right (R) and left (L) valves of the black clam from Bahia Falsa, B.C. and Bahia de Guasimas, Sonora, at the beginning of the observation period (Al and the end (Bl. N = number of clams studied; n = number of Polydora sp.; Mean Zl = Mean of the percentages of occupation of worms in both valves; Mean = mean number of worms in both valves; SE = standard error of the mean number of worms in both valves; Range = minimum and maximum number of Polydora sp. found in both valves. POLYDORA SP. AND ChIONE FLUCTIFRAGA 87 C 50 o 'O 40 Mean +4.472 * SE ^_^ Mean + SE Mean - SE Mean - 4.472 * SE □ Mean 0 34-36 37-39 40-42 43-45 46-48 49-51 52-54 55-57 Size (mm) Figure 3. Relationship between the size of the clams and the number of worms infesting their shells. number of living worms, and worm holes per zone in the shell of the clams were counted and compared to the initial numbers at the end of the period. Statistics Mean shell length and infestation between live and dead clams were compared using t and the Kruskal-Wallis tests. A nested effect model was carried out to determine infestation differences between: ( 1 ) clams from Bahia Falsa and Bahia de Guasimas; and (2) valve side: and (3) among valve zone, assuming that the num- ber of worms per zone was nested to the corresponding valve, which was also nested to the corresponding bay. Finally, another nested effect model was applied to the study in aquaria conditions. The initial number of worms was incorporated as a covariate to analyze whether the number of worms changed at the end of the study. RESULTS The mean size of live clams from Bahia Falsa was 43.3 mm (SE 0.48) and of dead clams was 46.4 mm (SE 0.56), the difference was significant (r = -39. p < .01 ). The mean size of clams from Bahia de Guasimas was 32.9 mm (SE 0.2). Holes in the shells were occupied by a spionid polichaete from the genus Polydora, and the mean size of worms was 37.7 mm (SE 5.68). Their morphological characteristics agree with the descrip- tion of Polydora limicola (Anenekova): prostomium weakly in- cised along anterior margin, caruncle extending to setiger 3. 4 eyes present: palps and dorsum of anterior setigers with black pig- mented bands or without bands; major spines of setiger 5 with a small, triangular lateral tooth; posterior notopodial spines absent and pygidium disclike with dorsal notch. However, its borrowing behavior suggests the species P. ciliata. (Johnson) (See Blake 1996; 173). The worm produced U-shaped channels, which were filled with compacted mud. The burrows were extended by "chim- neys" composed of detritus (algae remnant), which protruded from the surface of the clams' valves. Worms were very active, their palps were continuously extended from the burrows. Polydora prevalence in live and dead clams from Bahia Falsa was 48.6 and 66%, respectively. The number of worms per host ranged from 1 to 25 in live clams and from 1 to 48 in dead clams, and infestation differences were not significant, /-test (/ = 1.345, p = .180), Kruskal-Wallis test (X2 = 0.110. p = .740). In gen- eral, there was a trend of more Polydora sp. in larger clams relative to small clams (Fig. 3), however, this trend was not significant. f-test (F = 1.750. p = .112). Kruskal-Wallis test (\2 = 10.600. p = .157). Figure 4. X-ray photograph of several black clams showing different degrees of damage in the shell; note the channels are touching each other in heavily infested clams and the borders of the shells are de- stroyed. Worm prevalence in clams from Bahia de Guasimas was 15%, and the number of worms per host ranged from 2 to 15. Between 95 and 97% of the observed worms were placed in the Zl of live and dead clams from Bahia Falsa, respectively. The corresponding intensity of infestations were 5 and 3% in Z2. there were no worms in Z3. The distribution of Polydora sp. in clams from Bahia de Guasimas was 54.5% in the Zl, 18.2% in Z2, and 27.3% in Z3. The right valve was slightly more infested than the left valve. The results of the nested effect model confirmed that the infestation in clams from Bahia Falsa was greater than in clams from Bahia de Guasimas (F = 229.370. p < .0001): the model also showed that there was a greater infestation in the right than in the left valve (F = 6.390, p = .0017): finally, it was confirmed that the infestation per valve zone was greater in Zl than Z2 and Z3 (F = 72.660. p < .0001). The damage on the shell depends upon the number of worms and size of the channels. Figure 4 shows different degrees of infestation and associated damage. The channels often extended toward the middle of the valve (Z2) were in close proximity with one another, resulting in a very brittle shell. The shell of heavily infested clams was often broken in the siphon area, hindering valve closure (Fig. 5). In aquaria, where clams could not burrow into substrate. Poly- dora sp. showed a slight tendency to spread on all the surface of 88 Caceres-Martinez et al. Figure 5. Shell of Chione fluctifraga infested by Polydora sp. around the siphon area; note the holes in the area (Al, heavy infestation results in a very brittle shell that is easily broken (B). both valves of the host. Mean percentage of occupation of Zl varied from 94.2% at the beginning of the observation period to 88.3% at the end of the observation period in clams from Bahia Falsa, and from 54.5% at the beginning of the observation period to 43.4% at the end of the observation period in clams from Bahia de Guasimas. The distribution of worms was similar in both valves (Fig. 2). There was an increase in the mean number of worms on the clams after the observation period from 9 ( 1-25) to 15.7 (6-37) worms in organisms from Bahia Falsa, and from 5.2 (2-15) to 9.5 (2-38) worms in clams from Bahia de Guasimas (Fig. 2). The results of the nested effect model showed that infestation in clams from Bahfa Falsa and Bahia de Guasimas was slightly different (F = 3.070. p = .081); the same model revealed that infestation differences between both valves were not significant (F = 0.731, p = .482). However, infestation differences among valve zone showed that it was greater in Zl than Z2 and Z3 (F = 9.250. p < .0001 ): finally, it was corroborated that there was an increase in the number of worms at the end of the study ( F = 1 38.53, p < .000 1 ). Approximately 30% of the clams from Bahia Falsa and 1 1 % of the hosts from Bahia de Guasimas died during this observation period. Worm larval stages were observed during water exchange. The number of living worms at the end of the study from clams be- longing to Bahia Falsa was 58 and in clams from Bahia de Guasi- mas was 35. Temperature ranged from 24 to 25.5°C in summer and from 19 to 21°C in winter. DISCUSSION The polydorid species Polydora limicola and P. ciliata have been found in the northwest Pacific (Radashevsky 1993). Polydora limicola has been found in southern California (Hartman 1961, Blake 1996). However, the authors recognize that there are diffi- culties in their systematics. In accordance with Blake ( 1996), Poly- dora limicola is virtually identical to that of the well-known and widely distributed shell and limestone borer. P. ciliata. Manchenko and Radashevsky ( 1994) evaluated genetic differences between P. limicola and the superficially indistinguishable shell borer (P. cf. ciliata) in the Sea of Japan and found that, although there were no morphological differences, there were clear differ- ences in 10 of 27 gene loci surveyed. These genetic differences support the separation of P. ciliata and P. limicola. which formerly had been based strictly on habitat (Blake 1996). In this sense, the species that we found in this study could be P. ciliata: however, detailed molecular genetic studies are needed to clarify its identity . This is the first record of Polydora sp. associated with the siphon area of the black clam Chione fluctifraga; however, a simi- lar observation has been recorded in C. stutchburyi infested by the polychaete Boccardia acits in Wellington Harbour, New Zealand. This worm is a common and conspicuous epibiont of the cockle C. stutchburyi. The U-shaped burrow usually follows the curve of the shell growth lines. The boring is not lined with sand apart from a sand-grain partition at the U-bend and a short external sand-grain chimney. The external chimney extends out around the siphons of the cockle, which lies buried just beneath the sediment surface (Read 1975). The polychaete Boccardia chilensis has also been recorded boring in Chione stuchburyi shells in association with Broccardia syrtis [Read (1975)]. The high percentages of worms sited around the siphon area may suggest a specialized relationship between worm and host. The specialized relationship between spionids and their hosts has been described in Polydora commen- salis: it lives in a shallow burrow excavated along the columnella of the gastropod shell occupied by a hermit crab. This specialized species has short palps with an unusually narrow food groove that seems to be adapted to capturing food particles stirred up or sus- pended by the activities of the hermit crab (Blake 1996). Our results suggest that the relationship between the worm and the clam seems to be less specialized, because the worm may be sited out of the siphon area if the surface of the clam is available, as in aquaria conditions. Moreover. Polydora sp. remain alive on the shell after host death. The trend of more worms around the siphon area between clams from Bahia Falsa and those from Bahia de Guasimas suggest that, in the latter bay, clams are more exposed to colonization by Polydora sp. than in the former. Worm prevalence could be related to the type of substrate and with particular envi- ronmental factors of the two embayments we examined. Its place- ment, exclusively around the siphon aperture, allows the worm to feed on the particles inhaled or expelled by the clam, resulting in an advantageous position relative to other surface areas of the shell, if available. We observed great motility of worm palps while the clam protruded its siphons. The preference of worms for the right valve recorded in the field study remains unknown. The prevalence and number of worms per host in studied clams could be related to age and size of the clams. Possibilities of infestation by Polydora sp. in older clams are higher than in younger clams, because the former have had more encounters with POLYDORA SP. AND CHIONE FlUCTIFRAGA 89 the worms. In addition, larger surfaces provide more area for bur- rowing worm colonization. This observation could explain differ- ences in prevalence and number of worms per host between larger and older clams from Bahfa Falsa relative to smaller and younger clams from Bahi'a de Guasimas. The different environmental con- ditions of both bays may also play an important roll in prevalence and abundance of the worm. In aquaria conditions, the increase in number of worms through time and the presence of larval stages indicated reproduction and settling of the worm species. Temperature is one of the primary factors for determining the abundance of Polydora sp. (Lauckner 1983); in other words, generation time, reproduction and, hence, transmission. In this study, temperature was maintained near the values recorded in Bahi'a Falsa (see Caceres-Marti'nez et al. 1998) and Bahi'a de Guasimas (Arreola 1998), this supported the repro- duction, setting, and the increase in the number of worms recorded in this study. However, specific studies on temperature in relation to reproduction, growth, and transmission are needed. The mean number of worms per host (initial and final) was slightly higher in larger and older clams from Bahi'a Falsa, than in those from Bahi'a de Guasimas. This observation also supports the observed rela- tionship of surface area and intensity of worm infestation (Fig. 2). Mortality of both host and worms was detected at the end of the observation period. This could be related to deterioration of aquarium conditions ( frequency of water renovation and nutrition). However, specific studies on clam mortality in relation to the presence of this worm are needed. Heavy infestation may result in severe damage to the clam shell, a brittle shell border may increase the potential for mortality because of enhanced predation as a result of holes in the valves, and problems handling the clam for packing. ACKNOWLEDGMENTS M. C. Veronica Rodriguez identified the worms and Dr. M. A. del Rio Portilla took the photographs. Vicente Guerrero provided us with the clams and logistic support during sampling in Bahfa Falsa. This work was supported by CICESE # 623106. LITERATURE CITED Arreola. L. A. 1998. Grupo de Zonas Costeras del CIBNOR Unidad Guay- mas. Serie de datos ambientales de las Bahi'as del Estado de Sonora. Cento de Investigaciones Biologicas del Noroeste. Guaymas. Sonora. Mexico. Blake, J. A. 1996. Family Spionidae. Grube. 1850. pp. 81-223. In: Taxo- nomic Atlas of the Santa Maria Basin and Western Santa Barbara Channel, vol. 6. The Annelida. Part 3. Polichaeta: Orbiniidae to Cos- suridae. Santa Barbara Museum of Natural History. Santa Barbara. California. Caceres-Martfnez. J.. P. Macias-Montes de Oca & R. Vasquez-Yeomans. 1998. Polydora sp. infestation and health of the pacific oyster Cras- sostrea gigas cultured in Baja California, NW Mexico. J. Shellfish Res. 17:259-264. Daro. M. N. & P. Polk. 1973. The autoecology of Polydora ciliata along the Belgian coast. Neth. J. Sea Res. 6:130-140. Haderlie E. C. & D. P. Abbott. 1980. Bivalvia: the clams and allies, pp. 355-411. In: R. H. Moms. DP. Abbott, and EC. Haderlie. (eds.). Intertidal Invertebrates of California. Stanford University Press. Stand- ford. California. Handley, S.J. 1995. Spionid polychaetes in Pacific Oysters. Crassostrea gigas (Thunberg) from Admiralty Bay. Marlborough Sounds, New Zealand. A'ZV. Mar Freshwater Res. 29:305-309. Handley, S. J. & P. R. Bergquist. 1997. Spionid polychaete infestations of intertidal pacific oyster Crassostrea gigas (Thunberg). Mahurangi Har- bour, northern New Zealand. Aquaculture 153:191-205. Hartman, O 1961. Polychaetous annelids from California. Part III. Sys- tematica: Polychaetes. Allan Hancock Pacific Expeditions 27:1-93. Kent. R. M. L. 1979. The influence of heavy infestations of Polydora ciliata on the flesh content of Mytilus edulis. J. Mar. Biol. Ass. U.K. 59:289-297. Korringa. P. 1951. The shell of Ostrea edulis as a habitat. Arch, neerlan- daises zoologie 10:32-136. Lauckner. G 1983. Diseases of mollusca: bivalvia. pp. 477-1038. In: O. D. Kinne (ed.). Diseases of Marine Animals. Biologische Anstalt Hego- land. Hamburg. Federal Republic of Germany. Light. W.J. 1978. Spionidae polychaeta annelida. pp. 1-211. In: W. L. Lee (ed.). Invertebrates of the San Francisco Bay Estuary system. Cali- fornia Academy of Sciences. Manchenko. G. P. & V. 1. Radashevsky. 1994. Genetic differences between two allopatric sibiling species of the genus Polydora (Polychaeta: Spionidae) from the western Pacific. Bioch. Syst. Ecol. 22:767-773. Martinez-Cordova. L. R. 1988. Bioecologia de la almeja negra Chione fluctifraga (Sowerby, 1853). Rev. Biol. Trop. 36:213-219. Martinez-Cordova. L. R. 1996. Contribution to the knowledge of the mal- acological fauna of four costal lagoons in the state of Sonora. Mexico. Ciencias Marinas 22:191-203. Radashevsky, V. I. 1993. Revision of the genus Polydora and related gen- era from the Northwestern Pacific (Polychaeta: Spionidae). Publ. Sew Mar. Biol. Lab. 36:1-60. Read. G. B. 1975. Systematics and biology of polydorid species (Polycha- eta: Spionidae) from Wellington Harbour. J. Roy. Soc. of NZ 5:395- 419. Sato-Okoshi. W. Y. Sugawara & T. Nomura. 1990. Reproduction of the boring polychaete Polydora variegrata inhabiting scallops in Abashin Bay, North Japan. Mar. Biol. 104:61-66. Williams. C. S. 1968. The influence of Polydora ciliata (Johnston) on the degree of parasitism of Mytilus edulis L. by Mytilicola intestinalis Steur. /. Animal Ecol. 37:709-712. Journal of Shellfish Research, Vol. 18. No. 1, 91-97, 1999. ADHESION OF VIBRIO TAPETIS TO CLAM CELLS LOURDES LOPEZ-CORTES, ANTONIO LUQUE, EDUARDO MARTINEZ-MANZANARES, DOLORES CASTRO, AND JUAN J. BORREGO Department of Microbiology, Faculty of Sciences, University of Malaga, 29071 -Malaga, Spain ABSTRACT The adhesive properties of Vibrio rapetis, the causative agent of the brown ring disease affecting cultured clams, were determined considering both the contribution of bacterial surface hydrophobicity and the attachment capability to different animal cells. Hydrophobicity of V. tapetis strains was evaluated by means of three different methods, most of the strains being highly hydrophobic for any of the methods used. V. tapetis showed higher adhesion capability toward the clam cells used (hemocytes and mantle cells), as compared to several fish cell lines. No significant relationship was obtained between hydrophobicity and cell adhesion, which suggests the existence of adhesion-specific mechanisms. In addition, different bacterial structures were investigated as potential adhesins of V. tapetis, including hemagglutinins, pili, flagella, and outer membrane proteins. KEY WORDS: Vibrio tapetis, brown ring disease, adhesive capabilities, hydrophobic interaction, clam cells INTRODUCTION Vibrio tapetis is the causative agent of brown ring disease IBRD), an epizootic disease that affects cultured clams (Tapes philippinarum and T. decussatus). Although experimental repro- duction of BRD symptoms in healthy clams has been achieved by means of V. tapetis inoculation (Paillard and Maes 1990, Castro 1994. Novoa et al. 1998). the precise mechanisms involved in the in vivo infection have not yet been well established. Bacterial attachment and ulterior colonization of the clam periostracal lamina seem to be the first steps in the pathogenesis of V. tapetis (Paillard and Maes 1995a, Allam et al. 1996). The colo- nization and disruption of the periostracal lamina provoke the bac- terial accumulation on the inner surface of the clam shell, thereby producing the conchiolin deposit (Paillard and Maes 1995b), which constitutes the main gross symptom of this disease. In BRD. as in other fish and shellfish diseases, bacterial adhe- sion to appropriate host surfaces is a key factor for infection es- tablishment (Daly and Stevenson 1987, Santos et al. 1991). How- ever, little is known about the factors contributing to V. tapetis adhesion to host surfaces. Several potential adherence factors have been described for Vibrio species, including surface proteins, hemagglutinins, and several types of pili (Jonson et al. 1991, Spe- randio et al. 1995). These bacterial surface structures, named ad- hesins, interact with a broad variety of molecular host-cell recep- tors (Iijima et al. 1981. Christensen et al. 1985, Nakasone and Iwanaga 1990). On the other hand, it has been reported that hy- drophobic interactions in addition to hemagglutinating capabilities could be responsible for bacterial adhesion to animate and inani- mate surfaces (Bruno 1988. Clark et al. 1989. Savage 1992, Vazquez-Juarez et al. 1994). However, several Vibrio species showed an ability to adhere to host cells and cell lines, regardless of their degree of hydrophobicity (Santos et al. 1991 ). The aim of this work is to study the adhesion properties of V. tapetis, consid- ering both the contribution of bacterial surface hydrophobicity and the specific attachment to different cells. MATERIAL AND METHODS Microorganisms and Culture Conditions Twenty-seven strains of V. tapetis were used for the hydropho- bicity studies. Bacterial strains were grown in tryptone soya broth or agar (Oxoid Ltd.. Basingstoke, Hampshire. UK) supplemented with 1.5% NaCl (TSBS and TSAS. respectively), and incubated at 22°C for 18 h. Eight V. tapetis strains, representative of each V. tapetis group established previously by Borrego et al. (1996b) and Castro et al. (1997) were used for the adhesion experiments to cells. Bacteria were grown in TSBS for 18 h at 22°C. resuspended in sterile buffered saline (BS). and washed twice by centrifugation (at 24.000 x g for 5 min at 4°C). Bacterial pellets were resus- pended in the same BS and adjusted at a bacterial concentration of 108 bacteria/mL. Hydrophobicity Assays To test the hydrophobic capabilities of V. tapetis strains, three different assays were performed: microbial adhesion to hexa- decane (MATH); salt aggregation test (SAT); and nitrocellulose adhesion test (NCF). MATH was performed as described by Rosenberg et al. (1980). Bacteria were centrifuged at 4,000 x g for 10 min at 4°C, washed, and resuspended in phosphate urea magnesium sulphate (PUM) buffer (22.2 g/L K2HP04 • 3H20. 7.26 g/L KH,P04. 1.8 g/L urea. 0.02 g/L MgS04-7H20. pH 7.1). or phosphate buffered saline (PBS) (0.02 M. pH 7.2) to an absorbance at 400 nm of 0.9-1.1. Bacterial suspension aliquots (2 mL) were then transferred to clean round-bottom test tubes, and 0.3 mL n-hexadecane (Sigma Chemi- cal Co., St. Louis. MO, USA) were added and incubated for 10 min. After the mix was homogenized for 2 min, the hydrocarbon phase was allowed to rinse completely, and the aqueous phase was removed to determine the absorbance at 400 nm. The percentage of adhesion to hydrocarbons was calculated using the following ex- pression: Adhesion (%) = [A4(MI (initial bacterial suspension) - A40n (aqueous phase)]/[A4m (initial bacterial suspension)] x 100. The ability of the bacteria to bind to nitrocellulose filters (NCF) was determined according to the technique described by Lachica and Zink ( 1984). Bacterial cultures were centrifuged at 4,000 x g (10 min at 4°C), washed as above, and resuspended in saline solution (0.85% NaCl. pH 7.2) at an absorbance of 1 at 600 nm. Suspensions were filtered through a 13-mm NCF (type CS, 8.0- u,m pore size) (Millipore Corp.. Bedford. MA, USA). Optical den- sity of the filtrates was measured at 600 nm. and the percentage of adhesion was expressed as: Adhesion (%) = [A60l) (initial bacte- 91 92 Lopez-Cortes et al. rial suspension! - A600 ( filtrate )]/[A600 (initial bacterial suspen- sion!] x 100. The SAT, described by Lindhal et al. ( 1981 ), is based on bac- terial precipitation in presence of salts. Bacterial cultures were centrifuged at 4.000 x g, washed, and resuspended in PBS (0.002 M. pH 8.6) to achieve a concentration of 5 x 10'' bacteria/mL. Then, 30-p.L aliquots of bacterial suspensions were mixed with equal volumes of decreasing molarities of buffered ammonium sulfate solutions ranging between 0.05 and 4 M. Hydrophobicity was expressed as the lowest molarity of ammonium sulfate that produced visual clumping. Kendall rank coefficients were calcu- lated to determine the correlation between the different hydropho- bicity tests assayed. Hemocytes and Mantle Cells Collection Hemolymph was taken from the posterior adductor muscle of two clam species. Tapes decussatus and T. philippinarum, using a 20-gauge needle attached to a 3-mL syringe, through a hole per- formed in the shell margin of the clams. Then, the collected hemolymph was diluted 1 :3 in a modified anti-aggregant Alsever solution (MAS) (20.8 g/L glucose. 8.0 g/L sodium citrate. 3.36 g/L EDTA. 22.5 g/L NaCl, and 100 u-L/L distilled water). Hemolymph of 5 adult specimens of each clam species was pooled and the number of hemocytes was estimated using a Coulter-counter. Be- fore the adhesion assays, hemocyte suspension in MAS was cen- trifuged at 400 x g for 10 min. supernatant removed, and hemocytes were resuspended in BS (0.58 m NaCl. 13 mm KG, 13 mm CaCL, 26 mm MgCL. 0.54 mm Na2P04. 50 mm Tris-HCl. pH 7.4). Then, the cells were fixed with 3.7% formaldehyde for 20 min at 4°C. Formaldehyde was removed by centrifugation (400 x g for 10 min). and the pellet resuspended in BS at a concentration of 10"cells/mL. Mantle cells were collected from healthy specimens of both clam species. Briefly, the clams were opened, and the mantle was extracted in aseptic conditions and washed for 15 min in BS. for 20 min in an antibiotic solution (Sigma. 10.000 IU/mL penicillin. 10 mg/mL streptomycin and 25 u.g/mL amphotericin) 10-fold diluted in PBS, and finally washed for 15 min in BS supplemented with 2.5% trypsin. Mantle tissue was disrupted using Pasteur pipettes, centrifuged at 300 x g for 5 min at 4°C. and the pellet was resus- pended in saline solution (0.55 m NaCl in distilled water). Mantle cells were isolated using a continuous gradient of Percoll (Amer- sham Pharmacia Biotech GmbH. Barcelona. Spain) previously prepared by centrifugation (at 25.000 x g for 20 min at 4°C) of a 60% percoll in saline solution. Cells were separated by centrifu- gation at 10.000 x g for 10 min at 4°C. The band in the percoll gradient that contained the cells was collected, washed twice in BS (at 300 x g for 5 min at 4°C), and fixed with 3.7% formaldehyde for 20 min at 4°C. Then, the formaldehyde was removed by cen- trifugation (400 x g, for 10 min), and the pellet was resuspended in BS at a concentration of 106 cells/mL. Cell Adhesion Assays The adhesion of V. tapetis to hemocytes or mantle cells of both clam species was evaluated by two different methods: the adhesion method described by Kumazawa et al. (1991), and by an ELISA test developed in the present study. Clam cells and V. tapetis were incubated at a concentration of 1 x 10" cells/mL and 1 x 10x bacteria/mL, respectively, in BS at room temperature (about 20°C) for 2 h with gentle aeitation. After the incubation, the nonadhered bacteria were removed by three cycles of centrifugation (at 300 x g for 5 min. 4°C) in BS. and the final pellet was resuspended in 500 p.L of BS and fixed with 0.7% formaldehyde overnight. Af- terwards, volumes of 100 u.L were deposited in microplate wells to perform the ELISA test, and volumes of 200 uL were disposed in slides, stained with Giemsa and observed under light microscopy. In the indirect ELISA test, an anti-V. tapetis serum raised in rabbit (Castro et al. 1995) was used as the first antibody, and antirabbit IgG labeled with peroxidase (Sigma) as the second an- tibody. Mixtures of bacteria and clam cells without the first anti- body, and clam cells with the first and second antibodies, were used as controls. Adhesion to Fish Cell Lines Three fish cell lines were used for adhesion assays, Chinook salmon embryo (CHSE), epithelioma papullosum of carp (EPC) and SAF-1 derived from fibroblast of gilt-head seabream fins. Cells were maintained in Eagle's minimal essential medium (MEM) (Gibco Life Technologies. Paisley, UK) or, in the case of SAF-1 cell line, in L-15 Leibovitz medium (Gibco) supplemented with 2% glutamine. both containing 10% fetal calf serum and antibiotics (1% penicillin/streptomycin). Semieonfluent monolay- ers were grown on 24-multiwell plastic dishes with glycerol- treated coverslips (12-mm diameter). Cell monolayers were fixed with 3.7% formaldehyde for 20 min at 4°C, and washed thor- oughly with PBS. To perform the adhesion assays, bacterial suspensions contain- ing 10s bacteria/mL were placed in the multiwell dishes containing the cell-coated coverslips and incubated at 20°C with gentle shak- ing for 2 h. After being washed thoroughly with PBS. coverslips were air dried and fixed with formaldehyde for 20 min. Then, coverslips were stained with crystal violet, mounted onto micro- scope slides, and examined under light microscopy. The adherence to fish cell lines was also evaluated using the ELISA test described above. Fish cell monolayers were trypsinized and resuspended in fresh medium without antibiotics. Then, the cells were fixed and washed as described, and resuspended in PBS at a concentration of 10h cells/mL. Adhesion assays and the ELISA test were conducted as mentioned for clam cells. Hemagglutination Tests Hemagglutination was determined using rat, horse, rabbit, and human erythrocytes according to the technique described by Larsen and Mellergaard ( 1984). Equal volumes ( 100 u.L) of bac- terial suspension (109 bacteria/mL) and erythrocyte suspension (3%, v/v) in PBS (0.01 m. pH 6.8) were mixed on a 96-well plate, and incubated at room temperature for 1 h. As negative controls, erythrocyte suspensions in PBS and bacterial suspensions in PBS were used. The test was considered negative if visible agglutina- tion did not occur within 10 min. Inhibition of hemagglutination was performed by mixing the bacterial suspensions with 10. 25. 50. 75. and 100 mm solutions in PBS of D-mannose. D-fucose. L-fucose. D-glucose. D-galactose. D-fructose. D-lactose, and raffinose (Sigma). A negative control of erythrocytes plus sugar in PBS was used. Transmission Electron Microscopy ITEM) The arrangement of flagella and fimbriae was examined under TEM in 24-h V. tapetis grown in TSAS. Briefly, the samples were fixed with 2.5% glutaraldehyde in 0.01 m cacodylate buffer (pH Adhesion of Vibrio tapetis 93 7.2) for 2 h at 4°C. Then, they were stained with 1% uranyl acetate (pH 4.5) for 45 s on copper grids (400 mesh) covered with form- var. dried, and examined under TEM. Outer Membrane Protein Analyses Outer membrane protein (OMP) analyses were carried out by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). following the method described previously by Cas- tro et al. ( 1996). Briefly, bacterial cells were disrupted by sonica- tion (in an ice bath with 4 pulses of 30 s at 50 W). and the cellular envelopes were sedimented by centrifugation (100.000 x g for 1 h at 4°C). Cytoplasmic membranes were selectively solubilized with sodium lauryl sarcosinate (Sarkosyl, Sigma), and outer membranes (OM) were sedimented by centrifugation as above. OM samples were electrophoresed in discontinuous polyacryl- amide-SDS gels (4.5-12.5%). using the Laemmli"s technique (1970). Proteins were detected in the gels by staining with Coo- massie brilliant blue (Sigma). TABLE 1. Determination of cell surface hydrophobicity of Vibrio tapetis strains using different methods RESULTS Hydrophobicity Different results have been obtained depending on the method used to estimate cell surface hydrophobicity of V. tapetis strains (Table 1 ). Most of the isolates (77.8%) aggregated in the presence of less than 1 M ammonium sulfate. Adhesion to nitrocellulose filters yielded the highest values of hydrophobicity for most of the bacterial isolates, percentages of adhesion ranging between 94.9 and 99.9%. A variable degree of adhesion to n-hexadecane has been observed, percentages of adhesion ranging between 20. 1 and 64.6% when PUM buffer was used an aqueous phase and between 21.4 and 63.8% when PBS was used. The criteria of hydrophobicity proposed by Santos et al. ( 1990) and Lee and Yii (1996) were used to evaluate the hydrophobicity of the V. tapetis strains tested (Table 2). According to the criteria applied, 11.1 and 40.8% of the isolates were highly hydrophobic with MATH assay, using PUM and PBS buffers, respectively. However, 77.8% of the strains showed strong hydrophobicity with SAT and 100% with NCF assay (Table 2). Most of the isolates were included in the group of moderate hydrophobicity with MATH (88.9% for PUM buffer, and 59.2% for PBS buffer), and only 22.2% of the strains were included using the SAT test. Kendall rank coefficients have shown the presence of signifi- cant correlation (p < .05) between MATH assays in the presence of PUM and PBS buffers (p = .026). These findings suggest that adhesion to n-hexadecane is not significantly influenced by the use of PUM or PBS buffers as aqueous phase. On the contrary, no significant correlation was obtained between the rest of tests as- sayed (p > .05), with significance degree of p = .393 for MATH (PUM) versus SAT, p = .884 for MATH (PUM) versus NCF. p = .440 for MATH (PBS) versus SAT. p = .491 for MATH (PBS) versus NCF, and p = .348 for SAT versus NCF. Cell Adhesion Assays V. tapetis adhesion to different cell systems, such as clam hemocytes, mantle cells of clams, and fish cell lines, has been evaluated by two different approaches, the microscopic determi- nation of cellular adhesioin percentages and by an EL1SA (Tables 3 and 4). The results obtained varied depending both on the cel- lular system used and the V. tapetis strains tested. Thus, all the V. MATH" NCFh Strains PUM PBS SATC V. tapetis 2.1 47.5 61.0 99.1 0.50 V. tapetis 2.3 44.9 60.0 99.2 1.00 V. tapetis 8.1 47.3 58.9 99.9 0.46 V. tapetis 8.3 42.9 63.8 99.5 0.93 V. tapetis 8.4 32.2 55.7 99.5 0.70 V. tapetis 8.5 35.4 62.6 99.1 0.46 V. tapetis 8.6 22.0 22.3 99.5 0.80 V. tapetis 8.7 29.6 37.6 99.4 0.70 V. tapetis 8.17 24.8 35.4 99.9 1.20 V. tapetis 8.19 51.3 48.5 99.4 0.80 V. tapetis 9.3 54.0 53.3 99.4 0.70 V. tapetis 9.4 30.4 44.3 99.9 0.93 V. tapetis 9.5 46.2 53.7 99.7 0.66 V. tapetis 9.7 46.0 58.0 99.4 0.93 V. tapetis 11.1 64.7 50.0 99.1 1.00 V. tapetis 11.2 26.0 21.4 99.9 0.80 V. tapetis 11.4 29.1 56.3 97.9 0.93 V. tapetis IS- 1 20.1 37.4 94.9 0.86 V. tapetis IS-5 46.9 33.5 99.4 0.80 V. tapetis IS-7 9.7 44.8 98.1 0.83 V. tapetis IS-8 27.0 35.6 99.8 0.80 V. tapetis IS-9 46.2 50.7 99.9 0.80 V. tapetis CECT 4600T 20.7 46.5 99.6 1.00 V. tapetis 1703 23.3 31.4 98.9 0.80 V. tapetis 6301 46.0 42.0 99.5 0.66 V. tapetis 0202 40.9 44.5 99.1 1.00 V. tapetis 0705 26.2 44.4 99.0 1.00 J Percentage of adherence to n-hexadecane. b Percentage of adherence to nitrocellulose filters. c Lowest molarity of ammonium sulphate producing visible aggregation. CECT: Spanish Type Culture Collection. 1 Type strain. tapetis strains did not adhere to the fish cell lines CHSE and EPC. but the adhesion percentage to SAF- 1 cells ranged from 2 to 96% (Table 3). These values contrast with the values obtained for clam cells (hemocytes and mantle cells), varying between 68 and 100% (Table 3). Only two V. tapetis strains (8.6 and 0202) showed similar adhesion rates, regardless of the type of cells used. No significant differences (p > .05) were obtained in the adhesion capability of V. tapetis strains depending on the origin of the clam cells used, except in the case of the strain 1 1 .2 for hemocytes of T. decussatus and T. philippinamm (89% vs. 68%), and the strains 2.1 (76% vs. 100%) and CECT 4600T (77% vs. 97%) for mantle cells of both clam species. Mean numbers of adhered bacteria per cell are also given in Table 3. SAF-1 cell line proved to be a poor system for V. tapetis adhesion, with mean values ranging between 1 .5 and 6.8 adhered bacteria per cell. In contrast, mantle cells and clam hemocytes were better matrix systems for V. tapetis adhesion, ranging be- tween 5.8 and higher than 25 adhered bacteria per hemocyte. and between 10.4 and 22.8 adhered bacteria per mantle cell. In this case, no significant differences were detected between the origin of the cells (source and species), although significant differences were obtained between the V. tapetis strains tested, strain 11.2 being the least adherent to hemocytes (Table 3). 94 LOPEZ-CORTES ET AL. TABLE 2. Hydrophobicity degree of Vibrio lapetis strains using SAT, NCF, and MATH assays according to the criteria proposed by Santos et al. (1990) and Lee and Yii (1996) Hydrophobicity Percentage of Test Values degree V. lapetis Isolates SAT <1.0M Strong 77.8 1.1-2.0 M Moderate 22.2 2.1-1.0 M Weak 0 >4.0M Negative 0 NCF >75% Strong 100 50-75% Moderate 0 <50% Negative 0 MATH >50% Strong 40.8 (PBS)J 20-50% Moderate 59.2 <20% Negative 0 MATH >50% Strong 11.1 (PUM)b 20-50% Moderate 88.9 <20% Negative 0 1 PBS used as aqueous phase. h PUM buffer used as aqueous phase. Table 4 expresses the results of V. tapetis adhesion using an ELISA technique. In agreement with the results given in Table 3. V. tapetis strains possessed a low adhesion rate for SAF-1 cells, and even strain IS-7 was nonadherent. Adhesion to clam cells varied depending on the strains considered: thus, significant dif- ferences were found between the adhesion to clam hemocytes of both clam species for V. tapetis CECT 4600T. No correlation was obtained between the hydrophobicity of the strains and their adhesion to several cells, except in the case of MATH (PUM) hydrophobicity and adhesion to hemocytes of T. philippinarutn determined by ELISA (data not shown). Hemagglutination and V. tapetis Appendages None of the V. tapetis strains presented hemagglutination of horse, rabbit, and human erythrocytes; on the contrary, all of them agglutinated rat erythrocytes, and the hemagglutination was not inhibited by any of the sugars tested at the different concentrations assayed. The presence of appendages on the V. tapetis surface was de- termined by TEM. In all the V. tapetis strains tested, the presence of a sheathed polar flagellum was recorded (Fig. 1 1, and sometimes several lateral flagella were also observed. Piliation of the strains or fimbria-like structures were not observed in the bacterial cul- tures tested. However, visualization of isolated and purified pili has not been performed yet. Analysis of the Cellular Components of the Outer Membrane The electrophoretic analyses of OMP showed that all V. tapetis strains tested, except the strain 0202. present the same band pat- tern, expressing proteins of molecular weight ranging between 78 and 15 kDa. The profile of these strains is dominated by a major outer membrane protein (MOMP) of an estimated molecular weight (MW) of 35 kDa. In the case of the strain 0202 the MOMP was of 37 kDa. and presented a high MW protein of 94 kDa (Fig. 2). DISCUSSION In BRD. an epizootic disease affecting cultured clam species, mainly T. philippinarum, V. tapetis is predominantly detected on the clam periostracal lamina (Allam et al. 1996). The capacity of this pathogen to adhere to periostracum is obviously an essential step for the bacterial colonization. However, the mechanisms by which bacterial cells adhere to this clam surface has not been elucidated completely. One hypothesis for the mechanisms by which V. tapetis adheres to clam tissues involves the concept that filamentous appendages characterized as pili (Paillard and Maes 1995a) bind the cells to periostracal lamina. However, the presence of these bacterial appendages were visualized only in some colo- nizing V. lapetis in diseased clams (Borrego et al. 1996a). On the other hand, Arp ( 1988) pointed out that before the bacterial adhe- sion, it is necessary for the bacteria to maintain their position along a mucosal surface by establishing small numbers of noncovalent bonds between the bacterial and mucosal surfaces. These bonds depend on several physicochemical mechanisms, the hydrophobic interactions being the most important (Rosenberg and Kjelleberg 1986). For these reasons, we suggest that the adhesion of V. tapetis to clam tissues may be governed by two different but complemen- tary mechanisms: ( 1 ) physicochemical forces of adsorption; and (2) specific adhesion depending on adhesive bacterial structures TABLE 3. Adhesive capabilities of Vibrio tapetis strains to different cell systems Adhesion Percentages Mean of Adhered Bacteria per Cell SAF-1 Hemocytes Mantle Cells SAF-1 Hemocytes Mantle Cells Strains T.d" T.p" T.d T.p T.d T.p T.d T.p 2.1 ~> 88 91 76 100 1.9 6.6 11.2 21.0 21.3 8.6 96 100 94 88 98 6.8 11.7 10.9 19.0 17.6 9.4 25 98 94 90 99 5.1 >25 12.1 15.1 22.8 11.1 18 88 100 96 96 1.6 10.4 15.2 21.7 10.5 11.2 46 89 68 90 91 2.2 6.6 8.1 21.1 13.7 IS-7 58 89 99 94 98 1.5 >25 16.1 17.9 16.5 CECT 4600T 59 100 95 77 97 3.7 11.3 5.8 10.4 17.4 0202 96 100 100 90 97 6.6 >25 15.2 14.6 20.1 1 Tapes decussatus. ' Tapes philippinarum. Adhesion of Vibrio tapetis 95 TABLE 4. Adhesive capabilities of Vibrio tapetis strains to different cell systems using an ELISA technique SAF-1 Hemocytes Mantle Cells Strains T. decussatus r. philippinarum T. decussatus T. philippinarum 2.1 0.21 ' 0.23 0.24 0.28 0.34 8.6 0.19 0.22 0.25 0.28 0.33 9.4 0.18 0.23 0.24 0.28 0.31 111 0.17 0.24 0.24 0.24 0.30 11.2 0.15 0.20 0.26 0.26 0.30 IS-7 0.09 0.20 0.25 0.29 0.29 CECT 4600T 0.22 0.17 0.38 0.24 0.30 0202 NTh NT NT NT NT ' Absorbance units at 450 nm. ' NT: Not tested because of the lack of specificity of the antiserum against V. tapetis to strain 0202. that interlock in a stereospecific manner with complementary structures on the opposing surface. Cellular hydrophobicity is known to be associated with the capacity of microbial cells of many taxonomic groups to adhere to surfaces of numerous types, including those of animal tissues (Doyle and Rosenberg 1990). The results obtained in the present study demonstrate that even strains of V. tapetis closely related at biochemical, serological, and molecular levels (Borrego et al. 1996b, Castro et al. 1996, Castro et al. 1997) may vary signifi- cantly in surface hydrophobicity as measured by any of the three assays used (Table 1 ). As with other bacteria, consistency of re- sults among the three hydrophobicity assays was only observed in strains with very hydrophilic surfaces (Mozes and Rouxhet 1987, Sorogon et al. 1991). Differences in the distribution of hydropho- bic components on the bacterial surface may account for these disparities. Although the SAT may provide a measure of overall surface hydrophobicity. the MATH and NCF assays may indicate the presence of hydrophobic domains on an otherwise hydrophilic cell surface (van der Mei et al. 1987). The lower values of hydro- phobicity obtained for V. tapetis strains using the MATH assay as compared to SAT and NCF techniques has been reported previ- ously (Sorogon et al. 1991). These authors pointed out the possi- '7\ -n*A^ '» 4 . ■L *i ^^^ '• 1 fim bility that cell surfaces of the bacteria tested were modified by the hydrophobicity assay procedures. Thus, hexadecane used in the MATH assay may extract constituents from the bacterial envelope (Dillon et al. 1986. Rosenberg and Kjelleberg 1986). Methods for measuring bacterial hydrophobicity differ some- what in the precise properties they measure, and different types of interactions are considered when different methods to estimate hydrophobicity are used. Thus, Dickson and Koohmaraie (1989) observed that the relative hydrophobicity estimated by hydropho- bic interaction chromatography. MATH and contact angle mea- surements for different bacterial species was dependent on the specific method tested. The results obtained in the present study for V. tapetis strains isolated from diseased clams show that the experimental conditions imposed by the different methods used influence the observed hydrophobicity interactions to some degree. Specific hydrophobicity assays may be useful predictors of adhesion for closely related strains of certain bacterial species (Martin et al. 1997). In the case of V. tapetis strains isolated from diseased clams, the results obtained in the hydrophobicity tests indicate that adhesion of these bacteria cannot be. explained by hydrophobic interactions alone. Rather, adhesion is likely to be mediated by an interplay of hydrophobic and hydrophilic surface components (Dickson and Siragusa 1994), or specific interactions 205 Figure 1. Electron micrograph of negative stained Vibrio tapetis CECT 4600 cells, showing the presence of a single sheathed polar flagellum. 94 67 43 30 Figure 2. Electrophoresis in SDS-polvacrvlamide gels of the outer membrane proteins from several Vibrio tapetis strains; lanes 1 to 5, strains CECT 4600T, 2.1, 11.2, IS-7, and 0202. respectively. In lanes M, the molecular weight (in kDa) of standard proteins is indicated. 96 LOPEZ-CORTES ET AL. between bacterial adhesins and cellular receptors (Christensen et al. 1985). For many pathogenic bacterial strains, mucosal attachment is mediated by specific bacterial surface appendages. Pili are consid- ered as the most relevant adhesins and colonization factors of host tissue surface. Many types of bacterial pili have been recognized in Vibrio species, such as toxin-coregulated pili (TCP), mannose- sensitive hemagglutinin (MSHA) pili, core-encoded pili (Cep), ac- cessory colonization factor (Acf), NAGV 14, Na2, and Ha7 pili (Yamashiro and Iwanaga 1996). In the present study, we have not detected the presence of these appendages on V. lapetis surface, under TEM examination (Fig. 1). However, all V. tapetis cells examined showed the presence of a sheath flagellar structure, which may be important for attachment to host tissues (Attridge and Rowley 1983). In contrast. Paillard and Maes ( 1995a) reported that fimbria-like appendages were present in the V. tapetis colo- nizing the periostracal lamina of diseased clams. These contradic- tory results lead us to suggest the role of an unknown factor that promotes the fimbria synthesis in V. tapetis infection. Iijima et al. (1981) suggested that V. parahaemolyticus synthetize a cytotoxic factor that degenerates epithelial cells of the host and promotes its adherence. On the other hand, nutrient-limiting conditions may enhance the bacterial adhesion (Dai et al. 1992) or induce the formation of adherent structures (McCarter and Silverman 1989, Nakasone and Iwanaga 1990). The importance of the flagellum as a potential virulence factor has been demonstrated for several bac- terial species. This structure has been involved in pathogenicity as either a motility organelle or an organelle that carries an adhesive component, both roles providing an advantage to the bacterium for its invasive capabilities (Norqvist and Wolf-Watz 1993, Milton et al. 1996). The agglutination of bacteria to erythrocytes has been proposed as an efficient in vitro system to demonstrate the bacterial adhesive activity (Santos et al. 1990) and as a system to characterize the type of adhesins (Evans et al. 1980. Zunino et al. 1994). None of the V. tapetis strains tested showed hemagglutination of horse, rabbit, and human erythrocytes, but all of them were positive for rat erythrocytes. This apparent contradiction in the hemagglutina- tion results obtained may be explained by the fact that the hem- agglutination depends on the presence of specific receptors on the erythrocyte surface, and such receptors contain oligosaccharides that varied depending on the animal species (Jones and Freter 1976). Previously, adhesion to clam cell systems by V. tapetis has not been demonstrated. Therefore, this study constitutes the first report of adhesion to hemocytes and mantle cells of two clam species. All the V. tapetis strains tested showed a higher degree of adhesion to clam cells, both hemocytes and mantle cells, than to fish cells (Table 3). These findings suggest the existence of a host or tissue specificity. According to Christensen et al. ( 1984), the adhesion of a particular bacterial species may vary considerably depending on the host species, physiology, phenotype. and tissue. Miller and Mekalanos (1988) described the role of two outer membrane proteins. OmpU and OmpT. as colonization factors of V. cholerae. Later. Sperandio et al. ( 1995) verified that OmpU acts as an adherence factor involved in the colonization of epithelial cells by V. cholerae. The amino-terminal amino acid sequence of OmpU was similar to the sequences of Haemophilus influenzae HMW1 and HMW2 adhesins. and shared also similarities with the Bordetella pertussis FHA. As it can be seen in Fig. 2. all the V. tapetis strains tested presented the same OMP profile, except for 0202 strain. The MW of OmpU (32-38 kDa) is similar to the major OMP (MOMP) detected in V. tapetis (35-37 kDa), which induces to speculate about a similar function of this protein in V. tapetis strains. To verify this hypothesis, further studies of adhesion to cells using isolated MOMP and inhibition with anti-MOMP anti- serum are necessary. In short, the adhesion mechanisms of V. tapetis are complex and depend on different processes that act in several steps. In a hypothetical model, V. tapetis is directed to specific substrate of the clam tissue by means of their motility organelles, which also help the bacterium adhere to clam cells. Then, the hydrophobic forces maintain the bacterial position along a mucosal surface by establishing small numbers of non-covalent bonds, and, finally, bacterial adhesins or surface proteins of V. tapetis interlock spe- cific attachment with complementary structures on the opposing surface. ACKNOWLEDGMENTS This study was supported by a grant from the Direction Gen- eral de Ciencia y Tecnologi'a (DGICYT) (No. PB-95-0467). We thank Miss M. J. Navarrete for her help in the English revision of the manuscript. LITERATURE CITED Allam. B., C. Paillard & P. Maes. 1996. Localization of the pathogen Vibrio PI in clams affected by brown ring disease. Dis. Aquat. Org. 27:144-155. Arp, L. H. 1988. Bacterial infections of mucosal surfaces: an overview of cellular and molecular mechanisms, pp. 3-27. In: J. A. Roth (ed.). Virulence Mechanisms of Bacterial Pathogens. American Society for Microbiology, Washington, DC. Attridge, S. R.. & D. Rowley. 1983. The role of flagellum in the adherence of Vibrio cholerae. J. Infect. Dis. 147:864-872. Borrego. J. J.. A. Luque. D. Castro. J. A. Santamaria & E. Martinez- Manzanares. 1996a. Virulence factors of Vibrio PI. the causative agent of the brown ring disease in the Manila clam. Ruditapes philippinarum. Aquat. Living Resour. 9:125-136. Borrego. J. J., D. Castro. A. Luque, C. Paillard. P. Maes. M. T. Garcia & A. Ventosa. 1996b. Vibrio tapetis sp. nov.. the causative agent of the brown ring disease affecting cultured clams. Int. J. Syst. Bacterial. 46:480-484. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 72:248-254. Bruno. D. W. 1988. The relationship between autoagglutination. cell sur- face hydrophobicity. and virulence of the fish pathogen Renibacterium salmoninarum. F EMS Microbiol. Lett. 51:135-140. Castro. D. 1994. Patologi'as de origen bacteriano en almejas cultivadas Tapes decussatus y T. phillipinarum. Ph.D. thesis. University of Malaga. Spain. Castro, D.. A. Luque, J. A. Santamaria, P. Maes. E. Maninez-Manzanares & J. J. Borrego. 1995. Development of immunological techniques for the detection of the potential causative agent of the brown ring disease. Aquaculture 132:97-104. Adhesion of Vibrio tapetis 97 Castro, D., J. A. Saniamaria. A. Luque, E. Martinez-Manzanares & J. J. Borrego. 1996. Antigenic characterization of the etiological agent of the brown ring disease affecting manila clams. Syst. Appl. Microbiol. 19:231-239. Castro, D.. J. L. Romalde, J. Vila. B. Magarinos. A. Luque & J. J. Borrego. 1997. Intraspecific characterization of Vibrio tapetis strains by use of pulsed-field gel electrophoresis, ribotyping, and plasmid profiles. Appl. Environ. Micribiol. 63:1449-1452. Christensen. G. D., W. A. Simpson & E. H. Beachy. 1984. Bacterial ad- herence in infection, pp. 6-23. In: G. L. Mandell, R. G. Douglas & J. E. Bennett (eds.). Principles and Practice of Infectious Diseases, 2nd ed. John Wiley & Sons. New York. Christensen. G. D.. W. A. Simpson & E. H. Beachy. 1985. Adhesion of bacteria to animal tissues: complex mechanisms, pp. 297-305. In: D. C. Savage and M. Fletcher (eds.). Bacteria Adhesion. Plenum Press, New York. Clark, R. B„ F. C. Knoop. P. J. Padgitt. D. H. Hu, J. D. Wong & M. J. Janda. 1989. Attachment of mesophilic aeromonads to cultured mam- malian cells. Cnrr. Microbiol. 19:97-102. Dai. J. Ft., Y. S. Lee & H. C. Wong. 1992. Effects of iron limitation on production of a siderophore, outer membrane proteins, and hemolysin and on hydrophobicity, cell adherence, and lethality for mice of Vibrio parahaemolyticus. Infect. Imniun. 60:2952-2956. Daly, J. G. & R. M. W. Stevenson. 1987. Hydrophobic and hemaggluti- naling properties of Renibacterium salmoninarum. J. Gen. Microbiol. 133:3575-3580. Dickson. J. S. & M. Koohmaraie. 1989. Cell surface charge characteristics and their relationship to bacterial attachment to meat surfaces. Appl. Environ. Microbiol. 55:832-836. Dickson. J. S. & G. R. Siragusa. 1994. Cell surface charge and initial attachment characteristics of rough strains of Listeria monocytogenes. Lett. Appl. Microbiol. 19:192-196. Dillon, J. K„ J. A. Fuerst, A. C. Hayward & G. H. G. Davis. 1986. A comparison of five methods for assaying bacterial hydrophobicity. J. Microbiol. Meth. 6:13-19. Doyle. R.J. & M. Rosenberg (eds.). 1990. Microbial cell surface hydro- phobicity. American Society for Microbiology. Washington. DC. Evans. D. J. Jr.. D. G. Evans. L. S. Young & J. Pitt. 1980. Hemagglutina- tion typing of Escherichia colt: definition of seven hemagglutination types. J. Clin. Microbiol. 12:225-242. Iijima. Y., H. Yamada & S. Shinoda. 1981. Adherence of Vibrio para- haemolyticus and its relation to pathogenicity. Can. J. Microbiol. 27: 1251-1259. Jones. G. W. & R. Freter. 1976. Adhesive properties of Vibrio cholerae: nature of the interaction with isolated rabbit brush-border membranes on human erythrocytes. Infect. Iininnn. 14:240-245. Jonson, G.. J. Holmgren & A. M. Svennerholm. 1991. Identification of a mannose-binding pilus on Vibrio cholerae El Tr. Microb. Pathog. 11: 433-141. Kumazawa. N. H.. T. Tanigawa, Y. Tanaka. H. Osatake & K. Tanaka. 1991. In vitro attachment of Vibrio parahaemolyticus to hemocytes of two gastropod mollusks. J. Vet. Med. Sci. 53:297-300. Lachica. R. V. & D. L. Zink. 1984. Plasmid-associated cell surface charge and hydrophobicity of Yersinia enterocolitica. Infect. Iininitn. 44:540-543. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685. Larsen. J. L. & S. Mellergaard. 1984. Agglutination typing of Vibrio an- guillarum isolates from diseased fish and from the environment. Appl. Environ. Microbiol. 47:1261-1265. Lee. K. K. & K. C. Yii. 1996. A comparison of three methods for assaying hydrophobicity of pathogenic vibrios. Lett. Appl. Microbiol. 23:343-346. Lindhal. M.. A. Fans. T. Wadstrom & S. Hjerten. 1981. A new test based on "salting out" to measure relative surface hydrophobicity of bacterial cells. Biochim. Biophys. Acta 677:471-476. Martin, M. L., Y. Benito. C. Pin, M. F. Fernandez, M. L. Garcia. M. D. Selgas & C. Casas. 1997. Lactic acid bacteria: hydrophobicity and strength of attachment to meat surfaces. Lett. Appl. Microbiol. 24:14-18. McCarter. L.. M. Silverman. 1989. Iron regulation of swamer cell differ- entiation of Vibrio parahaemolyticus. J. Bacterial. 171:731-736. Miller. V. L. & J. J. Mekalanos. 1988. A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer mem- brane proteins and virulence determinants in Vibrio cholerae requires toxR. .1. Bacterial. 170:2575-2583. Milton. D. L.. R. O'Toole, P. Horstedt, & H. Wolf-Watz. 1996. Flagellin A is essential for the virulence of Vibrio anguillarunu J. Bacterial. 178: 1310-1319. Mozes, N. & P. G. Rouxhet. 1987. Methods for measuring hydrophobicity of microorganisms. J. Microbiol. Melli. 6:99-112. Nakasone. N. & M. Iwanaga. 1990. Pili of Vibrio cholerae non-Ol. Infect. biiniun. 58:1640-1646. Norqvist, A. & H. Wolf-Watz. 1993. Characterization of a novel chromo- somal virulence locus involved in expression of a major surface flagel- lar sheath antigen of the fish pathogen Vibrio anguillarum. Infect. In, mini 61:2434-2444. Novoa, B.. A. Luque, D. Castro, J. J. Borrego & A. Figueras. 1998. Char- acterization and infectivity of four bacterial strains isolated from brown ring disease-affected clams. J. Invertebr. Pathol. 71:34 — i\. Paillard. C. & P. Maes. 1990. Etiologie de la maladie de I'anneau brun chez Tapes philllipinaruni: pathogenicite d'un Vibrio sp. C.R. Acad. Sci. Paris 310:15-20. Paillard. C. & P. Maes. 1995a. The brown ring disease in the Manila clam. Ruditapes philippinarum. I. ultrastructural alterations of the periostra- cal lamina. J. Invertebr. Pathol. 65:91-100. Paillard, C. & P. Maes. 1995b. The brown ring disease in the Manila clam. Ruditapes philippinarum. 11. microscopic study of the brown ring syn- drome. J. Invertebr. Pathol. 65:101-1 10. Rosenberg. M.. D. Gutmck & E. Rosenberg. 1980. Adherence of bacteria to hydrocarbons: a simple method for measuring cell-surface hydro- phobicity. FEMS Microbiol. Lett. 9:29-33. Rosenberg. M. & S. Kjelleberg. 1986. Hydrophobic interactions: role in bacterial adhesion. Adv. Microb. Ecol. 9:353-393. Santos, Y.. I. Bandin. T. P. Nieto, D. W. Bruno. A. E. Ellis & A. E. Toranzo. 1990. Comparison of the cell surface hydrophobicity of bac- terial fish pathogens by different procedures, pp. 101-115. In: F. O. Perkins and T. C. Chen (eds.). Pathology in Marine Science. Academic Press. New York. Santos, Y., 1. Bandin. T. P. Nieto, J. L. Barja, A. E. Toranzo & A. E. Ellis. 1991. Cell-surface-associated properties of fish pathogenic bacteria. J. Aquat. Amm. Health 3:297-301. Savage. D. C. 1992. Growth phase, cellular hydrophobicity. and adhesion in vitro of lactobacilli colonizing the keratinizing gastric epithelium in the mouse. Appl. Environ. Microbiol. 58:1992-1995. Sorongon, M. L„ R. A. Bloodgood & R. P. Burchard. 1991. Hydrophobic- ity. adhesion, and surface-exposed proteins of gliding bacteria. Appl. Environ. Microbiol. 57:3193-3199. Sperandio. V., J. A. Giron. W. D. Silveira & J. B. Kaper. 1995. The OmpU outer membrane protein, a potential adherence factor of Vibrio cho- lerae. Infect. Iiiiiiiiin. 63:4433-4438. van der Mei. H. C, A. H. Weerkamp & H. J. Busscher. 1987. A compari- son of various methods to determine hydrophobic properties of strep- tococcal cell surfaces. J. Microbiol. Meth. 6:277-287. Vazquez-Juarez, R., T. Andlid & L. Gustafasson. 1994. Cell surface hy- drophobicity and its relation to adhesion of yeasts isolated from fish gut. Colloids Surf. 2:199-208. Yamashiro. T. & M. Iwanaga. 1996. Purification and characterization of a pilus of a Vibrio cholerae strain: a possible colonization factor. Infect. biimiin. 64:5233-5238. Zunino. P.. C. Piccini & C. Legnani-Fajardo. 1994. Flagellate and non- flagellate Proteus mirubilis in the development of experimental urinary tract infection. Microb. Pathog. 16:379-385. Journal of Shellfish Research, Vol. IX. No. 1. 99-105, 1999. ASSESSING MANIPULATIONS OF LARVAL DENSITY AND CULLING IN HATCHERY PRODUCTION OF THE HARD CLAM, MERCENARIA MERCENARIA CARRIE J. DEMING AND MICHAEL P. RUSSELL* Biology Department Villanova University Villanova, Pennsylvania 19085-1699 ABSTRACT Studies of the hard clam. Mercenaria mercenaria, indicate that the hatchery practice of larval culling may be coun- terproductive because of an inverse relationship in growth between larvae and postsettlement juveniles. The effects of larval culture manipulations were explored with two parallel studies: one set up in a commercial hatchery and the other in the laboratory. Both laboratory and hatchery studies started with the same cohort of larvae produced from spawning hatchery broodstock. Except for culling, these larvae were raised using standard practices in the hatchery for the first 10 days. On the tenth day of development, the cohort was sieved through 105-p.m mesh and separated into two treatments, large larvae and small larvae. In the hatchery, these samples were followed through settlement, and growth was monitored for 1 74 days postfertilization. In the laboratory, the two samples of larvae were raised under high-density (20 larvae/mL) or low-density (4 larvae/mL) conditions and thus assigned to one of four larval treatments: large/high-density, large/low-density, small/high-density. and small/low-density. These treatments were replicated (10 times each) to yield a 2 x 2 factorial, randomized block, repeated-measures analysis of variance (ANOVA) experiment. Growth in the laboratory was monitored for 276 days. Both the hatchery and laboratory studies indicate that larval culling does not increase productivity and may be counterproductive. Furthermore, larval density has an effect on subsequent juvenile growth. Larvae raised at low density produced larger clams in both the small and large larval treatments. KEY WORDS: Mercenaria mercenaria. hard clam, larval culling, aquaculture. growth INTRODUCTION ... it is reasonable to question the merits of the larval culling practices carried out worldwide in bivalve hatcher- ies . . . a substantial proportion of the lar\>▼ Larval Corral Downweller Upweller Figure 1. Experimental units. Each replicate was a 15-cm segment of PVC pipe with three 4-cm sections cut out of the bottom to form a tripod support. As the cohort developed, the size of the mesh on the bottom was increased (Table 1) to enhance flow within the approximately 8(10 ml. volume held by each unit. The units are drawn to scale; however, the sea table and upwelling table are not — all 40 experimental units were in the sea table and then moved to the upwelling table (which was held in the sea table). The thick arrows represent air and water flow. Flow rates were monitored regularly and adjusted to no less than 1 L/min in the downwellers and upwellers. A. Larval Corral. Filtered air entered the bottom of the experimental units through a 2-mm plastic tube positioned approximately 2 cm from the bottom and was regulated to I bubble/s. Seawater circulated within the unit and between the unit and the sea table through the bottom mesh ( 105 fan). B. Downweller. Seawater entered through a 1-cm tygon tube positioned directly above each experimental unit. C. Upweller. All 40 experimental units were placed in the same blocked arrangement, in an upwelling table nested within the sea table. Each unit was plumbed near the top with a sealed tygon connector attached to the upwelling table. Seawater tlowed into the upwelling table, up through the bottom of each unit, and out to the sea table via the tygon connector. 102 Deming and Russell ization) and used as a covariate in the repeated-measures ANOVA to account for the effects of juvenile density on growth. In addi- tion, an ANOVA was performed on the means of SL from each replicate on the last set of measurements (day 276 postfertilization) with average postsettlement density as the covariate. Differences in survivorship were tested using a randomized block ANOVA on the number of individuals in each replicate on the last sampling date (start density / final density). All analyses were performed using SAS (1989). RESULTS There was a significant difference (p < .0001. / = 18.06) in SL between the large and small larval treatments on day 10 postfer- tilization (Fig. 2A). On this date, approximately 5% of the culture was in the pediveliger stage of development, and all pediveligers were part of the large larval treatment. Settlement occurred earlier in the hatchery-reared clams (days 17-20 postfertilization) than in the laboratory experiment (days 23-25 postfertilization. Table 1 ). Although we tracked the growth of clams in the hatchery phase of the experiment (Fig. 3). we did not compare the two groups sta- tistically, because this part of the experiment was pseudoreplicated (Hurlbert 1984). At the end of the hatchery study ( 174 days post- fertilization), there did not seem to be a difference in SL between the two larval size groups; the mean size of clams from the large larval treatment was 4.21 mm. and the mean size of clams from the small larval treatment was 4.31 mm (Fig. 2B). The repeated-measures ANOVA revealed that there were no significant differences in SL among the four treatments attribut- able to larval size or larval density (0.10 >p> .05. Table 2) over the course of the 276-day experiment. However, the ANOVA from the last sample revealed that although larval size was not signifi- cant (p = .22, F = 1.69) larval density was significant (p = .02. F = 7.65). In both the large and small larval treatments, larvae raised at lower densities produced bigger juveniles at the end of the experiment on day 276 postfertilization (Fig. 2C). This difference in growth cannot be attributed to differences in postsettlement density (p = .39. Table 2). For most of the experiment, postsettle- ment juvenile density was highest in the large/low-density larval treatment (Fig. 4). and this treatment produced the biggest juve- niles (Fig. 2C). Figure 4 shows the average number of individuals per replicate in the four treatments over the course of the experiment. Mortality levels were highest among all treatments between days 21 and 42 postfertilization. For most of the experiment (days 42-270). sur- vivorship remained constant among all four treatments. Mortality was highest in the high larval density treatments and lowest in the large/low-density larval treatment (p < .0001, F = 4.12, Fig. 4). Larval Separation (June 15. 1996) 0.185 0.180 - bb 0.175 c "S 0.170 C/3 0.165 - 0.160 * 4.5 4.4 4 3 42 - 4.1 4.0 B C Hatchery Laboratory (November 26. 1996) (March 15. 1997) 6.00 -i 5.75 - 0 5.50 5.00 - 4.75 0 A 6 Small Large Small Large Small Large Larval Size Figure 2. Means (± 2 standard errors) of shell lengths at the beginning and end of the hatchery and laboratory phases of the study. The circles represent the small larval treatment; larvae that passed through the 105-u.m mesh and were caught on the 70-pm mesh on June 15, 1996. The triangles represent the large larval treatment: larvae caught on the 105 um mesh. A. Initial separation. These samples (n = 200 each for large and small) are significantly different (p < .0001, r-test). B. Final measurements of hatchery-reared samples on day 174 postfertilization (n = 200 each for large and small). No statistical analysis was performed on the hatchery-reared samples, because the design of this experiment was pseudoreplicated (Hurlbert 1984. see text). C. Final measurements of laboratory samples. These values are means of replicates (n = 10 for small larvae at low density, and n = 9 for the other three treatments, see text). There were 25 clams measured from each replicate. The open symbols (Lo) represent larvae raised at low density (4/mL) and the shaded symbols (Hi) represent larvae raised at high density (20/mL). Within larval size treatment, the low -density larval treatments produced larger postsettlement juveniles. There was no difference in SL associated with larval size treatment (p < .05, see text) on this date. Larval Hard Clam Density and Culling 103 4.0 - E 3-° I 00 a 1.0 T J T T ■ ( < < T j- - t' • S k ■ 3 **** * 25 45 65 85 105 125 Age (days post-fertilization) Figure 3. Means and standard deviations (±1) of shell lengths from the two experimental groups raised in the hatchery between the initial separation and the final measurements at the end of the experiment. The circles represent the small larval treatment and the triangles represent the large larval treatment (n: 2(10-1,600). DISCUSSION Our results are consistent with previous work showing either no correlation or a negative correlation between larval and juvenile growth rates in mollusks (e.g.. Newkirk and Haley 1982. Newkirk and Haley 1983. Stromgren and Nielsen 1989. Pechenik et al. 1996). Hilbish et al. (1993) found no evidence for "positive genetic- covariation between larval and juvenile growth" in M. mercenaria and recommended against larval culling as a method for enhancing juvenile growth and improving hatchery productivity (Hilbish et al. 1993. p. 102). In contrast, Heffernan et al. ( 1991 ) found a strong negative relationship between larval and juvenile growth, but they also questioned "the merits of hatchery culling practices for smaller larvae" (Heffernan et al. 1991. p. 199). We did not address the issue of heritability of growth characteristics in this study. Instead, we focused on the practical issues and consequences of 15 o o 3 x o "el c S S 9 - left y-axis right y-axis ► 10 14 21 30 42 -//- 1000 Slid ■600 400 200 ns u c 0> a. .05). Missing data for group II are attributable to oysters lost (see Materials and Methods). Comparative Growth of Oysters Reared at Low and High Salinity hi TABLE 2. Probabilities obtained with two-way ANOVA tests on data for condition indices and glycogen concentrations values of eastern oysters in relation to site and culture method, for each of four groups (I: 2-year-old oysters in 1993, II: 4-year-old oysters in 1993, III: 2-year-old oysters in 1994, and IV: 4-year-old oysters in 19941. Probabilities Before Spawning After Spawning Before Winter Group Group C roup Factors I II III IV I II III IV I II III IV Condition Indices Site .040* .006* .006* .019* .571 .012* .102 .002* .004* .000* .000* Culture method .002* .000* .797 .003* 1.000 .017* .043* .017* .017* .801 1.000 Site* method .100 .114 .512 .517 .909 .352 Glycogen Concentrations .602 .263 .077 .339 .844 Site .188 .000* .092 .737 .027* .001* .001* .000* .000* .006* .000* Culture method .001* .197 522 .015* .187 .031* .032* .248 .000* .110 .660 Site* method .087 .125 .934 .002* .203 .010* .889 .624 .009* .045* .511 ■ Significant difference (p < .05). data not available. collected before winter revealed that condition indices were higher for oysters sampled from stations HSS than for those from stations HSB. LSS. and LSB, all of which were similar to each other. The condition indices were higher for oysters sampled from high- salinity stations (HSS and HSB) than for the low-salinity stations (LSS and LSB) for groups III and IV oysters. DISCUSSION We focused on similarities and differences in growth and body condition between 2- and 4-year-old oysters cultured in suspension and on the bottom in a high- and a low-salinity environment during two consecutive growing seasons. Indices employed in our work. 250 250 r Figure 5. Glycogen concentrations in oysters sampled from stations HSS I high-salinity surface I, HSB (high-salinity bottom), LSS (low -salinity surface), and LSB (low-salinity bottom) for groups I (2-year-old oysters in 1993). II (4-year-old oysters in 1993), III (2-year-old oysters in 1994), and IV (4-year-old oysters in 1994), before (BS) and after spawning (AS) and before winter (B\V). Sampling dates for each group are shoyyn in Table 1. Bars with similar letters are not significantly different from each other (p > .05). Missing data for group II are attributable to oysters lost (see Materials and Methods). 112 Bataller et al. Figure 6. Condition indices for oysters sampled from stations HSS thigh-salinity surface), HSB {high-salinity bottom). LSS (low-salinity surface), and LSB (low-salinity bottom) for groups I (2-year-old oysters in 1993), II (4-year-old oysters in 1993), III (2-year-old oysters in 1994). and IV (4-year-old oysters in 1994), before (BS) and after spawning (AS) and before winter (B\V). Sampling dates for each group are shown in Table 1. Bars with similar letters are not significantly different from each other (p > .05). Missing data for group II are attributable to oysters lost (see Materials and Methods). such as growth rate, glycogen concentration, and condition index have been used in the past to gauge the relative growth and health of oysters (Muniz et al. 1986. Brown and Hartwick 1988. Little- wood and Gordon 1988, Austin et al. 1993. Boghen et al. 1993). Our work is consistent with those of others who demonstrated that oysters cultured in suspension display superior growth as com- pared to other growout methods. Although this proved to be the case at each site, when the sites were grouped together, we noted that oysters cultured on the bottom at high-salinity (HSB) dis- played growth comparable to those reared near the surface at low- salinity (LSS). Salinity and temperature, acting either independently or in syn- ergy, are known to have greater effects on growth of the eastern oyster than do other factors (Butler 1949. Wells 1961. Alderdice 1972, Vernberg and Vernberg 1972). This, however, does not pre- clude the fact that such environmental variables as seston and primary production, may, either in combination with or separate from salinity and temperature, have an important effect on the physiology and growth of the eastern oyster ( Bayne and Newell 1983. Brown and Hartwick 1988. Dekshenieks et al. 1993. Shum- way 1996). The significant difference that has been observed between years for the PIM/POM. may partially explain the differences be- tween 1993 and 1994 for certain recorded values, such as growth, condition index, or glycogen content ratio (Galtsoff 1964, Bayne and Newell 1983, Wallace and Reinsnes 1985). Although there were no significant differences in primary pro- duction among stations, this does not preclude the fact that the nutritional value of individual species of phytoplancton occurring at a given site or station at any particular time, may. in the long run. represent a more critical factor in determining the suitability of a particular culture location (Haven 1960. Dunathan et al. 1969. Castell andTrider 1974). Establishment of an halocline. persisting for up to 4 weeks, in the high-salinity water column during May and June for HSS and HSB (Fig. 3 and Table 1 ). is attributable to the effects of succes- sive periods of melting snow and heavy rain (Environment Canada 1993. 1994). Although better growth is associated with higher salinity (Butler 1949. Chanley 1958, Galtsoff 1964) as observed at HSB. water temperature was lower (Fig. 3. points 1 to 5). and may. therefore, have had an opposite effect on oyster growth. The dif- ference in temperature between bottom and surface at HSB and HSS persisted during the early growing period, which may help explain the reason for better growth at HSS over the growing season. Salinity fluctuations (5-1 59^) were evident at LSB at the end of May and beginning of June (Fig. 3). and this may be attributable to heavy precipitation and tidal action. From early August to mid-October, a phase representing opti- mal growing conditions for oysters (Shumway 1996). all four sta- tions (HSS. HSB. LSS. and LSB) displayed comparable tempera- ture and salinity profiles. The 1994 data demonstrated tendencies similar to those reported for 1993, although differences in weather patterns resulted in a slight shift in temperature and salinity varia- tions. In all instances. 2-year-old oysters displayed superior growth to 4-year-olds, similar to findings reported by Carriker ( 1996). Interpretations of findings related to glycogen concentrations and condition index (dry meat weight/dry shell weight), were con- sidered for the three critical periods during the oysters' growing season: before spawning, after spawning, and before winter, as previously mentioned. For this study, animals were sampled at Comparative Growth of Oysters Reared at Low and High Salinity 113 specific dates on a biweekly basis. Therefore, it is possible that the data reported for glycogen and condition index do not necessarily represent the absolute maximum and minimum values. This may help to explain differences in recorded values between 1993 and 1994. The less favorable environmental conditions (Fig. 3) for oysters at stations LSS and LSB for groups I and II after spawning may be responsible for their reduced capacity to rebuild glycogen reserves and their increased vulnerability at a critical time during their growing cycle. In general, however, oysters grown in suspension at HSS and LSS, during the same period, displayed higher glycogen concentrations than those cultivated on the bottoms at HSB and LSB, respectively. This is consistent with our growth data. Results of the condition indices measured for oysters sampled from all stations for groups I to IV generally support the findings recorded for glycogen and growth (Fig. 6). One notable exception was that the condition index proved to be inconsistent with the glycogen data for oysters cultivated in the low-salinity environ- ment. This index may prove to be an effective tool that could correlate positively with reported glycogen concentrations, de- pending upon environmental conditions (Ingle 1949. Walne 1970, Gabbott and Stephenson 1974). The absence of a detectable rela- tionship between the condition index and glycogen concentration or even growth in certain instances may be attributable to several factors, ranging from low calcium levels in less saline waters to poor substrates and/or inferior quality and availability of food. These factors may, in tum, contribute to improper shell formation and may alter shell form and thickness, thus biasing the relevancy of condition index as based on conventional methods (Riley and Chester 1971, Wilbur and Saleuddin 1983). Moreover, a pro- nounced asynchronism in the growth rate of shell versus soft tissue may likewise influence the accuracy of interpretation of values for condition index, as has previously been demonstrated for mussels (Hilbish 1986, Rainer and Mann 1992). Various authors (Lucas and Beninger 1985, Brown and Hart wick 1988, Rainer and Mann 1992) demonstrated that static indices based on the ratio of dry meat weight/shell cavity volume or dry meat weight/shell weight are less efficient than such dy- namic indices as biochemical indicators. The value of such bio- chemical indicators as glycogen concentration, carbohydratemitro- gen, and carbon:nitrogen have been discussed by Mann (1978). Our results demonstrate that oysters grown in suspension under high-salinity conditions display growth and development superior to those cultured in a low-salinity environment. Despite the supe- riority of oyster growth recorded at station HSS. the effectiveness of oyster culture in less saline waters as determined from our study should not be overlooked. Depending upon the specific environ- mental and hydrographic conditions characterizing a given site, bottom culture may be comparable to surface culture, if not more appropriate. This possibility is supported by Newell et al. (1998). who demonstrated that the contribution of detritus as a source of nutrient for mollusks may be more important then the phytoplank- ton available in the upper portions of the water column. Such an outcome may explain the reason for superior growth of bottom- cultured oysters versus oysters grown in suspension. A comparable situation may explain the similarity in growth of oysters from stations LSS and HSB. The implications of our findings lend credence to the potential advantages of identifying new aquaculture sites, which may have been rejected for commercial oyster culture up to now. The im- portance of such studies is reinforced by the decreasing availability of traditional culture sites in Atlantic Canada because of excessive coastal development and enhanced organic pollution. Finally, the appropriateness of certain sites may be more ap- plicable for the culture of one age group versus another, as re- ported by Ortega and Sutherland 1992. Alternatively, less tradi- tional sites may also prove to be useful as secondary or provisional storage areas for established operations, particularly at certain times of the year. ACKNOWLEDGMENTS We are grateful to Drs. N. Bourne. R. Lavoie, and G. Miron for reviewing this paper and for their helpful suggestions. Cooperation provided by Aquaculture Acadienne Inc. and the Big Cove Band Council is much appreciated. We also thank C. Mallet for his assistance with some of the technical drawings. This work was partially financed by the Faculties of Graduate Studies of the Uni- versite de Moncton (Moncton, N.B.) and University of New Brunswick (Fredericton, N.B.) through grants awarded to the se- nior author. Financial assistance was also provided by the New Brunswick Department of Fisheries and Aquaculture. This study is part of the Richibucto Environment and Resource Enhancement Program. LITERATURE CITED Alderdice. D. F. 1972. Factor combinations, responses of marine poikilo- therms to environmental factors acting in concert, pp. 1659-1772. In: O. Kinne (ed.). Marine Ecology. Wiley-Interscience, London. Austin. H. D. S. Haven & M. S. Moustafa. 1993. The relationship between trends in a condition index of the American oyster, Crassostrea vir- ginica, and environmental parameters in three Virginia estuaries. Es- tuaries 16:362-374. Bayne. B. L. & R. C. Newell. 1983. Physiological energetics of marine mollusks. pp. 407-515. In: A. S. M. Saleuddin and K. M. Wilbur (eds.). The Mollusca. vol. 4. Physiology, Part 1. Academic Press, Lon- don. Boghen, A. D.. J. Allard & E. Bataller. 1993. Rapport final et recomman- dations sur le programme de monitoring pour la cote est du Nouveau- Brunswick. Centre de Recherche en Sciences de 1' Environment. Uni- versite de Moncton. Moncton. N.-B. mars 1993. Brown. J. R. & E. B. Hartwick. 1988. Influences of temperature, salinity, and available food upon suspended culture of the Pacific oyster, Cras- sostrea gigas. II. condition index and survival. Aquaculture 70:253- 267. Butler. P. A. 1949. Gametogenesis in the oyster under conditions of de- pressed salinity. Biol. Bull. 96:263-269. Carr. R. S. & J. M. Neff. 1984. Quantitative semi-automated enzymatic assay for tissue glycogen. Comp. Biochem. Physiol. 77B:447^449. Carriker. M. R. 1996. The shell and ligament, pp. 75-168. In: V. S. Kennedy, R. I. E. Newell and A. F. Eble (eds.). The Eastern Oyster Crassostrea virginica. Maryland Sea Grant College, College Park. MD. Castell, J. C. & D. J. Trider. 1974. Preliminary feeding trials using artificial diets to study the nutritional requirements of oysters Crassostrea vir- ginica. J. Fish. Res. Bet. Can. 31:95-99. Chanley. P. E. 1958. Survival of some juvenile bivalves in water of low salinity. Proc. Natl. Shellfish Assoc. 48:52-65. Dekshenieks. M. M.. E. E. Hofmann & E. N. Powell. 1993. Environmental effects on the growth and development of eastern oyster. Crassostrea virginica (Gmelin, 1791), larvae: a modeling study. J. Shellfish Res. 12:241-254. 114 Bataller et al. Dunathun. J. P., R. M. Ingle & W. K. Havens. Jr. 1969. Effects of artificial foods upon oyster fattening. Florida Department of Natural Resources. Marine Research Laboratory Report 58:1-39. Environment Canada. 1993. Atmospheric Environment Service, climat. archive. Halifax. NS, Canada. Environment Canada. 1994. Atmospheric Environment Service, climat. archive. Halifax. NS. Canada. Gabbott. P. A. & R. R. Stephenson. 1974. A note on the relationship be- tween dry meat weight condition index and glycogen content of adult oysters. Ostrea edulis L.. kept in the laboratory. /. Cons. 35:359-361. Galtsoff, P. S. 1964. The American oyster Crassostrea virginica Gmelin. Fish. Bull. 64:1-480. Haven. D. S. 1960. Seasonal cycle of condition index of oysters in the York and Rappahannock Rivers. Proc. Nat. Shellfish Assoc. 51:42-66. Hilbtsh, T. J. 1986. Growth trajectories of shell and soft tissue in bivalves: seasonal variation in Mytilus edulis L. J. Exp. Mar. Biol. Ecol. 96:103- 113. Ingle. R. M. 1949. A comparative study of oyster condition. Science. 109: 593. Lawrence, D. R. & G. I. Scott. 1982. The determination and use of con- dition index of oysters. Estuaries 5:23-27. Littlewood. D. T. J. & C. M. Gordon. 1988. Sex ratio, condition, and gly- cogen content of raft cultivated mangrove oysters Crassostrea rhizo- phorae. J. Shellfish Res. 7:395-399. Lucas, A. & Beninger, P. G. 1985. The use of physiological condition indices in marine bivalve aquaculture. Aquaculture 44:187-200. Mann. R. 1978. A comparison of morphometric. biochemical, and physi- ological indices of condition in marine bivalve mollusks. pp. 484-497. In: J. H. Thorp and J. W. Gibbons (eds.). Early and Environmental Stress in Aquatic Systems. United States Department of Energy. Sym- posium Series (conf. -771 1 14). Muniz. E. C. S. A. Jacob & M. M. Helm. 1986. Condition index, meat yield, and biochemical composition of Crassostrea hrasiliana and Crassostrea gigas grown in Cabo Frio. Brazil. Aquaculture 59:235- 250. Newell. C. R.. D. E. Campbell & S. M. Gallagher. 1998. Development of the mussel aquaculture lease site model MUSMOD: a field program to calibrate model formulations. Aquaculture 219:143-169. Ortega. S. & J. P. Sutherland. 1992. Recruitment and growth of the eastern oyster. Crassostrea virginica, in North Carolina. Estuaries 15:158- 170. Rainer. J. S. & R. Mann. 1992. A comparison of methods for calculating condition index in eastern oysters. Crassostrea virginica, (Gmelin. 1791). J. Shellfish Res. 1 1:55-58. Riley. J. P. & R. Chester. 1971. Introduction to marine chemistry. Aca- demic Press, London. 465 pp. Shumway. S. E. 1996. Natural environmental factors, pp. 467-513. In: V. S. Kennedy. R. I. E. Newell, and A. F. Eble (eds.). The Eastern Oyster Crassostrea virginica. Maryland Sea Grant College. College Park. MD. Strickland. J. D. H. & T. R. Parsons. 1972. A Practical handbook of sea- water analysis. Fish. Res. Board. Can. Bull. 167:185-199. Vernberg. W. B. & F. J. Vernberg. 1972. Environmental physiology of marine animals. Springer- Verlag. Berlin. 346 pp. Wallace. J. F. & T. G. Reinsnes. 1985. The significance of various envi- ronmental parameters for growth of the Iceland scallop. Chlamys is- landica (Pectimdae). in hanging culture. Aquaculture 44:229-242 Walne. P. R. 1970. The seasonal variation of meat and glycogen content of seven populations of oysters, Ostrea edulis L., and a review of the literature. Fish. Invest. London. 26:1-35. Wells. H. W. 1961. The fauna of oyster beds, with special reference to the salinity factor. Ecol. Monogr. 31:239-266. Wetzel, R. G. & G. E. Likens. 1979. Limnological analyses. W. B. Saun- ders. Philadelphia. Widdows, J.. P. Donkin, P. N. Salkeld. J. J. Cleary. D. M. Lowe. S. V. Evans & P. E. Thomson. 1984. Relative importance of environmental factors in determining physiological differences between two popula- tions of mussels (Mytilus edulis). Mar. Ecol. Prog. Ser. 17:33—47. Wilbur. K. M. & A. S. M. Saleuddin. 1983. Shell formation, pp. 236-287. In: A. S. M. Saleuddin and K. M. Wilbur (eds.). The Mollusca: Physi- ology, vol. 4, part 1. Academic Press. New York. Zar. J. H. 1984. Biostatistical analysis. Prentice Hall. Upper Saddle River. NJ. 718 pp. Journal of Shellfish Research, Vol. 18. No. 1. 115-120, 1999. HIGH-RESOLUTION ANALYSIS OF KARYOTYPES PREPARED FROM DIFFERENT TISSUES OF THE EASTERN OYSTER CRASSOSTREA VIRGINICA Q. ZHANG,1 G. YU,2 R. K. COOPER,1 AND T. R. TIERSCH2 'Department of Veterinary Science Aquaculture Research Station Louisiana Agricultural Experiment Station Louisiana State University Agricultural Center Baton Rouge, Louisiana 70803 ABSTRACT Establishment of chromosome identity is the first step of physical genome mapping. This step is hindered hy a lack of handing techniques and size homogeneity in the chromosomes of the eastern oyster, Crassostrea virginica. In this study, chromosomes prepared from adult gill, ripe gonad tissues, and embryos were analyzed with a computer-based measurement system. Chromosomes from embryos were elongated with excellent morphology and identifiable secondary constrictions, although homologous pairs were difficult to establish because of asymmetric sizes found between homologues. Meiotic chromosomes at the stages of diakinesis (testis) and pachytene (ovary) offered distinct advantages for karyotyping. These chromosome bivalents possessed a haploid chromosome number (n = 10) with transverse chromomere bands analyzable by microdensitometry. Chromosomes derived from gill tissue were highly condensed and few spreads were analyzable. Idiograms of each chromosome were developed in this study based on size, centromere position, and chromomere bands. These results indicate that mitotic and meiotic cells are each important for the study of chromosomes of the eastern oyster and that computer-assisted analysis will be useful for establishment of karyotypes and idiograms. KEY WORDS: Chromomere bands, microdensitometry, karyotype. Crassostrea virginica INTRODUCTION Identification of individual chromosomes is the prerequisite step for in situ detection and location of specific DNA sequences. The genome of most oyster species is composed of 10 metacentric and submetacentric chromosomes, comprising three groups based on size (Longwell and Stiles 1967). Techniques need to be devel- oped for further identification of specific chromosomes within the oyster genome. Identification of individual mammalian chromo- somes has been accomplished by techniques such as Giemsa (G)- banding or reverse (R)-banding: however, these techniques do not produce consistent banding patterns in lower vertebrates (Zhang and Tiersch 1998a) and invertebrates, including oysters. It is well established that the quality of chromosome prepara- tions is directly related to the source material. Embryos and so- matic tissues (such as gill filament) are most commonly used for cytogenetic analysis in oysters. Karyotypes of the eastern oyster, Crassostrea virginica, based on somatic (Rodriguez-Romero et al. 1978) and meiotic chromosomes (Longwell et al. 1967), have been described. However, none of these studies examined structural markers on the chromosomes: therefore, chromosomes of similar size could not be distinguished in these studies. Although "G-like" chromosome bands were studied in this species (Rodriguez- Romero et al. 1979), routinely useful banding patterns and analysis techniques have not been established. In fish, meiotic chromosomes (at early prometaphase I) without colchicine treatment, are typically extended with knot-like struc- tural markers (chromomeres). which are useful for identifying in- dividual chromosomes (Yu et al. 1994). Most spreads prepared from ripe gonad tissue are composed of chromosome bivalents, and no pairing is required to establish a karyotype. With the as- sistance of microdensitometric methods (Zhang and Tiersch 1998b). these chromomere bands could be identifiable as markers for chromosomes of oysters. In the present study, we developed haploid karyotypes for the eastern oyster from meiotic chromosomes (in diakinesis and pachytene stages). By computer-based measurements and statisti- cal analysis, we compared these karyotypes with diploid karyo- types prepared from gill and embryonic cells, and documented naturally occurring chromomere structures useful for chromosome identification. MATERIALS AND METHODS Materials Eastern oysters were obtained from Grand Isle. Louisiana, and were maintained in an indoor recirculating system until use (Buchanan et al. 1998). Twenty ripe females and 15 ripe males were used to sample gonad and gill tissues in this study. Embryos were produced by artificial fertilization based on methods de- scribed in Paniagua-Chavez et al. (1998). Chromosome Preparation Gonad Tissue About 0.2 g of testis or ovary tissue were removed and cut into -2-mirr pieces, which were placed in 0.9% sodium citrate (pre- pared in deionized water) with continuous swirling for 5-7 h. The tissue fragments were fixed twice for 20 min with cold Carnoy's fixative I (methanol:chloroform:acetic acid: 6:3:1). and fixed three times for 20 min in cold Carnoy's fixative II (methanohacetic acid; 3: 1 ). The tissues were ground and passed through a 70-u, cell strainer. Cells were collected, pelleted, and resuspended in Car- noy's fixative II overnight. Chromosomes were prepared on mi- croscope slides using standard air-drying methods. Embryos Embryonic cells collected 5 h after fertilization were used in this study. Techniques for preparation of chromosomes of oyster embryos have been described in Guo and Allen ( 1997). Embryos were concentrated on a 15-p. nylon filter after the colchicine treat- 116 Zhang et al. Figure 1. Preparation of mitotic (somatic) metaphase chromosomes from eastern oyster (C. virginica); (a) gill; and (b) male gonad tissue; bar = 10 u. merits. Chromosomes were prepared using standard air-drying methods. Gill Filaments Gill tissues were incubated for 3 h in the medium JL-ODRP-4 developed for culture of oyster cells (J. LaPeyre pers. comm.). Colchicine was added to the medium at a final concentration of 0.01%. Chromosomes were prepared with the same method used for gonad tissues, except that hypotonic (sodium citrate) treatment was 30 min. Computer-Assisted Karyotyping Chromosome images were digitized with a 24-bit video capture board (Imaging Technology Inc.. Bedford. MA) using a light mi- croscope (Microphot-SA. Nikon Inc.) equipped with a high- resolution RGB color video camera (model A206A, Microimage Video Systems Co.. Inc.. Boyertown. PA). The Optimas® com- puter software packages (Bioscan. Inc.. Edmonds, WA). a Win- dows™ based application, was used to capture and process the chromosomal images. Total length, arm length, and banding pat- terns (chromomeres) of chromosomes were measured with the linear functions of Optimas (Zhang and Tiersch 1998b). Meiotic chromosomes (n = 10) were arranged by order of descending size. For comparisons, mitotic chromosomes (2n = 20) were ordered by size, and pairs were established based on relative length and ratio of short arm to entire length (i.e., centromeric index) (Levan et al. 1964). Idiograms were created using Microsoft Powerpoint™ for each chromosome based on size, centromere position, and banding patterns. Repeated measurements (n = 10) were taken from represen- tative spreads, and statistical analysis (see below) was performed to estimate technical variation using data derived from individual spreads, and for biological variation using data derived from dif- ferent spreads of a particular tissue type or from different tissue types. Statistical Analysis One-way analysis of variance (ANOVA) was performed to compare: (1) 10 repeated measurements of total length of chro- mosomes within a representative spread, used to examine the ac- curacy of the image analysis system; (2) relative length of indi- vidual chromosomes derived from different spreads from the same tissue source (embryo); and (3) relative length of individual chro- mosomes derived from four different tissue sources (gill, embryo. testis, and ovary). The relative length data were transformed to arcsine square root values before analysis. Data were collected Mitotic (embryo) Diakinesis (testis) Pachytene (ovary) Idiogram CI mean±SD 10 MrainixKJix ))Ultliw Tjjgj^mj SIM i Mil 43.3±2 1 44.5 ±3.5 35.2±3.4 41.2±1.4 43.5±2.1 35.1±38 40.4±1.2 38.7±1.3 32 4±4.1 41 5±2.3 J A/ ^ .05) (Table 3). DISCUSSION In this study, we found that chromosomes derived from four tissue types had distinct morphological features providing different options for year-round genetic studies of the eastern oyster (Tables 4. 5). The mitotic activity of oyster somatic cells is low. and there are no methods available for stimulation of mitosis in oyster so- matic cells in vivo or in vitro (Cornet 1993). Although an increase of colchicine concentration increased the number of spreads ob- served, the resultant chromosomes were short, less distinct in size, and not suitable for use in physical mapping. Chromosomes pre- pared from embryos were elongated and provided images of high resolution. Chromosome structures such as secondary constric- tions were recognizable on these chromosomes. However, heavy background caused by adherent materials limited subsequent analysis. Techniques have been developed for removing yolk ma- terials from preparations of oyster chromosomes derived from cleaving eggs (Longwell and Stiles 1968). In this study, we found that 50% acetic acid could eliminate most background materials (data not shown). Nevertheless, the effects of these treatments on the quality of chromosomes for use in physical gene mapping needs to be evaluated. In this study, a new protocol was introduced for preparation of bivalent chromosomes from ripe gonad tissue of the eastern oyster, which omitted colchicine treatment and included a prolonged hy- TABLE 4. Ploidy level, availability, division stage, and spreads per slide of four tissue types used in this study; the number of spreads observed on each microscope slide was used to estimate mitotic index of each tissue type of eastern oyster. Tissue Ploidv Availahilitv Division Stage Spreads per Slide" Gill Embryo Testis Ovary 2N 2N IN. 2N IN. 2N Year-around Spawning season Spawning season Spawning season Mitotic metaphase Mitotic metaphase Mostly diakinesis Mostly pachytene Low High a Low. <5 spreads per slide; high. >30 spreads per slide, and ■ cell division dependent on season. High-Resolution Analysis of Eastern Oyster, Karyotypes 119 TABLE 5. Suitability of chromosomes derived from different tissues of eastern oyster for physical genome mapping. Centromeres Chromomere Pairs Tissue Elongation Dispersal Resolution Identified Bands Identified Mapping Gill Short Separated Low Yes No 3-5 Not suitable Embryo Long Separated High Yes No 7 Suitable Testis (diakinesis) Intermediate Less separated High No Yes 10 Suitable Ovary (pachytene) Very long Overlapping Highest No Yes 10 Suitable potonic treatment step. Bivalent haploid spreads yielded the ad- vantages of reduced chromosome numbers, no need for pairing of homologous chromosomes, and identifiable chromomere bands on most chromosomes. However, these chromosomes were difficult to disperse, especially for the pachytene bivalents. Although incu- bation of gonad tissue with colchicine would facilitate dispersal of bivalent chromosomes, the chromomere bands on these chromo- somes would become less identifiable (Yu et al. 1994). Further identification of the advantages and disadvantages of chromo- somes from different source tissues is essential for analysis of oyster chromosomes and establishment of a representative karyo- type in this species for such activities as physical mapping of genes. In this study, the computer-based image analysis system pro- vided a reproducible and objective method for measurement of chromosomes. The difference (percentage variation) among mea- surements of individual chromosomes was less than 5%. which accounted for technical variation including manual error. Variation caused by biological factors was analyzed by comparing different spreads of the same tissue type or spreads of different tissue types. The difference in average relative length was found to be as high as 9% among different spreads from individual embryos. These differences were largely derived from the continuous changes of chromosome morphology produced during the cell cycle. On the other hand, results of this study demonstrated that our measure- ment system is highly sensitive and capable of detecting minor differences among individual chromosomes. These results suggest that methods need to be developed ( such as use of marker chro- mosomes) for standardization of the eastern oyster karyotype. Pairing of homologous chromosomes based solely on size and centromeric index was difficult for oyster chromosomes prepared from embryonic cells. Several chromosome pairs were asymmetric in size, centromeric index, and secondary constrictions. A typical example was chromosome 1. the largest metacentric chromosome, which displayed differences in relative length between homo- logues of 0.2 to 0.4% and in presence or absence of secondary constriction. These asymmetric features could be attributable to normal development of embryos, abnormal development, or inte- gration of features of the genomes of the parents. We did not observe such asymmetric features on chromosomes of other cells, although these chromosomes were highly condensed. Some differences were found between the karyotype developed previously for embryos of C. virginica (Longwell and Stiles 1968) and the one developed in this study. Only three submetacentric chromosomes were found in this study, as compared with four submetacentrics described previously. Also, the second largest chromosome was found to be metacentric rather than submetacen- tric, as described previously. Reasons for these differences include differences in measurement techniques, variation in the contraction of chromosomes used in each study, and genetic polymorphism among populations of the eastern oyster. Development of such methods as C-banding (for identification of constitutive hetero- chromatin) would be helpful for resolving these differences by identifying the location of centromeres. In summary, most chromosomes (seven or eight) prepared from embryos could be identified based on morphological measure- ments; however, the pairing of homologous chromosomes was difficult. Thus, embryos are a convenient source material for chro- mosome preparation, although their utility for high-resolution analysis is limited. Meiotic chromosomes, although less frequently studied, were useful, given their reduced numbers and the presence of unique, naturally occurring bands on each diakinesis chromo- some. This study provided information about the composition of the eastern oyster genome. Techniques developed for meiotic chromo- somes and idiograms based on these chromosomes have proved to be valuable for physical genetic mapping of oysters in our labo- ratories (unpublished data). Integrative analysis of mitotic meta- phase chromosomes and meiotic bivalent chromosomes will be a useful tool for specific identification of chromosomes in the oyster genome until other techniques become available for mollusks. ACKNOWLEDGMENTS This study was supported by a USDA special grant and the Louisiana Sea Grant College Program. We thank J. Buchanan and C. Paniagua for technical assistance with oyster spawning, and J. LaPeyre. LSU Department of Veterinary Science, for providing the JL-ORPD-4 medium. We thank A. Pani for performing mea- surements of the chromosomes. This manuscript was approved by the Director of the Louisiana Agricultural Experiment Station as number 98-66-0542. LITERATURE CITED Buchanan. J. T„ G. S. Roppolo. J. E. Supan & T. R. Tiersch. 1988. Con- ditioning of eastern oysters in a closed, recirculating system. J Shellfish Res. 17:1183-1189. Comet. M. 1993. A short-term culture method for chromosome preparation from somatic tissues of adult mussel (Mytilus ednlis). Experientia 49: 87-90. Guo. X. & S. K. Allen. Jr. 1997. Fluorescence in situ hybridization of vertebrate telomere sequence to chromosome ends of the pacific oyster. Crassostrea gigos Thunberg. J. Shellfish Res. 16:87-89. Le\an. A.. K. Fredga & A. Sandberg. 1964. Nomenclature for centromeric positions on chromosomes. Herecliias 52:201-220. Longwell, A. C. & S. S. Stiles. 1968. Removal of yolk from oyster eggs by 120 Zhang et al. Soxhlet extraction for clear chromosome preparation. Stain Technol. 43:63-68. Longwell, A. C. S. S. Stiles & D. G. Smith. 1967. Chromosome comple- ment of the American oyster Crassostrea virginica. as seen in meiotic and cleaving eggs. Can J. Genet. Cytology 9:845-856. Paniagua-Chavez. C. G.. J. T. Buchanan & T. R. Tiersch. 1998. Effect of extender solutions and dilution on motility and fertilizing ability of eastern oyster sperm. J. Shellfish Res. 17:231-237. Rodriguez-Romero, F.. M. Uribe-Alcocer & A. Laguarda-Figueras. 1978. Cytogenetic study of an oyster population of the species. Crassostrea virginica Gmelin. from the coast of Tabasco, Mexico. Japan. J. Ma- lacol. 37:83-86. Rodnguez-Romero. F.. A. Laguarda-Figueras. M. Uribe-Alcocer & M. L. Rojas-Lara. 1979. Distribution of "*G" bands in the karyotype of Cras- sostrea virginica. Venus 38:180-184. Yu. Q.. L. Fan. J. Cui. X. Ren. K. Li & X. Yu. 1994. High-resolution G-binding and idiogram on pachytene bivalents of rice field eels 'Monopterus alhus Zuiew). Sci. China 9:1090-1096. Zhang. Q. & T. R. Tiersch. 1998a. Standardization of the channel catfish karyotype with localization of constitutive heterochromatin and restric- tion enzyme banding. Trans. Am. Fisheries Soc. 127:551-559. Zhang, Q. & T. R. Tiersch. 1998b. Identification and analysis of weak linear banding patterns of fish chromosomes with a densitometry method. BioTechniques 24:996-997. Journal of Shellfish Research. Vol. 18. No. 1. 121-125, 1999. GENETIC DIVERSITY IN THE EASTERN OYSTER (CRASSOSTREA VIRGINICA) FROM MASSACHUSETTS USING THE RAPD TECHNIQUE BETH M. HIRSCHFELD,' ARUN K. DHAR,1 * KARL RASK,2 AND ACACIA ALCIVAR-WARREN1 f 'Department of Environmental and Population Health. Tufts University School of Veterinary Medicine. North Grafton. Massachusetts 01536 : Barnstable. Massachusetts 02630 ABSTRACT The random amplified polymorphic DNA (RAPD) technique was used to examine the genetic variability in eastern oysters (Crassostrea virginica) from four wild-naturalized stocks (Wellfleet. Wareham River. East Wareham/Onset and Barnstable Harbor) and one cultured stock (Cotuit) in Massachusetts. Initially. 20 oligonucleotide primers (Kits A. B. C. G. M. and Z; Operon Technologies Inc.. Alameda. California) were screened and 10 were selected to amplify DNA from 79 samples representing these five sites. A total of 90 DNA bands ranging in size from 205 to 1.400 base pairs (bp) were scored. The highest level of polymorphisms were detected in samples from Barnstable Harbor (74%) followed by Wellfleet (71%). Wareham River (70%). East Wareham/Onset (62%). and Cotuit (54%). There were significant differences in polymorphisms between three wild-naturalized (Barnstable Harbor. Wellfleet. and Wareham River) and cultured samples. East Wareham/Onset samples did not show significant differences with the cultured population. Potential site-specific RAPD markers were identified with primers OPA10. OPA17. OPC10. OPG06, and OPM18. The frequency of these site-specific RAPD markers varied among the sample collection sites. In addition, four unique alleles (OPA 17-900 and 875 bp: OPC 10-450 bp. and OPG06-270 bp) were identified in samples from Barnstable Harbor. Wareham River, and East Wareham/Onset. Our data provide baseline information on the genetic variation in cultured and wild oyster stocks in Massachusetts and may be useful for future management of the resource. KEY WORDS: Eastern oyster. Crassostrea virginica, RAPD. genetic diversity INTRODUCTION The eastern oyster (Crassostrea virginica) is a benthic murine species with a planktonic larva and a sessile adult stage (Hedge- cock 1995). The natural habitat ranges from St. Lawrence Bay. Nova Scotia, through the Gulf of Mexico to the Yucatan Peninsula, Mexico and into the West Indies (Galtsoff 1964). The genetic- structure of C. virginica across its entire habitat has been examined using different molecular techniques (Buroker 1983. Reeb and Avise 1990. Karl and Avise 1992, King et al. 1994. McDonald et al. 1996. Small and Chapman 1997). Allozyme polymorphism analyses indicated a high gene flow in C. virginica populations from Cape Cod. Massachusetts (MA) to the Gulf of Mexico (Bu- roker 1983). although restriction fragment length polymorphisms (RFLPs) of whole mitochondrial DNA and some (but not all) nuclear DNA markers suggested a genetic break in eastern Florida separating Atlantic populations from those of the Gulf of Mexico (Reeb and Avise 1990, Karl and Avise 1992, King et al. 1994. McDonald et al. 1996). Oyster populations from Laguna Madre, Texas, constitute a genetically divergent group from other popu- lations caused by differential selection operating in this hyperha- line environment (King et al. 1994). Mitochondrial RFLPs using 16S ribosomal RNA gene has been used to distinguish C. virginica from two closely related Asian oyster species (C. gigas and C. ariakensis) (O'Foighil et al. 1995). Sequence data revealed that Asian oyster species showed higher levels of similarity to each other (95%) than to C. virginica (84-86%). Recently, random amplified polymorphic DNA (RAPD) technique was used to dis- tinguish species in the marine bivalve genus Donax and evaluate *Present address: Super Shrimp Group Inc.. 1545 Tidelands Ave., Nation- al City, CA 91977. tCorresponding author: aalcivar@opal.tufts.edu its biogeography (Adamkewicz and Harasewych 1996). RAPD technique requires no prior knowledge of the genome, uses very little DNA. detects high levels of polymorphisms, analyzes a large portion of the genome in a short time, and is faster than such molecular techniques as RFLP and microsatellites. For these rea- sons, RAPD technique has become a powerful tool to assay the genetic variation of populations (Hadrys et al. 1992). Although the genetic variability of natural oyster populations has been evaluated in different studies, information on the genetic diversity of hatch- ery stocks and its comparison to native populations is limited. In recent years, native and cultured populations of oysters in the northeastern United States have suffered periodic heavy losses from diseases such as Dermo (Perkinsus marinus). MSX (Hap- losporidium nelsoni). and Juvenile Oyster Disease (Gaffney and Bushek 1996, Lewis et al. 1996). It is possible that the increase in disease incidence in oysters may be caused by reduced genetic diversity or the introduction or transfer of stocks (Ford 1992). In oyster hatcheries, broodstock are sometimes selected primarily on the basis of such morphological features as size, shell character- istics, color, and growth. In selected breeding programs, only a small number of broodstock are generally used, enhancing the possibility of decreased genetic variation in cultured, as compared to wild, stocks (Wilkins 1976). In this study, the RAPD technique was used to evaluate the genetic variability in one cultured and four wild populations of eastern oysters from different geographic- locations in Cape Cod areas in Massachusetts. MATERIALS AND METHODS Sample Collection Oyster samples were collected from one cultured and four wild populations in Cape Cod, Massachusetts (Table 1, Fig. 1 ). Initially, 143 samples (30 each from Cotuit. Wellfleet. and Wareham River. 29 from Barnstable Harbor, and 24 from East Wareham/Onset) 121 122 HlRSCHFELD ET AL. were collected. However, good quality DNA (undegraded/high molecular weight DNA) was obtained for only the 79 samples (Table 1) used for the RAPD assay. The cultured population (Co- tuit) refers to hatchery oysters purchased by the growers (Cotuit Oyster Co.. Inc. I and maintained for the entire lifespan of the oyster until sale. These oysters originated from Ocean Pond Corp. in Fishers Island. New York and were seeded in Cotuit Bay in 1995. Wild populations were represented by oysters from the areas of Wellfleet. Wareham River, East Wareham/Onset. and Barn- stable Harbor. All four locations have received oyster transplants from other areas along the east coast over the last 50 years, which have naturalized with any previously existing oyster populations. The last imports were in the mid-1980s. Thus, the existing stocks are best described as wild naturalized stocks. Among these wild naturalized populations. Onset has been the least diluted by im- portation. There was no significant native population in Barnstable Harbor until oysters were imported for aquaculture and municipal purposes from Connecticut in the early 1980s. These oysters sub- sequently spawned, and the offspring survived to contribute the present-day population in this new area. East Wareham/Onset Wareham River *~- .29) from the East Wareham/ Onset wild population but differed significantly from the Wellfleet (p < .02). Wareham River (p < .03). and Barnstable Harbor (p < .01) populations. Overall, the cultured population was the least polymorphic among the five populations. In a study involving allozyme polymorphisms. Wilkins (1976) reported that cultured samples exhibited fewer polymorphisms as compared to wild stocks. However. English et al. (1997) reported high genetic varia- tion in C. gigas (mean polymorphism: 0.70) both farmed and es- tablished Tasmanian populations. The discrepancy in polymor- phism levels in hatchery stock reported in the present study, as compared to earlier reports, could be attributable to the difference in the number of parental oysters used for spawning. Nevertheless, our data suggest the need to use a large number of broodstock for spawning to enhance the diversity of cultured stocks. Minimal genetic differentiation has been reported for many benthic marine species with planktonic larval stages (Buroker 1983. Palumbi and Wilson 1990). Analysis of the genetic structure of C. virginica from Cape Cod. Massachusetts through Browns- ville. Texas using allozyme markers revealed very high similarities (99%) among geographically separated populations, suggesting a Cultured Wild Cotuit Wareham River East Wareham/Onset Barnstable Harbor M Figure 2. Composite figure of RAPD fingerprints of oyster samples obtained from five different sites in Cape Cod, Massachusetts analyzed with primer OPA17. M = 100 bp molecular weight marker of Gibco BRL. 124 HlRSCHFELD ET AL. TABLE 3. The frequency of region-specific potential RAPD markers for oyster samples collected from five different locations in Cape Cod Massachusetts. East Wareham/ Cotuit Wellfleet Wareham River Onset Barnstable Harbor RAPD Marker" (n = 15) in = 16) (n = 16) In = 16) (n = 16) OPA 10-250 0.60 0.50 0.31 0.00 0.75 OPA 17-675 0.00 0.31 0.38 0.69 0.56 OPA 17-575 0.00 0.38 0.50 0.44 0.56 OPA 17-550 0.00 0.25 0.19 0.38 0.19 OPA 17-435 0.67 0.00 0.50 0.31 0.75 OPC 10-800 0.20 0.56 0.56 0.88 0.56 OPG06-350 1.00 0.38 0.31 0.50 0.56 OPM 18-350 0.93 0.69 0.50 0.69 0.20 OPM18-310 0.80 0.50 0.69 0.19 0.33 1 Primer name followed by the size of DNA band amplified. high gene flow (Buroker 1983). However. Reeb and Avise ( 1990) reported a distinct genetic break in oyster populations from the Atlantic coast and Gulf of Mexico while examining the RFLP pattern of whole mitochondrial DNA. The genetic discon- tinuity between oyster populations from the Atlantic Coast and the Gulf of Mexico was further confirmed by Karl and Avise (1992). using single-copy nuclear DNA markers. The discordance between the allozyme. mitochondrial, and single-copy nuclear DNA mark- ers was attributed to balancing selection that prevented the accu- mulation of allozyme differences (Karl and Avise 1992). In con- trast, using anonymous nuclear DNA polymorphisms. McDonald et al. ( 1996) failed to identify any genetic discontinuity between oysters from Panaca, Florida and Charleston. South Carolina on the Gulf and Atlantic Coasts, respectively. Nevertheless, all these studies indicated a high within-population homogeneity in Atlantic Coasts including Cape Cod. Massachusetts regardless of the ge- netic markers used for the study (Buroker 1983, Reeb and Avise 1990, Karl and Avise 1992). Our RAPD data are. therefore, concordant with earlier studies indicating high genetic simi- larity within oyster populations in Massachusetts. The number of samples collected from each location and the number of prim- ers screened in this study were limited. Therefore, more sam- ples should be tested with additional primers to confirm the genetic homogeneity of oyster samples from Cape Cod. Massa- chusetts. Nine RAPD markers showed differences in their frequency among populations (Table 3). Five of these markers (OPA10250, OPA17675. OPA17575, OPA17550, and OPA17435) could serve as potential region-specific markers. A 250 bp DNA band amplified by the primer OPA 10 was absent in East Wareham/Onset sam- ples and present in the remaining populations at different frequen- cies (0.31-0.75). Similarly, the 675 bp. 575 bp. and 550 bp DNA bands amplified by OPA 17 were absent in the cultured Cotuit samples, and a 435 bp band amplified by OPA- 17 was absent in the Wellfleet samples. RAPD also identified four unique alleles that include a 900 bp and a 875 bp DNA bands amplified by the primer OPA 17. a 270 bp band amplified by OPG06 primer. and a 450 bp band amplified by OPC10 primer (Table 4). Because we tested limited numbers of samples from each of the five geographic locations, it would be interesting to see if these region- specific as well as the unique markers can still be reproduced when a large number of samples are tested from all of these locations. In recent years, RAPD markers have been used to assess the population genetic variation in many plant and animal species (Hadrys et al. 1992). However, its application in population ge- netic studies in aquaculture species is still limited (Gomes et al. 1998). The present study confirms the potential of RAPD markers for population genetic studies in shellfish and provides baseline information on the genetic diversity of wild and cultured popula- tions of oysters in areas of Cape Cod, Massachusetts. Follow-up studies should be performed to determine if, indeed, all cultured stocks are low in genetic diversity or if environmental disturbance leads to lower levels of variation. Future efforts should also be made to examine the relationship between population genetic variation and the prevalence of diseases in oysters in these areas. TABLE 4. Unique alleles identified by RAPD technique and their frequency (in parenthesis) in oyster populations from five different locations in Cape Cod Massachusetts. East Wareham/ DNA Band Cotuit Wellfleet Wareham River Onset Barnstable Harbor Primer (bp) (n = 15) (n = 16) (n = 16) (n = 16) (n = 16) OPA 17 900 1 (0.06) OPA 17 875 1 (0.06) 1 (0.06) OPC 10 450 2(0.13) 2(0.13) OPG06 270 1 (0.06) 2(0.13) 2(0.13) Genetic Diversity of Eastern Oysters from Massachusetts 125 LITERATURE CITED Adamkewicz. S. L. & M. G. Harasewych. 1996. Systematics and biogeog- raphy of the Genus Donax (Bivalvia: Donacidae) in eastern North America. .4m. Malacolog. Bull. 13:97-103. Buroker, N. E. 1983. Population genetics of the American oyster Crasso- strea virginica along the Atlantic Coast and the Gulf of Mexico. Mar. Biol. 75:99-112. English. L. J.. G. B. Maguire & R. D. Ward. 1997. Genetic variation in Australian wild and hatchery populations of the Pacific oyster (Cras- sostrea gigas Thunberg) using allozyme. mitochondrial, and microsat- ellite techniques. In: Sixth International Symposium on Genetics in Aquaculture. University of Stirling. Scotland. 24-28 June. Ford, S. E. 1992. Avoiding the transmission of disease in commercial culture of mollusks. with special reference to Perkinsus marinus (Dernio) and Haplosporidium nelsoni (MSX). J. Shellfish Res. 1 1:539- 546. Gaffney, P. M. & D. Bushek. 1996. Genetic aspects of disease resistance in oysters. J. Shellfish Res. 15:135-140. Galtsoff. P. S. 1964. The American oyster Crassostrea virginica Gmelin. Fish Bull. U.S. 64: 1 -+80. Garcia, D. K.. M. Faggart. L. Rhoades. J. Wyban. W. Carr, J. Sweeney, K. Ebert & A. Alcivar-Warren. 1994. Genetic diversity of cultured Pe- naeus vannamei shrimp using three molecular genetic techniques. Mol. Mar. Biol. Bioleclmol. 3:270-280. Gomes. C. R. B. G. Dales & H. A. Oxenford. 1998. The application of RAPD markers in stock discrimination of the four-wing flying fish. Hirundichthys affinis in the central western Atlantic. Mol. Ecol. 7: 1029-1039. Hadrys. H.. M. Balick & B. Schierwater. 1992. Applications of random amplified polymorphic DNA (RAPD) in molecular ecology. Mol. Ecol. 1:55-63. Hedgecock, D. 1995. The cupped oyster and the Pacific oyster, pp. 1 15- 137. In: J. Thorpe. G. Gall, J. Lannan and C. Nash (eds.). Conservation of Fish and Shellfish Resources Managing Diversity. Academic Press. New York. Karl. S. A. & J. C. Avise. 1992. Balancing selection at allozyme loci in oysters: implications from nuclear RFLPs. Science 256:100-102. King, T. L... R. Ward & E. G. Zimmerman. 1994. Population structure of eastern oysters (Crassostrea virginica) inhabiting the Laguna Madre, Texas, and adjacent bay systems. Can. J. Fish. Aqu. Sci. 51:215-222. Lewis. E. J.. A. C. Farley. A. Baya & E. B. Small. 1996. Juvenile Oyster Disease — Transmission and bacteriological studies. Proceedings of the Annual Meeting of the National Shellfisheries Association. April 14- 18. 1996. Baltimore. MD. p. 516. (Abstract). McDonald. J. H.. B. C. Verrelli & L. B. Geyer. 1996. Lack of geographic variation in anonymous nuclear polymorphism in the American oyster, Crassostrea virginica. Mol. Biol. Evol. 13:1114—1118. O Foighil, D.. P. M. Gaffney & T. J. Hilbish. 1995. Differences in mito- chondrial 16S ribosomal gene sequences allow discrimination among American [Crassostrea virginica (Gmelin)] and Asian [C. gigas (Thun- berg). C ariakensis Wakiya] oyster species. J. Exp. Mar. Biol. Ecol. 192:211-220. Palumbi, S. R. & A. C. Wilson. 1990. Mitochondrial DNA diversity in the sea urchins Strongylocentrotus purpuratus and S. droebachiensis. Evo- lution 44:403-415. Reeb. C. A. & J. C. Avise. 1990. A genetic discontinuity in a continuously distributed species: mitochondrial DNA in the American oyster, Cras- sostrea virginica. Genetics 124:397^106. Small. M. P. & R. W. Chapman. 1997. Intraspecific variation in the 16S ribosomal gene of Crassostrea virginica. Mol. Mar. Biol. Biotechnol. 6:189-196. Wilkins, N. P. 1976. Genetic variability in marine bivalvia: implications and applications in molluscan mariculture. pp. 549-563. In: G. Per- soone and E. Jaspers (eds.). Proceedings of the 10th European Sym- posium on Marine Biology vol. 1. September 17-23. 1975. Ostend. Belgium. Universa Press. Wetteren. Belgium. Journal of Shellfish Research, Vol. 18. No. 1. 127-131. 1999. EFFECTS OF AGE, SIZE, AND SEASON ON GROWTH OF SOFT TISSUE IN THE OYSTER CRASSOSTREA GIGAS (THUNBERG, 1793) J. CIGARRIA Unidad de Zoologia, Departamento Biologia de Organismos y Sistemas, Universidd d'Uvieu. Principal! d'Asturies, Spain ABSTRACT This study describes potential effects of age, size, and season on the relative allocation of resources to meat and shell growth in Crassoslrea gigas. Two oyster size groups (slow- and fast-growing oysters) of the same cohort were followed over 2 years. Analysis of covariance detected significant differences in meat weight between groups, seasons, and years. Fast-growing oysters have more meat per gram of shell than do slow-growing, suggesting greater efficiency of the conversion of consumed energy to tissue in the former. Seasonal changes in both sets in meat were related to productivity pattern of phytoplankton, with the highest values of chlorophyll a in early summer and minimum in early winter. Differences observed between years in meat content of oysters are attributable both to variability in environmental conditions and to the effect of age. KEY WORDS: Crassostrea gigas. Japanese oyster, growth, size, age. Eo estuary INTRODUCTION The rapid growth, high meat yield, and relative robustness of the Pacific oyster. Crassostrea gigas (Thunberg 1793) has led to its introduction into many parts of the world as a successful species for aquaculture (Brown 1988). Of the nearly one million metric- tons of oysters (genus Ostrea and Crassostrea) landed in the world in 1990 (1.2% of the total aquaculture production). 87% were C. gigas (FAO 1992). Oysters are often marketed in their shells, but meat is the primary marketable component, so any oyster culture operation should aim to maximize the energy allocated to meat growth and to obtain an optimum meat/shell ratio. Precise knowl- edge of the temporal relationship between shell and tissue produc- tion is of great importance identifying optimal harvesting times of cultured bivalves (Lucas and Beninger 1985). It is important in the management of oyster culture operations to understand the genetic and environmental factors influencing energy resource allocation (Raubenheimer and Cook 1990). The performance of Pacific oysters under different environ- mental conditions has been assessed through measurements of growth rate, condition index, and survival rate (Quayle 1969, Askew 1972, Brown and Hartwick 1988a, Brown and Hartwick 1988b); however, no study dealing with adult oysters has explored how the relationship of meat-shell weight changes with the com- bined effects of age, season, and size. Age-specific responses may reflect variations in the allocation of assimilated energy into the such metabolic compartments as reproduction, storage, shell, or somatic tissues (Zandee et al. 1980). Seasonal changes in meat weight result from the storage and utilization of food reserves in relation to the complex interactions of food availability and tem- perature with growth and reproductive processes (Dare and Ed- wards 1975. Bayne and Newell 1983). The size effect represents changes in the relation of meat-shell depending upon the weight of the group. Usually, samples from different seasons contain indi- viduals of similar size, and no information is obtained from size- class differences, assuming that individuals of different sizes re- spond similarly to changing environmental conditions (Mallet and Carver 1993). In addition, the timing of shell and tissue growth may vary substantially among different size or age classes (Peter- son and Fegley 1986). Therefore, the objectives of this study were to determine temporal trends in meat production, and developing size, age, and season-specific relationships for shell and meat growth by monitoring growth rates in slow- and fast-growing oys- ters belonging to the same cohort over 2 years. MATERIAL AND METHODS One set of Japanese oysters (two million individuals. 0.6 g mean live weight) from a Spanish hatchery was planted in Sep- tember 1992 in the intertidal zone of the Eo estuary (Fig. 1). In January 1994, size sorting was carried out using a commercial oyster grader. Then, the oyster set was separated into three groups depending upon whole oyster weight: the slow-growing oysters, medium-growing oysters, and the fast-growing oysters of the co- hort (Fig. 2). Only the slow-and the fast-growing oysters were sampled in the subsequent analyses. Both groups were reared for 2 years in standard oyster bags ( 1 x 0.5 m, 14 mm diameter) at densities of approximately 100 oys- ters per bag, arranged in a line on metal tables (60-cm high and 3-m length), in the middle of the commercial oyster culture area. The bags were turned over every 4 months to prevent development of fouling organisms. At the same time, they were rotated to avoid the biased effect of some food concentration gradient. The sam- pling procedures were as follows: 15 oysters from each set were randomly collected every 10 weeks during a 2-year period. 0\ sters were washed, weighed, and the meat was removed and immedi- ately dried at 60°C to a constant weight (48-72 h) and reweighed to the nearest 0.01 g. The shell was air dried for 4S h and weighed to the nearest 0.1 g. Seawater temperature and salinity were mea- sured at 1 to 3-m depth every day at the culture site. Daily tem- perature and salinity measurements were recorded as monthly means (Fig. 3) in order to make the representation clearer. Statistical Analysis A three-way analysis of covariance (ANCOVA) was used to determinate the effect of age. size, and season on dry meat weight, with shell weight as a changing covariate. following Walne and Mann (1975), who recommended the use of weight rather than linear dimensions in the analysis of seasonal growth of oysters, because they are irregular in shape. In ANCOVA. the assumption of equal slopes was first tested, and without-slope heterogeneity intercepts (adjusted means) were tested. The samples were col- lected each year in January. March, June, August, and November. so this factor was treated as a fixed effect. Size (slowest- and 127 128 Cigarri'a CANTABRIA.N SEA Figure 1. Eo estuary: culture takes place in the Linera inlet. fastest-growing oysters) and age (2 and 3 years old) were also treated as fixed effects. The appropriate F statistics were calculated following Zar ( 1984). At the beginning of the experiment, both the covariate and dependent variable of the two sets were clearly dif- ferent. Obviously, a comparison of treatment effects using the unadjusted weight means would, therefore, be invalid, because the weight means were already different before the effects were stud- ied. By comparing adjusted means, this problem was overcome (Huitema 1980). Data were logarithmically transformed and tested for normality (Lilliefors test, p > . 1 ( and homogeneity of sample variances (Bart- lett test, p > .05). An allometric regression equation using the least-squares method, was then used to relate dry meat weight M (g) to shell weight S (g) for each sample: Ln ( 100*M) = Ln a + b Ln (100*S). Estimates of adjusted dry meat weight were detransformed for clarity. These were obtained by taking antilogs of the previously log-transformed data. To reduce the bias inherent in the logarith- mic transformation, the antilogs were multiplied by a correction factor (CF) calculated from the standard error of the estimate (SEE) (Sprugel 1983) as simply taking the antilog of a log-normal mean will underestimate the true mean value: CF = exp (SEE2/2). RESULTS The meat/shell regression slopes were not significantly differ- ent (all ps > .05) for the two oyster classes. The adjusted means for the dry meat differed between the two size groups (ANCOVA. F, 279 = 7.65, p < .01). among seasons (ANCOVA. F4 279 = 42.18. p < .001) and between years (ANCOVA, F, 279 = 8.28. p < .01 ). None of the interactions was significant (ANCOVA. all ps > .05). The adjusted dry meat weight of a standard oyster with a shell weight of 39.3 g (the grand covariate mean) showed a similar pattern in the 2 years (Fig. 4). The minimum dry meat weight was found in winter in both years. Thereafter, both size groups rose steadily to reach maximum dry meat values in July, after which time, tissue weight exhibited a continuous decline until March. The increase in adjusted meat weight was coincident with rising water temperatures (Fig. 3), although oysters achieved their maxi- mum meat weight before the peak of water temperature (August). Slow- and fast-growing oysters showed a similar pattern of change in dry meat weight throughout the 2 years, although the fast-growing oysters had significantly more meat than did the slow-growing oysters. An age effect was also significant. The adjusted meat weight of 2- and 3-year-old oysters are calculated on the basis of the grand covariate mean (39.3 g). which allows ex- amination of differences in age without the influence of weight. The linear regression equations of each group were used to obtain adjusted means of dry meat weight for each date (Fig. 4). using the common within-group slope (1.006) obtained from the ANCOVA. Uncoupled rates of growth of shell and tissue may severely affect analysis of seasonal variation in growth in bivalve popula- tions, resulting in apparent changes of adjusted weight that do not necessarily correspond to real losses or gains of weight by the animals (Hilbish 1986. Borrero and Hilbish 1988). In this experi- ment, the growth rates of shell and meat for the two size groups were highly correlated (Set Slow: r = 0.9437. p < .001: Set Fast: r = 0.8474. p < .001). DISCUSSION In allometric studies, it is assumed that a linear log-log rela- tionship exists between two body parts, and that this relationship persists throughout stable growth periods. A structural shift (i.e.. a change in the value of the slope) may. however, occur in this log-linear relationship, as the organism enters a new phase in its life history (Vahl 1984). No differences among slopes of the dry meat-shell weight relationship were found in the studied weight range (from 7 to 160 g, live weight), which means that no changes were detected in the relative allocation of available resources away from meat to shell growth or vice versa. Season Effect Changes in soft tissue weight are typically associated with sea- sonal variation in food availability (Bayne and Newell 1983). the reproductive cycle (Bayne and Worrall 1980) and patterns of en- ergy storage and mobilization (Barber and Blake 1981). In this experiment, both size groups (slow- and fast-growing oysters) showed similar seasonal changes in adjusted dry meat weight. This pattern resembles the productivity pattern of phytoplankton in the study area, with a major peak of chlorophyll a in early summer (2.4 mg/m3) and minimum values in early winter (0.46 rag/m3) (Fundacion Torres-Quevedo 1990). Therefore, the period of meat increase seems to result from increased feeding activity in re- sponse to the spring increase in phytoplankton and temperature, a finding that is in agreement with previous descriptions for the species (e.g.. Malouf and Breese 1977. Heral et al. 1983, Ruiz et 25 20 > 15 10 * B Stow gfow ing oysters D ^rted grow mg oysters Q Fast grow mg oysters J 30-35 40-45 5055 Weight (g) Figure 2. The weight distribution of the oyster population after sub- division into three groups: slow-, medium-, and fast-growing oysters in January 1994. Age, Size, and Season Effects on Oyster Growth 129 Dec-93 Mar-94 Jun-94 Sep-94 dec-94 Mar-95 Jun-95 Sep-95 dec-95 Figure 3. Temperature (°C) and salinity (Sir) in the Eo estuary. al. 1992). The meat increase in adult oysters, in contrast to the increase in young oysters, seems to be related mainly to the for- mation of reproductive tissue (Berthome et al. 1986. Heral 1989), thus the seasonal pattern of meat weight previously described may be primarily attributable to gamete production. Uncoupled rates of meat and shell growth affect "adjusted- weight cycles", where soft tissues are statistically adjusted to a covariate using ANCOVA (Hilbish 1986). In the present study, rates of shell and meat growth were highly correlated; therefore, growth of meat and shell were coincident in both size groups. Seasonal variation in available resources creates a large degree of concordance among growth rates of various body components. even if allocation among components changes seasonally (Peter- son and Fegley 1986). However, seasonal shifts in allocation of resources may prove to be general among bivalve mollusks, al- though different studies on bivalve mollusks have provided con- flicting results (Borrero and Hilbish 1988). In any case, the bias introduced in adjusted means attributable to uncoupled rates of growth (Hilbish 1986) has a substantially effect when somatic grow th is the main cause of meat weight fluctuations, but not when variations attributable to gametogenesis widely exceed somatic growth (Larouelle et al. 1994). Interannual Variability (Age Effect) Differences observed between years are attributable both to interannual variability in environmental conditions and to the ef- fect of age. Therefore, comparisons between years can only be referred to as interannual variability. Variation in meat production (mainly variation in gonad production in adult oysters) between years may originate from variable environmental conditions more than the effect of age. Variability may reflect the ability of this species to adapt the reproductive investment according to environmental conditions, because gamete production is strongly influenced by minor changes in the available food supply and the yearly temperature regime (MacDonald and Thompson 1985. Shumway 1996). There- fore, it would be an appropriate reproductive strategy to invest any surplus energy in gametes when more favorable conditions appear (Hofmann et al. 1992. Ruiz et al. 1992). In this sense. Brown and Hartwick (1988a. Brown and Hartwick 1988b) found that, in com- parison to high-growth sites, oysters at medium-growth sites have lower dry meat weight to shell weight ratios, which suggests that preferential shell thickening may be more energetically efficient under conditions of long-term low food availability. Size Effect Meat weight differed between the two size-groups, although no differences appeared in the timing of growth in shell or meat. This means that, for standard oyster (39.3 g shell weight), the fast- growing oyster has more meat per gram of shell than the slow- growing oyster. It is well known that pronounced differences in 3,5 S 3 i 2,5 n o E £• Q 2 Slow growing oysters Fast growing oysters 1,5 jarv94 mar-94 jun-94 aug-94 nov-94 Jan-95 mar-95 Jun-95 aug-95 nov-95 jan-96 Figure 4. Temporal variation of dry meat weight (g) for slow- and fast-growing oysters during 2 years. 130 ClGARRIA 180 160 140 S 120 f> 100 I 80 o | 60 40 20 Jan-94 Mar-94 Jun-94 Aug-94 Nov-94 Jan-95 Mar-95 Jun-95 Aug-95 Nov-95 Jan-96 Figure 5. Temporal variation of whole weight (±1 SE) of the two size groups during 2 years. size appear in oysters of the same cohort reared under uniform conditions (Walne 1958). These differences were found to be cor- related with the degree of heterozygosity (e.g.. Sigh and Zouros 1978, Koehn and Shumway 1982. Alvarez et al. 1989); whereas, others suggest that physiological variation in growth rate (i.e.. the variation not attributable to the environmental factors) may be attributable as much to variation in growth stimulating peptides as to genetic variation at metabolically important loci or multilocus heterozygosity (Painter and DiMiehelle 1990. Hedgecock 1995). Individual variation in growth rates can be attributed to variation in available food and/or variation in efficiencies with which food derived from available resources can be used for growth. The two size groups in this experiment were cultured at the same density (as number of oysters per bag) and location; thus, fast growing oysters may reflect a greater efficiency of the conversion of con- sumed energy to tissue, faster rates of feeding, or reduced energy requirements for maintenance metabolism (Fig. 5) (Hawkins and Day 1996). For example, in Triostrea chilensis, the net absorption efficiency increases with individual size, explaining at least part of the size differences found within the same cohort (Vergara et al. 1992). On balance, the relation between heterozygosity and growth rate is usually a modest one (r = 0.05); therefore, the association between heterozygosity and fitness traits remain to be determined (Gaffney 1996). The proportion of total production expended on gamete output in the Japanese oyster (C. gigas) rose from 18% in 1-year-old oysters to 84% in the adults (Heral 1989). Hence, as the oyster grows, most of the energy derived from ingestion is partitioned to reproductive processes (Dame 1976, Heral 1989. Thompson et al. 1996). so differences in meat contents are mainly related to dif- ferences in reproductive tissue production. This leads logically to the conclusion that fast-growing oysters are able to invest more in reproductive activity and that this may confer fitness advantages. Because fitness can be broken into two major components, the total number of offspring produced and the quality of these off- spring (Falconer and Mackay 1996). it seems that large size oysters have greater reproductive success. ACKNOWLEDGMENTS The original manuscript was improved by the constructive comments of Dr. A. "Elpidio" Valdes. Dr. A. G. Nicieza, and A. Ojanguren (Univ. Oviedo. Spain). P. Becker (Little Skookum Shellfish Growers. U.S.A.), D. McGoldrick (CSIRO, Australia). R Reed (Reed Mariculture Inc.. U.S.A.), D. Mills (Darwin Aquacul- ture Center). Dr. R. Rheault (Moonstone Oysters, U.S.A.). G. Krause. B. Paust (Univ. of Alaska, U.S.A.). I am especially grate- ful to Dr. Sandra E. Shumway (Southampton College, U.S.A.) and R. Gerwien (Bates College, U.S.A.). This work was funded in part by CULTIVOS MARINOS S.A. from Castropol (Asturies. Spain). LITERATURE CITED Alvarez, G.. C. Zapata. R. Amaro & A. Guerra. 1989. Multilocus heterozy- gosity at protein loci and fitness in the European oyster. Ostrea edulis L. Heredity 63:359-372. Askew. C. G 1972. The growth of oysters. Oslrea edulis and Crassostrea gigas, in Emsworth Harbour. Aquaculture \ ;237-259. Barber, B. J. & N. J. Blake. 1981. Energy storage and utilization in relation to gametogenesis in Argopecten irradians concentricus. J. Exp. Mar. Biol. Ecol. 52:121-134. Bayne, B. L. & C. M. Worrall. 1980. Growth and production of mussels Mytilus edulis from two populations. Mar. Ecol. Prog. Ser. 3:317-328. Bayne. B. L. & R. C. Newell. 1983. Physiological energetics of marine molluscs, pp. 407-515. In: A. S. M. Saleuddin and K. M. Wilbur(eds.). The Mollusca. vol. 4. part 1. Academic Press, New York. Berthome, J. P., J. Prou & A. Bodoy. 1986. Performances de croissance de l'huitre creuse, Crassostrea gigas (Thunberg) dans le bassin d'elevage de Marennes-Oleron entre 1979-1982. Haliotis 15:183-192. Borrero. F. J. & T. J. Hilbish. 1988. Temporal variation in shell and soft tissue growth of the mussel Geukensia demissa. Mar. Ecol. Prog. Ser. 42:9-15. Brown. J. R. 1988. Multivariate analvses of the role of environmental factors in seasonal and site-related growth variation in the Pacific oys- ter Crassostrea gigas. Mar. Ecol. Prog. Ser. 45:225-236. Brown. J. R. & E. B. Hartwick. 1988a. Influences of temperature, salinity, and available food upon suspended culture of the Pacific oyster. Cras- sostrea gigas. I. absolute and allometric growth. Aquaculture 70:231- 251. Brown, J. R. & E. B. Hartwick. 1988b. Influences of temperature, salinity, and available food upon suspended culture of the Pacific oyster. Cras- sostrea gigas. II. condition index and survival. Aquaculture 70:253- 267. Dare, P. J. & D. B. Edwards. 1975. Seasonal changes in flesh weight and biochemical composition of mussels (Mytilus edulis L.) in the Conwy Estuary, North Wales. J. Exp. Mar. Biol. Ecol. 18:89-97. Dame, R. F. 1976. Energy flow in an intertidal oyster population. Estua. Coast. Mar. Sci. 4:243-253. Falconer, D. S.& F. C. Mackay. 1996. Introduction to quantitative genetics. Longman Group, Essex. FAO Fishery Information. 1992. Aquaculture Production 1984-1990. FAO Fisheries Circular 815. Rev. 4. Fundacion Torres Quevedo. 1990. Analisis de las condiciones morfodi- Age, Size, and Season Effects on Oyster Growth 131 namicas de la Rfa del Eo. fase II. informe final. Consejeria de Medio Rural y Pesca, Principau d'Asturies, Unpublished. Gaffney. P. M. 1996. Biochemical and population genetics, pp. 423—437. In: V. S. Kennedy, R. I. E. Newell and A. F. Ehle (eds.). The Eastern oyster, Crassosirea virginica. Maryland Sea Grant College. Maryland. Hawkins. A. J. S. & A. J. Day. 1996. The metabolic basis of genetic dif- ferences in growth efficiency among marine animals. J. Exp. Mar. Biol. Ecol. 203:93-115. Hedgecock, D. 1995. The cupped oyster and the Pacific oyster, pp. 1 15- 137. In: J. E. Thorpe, G. A. E. Gall, J. E. Lannan, andC. E. Nash (eds.). Conservation of Fish and Shellfish Resources: Managing Diversity. Academic Press. London. Heral, M., J. M. Deslous-Paoli & J. M. Somin. 1983. Transferts energet- iques entre l'huitre Crassostrea gigas et la nourriture potentielle dis- ponible dans un bassin ostreicole: premieres approaches. Oceani.s 9: 169-194. Heral, M. 1989. Traditional oyster culture in France, pp. 295-336. In: G. Bamabe (ed.l. Aquaculture, vol I, Tech. et Docum, Lavoisier, Paris. Hilbish, T. J. 1986. Growth trajectories of shell and soft tissue in bivalves: seasonal variations in Mytilus edulis. L. J. Exp. Mar. Biol. Ecol. 96: 103-113. Hoffmann E. E., E. N. Powell, J. M. Klinck & E. A. Wilson. 1992. Mod- eling oyster populations III. critical feeding periods, growth, and re- production. J. Shellfish Res. 11:399-416. Huitema. B. E. 1980. The analysis of covariance and alternatives. John Wiley & Sons, New York. Koehn, R. K. & S. E. Shumway. 1982. A genetic/physiological explanation for differential growth rate among individuals of the American oyster. Crassostrea virginica (Gmelin). Mar. Biol. Lett. 3:35-42. Larouelle, F.. Guillou. J. & Y. M. Paulet. 1994. Reproductive pattern of the clams, Ruditapes decussatus and R. Philippinarnm on intertidal flats in Brittany. J. Mar. Biol. Ass. U. K. 74:351-366. Lucas. A. & P. G. Benmger. 1985. The use of physiological condition indices in marine bivalve aquaculture. Aquaculture 44:187-200. MacDonald, B. A. & R. J. Thompson. 1985. Influence of temperature and food availability on the ecological energetics of the giant scallop Pla- copeclen magellanicus. II. reproductive output and total production. Mar. Ecol. Prog. Ser. 25:295-303. Mallet. A. L. & C. E. Carver. 1993. Temporal production in various size groups of the blue mussel. J. Exp. Mar. Biol. Ecol. 170:75-89. Malouf, R. E. & W. P. Breese. 1977. Seasonal changes in the effects of temperature and water flow rate on the growth of juvenile Pacific oysters Crassostrea gigas (Thunberg). Aquaculture 12:1-13. Painter, K. T. & L. DiMichelle. 1990. Growth of tray-cultured oysters (Crassostrea gigas Gmelin) in the Chesapeake Bay. Aquaculture 87: 289-297. Peterson. C. H. & S. R. Fegley. 1986. Seasonal allocation of resources to growth of shell, soma, and gonads in Mercenaria mercenaria. Biol. Bull. 171:597-610. Quayle. D. B. 1969. Pacific oyster culture in British Columbia. Fish. Res. Board Can. Bull. 169:1-192. Raubenheimer. D. & P. Cook. 1990. Effects of exposure to wave action on allocation of resources to shell and meat growth by the subtidal mussel, Mytilus galloprovincialis. J. Shellfish Res. 9:87-93. Ruiz. C, M. Abad. F. Sedano. L. O. Garcfa-Martin & J. L. Sanchez. 1992. Influence of seasonal environmental changes on the gamete production and biochemical composition of Crassostrea gigas (Thunberg) in sus- pended culture in El Grove. Galicia. Spain. J. Exp. Mar. Biol. Ecol. 155:249-262. Shumway, S. E. 1996. Natural environmental factors, pp. 467-503. In: V. S. Kennedy. R. I. E. Newell and A. F. Eble (eds.). The Eastern oyster, Crassostrea virginica. Maryland Sea Grant College, Maryland. Singh. S. M. & E. Zouros. 1978. Genetic variation associated with growth rate in the American oyster (Crassostrea virginica). Evolution 32:342- 353. Sprugel, D. G. 1983. Correcting for bias in log-transformed allometric equations. Ecology 64:209-210. Thompson. R. J.. R. I. E. Newell. V. S. Kennedy & R. Mann. 1996. Re- productive processes and early development, pp. 335-364. In: V. S. Kennedy, R. I. E. Newell and A. F. Eble (eds.). The Eastern oyster. Crassostrea virginica. Maryland Sea Grant College. Maryland. Vahl. O. 1984. Size-specific reproductive effort in Chlamys islandica: reproductive senility or stabilizing selection? Proceedings of the 19th European Marine Biology Symposium. Plymouth England, pp. 521- 527. Vergara. A. M., J. E. Toro & K. A. Pashke. 1992. Relaciones entre vari- ables fisiologicas, heterocigosidad, y eficiencia de crecimiento en Tri- ostrea chilensis. (Philippi 1845). Rev. Biol. Mar. 27:283-297. Walne, P. R. 1958. Growth of oysters (Ostrea edulis L.). /. Mar. Biol. Assoc. U. K. 37:591-602. Walne. P. R. & R. Mann. 1975. Growth and biochemical composition in Ostrea edulis and Crassostrea gigas. pp. 587-607. Proceedings of the 9th European Marine Biology Symposium. University Press. Aber- deen. Scotland. Zandee. D. L, J. H. Kluytmans, W. Zurburg, H. Pieters. 1980. Seasonal variations in biochemical composition of Mytilus edulis with reference to energy metabolism and gametogenesis. Neth. J. Sea. Res. 14:1-29. Zar. J. H. 1984. Biostatistical analysis. Prentice-Hall International, Lon- don. Journal of Shellfish Research, Vol. IS, No. 1. 133-138. 1999. USE OF L-DOPA AND SOLUBLE BACTERIAL PRODUCTS TO IMPROVE SET OF CRASSOSTREA VIRGINICA (GMELIN, 1791) AND C. GIGAS (THUNBERG, 1793) MARIANNE WALCH,1 2 RONALD M. WEINER,2,3 RITA R. COLWELL,2,3 AND STEVEN L. COON4'* Naval Surface Warfare Center, Carderock Division, Bethesda, Maryland 20817 'Center of Marine Biotechnology, MBI, University of Maryland, Baltimore, Mankind 21202 Department of Cell Biology and Molecular Genetics, University of Maryland, College Park, Maryland 20742 4 Department of Biology, University of Maryland, College Park, Mankind 20742 ABSTRACT Solutions of L-dihydroxyphenylalanine (L-DOPA) and culture supernatants of the marine bacterium Shewanella col- welliana were used to treat competent Crassostrea larvae before transferring them to setting tanks. Setting trials were done in laboratory containers and on a larger scale in a Chesapeake Bay hatchery. Hatchery results showed a consistent and repeatable improvement in set of C. Virginia) after treatment with either L-DOPA or bacterial supernatants. Laboratory results were more variable, and differences between treatments were less pronounced. Pretreatment of larvae with theophylline led to a more rapid onset of settlement behavior in L-DOPA treatments but not in supernatant treatments, and had no clear effect on over-all set on cultch. KEY WORDS: Crassostrea gigas, Crassostrea virginica. oyster set. set cues, bacteria. Shewanella colwelliana, L-DOPA. theoph- ylline INTRODUCTION Microbial surface films and their products are known to medi- ate the settlement and metamorphosis of a variety of marine in- vertebrates (Mitchell 1984. Maki and Mitchell 1986). Specific bac- terial products have been implicated in the induction of settlement and morphogenesis of scyphozoans (Neumann 1979), polychaete worms (Kirchman et al. 1982a, Kirchman et al. 1982b). bryozoans (Brancato and Woollacott 1982), sea urchins (Cameron and Hine- gardner 1974), and oysters (Weiner et al. 1985, Weiner et al. 1989). Metamorphosis of the red abalone. Haliotis albescens, is induced by metabolic products of the crustose coralline algae on which they set (Morse et al. 1979). Inductive factors may be such soluble compounds as peptides or protein conjugates (Fitt et al. 1987, Siddall 1982) or fatty acids (Pawlik 1986). or they may be such insoluble components of the microbial film itself as polysac- charides or glycoproteins (Kirchman et al. 1982b). Whether larvae that respond to bacterial products require direct contact with mi- crobial films or sense soluble cues released from them is not known for most species. Previous studies have shown that larvae of Crassostrea vir- ginica and C. gigas undergo settlement behavior and metamor- phosis in response to the presence of the neurotransmitter precur- sor L-3,4-dihydroxyphenylalanine (L-DOPA) (Coon et al. 1985, Coon et al. 1990a). Settlement and metamorphosis of these oysters also is enhanced by compounds produced by the marine bacterium. 'Present address: Laboratory of Developmental Neurobiology, National Institutes of Health, Bethesda, Maryland 20892 USA. initially classified as Alteromonas colwelliana (Weiner et al. 1988) and now classified as Shewanella colwelliana (Coyne et al. 1989). which was isolated from oyster setting tanks (Weiner et al. 1985, Weiner et al. 1989). One or more soluble cues, as well as an extracellular polysaccharide component of S. colwelliana films seem to be involved in this process, but the exact mechanism of induction by bacterial products is not yet clear (Weiner et al. 1989. Bonar et al. 1990). Fitt et al. (1989) demonstrated that oyster larvae are able to respond behaviorally and complete metamorphosis after exposure to culture media supernatants of several species of bac- teria, including S. colwelliana. There has been much speculation about the biotechnological potential of controlling specific microbial-invertebrate interac- tions in aquaculture (Bonar et al. 1986). Use of exogenous micro- bial or "natural" cues to enhance settlement or growth of such economically valuable species as oysters could contribute signifi- cantly to increased fisheries production. Until now, however, suc- cessful use of microbial products to improve shellfish yields has not been commercially utilized. In this report, we extend laboratory findings (Coon et al. 1985. Walch et al. 1989, Bonar et al. 1990. Fitt et al. 1990) to the hatchery scale and demonstrate that exogenous inducers of oyster settlement behavior can, in fact, be used to increase spat set. We have used solutions of L-DOPA or S. colwelliana supernatants to treat competent larvae before transferring them to setting tanks. Setting trials were done on a small scale in laboratory containers as well as on a larger scale in a Chesapeake Bay hatchery. The potential and limitations of this technology for use in oyster hatch- eries is discussed. 133 134 Walch et al. MATERIALS AND METHODS Oyster Larvae Competent larvae of the Eastern oyster. Crassostrea virginica, were obtained from the Maryland Department of Natural Re- sources (DNR) oyster hatchery at Deal Island, Maryland. Larvae of the Pacific oyster, Crassostrea gigas, were obtained from the Coast Oyster Company in Quilcene. Washington. Larvae used in the in vitro experiments were maintained in the laboratory, em- ploying procedures described by Coon et al. (1990a). Chemicals L-3.4-dihydroxyphenylalanine (L-DOPAl and theophylline were obtained from Sigma Chemical Co. (St. Louis. Missouri). L-DOPA was dissolved in 0.005 N HC1 at a concentration of 10"3 M prior to use, then was diluted with 0.2-p.m filtered seawater to 10~* M or 10"5 M for treatment of the larvae. Theophylline ( 10^3 M) was dissolved directly in seawater. Bacterial Supernatants The bacterial species Shewanella colwelliana, strain LST-D (Weiner et al. 1985. Weiner et al. 1989b). was grown at 25°C in 500 mL Marine Broth 2216 (Difco) in 2-L baffled flasks on a rotary shaker. Cultures were harvested in the late logarithmic stage of growth (2 days), centrifuged at 10.000 g for 15 min, and the supernatants were frozen for use in larval treatments and behav- ioral assays. Behavioral Assays Assays of settlement "search" behavior (Coon et al. 1985) were performed in Costar 24-well tissue culture plates as described by Coon et al. (1990a). Twenty to 40 larvae were placed in each well with 500 (0.1 of 0.45 u.m filtered seawater (159£c salinity for C. virginica, 30%r for C. gigas). Bacterial supernatants or L-DOPA were added to the wells in varying concentrations, and the number of larvae swimming or crawling with the foot extended during a 30-s period was recorded at 5 to 10-min intervals for up to 1 h. Results are expressed as the mean percentage behavior of three replicates. In Vitro Setting Experiments Clean. 8-inch diameter glass bowls were prepared with 1 liter of filtered seawater. at the appropriate salinity, and equal numbers of aged C. virginica shells were selected to be approximately equal in size (5x8 cm) and relatively flat. They were washed lightly with a clean brush and tap water before use. Competent larvae were treated in beakers containing 500 to 1.000 larvae in 500 mL of seawater with the following additions: (1) control, no additions; (2) L-DOPA. 10"4 M (10 min); (3) L- DOPA. 10"5 M (30 min); (4) theophylline. 10~3 M (5 min), and L-DOPA. 10"4 M (additional 10 min); (5) S. colwelliana super- natant. 25% (10 min); and (6) theophylline. 10~3 M (5 min), and S. colwelliana supernatant, 25% (additional 10 min). After treatment for the specified length of time, the larvae were sieved out of the beakers, rinsed with seawater, and distributed as evenly as possible into the prepared bowls. The bowls were cov- ered and left at 20°C for 4 days. The number of spat set on each shell was counted with a dissecting microscope. In another experiment, four 5-gallon glass aquaria were pre- pared containing 1-p.m filtered seawater and clean oyster shell cultch in nylon mesh bags (100 similarly sized and shaped shells per tank). C. gigas larvae were treated, as described above, in beakers containing 500 mL seawater. with the following additions: (1) control, no additions; (2) L-DOPA. 10~4 M (10 min): (3) the- ophylline. 10~3 M (5 min). and L-DOPA. 10"4 M (additional 10 min); and (4) L-DOPA. 10"5 M (30 min). Treated larvae were transferred to the aquaria (final density 3^1 larvae per mL). Tanks were aerated gently, and larvae were fed during set. The number of spat on each shell was counted after 4 days. Hatchery Setting Experiments Hatchery trials were conducted at the Maryland DNR Oyster Hatchery at Deal Island, Maryland. Setting was done in 2.4 x 1.2 x 0.3-meter wooden trays containing course-filtered, aerated bay water. C. virginica larvae were treated before set in the manner described above and distributed as evenly as possible into prepared setting tanks. Approximately 500.000 larvae were added to each tank, giving a concentration of about 40 larvae per Liter. Larvae were fed cultured Isochrysis galbana regularly during set. In the first set of hatchery trials. 300-400 clean, similarly sized and shaped aged shells were arranged interior side up in a mono- layer on the bottom of each setting tray. Larvae were treated as follows and distributed into separate tanks. Experiment 1 ( 1 ) control, no treatment (2) theophylline. 1CT3 M (5 nun), and L-DOPA. 10"4 M (ad- ditional 10 min) Experiment 2 ( 1 ) Control, no treatment (2) Theophylline. 10"3 M (5 min). and L-DOPA. 10~4 M (ad- ditional 10 mini (3) S. colwelliana supernatant. 259c (10 min) After 4 days, the shells were removed from the tanks and washed gently under a stream of seawater to remove unattached larvae. The number of spat on all shells was counted. The following summer, an additional experiment was con- ducted using clean, aged oyster shell cultch in nylon mesh bags (100-150 shells per bag), which more nearly mimieks routine hatchery practices. Larvae were treated with 10"4 M L-DOPA for 10 min and distributed in one of the setting tanks. Untreated larvae were added to the second tank as a control. After set. the cultch were placed in a growout raceway. Spat on both sides of 120 random shells from three bags in each tank were counted after about 4 weeks. RESULTS Behavioral Assays 10~4 M L-DOPA induced settlement behavior within 20 min in nearly 100% of Crassossrea larvae. 10~5 m L-DOPA also induced behavior, but more slowly and at a lower level (70% of larvae behaving after 30 min) (Fig. 1 A). Treatment of larvae for 10 min with various concentrations of bacterial supernatants also induced a hish level of settlement behavior (Fig. IB). Behavior was sus- Oyster Set Using L-DOPA and Bacterial Products 135 TABLE 1. Results of in vitro setting experiments in glass bowls. Time, minutes Time, minutes Figure 1. Induction of settlement behavior in Crassostrea gigas larvae by L-DOPA and bacterial supernatants. (A) L-DOPA treatments: sea- water control (open circles), 1(T5 M L-DOPA (filled circles), 10~4 M L-DOPA (open triangles), theophylline + 10"* M L-DOPA (fdled tri- angles). (B) Bacterial supernatant treatments: theophylline control (open circles), 12.5% supernatant (filled circles), 25% supernatant (open triangles), 50% supernatant (filled triangles), theophylline + 25% supernatant (open squares). tained over the longest period when the supernatants were diluted to 25 or 12.5%. Pretreatment of larvae for 5 min with 10 M theophylline caused them to respond faster to L-DOPA. Maximum settlement behavior was observed in theophylline-treated larvae at least 10 min sooner than in those treated with I0"4 M L-DOPA alone (Fig. 1A). Theophylline pretreatment had no significant effect on settle- ment behavior induction by bacterial supernatants (Fig. IB). In Vitro Setting Experiments Results of small-scale setting experiments in glass bowls are shown in Table 1. For both C. virginica and C. gigas. the mean spat set for all treatments were higher than the controls. In the case of C. gigas, these means were found not to be significantly dif- ferent (p > .05. single classification analysis of variance (ANOVA)]. Treatment of C. virginica larvae, however, did result in significant differences in spat set (p < .001 ). The largest increase in spat set was obtained by treatment with 10"' M L-DOPA for 30 min. When C. gigas larvae were set on cultch in glass aquaria, again, the means for spat set were higher in all treatments than in the Spat jer Shell Time (mean i SE , n = 6) Larval Treatment C. virginica* C. gigas f Control 10 min. 2.0 ± 0.8ab 68.0 ±23.1 lCT1 M L-DOPA 10 min. 14.0 ± 3.3a 102.3 ± 19.8 irr5 m l-dopa 30 min. 70.3 ± 19.7 120.0 + 41.8 Theophylline 5 min. + + ltr4 M L-DOPA 10 min. 6.8 ± 4.0','b 122.3 ±29.6 S. colwelliana supernatant 10 min. 15.5 ±3.8" 69.5 ± 16.5 Theophylline + S. colwelliana 5 min. + supernatant lo min. 24.2 ± 6.0" b 99.5 ± 26.7 * Treatment means with the same superscript letter are not significantly different. t No treatment means for C. gigas are different from the control. control (Table 2). ANOVA showed only the 10~4 m DOPA treat- ment to be significantly better than the control (p < .05). The three larval treatments were not significantly different from one another. Hatchery Setting Experiments Results of hatchery setting trials conducted during both seasons demonstrated that treatment of larvae with L-DOPA. with theoph- ylline and L-DOPA. or with S. colwelliana supernatants signifi- cantly improved set of C. virginica on cultch (Fig. 2. Table 3). In the experiment done the second summer, in which larvae were treated for 10 min with I0"4 M L-DOPA and set on bagged cultch. the treatment increased set by about 30% over untreated controls (p < .05, Mann-Whitney U-test) (Fig. 2). This figure also shows that the outside surfaces of the cultch consistently had more spat on them than the interior surfaces. This was true in nearly all of our setting experiments. Setting trials conducted the first season demonstrated very sig- nificant increases in spat set. Normal spat set that summer was quite low, with cultch in control tanks (no larval treatment) having means of less than one spat per shell and a maximum of 10 per shell (Table 3). In other tanks containing the same cohorts of larvae treated with theophylline and 10"4 m L-DOPA for 10 min. mean spat set was 3=12 per shell, with some shells having more than 100 spat (Table 3, expt. 2). Treatment of larvae with a 1:3 dilution of 2-day-old S. colwelliana culture supernatant led to an increase in spat set nearly identical to that of the L-DOPA treat- TABLE 2. Results of in vitro C. gigas set on cultch bags in aquaria. Spat per Shell Larval Treatment (Mean ± SE, n = 100)* Control 10"4 M L-DOPA 10-5 M L-DOPA Theophylline + 10" M L-DOPA 19.3 ± 1.9a 28.0 ± 2.2" 24.6 ±23'" 23.5 ± 2.2ab * Treatment means with the same superscript letter are not significantly different. 136 Walch et al. ment (Table 3, expt. 2). These increases were statistically signifi- cant (p < .01. Mann-Whitney U-test). DISCUSSION Many environmental factors have been shown to influence oys- ter settlement and metamorphosis, including naturally occurring chemicals isolated from oyster habitats or from oyster tissues themselves (Crisp 1967, Degens et al. 1967. Veitch and Hidu 1971, Waite and Anderson 1980. Bonar et al. 19891. A distinction can be made between factors that promote settlement (a reversible behavior) and those that induce metamorphosis (irreversible mor- phogenesis) (Bonar et al. 1990). Coon et al. (1985) have shown that L-DOPA induces settlement behavior in oyster larvae. The response consists of swimming with the foot extended and crawl- ing on the substratum. The relationship of L-DOPA to the pre- sumed natural inducer is unknown, but it consistently induces a sequence of settlement behaviors in >90% of competent Crassos- trea larvae (Coon et al. 1985. Coon et al. 1990a. and Table 1). It also has been demonstrated that one or more soluble factors produced by certain marine bacteria cue oyster larval search be- havior (Fitt et al. 1989. Weiner et al. 1989b. Fitt et al. 1990). Surface films of one particular bacterium. S. colwelliana, are es- pecially active in promoting set on a variety of substrata (Weiner et al. 1985. Weiner et al. 1989b). Studies in our laboratories have shown that a number of metabolites produced by S. colwelliana; that is. ammonia and L-DOPA. induce settlement behavior in Crassostrea larvae (Dagasan and Weiner 1988, Walch et al. 1989. Weiner et al. 1989b). The major inducer present in S. colwelliana laboratory culture supernatants has been shown to be ammonia (Coon et al. 1990b. Walch et al. in preparation). A model of oyster set proposes that these soluble cues induce competent larvae to drop to the substratum and begin sampling the surface, looking for an appropriate settlement site. Secondary cues associated with the surface, such as specific exopolysaccharides. are required for ce- 80 70 i> 60 ss CO i. 50 cu 40 a, CO c 30 a a> 2 on 10 5 o 3 o Control L-DOPA Treatment of Larvae Figure 2. Results of the hatchery setting experiment evaluating the effect of L-DOPA treatment on spat set using shell cultch in bags. Data represent mean no. spat per shell ± SE (n = 120). TABLE 3. Results of hatchery setting experiments evaluating set of C. virginica on cultch after pretreatment of larvae, numbers represent mean no. spat per shell ± SE (n). Experiment #1* Experiment #2* Control Theophylline + ltr* M DOPA 5. colwelliana supernatant 0.2 ± 0.04 (380)" 2.1 ±0.8 (377)b 0.6 ± 0.08 (343)a 12.2 ±1.2 (345)b 16.0 ±2.1 (309)b * Treatment means with the same superscript letter within an experiment are not significantly different. mentation and metamorphosis (Bonar et al. 1990, Weiner et al. 1989a. Weiner et al. 1989b). The effectiveness of such surface cues may be enhanced by such additional bacterial products as homogentisic acid (Coon et al. 1994, Weiner et al. in preparation). Given that L-DOPA and soluble bacterial metabolites induce settlement behavior in oyster larvae and that this behavior nor- mally is a prerequisite for cementation and metamorphosis, it was hypothesized that exposure of larvae to these chemicals would increase total spat set. Our results confirm this, at least for the conditions described here. Other studies have shown that L-DOPA induces metamorphosis in C. gigas (Henderson 1981. Cooper 1983. Coon et al. 1985). In general, however, they have failed to demonstrate increases in percentage set on cultch surfaces. Cooper ( 1983), for example, found that, although L-DOPA treatments con- sistently increased the percentage of larvae completing metamor- phosis (20-30% increase relative to controls), many of these larvae underwent morphogenesis without attachment, sometimes result- ing in a net decrease in total spat set on the cultch. Results of hatchery experiments described here show a consis- tent and repeatable improvement in set of C. virginica after treat- ment with either L-DOPA or bacterial culture supernatants. Labo- ratory results were more variable, and differences between treat- ments were, in general, less pronounced. Differences between treatments also tended to be greater for C. virginica than for C. gigas; this may be related to our experience that C. gigas normally has a higher percentage set than C. virginica. The response of larvae to soluble inducers has been shown to vary considerably with larval age and condition as well as a number of environmental factors (Cooper 1983, Coon et al. 1990a. Fitt et al. 1990). This undoubtedly contributed to the variability inherent in our own experiments, as well as differences between our results and those of other researchers. In general, improvement of set was most dramatic when the larvae were otherwise setting poorly, as was the case with C. virginica during the summer in which the first set of trials was conducted (Table 3). Although C. virginica and C. gigas react qualitatively the same to soluble cues (Coon et al. 1990a), our studies have consistently shown that C. virginica did not set as well as C. gigas in any experimental regime (Weiner et al. 1989b). In the hatchery trials described here, a single concentration of L-DOPA (10-4 M) was used for larval treatments. Laboratory studies have indicated that, although larvae respond to 10~5 M DOPA more slowly and in somewhat lower numbers (Fig. 1A). they tend to spend more time exhibiting settlement behavior than those exposed to higher concentrations. This may be an advantage in hatchery setting systems, resulting in more even and consistent set throughout the cultch bags. Behavioral assays (Fig. 1A) demonstrate the rationale for pre- treating larvae with theophylline before exposing them to L- Oyster Set Using L-DOPA and Bacterial Products 137 DOPA. Theophylline is a caffeine analog known to inhibit the enzyme adenylate cyclase, resulting in higher intracellular pools of cyclic AMP. Cyclic AMP may be involved in mediating the L- DOPA-indticed behavioral response in Crassostrea, but this has not been proved (Bonar et al. 1990). Our results showed that pre-exposure of larvae to theophylline caused a much faster be- havioral response relative to L-DOPA treatments alone (Fig. 1A). Furthermore, we observed that the theophylline-treated larvae were "stickier": they spent much more time crawling than those treated only with L-DOPA. Despite the positive effect on settle- ment behavior, we were surprised that comparisons of theophylline treatments and L-DOPA controls failed to demonstrate any sig- nificant differences in total set (Tables 1 and 2), although casual observations revealed substantially more nonspecific set on the bottoms of the tanks when larvae were treated with theophylline before L-DOPA. It is interesting to note that theophylline did not have a signifi- cant effect on the behavior of larvae exposed to S. colwelliana supernatants (Fig. IB). The behavioral response of Crassostrea larvae to bacterial culture supernatants and to L-DOPA seem to be similar, but the primary inducer in the supernatants away from the biofilms has been shown to be ammonia rather than L-DOPA. Whereas L-DOPA-induced settlement behavior is a receptor- mediated response and may involve cAMP as a second messenger. ammonia-induced behavior seems to be a nonspecific response to intracellular pH changes (Bonar etal. 1990. Coon et al. 1990b) that would not be sensitive to cAMP levels. The results reported here indicate that treatment of competent oyster larvae with exogenous inducers of settlement behavior may be a useful technology for improving spat yields in hatcheries. This is the first report of the use of bacterial metabolites to increase set in the field. Current work focuses on optimization of treatment conditions and evaluation of growth and survival of the treated spat. Work is also in progress to improve the effectiveness and yield of soluble bacterial inducers through biotechnological ap- proaches. ACKNOWLEDGMENTS This project was supported by Grant S- 124-88-008 from the Maryland Department of Natural Resources, by Grant #NA88- AAD-00014 from the Maryland Sea Grant College, by University Research Initiative Grant #N00014-86K-0696 from the Office of Naval Research and by a University of Maryland Graduate School Semester Research Award to R.M.W. We thank Coast Oyster Company and the Maryland Department of Natural Resources for providing oyster larvae and facilities for these experiments. We also acknowledge William Schafer for his help with statistical analvses. LITERATURE CITED Bonar. D. B., R. M. Weiner & R. R. Colwell. 1986. Microbial-invertebrate interactions and potential for biotechnology. Microb. Ecol. 12:101- 110. 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. Brancato. M. S. & R. M. Woollcott. 1982. Effect of microbial films on settlement of bryozoan larvae iBugula simplex, B. stolonifera, B. tur- rila). Mar. Biol. 71:51-56. Cameron. A. M. & R. T. Hmegardner. 1974. Initiation of metamorphosis in laboratory-cultured sea urchins. Biol. Bull. 146:335-342. Coon. S. L.. D. B Bonar & R. M. Weiner. 1985. Induction of settlement and metamorphosis of the Pacific oyster Crassostrea gigas (Thunberg) by L-DOPA and catecholamines. J. Exp. Mar. Biol. Ecol. 94:211-221. Coon. S. L., W. K Fitt & D. B. Bonar. 1990a. Competence and delay of metamorphosis in the Pacific oyster, Crassostrea gigas (Thunberg). Mar. Biol. 106:379-387. Coon. S. L.. M. Walch. W. K. Fitt. D. B. Bonar & R. M. Weiner. 1990b. Ammonia induces settlement behavior in oyster larvae. Biol. Bull. 179: 297-303. Coon. S. L.. S. Kotob. B. Jarvis. S. Wang. W. C. Fuqua & R. M. Weiner. 1994. Homogentisic acid is the product of Mel A which mediates mel- anogenesis in the marine bacterium, Shewanella colwelliana D. Appl. Environ. Microbiol. 60:3006-3010. Cooper. K. 1983. Potential for application of the chemical Dopa to com- mercial bivalve setting systems. J. Shellfish Res. 3:1 10-1 1 1 . Coyne. V. E.. C. J. Pillidge, D. D. Sledjeski, H. Hon. B. A. Ortiz-Conde. D. G. Muir. R. M. Weiner & R. R. Colwell. 1989. Reclassification of Alteromonas colwelliana to the genus Shewanella by DNA:DNA hy- bridization, serology, and 5S ribosomal RNA sequence data. System. Appl. Microbiol. 12:275-279. Crisp. D.J. 1967. Chemical factors inducing settlement in Crassostrea virginica Gmelin. J. Animal Ecol. 36:329-336. Dagasan. L. & R. M. Weiner. 1988. Characterization and in vitro transla- tion of an atypical tyrosinase from a periphytic marine bacterium. Abstracts of 3rd Annual American Society of Microbiology Confer- ence of Biotechnology, p. 43. Degens. E.. W. Spencer & R. H. Parker. 1967. Paleobiochemistry of mol- luscan shell proteins. Comp. Biochem. Physiol. 20:553-579. Fitt. W. K... D. K. Hofmann. M. Wolk & M. Rahat. 1987. Requirement of exogenous inducers for metamorphosis of axenic larvae and buds of Cassiopea andromeda (Cnidaria: Scyphozoa). Mar. Biol. 94:415-422. 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 influencing bac- terial production of inducers of settlement behavior of larvae of the oyster Crassostrea gigas. Microb. Ecol. 17:287-298. Fitt. W. K„ S. L. Coon. M. Walch. R. M. Weiner. R. R. Colwell & D. B. Bonar. 1990. Settlement behavior and metamorphosis of oyster larvae of Crassostrea gigas in response to bacterial supernatants. Mar. Biol. 106:389-394. Henderson. B. A. 1981. Practical methods of handling and setting eyed Pacific oyster larvae. /. Shellfish Res. 2:1 19-120. Kirchman, D.. S. Graham. D. Reish & R. Mitchell. 1982a. Bacteria induce settlement and metamorphosis of Janua (Dexiospira) brasiliensis Grube (Polychaeta: Spirorbidae). J. Exp. Mar. Biol. Ecol. 56:153-163. Kirchman. D.. S. Graham. D. Reish & R. Mitchell. 1982b. Lectins may mediate in the settlement and metamorphosis of Januaf Dexiospira) brasiliensis Grube Polychaeta: Spirorbidae). Mar. Biol. Lett. 3:131— 142. Maki, J. S. & R. Mitchell. 1986. The function of lectins in interactions among marine bacteria, invertebrates, and algae. In: D. Mirelman (ed.). Microbial Lectins and Agglutinins: Properties and Biological Activity. John Wiley, New York. Mitchell. R. 1984. Colonization by higher organisms, pp. 189-200. In: K. C. Marshall (ed.). Microbial Adhesion and Aggregation. Springer Verlag, Berlin. Morse. D. E.. N. Hooker, H. Duncan & L. Jensen. 1979. Gamma- aminobutyric acid, a neurotransmitter, induces planktonic abalone lar- vae to settle and begin metamorphosis. Science 204:407-410. Neumann. R. 1979. Bacterial induction of settlement and metamorphosis in the planula larvae of Cassiopea andromeda (Cnidaria: Scyphozoa. Rhizostomeae). Mar. Ecol. Prog. Ser. 1:21-28. Pawlik. J. R. 1986. Chemical induction of larval settlement and metamor- phosis in the reef-building tube worm Phragmatopoma californica (Sa- bellariidae: Polychaeta). Mar. Biol. 91:59-68. 138 Walch et al. Siddall. S. E. 1982. Biological and economic outlook for hatchery produc- tion of juvenile queen conch. Proc. Gulf Caribb. Fish Inst. 36:46-52. Veitch, F. P. & H. Hidu. 1971. Gregarious setting in the American oyster Crassostrea virginica Gmelin: 1 . properties of a partially purified "set- ting factor." Chesapeake Sci. 12:173-178. Waite, J. H. & S. O. Anderson. 1980. 3.4-dihydroxyphenylalanine in an insoluble shell protein of Mytilus eclulis. Biochem. Biophys. Acta 541: 107-114. Walch. M., L. Dagasan, S. Coon. R. Weiner. D. Bonar & R. R. Colwell. 1989. Identification of soluble microbial products that induce settle- ment behavior of oyster larvae {Crassostrea spp.). J. Shellfish Res. 7:551. Weiner, R. M. A. M. Segall & R. R. Colwell. 1985. Characterization of a marine bacterium associated with Crassostrea virginica (the Eastern oysterl. Appl. Environ. Microbiol. 49:83-90. Weiner R. M.. V. E. Coyne. P. Brayton. P. West & S. F. Raiken. 1988. Alteromonas-colwelliana sp-nov. an isolate from oyster habitats. Int. J. System. Bacteriol. 38:240-244. Weiner. R. M. M. Walch, C. Fuqua. D. Sledjeski, L. Dagasan & S. Coon. 1989a. Molecular cues of Crassostrea set that are synthesized by bac- teria. J. Shellfish Res. 8:459. Weiner, R. M„ M. Walch. M. P. Labare, D. B. Bonar & R. R. Colwell. 1989b. Effect of biofilms of the marine bacterium Alteromonas col- welliana (LST) on set of the oysters Crassostrea gigas (Thunberg, 1793) and C. virginica (Gmelin, 17911. /. Shellfish Res. 8:117-123. Journal of Shellfish Research. Vol. 18. No. 1. 139-146, 1999. GROWTH AND BIOCHEMICAL COMPOSITION OF CRASSOSTREA GIGAS (THUNBERG) AND OSTREA EDULIS (LINNE) IN TWO ESTUARIES FROM THE NORTH OF PORTUGAL MARIA JOSE ALMEIDA, JORGE MACHADO, AND JOAO COIMBRA Laboratorio de Fisiologia Aplicada and CIMAR — Centro de Investigacdo Marinha e Ambiental, Institute de Ciencias Biomedicas Abel Salazar, Largo do Prof. Abel Salazar 2. 4050 Porto. Portugal ABSTRACT Oysters C. gigas and O. edulis were grown in hanging cultures at river Lima estuary and Ria of Aveiro. in the north of Portugal. Growth and biochemical composition were recorded between May 1990 and October 1991. Temperature, salinity, and particulate matter in the water, at both stations, were monitored at the same time. No differences were found in the environmental parameters between stations, although the oysters presented some differences, mainly in biochemical composition. C. gigas at both locations showed some shell thickening, evaluated through the shell thickness index (STI). The presence of butyltin compounds in the water seems to be the most reasonable explanation for the anomalous thickening and the highest constraint to the development of commercial oyster farms. Both stations have good trophic conditions for culturing the two species. KEY WORDS: biochemical composition, Crassostrea gigas. Lima estuary. Ostrea edulis. oyster culture, Ria of Aveiro, TBT INTRODUCTION Oyster production in Portugal was very important at the end of 1960s, especially that of the Crassostrea angulata species, known as the Portuguese oyster. This species, which is very similar to the Japanese oyster Crassostrea gigas (their hybrids originate from similar progeny) (see review by Gaffney and Allen 1993). was intensively exploited, mainly at the river Sado estuary, reaching a record number of exports at 10,448 tons in 1968 (Vilela 1975). In the 1970s, sanitary problems and an increase in industrial pollution at the river Sado estuary resulted in a dramatic decrease in Portu- guese oyster stocks. Some years before, similar condition affected oysters from river Tejo estuary. At the same time, in France, high mortalities in the Portuguese oysters, caused by gill disease (which still persists), led to a crisis in the French oyster industry (the largest in Europe). In Portugal, oyster production practically disappeared, but in France, oyster culture quickly recovered, with the introduction of C. gigas from Japan. C. gigas is being cultured in increasing numbers worldwide. The attraction of this species is its efficiency as a filter feeder, its fast growth rate, and its tolerance of a wide range of physical conditions, such as temperature, salinity, and silt load in the water (Quayle 1969, Bardach et al. 1972, Shpigel and Blaylock 1991). On the north coast of Portugal, natural stocks of the European oyster (O. edulis. Linne) can be found in some areas where there is potential for cultivation of both O. edulis and C. gigas. Sheltered coastal areas allow aquaculture development; however, at these sites, several antropogenic substances are discharged. Tributyltin (TBT) is one of the most toxic substances introduced into the marine environment. It is the main active component in organotin- based antifouling ship paints (Smith and Smith 1975, Zuekerman et al. 1978). Organotins are primary cause of oyster shell defor- mations (chambering and thickening), first reported in C. gigas at the Arcachon bay, France by Alzieu et al. ( 1980) and Alzieu et al. (1982). The same deformations were also observed by Waldock and Thain ( 1983) along the east coast of Britain. Organotin compounds were suspected as being responsible for anomalous shell thickening in the oysters from the Tejo estuary in Portugal (Andrea et al. 1983). Phelps (1993) studied TBT impact on C. angulata in some Portuguese estuaries, using the shell thick- ness index (STI) as indicator and noticed that oysters from the Sado estuary had high levels of TBT contamination. The present study describes C. gigas and O. edulis growing performances, biochemical composition, and STI at two estuaries on the northern coast of Portugal between June 1990 and October 1991. Because local variations in water quality can significantly affect the life cycle and reproduction of oysters in coastal areas (Heral et al. 1984, Heral et al. 1987, Brown and Hartwick 1988), water temperature, salinity, and total particulate matter were also measured. MATERIALS AND METHODS This experiment occurred at the Lima estuary and Ria of Aveiro in northern Portugal (Fig. 1 ). The Lima River has 2,480 km2 hydrological basin. Its estuary has substantial ship repairing industries. The experimental system located at Lima estuary was a 15-m longline, stainless-steel cable hanging near the surface. Floatation was achieved by plastic barrels tied along the cable, which was anchored by cement blocks. Ria of Aveiro is a seawater lagoon with a total area of 47 km2 and a freshwater inflow of 3 to 60 m3/s_1, caused by seasonal precipitation and runoff phenomena. The experimental system at Aveiro was a small raft (2 x 2 m) of galvanized tubing with two plastic floats. The raft, anchored by cement blocks, was placed near the sea entrance, at Ria de Aveiro's main channel (Fig. 1). At both stations, oysters were placed in pill-shaped baskets, as described elsewhere (Almeida et al. 1997). The experiment at Lima estuary was destroyed by a storm in September 1991. Seawater was sampled every 2 weeks at each site. Temperature, salinity, and particulate matter were measured as described else- where (Almeida et al. 1997). Water current was measured with a current flowmeter BFM001. Crassostrea gigas juveniles with a mean shell height of 20.1 ± 6.4 mm came from Marennes-Oleron (France) and were obtained from natural spatfall. Ostrea edulis juveniles with a mean of 33.1 139 140 Almeida et al. LIMA ESTUARY Figure 1. Location of the two stations: (1) Lima estuary; and (2) Ria of Aveiro. ± 5.2 mm length came from "Ostreira" a galician (Spain) hatchery. About 1.000 oysters, all from the same original stock, were placed at each site. Twice a month, all the baskets were agitated in the water to remove accumulated silt and feces. Every 45 days, a sample of 30 oysters were picked at random from each station for further analyses. In the laboratory, oysters were scrubbed under running tap water to remove encrusting organisms. Linear dimensions of in- dividuals were measured, and mean live weight was determined. Oysters were then opened and. tissues were excised. Wet meat weight was determined after superficial drying the extracted meat with absorbent paper. The flesh of 15 oysters were pooled and homogenized. Marine invertebrates in the field frequently have variable biochemical composition, and the use of pooled tissues from many individuals to determine the average composition may provide useful information (Giese 1967). Dry meat weight, ash, proteins, glycogen, and lipids were determined as described else- where (Almeida et al. 1997). Monthly instantaneous growth rate (G30) was calculated as G30 = ([log c(Z'+l/Z') / (log eD)] x 30. where Z1+1 is the mean shell height (cm) of the current month, Z' is the mean shell height of the previous month, and D is the number of days between observations (Ricker 1975). Condition index (CI) was calculated from the dry weights of meat and shell according to the formula CI = dry meat weight (mg)/dry shell weight (g) (Walne and Mann 1975). Thickness of the superior valve in C. gigos was measured with vernier callipers, after being cut at its longest axis with an esmeril disk adapted to a minidriller. Shell thickness index was determined according to the formula STI = L/T. where L is the length of the superior valve, and T is its thickness. Dyrynda (1992) compared the three most common methods to determine the STI and sug- gested that STI reflects abnormal thickening with greater accuracy and is easier to use. Analysis of variance (ANOVA) was performed using the STATISTICA 4.5 (Windows 95) statistical package. ANOVA as- sumptions were tested with Levene and Kruskal-Wallis tests. Par- ticulate inorganic material (PIM) and particulate organic material (POM) data were transformed (log x). Arcsine transformation was Growth and Biochemistry of C. gigas and O. edulis 141 carried out on biochemical data, which were compared as percent- ages. RESULTS Environmental Conditions Figure 2b shows seawater temperatures at both stations from May 1990 to October 1991. At station Lima, temperatures ranged from a minimum of 8°C in February to a maximum of 23°C in September of 1991. At station Aveiro, temperatures ranged from a minimum of 9.5°C in January of 1991 to a maximum of 2 1 ,5°C in May 1990. Mean temperatures at stations Lima and Aveiro during the experimental period were 15.1 ± 3.4 and 14.9 ± 3.2°C, respec- tively. Salinity values also showed minor differences between sta- tions (Fig. 2a). The mean salinity was 33. 1 ±3.6 and 33. 1 ± 3.9 ppt (n = 35) for Lima and Aveiro, respectively. The current flow varied between 0.12-0.40 m/s~] and 0.07-0.84 m/s"1 at Lima and Aveiro. respectively. POM values varied within 1.0-9.5 and 1.0-10.0 mg/r1 at sta- tions Lima and Aveiro. respectively (Fig. 2c). The mean POM content was 3.3 ± 1.8 and 3.2 ± 1.6 mg/r1 (n = 35) at Lima and Aveiro. PIM values varied within 2.0-51.0 and 4.3-70.3 mg/r1 at Lima and Aveiro. respectively (Fig. 2d). Extreme seasonal maxima were recorded in October of 1 99 1 at Lima and in May of 1 99 1 at Aveiro. The mean PIM content was 13.2 ± 8.9 and 14.6 ± 1 1.6 mg/r1 at Lima and Aveiro. Oyster Growth Crassostrea gigas Mean live weight (g) for C. gigas at both stations, from May 1990 to October 1991 is plotted in Figure 3a. At the end of the experimental period, oysters had a mean weight of 75 ± 16 and 83 ±21 g at stations Lima and Aveiro. Figure 3b presents the height instantaneous growth rate for both stations and shows some fluc- tuations over the study period. Growth rate was higher during the first 4 months of culture. From November to April, the growing rate was zero, after which, the growth showed a small recovery. Oysters at stations Lima and Aveiro reached a maximum dry tissue weight of 3.0 and 4.2 g. respectively, in July of 1991. After July and until the end of the experimental period, there was a decrease in dry tissue weight at both stations (Fig. 3c). Condition index values were very high in the summer of 1991 at station Aveiro, with a maximum of 85. At station Lima, the highest values were registered in June and July 1991. with a maximum of 103 (Fig. 3d). Seasonal variations on ash content are similar in both stations, with values ranging from 9.4 to 15.0%. Meat water content is significantly lower (p < .05) at station Lima than at station Aveiro (Fig. 3e). Mean percentages of meat water content are 80.0 ± 3.5 and 85.8 ± 4.0 at Lima and Aveiro. Meat protein, carbohydrate, and lipid, as a percentage of the ash-free dry weight (% AFDW), are shown in Fig. 3. from January to October 1991. Lipid values at station Lima were more or less constant throughout the year. At station Aveiro, lipid percentage decreased from January to May with a minimum value of 1.0%, recovering to 8.5% in July (Fig. 3f). Mean lipid content at station Aveiro was significantly lower (p < .05) than at station Lima. Carbohydrate content was also different throughout the summer, when comparing oysters from both stations. Although at station Lima, carbohydrate levels were high during the summer, with a maximum value of 28.8% in July; at station Aveiro, carbohydrate content was lower during the summer, with values around 6.4% (Fig. 3g). Protein content was more or less constant throughout the year, being significantly lower (p < .01) at Lima. Mean protein content was 62.3 ± 3.8 and 70.0 ± 3.4% at stations Lima and Aveiro, respectively (Fig. 3h). An STI of mean gravity (between 5 and 10) (Alzieu and Port- mann 1984) was recorded in 30% of the oysters sampled at both a 40 30 20 10 0 apr jul 1990 oct jan Lima — apr jul oct 1991 Aveiro apr Jul 1990 jul oct 1991 Figure 2. Seasonal variation in salinity: (a); maximum and minimum temperatures (b); particulate organic matter (POM) (c): and particulate inorganic matter (PIM) (d) at the two stations from May 1990 to October 1991. 142 Almeida et al. apr jul oct jan apr Jul oct 1990 1991 O Lima O Aveiro 8 oc 6 E 5 O t a. 3 i o D Or < u 40 30 20 10 - 90 80 70 60 50 40 dec feb apr jun aug oct dec 1991 dec feb apr jun aug oct dec 1991 Figure 4. Seasonal variation on O. edulis live weight (a); instantaneous growth rate in height (G10) (b); dry meat weight (c); condition index (d); meat water content (e); lipid as a percentage of the ash-free dry weight (AFDW) (f); carbohydrate (gl; and protein (hi reared at Lima estuary and Ria of Aveiro. Seasonal variations on ash content were similar in both stations, except in June, when ash values were 9.2 and 16.7% at stations Lima and Aveiro, respectively. These were, at the same time, the lowest and highest values registered at both stations. Meat water content was significantly lower (p < .01) at station Lima. Mean percentages of meat water content were 79.5 ± 2.4 and 89.6 ± 0.7 at Lima and Aveiro, respectively (Fig. 4e). Lipid values at station Lima decreased from January, with a mean of 7.3%. to the end of the experimental period, with 2.3%. At Aveiro. there was a de- crease in lipid content in the spring and the beginning of summer, with a recovery at the end of summer and autumn (Fig. 4f). Car- bohydrate values were very similar between stations, being more or less constant from January until June, with a maximum value of 37% in September (Fig. 4g). Protein seasonal variation was similar between stations, de- creasing from January to the end of the study. Mean protein con- tent was 60.2 ± 5.9 and 65.9 ±6.1% at stations Lima and Aveiro. respectively (Fig. 4h). 144 Almeida et al. DISCUSSION None of the environmental parameters evaluated showed dif- ferences between stations. Nevertheless, some differences were observed between the oysters at Lima and Aveiro at the biochemi- cal composition level. These differences were more accentuated in C. gigas than in O. edidis. One explanation for this could be the phytoplankton composition. The food value of different algal spe- cies has been evaluated for spat and juveniles of several bivalves (Enright et al. 1986. Laing and Millican 1986. Whyte et al. 1989, Whyte et al. 1990. Ferreiro et al. 1990). However, little is known about the effects of different food quality on growth of adult oys- ters. The relationship of bivalves to their food sources in the field is difficult to demonstrate, but seston quality and quantity have been related to growth and reproduction (Page and Hubbard 1987. Thompson and Nichols 1988, Utting 1988). Deslous-Paoli and Heral (1988) state that for C. gigas, stored energy, mainly as glycogen, is linked with the quality of food available during phy- toplanktonic blooms in spring and autumn. Bacteria content in the water could be another explanation for these differences. Oysters from the Lima estuary have a signifi- cantly higher bacterial content than oysters from Aveiro (unpub- lished data). This must be related to the lack of treatment of urban and industrial waters in this area. Bacteria are known to be one of the sources of bivalve diet (Brown et al. 1996). C. gigas at station Lima has higher lipid and lower protein content than oysters from Aveiro. C. gigas from Aveiro show a typical pattern of lipid and carbohydrate variation related to the reproductive cycle. This pattern, in which transformation of gly- cogen reserves into lipids intervened for the formation of gametes (Gabbott 1976, Lubet 1976), could be seen during the spring. Such a definite pattern was not observed in oysters from Lima. The variation in the meat water content was characterized by a decrease simultaneous with the build up of gametes and the accu- mulation of biochemical reserves. After the spawning period, meat water content began to increase. Curiously, meat water content was significantly lower in oysters from the Lima estuary, for both species. High water content is frequently associated with a low quality meat (Haven 1962. Shaw et al. 1967. Deslous-Paoli and Heral 1988), which is not the case. A minimum commercial weight of 60 g was attained in May and July 1991, more or less in 1 year of growth, in C. gigas from Aveiro and Lima, respectively. Dry meat weight values at station Lima are similar to those presented by Deslous-Paoli and Heral ( 1988) at Marennes-Oleron (France), concerning 2-year-old oysters. Data from Ruiz et al. ( 1992a) at Ria de Arosa (Galicia. Spain) indicate similar environ- mental conditions to those found in this study. These authors ob- served, in oysters of the same species, a first cycle of gonad growth in spring, with spawning in June to July. After this first spawning season, there was a rapid recovery during summer, with a second gametogenic cycle. They found oysters with completely developed gonad in mid-October. This second gametogenic cycle was asso- ciated with the greatest annual phytoplankton blooms, at the end of summer and the beginning of autumn. At Aveiro. we noticed that chlorophyll a levels are high throughout the summer (unpublished data). This does not happen at Lima estuary. The second gameto- genic cycle seems to reflect C. gigas' capacity to respond to the influence of seasonal phytoplankton blooms, as described by Lubet (1976). Data from Marennes-Oleron Bay (Atlantic French coast) indicate that oysters show only one dry weight maximum in July to August (Deslous-Paoli and Heral 1988). as observed in the oysters from Lima estuary. C. gigas growth, in the two locations described here, was sig- nificantly higher than the growth of oysters in earthen ponds (Almeida et al. 1997). Open sites allow the oysters to filter greater water volumes and, consequently, to ingest greater amounts of food. Shell thickness in C. gigas from both stations was probably caused by organotin compounds from antifouling paints used in boats. Both stations are located in channels with high maritime traffic, and near station Lima, there is also a ship-building yard. First indications of tributyltin effects in mollusks were from Pa- cific oysters, in the Atlantic French coast, near to recreational harbors (Alzieu et al. 1980). Studies on the effect of organotins in oysters indicated TBT as the agent responsible for shell thickening. This led to prohibition of the use of organotin compounds in boats smaller than 25 m in France and other European countries. C. gigas shell thickening is a readily and distinctly identifiable effect. which has been used with success in the biomonitoring of TBT in the UK. Ireland. France, the USA. Australia, and New Zealand (Alzieu et al. 1986. Stephenson et al. 1986, Wolniakowski et al. 1987, Batley et al. 1989. Ebdon et al. 1989, King et al. 1989, Minchin et al. 1996). Cortez et al. (1993) measured sediment con- lamination in butyltin compounds at Ria of Aveiro and found a degree of contamination considered to be a medium-high level at some sites. Butyltins were measured in the fine-grain fraction (<60 p.m). which can be ingested by filter-feeding organisms and is easily resuspended and transported away. Temperature and salinity levels, at both locations, prevalent during this study were compatible with O. edulis physiological requirements (Walne 1974. Mann 1979. Spencer 1988). Our growth rates on O. edulis were higher than those reported by Perez-Camacho and Beiras (1989) at Ria de Arosa and Cano and Rocamora ( 1996) at Mar Menor, in Spain, using spat with the same initial size. A few O. edulis specimens collected in the summer were car- rying larvae. The developmental zero value (the temperature below which no evidence of gonad development is found) is set around 7°C (Mann 1979. Wilson and Simons 1985). Both in Ria of Aveiro and the Lima estuary, the temperature never falls to a level low enough to interrupt gametogenesis. Similar observations were made by Ruiz et al. (1992b) for the Galician coast (Spain) and by Lubet (1976) for the Atlantic coast of France. Nevertheless, gly- cogen and lipid seasonal variation are not coincident with a ga- metogenic cycle. Abad et al. (1995) describe an increase in total lipid level throughout the gonad maturation period, in spring and early summer for adult O. edulis in the Galician coast. Oysters from Lima estuary show a decrease in lipid levels from the winter to the summer; whereas, at Aveiro, the highest lipid levels are in March and September through October. In terms of glycogen, the seasonal variation at both sites is similar, with an increase from April until September. Walne ( 1970) describe a decline in the glycogen content during the breed- ing season. The CI. observed throughout the experimental period for the European oyster, is considered low according to Walne (1974). Even in the summer months, the CI was below 70 in oysters from Lima estuary, those showing a higher CI (an average quality is indicated by an index >80). One possible explanation for the low performance of both populations can be the effect of TBT in the digestive gland. Axiak et al. ( 1995) exposed O. edulis to low levels of TBT (10 ng/T1) and observed a significant digestive cell atro- phy. In terms of shell morphology, we found no alterations in O. Growth and Biochemistry of C. gigas and O. eduus 145 edulis. No appreciable shell thickening was ever reported in the European oyster. The only constraint to oyster culture at these locations was shell thickening of C. gigas. If a commercial oyster farming business is to be developed at these sites, a more detailed study about the effect of TBT compounds is needed because this poses serious problems for such endeavors. The survey of Portuguese coastal environments demonstrates the occurrence of different degrees of butyltin contamination (Quevauviller et al. 1989, Cortez et al. 1993). The most contami- nated sites are in enclosed bays or estuaries with high TBT inputs from harbors and shipyards. The effect of high levels of contami- nation on molluscan species should encourage government action in Portugal to prohibit use of TBT-containing paints, as has been adopted in many other European countries. ACKNOWLEDGMENTS We thank Carlos Manuel, Joao da Guia, Bladmiro Couteiro. and FORPESCAS from Viana do Castelo for the help provided in the mantainance of the experimental structures. This work was supported by a JNICT grant (Junta Nacional de Investigacao Ci- entifica e Tecnologica). REFERENCES Abad, M. C. Ruiz. D. Martinez. G. Mosquera & J. L. Sanchez. 1995. Seasonal variations of lipid classes and fatty acids in flat oyster, Oslrea edulis, from San Cibran (Galicia, Spain). Comp. Biochem. Physiol. HOC: 109-1 18. Almeida, M. J., J. Machado & J. Coimbra. 1997. Growth and biochemical composition of Crassostrea gigas (Thunberg) at three fishfarm earthen ponds. J. Shell fish Res. 16:455-462. Alzieu. C. M. Heral. Y. Thibaud, M.J. Dardignac & M. Feuillet. 1982. Influence des peintures anti-salissures a base d'organostanniques sur la calcification de la coquille de l'huitre Crassostrea gigas. Rev. Trav. Inst. Peek. Maul. 45:101-116. Alzieu. C. & J. E. Portmann. 1984. The effect of tnbutyltin on the culture of C. gigas and other species. Proceedings of the 50th annual shellfish conference. 15-16 May 1984. pp. 87-104. Alzieu, C. J. Sanjuan. J. P. Deltreil & M. Borel. 1986. Tin contamination in Arcachon Bay: effects on oyster shell anomalies. Mar. Polhtt. Bull. 17:494-498. Alzieu. C. Y. Thibaud. M. Heral & B. Boutier. 1980. Evaluation des risques dus a l'emploi des peintures antisalissures dans les zones con- chylicoles. Rev. Trav. Inst. Peches Marit. 44:301-348. Andreae. M. O.. J. T. Byrd & P. N. Froelich. 1983. Arsenic, antimony, germanium, and tin in the Tejo Estuary. Portugal: modelling a polluted estuary. Environ. Sei. Technol. 17:731-737. Axiak, V., M. Sammut. P. Chircop, A. Vella & B. Mintoff. 1995. Labo- ratory and field investigations on the effects of organotin (tributyltin) on the oyster, Ostrea edulis. Sci. Total Envir. 171:1 17-120. Bardach. J. E., J. H. Ryther & W. O. McLamey. 1972. Aquaculture: the farming and husbandry of freshwater and marine organisms. Wiley Interscience. New York. 886 pp. Batley. G. E., C. Fuhua, C. E. Brocklebank & K. J. Flegg. 1989. Accumu- lation of tributyltin by the Sydney rock oyster Saccostrea commercia- lis. Aus. J. Mar. Freshwater Res. 40:49-54. Brown, M. R.. S. M. Barret. J. K. Volkman. S. P. Nearhos. J. A. Nell & G. L. Allen. 1996. Biochemical composition of new yeasts and bacteria evaluated as food for bivalve aquaculture. Aquaculture 143:341-360. Brown, J. R. & E. B. Hartwick. 1988. Influences of temperature, salinity, and available food upon suspended culture of the Pacific oyster, Cras- sostrea gigas. I. absolute and allometric growth. Aquaculture 70:231- 251. Cano, L. & J. Rocamora. 1996. Growth of the European flat oyster in the Mediterranean Sea (Murcia, SE Spain). Aquacult. Int. 4:67-84. Cortez, L„ P. Quevauviller. F. Martin & O. F. X. Donard. 1993. Survey of butyltin contamination in Portuguese coastal environments. Environ. Pollut. 82:57-62. Deslous-Paoli, J. M. & M. Heral. 1988. Biochemical composition and en- ergy value of Crassostrea gigas (Thunberg) cultured in the bay of Marennes-Oleron. Aquat. Living Resour. 1, 239-249. Dyrynda, E. A. 1992. Incidence of abnormal shell thickening in the Pacific oyster Crassostrea gigas in Poole Harbour (UK), subsequent to the 1987 TBT restrictions. Mar. Pollut. Bull. 24:156-163. Ebdon. L.. K. Evans & S. Hill. 1989. The accumulation of organotins in adult and seed oysters from selected estuaries prior to the introduction of U.K. regulations governing the use of tributyltin based antifouling paints. Sci. Total Environ. 83:63-84. Enright. C. T.. G F. Newkirk. J. S. Craigie & J. D. Castell. 1986. Evalu- ation of phytoplankton as diets for juvenile Ostrea edulis L. J. Exp. Mar. Biol. Eeol. 96:1-13. Ferreiro. M. J.. A. Perez-Camacho, U. Labarta. R. Beiras. M. Planas & M. J. Fernandez-Reiriz. 1990. Changes in the biochemical composition of Ostrea edulis larvae fed on different food regimes. Mar. Biol. 106: 395—40 1 . Gabbott. P. A. 1976. Energy metabolism, pp. 121-206. In: B. L., Bayne (ed.). Marine Mussels: Their Ecology and Physiology. Cambridge Uni- versity Press, Cambridge, UK. Gaffney, P. M. & S. K. Allen. 1993. Hybridization among Crassostrea species: a review. Aquaculture 1 16:1-13. Giese, A. C. 1967. Some methods for study of the biochemical constitution of marine invertebrates. Oceanogr. Mar. Biol. Ann. Rev. 5:159-186. Giese, A. C. 1969. A new approach to the biochemical composition of the mollusk body. Oceanogr. Mar. Biol. A. Rev. 7:175-229. Haven. D. 1962. Seasonal cycle of condition index of oysters in the York and Rappahannock Rivers. Proc. Nat. Shellfish. Assoc. 51:42-66. Heral, M., J. M. Deslous-Paoli, J. Prou & D. Razet. 1987. Relations entre la nourriture disponible et al production de mollusques en milieu es- tuarien: variabilite temporelle de la colonne d'eau. Haliotis 16:149- 158. Heral, M., J. M. Deslous-Paoli. D. Razet & J. Prou. 1984. Essai de mise en evidence in situ de parametres biotiques et abiotiques de l'eau et de 1' interface eau-sediment intervenant dans la production de l'huitre Crassostrea gigas. Oceanis 10:465-475. King, N., M. Miller & S. De Mora. 1989. TBT levels of seawater, sedi- ment, and selected marine species in coastal Northland and Auckland. New Zealand. New Zealand J. Mar. Freshwater Res. 23:287-294. Laing, I. & P. F. Millican. 1986. Relative growth and growth efficiency of Ostrea edulis L. spat fed various algal diets. Aquaculture 54:245-262. Lubet. P. 1976. Ecophysiologie de la reproduction chez les mollusques lamellihranches. Haliotis 7:49-55. Mann, R. 1979. Some biochemical and physiological aspects of growth and gametogenesis in Crassostrea gigas and Ostrea edulis grown at sus- tained elevated temperatures. J. Mar. Biol. Ass. UK 59:95-1 10. Minchin. D.. E. Stroben, J. Oehlmann, B. Bauer, C. B. Duggan & M. Keatinge. 1996. Biological indicators used to map organotin contami- nation in Cork Harbour, Ireland. Mar. Pollut. Bull. 32:188-195. Page, H. M. & D. M. Hubbard. 1987. Temporal and spatial patterns of growth in mussels Mytilus edulis on an offshore platform: relationships to water temperature and food availability. J. Exp. Mar. Biol. Ecol. 111:159-179. Perez-Camacho. A. & M. J. Beiras. 1989. Experiencias de cultivo de ostra Ostrea edulis L. en batea en el Rio Piedras (SO de Espafia) y en la Ria de Arousa (NO de Espafia). Acuicultura Intermareal Instituto de Cien- cias Marinas. Andalucia. Espafia. pp. 93-101. Phelps. H. L. 1993. Tributyltin and Portuguese estuaries. Congress on ec- otoxicology and environmental chemistry — a global perspective. Lis- boa. Portugal. 28-31 March, pp. 13. 146 Almeida et al. Quayle. D. B. 1969. Pacific oyster culture in British Columbia. Bull. Fish. Res. Board Can. 192 pp. Quevauviller. P.. R. Lavigne, R. Pinel & M. Astruc. 1989. Organotins in sediments and mussels from the Sado estuarine system (Portugal). En- viron. Pollut. 57:149-166. Ricker, W. E. 1975. Computation and interpretation of biological statistics offish populations. Bull. Fish. Res. Board Can. 191: 382 pp. Ruiz, C, M. Abad. F. Sedano. L. O. Garcia-Martin & J. L. Sanchez-Lopez. 1992a. Influence of seasonal environmental changes on the gamete production and biochemical composition of Crassostrea gigas (Thurn- berg) in suspended culture in El Grove. Galicia, Spain. J. Exp. Mar. Biol. Ecol. 155:249-262. Ruiz. C, D. Martinez. G. Mosquera. M. Abad & J. L. Sanchez. 1992b. Seasonal variations in condition, reproductive activity, and biochemical composition of the flat oyster, Ostrea edulis, from San Cibran (Galicia, Spain). Mar. Biol. 112:67-74. Shaw. W. N„ H. S. Tubiash & A. M. Barker. 1967. Freeze drying for determining total solids in shellfish. J. Fish. Res. Board Can. 24:1413- 1417. Shpigel. M. & R. A. Blaylock. 1 99 1 . The Pacific oyster. Crassostrea gigas. as a biological filter for a marine fish aquaculture pond. Aquaculture 92:187-197. Smith P. & L. Smith. 1975. Organotin compounds and applications. Chem. in Britain 11:208-212. Stephenson, M. D.. D. R. Smith. J. Goetzl. G. Ichikawa & M. Martin. 1986. Growth abnormalities in mussels and oysters from areas with high levels of tributylin in San Diego Bay. pp. 1246-1251. In: Proceedings of Oceans'86 Conference, vol. 4. Organotin Symposium, Washington. DC. Spencer. B. E. 1988. Growth and filtration of juvenile oysters in experi- mental outdoor pumped upwelling systems. Aquaculture 75:139-158. Thompson, J. K. & F. H. Nichols. 1988. Food availability controls seasonal cycle of growth in Macoma baltica (L.) in San Francisco Bay. Cali- fornia. J. Exp. Mar. Biol. Ecol. 116:43-61. Utting. S. D. 1988. The growth and survival of hatchery-reared Ostrea edulis L. spat in relation to environmental conditions at the on-growing sites. Aquaculture 69:27-38. Vilela. H. 1975. A respeito de ostras. biologia-exploracao-salubndade. Notas e Estudos. Secretaria de Estado das Pescas. Waldock. M. J. & J. E. Thain. 1983. Shell thickening in Crassostrea gigas: organotin antifouling or sediment induced. Mar. Pollut. Bull. 14:41 1- 415. Walne, P. R. 1970. The seasonal variation of meat and glycogen content of seven populations of oysters Ostrea edulis L. and a review of the literature. Fishery Investigations London, Series II, Number 3. Walne. P. R. 1974. Culture of bivalve mollusks: fifty years experience at Conwy. Whitefriars Press Ltd.. London. Walne. P. R. & R. Mann. 1975. Growth and biochemical composition in Ostrea edulis and Crassostrea gigas. pp. 587-607. In: H. Barnes (ed.). Proceedings of the 9th European marine biology symposium. H. Ab- erdeen University Press. Aberdeen. Scotland. Whyte. J. N. C. N. Bourne & C. A. Hodgoson. 1989. Influence of algal diets on biochemical composition and energy reserves in Patinopecten yessoensis (Jay) larvae. Aquaculture 78:333-347. Whyte, J. N. C. N. Bourne & C. A. Hodgoson. 1990. Nutritional condition of rock scallop. Crassadoma gigantea (Gray), larvae fed mixed algal diets. Aquaculture 86:25-40. Wilson, J. H. & J. Simmons. 1985. Gametogenesis and breeding of Ostrea edulis on the West Coast of Ireland. Aquaculture 46:307-321. Wolniakowski, K. U.. M. D. Stephenson & G. Ichikawa. 1987. Tributyltin concentrations and Pacific oysters deformations in Coos Bay, Oregon, pp. 1438-1442. In: Proceedings of the Oceans "87 Conference, vol. 4. Organotin Symposium, Halifax, Canada. Zuckerman. J. J.. R. P. Reisdorf. H. V. Ellis & R R. Wilkinson. 1978. Organometals and organometalloids: occurrence and fate in the envi- ronment. American Chemical Society, Washington. DC. pp. 388—124. Journal of Shellfish Research. Vol. 18. No. 1, 147-153, 1949. FOUNDER EFFECT, GENETIC VARIABILITY, AND WEIGHT IN THE CULTIVATED PORTUGUESE OYSTER CRASSOSTREA ANGVLATA LAUREANA REBORDINOS,1 PEDRO GARCIA,2 AND JESUS M. CANTORAL1 Laboratorio de Genetica y Microbiologic., CA.S.E.M.. Universidad de Cadiz. Cadiz. Spain 'Area de Genetica, Universidad de Leon, 24071 Leon, Spain ABSTRACT Existence of genetic variability is a prerequisite for successful implementation of breeding programs, and clarification of the relationships in such programs to quantitative traits is of great economic interest. We have studied the relationship between multilocus heterozygosity and/or allozyme genotypes and weight in the Portuguese oyster Crassostrea angulata (Lamark). Two cohorts were obtained in a commercial hatchery by mass-spawning from wild oysters. Loss of genetic variability was shown in cultured oysters as compared with the wild population because of a founder effect caused by a low effective population size. Significant effects on growth rate were detected for the Me-2. Xdh, Lap. Pgm. and Est loci. However, these effects were not retained in the two cohorts, nor in the two ages of the same cohort, nor were differentiated effects detected in weight classes of the same age. At the same time, differences between genotypes were not associated with differences between heterozygous and homozygous genotypes. Positive correlations between multilocus heterozygosities and growth rate, as well as significant differences between mean body weights for different degrees of heterozygosity, were found only in the largest weight class. Moreover, significant results were obtained when the mean weight of different heterozygosity classes for total weight, body weight, and shell weight were compared only in the oysters selected for their larger size. This result points to the isozymes as markers for quantitative traits and confirms the existence of heterosis in C. angulata. indicating the possibility of establishing breeding programs based on the maintenance of inbred lines and crossing them to obtain hybrid vigor. KEY WORDS: oysters, genetic variability, weight, Crassostrea angulata, isozymes INTRODUCTION The commercially exploited Crassostrea angulata, also known as Portuguese oyster, is an oyster species naturally found on the coast of the southwest Iberian Peninsula, mainly in river mouths (Michinina and Rebordinos 1997). As natural populations became exhausted, the industry was forced to turn toward hatchery rearing of oysters. Nevertheless, some important aspects of shellfish hatchery practices must be considered, because they interact with the genetics of cultured species. In this sense, although progenitors from natural beds had been the source of seed for this culture, low effective population sizes caused the loss of rare alleles and a reduction in heterozygosity. Very often, oysters grown for hatch- ery seed are used as broodstock to produce the next generation, resulting in closed and small populations becoming propagated, which show the effects of inbreeding depression. In some cases, development of these populations is merely the result of general husbandry procedures. However, in other cases, rearing of popu- lations from a reduced number of progenitors is intended to achieve some kind of genetic improvement in important traits (growth rate, healthy appearance, hard shells, etc.) without knowl- edge or evaluation of the validity of these programs that lack genetic controls. The existence of genetic variability is a prerequisite for the successful application of breeding programs, and its estimation is easy for qualitative traits. In contrast, for quantitative traits, such an estimation is not trivial. However, quantitative traits are often more important for selective breeding. Therefore, it is important to determine the genetic variability of traits and clarify whether a relationship between traits can be established. In this way, the genetic component of quantitative traits could be inferred from such markers as isozymes (Hedgecock et al. 1995). Relationships between heterozygosities and/or allozyme geno- types and such quantitative traits as viability, growth rate, have been described for marine mollusks (Mitton and Grant 1984, Zouros and Pogson 1994). Nevertheless, there is no clear correla- tion between the degree of heterozygosity and characteristics re- lated to fitness. In fact, results on the relationships with the growth and viability show that the studies reporting positive correlations only just outnumber those in which no correlations were found (Britten 1996). Some of these difficulties have been attributed to the use of unsuitable markers, because most of these studies were carried out using isozymes, but it is not completely established that allozymes (or at least some of them) are completely neutral mark- ers. Two main theories attempt to explain the positive correlations between the degree of heterozygosity and fitness. First is the over- dominance hypothesis, which considers the enzyme variants as responsible for the correlation effect (Koehn et al. 1988. Sarver et al. 1992); and second is the associative overdominance hypothesis, which considers the scored loci as markers for those involved in the effect (Zouros and Mallet 1989). Elucidating these two alternatives is important from the evolu- tionary point of view, but it also has clear relevance to breeding genetics. Our aims in this work are to compare the genetic vari- ability within a wild population and some cultivated populations reared from the wild population, as well as to determine differen- tial contribution to growth rate (determined as weight reached at a given time) in cultivated oysters (Crassostrea angulata). Genetic 147 148 Rebordinos et al. data in natural populations of the Portuguese oyster are scarce (Michinina and Rebordinos 1997). and. as far as we know, there are no published reports relating quantitative traits to isozymic markers. MATERIALS AND METHODS Populations Sampled Hatchery-propagated oyster stocks were maintained in the commercial hatchery AMALTHEA S.A. (Cadiz, Spain). A group of 20 adult oysters (selected for their large size and healthy ap- pearance) were taken from a natural bed (wild population) located on the Atlantic coast of Cadiz on May 1993 and kept in a small flowthrough seawater system. Before spawning, some oysters were opened to check on gonadal maturation. Spawning was in- duced by raising the temperature from 20 to 28-30°C. Eggs were collected and placed in seawater at 25°C and then were counted and fertilized by adding several milliliters of dense sperm suspen- sion; samples of the sperm-egg suspension were immediately ex- amined microscopically, and. if necessary, more sperm were added to achieve a bound-sperm:egg ratio of about 10 (Dupuy 1974). Larval cultures were kept at a density of 10 larvae/mL for 10 days and then changed to two larvae/mL. After metamorphosis, oysters were transferred to tanks supplied with a flow of seawater. Two cohorts were produced from the same progenitors. One was studied at 1 8 months old ( 1 8 m ) and the other at both 8 and 1 2 months old (8 m and 12 m, respectively). During 4 months, oysters of the second cohort were subjected to overcrowded conditions to obtain harder shells, which are more valuable from the commercial point of view. The analysis at 12 months was performed on two different groups established by the hatchery, one of which was the group selected because of the large specimen size (12 mL). for additional growout. and the other contained the smallest samples (12 ms). which we chose to be discarded. Once in the laboratory, oysters were numbered and weighed individually (total weight), then the shells were opened, the animal was removed, and weighed (body weight), and the shells were also weighed (shell weight). Before weighing, the samples were placed on filter paper to release most of the water contained within the shells. A sample of 62 individuals from the natural bed was also collected, and upon arrival in the laboratory, the animals were removed from their shells and tissue samples (hepatopancreas) from wild and cultivated (after being weighed) oysters were frozen in liquid nitrogen and stored at -70°C until analysis. Electrophoresis and Genetic Interpretations The allozyme genotype of each oyster was obtained for eight polymorphic genes: Est (esterase; E.C. 3.1.1.2.). Lap (leucine ami- nopeptidase; E.C. 3.4.11.1). Mdh-I and Mdh-2 (malate dehydro- genase; E.C. 1.1.1.37). Me-2 (malic enzyme; E.C. 1.1.1.40). Pgi (phosphoglucoisomerase: E.C. 5.3.1.9), Pgm (phosphoglucomu- tase; E.C. 5.4.2.2). and Xdh (xanthine dehydrogenase; E.C. 1.1.1.204). Electrophoretic techniques employed were those de- scribed in Michinina and Rebordinos (1947). Activity zones were named by abbreviations followed by a number denoting from slower to faster migration. Loci were des- ignated by italicized abbreviations with numerical suffixes corre- sponding to isozymes in order of ascending electrophoretic mo- bility . The most frequent allele for each locus was designated 100. "^■Expected ' "etExpecli-J' and alternative alleles were labeled according to their rates of migration relative to that of the most common allele. Null alleles (identified by absence of any enzymatic activity) were referred to by the number 00 (Michinina and Rebordinos 1997). Data Analysis Allele frequencies were obtained by direct count of the pres- ence or absence of bands from electrophoretic phenotypes, and numerical analyses were performed using the program Biosys-1 (Swofford and Selander 1981). Genotype frequencies were tested for Hardy-Weinberg equilibrium, and the degree of heterozygote deficiency relative to the Hardy-Weinberg expectations was re- corded as the D index = Het.observed where Het. is the number of heterozygotes. An estimation of the effective size (Ne) of the founder popu- lation was obtained from the reduction of expected heterozygosity with respect to the parental population, by the equation H, = H„( l-l/2Ne)' (Crow and Kimura 1970). where H, is the heterozy- gosity at time 1. 1 is the number of generations, and Hn is the initial heterozygosity. In the case described here, t = 1 H, is the average heterozygosity in the cultivated population studied, and H(> is the average heterozygosity in the natural population from which the broodstock was collected. Growth rate is one of the most commercially important traits, and it can be estimated as different weights reached by individuals of same age. grown in the same conditions. The relationship with each locus was tested in two different ways; first the mean weight of each genotype, and second the mean weight of homozygotes and heterozygotes were compared using one and two-way analysis of variance (ANOVA). A basic assumption underlying the analysis of variance is the equality of variances, and this assumption was checked using Lev- ene*s test. When the test resulted in a significant value, we per- formed the Welch and Brown-Forsythe procedures, which are two alternate tests of the null hypothesis of equality of group means in which the group variances are not assumed to be equal. Most statistical tests were performed with the BMDP program package (Dixon et al. 1990). RESULTS Genetic Structure of Populations Frequencies of allozyme alleles in each population are shown in Table 1. and heterozygosity estimates and measures of Hardy- Weinberg deviations are shown in Table 2. Some alleles present in the natural population were not present in reared populations. Thus, Mdhl and Mdhl showed two alleles in the natural popula- tion but became monomorphic in the oldest cultivated populations. Alleles Lap1? and Pg/?i138 from the wild population were absent in most of the cultivated ones. Nevertheless, a number of new alleles that were absent in the wild population did appear in cultivated ones. For example. Lap'22 was present in 8 m, 12 ms, and 12 mL; Lap00 appeared in 8 m. and Pgi'*7 appeared only in 8 m, probably because of a sampling error. Although contamination is very un- likely, because no spawning or larval settlement occurs in winter in natural populations of Crassostrea angulata, the hatchery did not install collector controls to check larval settlement, and, for this reason, external origin cannot be completely discounted for these new alleles. The average allele number (Table 1) ranged from 3.14 ± 0.40 in the 8 m population to 2.43 ± 0.53 in 18 m, Genetic Variability and Weight in C. angulata 149 TABLE 1. Allele frequencies in two cohorts of C. angulata, one analyzed both at 8 (S ml and 12 months old (12 nil and the other at IS months old, and in the source population {wild I. At 12 months old, two weight classes of oysters were studied: the smallest ( 12 ms) and the largest ( 12 ml. I. TABLE 2. Expected and observed averages heterozygosities (h„ and hj and deviations from Hardy-Weinberg proportions (1)1 in two cohorts of C. angulata, one analyzed both at 8 (8 ml and 12 months old (12 m) and the other at 18 months old, and in the source population (wild). At 12 months old two weight classes of oysters were studied: the smallest (12 ms) and the largest (12 mL). Locus Allele Wild 62 8 m 51 12 ms 69 12 mL 5(1 18 m 49 (n) Locus Wild 8 111 12 ms 12 mL 18 m Lap Lap 73 04 23 0.010 0.000 0.000 0.000 K 0.452 0.377 0.484 0.363 0.058 91 (1711 0.173 0.255 0.210 0.030 h„ 0.333 0.250 0.451 0.366 0.020 100 04 67 0.769 0.667 0.770 0.970 D -0.263"*** -0.081 -0.069 0.008 -0.656* 122 0.000 0.029 0.078 0.020 0.000 Mdh-1 00" 0.000 0.019 0.000 0.000 0.000 \. 0.065 0.056 0.028 11111 mn Mdh-I h,, 0.068 0.058 0.028 85 0.034 0.029 0.014 0.000 0.000 D 0.035 0.030 0.014 100 0.966 0.971 0.986 1.000 1.000 Mdli-2 Mdh-2 he 0.017 0.019 mn 11111 mn 100 0.992 0.990 1.000 1.000 1. 000 h„ 0.017 0.019 118 0.008 0.010 0.000 0.000 0.000 D 0.009 0.010 Me-2 Me-2 90 0.108 0.029 0.359 0.190 0.020 K 0.183 0.224 0.573 0.567 0.096 ]()() 0.650 0.875 0.535 0.590 0.950 h„ 0.507 0.212 0.479 0.680 0.100 110 0.242 0.096 0.106 0.220 0.030 D 0.639*** -0.033 -0.165°** 0.198* 0.040 Pgi Pgi 11 0.086 0.183 0.021 0.030 0.260 \. 0.158 0.359 0.041 0.058 0.385 100 0.914 0.779 0.979 0.970 0.740 h„ 0.172 0.442 0.042 0.060 0.520 157 0.000 0.038 0.000 0.000 0.000 D 0.094 0.284 0.022 0.03 1 0.351 Pgm Pgm 94 0.057 0.000 0.042 0.040 0.030 he 0.755 0.548 0.642 0.452 0.695 96 0.131 0.019 0.028 0.020 0.140 h„ 0.787 0.558 0.535 0.480 1.000 100 0.230 0.615 0.542 0.720 0.460 D 0.043 0.045 -0.167 0.061 0.438*** 114 0.189 0.240 0.197 0.070 0.230 Xdh 132 0.369 0.125 0.141 0.150 0.140 he 0.572 0.659 0.619 0.657 0.623 138 0.025 0.000 0.049 0.000 0.000 h„ 0.042 0.654 0.423 0.620 0.420 Xdh D —0 927*** -0.008 -0.318** -0.056'** -0.326J** 99 0.438 0.298 0.303 0.260 0.220 Est 100 0.479 0.404 0.500 0.400 0.500 K nd 0.557 0.739 0.699 0.676 101 0.083 0.298 0.197 0.340 0.280 h„ 0.565 0.594 0.600 0.580 Est D 0.363 -0.196"** -0.141 -0.142a** 94 nd 0.083 0.210 0.090 0.120 Mean value 100 0.625 0.319 0.380 0.450 D -0.045 0.076 -0.091 0.004 -0.037 107 0.181 0.283 0.330 0.300 SD 0.154 0.149 0.115 0.093 0.328 103 0.1 11 0.188 0.200 0. 1 30 He 0.361 0.323 0.341 0.300 0.265 P95 7 1 ,43 7 1 .43 57.14 57.14 57.14 SD 0.107 0.091 0.114 0.105 0.113 na 3.00 3.14 2.86 2.57 2.43 H„ 0.229 0.357 0.319 0.347 0.330 (SE) (0.53) <0.14) (0.59) (0.53) (0.53) SD 0.102 0.091 0.089 0.103 0.128 J Null allele, identified by absence of any enzymatic activity; n: sample size; nd: not determined; P95: percentage of loci polymorphic at 95%; na: average number of alleles per locus; SE: standard error. H0 and He are the averages of h0 and hc over loci:a not significant after pooling; mn: monomorphic loci; SD: standard deviation. * p < .05; **p < .01; ***p < .001. although these differences were not statistically significant (t = 1.07; df = 12; p > .05). The proportion of polymorphic loci (P,,5) was highest in the wild and 8 m populations (71.43%) and least in the other three cultivated populations (P = 57.14). Nevertheless, the major differences between populations were observed when allelic frequencies between populations were considered, and the contingency \2 analysis at all loci was highly significant (p < .001). except in the cases of the Mdhl and Mdh2 loci, where this analysis was not significant. Nine of the equilibrium tests carried out to determine devia- tions from Hardy-Weinberg proportions (D) were significant for cultivated oysters, mainly at 12 and 18 months. Considering that only one of the tests would be expected to be statistically signifi- cant by chance, there is a considerable deviation from equilibrium conditions. Significant x2 tests could be caused by low individual numbers in some phenotypic classes. Therefore, a further \2 test was calculated after pooling alleles in three classes: homozygotes for the most common allele, frequent-infrequent heterozygotes and a third class composed of infrequent homozygous and other heterozygous genotypes. Loci Est, Me-2, and Pgm fit the equilib- rium conditions better when pooled classes are used, meaning that after pooling, three out of nine significant tests became nonsignif- 150 Rebordinos et al. icant. Hence, these cases could be explained by a high number of alleles present in low frequencies, although all scored alleles were present at high frequencies (Table 1 ). as compared to the low frequencies described in natural populations of mollusks in general (Sarver et al. 1992) and C. angulata in particular (Michinina and Rebordinos 1997). On the other hand, only in two cases the lack of equilibrium was attributable to an excess of heterozygotes, the rest being at- tributable to an excess of homozygous genotypes. The natural population from which the progenitors were taken showed signifi- cant departures from Hardy-Weinberg proportions at three loci. One of the loci became nonsignificant after pooling of alleles (Lap) and another one yielded a homozygous excess (Me-2: D = 0.639). As far as D values in every locus are concerned, they were negative in all cases in both cultivated and wild populations in the cases of Lap and Xdh loci. The Lap locus showed null homozy- gotes in the 8 m population, although they were not detected in the other populations or in the wild population, probably because of the low number of parental oysters used as progenitors. Mean D values (± standard deviation) were positive in 8 m and 12 mL (D = 0.076 ± 0.149 and 0.004 ± 0.093. respectively); however, the 18 m population yielded a negative value (D = -0.037 ± 0.328), similar to that of the wild population (D = -0.045 ± 0.154). and 12 ms gave the highest negative D value (D = -0.091 ± 0.115). The average expected heterozygosity (He) was higher in 1 2 ms (0.39 1 ±0.111) than in wild population (0.361 ± 0.107), and also higher than in the other three populations (Table 2). However, when average observed heterozygosities (Ho) were compared, the lowest value was given by the wild population (0.229 ± 0.102) followed by the small weight class of 12 months (0.319 ± 0.089); the populations of 8 months and large weight class of 12 months and 18 months showed slightly higher values (Table 2). If we consider the over-all results on the occurrence of hetero- zygote excess at Pgi and Pgm, it can be seen that there is an absence in the cultivated populations of some alleles present in the natural population, marked differences in allele frequencies be- tween the wild and the cultivated populations, and a lower value of the average heterozygosity (He) in the cultivated populations as compared with wild ones. All these facts point to a founder effect on hatchery populations. The effective size of the founder popu- lation estimated by the reduction in average heterozygosity was two. This means a number of actual contributors to the progeny that is one order of magnitude lower than the total number of oysters set to spawn, and a similar situation has been described for other species (Hedgecock and Sly 1990). Average genetic distances and genetic identities, calculated over the seven loci studied on the five populations (Table 3). gave a maximum divergence value of D = 0.991 in the comparison of TABLE 3. Unbiased genetic identities above diagonal and unbiased genetic distances below diagonal (Nei 1978) between a wild population and four cultivated ones reared from it of Crassostrea angulata. TABLE 4. Descriptive analysis of weight in grams in two cohorts of C. angulata, one analyzed both at 8 (8 m) and 12 months old (12 m) and the other at 18 months old. At 12 months old, two weight classes of oysters were studied: the smallest (12 ms) and the largest (12 mL). Population Wild 8m 12 ms 12 mL 18 in Wild 0.890 0.919 0.892 0.849 8 m 0.116 0.967 0.981 0.991 12 ms 0.085 0.033 0.989 0.945 1 2 mL 0.115 0.020 0.011 0.955 18 m 0.163 0.009 0.057 0.046 Populations/Traits N M SD SE cv 8 m Total weighl 51 8.400 3.033 0.425 0.361 Body weight 51 2.108 0.934 0.131 0.443 Shell weight 51 6.293 2.178 0.305 0.346 12 ms Total weight 69 5.905 1 .505 0.181 0.255 Body weight 69 0.812 0.265 0.032 0.326 Shell weight 69 5.093 1.295 0.156 0.254 12 mL Total weight 49 11.676 3.174 0.453 0.272 Body weight 49 1.467 0.527 0.075 0.359 Shell weight 50 10.236 2.737 0.387 0.267 18 m Total weight 49 49.558 7.769 1.110 0.157 Body weight 49 8.566 1.715 0.245 0.200 Shell weight 49 40.992 6.658 0.951 0.162 N: number of individuals; M: mean value; SD: standard deviation; SE: standard error: CV: coefficient of variation. 8 and 18 m. and a maximal identity value of I = 0.163 between the wild population and 18 m. Variability and Weight Descriptive statistical data regarding the distribution of total weight, body weight, and shell weights of the samples are given in Table 4. Shapiro-Wilks tests using the mean and deviation values from the samples were used to assess whether the distribution of data was normal. Oyster data of 12 m were found not to be normal at any of the weight traits. In contrast, for the rest of the ages, the total weight and shell weight were found to be distributed nor- mally, and, in addition, body weight at 8 months was also normal. Logarithmic transformation of data did not improve the adjustment to normality, and so this transformation was not used. Twelve-months-old oysters came from the same cohort as those of 8 months, but were sampled 4 months later. These individuals (12 m) were subjected to overcrowded conditions to obtain harder shells, which became more resistant and. hence, healthier-looking in the long term. This is also the reason why 12-month-old oysters yielded a lower body weight than 8-month-old specimens. How- ever, the largest class showed the highest mean value for both total and shell weight. The mean total weight at 8 m was 8.40 g, and 4 months later, the value was 11.68 g. Most of this increase was because of the increase in shell weight that resulted from the overcrowded conditions, because the mean body weight after this time period was lower. The mean body weight of 1 8-month-old oysters was four times that of the 8 m specimens. The coefficient of variation in the four groups studied was higher for the body weight variable, and it always decreased with age. with the value being 0.443 for body weight for 8 month and 0.200 for 1 8-month- old oysters. The effect of the genotypes at individual loci on growth rate was assessed by one-way ANOVA on total weights of individuals Genetic Variability and Weight in C. angulata 151 TABLE 5. Analysis of variance for total weight within genotypes (Genotypes) and comparing total weight of homozygotes versus heterozygotes (Horn. vs. Het.) for eight allo/.ymic loci in four cultivated populations of C.angulata with degrees of freedom (df), F-value (F), and probability of fitting to null hypothesis (pi. 'Significant under Welch's and Brown-Forsythe's tests; nd: not determined; na: not applicable. 8 m 12 ms 12 mL 18 m Locus F P df F P df F P df F P df Lap Genotypes 0.71 .620 5 0.23 .918 4 2.17 .105'' 3 na Horn. vs. Het. 0.77 .471 1 0.31 .582 1 1.27 .266 1 Mdh-1 Genotypes 0.01 .927 1 0.17 .685 1 na na Horn. vs. Het. 0.01 .927 1 0.20 .658 1 Me-2 Genotypes 2.92 .045* 3 0.13 .970 1 1.17 .337 4 3.08 .056 4 Horn. vs. Het. 2.43 .125 1 0.09 .772 1 0.62 .434 1 5.74 .021* 1 Pgi Genotypes 0.95 .392 2 1.26 .266 1 0.03 .853 1 0.01 .927 1 Horn. vs. Het. 0.02 .885 1 1.26 .266 1 0.03 .853 1 0.01 .911 1 Genotypes 2.13 .080a 5 0.82 .619 1 0.63 .732 7 0.63 .674 7 Horn. vs. Het. nd nd nd nd Xdh Genotypes 0.65 .662 5 0.64 .669 5 4.79 .003** 4 0.30 .909 4 Horn. vs. Het. 0.00 .997 1 0.40 .530 1 3.26 .077 1 0.01 .910 1 Est Genotypes 0.82 .548 5 1.20 .311" 9 1.13 .365 8 1.37 .238 8 Horn. vs. Het. 0.07 .797 1 2.12 .150 1 1.18 .283 1 1.29 .261 1 for the four studied populations (Table 5). Interactions were sig- nificant at loci Me-2 at 8 m and Xdh at 12 mL. All variances were tested by a Levene's test, which gave nonsignificant results with the exception of locus Pgm at 8 m; in this case, the Welch test was also significant. At the same time. Lap at 12 ml gave rise to a Welch test with values of F = 18.61, resulting in p = .001 and Brown-Forsythe test F = 3.16 and p = .071. In addition, for for, the Welch test at 12 ms gave values F = 6.52 and p = .001. Further ANOVA analysis was conducted considering weights of heterozygotes versus homozygotes at each locus (Table 5). A significant value was detected at 18m for locus Me-2 and for locus Xdh in 12 mL p value was close to the significance level. In both cases, the mean body weight of heterozygous genotypes was higher than in homozygous genotypes. To study the relationship between multilocus heterozygosity and growth rate, the correlation of individual heterozygosity (the number of loci for which one oyster was heterozygous) and indi- TABLE 6. Correlation coefficients between individual heterozygosity and weight (total weight, body weight, and shell weight) in two cohorts of C. angulata, one analyzed both at 8 (8 m) and 12 months old (12 m), and another at 18 months old. At 12 months old, two weight classes of oysters were studied: the smallest (12 ms) and largest (12 mL). Traits 8 in 12 ms 12 mL 18 m Total weight Body weight Shell weight •0.001 -0.160 0.010 -0. 1 89 ■0.006 -0.146 0.208 0.271 0.180 -0.116 -0.172 -0.098 vidual weight was computed (Table 6). None of the correlations was statistically significant, although two different trends can be deduced from Table 5. The two populations that were not selected (i.e., 8 m and 18 m) and the one selected for small specimen size (12 ms) showed values close to zero, indicating no correlation or a negative one. However, the population subjected to size selection (12 mL) yielded positive values and, for total weight and body weight, were close to being significant (r = 0.208 and r = 0.271, respectively), because the significant level at p < .05 started at r = 0.285. Moreover, the existence of a relationship between multi- locus heterozygosity and weight was examined by grouping individuals into classes according to the number of heterozygous loci in each individual and calculating the variance between weights of classes with different number of heterozygous loci. ANOVA values were not significant in any case, although in the 12 mL class, low p values were obtained and were close to sig- nificance in the case of body weight. When mean weights of each heterozygous class were compared by a Student's Mest, statisti- cally significant (p < .05) results were found in 12 mL in the following cases: ( 1 ) The mean total weight of heterozygotes for two loci was 10.188 g and, for the class with three loci 12.960 g. being p = .018. (2) The mean body weight for the class with two loci heterozy- gotes was 1.155 g. For the class with three, the value was 1.642 g and for four 1.768 g. The p value between classes two and three was .009 and for classes two and four it was .004. (3) The mean shell weight for the group exhibiting two het- erozygous loci was 9.214 g and for that exhibiting three such loci, it was 11.318 g. The p value was .036. 152 Rebordinos et al. DISCUSSION Growth rate is a very economically important trait, and it can be estimated as differences in weights reached by individuals from the same cohort grown under uniform conditions. Its relationship. if any, to gene markers would be very useful in any breeding program, but its estimation is not trivial because of the character- istics of the quantitative traits and also because methods and ex- perimental designs used in research laboratories are likely to be different from procedures used in commercial hatcheries. This means that results cannot always be easily extrapolated. Hence, we decided to address this problem by checking, from a genetic point of view, oysters that had been classified as valuable for a hatchery and looking for relationships between allozymic genotypes and differential weights in individual oysters. Loss of genetic diversity following hatchery culture of organ- isms has been widely reported and is caused by the effective popu- lation size of hatchery "mass spawning." Several factors contribute to this result, such as the sex ratio of progenitors, the mating behavior, physiology, and genetic organization (Beaumont 1994). Our results show a decrease in genetic variability of both the average allele number, and at the polymorphism level, and these are likely to be because of the small number of oysters (20) used as broodstock. Gosling ( 1982) estimated that at least 45 individuals (sex ratio of about 1:1) are needed to provide a 99% chance of retaining alleles at a frequency of 0.01 at a locus. Moreover, the estimated effective population size was estimated to be one order of magnitude lower than the number of progenitors, or. more likely, as low as two. Discrepancies between apparent and effec- tive population size could be attributable to an insufficient number of progenitors used in the spawns that produced the cultivated populations and/or unequal spawning success among the chosen progenitors. The main practical meaning of this result is that hatch- ery managers should control the number of spawnings from the total number of oysters induced to spawn. As far as the effects of loci on growth rates are concerned, we found only two (five, if we consider the significant values rendered by the Brown-Forsythe test) out of 28 cases producing significant effects on applying Welch's test. These were Me-2 on 8 m (p < .05) and Xdh (p < .01 ) on 12 niL and Lap. Pgm. and Est (p < .05) on 12 niL, S m and 12 ms, respectively. To accept that a real dependence of the growth rate on some isolated isozymic genes exists, we would expect a generalized effect on related species, populations of the same species, or at least individuals of the same cohort at different ages. However, none of these circumstances have been reported. First of all. we cannot compare our results to other results obtained for C. angulata species because of lack of relevant data. However, results published on C. virginica (Gmelin) species describe variable correlations between weight class and loci (Zouros and Pogson 1994. Foltz and Chatry 1986). Although methods described in the aforementioned papers were slightly dif- ferent from ours, there should not be much difference, because all data are referred to specific enzymatic activities. Published data indicated that some biochemical characteristics in relation to qua- ternary structure and number of genes controlling each activity are conserved between taxa located phylogenetically far away. For this reason, it would be expected that extended relationships between specific genes and traits that are important in hatcheries would also have been conserved (Richardson et al. 1986). At the same time, such effects were not retained when weights of heterozygotes and homozygotes were compared at those loci affecting growth rate. Thus, this effect was only shown for Me-2 at 18 m (p < .05), and in 12 mL. the gene Xdh. reported as highly significant when comparing genotypes, was nearly significant when the weight of homozygous individuals was compared to the larger one of heterozygous individuals (p = .077). Moreover, if a consistent effect is caused by some specific loci, this effect should have been retained with age (at least) in indi- viduals originating from the same cohort (8 m and 12 m). and/or differential effects should have been found between the lowest and highest weight classes of the same age ( 12 ms. 12 mL). However, these conditions did not occur, and no differences were observed between the two weight classes. Another interesting characteristic of genes contributing to quantitative traits in marine bivalves is the deficit of heterozygotes. although there is no definitive explanation for this observation (Gaffney et al. 1990). The deficit of heterozygotes has been ex- plained by the presence of null alleles, but also the relationship between heterozygote deficiencies and heterozygosity-fitness cor- relations has been attributed to selection against heterozygotes during the larval phase and overdominance at allozyme loci at the adult stage (Zouros and Pogson 1994). A different explanation considers the allozyme markers as neutral, but related to fitness, and this effect is seen after inbreeding and is also responsible for heterozygote deficiencies. Relationships between allozyme loci and overdominant or deleterious alleles could be established either by gametic disequilibrium caused by a small effective population size, or by partial inbreeding, resulting in a generalized homozy- gosity producing inbreeding depression (Zouros 1993. David et al. 1995). It has been claimed that the use of real neutral markers could help to answer these questions. In fact, microsatellites can provide indicators of selection processes at linked loci attributable to link- age disequilibrium (Slaktin 1995). By studying microsatellite in different larval stages, Bierne et al. ( 1998) found that these mark- ers cosegregated with fitness-associated genes and significant mul- tilocus heterozygosity-growth correlations were recorded at all stages, ruling out the hypothesis of differential selection between larval and juvenile or adult stages. When relationships between heterozygosities and weight were analyzed, positive correlation and nearly statistical significances were detected only for 12 mL. showing an important difference between all of them and pointing to the expression of the selective effect acting upon selected individuals (more heterozygous). This result was corroborated by the analysis of variance carried out within different classes of heterozygosity in C. angulata. which rendered lower p values and even a significant value (p < .001) under the Welch's test for body weight on this class. In addition, when a Student's /-test was performed, significant results were only obtained when comparing the weight of different heterozy- gosity classes for total weight, body weight and shell weight only in 1 2 mL. When the mean observed heterozygosities were com- pared between populations, a higher value was detected in the largest weight class (0.347 ± 0.103) as compared to the smallest one (0.319 ± 0.089), as expected. However, the value shown at 8 months was higher (0.357 ± 0.091 ): the decrease was probably the response to the selection caused by the stressful situation of oysters living in overcrowded conditions over 4 months. Finally, the dif- ference between the average heterozygosity values of wild popu- lation (0.229 ± 0.102) and the considerably higher values found in all cultivated populations could be explained by the use of selected progenitors (by large size and healthy appearance) from the wild Genetic Variability and Weight in C. anguiata 153 population (quantitative traits were not determined in wild popu- lation, because the age of these oysters was unknown). Differences between populations are even more remarkable when expected average heterozygosities are also considered (Table 2). All these results would reinforce the role of isozymes as markers and, most importantly, as valid markers for quantitative traits. Nevertheless, in the case of allozymes. it might also occur that both effects of linkage disequilibrium and overdominance would be operating within the genome, although it is very difficult to provide direct experimental evidence to estimate how much of the effect might be attributable to overdominance. or dominance at the scored loci, and how much could be attributable to other linked genetic conditions. Our results point to the associative hypothesis and the role of isozyme acting as markers rather than playing a direct role. It is likely that this means that they can be used as markers for quantitative traits and, more importantly, confirm the existence of heterosis in Crassostrea ssp. This phenomenon could be very valuable for breeding in aquaculture and also for estab- lishing breeding programs in oysters based on the maintenance of inbred lines of oysters and crossing them to obtain hybrid F, showing hybrid vigor. ACKNOWLEDGMENTS We thank I. Moreno for technical assistance. L. Saenz de Miera for help with the computed programs and S. A. Amalthea. (Cadiz- Spain) for providing us with the cultivated oysters. This study was supported by the Project UCA GRPRE-94/03 of the University of Cadiz and the CV-219 from the Junta of Andaluci'a. LITERATURE CITED Beaumont. A. R. 1994. Genetics and aquaculture. pp. 467—186. In: A. R. Beaumont (ed.). Genetics and Evolution of Aquatic Organisms. Chap- man & Hall. London. Bierne, N.. Launey, S.. Naciri-Graven. Y. & Bonhomme. F. 1998. Early effect of inbreedings as revealed by microsatellite analyses on Ostrea edulis larvae. Genetics 148:1893-1906. Britten. H. B. 1996. Meta-analyses of the association between multilocus heterozygosity and fitness. Evolution 50:2158-2164. Crow. J. F. & Kimura. M. 1970. An introduction to population genetics theory. Harper & Row, New York. David, P.. Delay, B.. Berthou P. & Jarne. P. 1995. Alternative models for allozyme-associated heterosis in the marine bivalve Spisula oralis. Ge- netics 139:1719-1726. Dixon. W. J.. Brown. M. B.. Engelman. L. & Jennrich, R. I. 1990. BMDP statistical software manual, vol. 1. University of California Press, Berkeley. California. Dupuy. J. L. 1974. Translation of mariculture research into a commercial oyster seed hatchery, pp. 677-685. In: Proceedings of the 9th Annual Conference of the Marine Technology Society. Foltz. D. W. & Chatry, M. 1986. Genetic heterozygosity and growth rale in Louisiana oysters ( Crassostrea virginica). Aquaculture 57:261-269. Gaffney, P. M.. Scott, T. M., Koehn. R. K. & Diehl, W. J. 1990. Interre- lationships of heterozygosity, growth rate, and heterozygote deficien- cies in the coot clam. Mulinia lateralis. Genetics 124:687-699. Gosling. E. M. 1982. Genetic variability in hatchery-produced Pacific oys- ter (Crassostrea gigans Thunberg). Aquaculture 26:273-287. Hedgecock. D. & Sly. F. 1990. Genetic drift and effective population sizes of hatchery-propagated stocks of the Pacific oyster, Crassostrea gigas. Aquaculture 88:21-38. Hedgecock. D.. McGoldnck, D. J. & Bayne. B. L. 1995. Hybrid vigor in Pacific oysters: an experimental approach using crosses among inbred lines. Aquaculture 137:285-298. Koehn, R. K.. Diehl, W. J. & Scott. T. M. 1988. The differential contribu- tion by individual enzymes of glycolysis and protein catabolism to the relationship between heterozygosity and growth rate in the coot clam, Mulinia lateralis. Genetics 118:121-130. Michinina. S. R. & Rebordinos. L. 1997. Genetic differentiation in marine and estuarine natural populations of Crassostrea anguiata. Mar. Ecol. Prog. Ser. 154:167-174, Mitton. J. B. & Grant. M. C. 1984. Associations among protein heterozy- gosity, growth rate, and developmental homeostasis. Ann. Rev. Ecol. Syst. 15:479^199. Nei. M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 98:583-590. Richardson, B. J.. Baverstock, P. R. & Adams, M. 1986. Allozyme elec- trophoresis: a handbook for animal systematics and population struc- ture. Academic Press. Sydney. Sarver. S. K, Katoh, M. & Foltz, D. W. 1992. Apparent overdominance of enzyme specific activity in two marine bivalves. Genetica 85:231-239. Slatkin. M. 1995. Hitchhiking and associative overdominance at a micro- satellite locus. Mol. Biol. Evol. 12:473-180. Swofford, D. L. & Selander. R B. 1981. BIOSYS-1: a FORTRAN pro- gram for the comprehensive analysis of electrophoretic data in popu- lation genetics and systematics. J. Hered. 72:281-283. Zouros, E. 1993. Associative overdominance: evaluating the effects of inbreeding and linkage desequilibrium. Genetica 89:35^16. Zouros, E. & Pogson, G. H. 1994. The present status of the relationship between heterozygosity and heterosis, pp. 135-147. In: A. R. Beau- mont (ed.). Genetics and Evolution of Aquatic Organisms. Chapman & Hall, London. Zouros. E. & Mallet, A. L. 1989. Genetic explanations of the growth/ heterozygosity correlation in marine mollusks. pp. 317-324. In: J. S. Ryland and P. A. Tyler (ed.). Reproduction. Genetics, and Distributions of Marine Organisms. Proceedings of the 23rd European Marine Biol- ogy Symposium. Olsen and Olsen, Fredensborg, Denmark. Zouros. E. & Pogson. G. H. 1994. The present status of the relationship between heterozygosity and heterosis, pp. 135-146. In: A. R. Beau- mont (ed.). Genetics and Evolution of Aquatic Organisms. Chapman & Hall. London. Journal of Shellfish Research, Vol. IS. No. I, [55-158, 1999. GROWTH OF PTERIA COLYMBUS (RODING, 1798) IN SUSPENDED CULTURE IN GOLFO DE CARIACO, VENEZUELA CESAR J. LODEIROS,1 JOSE JESUS RENGEL,1 AND JOHN H. HIMMELMAN2 lDepartamento de Biologia Pesquera, Institute Oceanogrdfico de Venezuela, Universidad de Oriente, Cunumd 6101. Venezuela zDepartement de Biologie et G1ROQ. Universite Laval, Quebec, Canada G1K 7P4 ABSTRACT Over a 10-month period, we examined the growth of shell height and width, and of mass of the shell and tissues, of juveniles of Pteria colymbus (initially measuring 13.5 mm in shell height) placed in a suspended culture at 8-m depth at Turpialito in the Golfo de Cariaco, Venezuela. A high growth rate was observed for all body parameters. Shell parameters increased rapidly during the first 5 months and then at lower rates until the end of the study, and total tissue mass increased at a rapid and almost steady rate throughout the study. Mortality during the study was negligible. The maximum size predicted by the von Bertalanffy growth equation (L, =71 mm) was near the maximal size observed in natural populations. P. colymbus seemed little affected by marked changes in environmental conditions during our study, because it showed continuous rapid growth rate. The rapid growth, low mortality, and availability of natural spat indicate that P. colymbus could be an excellent species for aquaculture in this region. INTRODUCTION Although pearls are cultured in only a few species, all bivalve species secreting large amounts of iridescent nacreous material have the potential for use in pearl culture. The family Pteridae includes a number of such species, Pinctada margaritifera, Pinctada maxima, Pinctada mazathantica, and Pteria sterna, which have been exploited in tropical areas (Coeroli et al. 1984, Alagarswami et al. 1989, Gaytan-Mondragon et al. 1993, Monte- forte and Garcia-Gasca 1994, Shirai 1994). In northeastern Ven- ezuela, beginning with the Spanish colonization, considerable ex- ploitation of the mother-of-pearl oyster Pinctada imbricata has been made on natural banks, particularly near Margarita Island ("the island of pearls'") and Verginelli and Prieto (1991) report its growth in a natural population. Another Caribbean bivalve pro- ducing a thick nacreous layer is the winged pearl oyster Pteria colymbus (Roding. 1798). which presently is being considered as a potential species for aquaculture in Venezuela and Colombia (Borrero 1994, Marquez 1996. Velasco and Borrero 1996). P. colymbus occurs along the coasts of North Carolina. Florida, and Texas, in Bermuda and the West Indies, and also as far south as Brazil (Abbot 1974. Diaz and Puyana 1994). In the Golfo de Cariaco in northeastern Venezuela, it is common in shallower waters attached to soft corals and suspended structures, such as the collectors and pearl nets used for bivalve culture. The present study examines the growth and survival of P. colymbus in sus- pended culture over a 10-month period. METHODS The study was conducted at Turpialito in the Golfo de Cariaco, northeastern Venezuela (Fig. 1) and ran from 29 December 1993 to 29 October 1994. We began by placing 300 juveniles of P. colymbus, measuring 13.5 mm in mean shell height (95% CI = 2.09), in 6-mm mesh pearl nets suspended at 8-m depth from a longline. The juveniles originated from a natural settlement that had occurred on pearl nets in which the scallop, Euvola ziczac, was being cultured. At periodic intervals throughout the study (dates given in Table 1 ), we collected all oysters to determine the pro- portion of living individuals and, at the same time, randomly se- lected 10 individuals for determinations of shell height (the maxi- mum distance between the dorsal hinge and ventral margin), shell width (maximal distance through the animal from the outermost portion of the left and right valves), and the dry mass of the shell and tissues (drying at 60°C for =2 days). At each sampling date, the remaining oysters were returned to the sea in new pearl nets. As the oysters increased in size, we decreased their density in the experimental pearl nets (Table 1; scallops not needed to complete 40° 30°- 20° 10° ■ Gulf of , Mexico \)~» -™. Caribbean Sea * *^^N/>t£Venezuelas 10°35' Golfo de Cariaco Turpialito 10' 64°00' 50' 40' Figure 1. Location of study site at Turpialito in Golfo de Cariaco, northeastern Venezuela. 155 156 LODEIROS ET AL. TABLE I. Mortality rates as documented on different sampling dates during our study of the growth of Pleria colymbus in suspended culture in Golfo de Cariaco. Set Out on Pre' /ious Date Percentage Mortality/ No. Density Total No. Number Sampling Date Nets /Net Oysters Dead Day 29 December 1993 (start of experiment) 1 February 1994 3 100 300 3 0.010 28 February 1994 7 40 280 2 0.007 26 March 1994 6 40 240 0 0.000 26 April 1994 5 40 200 0 0.000 17 May 1994 4 40 160 0 0.000 23 June 1994 10 15 150 0 0.000 23 August 1994 9 15 135 0 0.000 29 October 1994 8 15 120 0 0.000 the desired density in the pearl nets were discarded), so that the oysters covered about 30—40% of the floor of the nets, as recom- mended for scallop culture by Ventilla (1982). During the study, temperature was continuously recorded at 8-m depth in the experimental site using a SEALOG electronic thermograph (Vemco Ltd.. Halifax. Nova Scotia. Canada), and other environmental factors were determined at weekly intervals during the first 7 months. (We were unable to take seston samples during the last 3 mo.) Water samples were collected using 5-L Niskin bottles and transferred to opaque containers for transport to the laboratory, where the seston was analyzed. Samples (500-1000 mL) were filtered through precombusted 0.7-u.m MFSF filters, which were then frozen at -20°C for later analyses. The masses of the organic and inorganic sestonic fractions (suspended particles) were determined gravimetrically. and chlorophyll a concentrations were obtained using the spectrophotometry' method (Strickland and Parsons 1972). The general pattern of increase in shell height over time (in 70 60- 50- 40 30 20- 10- Shell height (mm) 0 18 15 12 9 6- 3- A SO 0 1994 Shell mass (g) c D I ] " F 'm'a'm' J ' J 'A'S 'O' 1994 30 1 25 20 15 101 5 Shell width (mm) 0 dM'f'm'a'm1 p j'a's'o' 2.5 i 2.0 1.5 1.0 0.5-1 0 Dl J ' F 'M'A'M' J ' J 'A'S 'O' 93 1994 Von Bertalanffy curve 0 2 4 6 8 Time (mo) Figure 2. Changes in shell height (A) and width (B) and in the mass of shell (C) and tissues (D) between 29 December 1993 and 29 October 1994 and the calculated von Bertalanffy growth curve (E), for juveniles of Pteria colymbus maintained in suspension culture at 8-m depth at Turpialito in Golfo de Cariaco, Venezuela. Vertical bars represent standard deviations. Pthria colymbus Growth in Suspended Culture 157 months) were described using the von Bertalanffy growth function and the K growth constant, and asymptotic size, L^, was deter- mined using the Gulland and Holt plot procedure (Sparre and Ursin 1989). RESULTS Both shell and tissue growth were rapid, but they showed dif- ferent patterns (Fig. 2). Growth in shell height and shell mass were greatest at the beginning of the study (9.1 mm mo"1 and 1.8 g mo-1, respectively, for the period 29 December 1993 to 17 May 1994), but decreased in the latter part of the study (by 81% for shell height and 48% for shell mass). The rate of increase in shell width was similar to the increases in shell height and mass, except that the decrease in the latter months was less marked. In contrast, the rate of increase in tissue mass was almost constant throughout the study, about 0.18 g mo"1. The mortality rate was extremely low. there being only 3 of 300 oysters dead on the first sampling date, 2 of 280 on the second date, and none dead on all subsequent dates (Table 1 ). The general increase in shell height over time was well described by a von Bertalanffy growth curve, with a K growth rate of 0.28 and an asymptotic size (L„) of 71 mm (Fig. 2). The thermograph record showed a slow temperature decrease from December 1993 (25°C) to early March 1994 (21.5°C) and then increased in April (to 24.3°C) and early July (to 25.0°C). The most marked temperature change during the study was in late August, when temperature increased from 22.5 to 27.0°C, and the level further increased to 28.5°C in late October (Fig. 3A). The high temperature levels during the last 3 months seemed to have no effect on the growth of P. colymbus, because the increase in tissue mass continued at the same rate throughout September and Octo- ber, and the slowing in shell growth had begun in June; thus, several months before the abrupt temperature increase in August. Slowing in shell growth was likely part of the normal growth curve. High temperatures also did not affect survival, because no mortality occurred after February. From December to July, we documented fluctuating, but gen- erally high, levels of total seston, its organic content, and chloro- phyll <7 (Fig. 3B, C. D). Chlorophyll a values likely dropped during the period of elevated temperatures in late August through October (when we did not collect seston samples), because studies during previous years demonstrated that primary productivity falls during the thermal stratification (caused by reduced upwelling) that de- velops in these months (Lodeiros and Himmelman 1994, Lodeiros 1996). Regular continuation of growth curves through September and October suggested that growth of P. colymbus did not decrease in response to decreased phytoplanktonic food availability. Fur- thermore, mortality was null during this period. DISCUSSION Our study suggests that the growth and survival of P. colymbus is little affected by the variations in external environmental factors in the Golfo de Cariaco or by experimental manipulations. P. colymbus showed negligible mortality and rapid growth of all body parameters throughout the 10-month study period. The study pe- riod included the full range of environmental conditions typical of the gulf (Lodeiros 1996), from strong wind-driven upwelling and associated low temperatures and high primary productivity (Janu- ary to March) to a well-developed thermal stratification and low productivity (September to October). The asymptotic size pre- dicted by the von Bertalanffy curve (Fig. 2E, 71 mm) was similar 18- C Organic seston (mg L"1) 12- ilk A A 1 6 - \ ~h^N n - 1 1 \l V V ^\1 -i 1— — i 1 1 1 1 1 r D Chlorophyll a ([ig V1) M J 1994 Figure 3. Variations in temperature (A), total seston (B), the organic fraction of the seston (C) and in chlorophyll a concentration |D( at the culture site at 8-m depth at Turpialito in the Golfo de Cariaco, Ven- ezuela. to the maximum size P. colymbus attains on a natural bank in the northern Golfo de Cariaco and on suspended structures (=70-75 mm. pers. obs.). Furthermore, this suggests that the oysters were not stressed during our study. P. colymbus seems better adapted to culture conditions in the Golfo de Cariaco than other bivalves. For example, concurrent growth trials with the scallop Lyropecten (No- dipecten) nodosus at the same study site showed decreases in growth and increases in mortality associated with periods of high temperatures, red tide blooms, and abundant fouling (Lodeiros et al. 1998). A complete mortality of L. nodosus occurred during August, simultaneous with a sharp rise in temperature and a 158 LODEIROS ET AL. bloom of the toxic diatom Gymnodinium catenation. In earlier studies at Turpialito. the scallop Euvola ziczac was found to be highly susceptible to high temperatures and fouling (Lodeiros and Himmelman 1994. Lodeiros and Himmelman 1996, Lodeiros 1996). Fouling organisms colonized the shells of P. colymbus but seemed to have little effect on growth and survival. Throughout our study. P. colymbus showed aggregating behav- ior; that is. individuals byssally attached to one another, as is characteristic of mussels. This suggests that P. colymbus is adapted to living at high densities and, furthermore, that mussel culture techniques may be suitable for this species. Spat of P. colymbus (5-15 mm shell length) are readily collected on spat collectors throughout the year (Marquez 1996). Velasco and Borrero (1996) studied P. colrmbus in the Colombian Caribbean and observed rapid growth in suspended culture but high mortality (51% by the end of a 5-mo study) and low spat settlement. In our study, growth rates were higher than in the latter study, and mortality was neg- ligible. The rapid growth, low mortality, and high availability of spat indicates P. colymbus could be an excellent species for aqua- culture in the Golfo de Cariaco. Studies are required to optimize culture techniques (types of enclosures, stocking density, culture depth, and potential for using mussel culture techniques) and to determine whether the time of spat deployment affects growth rates and survival. Furthermore, studies are needed to develop techniques for inducing pearl formation. Our observations suggest that the tissues have a favorable texture and taste, but further studies are required to determine the marketability of the tissue for human or animal consumption. ACKNOWLEDGMENTS This study was made possible by provision of facilities by the Instituto Oceanografico de Venezuela. Universidad de Oriente and was supported by grants from the Consejo Nacional de Investiga- ciones Cientificas de Venezuela (CONICIT) to C.J. L. and the National Sciences and Engineering Research Council (NSERC) of Canada to J. H. H. We are indebted to M. Nunez for his help with the field work. LITERATURE CITED Abbott. R. T. 1974. American seashells. 2nd ed. Van Nostrand Reinhold. Ltd.. New York. 663 pp. Alagarswami. K., S. Dharmaraj. A. Chellam & T. S. Velayudhan. 1989. Larval and juvenile rearing of black-lip pearl oyster. Pinctada marga- ritifera (Linnaeus). Aquaculture 76:43-56. Borrero. F.J. 1994. Potential of pearl oyster culture on the Colombian Caribbean. (Proceedings of Pearls '94. International Pearl Conference. Honolulu. Hawaii. 14-19 May 1994.)/ Shellfish Res. 13:331-332. Coeroli. M.. D. De Gaillande & J. P. Landret. 1984. Recent innovations in cultivation of mollusks in French Polynesia. Aquaculture 39:45-67. Diaz, J. M. & M. Puyana. 1994. Moluscos del Caribe Colombiano. Un catalogo ilustrado. COLCIENCIAS. Fundacion Natura and INVE- MAR. Bogota Colombia. 367 pp. Gaytan-Mondragon. I.. C. Caceres-Martinez & M. Tobias-Sanchez. 1993. Growth of the pearls oysters Pinctada mazatlanlca and Plena sterna in different culture structures at La Paz Bay. Baja California Sur. Mexico. J. Aquacult. Soc. 24:341-546. Lodeiros. C. J. 1996. Influence des facteurs environnementaux sur la crois- sance du petoncle tropical Euvola (Pecten) ziczac (L. 1758) cultive en suspension au Golfo de Cariaco, Venezuela. Ph.D. thesis. Department of Biology, Laval University, Quebec, Canada. 83 pp. Lodeiros, C. J. & J. H. Himmelman. 1944. Relations among environmental conditions and growth in the tropical scallop Euvola {Pecten) ziczac (L.) in suspended culture in the Golfo de Cariaco. Venezuela. Aqua- culture 119:345-358. Lodeiros. C.J. & J. H. Himmelman. 1996. Influence of fouling on the growth and survival of the tropical scallop. Euvola iPecten) ziczac (L. 1758) in suspended culture. Aquaculture Res. 27:749-756. Lodeiros. C. J.. J. Rengel. L. Freites & J. H. Himmelman. 1998. Growth and survival of the tropical scallop Lyropecten (Nodtpecten) nodosus maintained in suspended culture at three depths. Aquaculture 165:41- 50. Marquez. B. 1996. Variacidn estacional de la fljacidn de la ostra negra Pteria colymbus (Roding. 1798) (Bivalvia: pteriidae) a diferentes pro- fundidades en la localidad de Turpialito. Golfo de Cariaco. Edo. Sucre. Venezuela. Licenciatura thesis. Biology Department, Universidad de Oriente. Cumana. Venezuela. 69 pp. Monteforte. M. & A. Garcia-Gasca. 1994. Spat collection studies on pearl oysters Pinctada mazatlanlca and Peteria sterna (Bivalvia. pteridae) in Bahia de La Paz. South Baja California. Mexico. Hydrobiologia 291: 21-34. Slnrai. I. 1994. Pearls and pearl oysters of the world. Marine Planing Co.. Ltd.. Japan. 109 pp. Sparre. P. & E. Ursin. 1989. Introduction to tropical fish stock assessment. Part I — manual. FAO Fisheries Tech. Paper. 306/1. Strickland. J. D. H. & T. R. Parsons. 1972. A practical handbook of sea- water analysis. Bull. Fish. Res. Board. Can.. No. 167. 2nd ed. 310 pp. Velasco. L. A. & F.J. Borrero. 1996. Cultivo experimental de la ostra perlffera alada Pteria colymbus (Bivalvia: pteridae) en cultivo suspen- dido. pp. 224-231. In: A. Silva and G. Merino (eds.). Communica- ciones Breves del IX Congreso Latinoamericano de Acuacultura y II Simposio de Avances y Perspectivas de la Acuacultura en Chile. Uni- versidad Catolica del Norte y Asociacion Latinoamericana de Acua- cultura. Coquimbo, Chile. Ventilla. R. F. 1982. The scallop industry in Japan. Adv. Mar. Biol. 20: 310-382. Verginelli. R. & A. Prieto. 1991. Produccion secundaria de Pinctada un- bricata (Roding. 1798) (Pterioida: pteriidae) en una poblacidn del Golfo de Cariaco. Venezuela. Act. Clem. Venez. 42:138-144. Journal of Shellfish Research, Vol. 18, No. 1. 159-167, 1999. GROWOUT OF BLACKLIP PEARL OYSTERS, P1NCTADA MARGARITIFERA COLLECTED AS WILD SPAT IN THE SOLOMON ISLANDS KIM J. FRIEDMAN1 AND PAUL C. SOUTHGATE1 lAquaculture Department, James Cook University, Townsville, Queensland 4811, Australia 2ICLARM Coastal Aquaculture Centre. Honiara. Solomon Islands ABSTRACT This study assessed growth and survival of juvenile blacklip pearl oysters (Pinctada margaritifera) in a number of intermediate culture systems: lantern nets, panel nets, perforated plastic trays, and attached to ropes enclosed by mesh. Juveniles with initial dorsoventral measurements of 8.3 to 51.5-mm increased in size by 20.4 to 24.8-mm in 3 months, and 30.7 to 36.5-mm in 5 months Growth rates of juvenile P. margaritifera cultured in the open reef systems of the Solomon Islands compared favorably with those reported from the established pearl culture operations in French Polynesia and the Cook Islands Initial experiments showed that survival of oysters in lantern nets in shallow reef areas was poor as a result of predation by fish and invertebrates. Siting of culture systems in deeper water decreased mortality by fish, although predation by invertebrates that recruited from plankton was still a potential problem. In seneral. there were no significant differences in growth or survival between juveniles held in lantern nets and panel nets; however, lantern nets were more difficult to clean and inspect for predators. Juvenile growth and survival did not differ significantly (p > .05) between panel nets and trays after 5 months, although the rigid trays were easier to clean of fouling organisms. Juveniles placed loosely into trays tended to aggregate, and rates of growth and survival of oysters glued separately into trays were significantly °reater (p <.05) than those for oysters placed loosely into trays. There was no significant difference in growth between oysters glued into trays and those glued onto ropes and enclosed behind plastic mesh. Overall, this study shows that important criteria of the growout units needed for the intermediate culture of P. margaritifera in the Western Pacific include ease of cleaning and access for regular inspection and removal of predators. KEY WORDS: Pearl oyster, spat. Pinctada margaritifera, intermediate culture, growth, survival, predation INTRODUCTION Protocols developed by the Japanese for collection and culture of the pearl oyster, Pinctada fiicata martensii, have been adapted in French Polynesia for blacklip pearl oysters. P. margaritifera (Coeroli and Mizuno 1985). Since adapting and developing their own spat collection and growout techniques and stimulating mar- ket demand for cultured "black" pearls, there has been rapid growth in round pearl production in Polynesia. For example, an- nual value from the export of loose "black" pearls from French Polynesia presently stands in excess of US$145m (Fassler 1995, Remoissenet 1996. Doubilet 1997. N. Sims pers comm. 1998). Oysters used for the production of black pearls are collected as spat from substrates (spat collectors) deployed in the surface wa- ters of "closed" and semiclosed atoll lagoons in Polynesia (Sims 1992, Sims 1993a). In French Polynesia, spat collectors are gen- erally deployed for 6 months (Coeroli et al. 1984, Cabral et al. 1985, Lintilhac 1987); however, this period is extended in some French Polynesian lagoons (Preston 1990) and, in Manihiki atoll in the Cook Islands, collectors are immersed for up to 2 years (J. Lyons pers comm. 1997). P. margaritifera are generally harvested from collectors when they are large enough [65-90 mm dorsoven- tral measurement (DVM) Nicholls 1931), to be hung from dropper ropes or "chaplets" ( AQUACOP 1982. Preston 1990, J. Lyons pers comm. 1997). Oysters are drilled through the hinge of their shell and attached to the chaplets with wire or monofilament fishing line. Chaplets are then connected to submerged longlines. and this is the predominant method of holding adult P. margaritifera in Polynesia. The success of the Polynesian pearl culture industry has not ICLARM Contribution No. 1475. gone unnoticed by other small island nations in the Pacific (Lucas et al. 1998). which historically have relied on a more modest income from the sale of P. margaritifera shell for its nacre or mother-of-pearl (MOP) (Richards et al. 1994. Gervis and Sims 1992). However, not all nations with stocks of P. margaritifera, have access to "closed" atoll lagoons. For example, in the central- western Pacific, most reefs fringe high islands or occur in shallow, sublittoral areas (Wells 1988) with few "closed" atolls. Between 1994 and 1997, Friedman et al. (1998) conducted trials to adapt spat collection and culture protocols used in Polynesia for collec- tion and growout of P. margaritifera in the "open" reefs of the Solomon Islands. Their study found that commercial quantities of P. margaritifera spat could be collected from open reefs at certain sites and at certain times (Friedman and Bell 1996. in review. Friedman et al. 1998). The study also showed that collectors har- vested after 6 months held large numbers of dead spat and that greater numbers of live spat could be amassed if collectors were harvested after 3 to 4 months (Friedman and Bell in review b). Because P. margaritifera were removed from collectors at a small size (10 to 30-mm DVM), and invertebrate and fish preda- tors are widespread in Solomon Islands (Friedman et al. 1998. Friedman 1998). it is essential lo nurse juveniles until they attain a "size refuge" (Coeroli et al. 1984). This process has been termed "intermediate culture" (Ventilla 1982). because it covers the cul- ture stage between spat collection and transfer of oysters to chap- lets. Because there is less emphasis on rearing juvenile pearl oys- ters in the "closed" atolls of Polynesia, there is a paucity of infor- mation relating to this stage in the culture process. Therefore, the aim of this study is to compare growth and survival of P. marga- ritifera in a number of intermediate culture systems. The informa- tion generated by this study will not only assist in developing appropriate culture protocols for pearl oysters in open reef sys- 159 160 Friedman and Southgate terns, but will be valuable for those developing nursery culture systems for hatchery-reared P. margaritifera spat (Southgate and Beer 1997). MATERIALS AND METHODS There were two phases of this study. In Phase One ( 1994 and 1995). spat were harvested from collectors deployed at 21 sites spanning over 500 km of the Solomon Islands (Fig. 1 ). Oysters were grown out at nine of these sites (Fig. 1) on submerged lon- glines (Fig. 2). In Phase Two (1996 and 1997), spat were collected at 36 sites within the Western Province of the Solomon Islands, and grown out at one site in Gizo lagoon (Fig. 1). When spat collectors were harvested, live oysters were placed in a number of intermediate culture units; lantern nets, panel nets, trays, and glued onto ropes (Fig. 3). Lantern nets consisted of a maximum of eight platforms surrounded with 6- or 12-mm netting. The platforms were positioned on a frame before being stocked with juveniles and enclosed with mesh. The size of the mesh used to contain the juveniles depended on the size of the oysters in the experiment. Panel nets were constructed from a galvanized wire frame covered with 6- or 12-mm netting. They had five horizontal rows that could be accessed to insert juveniles from holes cut into one side of the netting (Fig. 3). Trays were made from stiff per- forated plastic (8.5-mm mesh) and had removable lids that allowed for inspection and removal of predators (Fig. 3). In a fourth culture unit, oysters were glued to 4-mm rope enclosed behind stiff plastic- mesh ( 18-mm mesh size) (see Fig. 3). Phase One Growth of Juveniles in Lantern Nets at 3— 4-m Depth P. margaritifera juveniles (936 individuals) were cultured at nine sites throughout the Solomon Islands, in lantern nets sus- pended at 3 to 4-m depth, on longlines in shallow water reef areas (8 to 25-m depth). Lantern nets were checked every 3 months to remeasure juveniles, remove predators, clean or change the meshes, and record survival. Growth of Juveniles of Different Sizes in Lantern Nets at 6-m Depth In April 1995, an experiment was set up to assess the effects of two husbandry regimes on the growth and survival of two sizes of juveniles (10 to 25-mm DVM and 26 to 55-mm DVM). Lantern nets were suspended from a longline set at 6 m in Gizo lagoon, outside the reef flat in front of ICLARM's Nuse Tupe Research Station (NTRS) (Fig. 1). The longline was deployed in deeper water (25-30 m), running parallel to the reef edge (-15 m out from the reef edge). For each husbandry regime, eight (4x2 juvenile sizes) replicate lantern nets, each holding eight oysters (four juve- niles per platform) were deployed. All oysters were marked indi- vidually with glued tags (n = 128). In the first husbandry regime, meshes alone were cleaned of predators, epibionts. and fouling. In the second regime, both oysters and meshes were cleaned. This process was carried out fortnightly for 6 months, after which the second cleaning regime was used for both sets of lantern nets. Growth (DVM) and survival of juveniles was recorded after 3, 6, and 12 months. Phase Two For this phase of the study, longlines for growout of juveniles were set at a depth of 9-12 m. in 35 to 45-m of water, just to the northeast of NTRS (Fig. 1). Longlines were sited >50 m from fringing reef, over sandy substrate. The area chosen for deploy- ment of longlines was within a section of Gizo lagoon that was approximately 1 knr. had a mean depth of -40 m, and had nu- merous passages and submerged reefs linking the lagoon to the open ocean. In Phase Two. the timing of husbandry checks was increased to 2-3 times a month; SCUBA divers brushed off algal fouling and manualU removed predators from growout units. Comparison of Growth and Survival of Juvenile Oysters in Lantern Nets and Panel Nets In Phase One, lantern nets had two deficiencies for the growout of juveniles: ( 1 ) larvae of invertebrate predators and particulate matter settled onto the platforms within the nets, and these plat- 7°S. 9°S.. 11" S-- ICLARM's*' ^fc,-.' Nusa Tupe *■•■ _ Fieldstation Gela Islands &$> VVa Gaudalcanal Malaita SOLOMON ISLANDS Makira 161°E 157° E 159° E Figure 1. Twenty-one sites (triangles) where spat of P. margaritifera were collected and cultured (boxed triangles! in Phase One and, Gizo Lagoon (insert), where spat where held for growout trials in Phase Two of the study. Growout of Blacklip Pearl Oysters in the Solomons 161 Substrate Figure 2. Longline system used to suspend growout units for juvenile P. margaritifera. forms were difficult to access for cleaning; and (2) the flexible mesh on the sides of the nets made removal of algal fouling dif- ficult. Panel nets (Fig. 3) were trialed in an attempt to overcome some of these problems. Three experiments to compare growth and sur- vival of juvenile P. margaritifera held in lantern nets and panel nets were conducted from 29 March to 29 September 1996 (Table 2). For the three experiments, juveniles of different mean sizes were used (DVM of 16- 24-, and 33-mm). To ensure that oysters were not lost through the meshes, juveniles of the smallest size class were enclosed behind 6-mm mesh; whereas, it was possible to use 12-mm mesh for the two experiments involving larger spat. Growth (DVM) and survival were recorded when units were re- moved from the water after 3 months. Comparison of Growth and Survival of Juvenile Oysters in Panel Nets and Trays Because panel nets also proved difficult to keep clean, an ex- periment was conducted using rigid perforated plastic trays that were easier to brush clean of algae. The first experiment compared growth and survival of juveniles of 24 mm in panel nets and trays and was run between 29 May 1996 and 29 August 1997 (Table 2). Growth (DVM) and survival were recorded when units were re- moved from the water after three months. A second experiment was run from 29 January to 29 June 1 997 (Table 2). In this experiment, we recorded growth (DVM), wet weight, and survival of 1 1 mm juveniles when units were removed from the water after 5 months. Different colored threads were glued to 10 oysters per replicate at the start of the experiment, so that individual growth rates could be calculated. Use of Cyano-Acrylate Glue in Intermediate Culture Although management of plastic trays (Fig. 3) was faster and easier than panel nets, juveniles tended to form aggregations or '"clumps" (Southgate and Beer 1997) in the trays. Juveniles "trapped" within these clumps had stunted development. Cyano- acrylate glue (Loctite 454 gel®) was assessed as a means of fixing juveniles to the sides and bottom of trays to prevent clumping and to determine whether spacing of oysters affected growth. Glue was also used to fix juveniles onto rope, which was then surrounded by stiff mesh of large size (Fig. 3). The mesh was too coarse to hold "loose" oysters but was stiff enough to be brushed clean of algal fouling. Between 29 April and 29 September 1997. both growth (DVM and wet weight) and survival of juveniles were compared among four growout units: panel nets: trays with juveniles loosely added: trays with juveniles glued in place; and ropes with juveniles at- tached (glued) enclosed in large mesh. For each of the 10 replicates (25 juveniles per replicate), different colored threads were glued to 10 oysters at the start of the experiment so that individual growth rates could be calculated when growth units were removed from the water after 5 months. Lantern Net Panel Net Perforated Tray Enclosed Chaplet (platforms 40 cm dia ) (75x45cm) (60x35x10cm) (150 cm long, 15 cm dia.) Figure 3. Units used for intermediate culture of P. margaritifera juveniles removed from spat collectors. 162 Friedman and Southgate Growth (DVM and wet weight) and survival of juveniles at- tached to trays and ropes were also compared in a second experi- ment deployed between 2 May and 2 October 1997 (Table 2). Again, threads were used to mark juveniles individually. General Growth Rates of Juveniles in Intermediate Cultue In 1996 and 1997, several different batches of oysters of dif- ferent initial size, were reared in the intermediate culture systems described above. Growth (DVM) was measured after 3 and/or 5 months, and growth trajectories were plotted. Analysis of Data To examine differences in survival of juveniles reared in two types of culture units, /-tests were used in comparisons of live oyster number. In comparisons of survival among more than two types of culture units, a one-way analysis of variance (ANOVA) was used to analyze survival data. To compare growth of juveniles marked individually, growth measurements from individual oysters were compared by /-test and. when appropriate, one-way ANOVA. To compare growth of juveniles in lantern nets, panel nets, and trays, the final size of juveniles were compared in each experiment using /-tests. In these tests, only subunits (e.g.. platforms in lantern nets and rows in panel nets) and whole trays, which had 100% survival, were used. Before /-tests or ANOVA. data were checked for homogeneity of variance using Levene's or Cochran's test, respectively, and transformed to log10(x+l) to meet this assumption, where neces- sary. Significant differences between means were identified using Tukey's HSD test. RESULTS Phase One Growth of Juveniles in Lantern Nets at 3 to 4-m Depth Spat removed from collectors immersed for 6 months had a mean DVM of 20.4 mm ± 0.4 SE (n = 936). and a range of 3-61 mm. Annual growth increments of juveniles of 10-100 mm DVM. grown in lantern nets suspended from shallow water longlines, are shown as a Ford Walford plot in Fig 4. The "growth performance 220 200 180 160 140 120 =g 100 + e. so hi. i 40 . 20 - 0 ^r- Loo = 183.45 0 20 40 60 80 100 120 140 160 180 200 220 L(t) Figure 4. A Ford-Walford plot of annual growth of P. margaritifera juveniles (n = 101) held in lantern nets at eight sites spanning 500 km of Solomon Islands. indicator" or 4>' value, which can be calculated from the K and L . derived from this plot (4>' = log K + 2 log hv. Munro and Pauly 1984) was 4.39. Survival of spat grown in lantern nets in shallow reef areas was poor and averaged 36.2% ± 8.4 SE. (n = 9 sites). Growth and Survival of Juveniles of Different Sizes in Lantern Nets at 6-m Depth Both sizes of juveniles held in lantern nets at 6 m grew at rates similar to those held at 3 to 4 meters (y = 0.49x + 89.03. r = 0.29, 4>' = 4.34). Because of heavy mortality, it was impractical to analyze the differences in growth between husbandry regimes in this experiment. Only 49 (38.3%) of the 128 juveniles remained alive at the end of 6 months. Fish associated with a nearby reef caused the mortality, accessing the platforms by ripping the mesh of the lantern nets. There were no significant differences in survival between the two husbandry treatments (Table 1 ). Mortality was. however, sig- nificantly greater than smaller-sized juveniles (Table 1). The de- cline in abundance of live juveniles over the course of this experi- ment was greatest in the first three months for the smaller size classes (Fig. 5). Phase Two Comparison of Growth and Survival of Juvenile Oysters in Lantern Nets and Panel Nets Growth of oysters was significantly greater in lantern nets for the smallest size class of juveniles, but no significant difference in growth was detected between the two growout units for the two larger size groups (Table 2. Fig. 6). Survival for juveniles held in lantern nets and panel nets for 3 months was 59.5 and 96.0%. respectively (Fig. 7). Survival of oysters was significantly greater in panel nets for the smallest size class of juveniles, but no sig- nificant difference in survival was detected between the two gro- wout units for the two larger size groups (Table 2. Fig. 7). Survival for the two larger size groups of juveniles was good (mean 86.9%). corresponding to a period when settlement of predators (e.g.. Cy- matium spp) was low. TABLE 1. Survival (%) at 3, 6. and 12 months of small (10 to 25-nim DVM) and large (26 to 55-mm DVM). P. margaritifera juveniles deployed in lantern nets and subjected to two husbandry regimes. Husbandry Regime Clean Mesh and Oysters Clean Mesh Only 3 Months 6 Months 36.0" 54.8" 34.4" 42 V Size of Oysters Small Large (10-25 mm) (26-55 mm) 3 Months 6 Months 1 Year 36.0" 54.8" 25.0" 51.6b 15.6" 36.0h Means with the same superscript do not differ significantly by /-test (p<.05). Growout of Blacklip Pearl Oysters in the Solomons 163 0. 100 80 60 20 (d) 60 9% Months Figure 5. Percentage survival of/', margaritifera juveniles held in lan- tern nets after 3, 6, and 12 months. Oysters were divided into size classes: 10-19 mm (al. 20-29 mm (bl, 30-39 mm (c). and 40 mm and above (d). Comparison of Growth and Survival of Juvenile Oysters in Panel Nets and Trays In the first experiment, juveniles in panel nets (46.97-mm ±1.1 SE DVM) were significantly larger than juveniles held in trays (43.59-mm ± 0.7 SE DVM) after 3 months (Table 2). Survival in panel nets and trays averaged 94 and 969c. respectively, and did not differ significantly (Table 2). In the second experiment, there was no significant difference in growth (DVM) of juveniles in panel nets (29.24-mm DVM, ± 0.56 SE) and trays (3 1 .47-mm ± 1 .3 SE DVM) after 5 months (Table 2 ). There was also no significant difference (df 16. /-value = -0.196. p = .847) in wet weight of juveniles between panel nets (1 1.68 g ± 0.47 SE) and trays (1 1.93 g ± 1.33 SE) after 5 months. Despite high mortality of oysters in two replicate trays because of preda- tion by Cymatium spp. gastropods, there was no significant differ- ence in survival between panel nets (70.0 % ± 4.0 SE. n = 10) and trays (60.5% ± 10.7 SE. n = 10) (Table 2). Use of Cyano-Acrylate Glue in Intermediate Culture Growth (DVM) of glued oysters was significantly greater than that of oysters placed loosely into trays or in panel nets (Table 3. Fig. 8). There was no significant difference in growth increment between juveniles glued to ropes and juveniles glued to trays after 5 months (Table 3. Fig. 8). Survival was greatest for juveniles attached to rope enclosed behind mesh (88.4 9c ± 2.6 SE). Survival of oysters glued to trays (86.4 9c ± 2.4 SE) and in panel nets (82.4 % ± 3.0 SE) was also high. Oysters not glued into trays had the lowest survival (74.8 9c ± 4.9 SE). Analysis of survival at 5 months was significantly different among the four culture units tested (df 3,36, F = 3.20, p = .035), however post hoc analysis (Tukey's HSD) only distinguished significant differences between oysters attached to rope and oysters "loose" in trays. In the second experiment, average growth (DVM) of juveniles glued into trays (31.29 mm ± 1.71 SE) and onto rope (31.68 mm ± 0.64 SE) did not differ significantly (df 17. / value = 0.201. p = .843) after 5 months. Difference in survival of juveniles glued onto trays (71.2 % ± 6.10 SE) and glued onto rope (86.7 % ± 3.26 SE) was significant (df 17. t value = 2.16, p < .05); however, because the two datasets were heterogeneous by Levene's test (p = .01), this result should be viewed with caution. Growth Rates of Juveniles in Intermediate Culture The growth trajectory of 10 batches of oysters that entered intermediate culture at different sizes is shown in Figure 9. DISCUSSION In the Solomon Islands, "intermediate" culture is required to nurse juvenile pearl oysters collected from the wild before they can be hung on chaplets. Juvenile P. margaritifera grew well in the culture units tested in this study; batches of juveniles with initial DVM of 8.3 to 51.5 mm increased in size by 20.4 to 24.8 mm in 3 months and 30.7 to 36.5 mm in 5 months. These growth rates compare favorably with growth of P. margaritifera reported from "closed" and semiclosed atolls in the Pacific. For example, in Takapoto atoll, French Polynesia, juveniles of 40 to 50-mm DVM grew 30 mm in 6 months (Coeroli et al. 1984. Lintihac 1987); whereas, in the Cook Islands. Braley ( 1997) reported that hatchery produced P. margaritifera juveniles with a mean DVM of 10 mm grew approximately 16.4 mm DVM in 3 months. On the other hand, Sims ( 1993b) presented size-at-age data for 9-month-old P. margaritifera from Manihiki atoll as approximately 81 mm DVM. which indicate fast growth of juveniles in a "closed" atoll envi- ronment. However, the age of spat used in Sims's study may initially have been underestimated, because "median date" or "heaviest fall" was used to estimate the age of spat removed from collectors prior to the beginning of the growth study. In the open reef systems of Dongonab Bay in Sudan. P. mar- garitifera juveniles collected as spat between 18 to 37-mm DVM grew by 13 to 24 mm and by 24 to 32 mm in 3 and 5 months of culture, respectively (Nasr 1984). However, in Sudan, there is little or no growth during winter (Nasr 1984). In Australia, hatchery- produced juveniles of P. margaritifera with a mean DVM of 13.9 mm, grew by 21.8 to 26.6 mm when held loosely in perforated plastic trays for 19 weeks (4.3 months) (Southgate and Beer 1997). Our study demonstrates that growth is also dependant on the method used to hold juvenile oysters; that is. juveniles held under different intermediate culture conditions grew at different rates. A major difference between the systems used in this study was the ability to separate oysters. For example, oysters held in panel nets or glued into trays and on ropes were prevented from clumping; whereas, juveniles that were able to move around freely tended to form aggregations, such as those reported by Crossland (1957). Southgate and Beer (1997), and Sims and Sarver (1998). Oysters in the units where clumping occurred exhibited highly variable growth rates as a result of increased competition for food and space. In another study on P. maxima, Taylor et al. (1997b) re- ported that this behavior promoted an increase in the prevalence of growth deformities in juveniles. Separation of juveniles in intermediate culture had the added advantage that units were easier to check for predators, because Cymatium spp. and crabs could not hide within clumps of oysters. The results of the trials where juveniles were stuck directly onto ropes highlighted the potential for holding juveniles behind meshes that were too large to contain oysters, but small enough to afford the growing juveniles some protection from fish. Although the use of cyano-acrylate glue gave some of the best growth rates in this study, adhesives are not a panacea for the problems en- 164 Friedman and Southgate TABLE 2. Results of experiments to assess growth and survival of P. margaritifera juveniles in various culture units in Phase 2 of the study (between March 1996 and October 1997). Oysters per Mean Starting Survival Growth (DVM) Experiment Replicates Replicate Size (DVM) df t-value P df t-value P a) Lantern vs. panel nets 29/3/96 to 29/6/96 12 20 16 mm 22 -3.77 0.001 250 3.81 <0.000 29/5/96 to 29/8/96 12 25 24 mm 22 -1.91 0.069 510 0.16 0.870 29/6/96 to 29/9/96 8 20 33 mm 14 -1.45 0.168 184 -0.66 0.507 Panel nets vs. trays 29/5/96 to 29/8/96 12 25 24 mm 22 0.52 0.607 460 -2.79 0.006 29/1/97 to 29/6/97 10 20 1 1 mm 18 0.82 0.418 16 -1.66 0.116 c) Trays (glued) vs. ropes (glued) 2/5/97 to 2/10/97 10 25 34 mm 17-' 2.16 0.045 133 0.64 0.52 1 JOne rope was lost in this experiment. countered in growout; cyano-acrylate is both expensive and diffi- cult to apply. Because the central south-western Pacific [e.g., Fiji, Vanuatu, Papua-New Guinea (PNG)] has few "closed," deep-water lagoons, pearl oyster growout under these conditions contrasts with gro- wout in the atoll lagoons of the eastern Pacific. Whereas atoll lagoons are surrounded by low-lying carbonate islands, the reef systems in the Solomon Islands are subject to relatively large inputs of nutrients and particulate matter from high islands (Littler et al. 1991). The higher nutrient load in the lagoons of Solomon Islands may have been a factor in the good growth rates recorded in this study (Yukihira 1998). However, the negative side of this is the increased algal fouling when compared to the relatively nutri- ent-poor atoll lagoons of Polynesia. In the Solomon Islands, meshes of culture units required regular cleaning; thereby, increas- ing labor needs. Although meshes required cleaning, there was relatively little fouling by such "cementing" organisms as bivalves and polychetes ("hard" fouling). Algal fouling is easier to remove than "hard" fouling, and regular brushing for this purpose may have inhibited recruitment and survival of hard-fouling organisms and other byssally attached bivalves, which have been shown to be a problem during growout of other pearl oyster species (Taylor et al. 1997a). In addition to being simpler to remove, algal fouling does not directly compete with juveniles for food resources and space. We also found that control of algal fouling was easier when culture units were made of stiff plastic meshes. This material was more practical and cost effective than flexible mesh, which could not be easily cleaned and had limited potential for reuse. Coeroli et al. (1984) reported that 30% of 6 to 12-month-old P. margaritifera juveniles were lost in culture in French Polynesia and stated that fishes from the family Balistidae and Tetraodon- tidae were the chief predators of pearl oysters. In initial trials in the Solomon Islands, fish devastated juveniles in intermediate culture when longlines were deployed too close to reefs (Friedman et al. 1996). Although there was no direct comparison between growout culture in shallow and deeper water in this study, the lack of broken shells in culture units on longlines placed farther from reefs, and evidence presented in other studies (Sims and Sarver 1995), supports the inference that juvenile pearl oyster culture conducted at a distance from reefs and in deeper water reduces predation by fish. For example, in the Marshall Islands 5.5% of juveniles were lost to fish predation on longlines set in deeper 40 35 30 E E 25 2 > 20 .c o O 15 10 5 0 100 80 - 40 20 - 16 mm 24 mm 33 mm 16 mm 24 mm 33 mm Size at start of grow-out (DVM) Figure 6. Growth (final DVM-mean DVM at start of experiment) in shell size (DVM mm) of P. margaritifera juveniles held for 3 months in lantern nets (shaded) and panel nets (open). Size at start of grow-out (DVM) Figure 7. Percentage survival (±SE) of P. margaritifera juveniles held in lantern nets (shaded) and panel nets (open) for 3 months. Columns marked with an asterisk differed significantly in /-tests. Growout of Blacklip Pearl Oysters in the Solomons 165 TABLE 3. Results of one-way ANOVA for effects of culture unit on a) increase in mean shell height (I)VM). h) increase in mean wet weight, and c) survival of juvenile P. margaritifera, grown for 5 months. df MS a) Growth (DVM mm) Culture unit 3 231.771 Residua] 36 7.911 h) Change in wet weight (gl Culture unit 3 423.122 Residual 36 19.822 c) Survival Culture unit 3 22.567 Residual 36 7.050 231.771 29.299 ,0000 21.346 3.201 .0000 .0350 The four culture units in this study were: panel nets; trays with juveniles loosely added: trays with juveniles glued in place; and ropes with juveniles attached (glued), enclosed in large mesh. a) E E > 'a> b) 'a) c) 75 > _ra "c CD o I— CD Q- 50 40 30 20 10 0 50 40 30 20 10 0 100 80 60 40 20 0 c c D D 1 T w*- ■^ ■-*,-£•■-' . EF EF water, as compared to 25.5% losses on shallow water longlines (Sims and Sarver 1995). In French Polynesia, pearl farmers hold young oysters behind galvanized wire mesh to isolate oysters from attacks by balistid fish (Lintilhac 1987). Smaller fish also had an effect on growth. In this and other studies (Sims 1993a; Southgate and Beer 1997). small balistids and tetraodontids '"grazed" on the non-nacreus shell margins of juvenile oysters in culture. In the Solomon Islands, this was only found to be significant at embayed inshore sites, and relocation of growout longlines to less turbid areas with greater water movement reduced or eliminated this problem. Once longlines were located in areas free from predation by reef fish, survival of juveniles was influenced primarily by inver- tebrate predation. Invertebrate predators, such as Cymalium spp. (Govan 1995), crabs, and flatworms (Newman et al. 1993. Taylor et al. 1997b) recruited into growout units from the plankton (Day- ton et al. 1989. Newman et al. 1993. Friedman and Bell 1996. Friedman 1998). The effect on the hydrodynamics of water flow has been presented as a major determinant of the success of culture unit design (Claereboudt et al. 1994). In the Solomon Islands, invertebrate predators were found in all growout units, but were more common in lantern nets and trays than in units that had smaller holding spaces (e.g.. panel nets) or less scope for reducing water flow (e.g.. glued ropes). This may have been caused by differences in water flow characteristics that could have influenced settlement of suspended particulate matter and the larvae of po- tential predators. Predators that settle within nondivided culture units (e.g., trays) had access to all the juveniles within that unit. On two occasions 100 E E > Q 0) N "5 CO Panel Trays Glued Glued nets trays rope Figure 8. Changes in mean (±SE) a) mean shell height, b) wet weight, and c) survival of juvenile P. margaritifera held in growout units after 5 months. Columns with the same letter do not differ significantly (p < .05) in post hoc tests (Tukey's HSD). Months in intermediate culture Figure 9. Growth trajectories for P. margaritifera juveniles of differ- ent sizes (±SE) grown in intermediate culture. Once oysters reach a size of approximately 65-mm DVM I marked on graph), they can be drilled for ear hanging. Trajectories that have been extended past the 5-month period (dashed lines), were added using real data from growth experiments involving larger juveniles. 166 Friedman and Southgate in this study. 20 juveniles ( = 11 mm DVM) were killed by a Cy- matium spp. within a single tray. To combat the problem of se- quential juvenile predation in intermediate culture, one of two strategies can be adopted: ( 1 ) juveniles could be separated into small groups by mesh barriers; or (2) units could be designed to make them easier to inspect for predators. The first strategy relies on the predator outgrowing the mesh size of a single compartment in the growout unit and being prevented from entering other com- partments of the culture unit. One such system has been success- fully trialed in the Marshall Islands, where settlement of Cymatium spp. had previously devastated hatchery-produced P. margaritifera juveniles in nursery culture (Sarver et al. 1998. Sims and Sarver 19981. In the Solomon Islands, the second strategy was adopted for three reasons: ( 1 ) because in intermediate culture of spat taken from collectors, there is no need to deal with very small spat (<6-mm DVM). which are difficult to handle and check for preda- tors: (2) multispaced culture units are difficult to set up and keep free of fouling; and (3 1 because staff in the Solomon Islands could quickly recognize settlement periods of such predators as Cyma- tium spp. and respond accordingly. Although no reliable seasonal trends in settlement of the main predator groups have been recog- nized in the Solomon Islands, divers increased the intensity of inspections when large numbers of newly settled predators were found. In this way. predators were removed from growout units when they were small — before they caused significant juvenile mortality. This study showed that there was a positive relationship be- tween the size of juveniles placed into intermediate culture and their survival. Also, Coeroli et al. (1984) suggested that oysters over 50-mm DVM were "resistant to attacks from predators." In the Solomon Islands, oysters that had reached -65-mm DVM could be removed from intermediate culture and grown on chaplets without a protective mesh covering. In summary, this study has shown that within the open-reef systems typical of the western Pacific: ( 1 ) the rate of growth of juveniles compared favorably to that reported for the closed-atoll lagoons of Polynesia. In the Solomon Islands, juveniles of 10-mm DVM placed in intermediate culture generally attained a size suit- able for transfer to chaplets (-65 mm DVM) in 8 months. Those entering intermediate culture at a size of 25- to 30-mm DVM were ready for moving to chaplets after 5-6 months; (2) the siting of longlines in deepwater decreased mortality attributed to fish asso- ciated with reef; (3) important characteristics of pearl oyster gro- wout units include ease of cleaning and access for regular inspec- tion and removal for invertebrate predators; and (4) separating oysters in intermediate culture resulted in more uniform growth. ACKNOWLEDGMENTS We thank Gideon Tiroba. and Ruth and Barley White Dunne for their assistance with the experiment. We are also thankful to Johann Bell and Andrew Beer for their instructive comments on the draft manuscript. This study was partially supported with fund- ing from the Australian Centre for International Agricultural Re- search (ACIAR). LITERATURE CITED AQUACOP. 1982. French Polynesia— Country Report, pp. 31-33. In: F. B. Davy and M. Graham (eds.). Bivalve Culture in Asia and the Pacific. Proceedings of Workshop Held in Singapore. 16-19 February 1982. International Development Research Centre. Ottowa. Canada. Braley. R. D. 1997. Final Report — Appendices |Binding B] Outer Islands Marine Resources Management Training Project. 23-24. Asian Devel- opment Bank. Cahral. P. K. Mizuno & Tauru. A. 1985. Preliminary data on the spat collection of mother-of-pearl (Pinctada margaritifera, bivalve, mol- lusk) in French Polynesia. Proceedings of the 5th International Coral Reef Congress. Tahiti 5:177-182. Claereboudt. M. R.. D. Bureau, J. Cote & J. H. Himmelman. 1994. Fouling development and its effect on the growth of juvenile giant scallops (Placopeaen magellanicus) in suspended culture. Aquacuhure 121: 327-342. Coeroli. M. & K. Mizuno. 1985. Study of different factors having an influence upon the pearl production of the black lip pearl oyster. Pro- ceedings of the 5th International Coral Reef Congress. Tahiti 5:551- 556. Coeroli. M. D. De Gaillande. J. P. Landret & AQUACOP (D. Coatanea). 1984. Recent innovations in cultivation of mollusks in French Polyne- sia. Aquacuhure 39:45-67. Crossland. C. 1957. The cultivation of the mother-of-pearl oyster in the Red Sea. Aust. J. Mar. Freshwater Res. 8:111-130. Dayton. P. K.. J. H. Carleton. A. G. 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. Doubilet. D. 1997. Black pearls of French Polynesia. Nat. Geo. 191:30-37. Fassler. C. R. 1995. New developments in pearl fanning. World Aquacult. 26:5-10. Friedman. K.J. 1998. Black pearl farming in the Pacific: the open-reef systems of Solomon Islands. Austasia Aquacult. 12:3-9. Friedman. K. J.. Bell, in review a. Variation in abundance of blacklip pearl oyster {Pinctada margaritifera Linne) spat from inshore and offshore reefs in Solomon Islands. Aquacuhure. Friedman. K. J.. Bell. J. in review b. Effects of differing immersion times on yields of the blacklip pearl oyster. Pinctada margaritifera (L). from spat collectors in Solomon Islands. Aquacuhure. Friedman. K. J. & J. Bell. 1996. Effects of different substrata and protec- tive mesh bags on collection of spat of the pearl oysters. Pinctada margaritifera (Linnaeus 1758) and Pinctada maculata (Gould 1850). J. Shellfish Res. 15:535-541. Friedman. K. J. Bell. M. Gervis & G. Tiroba. 1996. Progress on research on the potential of farming blacklip pearl oysters in Solomon Islands. South Pacific Comm. Pearl Oyster Information Bull. 9:1 1-13. Friedman. K. J. Bell & G. Tiroba. 1998. Availability of wild spat of the blacklip pearl oyster, Pinctada margaritifera. from "open" reef systems in Solomon Islands. Aquacullure 167:283-299. Gervis, M. H. & N. A. Sims. 1992. The biology and culture of pearl oysters (Bivalva: Pteriidae). ICLARM Study Review 21. ODA (Pub.). London. 49 pp. Govan. H. 1995. Cymatium muricinum and other ranellid gastropods: ma- jor predators of maricultured tridacnid clams. ICLARM Tech. Rept. 49. 136 pp. Lintilhac. J. 1987. Black Pearls of Tahiti. Royal Tahitian Black Pearl Book. Papeete. Tahiti. 53 pp. Littler. M. M.. D. S. Littler & E. A. Titlyanov. 1991. Comparisons of N- and P-limited productivity between high islands versus low carbonate atolls in Seychelles Archipelago: a test of the relative dominance para- digm. Coral Reefs 10:199-209. Lucas, J. S.. M. Ito. M. Kamatie. J. R. Norton. P. C. Southgate. B. Tioti & J. R. Whitford. 1998. Pacific Island Pearl Oyster Resource Develop- ment Project, pp. 112-115. In: Y. Zhoa.. F. Hu.. H. Zhou.. H. Cui. C. Yao., F. Din. and Y. Lu. (eds.). Proceedings of the 4th Asian Fisheries Forum. Asian Fisheries Society. Manila. Munro. J. L. & D. Pauly. 1984. Once more on the comparison of growth in fishes and invertebrates. Fishbyte 2:21. Growout of Blacklip Pearl Oysters in the Solomons 167 Nasr. D. H. 1984. Feeding and growth of the pearl oyster Pinctada mar- garitifera (L.) in Dongonab Bay. Red Sea. Hydrobiologia 110:241- 245. Newman. L. J., L. R. G. Cannon & H. Govan. 1993. Stylochus (Imogene) matatasi n. sp. (platyhelminthes polycladida): pest of cultured giant clams and pearl oysters from Solomon Islands. Hydrobiologia 257: 185-189. Nicholls, A. G. 1931. On breeding and growth rate of the black-lip pearl oyster (Pinctada margaritifera). Rept. Great Barrier Reef Commission 3:26-31. Preston. G. 1990. Pearl culture in three French Polynesian atolls 1986- 1987. South Pacific Comm. Pearl Oyster Inf. Bull. 1:10-12. Remoissenet, G. 1996. From the emergence of mortalities and diseases on Pinctada margaritifera to the effects on the pearling industry, pp. 371- 386. In: Present and Future of Aquaculture Research and Development in the Pacific Island Countries. Proceedings of the International Work- shop, 20-24. November. 1995. Ministry of Fish. Tonga, JICA. Richards. A. H.. L. J. Bell & J. D. Bell. 1994. Inshore fisheries resources of Solomon Islands. Mar. Poll. Bull. 29:90-98. Sarver, D. J.. N. A. Sims & R. P. Clarke. 1998. Pinctada nursery culture: overcoming the bane of Cyinatimn. South Pacific Comm. Pearl Oyster Inf. Bull. 11:31. Sims. N. A. 1992. Abundance and distribution of the black-lip pearl oyster. Pinctada margaritifera (L. ), in Cook Islands. South Pacific. Aust. J. Mar. Freshwater Res. 43:1409-1421. Sims. N. A. 1993a. Pearl Oysters, pp. 409-130. In: A. Wnght (ed.). Near- shore Resources of the South Pacific. IPS. Suva. Sims. N. A. 1993b. Size, age, and growth of the black-lip pearl oyster. Pinctada margaritifera (L.). (Bivalvia; Pteriidae). ./. Shellfish Res. 12: 223-228. Sims, N. A. & D. J. Sarver. 1995. Feasibility of commercial black pearl culture in Namdrick Atoll. Marshall Islands. Final Rept. Saltonstall- Kennedy Grant Program. 24 pp. Sims, N. A. & D. J. Sarver. 1998. Marshall Islands farming expansion: new hatchery, new atolls, new techniques. South Pacific Comm. Pearl Oys- ter Info. Bull. 11:6-7. Southgate. P. C. & A. C. Beer. 1997. Hatchery and early nursery culture of the blacklip pearl oyster {Pinctada margaritifera L.). J. Shellfish Res. 16:561-567. Taylor, J. J., R. A. Rose. P. C, Southgate & C. E. Taylor. 1997a. Effects of stocking density on growth and survival of early juvenile silverlip pearl oysters Pinctada maxima (Jameson), held in suspended nursery culture. Aquaculture 153:41-49. Taylor, J. J.. P. C. Southgate & R. A. Rose. 1997b. Fouling animals and their effect on the growth of silver-lip pearl oyster. Pinctada maxima (Jameson). Aquaculture 153:31—10. Ventilla, R. F. 1982. The scallop industry in Japan. Adv. Mar. Biol. 20: 310-381. Wells, S. M. 1988. Coral reefs of the world, vol. 3. Central and Western Pacific. United Nations Environment Programme (UNEP)/ International Union for Conservation of Nature and Natural Resources (IUCN). Gland. Switzerland. 329 pp. Yukihira. H. 1998. Feeding, energy budgets, and nutritional ecology of the pearl oysters, Pinctada margaritifera (Linnaeus) and P. maxima Jame- son. Ph.D. thesis, James Cook University. Australia. 177 pp. Journal of Shellfish Research, Vol. IS, No. I. 169-174. 1999. A NEW EPIZOOTIC OF HAPLOSPORIDIUM NELSONI (MSX), A HAPLOSPORIDIAN OYSTER PARASITE, IN LONG ISLAND SOUND, CONNECTICUT INKE SUNILA, JOHN KAROLUS, AND JOHN VOLK State of Connecticut Department of Agriculture Bureau of Aquaculturc Milford, Connecticut 06460 ABSTRACT Epizootic prevalences of MSX. Haplosporidium nelsoni, were detected in eastern oysters (Crassostrea virginica) at several sites in Long Island Sound. Connecticut. Diagnosis was based on histological sections. Twenty-two sampling stations repre- senting commercially important oyster grounds were tested. Three sampling stations were negative for MSX. the rest had prevalences from 7 to 89% (319 MSX positive specimens out of 653 processed). Two major seed areas (Housatonic River and Bridgeport Natural Bed) were uninfected, but all studied shallow « 1 5 feet) or deep (>15 feet) water growing areas were infected along the whole shoreline. Hatchery-raised seed, uninfected when imported to Connecticut, acquired infection in New Haven and Groton. Sporulation with acid-fast spores was detected in 1% of MSX-positive specimens. One specimen showed unusual pattern of sporulation with spores inside ciliated digestive duct cells and vesicular connective tissue in addition to digestive tubules. The percentage of potentially resistant specimens was low (<1%). A periodic range extension of an alternate host, usually restricted to the mid-Atlantic region of the U.S. coast, is suggested as the cause for the recent epizootics in Long Island Sound. KEY WORDS: Haplosporidium nelsoni, epizootic, eastern oyster, Crassostrea virginica, Long Island Sound INTRODUCTION MSX-disease, caused by a protozoan parasite Haplosporidium nelsoni, has caused heavy mortalities of the eastern oyster, Cras- sostrea virginica (Gmelin. 1791 ) in the Delaware Bay since 1957 (Haskin et al. 1966) and in the Chesapeake Bay since 1959 (An- drews and Wood 1967). MSX has been detected from Biscayne Bay, Florida, to Damariscotta River, Maine (Haskin and Andrews 1988). but has not been associated with mortalities in all areas. Systematic monitoring for prevalences of MSX in Long Island Sound had not been performed before the autumn of 1997. MSX was present in oysters (in 10 out of 1,337 oysters) collected 1966 to 1967 from New Haven Harbor, Connecticut, in Long Island Sound (Newman 1971). Hammonassett River, Clinton. Connecti- cut, had a prevalence of 32%, and Bridgeport had 18% in 1985 (Haskin and Andrews 1988). Oysters from the Hammonassett River were transplanted the year before to Cotuit, Massachusetts, where MSX-related mortality of 85% was estimated by the end of summer 1985 (Matthiessen et al. 1990). That same year, heavy MSX-related mortalities occurred on the south shore of Long Is- land Sound at the Flowers Hatchery, New York, where approxi- mately 90% mortalities were reported (Haskin and Andrews 1988). MSX-related mortality data from Connecticut's shoreline at that time was not available, but production fell from 243,883 bushels in 1984 to 69,721 bushels in 1987 (DEP 1995). Haskin and Andrews ( 1988) wrote: "A nagging question over the past 20 years has been. What is preventing massive epizootics, like those in the Chesa- peake and Delaware bays, from occurring in the eastern oyster areas extending from Great South Bay. Long Island, to the tip of Cape Cod?" We report a new epizootic of MSX in Long Island Sound. Connecticut. MATERIAL AND METHODS Connecticut's oyster industry is based on grounds leased from the state. Over 46,000 acres of underwater farms are cultivated. Every oyster is transplanted on the average four times before it reaches market size (3^1 years), and. consequently, every oyster is exposed to possible parasitic infection in several different sites during its lifetime. The oyster industry is based mostly on natural seed, but also hatchery-raised seed is deployed in some areas, especially in the eastern end of the state. Major seed areas are situated up such rivers as the Housatonic and Quinnipiac. and are also close to the shoreline in Bridgeport. These areas are prohibited from shellfishing for direct consumption. The oysters are trans- planted to clean water before marketing. Growing areas are clas- sified either approved (A), conditionally approved (CA). restricted (R), conditionally restricted (CR). or prohibited (P) according to National Shellfish Sanitation Program (1997). Sampling strategy for this study was to select different types of oyster grounds to represent the coastline: seed areas up the rivers and close to the shoreline, and shallow (<15 feet) and deep (>15 feet) water grow- ing areas; 22 sites altogether. Several samples were taken from oyster growing areas of economic interest, such as Norwalk. Sam- pling started in the end of August 1997. and the last sample in- cluded in this paper was accessed in the end of May 1998. Sam- pling stations from west to east are listed in Figure 1 and Table 1. Thirty oysters were collected from each sampling station (28 from two, 27 from one sampling station). Tissues were fixed in David- son's fixative (in 20%c artificial seawater). Six-p,m thick paraffin sections were stained with hematoxylin-eosin. Advanced MSX cases with migration of plasmodia to the digestive tubules, pos- sible prespores or spores were also stained with Ziehl and Harris' hematoxylin to detect acid-fast spores, according to Farley ( 1965). MSX infections were classified as initial, intermediate, advanced, terminal, or occult, according to Farley (1968). RESULTS Six hundred fifty-three (653) oysters were sectioned for detect- ing MSX. Three hundred nineteen (319) were diagnosed with MSX disease. Prevalences of MSX at different sampling sites, listed in Table 1. varied from 0 to 89%. Major seed areas. Hou- satonic River and Bridgeport Natural Bed. were uninfected at this point, but another major seed area, Quinnipiac River, sampled in spring 1998, had a prevalence of 53%. Shallow as well as deep water growing areas were infected with MSX. Norwalk Bed 1131 169 170 SUNILA ET AL. 73 00' Connecticut uinnipii Quinnipiac River Housatonic River N.Y. Fairfield Bridgtep' Westport Norwalk ;- Stamford , .^ Si* "K Branford ew Haven , Gumu Clinton New Haven Guilford ^_ + Haplosporidium nelsoni positive # uninfected Atlantic Ocean 20 Miles Figure 1. A map of sampling stations for detecting Haplosporidium nelsoni infection in Long Island Sound, Connecticut. was sampled at the depth of 47 feet with prevalence of 70%; Westport 1088 at 43 feet with prevalence of 83%. Hatchery-raised seed acquired infection in Groton (27% ) and two sites New Haven (43 and 29%). Seed, which was imported from Long Island, NY. has been uninfected when tested before import into Connecticut's waters. Infection stages were classified as follows. Initial Infection: Plasmodia Limited to Gill Epithelium First, plasmodia were detected between epithelial cells outside basement membranes. After that, plasmodia invaded hemolymph sinuses and interlamellar junctions. Hyaline hemocytes aggregated around parasitic infection areas. There was one case with initial infection originating from digestive epithelium (which might also be a secondary infection after sloughing of infected gill epithelial cells) and one case with infection starting from water tubes. Intermediate Infection: Isolated Plasmodia in the Digestive Diverticula in Addition to the Gill Infection Typical areas for plasmodia at this stage were the typhosole and vesicular connective tissue surrounding the stomach and intestine. Advanced Infection: Plasmodia Dispersed Throughout the Tissues Plasmodia have invaded the reproductive system and hemolymph vessels. There was usually a massive hyaline hemocyte response. Terminal Infection: Profuse Parasitemia in All Tissues This stage precedes death of the oyster. Tissue lysis may be present, and several specimens were sterilized by parasitic infil- tration to the follicles. Occult: Tissues with Physiological Presentation of MSX Infection hut No Plasmodia Detectahle in Section Classical signs of MSX infection were massive hyaline hemocyte response and perivascular hyaline hemocyte infiltration in the mantle (cuffing). Resistant: Moribund and Phagocytosed Plasmodia or Plasmodia Retained Outside Basement Membranes in Gills Distribution of different stages — initial, intermediate, ad- vanced, terminal, occult, or resistant, at different sampling stages are listed in Table 2. Distribution of stages through the sampling period from August 1997 to May 1998 remained surprisingly unchanged. Active in- fection seemed to continue through the winter in the samples; for example, in Branford 179, which was collected in March, but had full-blown MSX disease with initial and intermediate infections. Four cases out of 319 MSX-positive specimens had the histologi- cal characteristics of resistance; one had gill infection with mori- bund plasmodia outside the basement membrane in the gills, one HAPLOSPORIDIUM NELSON1 EPIZOOTIC IN OYSTERS 171 TABLE 1. Prevalence of Haplosporidium nelsoni at different sampling stations in Long Lsland Sound, Connecticut. Shell Sampling Length Prevalence Sampling Site Location Date (mm) (%) Characterization of Site Classification Stamford 494 4F:,0I.05'N: 73°31.96"W 01.02.98 61 0 Shallow-water growing area RR Norwalk 1131 41°02.42'N: 73°25.25'W 12.07.97 77 70 Deep-water growing area A Norwalk 173 41°03.38'N: 73 = 25. 27'W 04.29.98 73 37 Shallow-water growing area CA Norwalk 162 41°03.36'N:73°25.i2"W 12.07.97 74 87 Shallow-water growing area CA Norwalk 176 41°03.65'N: 73°25.00'W 04.29.98 82 47 Shallow-water growing area CA Norwalk Manresa Island 41°04.42'N: 73°24.55'W 12.17.97 80 57 Shallow-water growing area P Norwalk town grounds 41°04.85'N:73°23.55"W 09.16.97 91 89 Seed area RR Norwalk 40 41°04.57'N: 73°23.28'W 12.17.97 89 63 Shallow-water growing area RR Westport 1088 41°02.73'N:73°22.21"W 12.07.97 78 83 Deep-water growing area A Westport 254 41°04.50'N: 73°22.60'W 08.27.97 85 11 Shallow-water growing area CA Fairfield Ash Greek 41o09.2rN;73°14.00'W 12.16.97 78 7 Seed area P Bridgeport natural bed 41°08.50'N: 73°09.40'W 11.12.97 72 0 Seed area RR Housatonic River 41°11.00'N:73°07.52'W 11.12.97 80 0 Seed area P New Haven lot IB 41°15.47'N:72°55.26'W 01.08.98 43 43 Hatchery seed on-bottom culture RR New Haven lot 1A 4ri5.33'N:72055.56'W 01.08.98 46 29 Hatchery seed on-bottom culture RR Quinnipiac River 41018.75'N:72°53.I2'W 05.26.98 94 53 Seed area P Branford 179 41°15.73'N:72°45.63'W 03.02.98 87 83 Shallow-water growing area A Guilford East River 41°16.05'N:72o39.62'W 09.28.97 91 73 Seed area RR Clinton Hammonassett River 41°15.97'N:72°32.84'W 12.07.97 101 63 Seed area RR Clinton Cedar Island 4ri5.97'N:72°32.00"W 10.14.97 104 83 Seed and shallow-water growing area RR Clinton Hammock River 41°16.05'N:72°31.30"W 12.07.97 97 70 Seed and shallow-water growing area RR Groton Pine Island Bay 41o18.99'N:73°03.58'W 01.09.98 65 27 Hatchery seed off-bottom culture CR A = approved; CA = conditionally approved; R = restricted; CR = conditionally restricted ; P = prohibited. with moribund plasmodia outside the basement membrane and a few phagocytosed plasmodia inside interlamellar junction, and two with moribund and phagocytosed plasmodia inside a hemolymph vessel. Sporulation rate was three out of 319 MSX-positive specimens (0.94%). Prespores and spores, which did not respond to acid-fast stain, were observed in two additional animals. Sporulation showed an unusual pattern in one of the specimens. In addition to sporulation in digestive cells in digestive tubules (Fig. 2). there were several foci of spores in the ciliated digestive duct cells (Fig. 3). Nests of spores inside digestive cells broke outside digestive tubules creating disseminated accumulations of spores in vesicular TABLE 2. Progression of Haplosporidium nelsoni at different sampling stations in Long Island Sound, Connecticut. Sample MSX Sampling Site Size Positive Initial Intermediate Advanced Terminal Occult Resistant Norwalk 1131 30 21 3 4 7 7 0 0 Norwalk 173 30 11 8 2 1 0 0 0 Norwalk 162 30 26 2 5 9 3 6 1 Norwalk 176 30 14 7 4 1 0 1 1 Norwalk Manresa Island 30 17 4 5 4 1 3 0 Norwalk town grounds 27 24 3 7 6 5 3 0 Norwalk 40 30 19 2 5 5 4 2 1 Westport 1088 30 25 2 7 7 0 9 0 Westport 254 28 3 3 0 0 0 0 0 Fairfield Ash Greek 30 2 0 2 0 0 0 0 New Haven lot IB 30 13 0 3 4 6 0 0 New Haven lot 1A 28 8 3 3 2 0 2 0 Quinnipiac River 30 16 3 5 2 4 0 0 Branford 179 30 25 4 5 13 3 0 1 Guilford East River 30 22 3 3 9 2 5 0 Clinton Hammonassett River 30 19 1 3 1 0 14 0 Clinton Cedar lsland 30 25 1 4 8 4 8 0 Clinton Hammock River 30 21 2 2 7 4 6 0 Groton Pine Island Bay 30 8 3 3 1 0 1 0 172 SUNILA ET AL. ; ', 2 « ♦ \*&i Figure 2. Sporulation of Haplosporidium nelsoni in the digestive diverticula of the eastern oyster. Note the presence of spores in the connective tissue (arrows) in addition to inside digestive cells and the lumen of digestive tubules. Paraffin section, Ziehl and Harris' hematoxylin for acid-fast spores. 400x. connective tissue (Fig. 2). Hemocytes were observed in the ve- sicular connective tissue with phagocytosed spores (Fig. 4). Usu- ally, spores maturated inside digestive cells and were released in the lumen of digestive tubules. Spores were also observed inside the lumens of digestive ducts and intestine. Plasmodia were ob- served in different stages of endomitosis (Fig. 5). Gross macroscopic signs of MSX infection as described by Farley (1968); mantle recession, conchiolin deposits, and pale di- gestive diverticula, were not explicit in these samples. Pale diges- tive diverticula were recorded in 85 of 259 (33%) MSX positive specimens, but also in 44 of 334 ( 13%) uninfected specimens (\2 = 28.6; df 1; p < .005. occult MSX positives not included). Mantle recession was defined as an unprotected marginal area of the shell, which became heavily fouled. Mantle recessions were not often noticed, possibly because of the recent outset of the epizootics, so they had not developed. One shell with conchiolin deposits — an indication of resistance — was found from the Fairfield Ash Creek oyster bed. This bed. unlike the others, had a light level of infec- tion and possibly represents a population with enzootic history of MSX. There were co-infections with several other organisms. The probability for a co-infection with Perkinsus marinus (based on Ray/Mackin tissue assay. Ray 1954) in MSX-positive (176/259. 68%, occult not included) or MSX-negative specimens (240/334: 72^'c ) showed no statistically significant difference (\2 = 0.29; df 1; NS). Nematopsis ostrearum was present in MSX-positive speci- mens (69/259. 27%). but also in uninfected specimens (81/334; 24%; x2 = 0.34; df 1; NS). The probability for co-infection with Stegotricha-Uke ciliates in the digestive diverticula was present in MSX-positive specimens (37/259. 14%). but also in uninfected specimens (40/334: 12%; x2 = 0.59: df 1; NS). Ciliates on gills, mantles, or palps (several different species from the family Ancis- trocomidae) were more likely to be found in uninfected ( 123/334, 37% ) than in MSX-infected specimens (53/259: 20%; x2 = 13.25; df 1; p < .005). This result was polarized by a high prevalence of a species of a small ciliate in one of the uninfected sites (Stamford, 37%) and the off-bottom site in Groton (53%). Several fouling organisms affected the shells — boring sponges (Cliona sp. several different species), and mud worms (Polydora websteri) were com- monly found. Cliona sp. were found on MSX-positive specimens (83/259; 32%) and uninfected specimens (120/334; 36%: x2 = 0.64; df 1; NS). Polydora websteri was found in uninfected (155/ 334; 46%) and MSX-infected specimens (92/259: 36%; x2 = 4.14: df 1. p < .025). This result was influenced by a 100% preva- lence in an uninfected Housatonic River seed sample. DISCUSSION The beginning of epizootics in the Chesapeake and Delaware Bays demonstrated that infection started in the bay proper and spread to seed areas or rivers later (Andrews and Wood 1967. Ford and Haskin 1982). A similar pattern may appear in the Long Island Sound epizootics. At the time of sampling, some major seed areas were uninfected, but some, sampled later, were infected (Table 1 ). and the uninfected seedbeds may acquire infection later. The lim- iting low salinity parameter does not necessarily exist in Long Island Sound. Major oyster grounds, both seedbeds and growing areas, are situated within optimum salinity levels for MSX. from 15 to 25% (Andrews 1979). In Delaware Bay. disease-free stocks Haplosporidium nelsoni Epizootic in Oysters 173 Figure 3. Unusual sporulation of Haplosporidium nelsoni inside digestive duct cells (arrows). Note ciliated digestive duct cells. Paraffin section, H & E. 600x. are obtainable from low-salinity rivers (Andrews 1979). According to Farley (1975). drought and accompanying higher salinity facili- tated spreading of the disease in the estuaries in the Chesapeake Bay. Management strategies for MSX and Dermo diseases in the Chesapeake Bay are based on using low-salinity sanctuaries in rivers for the limited oyster industry (Krantz and Jordan 1996). Water temperature cannot be considered to be critical to the ability of MSX to reach epizootic prevalences, because epizootics have been reported in the northeast in Massachusetts (Matthiessen et al. 1990), but not in MSX-positive areas south of the Chesapeake Bay, such as North and South Carolina, Georgia, or Florida (Haskin and Andrews 1988). ♦ ■ • -. v -* J - • . I ■ ••'• Figure 4. Phagocytosis of Haplosporidium nelsoni spores by hemocytes (arrow). Paraffin section, Ziehl and Harris' hematoxylin. l.OOOx. t Figure 5. Endomitosis of Haplosporidium nelsoni plasmodia (arrow). Paraffin section, H & E. 5(Mx. 174 SUNILA ET AL. Classification of different MSX stages was adopted from Far- ley (1968). because it has clinical implications and gives oppor- tunity for predicting mortality rates and expected life spans in the field, which is crucial for management purposes. Distribution of stages with serious infections decreasing at the end of a year (Ford and Haskin 1982. Farley 1975) was not apparent in this set of samples. On the contrary, infections seemed to rage through the oyster populations over the entire winter. Appearance of plasmodia underwent transformation from endomitotic "Kernstab" stages (Fig. 5) (Farley 1967) into passive, smaller plasmodia in some samples (the Clinton Hammonassett. and Hammock Rivers), but continued multiplication in some samples, such as Branford 179. The number of specimens with histological characteristics with potentially resistant animals was low, 4 of 653 animals (<\%). This might be attributable to the recent outset of Long Island Sound epizootics. Farley (1968) reported an increase of oysters with characteristics for resistance after 1.5 years or more of expo- sure to the disease. Haskin and Ford (1979) reported increasing survival with each year class in naturally resistant oysters in the beginning of the epizootics in the Delaware Bay. During the first year of exposure in 1957, oyster set had $4% mortality, 1958 set had 48% mortality, and 1959 set had 29% mortality. Sporulation rate (three cases of sporulation in 319 cases of MSX-positive specimens) in this article is higher than usually reported. Andrews ( 1979) reported fewer than one case of sporu- lation per 2,000 cases of MSX. Unusual sporulation in the diges- tive duct cells and in connective tissue (Figs. 2 and 3) was prob- ably caused by Haplosporidium nelsoni and not H. costale (SSO, Seaside Organism), because the spore size was 6x8 u.m and not 3x5 u.m. as reported for H. costale (Wood and Andrews 1962). Plasmodia in the specimen had size and characteristic capped nu- clei of H. nelsoni. Furthermore. H. costale sporulates in May to June (Andrews 1979), and the specimen in this article was col- lected in September. H. costale has not been reported in field samples collected from Long Island Sound. Connecticut. However. SSO-like haplosporidians were reported in oysters transplanted from New Haven, Connecticut to Tomales Bay, California in 1967 (Katkansky and Warner 1970). The Bureau of Aquaculture will monitor the prevalence of MSX annually. Long Island Sound experienced a MSX-epizootic in 1985 (Haskin and Andrews 1988). which was reflected in a collapse in production figures. After that. Connecticut's produc- tion rose from 69.721 bushels in 1987 to 893,964 bushels in 1992 (DEP 1995). At the same time (1987), state and private oyster companies started a massive culch program in Connecticut. This prevented habitat loss, which has drastically changed the ecosys- tem in other MSX epizootic areas. No other MSX enzootic areas, Chesapeake or Delaware Bay, have reported total recovery of oys- ter stocks. We propose the following hypothesis for the transmission of MSX. Cyclic patterns have been observed in disease activity in the Delaware Bay. with peaks every 6 to 8 years (Ford and Haskin 1982). Similar fluctuations have been observed in the Chesapeake Bay (Farley 1975). Cyclic patterns in stock densities with several year intervals are characteristic for several estuarine species. The alternate host may be such a species. According to our hypothesis. MSX is transmitted via an alternate host, which is restricted to the mid-Atlantic states. The alternate host's northern limit of distribu- tion is Long Island Sound, and periodically it establishes itself there and then withdraws south to the Chesapeake and Delaware Bays. If the present epizootic in Long Island Sound will pass by, as it did after 1985, this hypothesis is defensible. LITERATURE CITED Andrews. J. D. 1979. Oyster disease in Chesapeake Bay. Mar. Fish. Rev. 41:45-53. Andrews. J. D. & J. L. Wood. 1967. Oyster mortality studies in Virginia. VI. History and distribution of Minchinia mis, mi. a pathogen of oys- ters, in Virginia. Chesapeake Sci. 8:1-13. DEP 1995. State of Connecticut. Department of Environmental Protection, Marine Fisheries Unit. Annual Canvas Catch Data for Commercial Fisheries Landings. Farley. C. A. 1965. Acid-fast staining of haplosporidian spores in relation to oyster pathology. J. Invert. Pathol. 7:144-147. Farley. C. A. 1967. A proposed life cycle of Minchinia nelsoni (Hap- losporida. Haplosporidiidae) in the American Oyster Crassostrea vir- ginica. J. Protozool. 14:616-625. Farley. C. A. 1968. Minchinia nelsoni (Haplosporida) disease syndrome in the American Oyster Crassostrea virginica. J. Protozool. 15:585-599. Farley, C. A. 1975. Epizootic and enzootic aspects of Minchinia nelsoni (Haplosporida) disease in Maryland oysters. J. Protozool. 23:418—127. Ford, S. E. & H. H. Haskin. 1982. History and epizootiology of Haplospo- ridium nelsoni (MSX). an oyster pathogen in Delaware Bay, 1957- 1980. J. Invert. Pathol. 40:118-141. Haskin, H. H. & S. E. Ford. 1979. Development of resistance to Minchinia nelsoni (MSX I mortality in laboratory-reared and native oyster stocks in Delaware Bay. Mar. Fish. Rev. 41:54-63. Haskin. H. H., Stauber. L. A. & J. A. Mackin. 1966. Minchinia nelsoni n. sp. (Haplosporida, Haplosporidiidae): causative agent of the Delaware Bay oyster epizootic. Science 153:1414-1416. 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. Special Publ. 18:5-22. Katkansky, S. C. & R. W. Warner. 1970. The occurrence of a haplospori- dan in Tomales Bay. California. J. Invert. Pathol. 16:144 Krantz. G. E. & S. J. Jordan. 1996. Management alternatives for protecting Crassostrea virginica fisheries in Perkinsus marinus enzootic and epi- zootic areas. J. Shellfish Res. 15:167-176. Matthiesen. G. C. Feng. S. Y. & L. Leibovitz. 1990. Patterns of MSX (Haplosporidium nelsoni) infection and subsequent mortality in resis- tant and susceptible strains of the eastern oyster Crassostrea virginica (Gmelin. 1791). in New England. J. Shellfish Res. 9:359-365. National Shellfish Sanitation Program. Interstate Shellfish Sanitation Con- ference. 1997. Guide for the Control of Molluscan Shellfish. U.S. De- partment of Health and Human Services. Public Health Service. Food and Drug Administration. 406 pp. Newman. M. W. 1971. A parasite and disease survey of Connecticut oys- ters. Nat. Shellfish Assoc. 61:59-63. Ray, S. M. 1954. Biological studies of Dermocystidium marinum, a fungus parasite of oysters. The Rue Institute Pamphlet. Special Issue: 1-114. Wood, J. L. & J. D. Andrews. 1962. Haplosporidium costale (Sporozoa) associated with the disease of Virginia oysters. Science 136:710-71 1. Journal of Shellfish Research, Vol. IS. No. 1. 175-174, 1999. DISTRIBUTION OF THE TURBELLARIAN URASTOMA CYPRINAE ON THE GILLS OF THE EASTERN OYSTER CRASSOSTREA VIRGINICA NICOLE T. BRUN,1 ANDREW D. BOGHEN,1 AND JACQUES ALLARD2 Departement de Biologic Universite de Moncton, Moiuton Nouveau-Brunswick, Canada E1A 3E9 'Departement tie Mathematiques et Statistiques, Universite de Moncton, Moncton, Nouveau-Brunswick, Canada E1A 3E9 ABSTRACT Urastoma cyprinae has been reported from the gills of various bivalve species. In Atlantic Canada, it has been identified in the eastern oyster Crassostreu virginica. Recent studies have demonstrated that U. cyprinae can cause serious alterations to gill tissue in mussels. It has been suggested that, in oysters, the worm feeds on mucus secreted by the gills. Our work has revealed that U. cyprinae is attracted to oysters and that the source of attraction is mucus. The current study focuses on the distribution of U. cyprinae on the gills of oysters in relation to the presence and the state of mucus from different regions of the gills. Findings demonstrate that the worms occur throughout the gill surface, but that they are most heavily concentrated along the basal food tract, the major pathway for the transport of food particles to the labial palps. If mucus plays an important role in the parasite's survival, then U. cyprinae's preference for this location may be a way of ensuring intimate contact between itself and the mucus slurry characterizing this site. KEY WORDS: Urastoma cyprinae, turbellaria. eastern oyster. Crassostrea virginica, gills, oyster mucus INTRODUCTION MATERIALS AND METHODS Urastoma cyprinae Gruff occurs as a free-living organism in mud and algae (Marcus 1951, Westblad 1955) and on the gills of various molluscan bivalves (Fleming et al. 1981, Goggin and Can- non 1989, Noury-Srai'ri et al. 1990. Murina and Solonchenko 1991. Teias dos Santos and Coimbra 1995, Trotti et al. 1998). In Atlantic Canada, the "gill-worm" has been reported in the eastern oyster Crassostrea virginica Gmelin (Burt and Drinnan 1968, Fleming et al. 1981, Boghen et al. 1993). Robledo et al. (1994) have shown that U. cyprinae can in- duce pathological changes in the mussel Mytilus galloprovin- cialis Lamarck, resulting in disorganization of gill filaments and infiltration of hemocytes. Brun et al. (in press) have demon- strated that U. cyprinae is strongly attracted to the mucus that coats the gills of oysters, and Fleming (1986) has suggested that the worm may actively feed on this substance. Such findings bring into question to what extent U. cyprinae is a facultative commensal, as has previously been suggested (Burt and Drinnan 1968). The role of mucus in particle processing of bivalves has been studied extensively (Ward et al. 1993. Ward et al. 1994, Ward 1996. Beninger and St-Jean 1997). The attraction of U. cyprinae to mucus may depend upon such factors as its viscosity, composition, and transport velocity, which vary significantly between different areas of the gills (Ward et al. 1993. Ward 1996). Considering the affinity of U. cyprinae to mucus (Brun et al. in press) and the possibility that the worm feeds on this substance (Fleming 1986). a comparative study of its attraction to different regions of the gills where mucus is present in a form that is most readily accessible becomes an interesting consideration. In this paper, we examine the distribution of U. cyprinae on the gills of oysters in relation to some of the characteristics of the mucus on the different parts of the siills. as mentioned above. Twenty adult oysters infected with U. cyprinae were collected from Shippagan Bay (New Brunswick, Canada) on October 8, 1997. The animals were transported to the Universite de Moncton on ice and were kept at 6°C in a walk-in cold room for 2 days before the start of the study. Of the 20 molluscs collected, five were randomly chosen, and their right valves were removed (Fig. 1). The animals were indi- vidually dipped into liquid nitrogen (-196°C) for 30 s, which allowed immediate sacrifice of both oysters and worms. At the same time, it immobilized U. cyprinae in their respective positions on the gills, thereby facilitating worm counts at the specific loca- tion where they occurred. The oyster lamellae were divided into three zones: ventral, medial, and dorsal, from which U. cyprinae were identified and counted (Fig. 2). These zones generally correspond to the gill regions of bivalves occasionally referred to by other authors (Ward et al. 1993, Ward et al. 1994. Ward 1996. Beninger and St-Jean 1997). The surface area occupied by each of the three zones for each lamella was determined by measuring its length and width (Fig. 2). Based on these measures, boundaries were delineated so that the dorsal zone, representing the area beginning at the basal food tract and extending toward the ventral axis of the oyster, would occupy 10% of the surface of the lamella. The medial zone, commencing at the lower limit of the dorsal zone toward the ven- tral axis, represented 80% of the surface of the lamella, and the ventral zone, stretching from the lower limit of the medial zone to the marginal indentation of the marginal food groove, occupied the remaining 10% (Fig. 2). Both the ascending and the descending sides of the basal and ventral tracts were examined and treated as separate zones for each of the lamellae. The combined surface areas, therefore, occupied by all eight ventral, eight medial, and eight dorsal zones were 10. 80. and 10%. respectively, for each of 175 dorsal axis left valve right mantle left mantle ventral axis Figure 1. Schematic diagram of the oyster, after removal of the right valve, displaying the dorsal and ventral axes, left mantle, and raised right mantle exposing the gills. the five oysters. In our study, the mantle area was not used for determining the counting zones or for the calculation of worm numbers. Each oyster was partially thawed and examined in a semi- frozen state using an Olympus SZ30 stereomicroscope. This condition ensured minimum dislodging of the worms from their point of attachment on the gills during examination. Worm counts were conducted on successive demibranchs from each of the zones as defined above, using a hand-held counter, starting with the demibranch immediately adjacent to the right mantle (Fig. 1). A log-linear model (Agresti 1990) was employed using Systat8 8.0 for Windows® (SPSS 1998) to analyze the worm counts as a function of three factors: oysters, zones, and lamellae, and their interactions. This analysis allows the user to identify signifi- cant factors and estimate their importance with an adequate con- trol of Type 1 error. The alternative would be for the user to employ numerous chi-square tests on separate contingency dorsal axis Imfg ventral axis I 1 dorsal zone (10%) HO medial zone (80%) | ventral zone (10%) Figure 2. Schematic diagram of the four gill demibranchs of an oyster displaying the ascending and descending lamellae. Fnlargement of area in rectangle to show details of one of the two lamellae forming an oyster demibranch. Each lamella is divided into three zones from which Urastoma cyprinae were counted: ventral, medial and dorsal. The shaded areas, with corresponding percentages, represent the sur- face area of the lamella occupied by each zone (10. 80, and 10%, respectively), (a) ascending lamella (bftl basal food tract, (d) descend- ing lamella, (dbl) demibranch I, ldb2) demibranch 2, Idb3) demi- branch 3, (db4) demibranch 4. (mfg) marginal food groove, (mi) mar- ginal indentation. E •— o B 90 80- 70 60 5 0 4 0 30 20 10 0 ventral i q * i 9 □ medial v ] dorsal adadadadadadadadadadadad dbl dh2 db3 db4 dbl db2 db3 db4 dbl db2 db3 db4 P osition on the gills Figure 3. Box and whiskers plot representing the number of worms observed in oysters for each of the counting zones (ventral, medial, dorsal) of the ascending and descending lamellae for the four demibranchs. Due to the small number of oysters sampled, it is possible to identify single counts in most cases from the box and whiskers plot. Distribution of Urastoma cyprinae on Gills of Oysters 177 tables, but with less control on the over-all probability of error. For factors found to be significant in the log-linear model, the re- sults of chi-square tests are given as a simple measure of signifi- cance. In the log-linear model, the Pearson chi-square and the likeli- hood-ratio test are used to identify the most significant factors and Raftery's BIC (Bayesian Information Criterion) is employed to select the most appropriate model. The model is reported in a multiplicative form as follows: Fljk = Fn m, m. mk mM mlk m1|k. where i represents the oysters, j the zones, and k the lamellae. F0 is the reference frequency, and m represents the effects. An esti- mated m value close to 1 suggests that the effect of the individual factor or the interaction between two factors is small. The standard scores [lambda/sedambdal] for the parameters of the additive ex- pression of the model (ln Fl)k = H + \, + hf + \k + Ai( + Aik + \|k) were calculated and used with the Bonferroni correction test to determine their significance. However, for the sake of concise- ness, they are only reported as required for analysis. For each position, the counts for all oysters were added. These total counts were arranged in a simple two-way table (zones x lamellae) and analyzed using the chi-square test (Zar 1984). To make com- parison with other studies easier, mean counts per oyster are re- ported. RESULTS Grouped box and whisker plots (Fig. 3) represent the observed worm counts graphically. The level of infection/parasitism for the five individual oysters are 513, 514. 507. 279. and 466. respec- tively. The results of the log-linear model indicate that the test of fit is significant for both the Pearson chi-square (x2 = 189.3735, df = 56, p < .000005) and the likelihood-ratio test (x" = 199.4746. df = 56. p < .000005). A triple interaction factor m,|k was removed, because it was nonsignificant. The model retained, therefore, is F,jk = F„ m, mp mk m,, mlk m|k with a Raftery's BIC value of -233.4889. Table 1 shows that the three factors and their respective inter- actions are significant. The variable having the greatest impact on worm numbers is the zones (\2 = 1208.88). When considering the multiplicative effects for the latter, the results in Table 2 in- dicate that the worms are more abundant in the dorsal zones of the gills (2.688) as compared to the medial (1.345) and ventral (0.276) zones, respectively. The standard scores for the additive expression of the model show that each value within the zones is highly significant: ventral (-20.054); medial (6.881); and dorsal (25.246). In most instances, the multiplicative effects of the other two factors as well as all the interactions are close to 1, except for TABLE 1. Results of the log-linear model analysis. Removal of Term Model Without the Term from Model Term Tested Ln (MLE) Chi-sq P- df value Chi-sq df P- value Oysters -388.580 275.72 60 0.0000 76.25 4 0.0000 Zones -855.156 1208.88 58 0.0000 1009.40 2 0.0000 Lamellae -363.095 224.75 63 0.0000 25.28 7 0.0007 Zones-oysters -377.011 252.59 64 0.0000 53.11 8 0.0000 Lamellae-oysters -419.180 336.92 84 0.0000 137.45 28 0.0000 Zones-lamellae -441.028 380.62 70 0.0000 181.15 14 0.0000 a few values (Table 2) that represent particularly low or high worm counts. Table 3 shows the mean number of worms observed and ex- pected for each position. The results of the chi-square test com- paring the total (5 x mean number) number of worms observed with the expected number of worms observed at each position indicate that U. cyprinae are more numerous along the dorsal zones as compared to the medial and ventral zones. This is re- flected by the total number of worms counted in each of the three TABLE 2. Multiplieative effects for each of the three factors and their interactions. b. 1.257 1.355 1 .049 0.571 0.981 Ventral Medial Dorsal 0.276 1.345 mk 2.688 la Id 2a 2d 3a 3d 4a 4d 0.974 0.888 0.912 1.200 1.018 m,. 1.346 1.060 0.727 1 2 3 4 5 Ventral 1.009 1.264 0.800 0.987 0.992 Medial 1.122 1.134 1.099 0.747 0.958 Dorsal 0.883 0.698 1.137 1.357 1 .053 la 0.804 0.981 0.644 1 .334 1.478 Id 0.851 1.209 1.127 0.736 1.171 2a 1.330 1.499 0.731 1.006 0.682 2d 0.825 0.609 1.229 1.510 1.073 3a 0.990 1.062 1.165 1.016 0.810 3d 0.963 1.151 0.867 0.688 1.150 4a 1.183 0.981 0.938 0.894 1.029 4d 1 181 0.771 1.631 1.073 0.627 1 3 la Id 2a 2d 3a 3d 4a 4d 0.950 0.939 1.037 1 .304 1.369 0.888 1.029 0.640 1.266 0.567 1.382 0.687 1.175 0.821 1.163 1.307 0.831 1.878 0.675 1.116 0.621 1.371 0.835 1.196 m represents the effects, i the oysters, j the zones, and k the lamellae. The calculated reference frequency (F0) is valued at 1 1.516. Values significantly different from 1 using the Bonferroni correction test are indicated in boldface. a. Multiplicative effects for each of the five oysters, h. Multiplicative effects for the three counting zones, c. Multiplicative effects for the four ascending (a) and four descending (d) lamellae, d. Multiplicative effects for the interactions between the oysters and the zones, e. Multiplicative effects for the interactions between the oysters and the lamellae, f. Multiplicative effects for the interactions between the zones and the lamellae. 178 Brun et al. TABLE 3. Observed and expected mean number of worms for each of the three zones (ventral, medial, dorsal), and the differences in percentages between them for the ascending (a) and descending (d) lamellae for the four demibranchs (db). Negative values indicate an observed worm count lower than expected; positive values demonstrate a higher one. Zones Ventral Medial Dorsal Observed Expected difference Observed Expected difference Observed Expected difference Lamellae mean ± SEa mean (%) mean ± SE" mean (%) mean ± SEa mean (%) db 1 a 3.2+ 1.0 5.6 -43 20.4 + 6.7 45.6 -55 25.8 ± 3.9 5.6 +361 db 1 d 3.0 ± 0.8 5.6 -46 9.0 ±2.7 45.6 -80 53.8 ± 6.5 5.6 +861 db 2 a 5.0 ± 0.9 5.6 -11 13.2 ±2.1 45.6 -71 43.4 ± 8.6 5.6 +257 db 2 d 3.8± 1.1 5.6 -32 23.8 + 10.0 45.6 -48 20.0 + 2.8 5.6 +675 db 3 a 4.8 ± 1.9 5.6 -14 20.6 + 5.8 45.6 -55 20.0 ± 4.0 5.6 +257 db 3 d 4.4+ 1.6 5.6 -21 20.0 ±5.1 45.6 -56 60.4 ± 8.8 5.6 +979 db4 a 1.6 ±0.7 5.6 -71 17.0 ±5.3 45.6 -63 29.4 + 5.2 5.6 +425 db4d 3.8 ± 1.5 3.7 ± 0.4 5.6 -32 21.4 + 3.7 18.2 ± 2.0 45.6 -53 28.0 ±5.8 35.1 ± 3.0 5.6 +400 ' Standard error on the mean. zones: dorsal (1404). medial (727). ventral (148). Differences in percentages between the observed and expected worm counts fur- ther underline the preponderance of U. cyprinae in the dorsal versus the medial and ventral zones (Table 3). DISCUSSION Previous studies have shown that V. cyprinae are highly at- tracted to mucus on the gills of oysters (Brun et al. in press). Mucus is secreted by pallial organs and plays a predominant role in activities associated with bivalve-suspension feeding (Ward et al. 1994. Ward 1996). The use of mapping for distinguishing between particular re- gions of pallial surfaces has previous been applied in studies on mucocyte identification in bivalves (Bellinger et al. 1993, Be- ninger and St-Jean 1997). The current work employed the principle of gill mapping to determine whether there existed a relationship between infestation levels of U. cyprinae and specific areas on the gills. This approach contrasts with previous investigations on oys- ters (Fleming et al. 1981, Fleming 1986) and such other species as mussels (Robledo et al. 1994, Teias dos Santos and Coimbra 1995, Trotti et al. 1998). which reported the occurrence of the parasite in the host without specific reference to a niche. Statistical analyses of three variables (oysters, zones, lamellae) and the interactions between them, demonstrated that it was the zones that exerted the greatest influence on the distribution of U. cyprinae. It was also noted that the highest numbers of worms were consistently found along the dorsal regions as compared to the medial and ventral zones. Having already established that U. cyprinae display a strong attraction for mucus (Brun et al. in press), the existing situation raises the question as to why there is such a marked preference for the dorsal zones. Several possible explanations could be proposed. Mucus contained in the basal food tract consists primarily of mixed mucopolysaccharides (acid and neutral); whereas, the mar- ginal food groove is mostly composed of acid-dominant muco- polysaccharides (Beninger and St-Jean 1997). As a result, trapped particles in the dorsal regions are transported in a low-viscous mucus slurry (Ward 1996), at a velocity of more than twice that occurring in the ventral grooves (Ward et al. 1993). This differs from the material in the marginal food grooves, which is contained in a viscous mucus cord (Ward et al. 1994). The likelihood that the worms prefer to be in contact and possibility feed on the more "fluid" mucus in the dorsal zones is an attractive prospect, both from the point of accessibility and ease of uptake. Urastoma cyprinae possesses an oral-genital pore located in the posterior end (Cannon 1986). and a tegument that takes the form of a ciliated epithelium (Schmidt and Roberts 1989, Noury- Srai'ri et al. 1990). Although U. cyprinae obviously feeds with its oral-genital pore, the tegument may also play an important role in food acquisition and the transfer of material across the body wall (Noury-Srai'ri et al. 1990). The dorsal regions may. therefore, rep- resent choice sites where extensive contact between the body of U. cyprinae and the fast-moving, "fluid" mucus slurry prevails. In the oyster, the basal food tract is the predominant route for transportation of food in preparation for final sorting and ingestion (Ward et al. 1998). In contrast, mucus-bound particles contained in the marginal grooves may have a lower nutritional value (Ward et al. 1998) and are eventually rejected as pseudofeces (Newell and Langdon 1996, Ward 1996). If we assume that the worms do not feed solely on the mucus, but also rely on particles bound in the mucus, then U. cyprinae' s preference for the dorsal regions, where high-quality nutritional materials are present, can be expected. The dorsal regions may also provide added protection for the worms, because the latter would be less vulnerable to water currents when the oyster's valves are open. Finally, given the fact that U. cypri- nae are negatively phototactic (Burt and Bance 1981, Pike and Wink 1986). the dorsal regions may also provide the worms with a darker environment as compared to the ventral regions. Morpho- logical and behavioral studies are presently under way to under- stand the parasite's adaptations and possible host-parasite interac- tions better in the dorsal regions of the oyster's gills. ACKNOWLEDGMENTS We are grateful to Dr. J. E. Ward (Department of Marine Sci- ences, University of Connecticut. CT. USA). Dr. S. Reebs (De- partement de Biologie. Universite de Moncton. Moncton, NB. Canada), Dr. M. D. B. Burt (Huntsman Marine Science Centre, St. Andrews. NB. Canada), and Dr. B. A. MacDonald (Department of Biology. University of New Brunswick. Saint John. NB, Canada), for reviewing our manuscript and for their helpful comments and Distribution of Urastoma cyprinae on Gills of Oysters 179 suggestions. The authors also thank Dr. P. Ashrit (Department de Physique, Universite de Moneton. Moncton. NB. Canada) for all his helpful comments. Financial support to the senior author was provided in part by the Faculty of Research and Graduate Studies of the Universite de Moncton. This project is part of the Richibucto Environmental and Resource Enhancement Program. LITERATURE CITED Agresti A. 1990. Categorical Data Analysis. Wiley-Interscience, New York. 558 pp. Beninger, P. G. & S. D. St-Jean. 1997. The role of mucus in particle processing by suspension-feeding bivalves: unifying principles. Mar. Biol. 129:389-397. Beninger, P. G.. S. St-Jean, Y. Poussart & E. Ward. 1993. Gill function and mucocyte distribution in Placopecten magellanicus and Mytilus edulis (Mollusca: Bivalvia): the role of mucus in particle transport. Mar. Ecol. Prog. Ser. 98:275-282. Boghen. A. D., J. Allard & E. Bataller. 1993. Rapport final et recommen- dations sur le programme de monitoring pour la cote est du Nouveau- Brunswick. Centre de recherche en sciences de l'environnement, Uni- versite de Moncton. Moncton. Nouveau-Brunswick. 76 pp. Brun. N. T. B.. A. D. Boghen & J. Allard. in press. Attraction of Urastoma cyprinae Graff. 1882 (Turbellaria: Urastomidae) to the eastern oyster Crassostrea virginica. Dis. Aquat. Org. Burt. M. D. B. & G. W. Bance. 1981. Ultrastructure of the eye of Urastoma cyprinae (Turbellaria. Alloeocoela). Hydrobiologia 84:276. Burt, M. D. B. & R. E. Drinnan. 1968. A microturbellarian found in oysters off the coast of Prince Edward Island. J. Fish. Res. Bd. Can. 25:2521- 2522. Cannon. L. R. G. 1986. Turbellaria of the World — A Guide to Families & Genera. Queensland Museum, Brisbane. Australia. 132 pp. Fleming. L. C. 1986. Occurrence of symbiotic turbellarians in the oyster, Crassostrea virginica. Hydrobiologia 132:311-315. Fleming. L. C. M. D. B. Burt & G. B. Bacon. 1981. On some commensal turbellaria of the Canadian east coast. Hydrobiologia 84:131-137. Goggin. C. L. & L. R. G. Cannon. 1989. Occurrence of a turbellarian from Australian tridacnid clams. Int. J. Parasitol. 19:345-346. Marcus. E. 1951. Turbellaria Brasileiros (9). Boletins da faculdade filoso- fia ciencias e letras, Universidade Sao Paolo. Zoologia 16:5-215. Murina. G. V. & A. I. Solonchenko. 1991. Commensals of Mytilus gallo- provincialis in the Black Sea: Urastoma cyprinae (Turbellarial and Polydora ciliata (Polychaeta). Hydrobiologia 227:385-387. Newell. R. I. E. & C. J. Langdon. 1996. Mechanisms and physiology of larval and adult feeding, pp. 185-229. In: V. S. Kennedy, R. I. E. New- ell and A. F. Eble (eds.). Maryland Sea Grant College, College Park. Maryland. Noury-Srairi, N.. J. L. Justine & L. Euzet. 1990. Ultrastructure du tegument et des glandes sous-epitheliales de Urastoma cyprinae ("Prolecitho- phora"), Turbellarie parasite de Mollusque. Annales des Sciences Na- turelles, Zoologie. 13e Serie 11:53-71. Pike, A. W. & R. Wink. 1986. Aspects of photoreceptor structure and phototactic behavior in Platyhelnnnthes. with particular reference to the symbiotic turbellarian. Paravortex, Hydrobiologia 132:101-104. Robledo. J. A. F.. J. Caceres-Martinez. R. Sluys & A. Figueras. 1994. The parasitic turbellarian Urastoma cyprinae (Platyhelminthes: Urastomi- dae) from blue mussel Mytilus galloprovincialis in Spain: occurrence and pathology. Dis. Aquat. Org. 18:203-210. Schmidt, G. D. & L. S. Roberts. 1989. Foundations of Parasitology. Time Mirror/Mosby College Publishing. St. Louis. MO. 750 pp. SPSS Inc. 1998. Systat® 8.0 for Windows®: new statistics. SPSS Inc.. Chicago. 1086 pp. Teias dos Santos. A. M.. & J. Coimbra. 1995. Growth and production of raft-cultured Mytilus edulis L.. in Ria de Aveiro: gonad symbiotic infestation. Aquaculture 132:195-211. Trotti. G. C. E. M. Baccarani. S. Giannetto. A. Guiffrida & F. Paesanti. 1998. Prevalence of Myticola intestinalis (Copepoda: Mytilicolidae) and Urastoma cyprinae (Turbellaria: Hypotrichinidae) in marketable mussels Mytilus galloprovincialis in Italy. Dis. Aquat. Org. 32:145- 149. Ward, J. E. 1996. Biodynamics of suspension-feeding in adult bivalve molluscs: particle capture, processing, and fate. J. Invert. Biol. 115: 218-231. Ward. J.E., J. S. Levinton, S. E. Shumway & T. Cucci. 1998. Particle sorting in bivalves: in vivo determination of the pallial organs of se- lection. Mar. Biol. 131:283-292. Ward, J. E.. B. A. MacDonald & R. J. Thompson. 1993. Mechanisms of suspension feeding in bivalves: resolution of current controversies by means of endoscopy. Limnol. Oceanogr. 38:265-272. Ward, J. E„ R. I. E. Newell. R. J. Thompson & B. A. MacDonald. 1994. In vivo studies of suspension-feeding processes in the eastern oyster, Crassostrea virginica (Gmelin). Biol. Bull. 186:221-240. Westblad. E. 1955. Marine "Alloeocoels" (Turbellaria) from North Atlan- tic and Mediterranean coast. Arkiv for Zoologi 7:491-526. Zar, J. H. 1984. Biostatistical Analysis. 2nd ed. Prentice-Hall. Inc.. Engle- wood Cliffs. NJ. 718 pp. Journal of Shellfish Research, Vol. IS. No. 1. 181-183. 1999. FECUNDITY OF THE VELVET SPIDERCRAB STENOCIONOPS O VATA (BELL, 1835) (BRACHYURA: MAJIDAE) IN THE GULF OF CALIFORNIA, MEXICO MARCIAL VILLALEJO-FUERTE, BERTHA PATRICIA CEBALLOS-VAZQUEZ, MARCIAL ARELLANO-MARTINEZ, AND FEDERICO GARCIA-DOMINGUEZ CICIMAR-IPN, La Paz. B.C.S.. Mexico 23000 ABSTRACT The fecundity of the velvet spidercrab Stenocionops ovata from the Gulf of California was determined based on an incidental capture realized in the Gulf of California. Evidence suggested that adult females of 5. ovata can produce more than one batch of eggs in a breeding cycle. The number of eggs per brood by S. ovata range from 35. 1 87 eggs to 1 70.057 eggs with a mean of about 94.200 eggs. This species showed lower fecundity than other Majidae of deep water (e.g., Maiopsis panamensis) from the Gulf of California. Partial fecundity was positively correlated with both length of carapace without rostral horns (r = 0.655) and branchial width of carapace (r = 0.617), without a clear tendency of one measure to be a more useful morphometric measure in the determination of the relation with fecundity in this species. KEY WORDS: Fecundity, spidercrab, Majidae, Stenocionops, Gulf of California INTRODUCTION Stenocionops ovata (Bell 1835) has been captured from 15 to 275 m of depth over sandy-lime bottoms or over coral, shell, seaweed, or sponge substrates (Hendrickx 1995). According to Hendrickx (1989) this species is represented by abundant popula- tions in the central Gulf of California, between 60 and 100 m depth, where they are accessible to commercial drag nets. The great number of 5. ovata obtained by incidental capture by one trawler at Isla Danzante supports this idea. Therefore, great popu- lations exist in the gulf that may be commercially exploited. This species represents a potential resource not currently being ex- ploited, which merits further study (Hendrickx 1995). Studies on fecundity in a population with economic potential are of special interest because of its relation with the renewal intensity of the population. Moreover, it is a basic to our knowl- edge of the reproductive strategy and evolution of the population from what it is an essential component in the population dynamic (Garcta-Montes et al. 1987). Fecundity is an import parameter in crustaceans, among others, in determining the reproductive potential of a species and/or the stock size of a population (Medina and Fransozo 1997). Among brachyuran crabs, there is considerable variation in fecundity (Me- dina and Fransozo 1997). Body size of female brachyurans is the principal determinant of fecundity per brood and reproductive out- put (brood mass) (Hines 1988). The fecundity of crabs in deep water of the Gulf of California has not been determined, except for Maiopsis panamensis (Faxon 1893, Villalejo-Fuerte et al. 1998). In this work, the estimation of fecundity was carried out from the ovigerous mass. The intention of this paper is to present the first data about reproductive effort and fecundity of 5. ovata in the Gulf of California and their relation to the carapace length and width. We also present a comparison with the results of fecundity obtained for another Majidae of deep water (M. panamensis) from the Gulf of California. MATERIALS AND METHODS Three hundred ninety-one specimens of velvet spidercrabs S. ovata were incidentally captured in the northern extreme at lsla Danzantes, Gulf of California, Mexico (25°48'54"N and 1 1 1°15'45"W) in October 1994. Captures were made with gill nets of monofilament with 6-in mesh size at 180-m depth. Velvet spi- dercrabs obtained were identified with the Keys of Garth ( 1958). For each ovigerous female, length of carapace (mm) without rostral homs (CD. branchial width of carapace (CW). total weight, and the weight of the brood were registered. The pleopod structure with the egg mass was removed immediately after capture and was preserved in a 10% buffered solution of formaldehyde prepared with sea water. The eggs were removed from the pleopods by means of the method described by Choy ( 1985). They were stored in 7% formalin until they were processed. For the fecundity estimation, only 25 ovigerous females were considered, according to the criterion of Medina and Fransozo (1997). The fecundity estimation, considered in this paper as the number of eggs per brood, was done by means of the gravimetric method described by Bagenal (1978). The number of eggs per brood was determined by direct count- ing of samples of 0.1 g. The relative fecundity was determined as the number of eggs per gram of total weight. The reproductive effort had been estimated with the ratio of brood weight per body weight, because it has often been used as an index of current reproductive effort (Fukui and Wada 1986). To determine the re- lation of partial fecundity with length of carapace without rostral horns (CL) and branchial width of carapace (CW), dates were fitted to four models. Gonads were removed and fixed in a neutral formalin solution 10%. Gonad sections were taken and dehydrated in alcohol, embedded in paraffin, and sectioned at 7 u.m. Sections were placed on slides and stained with hematoxylin-eosin (Huma- son 1979). RESULTS From the 391 velvet spidercrabs obtained. 304 (77.7%) were males, and 87 were females (23.3%). The sex ratio for the total sample was 3.5 M: 1 F and differed significantly (p s.05) from the expected ratio of 1:1 . All the 87 females obtained were ovigerous. Ovigerous females reach 71.5-102.5 mm CL (90.1 mm mean. 7.5 SD) and 65.6-88.5 mm CW (76.6 mm mean. 5.84 SD). The ratio of brood weight to body weight ranged from 0.0375 to 0.1 1 12 (0.0835 mean, 0.0193 SD) and was not correlated with CW or CL (p > .05). Relative fecundity ranged from 236 eggs/g (90 mm CL. 82 mm CW) to 972 eggs/g (86 mm CL. 75 mm CW) with a mean of 587 eggs/g. Partial fecundity ranged from 35,187 eggs (90 mm CL. 82 mm 181 182 VlLLALEJO-FUERTE ET AL. TABLE 1. Results of the fits between partial fecundity with carapace length and carapace width. Model Geometric Exponential Linear Logarithmic With Carapace Length With Carapace Width Y = 5530.98XOOO69X r2 = 0.5361 Y = 2993.42e00377X r = 0.5341 Y = -178093.82 + 3018.66X r = 0.4904 Y = -1062616.4 + 257173. 25LnX r = 0.4736 Y = 6057.77X""°8IX r = 0.4599 Y = 3242.57e004,4X r = 0.4587 Y = -197944.2 + 3804.99X r = 0.4364 Y = -1142520 + 285099.48LnX r = 0.4285 CW) to 170.057 eggs ( 102 mm CL. 88.5 mm CW) with a mean of 94,200 eggs. Partial fecundity was significantly correlated (p < .05) with both length of carapace without rostral horns (CL) and branchial width of carapace (CW). demonstrating that fecundity increases with the increase in both length of carapace without rostral horns and branchial width of carapace. The equations of regression obtained for the four models are summarized in Table 1. The best fit between fecundity and CL or CW was with the geometric model (r = 0.5361 and r = 0.04599. respectively) (Figs. 1, 2). Histologically, the ovaries pre- sented mature and previtelogenic oocytes. DISCUSSION Although incidental captures of velvet spidercrabs were real- ized only in October 1994, ovigerous females were present. Ac- cording to Garth ( 1958), in the Gulf of California, the breeding of 5. ovata occurs in April, but the fact that ovigerous females had been encountered in October is evidence that S. ovata have a larger breeding cycle. However, an annual study of the reproductive bi- ology of the species is necessary to give conclusive statements. In addition, there is histologic evidence to suggest that S. ovata females can produce more than one batch of eggs in a breeding cycle. The occurrence of ripening ovaries simultaneous with brooded eggs has been interpreted as an indication of the potential to produce a second brood during the same breeding season (Pillay and Nair 1971). This condition has been reported for the panamic spidercrab M. panamensis (Villalejo-Fuerte et al. 1998), as well as for other brachyuran species (Ryan 1967, Pillay and Nair 1971. Perez 1990) and several anomuran species (Ameyaw-Akumfi 1975. Varadarajan and Subramoniam 1982). Reproductive effort wasn't correlated with CL or CW. but the fecundity was correlated with both CL and CW. Although the correlation between fecundity and CL (r = 0.655) was slightly higher than that obtained between fecundity and CW (r = 0.617). we do not consider that this difference is sufficient to believe that the CL is the more useful morphometric measure in the determi- nation of the relation with fecundity in this species, such as occurs for M. panamensis, in which the CW is the more useful morpho- metric measure in the determination of the relation to fecundity (Villalejo-Fuerte et al. 1998). The lower fecundity observed in 5. ovata (mean of 94,200 eggs) as compared to the sympatric spider- crab M. panamensis (mean of 676.039 eggs), may probably be attributed to size, considering that this species is smaller than M. panamensis. ACKNOWLEDGMENTS Thanks are due to Jorge and David Villalejo and Luis and Manuel Torres for supplying us with the specimens. We are also h -o c O o 18 16 ■■ 14 ■• 12 •■ 10 ■• 8 ■■ 6 ■■ 4 2 70 Y = 5530.98X00069X r2 = 0.5361 h — l — i — h H 1 1 1 1 1 H 75 80 85 90 95 100 105 Carapace length (mm) Observed — Calculated Figure 1. Relationship between partial fecundity and carapace length in ovigerous female Stenoeionops ovata from the Gulf of California. h ■D C O 18 -j- 16 ■■ 14 ■■ 12 ■• 10 •• 8 ■■ 6 ■• 4 ... 2 - 65 Y = 6057.77X00081X r2 = 0.4599 . -+- 70 75 80 85 Carapace width (mm) 90 Observed — Calculated Figure 2. Relationship between partial fecundity and carapace with in ovigerous female Stenoeionops ovata from the Gulf of California. grateful to Direccion de Estudios de Postgrado e Investigacion del Instituto Politeenico Nacional (IPN) for funding this work. We acknowledge the fellowships of Comision de Operacion y Fo- mento de Actividades Academicas del IPN to M. Villalejo-Fuerte and F. Garci'a-Domi'nguez. Fecundity of Stenocionops ovata 183 LITERATURE CITED Ameyaw-Akumfi. C. 1975. The breeding biology of two sympatric species of tropical, intertidal hermit crabs. Clibanarius chapini and C. senega- lensis. Mar. Biol. 29:15-28. Bagenal. T. 1978. Aspects of fish fecundity. In: Methods of Assessment of Fish Production in Fresh Waters. IBP Handbook 3. Blackwell Scien- tific Publications, Ltd.. London. Choy. S. C. 1985. A rapid method for removing and counting eggs from fresh and preserved decapod crustaceans. Aquaculture 48:369-372. Fukui Y. & K. Wada. 1986. Distribution and reproduction of four intertidal crabs (Crustacea. Brachyura) in the Tonda River Estuary. Japan. Mar. Ecol. Prog. Ser. 30:229-241. Garcfa-Montes. J. F.. A. Gracia & L. A. Soto. 1987. Morphometry, relative growth, and fecundity of the Gulf Crab. Callinectes si/nilis Williams. 1966. Ciencias Marinas 13:137-161. Garth, J. S. 1958. Brachyura of the Pacific Coast of America Oxyrhyncha. Allan Hancock Pacific Expeditions 2 1 . Part 1 . VJniversity of California Press, Los Angeles. 499 pp. Hendrickx. M. E. 1989. New distribution and size records of Maiopsis panamensis Faxon and Stenocionops ovata (Bell) (Crustacea: Deca- poda: Majoidea) in the Gulf of California. Mexico. Inv. Mar. CICIMAR 4:285-290. Hendrickx. M. E. 1995. Cangrejos. Majidae. pp. 607-612. In: Gui'a FAO para la Identification de Especies para los Fines de la Pesca. Pacifico Centro-Oriental. vol. I. Plantas e Invertebrados. U.N. Food and Agri- culture Organization, Rome. Hines, A. H. 1988. Fecundity and output in two species of deep-sea crabs. Geryon fenneri and Geryon quinquedens (Decapoda: Brachyura). J. Crust. Biol. 8:557-562. Humason. G. L. 1979. Animal tissue techniques. W. H. Freeman. San Francisco. 661 pp. Medina. M. F. & A. Fransozo. 1997. Fecundity of the crab Callinectes ornatus ordway. 1863 (Decapoda. Brachyura. Portumdae) from the Ubatuba Region. Sao Paulo. Brazil. Crustaceana 70. Perez, O. S. 1990. Reproductive biology of the sandy shore crab Matuta lunaris (Brachyura: Calappidae). Mar. Ecol. Prog. Ser. 59:83-89. Pillay, K. K. & N. B. Nair. 1971. The annual reproductive cycles of Uca annulipes, Portunus pelagicus, and Metapenaeus qffinis (Decapoda: Crustacea) from the southwest coast of India. Mar. Biol. 1 1:152-166. Ryan. E. P. 1967. Structure and function of the reproductive system of the crab Portunus sanguinolentus (Herbst) (Brachyura: Portunidae). II. The female system. Symp. Ser. Mar. Biol. Ass. India 4:151-164. lit: Perez, O. S. 1990. Reproductive biology of the sandy shore crab Matuta lu- naris (Brachyura: Calappidae). Mar. Ecol. Prog. Ser. 59:83-89. Varadarajan. S. & T. Subramoniam. 1982. Reproduction of the continu- ously breeding tropical hermit crab Clibanarius clibanarius. Mar. Ecol. Prog. Ser. 8:197-201. Villalejo-Fuerte. M., M. Arellano-Martinez & B. P. Ceballos-Vazquez. 1998. Fecundity of panamic spidercrab Maiopsis panamensis Faxon, 1893 (Brachyura: Majidae) in the Gulf of California. Mexico. / Shell- fish Res. 17:299-301. Journal of Shellfish Research, Vol. 18. No. 1. 185-192, 1999. DEVELOPMENT OF A RECRUITMENT INDEX FOR FORECASTING SEASONAL LANDINGS OF THE KUWAIT SHRIMP FISHERIES YIMIN YE, J. M. BISHOP, H. MOHAMMED, AND A. H. ALSAFFAR Kuwait Institute for Scientific Research Salmiya, 22017 Kuwait ABSTRACT The seasonal landings and preseason survey data for the Kuwait shrimp fishery were analyzed. Mean catch rates in the preseason surveys in the southern area in July and August might serve as a recruitment index for the fishery. A combined-species model and a single-species model for Penaeus semisulcatus explained 96 and 89% of the variance, respectively. These parametric models together with nonparametnc methods were then employed to forecast the 1997 to 1998 season's shrimp landings before the season opened. The nonparametric models produced better forecasts than the parametric models. Finally, a general discussion about parametric and nonparametric methods for fishery forecasting was presented. KEY WORDS: Recruitment index, shrimp, parametric method. INTRODUCTION Recruitment to a fishery is known to vary greatly between years, sometimes by orders of magnitude, in unpredictable ways. In a single-cohort fishery, such as the Kuwait shrimp fishery, variation in recruitment leads directly to changes in seasonal land- ings. A reliable recruitment index is not only useful for investi- gating stock-recruitment relationships, but also for forecasting seasonal landings. Establishing a stock-recruitment relationship is a top priority of fishery managers, because this relationship pro- vides guidelines for the maximum limits of fishing mortality for sustainable production. Forecasts of annual landings before the season opening are of considerable importance to commercial fish- ermen as well. This information is crucial for decisions concerning the best fishing areas and what their investment strategy should be with regard to fishing equipment. Such a forecast also provides valuable information to the fishery management authority that con- trols the number of licensed boats and results in opportunities for improved biological management, through planning to take advan- tage of strong recruitment events and to protect stocks when poor recruitment is expected (Walters 1989). Morgan and Garcia (1982) developed a recruitment index for the Kuwait shrimp fishery to investigate the possible stock- recruitment relationship. Their recruitment index was calculated by dividing the observed catch during a biological year by the estimated yield-per-recruit for that year. In this calculation of yield-per-recruit, there is an implicit assumption that the fishing effort coefficient is constant. However, a recent study (Ye and Mohammed 1999) found that catchability of shrimp has a high interannual variability in the Kuwait shrimp fishery. This variabil- ity may mask the real variation of recruitment and lead to an error in the stock-recruitment relationship (Garcia 1983). To solve this kind of problem, a recruitment index independent of fishing effort is required. To date, a preseason forecast for the shrimp fishery in Kuwait has not been attempted. In general, there are two types of fisheries forecasting. One involves the traditional parametric method, in- cluding methods of ordination and canonical analysis and univari- ate and multivariate linear, curvilinear, and logistic regressions. The other is the nonparametric method (Rice 1993). Traditional statistical methods have many disadvantages (James and McCul- loch 1990). and nonparametric methods, in which the predicted variable is described by a probability density function, have gained more favor (Evans and Rice 1988, Rice 1993). This study attempts to establish a recruitment index for the nonparametric method, Kuwait Kuwait shrimp fishery that best represents the variation in recruit- ment. Both parametric and nonparametric methods were employed to forecast seasonal landings. Finally, a general discussion about the difference between the parametric and nonparametric is made. MATERIALS AND METHODS Data Used To develop an index independent of fishing effort for the shrimp fishery, survey data are the first choice. Before the Gulf War, the Mariculture and Fisheries Department, Kuwait Institute for Scientific Research (KISR) carried out year-around surveys for 6 years from May 1985 to April 1990. Unfortunately, the computer database was destroyed during the Iraqi occupation. Although a great effort was made by KISR to rescue the original survey record files, some were permanently lost. Because a recruitment index mainly concerns the period before season opening, survey data for the period of April to August from 1985 and 1987 to 1989 were used together with the postwar preseason surveys. After the Gulf War. a preseason survey was initiated in 1993. Because of financial constraints, the preseason surveys were con- ducted for different numbers of months in different years, 4 months for some years and only 2 months for the others. Its survey design was also not consistent. Fixed stations were sampled in 1993 and 1994. but a random sample scheme was introduced af- terwards. Either of the two survey periods. 4 to 6 years depending upon the month for which the data were used, is too short for a rigorous time-series analysis. Therefore, the two surveys were pooled to- gether as a single time series after standardization. The prewar survey used the RV Bahith. a stern trawler (679 tons gross ton- nage) of KISR with one flat trawl of 30-meter foot rope between otter boards and a cod end of 50-mm stretched mesh. The postwar preseason survey employed Mutaheda 5, a Gulf of Mexico-type, double-rigged shrimp trawler leased from the United Fisheries of Kuwait. This boat towed two flat trawls of the same cod-end mesh size and foot rope length as RV Bahith. Because both boats and nets used were similar, a factor of 2 was used to standardize roughly the density index of the prewar survey. Monthly surveys from April to August sampled 10 to 17 sta- tions from Kuwait Bay to Umm Al-Maradem Island in the south- ern Kuwait waters (Fig. 1). One tow was made at each station lasting for 15 to 60 minutes, varying with catch level. All data were standardized to catch per boat-hour for analytical purposes. 185 186 YlMEN ET AL. * Southern Area Umm Al-Maradem 45' Figure 1. The Kuwait waters covered by the shrimp surveys and its stratification. Dots indicate the survey stations in 1993. The Kuwait shrimp fishery captures mainly three species: Pe- ndens semisulcatus, Metapenaeus ajfinis, and Parapenaeopsis stylifera. P. semisulcatus dominates in the central and southern areas, but all three species occur in the northern waters (Ye et al. 1996). Catch rates of either combined species or single species varied greatly from station to station. To improve the shrimp den- sity index and to account for geographical differences in distribu- tion, the entire survey area was stratified into three areas: Kuwait Bay. middle area, and southern area (Fig. 1) (Sparre and Venema 1992. Xuet al. 1995). The Kuwait shrimp fishery consists of two sectors: industrial trawlers and dhow trawlers. The industrial sector consists of three fishing companies. United Fisheries of Kuwait. Bubiyan Fisheries Company, and National Fishing Company, with a total of 35 fish- ing vessels. Its annual landings constituted about 75% of the total shrimp landings in Kuwait from 1985 to 1990 and dropped to 50% after the war (Ye et al. 1996). Each company provided KISR with monthly shrimp catch and effort data for each vessel. The artisanal fleet had more than 70 dhow boats after the war (Ye et al. 1996). This fleet sells shrimp at three local fish markets. The Kuwait Central Statistics Office (CSO) collects shrimp land- ings only from whole sales. Some fishermen, however, sell shrimp directly to retailers; thus, the CSO statistics are conservative esti- mates. KISR established a dockside interview system to collect dhow boats' landings since 1977 when KISR. in conjunction with the United Nations Food and Agriculture Organization, started the Shrimp Fisheries Management Project (van Zalinge 1981. Moham- med et al. 1998). A l-in-5 systematic sample of daily interviews with a random start date each season has been in effect: that is, every fifth day throughout each season, interviews were carried out with skippers of all dhow boats landed that day. and consisted of the following queries: total shrimp catch, days spent fishing, trawl tows per day. hours per tow, and fishing ground (for a detailed description of this interview sampling see van Zalinge 1981). The estimated precision, expressed by the estimated coefficient of variation: that is. standard deviation of the mean divided by mean, of the seasonal estimate of shrimp landings was 5.84% in 1980 to 1981 (van Zalinge 1981). This sampling program has been in use since 1980. Parametric Method Traditional parametric methods have concentrated on the de- velopment of parametric relationships (linear or curvilinear equa- tions) between the observed catch or recruitment and explanatory variables, such as a recruitment index or environmental factors. This method requires a perceived understanding of the physical or biological characteristics of the system that may affect the shrimp species. Its predictive ability has generally been readily variable with blackbox techniques, such as regression and time series analysis, often producing good forecasts of catch or recruitment (Noakes et al. 1989. Stacker and Hilborn 1981). The Kuwait shrimp fishery is a single cohort fishery, with seasonal total landings based mainly on recruitment from the same season (Morgan 1989). It seems straightforward that the density- index derived from a preseason survey is directly related to the yield of the coming season. However, thorough analysis is neces- sary to determine indices from which area and month have the best-fit relationship with seasonal total landings. In this study, a set of monthly mean catch rates from the sam- pling stations within each area during the study period was treated as a time series. Thus. 15 sets (3 areas x 5 months from April to August) were chosen as independent variables and seasonal catch as the dependent variable. A backward, stepwise multiple regres- sion analysis was used to identify the best variables. The prob- ability-to-remove value was set at 0.10 and the probability-to-add value at 0.05. Nonparametric Method Several probabilistic approaches have been applied in fishery forecasts over the last few years as alternatives to parametric mod- els (Evans and Rice 1988. Fried and Hilborn 1988, Noakes 1989, Rice 1993. Chen and Shelton 1996). Nonparametric methods allow the data to "speak for themselves" instead of merely having to conform to some arbitrary mathematical equation. If the assump- tions associated with the parametric methods are not satisfied or if the wrong parametric model has been selected, then these proba- bilistic or nonparametric approaches may produce more accurate forecasts than the competing parametric methods (Noakes 1989). Evans and Rice (1988) developed a modified "kernel estimator" for similar purposes, which showed that the nonparametric method was superior to the alternatives tested. The problem, however, is this: given a set of past observations of recruitment indices and seasonal landings, and the current re- cruitment index, estimate the current corresponding landings prob- ability distribution function (PDF). Thus, the landings PDF is a Shrimp Recruitment Index for Forecasting 187 Catch rate in 1997/98 5000 4000 ■ en | 3000 c CO _l ■ " x, * 2000 - ■ ■ ■ ■ 1000 0 ' r 10 20 50 60 70 30 40 Catch rate (kg/boat-hour) Figure 2. Illustration of nonparametric algorithms for assigning probabilities of occurrence of observed recruitment indices during a projection (scatter plot of species combined landings vs. mean survey catch rate in July). function of past observations and this year's recruitment index, and will, in general, be different if next year's recruitment index is different. The basic rule is to pay more attention to past observa- tions whose recruitment indices are close to this year's recruitment index. There are several algorithms for estimating landings PDF, but Evans and Rice ( 1988) found that the Cauchy algorithm gave the best results; it also seemed to present the least risk of going seri- ously wrong. Therefore, the Cauchy algorithm is also used here. W .=- 1 1 + (1) where W, is a weighting factor for ith observation. X, is difference between value of ith past recruitment index and this year's index value (Fig. 2), D is the "tuning" parameter, in the same unit as X;. When X, is small relative to D (sites are similar). W, is large; when X; is large relative to D (sites differ greatly). W, is small. Among the reasons for using nonparametric density estimation methods instead of parametric methods are lack of knowledge of the true distribution of errors and lack of smoothness in the un- known true functional relations. Therefore, to measure goodness- of-fit, the deviations of observations from the PDF median may be better than deviations from the PDF mean (Rice 1993). The optimum value of D is chosen through crossvalidation (Efron and Gong 1983, Hall and Marron 1990). Begin with a trial D much larger than the range of values of the index observations, and estimate the crossvalidated sum of squared values. Delete one observation in the historical data set. Use Eq. 1 to estimate the PDF for that observation using the value of the deleted recruitment index and all other pairs of observations in the historical dataset. Calculate the squared difference of the seasonal landings at the deleted site from the median of the estimated PDF of the landings. Repeat for each observation in the historical dataset. and sum the squared differences. This crossvalidated sum of squares measures goodness-of-fit. assuming the null hypothesis of no relationship (Rice 1993). RESULTS AND DISCUSSION Combined Species Model The Kuwait shrimp fishery is a multispecies fishery. In the commercial sector, fishers and managers of fishing companies and the management authority, a combined-species index or forecast for the coming season's success might be more important than single-species forecasts. As an initial step, mean catch rates (kg per boat-hour) of combined species from the three areas from April to August were used as independent candidate variables, and seasonal total landings were used as a dependent variable. A stepwise regression built the following model of two inde- pendent variables. a0 + atX, + a^X, (2) where C = yearly catch in kg; X, = mean catch rate in July from the southern area in kg per boat hour; X, = mean catch rate in August from the southern area in kg per boat hour; and a^ a,, and a2 are constants (Table I ). This model fits the data very well (R2 = 0.96. p = .0004. n = 8) (Fig. 3). r-statistics show that the intercept and two coefficients of the model are statistically signifi- cant (Table 1). A residual plot showed that there is no serious autocorrelation of the data. TABLE 1. Statistics of the forecast models. Model Constant Estimate p Statistic Combined species a*, 1255.5 0.001 a. 52.2 0.001 a2 24.2 0.028 R2 = 0.96 F = 54.33 n = 8 P. semisulcatus bo 325.1 0.140 b, 14.9 0.013 b2 15.3 0.074 R2 = 0.89 F = 21.04 n = 8 188 YlMEN ET AL. 6000 5000 4000 H 3000 2000 1000 1985 1987 1988 1989 1993 1994 1995 1996 Year Figure 3. Comparison between the observed and predicted seasonal landings of species combined shrimp in Kuwait waters. Traditionally. Kuwait Bay has been recognized as an important nursery area (Bishop and Khan 1991.Yeetal. 1996). Surprisingly, no significant correlation between the mean catch rate in Kuwait Bay and seasonal yield was found. The survey catch rate was highly variable, as expected in fisheries. Because of lack of sta- tistical consideration in the survey design, only two stations were sampled in Kuwait Bay each month in both the pre- and postwar surveys before 1997. This makes the index from Kuwait Bay more unreliable. To increase the precision of the catch rate estimates, the number of stations was increased to five in Kuwait Bay and seven in the middle and southern areas in the 1997 survey. The average catch rates with standard errors and percentage of P. semisulcatus are listed in Table 2. The most important conclusion from Table 2 is that the per- centage of P. semisulcatus in the catch of Kuwait Bay was very low. 10.3% in weight in July and 5.7% in August 1997. respec- tively. In contrast, the seasonal landings were dominated by P. semisulcatus, from 50% in bad seasons to 98% in good seasons (Siddeek et al. 1994). Although the shrimp densities of combined species in Kuwait Bay were much higher than in the other areas, the low percentage of P. semisulcatus in Kuwait Bay, combined with its small area, underplays the significance of its contribution to the seasonal total landings of shrimp. It may be more reasonable to say that Kuwait Bay is a more important and significant nursery ground for M. affinis and P. styiifera than for P. semisulcatus. Some major nursery grounds of P. semisulcatus may be located in the southern area rather than in Kuwait Bay (Bishop et al. 1994). Further investigation is needed in the future. Because P. semisulcatus dominates the shrimp catch and is more abundant in the central and southern area, a high correlation between the index from the southern area and total yield may be expected. The coefficient for the mean catch rate in July from the southern area (X,) doubled that for the August mean catch rate (X2) (Table 1 ). This hints that the stock in the southern area in July has a more important effect on seasonal landings than that in August, if the estimation is not biased. The decrease of importance of the southern index from July to August may be caused by the much larger standard errors associated with the mean catch rates in August (Table 2). Single-Species Model To develop a recruitment index for each single species, both the survey data and total landings were separated by species based on species composition data from commercial fishing and research survey. A comprehensive examination of the data of the three main species indicates that a single species model is possible only for P. semisulcatus. This may be because both the pre- and postwar surveys, because of regional conflicts, did not cover the northern waters where M. affinis and P. styiifera are dominant. Because fishing effort was also affected by regional conflicts, landings in Kuwait probably do not reflect their population magnitudes. Thus, no significant correlation between the survey density estimates and seasonal landings could be established. The same stepwise regression technique as for the combined- species model was used. A multivariate linear model was found. C = b„ + b,X, + b,X2 (3) TABLE 2. Average catch rates and their standard errors of the 1997 preseason shrimp survey n Species C ( Mean ombined ±SE) P. semisulcatus (Mean ± SE) % of P. semisulcatus July August July August July August Kuwait Bay Middle area Southern area 5 7 7 243.44 ±84 .81 10.34 ±4.81 25.91 ± 12.06 457.58 ± 97.22 11 1.60 ±68.60 48.51 ±41.57 25.02 ±7.89 26.21 ±9.79 9.22 + 3.92 12.17 ±6.09 25.64 ±11.89 48.49 ±41.57 10.28 89.15 98.96 5.73 10.90 99.97 Shrimp Recruitment Index for Forecasting 189 6000 1985 1987 1988 1994 1995 1996 1989 1993 Year Figure 4. Comparison between the observed and predicted seasonal landings of Penaeits semisulcatus in Kuwait waters. where C = total landings of P. semisulcatus; X, = mean catch rate of P. semisulcatus in the southern area in July; X, = mean catch rate of P. semisulcatus in the southern area in August; and, b0, b,, and b, are constants (Table 1 ). The model fits the data well with R" = 0.89 (Fig. 4). r-statistics show that the hypothesis of zero coefficient for X, could be rejected at p = .013. and that for X2 at p = .074. Although the latter is bigger than the traditional criterion of p = .05. we decided to keep it. because the time series is only 8 years (/-statistics are not very reliable for short time series data). By keeping X2, we will significantly reduce the standard deviation of the model, resulting in greater accuracy of the model forecasts. The last datum point in Fig. 4 fits the model more poorly than the others. This may be caused by imprecise species composition data from the industry sector in the 1996 to 1997 season. Species composition was estimated roughly by wholesale grade-size data, and the fit is expected to improve if more precise species compo- sition data are available. Recruitment Indices and Forecasts from Parametric Methods The above regression analyses show that the mean survey catch rates in the southern area in July and August can explain a very high percentage of variation in seasonal landings (Table 1 ) and can serve as an index of recruitment to the fishery in the coming season. The index could be calculated from July mean catch rates alone or July and August mean catch rates, depending upon the requirements of practical aspects. If only 1 month is to be used, then the July mean catch rate is preferred, because the coefficient of variability in August is much higher (Table 2). The increase of the coefficient of variability may be caused by the patched school- ing behavior of shrimp before recruitment (Matthews et al. 1994). In general, the time series is short, and the predictive ability of the models is expected to improve as more data become available. Test forecasts for the 1997 to 1998 season were made on the basis of the above-developed models before the opening of the shrimp season on 1 September. The indices from the southern area in July and August were used (Table 2). The point estimate of the species combined total catch was 3781 .6 tons with 95% confidence boundaries of 2802.1 to 476 1.0 tons (Table 3). The P. semisulcatus catch was predicted at 3684.5 tons with 95% confidence limits from 1751.5 to 5617.6 tons (Table 3). The observed catch of all species was 2551.5 tons in the 1997 to 1998 season, falling outside the 957c confidence interval. 2802. 1 to 4761.0 tons, forecast by the parametric method (Table 3). The landed catch of P. semisulcatus in this season was 2117.7 tons, much lower than the forecast of 3684.5 tons, although it was within the 95% boundaries of 1752 to 5618 tons (Table 3). Forecast from the Nonparametric The regression analysis above suggests that the best recruit- ment indices are the mean catch rates from the southern area in July and August. Although multivariate extrapolations of kernel estimators have been developed (Loftsgaarden and Quesenberry 1965). the multivariate methods are computationally demanding. As a first step, an univariate one was used here. Between these two indices, the July mean catch rates were chosen, because of their greater contribution (Table 1 ) and a lower standard error (Table 2). The estimated PDF is shown in Figure 5. The prediction for the 1997 to 1998 season, in fact, was a full PDF. not a single expected value. Therefore, the plot was probability (y-axis) versus catch (x-axis) for a given recruitment index, not the usual regression- based plot of expected value of catch across the range of the recruitment index. The full uncertainty of each prediction was an intrinsic part of the results. The median, the catch corresponding to the 0.5 cumulative proportion in Figure 5. of the 1997 to 1998 season"s total landings was estimated 3229.9 tons (Table 3), 14.5% TABLE 3. Comparison between observed shrimp landings and forecasts in the 1997 to 1998 season Species combined P. semisulcatus Parametric forecast (t) 3781.6 (2.802^1.761 f Nonparametric forecast (t) 3229.9 ( 1,600-1.000) Observed landings (t) 2551.5 3684.5(1,752-5,618) 2031.9(900-4,000) 2117.7 1 Numbers in bracket are 95% boundaries. 190 YlMEN ET AL. 1000 2000 4000 5000 3000 Catch (tons) Figure 5. The cumulative probability of species combined shrimp landings for the July 1997 recruitment index in the southern area. lower that the prediction from the regression model. The ogive suggests that the 1997 to 1998 season has a very low probability (<3%) of landings over 4,000 tons, and a probability of 5% of landings less than 1,600 tons. The nonparametric method suggests that both the lower and upper 95% boundaries are lower than the estimates given by the regression model. The estimated median of the P. semisulcatus catch in 1997 to 1998 was 2,031.88 tons (Table 3). The probability of landings over 4,000 tons was less than 3%, and there was a probability of about 5% that the total shrimp catch would be below 900 tons (Fig. 6). The actual species combined catch in 1997 to 1998 was 2,551.5 tons, 21% lower than the forecast 3,229.9 tons, but still within the 95% confidence interval of 1,600 to 4,000 tons forecast by the nonparametric method. The single-species catch of P. semisulcatus was 2.117.7 tons, very close to the nonparametric forecast of 2.031.9 tons (Table 3). General Discussion All the predictions have wide confidence intervals. This is at- tributed to the great variation of the survey data, which is usual in fisheries. The parametric forecasts are quite poor. Actual species combined landings in 1997 to 1998 fell outside the 95% confi- dence boundaries, and the parametric forecast for P. semisulcatus was 74% higher than what was landed, although its 95% confi- dence interval covered the real catch value (Table 3). In contrast, the nonparametric methods made better forecasts than traditional parametric methods. Both landed catches of combined species and single species of P. semisulcatus were within the 95% confidence intervals of forecasts. Impressively, the forecast for P. semisulca- tus was only 4% lower than the landed in the 1997 to 1998 season (Table 3). The parametric methods made poor forecasts, although they fitted the historical data very well (Figs. 3 and 4). The best esti- mator that provides the best fit to the historical data may not produce the best forecasts (Noakes 1989). One or two outliers (see the point at the upper right corner of Fig. 2) will affect parameter estimates of regression strongly, and hence, all predictions using this model. In the nonparametric methods, the outliers will be present as an extended limb of the PDF, but the influence of the outliers diminish quickly outside the area of the recruitment index 1 09 08 c °7 o I 0.6 o o. a) 05 > = 04 3 0.3 02 01 0 i r I 1000 2000 5000 6000 3000 4000 Catch (tons) Figure 6. The cumulative probability of Penaeus semisulcatus landings for the July 1997 recruitment index in the southern area. Shrimp Recruitment Index for Forecasting 191 axis where the exceptional observations lie. Predictions from non- parametric methods may be biased by distribution characteristics of the observations, but the predictions do not necessary preserve those characteristics. This may partly explain why the predictions of the parametric methods are higher than those of nonparametric methods (Table 3). The predicted PDF facilitates interpretation in the context of the biology of the population under study and differs from model- based analysis methods. The full PDF is rich with information and can be used in many ways. When testing a priori hypothesis, the slope of the PDF provides important information about the range of estimates consistent with a prediction. Relatively high prob- abilities of 20% of the combined species catch below 2.000 tons can easily be seen in Fig. 5. The direct estimation of the probability that a seasonal catch falls above or below some value given a special set of conditions may be useful in risk assessment. In this study, because of particular difficulties, only a short time series of observations are available. This will deteriorate the ef- fects of outliers on predictions. Although a few outliers do not introduce serious bias to all predictions from density estimation methods, bias is still a concern. Large numbers of observations are needed before kernel estimators provide completely unbiased es- timates of PDFs (Bowman 1985. Silverman 1986). The PDF's forecasts by the kernel methods are likely to contain some bias, but they are likely to be wrong by less than predictions from routine model fitting to a few dozen noisy datapoints (Evans and Rice 1988, Rice 1993). The parametric models used two variables, the mean catch rates of the southern area in July and August; the nonparametric ap- proach, however, employed only one. The nonparametric forecast results from one index, the mean southern catch rate in July, are quite promising. The cause-effect relationship between the sea- sonal landings and recruitment is neither simple nor knowable. There are many potential effects of such environmental variables as water temperature, salinity. Shatt Al-Arab discharge, or area of nursery habitat. In principle, the Cauchy method can be extended to two or more independent variables, but details have not been worked out (Rice 1993). Therefore, difficulties may exist for se- lecting one or two best sets of independent variables for nonpara- metric methods from a great number of variables. Parametric methods may well serve as an approach to identify independent variables primarily for nonparametric methods. ACKNOWLEDGMENTS This study was part of the Shrimp Fisheries Management Proj- ect (code FM01 IK) sponsored by the Kuwait Institute for Scien- tific Research. Special thanks go to all the staff who participated in the surveys. LITERATURE CITED Bishop. J. M.. A. Y. Kwarteng. S. Al-Yakoob. F. Abdeli, P. G. Jacob, K. Al-Matrouk & A. H. Alsaffar. 1994. Postwar survey and evaluation of Kuwait's shrimp nursery grounds with emphasis on the south of Ku- wait Bay. phase II. Kuwait Institute for Scientific Research Rep. KISR4748. Kuwait. 184 pp. Bishop. J. M. & M. H. Khan. 1991. Depth as a factor in abundance and size of juvenile penaeid shrimps in the absence of estuaries and marshes. Mar. Biol. 109:103-114. Bowman. A. W. 1985. A comparative study of some kernel-based non- parametric density estimators. J. Star. Comp. Simul. 21:313-327. Chen, X. H. & P. A. Shelton. 1996. A nonparametric forecast model of inshore Atlantic cod (Gadus morhua) landings based on biomass. cu- mulative landings, and water temperature. Can. J. Fish. Aquat. Sci. 53:559-562. Efron. B. & G. Gong. 1983. A leisurely look at the bootstrap, jackknife, and crossvalidation. Am. Slat. 37:36-48. Evans. G. T. & J. C. Rice. 1988. Predicting recruitment from stock size without the mediation of a functional relation. ICES J. Mar. Sci. 44: 111-122. Fried, S. M. & R. Hilborn. 1988. In-season forecasting of Bristol Bay, Alaska, sockeye salmon (Oncorhynchus nerka) abundance using Bayesian probability theory. Can. J. Fish. Aquat. Sci. 45:850-855. Garcia. S. 1983. The stock-recruitment relationship in penaeid shrimps fisheries: reality or artifacts and misinterpretations? Oceanogr. Trop. 18:25-48. Hall. P. & J. S. Marron. 1990. On variance estimation in nonparametric regression. Biometrika 77:415—419. James. F. C. & E. McCulloch. 1990. Multivariate analysis in ecology and systematics: panacea or Pandora's box? Ann. Rev. Ecoi. Syst. 21:129- 166. Loftsgaarden D. O. & C. P. Quesenberry. 1965. A nonparametric estimate of a multivariate density function. Arm. Math. Stat. 36:1049-1051. Mathews. C. P.. S. Kedidi. J. Al-Qader. I. Al-Qader. A. H. Radhi. N. I. Fita & A. Al-Yahya. 1994. Penaeus semisulcatus stocks of the western Gulf: effects of schooling and environmental variation on interstock modeling and management of Kuwaiti, Saudi Arabian and Bahrain stocks. Proceedings of the Technical and Consultation on Shrimp Man- agement in the Arabian Gulf, 6-8 November 1994, Al Khobar. Saudi Arabia, pp. 1—47. Mohammed. H. M. A., J. M. Bishop & Y. Ye. 1998. Kuwait's post-Gulf war shrimp fishery and stock status from 1991 to 1992 through 1995 to 1996. Rev. Fish. Sci. 6:253-280. Morgan, G. R. 1989. Separating environmental and fisheries effects in the recruitment of Gulf shrimp. Kuwait Bull. Mar. Sci. 10:51-60. Morgan. G. R. & S. Garcia. 1982. The relationship between stock and recruitment in the shrimp stocks of Kuwait and Saudi Arabia. Ocean- ogr. Trop. 17:133-137. Noakes, D. J. 1989. A nonparametric approach to generate inseason fore- casts of salmon returns. Can. J. Fish. Aquat. Sci. 46:2046-2055. Noakes, D. J., K. W. Welch, M. Henderson & E. Mansfield. 1989. A comparison of alternative methods for generating preseason forecasts of sockeye salmon (Oncorhynchus nerka) returns to the Adams and Chilko Rivers, British Columbia, Canada. North Am. J. Fish. Manag. 10:46-57. Rice, J. C. 1993. Forecasting abundance from habitat measures using non- parametric density estimation methods. Can. J. Fish. Aquat. Sci. 50: 1690-1698. Siddeek. M. S. M.. J. M. Bishop. M. El-Musa. A. R. Abdul-Gahffar & M. S. Abdulla. 1994. Possible reasons for increased landings of Ku- wait's green tiger shrimp (Penaeus semisulcatus) in the late 1980s. Proceedings of the 3rd Asian Fisheries Forum, 26-30 October 1992. Singapore, pp. 224-227. Silverman. B. W. 1986. Density estimation for statistics and data analysis. Chapman & Hall, New York. 169 pp. Sparre. P. & S. C. Venema. 1992. Introduction to tropical fish stock as- sessment. FAO Fisheries Technical Paper 306/1. 376 pp. 192 YlMEN ET AL. Stacker. M. & R. Hilborn. 1981. Shore-term forecasting in marine fish stocks. Can. J. Fish. Aquat. Sci. 38:1247-1254. van Zalinge, N. P. 1981. Kuwait's artisanal fisheries — a sampling pro- gramme for shrimp landings. Kuwait Institute for Scientific Research Rep. KISR396, Kuwait. 21 pp. Walters. C. J. 1989. Value of short-term forecasts of recruitment variation for harvest management. Can. J. Fish. Aqua!. Sci. 46:1969-1976. Xu. X., J. M. Bishop. H. M. A. Mohammed & A. H. Alsaffar. 1995. Esti- mation of the natural mortality rate of green tiger prawns Penaeus semisulcatns (De hann, 1844) in Kuwait waters using relative abun- dance data. J. Slieltfish Res. 14:179-184. Ye, Y. & H. M. A. Mohammed. 1999. An analysis of catchability varia- tion in the Kuwait shrimp {Penaeus semisulcatus) fishery. Fish. Bull. USA 97:702-712. Ye. Y.. H. M. A. Mohammed & J. M. Bishop. 1996. Shrimps resources and fisheries in Kuwait waters. Tech. Rep. Kuwait Institute for Scientific Research. 38 pp. Journal of Shellfish Research, Vol. 18, No. 1, 193-201. 1999. SKEWED SEX RATIO IN AN ESTUARINE LOBSTER (HOMARUS AMERICANUS) POPULATION W. HUNTTING HOWELL, WINSOR H. WATSON, III AND STEVEN H. JURY Department of Zoology and Center for Marine Biology University of New Hampshire Durham, New Hampshire 03H24 ABSTRACT A total of 19.485 lobsters were caught at eight sites in the estuarine and coastal waters of New Hampshire from 1989 to 1992. and their size and sex were determined. The sex ratio of lobsters caught farthest from the coast, in Great Bay. was heavily skewed in favor in males. Sex ratios in other estuarine and river sites were also skewed toward males, and there was a tendency for the number of males per female to decline as one moved down the estuary toward the coast, where the sex ratio was nearly 1:1. The single offshore site was dominated by females, with about 0.6 males for each female. There were also seasonal trends in the sex ratios in the upper estuarine sites, where the number of males per female tended to decline from summer through autumn. In general, differences in the sex ratios between sites were those of primarily adult lobsters larger than 80 mm carapace length (CL). At all sites, the sex ratio of lobsters smaller than this size was close to 1:1, whereas in the upper estuary the mean sex ratio of lobsters greater than 80 mm CL was more than 14:1. These data, in conjunction with seasonal variations of sex ratios, suggest that differential movements of adult male and female lobsters is the primary cause of skewed sex ratios in the Great Bay Estuary. KEY WORDS: Estuary, lobster, Homarus americanus, sex ratio INTRODUCTION The American lobster. Homarus americanus (Milne-Edwards), is broadly distributed in the western north Atlantic from Labrador to North Carolina (Squires 1990). Within this range, the species supports important commercial and recreational fisheries, particu- larly in New England and the Canadian Maritime provinces. Be- cause of their commercial importance, lobsters have received a considerable amount of attention from the scientific community (see reviews by Cobb and Phillips 1980a, Cobbs and Phillips 1980b, Factor 1995). Not surprisingly, most of these studies have focused on coastal and off-shore populations where lobsters are most abundant. Although lobsters are considered to be stenohaline, and gener- ally limited to marine (>25 ppt salinity) habitats (Dall 1970), there are smaller exploited populations found in estuarine habitats (Thomas 1968, Thomas and White 1969, Munro and Therriault 1983. Reynolds and Casterlin 1985. Vetrovs 1990). The physi- ological ecology and population structure of these lobsters is poorly understood. In recent years, we have been studying one such population located in the Great Bay Estuary of New Hamp- shire, USA (Jury et al. 1994a, Jury 1994b, Jury et al. 1995. Crossin et al. 1998, Watson et al. 1999). This system, located in the south- eastern portion of the state, is characterized by extensive mudflats separated by deep (10-20 m) channels, strong tidal mixing and flushing, and marked seasonal changes in temperature and salinity. Monthly mean temperatures can vary from 0-18°C at the coast, and from 0-25°C in the upper estuary (Loder et al. 1983). The system receives freshwater from seven rivers that drain an area of approximately 2400 km2. Salinities in the upper estuary may drop to 10-15 ppt in the spring, as freshets associated with snow and ice melt, and heavy rains enter the system. At the coastal terminus, average salinities are much more stable, typically ranging from 30-33 ppt (Loder et al. 1983). Among the data we have gathered is information on sex ratio by location, season, and size class. The sex ratio of many geo- graphically separate American lobster populations has been re- ported. Although most coastal lobsters populations that have been examined approximate the expected 1 : 1 ratio (Cooper 1970, Stew- art 1972. Krouse 1973. Cooper et al. 1975, Pecci et al. 1978), there are several instances where skewed ratios have been observed. These include reports of populations with more males than females (Briggs and Mushacke 1979, Munro and Therriault 1983, Karnof- sky et al. 1989). as well as reports of populations with more fe- males than males (Skud and Perkins 1969, Estrella and McKiernan 1989). Explanations for these skewed sex ratios have included differential catchability (Krouse and Thomas 1975. Fogarty and Borden 1980, Miller 1990, Tremblay and Eagles 1997). segrega- tion of the sexes by depth (Skud and Perkins 1969. Briggs and Mushacke 1979). differences in migratory behavior (Munro and Therriault 1983). physiological and behavioral differences be- tween the sexes (Jury et al. 1994a, Jury et al. 1994b), and fisheries regulations that protect some females (Estrella and McKiernan 1989). In this paper, we report consistent spatial differences in lobster sex ratio within a New England estuary, and differences in sex ratio between size classes of lobsters found in upper estuarine areas. MATERIALS AND METHODS The Great Bay estuarine system lies in the southeastern corner of New Hampshire. USA. It receives freshwater from seven rivers, which mixes with saltwater entering from the western Gulf of Maine. Lobsters were sampled at eight sites in the estuarine and coastal waters from 1989 to 1992 (Fig. 1). These spanned a dis- tance (by water) of approximately 37 km, ranging from Great Bay proper, which is about 26 km inland, to the Isles of Shoals, which lie 1 1 km offshore. The eight sites fall into three broader spatial categories, which we have arbitrarily designated as "estuarine" (Great Bay. Little Bay, Bellamy River), "riverine" (upper, mid-, and lower Piscataqua River), and "coastal" (Coast, Isles of Shoals). Along this line of sites, physical and chemical characteristics vary from those of a typical New England estuary (greatly fluctuating temperature and salinity, strong tidal mixing, soft substrate) to those of a typical New England coast (relatively stable temperature and salinity, less tidal current, and harder substrates of cobble and rock). 194 Howell et al. Maine - 43 06 * 43 00 70 48 70 42 Figure 1. Location of the study sites within the estuarine and coastal waters of New Hampshire. GB (Great Bay). LB (Little Bay), BR (Bellamy River), UPR (Upper Piscataqua River), MPR (Middle Piscataqua River), LPR (Lower Piscataqua River). CST (Coast), SHL (Shoals). All lobsters were caught in traps baited with herring and tended two to three times per week. Most were caught in our own traps as part of a larger study on estuarine lobsters, but many were caught by commercial lobstermen with whom we fished, and a small number were caught by the New Hampshire Department of Fish and Game. All traps from which we collected data were made of vinyl-coated wire, equipped with one or two escape vents (1 7/8" H x 6" W). and had either a single (research traps) or double parlor (commercial traps). Although winter sampling was limited because of upper estuarine ice cover and general lack of commercial fish- ing activity, we were able to sample all sites adequately during the spring (April-June), summer (July-September), and autumn (Oc- tober-December) in most years. All lobsters had their carapace length (CL) and abdomen width measured to the nearest millime- ter, all were molt-staged using external shell criteria and/or pleo- pods (Aiken 1973. Aiken 1980), and all were sexed by examining the first pair of pleopods (Templeman 1944). Most were also tagged, before release, with numbered modified sphyrion tags (Scarratt 1970). because in another part of the study, we were examinine movement and growth (Watson et al. in press). At each study site, except the Shoals, both temperature and salinity (YSI Meter Model 33) were measured at the surface each time our traps were hauled. In 1991. data were collected from surface and bottom waters. There was always <2°C and 2 ppt difference between surface and bottom values because of extensive vertical mixing (Loder et al. 1983). Lobster Sex Ratio 195 u B u Oh e Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 35 30 * 25 •a i tza HO 15 10 B i ' i ■ i ■ i ' i — ■ — i — > — i — ' — i — > — i — ' — i — ' — \ — ■ — i — Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month Figure 2. Mean <±SE) monthly temperatures (°C) and salinities (ppt) at the Great Bay and Coastal sites from 1989 to 1992. The null hypothesis of equal numbers of males and females was, in all cases, tested using chi square analysis. Comparisons of sex ratio between sites, and between seasons within sites, were done using one-way analysis of variance (ANOVA) followed by Tukey's posterior test. Mann-Whitney U tests were used to com- pare the sizes of males and females at each site within each year. The alpha level for all statistics was 0.05. RESULTS Mean monthly temperature and salinity, from 1989 to 1992, at two of our sites is depicted in Fig. 2. In the upper estuary (Great Bay site, GB), mean monthly temperatures were 3-7°C warmer than the Coastal (CST) site from April through October (Fig. 2a). Mean salinity in Great Bay was highest in late summer (approx. 27 ppt), and lowest (approx. 16 ppt) in the spring when freshwater input was more abundant because of heavy rains and snowmelt. At the Coastal site, the salinity was relatively constant throughout the year (Fig. 2b). As expected, both temperatures and salinities at the sites that occur between Great Bay and the Coast are intermediate to those depicted. Although complete temperature and salinity records are not available from the Isles of Shoals, the small amount of data available indicates this site is very similar to the Coastal site, located approximately 1 1 km away. Observed mean sex ratios, from 1989 to 1992. at each of the sampling sites are given in Table I . Although there was some interannual variation at each site, in each of the estuarine sites (Great Bay. Little Bay, Bellamy River), there were significantly more males than females in every year (p < .05). In the riverine sites (upper-, mid-, lower Piscataqua River) and at the Coast, the numbers of males and females were more nearly equal. However, even at these sites, there were significantly more males than fe- males in some years (Upper Piscataqua River 1991: mid- Piscataqua River 1989. 1990; lower Piscataqua 1991; Coast 1990). As with the other sites, there are some interannual variation in sex ratio at the Isles of Shoals, but in each year for which data were available, there were significantly more females than males (p < .05). When the data from all years and seasons were combined, sex ratio departed significantly from the expected 1:1 ratio in seven of 196 Howell et al. TABLE 1. Mean number of males per female at each sampling site from 1989 to 1992 and in all years combined. Year Site 1989 1990 1991 1992 All Years Great Bay Mean 6.92* 5.32* 5.88* 3.03* 5.29* SE 2.65 0.86 3.28 0.53 0.82 n 83 136 358 130 707 Little Bay Mean 1.81* 1.86* 2.04* 1.68* 1.85* SE 0.15 0.12 0.39 0.14 0.07 n 297 876 1,165 516 2,854 Bellamy R. Mean 1.32* 1.32* 2.30* 1.56* 1.63* SE 0.16 0.01 0.37 0.37 0.63 n 614 1.318 1,236 902 4,070 Upper Pise. R. Mean 1.27 1.29 1.39* 1.31* SE 0.21 0.19 0.20 0.04 n 486 316 724 0 1 .526 Middle Pise. R. Mean 1.42* 1.28* 0.95 0.93 1.14* SE 0.10 0.05 0.16 0.15 0.12 n 1.197 1.210 489 342 3,238 Lower Pise. R. Mean 0.89 1.02 1.44* 1.11 1 12 SE 0.16 0.06 0.08 0.02 0.12 n 278 480 244 284 1,286 Coast Mean 1.02 1.33* 1.22 1.00 1.18* SE 0.24 0.14 0.12 0.08 n 253 1.248 463 777 2,488 Shoals Mean 0.64* 0.52* 0.56* 0.57* SE 0.07 0.06 0.03 n 222 0 1,893 1.201 3,316 Mean and standard error within years is based on three seasons. Mean and standard error for all years is from all years and all seasons combined. (n) = number examined: * denotes a significant departure from a 1:1 sex ratio (chi square, p < .05). the eight sampling sites (Table 1). There were significantly more males than females in each of the five upper sites and at the coast (p < .05). and significantly more females than males at the Isles of Shoals (p < .051. In the remaining site (lower Piscataqua River) there were approximately equal numbers of males and females. Using the aggregate data from all years and seasons, we found that the mean number of males per female in Great Bay (5.29) was significantly higher (p < .001 ) than every other site. Although there was a tendency for the number of males per female to decline as one moved toward the coast (Table 1 ). there was no significant difference in the mean sex ratio among these other sites (p > .05). This same tendency was also observed in each of the three upper estuarine sites. Although there was some interannual variation in each of these sites, in each year except 1991. sex ratio declined as one moved down the estuary from Great Bay to Little Bay to the Bellamy River. Unlike the three upper estuarine sites, there was no obvious clinical trend in sex ratio found in the riverine and coastal sites. The Shoals site, however, had the lowest mean sex ratio of all sites in each of the years for which we had data (Table 1 ). In addition to the observed spatial differences in sex ratio, there were also some seasonal trends observed (Fig. 3). At the Great Bay site, there was a considerable amount of interannual variation within seasons. Although the mean number of males per female tended to be highest in the spring (6.37), and then to decline through the summer (5.45) and autumn (4.38). there were no sig- nificant differences (p > .05) between seasons. A similar, but less pronounced pattern was observed in Little Bay, but again, there were no significant differences between seasons (p > .05). Inter- estingly, at the Bellamy River site, the seasonal trend was reversed. Although there was no significant difference between seasons, sex ratio tended to increase from spring (1.42), to summer (1.50). to autumn (1.84). Seasonal trends were much less pronounced at the other five sites. Significant seasonal differences in sex ratio were found only at the upper Piscataqua River site, where the mean number of males per female was significantly higher (p < .001 ) in the spring (1.71) than in either summer (1.10) or autumn (1.13). between which there was no significant difference (p > .05). The mean size (CL) of male and female lobsters at each site and year is given in Table 2. In general, the mean size of males were significantly larger than that of females in the three upper estuarine locations. In the Piscataqua River sites, males and females were more similar in mean size. The only significant differences were found in the middle Piscataqua River, in 2 of the 4 years, and in the lower Piscataqua River, in 1 of the 4 years. In each of these instances, males were larger than females. At the coastal site, the mean size of males was larger than that of females in 1989: whereas the reverse was true in 1990. No difference in mean size was found in the remaining 2 years. Finally, at the Isles of Shoals site, where we had only 3 years of data, females were significantly larger than males in two of these ( 1989. 1992), but not in the other (1991). Sex ratio varied with size class at certain sites (Table 3). The site where the change in ratio with size class was most pronounced was in Great Bay, which is the site farthest up the estuary. In the other estuarine sites (Little Bay and Bellamy River), males also tended to dominate the larger size classes, but not to the same extent as in Great Bay. In the riverine sites, there tended to be more males than females in manv size classes, but the sex ratio was most Lobster Sex Ratio 197 u "3 a ,° '— H S CO U I Spring Q Summer E3 Autumn k as s - - & 1 1 h m m It ^■C^ BR UPR MPR LPR COAST SHOALS Sampling site Figure 3. Mean (±SD) number of males per female at each of the sampling sites in the spring, summer, and autumn. Means are based on the 4 years included in the study ( 1989 to 1992). The number of lobster examined (h) is given vertically above each error bar. GB (Great Bay), LB (Little Bay). BR (Bellamy River), UPR (Upper Piscataqua River), MPR (Middle Piscataqua River), LPR (Lower Piscataqua River). heavily skewed toward males in the largest (>85 mm CL) size classes. This was particularly true at the uppermost riverine site (upper Piscataqua), and, to a lesser extent, the middle Piscataqua site. At the coastal site, there was a less pronounced pattern of change in sex ratio with size class. There tended to be more males than females in the smaller size classes (<70 mm), about equal numbers of males and females in the 71-90 mm size classes, and about twice as many males as females in the largest size class (>90 mm). At the Isles of Shoals site, however, there were consistently fewer males than females in all of the larger size classes (>65 mm), and the same number, or more, males than females in the smaller size classes. TABLE 2. Mean (and standard deviation) carapace length (mm) of male and female lobsters at each site in each year, and in all years combined. 1989 1990 1991 1992 All Years Site M F M F M F M F M F Great Bav 80.7* 71.2 81.3* 75.5 83.7* 78.7 77.7 76.7 82.3* 77.4 SD = 9.6 16.7 7.2 7.2 8.1 5.8 7.8 2.8 8.3 7.5 Little Bav 78.1* 74.9 75.3 73.5 81.7* 78.3 78.3* 76.0 80.0* 76.0 SD = 6.7 8.6 6.4 7.2 6.9 6.3 7.7 7.5 7.3 7.3 Bellamv River 73.6* 72.0 74.4* 72.0 80.4* 76.0 79.8* 76.8 76.6* 73.3 SD = 10.7 10.4 8.4 8.4 8.2 7.0 8.5 6.8 9.2 8.7 Upper Piscat. R. 51.7 53.7 53.0 50.9 75.1 74.8 No data No data 57.6 57.2 SD = 14.7 14.2 13.3 13.4 9.5 6.5 16.2 15.6 M. Piscat. R. 72.2 71.7 74.9* 73.5 77.0* 75.3 78.1 70.4 73.9* 73.1 SD = 7.0 7.7 7.7 7.6 6.7 6.8 8.2 12.2 7.4 7.6 L. Piscat R. 76.5* 74.3 65.3 63.9 72.2 70.3 78.0 80.9 70.0 68.5 SD = 8.8 10.8 12.4 12.3 11.9 13.2 7.7 8.6 12.4 12.9 Coast 77.7* 76.5 7 1 .9* 74.4 75.5 75.2 75.1 67.5 74.1 74.9 SD = 5.5 5.1 13.2 11.8 9.7 10.4 3.9 16.4 11.3 10.6 Shoals 75.4* 80.1 No data No data 79.6 80.1 80.1* 81.4 79.5* 80.6 SD = 5.8 11.9 7.5 7.6 8.4 8.7 7.8 8.3 * Between male and female lengths within a year and site indicates that the mean lengths of the two sexes are significantly different (Mann-Whime> U test, p < .05). Sample sizes are given in Table 1. 198 Howell et al. TABLE 3. Mean (±SE| number of males per female in different size categories at each sampling size, based on data collected over 4 years. Carapace Upper Middle Lower Length (mm) Great Bay Little Bay Bellamy R. Pise. R. Pise. R. Pise. R. Coast Shoals = <40 1.00 1.21 ±0.49 1.52 ±0.22 0.71 ±0.20 0.93 + 0.05 41-15 1.00 0.77 ±0.19 1.29 + 0.17 0.67 ±0.29 0.55 ± 0.25 0.25 ±0.1 7 46-50 0.75 ± 0.25 1 .33 ± 0.47 0.90 ± 0.28 1.17 ±0.96 0.68 ±0.31 2.5 ± 0.90 1.5 ±0.35 51-55 1.00 1.85 ± 1.06 1.14 ± 0.19 1.25 ±0.43 1.45 ±0.22 0.69 ±0.19 1.31 ±0.17 3.50 ± 1.77 56-60 1.33 ±0.29 0.47 ± 0.22 1.27 + 0.14 1.39 ±0.23 0.96 ±0.31 0.76 ±0.15 1.23 ±0.14 1.50 ±0.35 61-65 1.75 ±0.48 1.42 ±0.27 1.16 =t 0.12 119 + 0.17 0.99 ±0.22 1.23 ±0.20 1 .65 ± 0.09 0.95 ±0.19 66-70 2.38 ± 0.69 1.45 ±0.15 1.34 ±0.19 1.14 ±0.26 1.21 ±0.09 1.02 + 0.18 1.35 ±0.26 0.75 ±0.16 71-75 3.00 ± 0.75 1.28 ±0.15 0.79 ± 0.09 1.17 ± 0.23 1.14 + 0.32 0.98 ± 0.25 0.97 ±0.10 0.60 ± 0.02 76-80 3.42 ± 0.70 1.94 + 0.20 1.38 ±0.22 1 13 ±0.06 1.16 ± 0.15 1.28 ±0.09 1.08 ±0.09 0.49 ±0.01 81-85 12.82 ±5.68 2.74 ±0.28 2.53 ±0.41 1.69 ±0.17 1.11 ±0.12 1.38 ± 0.18 0.98 ± 0.20 0.62 ± 0.03 86-90 9.05 ± 1.78 3.9 ± 0.34 5.12 + 2.03 7.00 ±0.00 3.05 + 1.04 1.79+ 121 0.93 ±0.17 0.58 ± 0.04 >90 14.5 ±5.48 6.43 ± 3.05 9.11 ±2.41 3.00 ±0.00 2.19 ± 1.28 0.80 ±0.14 2.2 ± 1.08 0.55 ± 0.03 DISCUSSION Results of this study indicate that this estuarine population of lobsters departs, in many ways, from the expected 1:1 sex ratio typical of coastal populations. Among the most consistent of our findings was the observed spatial difference in sex ratio. In each season of each year, the upper estuary had more males than fe- males. This skewed ratio tended to decrease, in a clinal fashion, as one moved down the estuary toward the coast, where the sex ratio approximated the expected 1:1 ratio. Surprisingly, this clinal trend continued outside the estuary, so that at the Isles of Shoals, which is located about 1 1 km from the mouth of the estuary, there were consistently more females than males (Table 1). Lobster populations with skewed sex ratios have been reported by others (Skud and Perkins 1969. Briggs and Mushacke 1979. Munro and Therriault 1983, Campbell and Pezzack 1985, Kamof- sky et al. 1989. Estrella and McKiernan 1989), and several expla- nations have been put forth to explain the disparity between num- bers of males and females. Fishery-related factors, including dif- ferential catchability of males and females (Miller 1990), and regulations that protect some (e.g., ovigerous, V-notched) females (Estrella and McKiernan 1989), can result in skewed sex ratios. Differential catchability of the sexes is an unlikely explanation for our results, because Becker ( 1994), who also worked in the Great Bay estuarine system, found sex ratios virtually identical to ours using SCUBA sampling. Furthermore, if males and females dif- fered in their trapability, as suggested by Fogarty and Borden (1980), Miller (1990). and Campbell (1992). we would have ex- pected skewed sex ratios in all of our study sites because the same types of traps, including identically sized escape vents, were used at all locations. It is possible, however, that the skewed ratio fa- voring females at the Isle of Shoals may have resulted from dif- ferences in mean size, and therefore, trapability. of the sexes. When data from the 2 years were combined, females were signifi- cantly larger than the males (Table 2). It has been suggested that female lobsters in some areas (Rhode Island) have a proportion- ately wider carapace width than similarly sized males (Fogarty and Borden 1980). This difference in body proportion between the sexes may not be geographically universal, however, because Krouse and Thomas ( 1975) found no significant differences in the carapace length-width ratios of males and females along the Maine coast. If females in our study area do have a proportionately wider carapace, they would not move as readily through escape vents, so it is possible that the larger mean size of the females at this site influenced sex ratio in the catch. It is also possible that regulatory protection of certain females may explain the sex ratio observed at the Isles of Shoals, where females outnumbered males. At this location, percentages of ovigerous and V-notched females (5-12%) are relatively high as compared to the estuary (<\%) (Howell. W. H. & W. H. Watson. Dept. of Zoology. Univ. of New Hampshire. Durham. NH 03824. Unpubl. data). Thus, both size and protective management may explain the preponderance of fe- males at the Isles of Shoals, but it is also possible that this site simply has a distinctive physical habitat that has resulted in an aggregation of females such as that reported by Campbell ( 1990). Ecological factors may also affect sex ratio. Skud and Perkins ( 1969) and Briggs and Mushacke ( 1979) found a segregation of the sexes by depth, while Karnofsky et al. (1989) suggested that in- traspecific competition effected sex ratio. Our sampling locations were similar in depth (=3-10 m), so it is highly unlikely that our observations resulted from segregation of the sexes by depth. It is also unlikely that intraspecific competition was a factor. Karnofsky et al. ( 1989) found that there were nearly twice as many males as females in a small, shallow cove in Buzzards Bay, MA, USA and that a disproportionately large proportion of the males were miss- ing one or more claws. The authors suggested that the cove may function as a refuge for injured males, and that these individuals had been displaced to this shallow water site by aggressive, in- traspecific competition for mating shelters. Moreover, they sug- gested that the relative paucity of females at their study site re- sulted from the preference of females for deeper areas where the dominant males held mating shelters. We have no information about where mating occurs in our study area, including the depth of mating shelters and whether or not the spatial distribution of females is affected by the distribution of dominant males. Thus, it is possible that the mechanism described by Karnofsky et al. ( 1989) may be applicable to this study, but it is doubtful, because we saw no indication that the proportion of males missing claws differed among sites (Howell and Watson). The spatial and temporal trends of the data in this study indicate that sex ratio may be associated with seasonally changing gradi- ents of salinity and/or temperature that are typical of northern estuaries. In particular, it is likely that male and female lobsters differ in their physiological and behavioral responses to salinity Lobster Sex Ratio 199 and/or temperature and that these differences result in the sex ratio patterns we observed. Water temperature affects many, if not all, aspects of lobster biology. In a laboratory study, Crossin et al. (1998) documented that lobsters are capable of sensing temperature, and that they behaviorally thermoregulate: seeking preferred temperatures and avoiding water that is either too warm or too cool. Results from related studies, also done in our laboratory, further suggest that males and females may respond differently to changing tempera- ture. In one study. 75% of females, but only 50% of males, exited their shelters as shelter temperature was increased (Jury. S. H. Dept. of Zoology. Univ. of New Hampshire, Durham, NH 03824. Unpubl. data); whereas in another study, in which males and fe- males were placed in a thermal gradient tank, males generally preferred wanner temperatures than females, particularly in the spring and fall, when ambient temperatures were seasonally lower (Jury. S. H. The effect of acclimation temperature and sex on the behavioral thermoregulation of the American lobster. Homarus americanus. In prep.). Although these data are preliminary, they suggest that males and females differ in their temperature prefer- ences, and that spatial and temporal differences in temperature could thus affect sex ratio. Although laboratory studies on temperature are relatively scarce, numerous field studies have documented that water tem- perature affects the temporal and spatial distribution of lobsters, and that males and females differ in their movements in response to seasonally changing temperatures (Munro and Therriault 1983. Roddick and Miller 1992, Lawton and Lavalli 1995. Estrella and Morrisey 1997). It has been suggested, for example, that seasonal onshore -offshore migrations are associated with temperature se- lection, and are adaptive for accelerating growth and egg devel- opment (Saila and Flowers 1968. Cooper and Uzmann 1971, Pez- zack and Duggan 1986. Estrella and Morrissey 1997). This may also be true, on a geographically smaller scale, for seasonal mi- grations that occur within New England estuaries, including Great Bay (Watson et al. 1999). Differential migration of the sexes, associated with seasonal changes in water temperature, can also effect sex ratio. Roddick and Miller (1992) found, for example, that males and females arrived at. and departed from, a small embayment in Nova Scotia in different months, and these differ- ences in seasonal movements resulted in skewed sex ratios. Adult females have also been reported to move to deeper water earlier in the autumn than males (Campbell and Stasko 1986. Robichaud and Campbell 1991 ). which results in temporal and spatial segregation of the sexes. Munro and Therriault ( 1983) found more males than females in estuarine locations in the Magdelaine Islands, and spec- ulated that this resulted from differential migration of the sexes. Both sexes left the estuaries as temperatures cooled in the autumn, but males were more likely to return in the spring as temperature increased. A similar situation may exist in the Great Bay Estuary. In a study concurrent with this one, Watson et al. (1999) docu- mented that lobsters tended to migrate up the Great Bay Estuary in the spring as temperatures increased, and down the estuary in the summer and autumn. Although Watson et al. saw no marked dif- ferences in the movements of males and females, their data were somewhat equivocal on this point, and they suggested that differ- ential movement of the sexes was possible. Munro and Therriault (1983) suggested that the reason for males returning earlier than the females was to take advantage of the wanner temperatures of the estuarine sites for molting. Indeed, they found that all males <75 mm CL molted twice each year. It has also been suggested that there are seasonal differences in the catchability of males and females, that these differences are caused by the two sexes molting at different times, and that dif- ferential catchability results in seasonally changing sex ratios (Tremblay and Eagles 1997). In the Great Bay Estuary, however, we saw no evidence that males and females molted at different times or in different locations (Howell and Watson unpubl. data). We conclude from this that there is no difference between the sexes in location and temperature of molting. Thus, although our skewed sex ratios may indeed be related to temperature-mediated differences in movement between the sexes, it seems unlikely that it is strongly correlated with molting, as suggested by Munro and Therriault (1983) and Tremblay and Eagles (1997). A number of laboratory and field studies have documented that salinity can also effect the temporal and spatial distribution of lobsters. Lobsters are considered to be poor osmoregulators (Dall 1970). and several previous field studies have shown that lobsters use behavioral mechanisms to avoid low salinities (Munro and Therriault 1983. Reynolds and Casterlin 1985. Maynard 1991. Jury et al. 1995). In a recent laboratory investigation Jury et al. ( 1994a) measured hemolymph osmolarity. oxygen consumption, heart rate and ventilation rate of lobsters under salinity regimes similar to those found in the Great Bay Estuary under spring runoff condi- tions. They found that exposure to decreasing salinity (from 20 to 10 ppt) caused an increase in oxygen consumption, heart, and scaphognathite rate. At the lowest salinity (10 ppt). females re- quired more energy than males to maintain the same hemolymph osmolarity. Females also recovered more slowly than males as salinities were subsequently increased. This study has been con- firmed by Houchens (1996). and extended to show that female lobsters suffer significantly more mortality than males when held at 5-10 ppt. For this reason, upper estuarine locations where sa- linities are the lowest, particularly in the spring, probably represent a stressful and potentially lethal environment for females. In a second set of experiments. Jury et al. (1994b) measured the be- havioral response of lobsters to reductions in salinity. When given a choice of salinity, females were more selective in their prefer- ence for higher salinity, and females found low salinities more aversive than did males. Results from these studies indicate that lobsters respond to changes in salinity, that male and female lob- sters differ in their physiological and behavioral responses, so that males find low salinity less aversive and less stressful. It is likely that these differences partially explain the observed skewed sex ratios found in this study. In general, we found an inverse rela- tionship between lobster sex ratio and salinity. Physiological and behavioral differences in the way each sex responds to salinity could also explain the seasonal trends in sex ratio that we ob- served. The number of males per female was highest in the spring in the upper estuary, when salinities were lowest, and then de- clined over summer as salinities increased. We believe that the observed reduction in sex ratio was caused by the arrival of more females as salinity increased in these areas. Aside from the physiological and/or behavioral reasons already discussed, it is possible that the observed spatial pattern in sex ratio may also relate to the reproductive biology of lobsters. Because lobster embryos and larvae are quite vulnerable to low (<14 ppt) salinity (Scaratt and Raine 1967, Charmantier et al. 1998. Forward 1989). relatively low salinity environments, such as those in the upper estuary, may be suboptimal for reproduction. Unpublished data on the distribution of ovigerous females in this study support this view (Howell and Watson). We cauaht and examined 8.153 200 Howell et al. female lobsters as part of this study, and 168 of these (2.06%) were ovigerous. Of these 168. only 43 were caught in the estuarine and riverine sites, and the remaining 125 were from the Coast and Shoals. The low incidence of ovigerous females in the estuary is similar to the situation reported for blue crabs in the upper Chesa- peake Bay by Hines et al. (1487). and it is likely that ovigerous females avoid the low salinity conditions of the estuary, because salinity is generally too low for larval survival. Note, however, that Munro and Therriault ( 1983) found a higher percentage of oviger- ous females (13-16%) in estuaries than they did at the coast (7%). The difference between their study and ours may have resulted from the fact that our upper estuarine salinities are typically as low as 10-15 ppt in any given year; whereas the lowest reported by Munro and Therriault was 22 ppt. We also found that sex ratio was more skewed in larger size classes (>80 mm CL) in all of our estuarine and riverine locations. Changes in American lobster sex ratio with size class have also been noted by Karnofsky et al. (1989). They found that females dominated the 50-59 mm CL size class, but that males were more numerous than females in size classes >60 mm CL. As a result, males were not only more common, they were also larger. We believe that the observed changes in sex ratio with size class are related to changes in mobility with size. Wahle and Steneck ( 1992) suggested that small lobsters (<=60 mm CL) are dependent on their shelters to avoid predation. but that this vulnerability is even- tually outgrown, and lobsters >=60 mm CL are able to move about more freely, because they are virtually immune to predation. Once this release has occurred, mobility generally increases as lobsters continue to increase in size (Campbell and Stasko 1986, Campbell 1989). The fact that both mobility and skewness in sex ratio in- crease with size class indicates that changes in sex ratio with size may result from differential movement of the sexes. When small, both sexes move little, and sex ratio is approximately 1:1. As size (and mobility) increase, males, which are more tolerant of low salinity than females, may travel further up the estuary, especially in the spring, resulting in the predominance of males in the larger size classes in this location at this time. Studies are currently underway to determine if the aforementioned differences in the behavior of male and female lobsters exist, even in the smaller size classes, or if they manifest themselves only as they reach sexual maturity. If the latter situation is true, it supports the view that the strongest influence on female migratory behavior in the estuary is related to reproduction and the seeking of appropriate habitats for hatching of larvae. In the Great Bay Estuary, ovary dissections indicate that approximately 50% of females have reached sexual maturity of 80 mm CL (Howell and Watson unpl. data), and it is in size classes greater than this that we observe the most skewed sex ratios. In summary, we believe that the skewed sex ratio patterns we observed in this study resulted from differential movement of the sexes: probably in response to salinity and temperature cues. Both sexes tend to move down the estuary in the summer and autumn. Males, which are more tolerant of low salinity and warmer tem- peratures, return to upper estuarine areas earlier than females in the spring, which accounts for the elevated sex ratio seen in these locations. Although some females move up the estuary as salinity rises, thereby making the sex ratio more nearly equal, more fe- males than males remain in the lower estuary, because they are less tolerant of low salinity and warmer temperatures, and/or because it is a more favorable (higher salinity) location to release their larvae. The fact that sex ratio is most skewed among the largest size classes, which are also the most mobile, supports our contention that skewed sex ratio in our study site results from differential movement of the sexes. ACKNOWLEDGMENTS We thank the numerous students, commercial fishermen, and NH Fish and Game personnel who participated in this research. Funding was provided by the University of New Hampshire Sea Grant program. This is publication #344 of the UNH Center for Marine Biology/Jackson Estuarine Laboratory series. LITERATURE CITED Aiken D. E. 1973. Proecdysis. setal development, and molt prediction in the American Lobster {Homarus americanus). J. Fish. Res. Board Can. 30:1337-1344. Aiken. D. E. 1980. Molting and growth, pp. 91-163. In: Cobb. J. S. and B. F. Phillips (eds.). The Biology and Management of Lobsters, vol. 1. Physiology and Behavior. Academic Press. New York. Becker, C. J. 1994. The influence of prey availability and habitat structure on the distribution of lobsters (Homarus americanus) in the Great Bay Estuary). M.S. thesis. University of New Hampshire, Durham. NH. 109 pp. Briggs, P. T. & F. M. Mushacke. 1979. The American lobster in western Long Island Sound. N.Y. Fish Game J. 25:59-86. Campbell, A. 1989. Dispersal of American lobsters. Homarus americanus. tagged off southern Nova Scotia. Can. J. Fish. Aquat. Sci. 46:1842- 1844. Campbell. A. 1990. Aggregations of berried lobsters (Homarus america- nus) in shallow waters off Grand Manan. eastern Canada. Can. J. Fish. Aquat. Sci. 47:520-523. Campbell. A. 1992. Characteristics of the American Lobster fishery of Grand Manan. New Brunswick. Canada. N. Amer. J. Fish. Manag. 12:139-150. Campbell. A. & D. S. Pezzack. 1986. Relative egg production and abun- dance of berried lobsters, Homarus americanus, in the Bay of Fundy and off southwestern Nova Scotia. Can. J. Fish. Aquat. Sci. 43:2190- 2196. Campbell, A. & A. B. Stasko. 1986. Movements of lobsters {Homarus americanus) tagged in the Bay of Fundy. Canada. Mar. Biol. 92:393- 404. Charmantier G. M. Charmantier-Daures. N. Boouaricha, L. P. Thuet. D. E. Aiken & J. P. Trilles. 1988. Ontogeny of osmoregulation and salinity tolerance in two decapod crustaceans: Homarus americanus and Pe- naeus japonicus. Biol. Bull. 175:102-110. Cobb. J. S. & B. F. Phillips. 1980a. The biology and management of lob- sters, vol. I. physiology and behavior. Academic Press. New York. 463 pp. Cobb. J. S. & B. F. Phillips. 1980b. Vie biology and management of lob- sters, vol. II. ecology and management. Academic Press. New York. 390 pp. Cooper. R. A. 1970. Retention of marks and their effects on growth, be- havior, and migrations of the American lobster. Homarus americanus. Trans. Am. Fish. Soc. 99:409-117. Cooper. R. A. & J. R. Uzmann. 1971. Migrations and growth of deep-sea lobsters. Homarus americanus. Science 171:288-290. Cooper. R. A.. R. A. Clifford & C. D. Newell. 1975. Seasonal abundance of the American Lobster. Homarus americanus. in the Boothbay Re- gion of Maine. Trans. Amer. Fish. Soc. 4:669-574. Crossin. G T.. S. A. Al-Ayoub. S. H. Jury. W. H. Howell & W. H. Wat- son. III. 1998. Behavioral thermoregulation in the American lobster. Homarus americanus. J. Exp. Biol. 201:365-374. Lobster Sex Ratio 201 Dull, W. 1970. Osmoregulation in the lobster Homarus americanus. J. Fish. Res. Board Can. 27: 1 123-1 130. Estrella, B. T. & D. J. McKieman. 1989. Cateh-per-unit effort and biologi- cal parameters from the Massachusetts coastal lobsters (Homarus americanus) resource: description and trends. NOAA Tech. Repl. NMFS. vol. 81. 21 pp. Estrella, B. T. & T. D. Morrissey. 1997. Seasonal movement of offshore American lobster. Homarus americanus, tagged along the eastern shore of Cape Cod. Massachusetts. Fish. Bull. U.S. 95:466-476. Factor. J. R. (ed.l. 1995. Biology of the lobster. Homarus americanus. Academic Press, New York. 528 pp. Fogarty, M. J. & D. V. D. Borden. 1980. Effects of trap venting on gear selectivity in the inshore Rhode Island American lobster. Homarus americanus. fishery. Fish. bull. U.S. 77:925-933. Forward. R. B. Jr. 1989. Behavioral responses of crustacean larvae to rates of salinity change. Biol. Bull. 176:229-238. Hines. A. H.. R. N. Lipcius & A. M. Haddon. 1987. Population dynamics and habitat partitioning by size, sex, and molt stage of blue crabs (Callinecles sapidus) in a subestuary of central Chesapeake Bay. Mar. Ecol. Prog. Ser. 36:55-64. Houchens. C. R. 1996. A comparison of the regulatory capabilities of estuarine and coastal populations of the American lobster, Homarus americanus. MS thesis. University of New Hampshire, Durham. NH. USA. Howell. W. H. & W. H. Watson. Department of Zoology. University of New Hampshire. Durham. NH. (unpublished data). Jury, S. H., M. T. Kinnison. W. H. Howell & W. H. Watson III, 1994a. Metabolic responses of lobsters exposed to reduced salinity. J. Exp. Mar. Biol. Ecol. 176:167-185. Jury. S. H„ M. T. Kinnison, W. H. Howell & W. H. Watson III. 1994b. The behavior of lobsters in response to reduced salinity. J. Exp. Mar. Biol. Ecol. 180:23-37. Jury. S. H.. W. H. Howell & W. H. Watson III. 1995. Lobster movements in response to a hurricane. Mar. Ecol. Prog. Ser. 119:305-310. Karnofsky, E. B., J. Atema & R. H. Elgin. 1989. Natural dynamics of population structures and habitat use of the lobster. Homarus ameri- canus. in a shallow cove. Biol. Bull. 176:247-256. Krouse. J. S. 1973. Maturity, sex ratio, and size composition of the natural population of American lobster. Homarus americanus. along the Maine coast. Fish. Bull. U.S. 71:165-173. Krouse, J. S. & J. C. Thomas. 1975. Effects of trap selectivity and some population parameters on size composition of the American lobster, Homarus americanus. catch along the Maine coast. Fish. Bull. U.S. 73:862-871. Lawton, P. & K. L. Lavalli. 1995. Postlarval, juvenile, adolescent and adult ecology, pp. 47-88. In: Factor. J. R. (ed.). Biology of the Lobster. Homarus americanus. Academic Press, New York. Loder. T. C, J. A. Love, C. E. Penniman & C. D. Neefus. 1983. Long-term environmental trends in nutrient and hydrographic data from the Great Bay Estuarine system. New Hampshire-Maine. UNH Mar Prog. UNH- MP-D/TR-SG-83-6. 65 pp. Maynard. D. R. 1991. Molting and movement of lobster {Homarus ameri- canus). in and adjacent to Malpeque bay. Prince Edward Island. Canada. National Shellfisheries Association Program and Abstracts of the 83rd annual meeting. 6/23-27. Portland. ME (abstract only), p. 286. Miller. R. J. 1990. Effectiveness of crab and lobster traps. Can. J. Fish. Aquat. Sci. 47:1228-1251. Munro. J. & J. C. Therriault. 1983. Migrations saisonnieres du homard {Homarus americanus) entre la cote et les lagunes des Iles-de-la- Madeleine. Can. J. Fish. Aquat. Sci. 40:905-918. Pecci, K. J., R. A. Cooper. C. D. Newell. R. A. Clifford & R. J. Smolowitz. 1978. Ghost fishing of vented and nonvented traps. Mar. Fish. RevAO: 9-43. Pezzack. D. S. & D R. Duggan. 1986. Evidence of migration and homing of lobsters (Homarus americanus) on the Scotian Shelf. Can. J. Fish. Acpiat. Sci. 43:2206-2211. Reynolds, W. W. & M. E. Casterlin. 1985. Vagile macrofauna and the hydrographic environment of the Saco River Estuary and adjacent wa- ters of the Gulf of Maine. Hydrobiologia 128:207-215. Robichaud. D. A. & A. Campbell. 1991. Annual and seasonal size- frequency changes of trap-caught lobsters (Homarus americanus) in the Bay of Fundy. J. Northw. All. Fish. Sci. 1 1:29-37. Roddick, D. L. & R. J. Miller. 1992. Spatial and temporal overlap of the American lobster (Homarus americanus) and sea scallop (Placopecten magellanicus) as related to the impact of inshore scallop dragging. Can. J. Fish. Aquat. Sci. 49:1486-1492. Saila, S. B. & J. M. Flowers. 1968. Movements and behavior of berried female lobsters displaced from offshore areas to Narragansett Bay. Rhode Island. J. Cons. Perm. Int. Explor. Mer 31:342-351. Scarratt, D. J. 1970. Laboratory and field tests of modified sphyrion tags on lobsters (Homarus americanus). J. Fish. Res. Board Can. 27:257-264. Scaratt, D. J. & G. E. Raine. 1967. Avoidance of low salinity by newly hatched lobster larvae. /. Fish. Res. Board Can. 24:1403-1406. Skud, B. E. & H. C. Perkins. 1969. Size composition, sex ratio and size at maturity of offshore northern lobsters. U.S. Fish. Wildl. Sen'., Sci. Kept.— Fish. 598:1-10. Squires, H. J. 1990. Decapod Crustacea of the Atlantic Coast of Canada. Can. Bull. Fish. Aquat. Sci. 221. Stewart. L. L. 1972. The seasonal movements, population dynamics, and ecology of the lobster. Homarus americanus, off Ram Island, Con- necticut. Ph.D. thesis. University of Connecticut, Storrs, CT. Templeman, W. 1944. Sexual dimorphism in the lobster (Homarus ameri- canus). J. Fish. Res. Board Can. 6:228-232. Thomas, M. L. H. 1968. Overwintering of American Lobsters. Homarus americanus, in burrows in Bidelord River. Prince Edward Island. /. Fish. Res. Board. Can. 25:2725-2727. Thomas. M. L. H. & G. N. White. 1969. Mass mortality of estuarine fauna at Bideford P.E.I, associated with abnormally low salinities. J. Fish. Res. Board Can. 26:701-704. Tremblay. M. J. & M. D. Eagles. 1997. Molt timing and growth of the lobster, Homarus americanus, off northeastern Cape Breton Island, Nova Scotia. J. Shellfish Res. 16:383-394. Vetrovs, A. 1990. The distribution of lobsters (Homarus americanus) in the Great Bay Estuary. MS Thesis. University of New Hampshire. Durham. NH, USA. 88 pp. Wahle. R. A. & R. S. Steneck. 1992. Habitat restrictions in early benthic life: experiments on habitat selection and in situ predation with the American lobster. J. Exp. Mar. Biol. Ecol. 157:91-1 14. Watson. W. H., A. Vetrovs & W. H. Howell. 1999. Lobster movements in an estuary. Mar. Biol. 134:65-75. Journal of Shellfish Research, Vol. 18. No. 1. 203-209. 1999. PARALYTIC SHELLFISH TOXINS IN MUSSELS AND ALEXANDRWM TAMARENSE AT VALDES PENINSULA, CHUBUT, PATAGONIA, ARGENTINA: KINETICS OF A NATURAL DEPURATION DARK) ANDRINOLO,1 NORMA SANTINELLI,2 SILVIA OTANO,2 VIVIANA SASTRE,3 AND NESTOR LAGOS1 1 Facultad de Medicina, Universidad de Chile, Lahoratorio Bioquimica de Membranas, Departamento de Fisiologia y Biofisica, Casilla 70005, Santiago 7, Chile 'Facultad de Ciencias Naturales, Universidad Nacional de La Patagonia, 9100 Trelew, Chubut, Argentina. ABSTRACT Paralytic shellfish toxin profiles of Alexandrium tamarense (Lebour) Balech and mussels (Aulacomya ater) contami- nated by the dinoflagellate, were obtained from eight sampling stations along the Valdes Peninsula. Chubut. Argentina. The samples were collected from November 1995 to May 1996. The data show, that the monitoring began after an outbreak during a bloom of A. tamarense. The highest cell densities were found in November 1995 at Bengoa ( 1.81 x 103 cells/L) and Larralde (1.2 x 103 cells/L). both stations are located in the San Jose Gulf. Occurrence of other species of phytoplankton are also reported. A. tamarense was never more than 2% of the total phytoplankton population. Low temperatures and high salinity were found in November 1995, when the highest A. tamarense cell density was observed. Using a postcolumn derivatization high-performance liquid chromatography (HPLC) analysis, the PSP toxin profiles of Patagonian coast phytoplankton and mussel samples were obtained for the first time. The average PSP toxin profile of over 30 mussel samples from all monitoring stations showed the gonyautoxins 1-4 (GTX 1-4) epimers to be the most abundant PSP toxins. These epimers were the most prevalent ones in the A. tamarense present in the phytoplankton samples analyzed. Other PSP toxins quantified in mussel samples were: STX, dcSTX. and C1-C4. NeoSTX was never found in mussel or phytoplankton samples. The highest toxicity in the phytoplankton samples was 490 frnol of PSP toxins/cell and mussels 631 p,g STX eq./lOO g. both of which were obtained in November. 1995. The decrease of the toxicity in the filter feeder Aulacomya ater, occurs following an exponential decay of the first order, showing that, in the San Jose Gulf. Valdes Peninsula, the natural depuration process of A. ater can be interpreted by a one-compartment model. According to the detoxification rate determined for A. ater, a native South American filter-feeder bivalve, can be classified as a moderate detoxifier. KEY WORDS: PSP. depuration. HPLC toxin profiles. Alexandrium tamarense, mussel, Patagonia, Argentina. Aulacomya ater INTRODUCTION The natural depuration of filtering organisms is a process that In the southern part of Argentina, the presence of paralytic has been little studied- and toxin kinetic studies are rare (Blanco et shellfish poisoning (PSP) had been associated with the occurrence al- 1997- BricelJ and Shumway 1998). Very recently, laboratory of Alexandrium sp.. with outbreaks of PSP attributed to A. results of teedln§ studies and concomitant field monitoring of catenella in the Argentine sector of the Beagle Channel (Bena- dinoflagellate concentrations and toxicity in bivalve tissues have vides et al. 1995) and A. tamarense from the southern Atlantic been implemented in Galicias Rias, Spain (Morono et al. 1998). coast (47°S) to Uruguayan shores (34°S) (Carreto et al. 1998). The This lyPe of study could- in the future- be used t0 develop predic- first recorded toxic bloom attributable to A. tamarense on the tive relationships between water column toxin concentrations and Argentine coast was documented at the Valdes Peninsula in 1980 Peak shellfish toxicities, showing the merits and efficacy of both (Carreto et al. 1996). Since that time, the phenomenon has oc- phytoplankton and shellfish monitoring. curred periodically in the spring and summer seasons, where the Depending upon their detoxification kinetics, bivalves have PSP producer A. tamarense, which increases its density in coastal been classified into two major groups: slow detoxifiers (e.g., Saxi- waters, had been the causative source of PSP toxins contained in doimts giganteus, Spisula solidissima, Placopecten magallanicus, filter-feeder bivalves (Carreto et al. 1996, Carreto et al. 1998). and Patinopecten yassoensis) and rapid-to-moderate detoxifiers Because such filter feeders as Aulacomya ater, the "cholga." (e.g., Mytilus edulis and Mya arenaria) (Bricelj and Shumway take dinoflagellates as food and concentrate such PSP producers as 1998). A. tamarense, they constitute a public health problem, as well as The development of dynamic models linking toxic cell concen- cause damage to the commercial shellfish industry worldwide trations and toxin accumulations in the filter-feeder body could (Hallegraeff 1993, Asakawa et al. 1994, Anderson et al. 1996: provide valuable tools for predicting the timing and duration of Lagos et al. 1996; Compagnon et al. 1998). Along the Patagonian toxic blooms (Blanco et al. 1997, Bricelj and Shumway 1998). coast, one of the most economically important species of bivalves Both one- and two-compartment models have been used to de- affected by quarantines because of PSP toxicity is the "cholga," scribe the detoxification kinetics of PSP toxins in different mussel which presents its highest population density in the infralittoral species (Blanco et al. 1997, Bricelj and Shumway 1998. Morono et zone and represents a commercial harvest of about 100 tons per al. 1998). Two depuration rates have been described: the initial one year from this area. is very fast, followed by a second, slower one (Lassus et al. 1989. 203 204 Andrinolo et al. Silvert and Cembella et al. 1995; Blanco et ai. 1997. Bricelj and Shumway 1998). The study of natural toxic events is fundamental for the appropriate management of affected areas to avoid health risks from PSP toxins. This paper describes a toxic event attributable to an outbreak of A. tamarense. Because the precise chemical composition of PSP toxins produced by this dinoflagellate in this area was not known, in this study we show the quantitative high-power liquid chroma- tography (HPLC)-FDL analyses of PSP toxin contents and PSP toxin profiles of Patagonia coast phytoplankton and mussel samples for the first time. Using these data, the natural depuration kinetics of A. ater, one of the native mussel species from the Valdes Peninsula was studied. Using a one-compartment model, it was possible to predict the time when peak toxicity was attained and the detoxification rate for A. ater at the San Jose Gulf. Valdes Peninsula was obtained. MATERIALS AND METHODS Six locations around the San Jose Gulf and two more in Nuevo Gulf (Chubut. Argentine Patagonian coast) were chosen for a field- sampling program. The presence of PSP toxins was investigated in the samples of phytoplankton and mussel samples collected from each location, from November 1995 to May 1996. The mussel and phytoplankton samples were obtained from principal A. ater har- vesting areas. The phytoplankton cell counts were done by filtering 2 L of seawater in a 20-p.m phytoplankton net. The phytoplankton samples were collected at a distance about 1 m from the natural mussel banks. These banks were generally found at between 10 to 15 m depth. Also, the surface and bottom water temperatures and the salinity of the seawater were measured using a multiparameter Cole-Parmer sensor. The turbidity of the sea water was measured using a Sechi disk. All mussel samples were extracted as described by the standard AOAC mouse bioassav method (Williams 1984). Pellets of A. tamarense cells were suspended in 0.0 IN HC1 and disrupted by sonication. For clean-up, the mussel and phytoplankton samples were passed through a cartridge column (Millipore Corp., Sep- Pack C18) and filtered using microcentrifuge filters (Millipore Corp.. Ultrafree-MC filters Units, 400 p.L. NMWL: 5.000). Toxin analyses were carried out on an HPLC with on-line fluorescent detection using ion pair chromatography with postcol- umn derivatization. as described previously (Oshima 1995a, Lagos et al. 1996). For HPLC, a Shimadzu LC-10AD liquid chromato- graph apparatus, on-line with a Shimadzu RF-551 spectrofluoro- metric detector was used. A silica-base reversed phase column (Prodigy 5 p.m C8, 4.6 x 150 mm. Phenomenex, CA, USA) was used for the analytical quantification. The oxidizing reagent and acid were pumped using a dual-head pump (model SP-D-2502, Nihon Seimitsu Kagaku Co., Ltd.). Toxin concentrations were de- termined by comparing the peak areas for each toxin with those of the standard. As external standard, pure PSP toxin solutions were prepared and calibrated by HPLC-FDL and HPLC-MS used in our laboratory. The PSP toxins were purified using preparative liquid chromatography starting from high PSP-contaminated shellfish collected in the southern fjords of Chile (Lagos et al. 1996. Compagnon et al. 1998; Andrinolo et al. 1998). The phytoplankton cells were counted using the Utermohl inverted microscope tech- nique (Utermbhl 1958). The taxonomic determinations were made according to E. Balech 1977. RESULTS AND DISCUSSION Until recently, on the Argentine Patagonian coast, all monitor- ing of PSP toxins by routine survey programs or during PSP out- breaks was carried out using the mouse bioassay to determine the total toxicity in shellfish (Carreto et al. 1996. Carreto 1998). Be- cause of the annual periodicity of this phenomenon, in collabora- tion with Dr. N. Santinelli and her group (Universidad Nacional de la Patagonia), we resolved to look for a paralytic shellfish toxin Figure 1. Sampling locations in the San Jose and Nuevo gulfs, Valdes Peninsula. Bivalve and plankton samples came from the principal mussel harvesting areas. The monitoring stations for samplings were: Riacho (42°25'S, 64 °36'W), Punta Logaritmo (42 24'S. 64°29'W), Larralde (42°25'S, 64°22'W), Punta Conos (4220'S, 6404'W), Bengoa (42°16'S, 64°05'W), Punta San Roman (4215'S, 64°14'W). Punta Pardelas (42°37'S, 64 16W1 and Cerro Avanzado (42 SOS, 6453'W). Paralytic Shellfish Toxins in Patagonian Mussels and Alexandr/um tamarense 205 u g EL z CTi r i oc (N — 3 2 O — O r- •C r- sO v© O OO 9- 5 S •? B- -S 8 II =3 fc •« a tj « g I1I< NI^I s ° * -^ « -s 5 U K -a; 3 G a. « "^- ^ g. ^ .^ P a. Q a. a. C | 6, g> 5 6' S <§ s; £ 206 Andrinolo et al. producing dinoflagellate bloom in this area, with the idea of study- ing a natural bloom at the Valdes Peninsula and also for the first time, to do quantitative HPLC-FDL analyses of PSP-toxin contents and obtain the PSP toxin profiles of Patagonia coast phytoplankton and mussels. Accomplishing this, successive samplings were car- ried out between November of 1995 and May of 1996 at eight monitoring stations around the Valdes Peninsula, located on the southern Atlantic coast of Argentina (Fig. I ). The qualitative phytoplankton surveys showed mainly diatoms, such as Chaetoceros spp., with Pseudonitzschia sp. as important components of this phytoplankton. In the San Jose Gulf, during the months of February and March, a great abundance of nanoplankton was observed as containing unindentified species and cryptoficeae of the genera Rhodomonas (Table 1 1. In May 1996, also in the San Jose Gulf, dinoflagellates constituted an important fraction of the total phytoplankton population, their main components being small, unidentified dinoflagellates. In the Punta Conos monitoring station, also on the San Jose Gulf, the presence of Prorocentrum lima was detected in only one sampling. Dinophysis acuminata and Prorocentrum micans, all potentially diarrheic shellfish poi- soning (DSP) species, were norma] components of the phytoplank- ton of both gulfs (Table 2). The PSP-producing dinoflagellate A. tamarense, was detected in the quantitative phytoplankton samples at six stations during the monitoring of November 1995. in the San Jose Gulf. The highest cell densities were registered at Bengoa station with 1.8 x 10 cell/L. A. tamarense was absent during the rest of the monitoring program. The population of A. tamarense never surpassed 2Vc of the total phytoplankton cells. In contrast, during the entire moni- toring period of this study, a total absence of A. tamarense was observed in the phytoplankton samples collected at the monitoring stations in Nuevo Gulf. Nevertheless, the analyzed mussel samples collected from this gulf showed toxicities as high as 631 p.g of STX equiv./lOO gr. This PSP toxin contamination must have come from another out- break occurring at least 2 months earlier. The study of the physicochemical parameters on both sides of the isthmus of Carlos Ameghino (between San Jose and Nuevo gulfs) showed no significant differences between them. Figure 2 shows the mean variation of temperatures and salinity of seawater on the surface and at the bottom (20-m depth), as well as the depth of the Sechi disk measured in each monitoring station. The lowest temperature values were registered in November 1995, with a minimum of 10°C (early spring) and an average of 14.5°C during the summer months of 1996. The salinity varied from 40.12 ± 0.42 g/L (mean ± SEM. n = 1 1 ) in November 1995 (late spring) to 33.88 ±0.1 g/L (mean ± SEM, n = 9 1 in February 1996 (mid- summer) and remained constant during the entire summer season. The Sechi disk had a average value of 5.4 ± 0.3 m (mean ± SEM, n = 24) during the monitoring program, showing a constant tur- bidity in both gulfs during the monitoring period. The toxin profiles found in phytoplankton samples from San Jose Gulf showed that the majority of the PSP toxins present were from the gonyautoxins group (GTXs). Among these. GTX 1-4 epimers were slightly predominant. From the group of the saxitox- ins (STXs). similar quantities of STX and dcSTX were found (Fig. 3). None of the GTX 5. neoSTX. and C1-C4 toxins were detected in any of the phytoplankton samples. The average total toxicity per cell of A. tamarense was 336 ± 142 fmoles of toxins per cell (mean ± SEM. n = 7). Phytoplankton samples that showed a complete absence of A. tamarense also showed an absence of PSP toxins. The average total toxicities per cell shown by A. tamarense collected in the San Jose Gulf, are within the ranges of cell toxicity reported for other A. tamarense cells and Alexandrium strains. from other locations, including the coast of Argentina. For ex- ample. A. tamarense kept in culture in the United States showed a cell toxicity of approximately 220 fmoles/cell (66 pg STX eq./cell. Bncelj et al. 1990). Oshima 1992. reported an average toxicity of TABLE 2. Occurrence of A. tamarense and other species of phytoplankton at Nuevo Gulf (Nov. 95-May 96). Phytoplankton Nov. 95 Feb. 96 Mar. 96 Mar. 96 May 96 May 96 Community P P P CA P CA Total Phytoplankton 13.412 192.787 215.256 191.384 42.605 25.856 Alexandrium tamarense Chaetoceros sp. 50.952 110.121 12.136 830 3.616 Thalassiossira spp. Asterionella japonica 54.612 Skeletonema costatum 11.715 Cylindrotheca closterium 9.372 36.408 Pseudonitzschia pseudodelicatissima 33.968 Pseudonitzschia spp. 778 507 Dinophysis acuminata 158 452 Prorocentrum micans 38 9.372 1.911 Prorocentrum lima Gyrodinium sp. 115 78 Raphydophyceae spp. 706 2.343 452 Flagellates 1.205 Rhodomonas sp. 6.068 19.034 Phaeocyslis sp. 13.587 Dinoflagellate spp. 3.544 452 P. Pardelas; CA. Cerro Avanzado. Paralytic Shellfish Toxins in Patagonian Mussels and Alexandrium tamarense 207 <0 » 5 S 1/1 Nov. Feb March Mav Figure 2. Average physicochemical parameters at Valdes Peninsula (Nov. 95-May 96). Surface temperature (open triangles. .1): bottom temperature (filled triangles. A); Surface salinity (open squares, D); bottom salinity (filled squares. ■). and Sechi disk (tilled bars). Errors bar represents SEM (n = 7). 595 fmole/cell in cysts of A tamarense isolated in sediments from ship ballast tanks, they also showed that this amount was sixfold higher than that of the natural population of vegetative cells. In Argentina. Benavides et al. 1995 reported a toxicity of 325 pg STX eq./cell for an Alexandrium strain, which corresponds approxi- mately to 1,083 fmol/eell, to now, the highest toxic Alexandrium strain reported (Benavides et al. 1995, Bricelj and Shumway 1998). Carreto et al. (1996). describes the presence of large quantities of C1-C2 in .4. tamarense cells isolated from the Mar del Plata coast and grown in batch cultures in f/2 medium. They also showed that the GTX 1-4 epimers were the predominant ones in the gonyaulax toxins; likewise. Anderson et al. (1996) describes a high proportion of C1-C2 in A. tamarense cells collected in the South China Sea and cultured under controlled conditions. More- over. Oshima 1992 reported that A. tamarense in culture showed a larger proportion of C1-C2 than the natural vegetative cells. The absence of C1-C2 in the phytoplankton samples analyzed in this study can be explained as differences among populations of the same strain or to the possibility that A. tamarense synthesize low quantities of C1-C2 under natural conditions, so that these PSP toxins could not be detected because of the low number of cells present in the natural plankton sample extracted. The differences detected in toxin profiles and the quantities of toxins produced per cell between natural and cultured cells reinforces the idea that important differences exist in the PSP production of A. tamarense, depending upon on whether it is subject to natural or laboratory conditions. The PSP toxin profile corresponding to the filter-feeder bivalve A. ater. collected in November 1995 (Fig. 4). was practically iden- tical to the profile shown in the natural phytoplankton samples (Fig. 3), where the only PSP toxin producing dinoflagellate unam- biguously identified was A. tamarense (Lebour) Balech (Balech 1977). A decrease in the relative levels of dcSTX and the definite preponderance of GTX 1-4 epimers, were the changes observed (Fig. 4). The similarities between the PSP toxin profiles of the filter-feeder bivalve A. ater and the phytoplankton samples clearly show that the dinoflagellate A. tamarense is the source of the PSP toxin contamination in this gulf (Fig. 5). LU _l o LU _l o LU o Tt t- in n n ^ ^ i i i i |_ y} ' ' ' 'Hr-Ht-|-0§P OOOOOOOOCD-o c (O X X I- I- CO ^-cnco^xXXXXco --.X ' ' ' 'F-F-r-r-l-0gP OOOOOOOOCD-o cCO X ■*f t- ud CO Cnj C fn XXXXXfe^x t-CMCO'3-r-r-r-r-r- 0 ml— OOOOCDOCDOO-o cCO Figure 3. Average PSP toxins profile of A. tamarense, San Jose Gulf, Valdes Peninsula, November 1995 (mean ± SEM, n = 7). Figure 4. Average PSP toxin profiles of the filter-feeder bivalve A. ater in November 1995 at San Jose Gulf. Valdes Peninsula (mean ± SEM, n = 9). Figure 5. Average PSP toxins profile found in A. ater throughout the 6 months of monitoring programs (Nov. 1995-May 1996) (mean ± SEM, n = 34 profiles). 208 Andrinolo et al. By the analyses of all bivalve toxin profiles recorded during the monitoring program, it was possible to obtain an average toxin profile of the bivalve A. ater from both gulfs. Although the toxic profile preserves the main characteristics of the two previous ones (Figs. 3, 4). the profile obtained from all samples showed the presence of small amounts of N-sulfocarbamoyl-1 1 -hydroxy- sulfate toxins (C1-C4). suggesting that these PSP toxins are prob- ably produced by A. tamarense cells and can be detected only when enough of these cells are concentrated, as normally occurs in filter-feeder bivalves. The GTXs continue being the most prevalent ones. Also, in some samples, trace quantities of GTX 5 were detected. STX was always present in significant quantities (Fig. 5). These data suggest that changes in the toxin profiles would be related to the depuration process by means of specialized and specific processes of transport as well as by metabolic transfor- mations occurring inside the filter-feeders (Shimizu and Yoshioka 1981. Sullivan et al. 1983; Oshima 1995b). The monitoring program coincided with a toxic event already underway, because toxicity of the samples diminished from an average of 437.22 ± 161 |xg of STX eq/100 g (mean ± SEM. n = 9) in November 1995 to 26.00 ± 20 p,g of STX eq/100 g (mean ± SEM, n = 6) in May of 1996 (Fig. 6). This figure shows that the natural depuration process takes place in the form of an exponen- tial decay. This process lasts for about 6 months after the causative dinoflagellate A. tamarense has disappeared from the phytoplank- ton community. Natural depuration in the filter-feeder bivalve A. ater occurs in the form of an exponential decay of the first order, indicating that the data can be interpreted by following a one-compartment model. This kinetic of natural depuration can be described according to the equation C = C0 * e~Kt. where C is the total toxicity of the contaminated filter-feeder at a given moment, C0 is the initial total toxicity (maxim toxicity reached), K is the apparent first-order detoxification rate, and t is time (Bricelj and Shumway 1998). According to this equation, the calculated maximum toxicity of A. ater should have been around 561 p,g of STX eq./lOO g in October 1996. Also, according to this equation, the natural detoxification process occurs at a depuration rate of 0.017 d~'. meaning that 50% detoxification occurred every 41 days in this area, under these conditions of bloom intensity, environment, oceanography, and physiology. According to this model, the contaminated A. ater in this harvesting area should reach safe toxicity limits (under 80 p,g STX eq./lOO g) at least 4 months after November 1995, which is. in effect, what occurred. In March 1996. the average total toxicity measured by HPLC in A. ater was 25 ± 7 p.g of STX eq./lOO g. o M 3. 200 ■ ■ — Co- 796 ; v - 0.017 * t - 1 y= 796 * e - -\ R=0.94 i i i > i i i a. & <• .* N<3 >J> $■ r(> $> rf $> rg> ^\ ^> <$ n£> DaVS New algal diet S. costatum •? V v v Days Figure 2. A, B. daily mean shell valve activities (expressed as percentages of oysters displaying open valves) when fed A. minutum and then /. galbana and T. suecica or S. costatum and T. weissflogii. Parameters used in the linear model are indicated. 214 Lassus et al. TABLE 1. Results of a change of algae on parameters monitored in the raceway. Oct. 1996 March 1997 Isoehrysis Tetraselmis Skeletonema Thalassiosira Shell valve activitv Food 50.0 (<0.0001) 14.7 (.0006) 48.2K.0001) 42.2 K.0001) Slope -0.96(0.0003) -0.52 1.02) Clearance rate Food 0.21 (.011) 1.8 K.0001) 1.6 1.0004) Slope -0.01 (.01) -0.1 K.0001) -0.1 (.002) Filtration rate Food 0.26 (.02) 0.9K.0001) 1.2 K.0001) Slope -0.02 (.005) -0.1 K.0001) -0.1 (.001) Biodeposition Food 6.4 (.009) 11.1 (.001) 7.5 1.002) Slope -1.3 K.0001) -0.9K.0001) For each experimental condition and parameter, the "food" effect value (FEV) is given together with the slope value of the time series indicated below. Only significant values (p < .05) are given, with their associated probabilities shown in parentheses. Exposure to A. minutum, then to a Nontoxic Diet Shell valve activities (Fig. 2). showed: ( 1 ) variations relating to the period of total renewal of seawater; (2) a mean daily activity involving only 40 to 50% of individuals for a diet composed of A. minutum; and (3) variable mean daily activities for the different nontoxic diets experimented after exposure to A. minutum: be- tween 80 and 90% of individuals were active with diatoms, around 70% with /. galbana and only 50% with T. suecica. This kind of graphical representation is quite valid for the other ecophysiological parameters [CR, FR. and biodeposition rate (BR)]. For this reason, only the food effect and feeding trend during the detoxification phase were considered. Seawater re- newal, although integrated into the model to avoid any interference with variations obviously linked to a switch in algal diet, was not a meaningful parameter in terms of physiological variations. The results (Table 1 ) clearly show a significant "food effect" when the A. minutum diet was replaced by flagellates or any dia- tom species. The early increase in shell valve activity was less pronouned with T. suecica but detectable (food effect value: 14.7). A significant decreasing trend in shell valve activity was observed (p = .0003) with Isoehrysis, and to a lesser extent with T. suecica (p = .02), but not with any of the diatom species used. The CR for oysters was low (0.10 to 0.20 L/h"') when they were fed first with A. minutum and then with two different flag- ellate diets. When diets composed of A. minutm were replaced with diatoms, the CR differed slightly. Conversely, there was no sig- nificant difference between A. minutum and Isoehrysis diets and a very slight effect of T. suecica (FEV : 0.21 L/h"1). The results were quite similar for FR but not for BR, for which food effect was more or less marked with T. suecica, S. costatum, or T. weissflogii, but not significant with Isoehrysis. A decreasing trend in BR was observed for each of the diatom diets. Comparison Between Two Populations Fed Toxic or Nontoxic Diets The two 45 oyster populations experimented in two different raceways were expected to be identical: animals of similar length and weight were randomly distributed in each tank, kept at the same temperature, light intensity, and photoperiod and fed con- tinuously with an identical seston supply (same TPM value). Nev- ertheless, as in the S. trochoidea versus A. minutum experiment /. trochoidea A. minutum ♦ < ♦ 12 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Days Figure 3. Oyster shell valve activity with S. trochoidea and A. minutum as food supply. Parameters used in the linear model are indicated. ALEXANDRIUM M1NUTUM AND FEEDING IN OYSTERS 215 (Fig. 3), two oyster populations fed only with A. minutum were compared. The results (Table 2) show a very significant increase (as compared to A. minutum) in shell valve activity or CR and BR when oysters were fed 5. trochoidea. However, in terms of a possible "raceway" effect on oysters fed only A. minutum, it is noteworthy that shell valve activity and CR were not affected; whereas, FR and BR displayed significant dif- ferences, although less pronounced than with the Scrippsiella/ Alexandrium experiment. DISCUSSION The first results obtained for short-term experimental contami- nation (6 h in an open circuit) (Bardouil et al. 1993) showed higher clearance rates in oysters fed with S. trochoidea than and A. minu- r/im-based diet. Mean time-averaged biodeposition rates and fil- tration rates were also much higher for 5. trochoidea than for A. minutum. Moreover, for parameters not evaluated here, the authors found that the toxic nature of A. minutum seemed to account for a lower absorption of this species than that of S. trochoidea. Finally, the high absorption rate of a more toxic strain, such as A. tama- rense seemed to be attributable to the greater ability of oysters to digest this species than A. minutum. Subsequent studies (Lassus et al. 1996) also seemed to demonstrate the resumption of filtering activity (after nearly total inhibition) in oysters exposed to A. fundyense for two 6-h periods before and after overnight fasting. The results obtained here under quite different conditions (higher flow rates, seston input kept constant, long-term exposure to algal foods, larger size of oyster population) showed a signifi- cant reduction in shell valve activity as well as CR, FR, and BR when oysters were fed with toxic A. minutum. rather than the nontoxic control, S. trochoidea. Nevertheless, these results must be considered with caution, especially for FR and BR. because some "raceway" effects may interfere. When oysters were fed for 8 to 15 days with toxic alga, the reduction in shell valve activity seemed to be stable over time. The transition to a nontoxic diet led to a very marked increase in shell valve activity with both diatoms and flagellates. In fact, an obser- vation that needs to be corroborated is the apparently different feeding behavior (if shell valve activity, as defined in this study, is regarded as a good indicator of feeding activity) between oysters fed diatoms or flagellates as detoxication foods. Whereas shell valve activity remained high throughout detoxication time, regard- less of the diatom species used, a decreasing trend was observed with flagellates, particularly with Isochrysis. thus foreshadowing a return to the previous lower activity. The impact of a change of food on other physiological param- eters can vary according to the type of algal food and the involved parameter. Generally speaking, FEV are much less pronounced than for shell valve activity. The biodeposition rate, as observed in a previous study (Lassus et al. 1996), seems to be the second most sensible parameter, after an Isochrysis diet. CR and FR were only slightly affected by the change of diet, even when FEV shell valve activity was high. From a practical point of view, the linear model used here not only allows the possible effects of water renewal to be taken into account, but also provides an opportunity to consider some inter- esting performances for specific diets, especially in terms of en- TABLE 2. Summary of the results, in October 1997, of the comparison between two raceways with the same alga as food {Alexandrium minutum), or with two different algae. Oct. 1996 March 1997 Oct. 1997 Alexandrium Alexandrium A. minutum minutum minutum l Raceway 1 ) vs. Species Raceway Raceway Scripsiella trochollidea Raceway 1 vs. 2 1 vs. 2 (Raceway 2) Shell valve <0(<.0001) activity Clearance rate <0(.0005) Filtration rate <0(.004) >0(.04) <0(.02) Biodeposition <0(.002) >0(.()l) <0(.002) The sign of the difference between the two raceways is indicated, with the associated probability in parentheses, when significant (p < .05). hanced and steady feeding activity; for example, a diet with high shell valve activities, high biodeposition rates and high FEV. An- other interesting feature for the diatom diet is that shell valve activity remained roughly constant during the detoxication period. The negative slope for flagellate diets may be attributed to ( 1 ) a temporary effect on either valval activity or BR (Tetraselmis). only detectable during the first days of food change; or (2) a more interesting efficiency of diatoms in oyster detoxication. In the case of populations exposed successively to a toxic and then a nontoxic diet, it is likely that the presence of toxins in tissues affects bivalve physiology after the toxic diet is stopped and during the period of nontoxic diet exposure. Any residual traces of toxin in oyster tissues might affect feeding behavior negatively. A. minutum has been used in only one experimental study in the literature (Bricelj and Cembella 1995). Moreover, laboratory of in situ studies have generally focused on toxin distribution per organ and the kinetics of accumulation and detoxification in bivalves known to have very long retention times for paralytic toxins, as in the case of the surf clams Spisula solidissima (Bricelj and Cem- bella 1995) and Paphies subtriangulata (Mackenzie et al. 1996). Because of a scarcity of information and considering the cur- rent worldwide spread of -A. minutum (northern and southern Eu- ropean coasts, Australia, New Zealand) it is essential to have a better knowledge of the different effects of this species on com- mercial bivalve physiology, particularly for oysters. ACKNOWLEDGMENTS This study could not have been performed without the financial assistance of the Ministry of the Environment (National Program on Toxic Algal Blooms) and the Poitou-Charentes Region. We are also grateful to E. Erard-Le Denn for supplying the AM 89 BM strain of Alexandrium minutum and M. Nourry for collecting oys- ters at the Bouin station. The experiments performed in the present study comply with the current laws applied in France. LITERATURE CITED Bardouil, M.. M. Bohec. S. Bourgrier. P. Lassus & P. Truquet. 1996. Ecophysiologic feeding responses of Crassostrea gigas (Thunberg) to inclusion of different proportions of toxic dinoflagellales in their diet. Ocean. Acta. 19:177-182. 216 Lassus et al. Bardouil. M., M. Bohec. M. Cormerais. S. Bougrier & P. Lassus. 1993. Experimental study of the effects of a toxic microalgal diet on feeding of the oyster Crassostrea gigas Thunberg. J. Shellfish Res. 12:417—122. Bricelj. V. M. & A. D. Cembella. 1995. Fate of gonyautoxins in surf clam. Spisula solidissima. grazing upon toxigenic Alexandrium. pp. 413—118. In: P. Lassus. G. Arzul, E. Erard-Le Denn. P. Gentien and C. Marcail- lou-Le Baut. (eds. ). Harmful Marine Algal Blooms. Lavoisier Intercept Publish., Paris. Bricelj, V. M.. J. H. Lee. A. D. Cembella & D. M. Anderson. 1990. Uptake of Alexandrium fimdyense by Mytilus edulis and Mercenaria merce- naria under controlled conditions, pp. 269-274. In: Graneli et al. (eds.). Toxic Marine Phytoplankton. Elsevier Science, New York. Erard-Le Denn, E. 1991. Alexandrium minutum (Dinophycees). pp. 83-90. In: Sournia et al. (eds.). Le phytoplancton nuisible des cotes de France. IFREMER/CNRS Publishers, Nantes, France. Erard-Le Denn. E. & C. Belin. 1997. L'extension du plancton toxique. Alexandrium minutum, inquiete les chercheurs. IFREMER-Lettre aux medias 44:5. Hawkins, A. J. S.. R. F. M. Smith. B. L. Bayne & M. Heral. 1996. Novel observations underlying fast growth of suspension-feeding shellfish in turbid environments: Mytilus edulis. Mar. Ecol. Prog. Ser. 131:179- 190. Hawkins, A. J. S.. R. F. M. Smith. S. Bougrier. B. L. Bayne & M. Heral. 1997. Manipulations of dietary conditions for maximal growth in mus- sels, Mytilus edulis. from the Marennes-Oleron Bay. France. Aquat. Living Resour. 10:13-22. Koroleff, F. 1969. Direct determination of ammonia in natural waters as indophenol blue. ICES, CM. C:9 Hydro. Comm. Lassus, P., M. Ledoux, M. Bardouil, M. Bohec & E. Erard-Le Denn. 1994. Kinetics of Alexandrium minutum Halim toxin accumulation in mussels and clams. Nat. Toxins. 2:329-333. Lassus. P.. D. J. Wildish. M. Bardouil. J. L. Martin, M. Bohec & S. Bou- grier. 1996. Ecophysiological study of toxic Alexandrium spp. effects on the oyster Crassostrea gigas. pp. 409—112. In: T. Yasumoto. Y. Oshima and Y. Fukuyo (eds.). Harmful and Toxic Algal Blooms. IOC of UNESCO. Ledoux, M„ J. M. Fremy. E. Nezan & E. Erard-Le Denn. 1989. Recent occurrence of paralytic shellfish poisoning (PSP) toxins from the north- western coasts of France. J. Shellfish Res. 8:486 (summary only). Mackenzie, L.. D. White. J. Adamson. 1996. Temporal variation and tissue localization of paralytic shellfish toxins in the New Zealand Tuatua (surf clam). Paphies subtriangulata. J. Shellfish Res. 15:735-740. Oshima, Y. 1995. Postcolumn denvatization liquid chromatographic method for paralytic shellfish toxins. J. AOAC Int. 78:528-532. Oshima. Y.. K. Sugino & T. Yasumoto. 1989. Latest advances in HPLC analysis of paralytic shellfish toxins, pp. 319-326. In: S. Natori. K. Hashimoto & Y. Ueno (eds.). Mycotoxins and Phycotoxins '88. Elsevier. Amsterdam. Provasoli. L. 1966. Media and prospects for the cultivation of marine algae. Culture and Collection of Algae: Proceedings of the U.S.-Japan Con- ference. Hakone. Japan. Sept. 12-15. The Japanese Society of Plant Physiologists. Shumway. S. E. 1989. Toxic algae, a serious threat to shellfish aquaculture. World Aquacull. 20:65-74. Shumway. S. E. & T. L. Cucci. 1987. The effects of the toxic dinoflagellate Protogonyaulax tamarensis on the feeding and behavior of bivalve mollusks. Aquat. Toxicol. 10:9-27. Shumway. S. E., T. L. Cucci. L. Gainey & C. M. Yentsch. 1985a. A pre- liminary study of the behavioral and physiological effects of Gonyaulax tamarensis on bivalve mollusks. pp. 389-394. In: Anderson et al. (eds.). Toxic Dinoflagellates. Elsevier Science. Amsterdam. Shumway, S. E.. T. L. Cucci. L. R. D. Newell & C. M. Yentsch. 1985b. Particle selection, ingestion, and absorption in filter-feeding bivalves. J. Exp. Mar. Biol. Ecol. 91:77-92. Shumway. S. E.. J. Barter & S. Sherman-Caswell. 1990. Auditing the impact of toxic algal blooms on oysters. Environ. Auditor. 2:41-56. Ward. J. E.. L. P. Sanford. R. I. E. Newell & B. A. MacDonald. 1998. A new explanation of panicle capture in suspension-feeding bivalve mol- lusks. Limnol. Oceanogr. 43:741-752. Journal oj Shellfish Research. Vol. 18. No. I, 217-222. L999. SURVIVAL OF TOXIC DINOFLAGELLATES AFTER GUT PASSAGE IN THE PACIFIC OYSTER CRASSOSTREA GIGAS THUNBURG MOHAMED LAABIR1 2 AND PATRICK GENTIEN2 1FREMER DEL. centre de Brest BP 70-29280 Plouzane, France ABSTRACT The oyster Crassostrea gigas drastically reduced its clearance rate immediately after exposure to the toxic dinoflagel- lates Alexandrium minutum Halim (AM89BM), Alexandrium tamarense MOG835, and two toxic naked strains of Gymnodinium mikimotoi. Two nontoxic thecate dinoflagellates: Alexandrium tamarense and Scrippsiella trochoidea (M93/12) were ingested but with clearance rates below those measured for the diatom Thalassiosira weissflogu (control). All of the four thecate dinoflagellates tested survived gut passage. They were usually found intact, immobile, and viable in the feces. These cells generally constituted up to 50% of the fecal material; they recovered their motility 24 h after incubation in favorable conditions. Egestion of such viable cells may influence the recurrence and duration of toxic blooms through reinoculation of the water column. The survival of these algae to passage through the gut of C. gigas could also enhance the risk of their transfer to virgin areas through displacement of these bivalves. KEY WORDS: Crassostrea gigas. dinoflagellates. thecae, feces, clearance rate, cell viability, gut passage INTRODUCTION It has been shown that the responses of mollusks to dinoflagel- lates are species-specific (Twarog and Yamaguchi 1974, Cucci et al. 1985, Gainey and Shumway 1988. Shumway 1990, Bricelj et al. 1992). The observed differences in the reaction of bivalves to dinoflagellates have not been well clarified. Differences may be related to toxic profiles of the algae, algal concentrations, cell palatability, and size, or to differences in digestive function (Wik- fors and Smolowitz 1995). On the other hand, the physiology of oysters is known to be adversely affected by feeding on toxic dinoflagellates. especially species belonging to the genus Alexan- drium, which produce paralytic shellfish poisoning (PSP) toxins (Shumway et al. 1985, Shumway and Cucci 1987, Lassus et al. 1996. Bricelj and Shumway 1998). Reduction or complete sup- pression of clearance rate and/or biodeposition of Crassostrea gi- gas exposed to toxic strains of Alexandrium sp. has been reported (Shumway 1990, Bardouil et al. 1993, Bardouil et al. 1996, Lassus et al. 1996). Gymnodinium spp. are often associated with large marine faunal kills (Gentien and Arzul 1990, Heinig and Campbell 1992). Also, Gyrodinium aureolum has been shown to reduce C. gigas larval survival (Helm et al. 1974, Smolowitz and Shumway 1997). In this study, we determined the impact of a toxic diet on feeding habits of C. gigas. To this effect, we report here clearance rates of C. gigas when fed each of six dinoflagellates of differing toxicity. Bricelj et al. (1993) hypothesized that the sustained bivalve filtration on toxic strains would enhance their transfer from the upper layer to the bottom, a process that may influence the dura- tion and fate of a bloom in coastal waters. Until now. few studies have addressed the problem of the survival of ingested dinoflagel- lates after their gut passage in wild or cultured mollusks. important filter feeders. We reported here the fate and survival capacity for thecate and athecate dinoflagellates after gut transit in C. gigas 'Present address: Universite du Littoral cote d'Opale Laboratoire Interdis- ciplinaire en Sciences de l'Environment. ELICO-CNRS 1NSU 8013. Cen- tre de la Recherche 32, avenue Foch, BP 59. F62930 Wimereux. France. :Laboratoire d'Hydrobiologie et d'Ecologie Generale. Faculte des Sciences. Universite Mohamed ler. Oujda. Morocco. involving extracellular digestion, acidic (pH < 5) and mechanical (crystalline style) actions. Cells surviving gut transit and present in the feces could play an important role in bloom maintenance and transfer in virgin environment (Carriker 1992, Scarrat et al. 1993). METHODS Oyster Preparation and Algal Cultures Adult Pacific oysters {Crassostrea gigas) ranging in size from 8 to 10 ± 2 (SD) cm were obtained from Normandy (France) from March to early June 1997. The oyters were maintained in filtered ( 10 p.m) running seawater (salinity: 35 psu) at a constant tempera- ture of 18°C for less than a fortnight. Before experiments, animals were placed in 1-u.m filtered seawater for 24 h to clear the gut contents. Experiments were conducted with the following unialgal diets: the diatom Thalassiosira weissflogii, four thecate dinoflagel- lates including toxic strains of Alexandrium minutum (Halim) Ba- lech ( AM89BM). and Alexandrium tamarense (MOG 835, Japan), and the nontoxic strains Alexandrium tamarense (Tamar Estuary, UK) and Scrippsiella trochoidea. We also tested two naked ich- tyotoxic dinoflagellate strains, Gymnodinium mikimotoi Tinduff 87 and Gymnodinium mikimotoi Tinduff 95, isolated in Brest Bay (Tinduff). All phytoplankton cells were cultured without agitation in f/2 medium (Guillard and Ryther 1962) at 17°C on a 12-h light: 12-h dark cycle at 90 u.E nT2 s-1. Nonaxenic algal cultures were collected in the late exponential to early stationary phase (10 to 15 days after inoculation). Ledoux et al. (1990) and Erard-Le Denn (1991) give mean diameter and toxic profiles for Alexan- drium minutum. which is often responsible for PSP shellfish, con- tamination along the Brittany coast (France). Toxicity of Gymno- dinium strains related to free or esterified polyunsaturated fatty acids composition are reported in Bodennec et al. (1995) and Arzul et al. (1995). Characteristics of the species tested and final cell concentrations in the beakers are reported in Table 1 Measurement of Clearance Rate The effect of diet was evaluated by quantifying clearance rate. This parameter was measured indirectly according to Coughlan (1969) by determining the decrease of microalgal concentrations attributable to oyster filtration over time. For simple and rapid 217 218 Laabir and Gentien TABLE 1. Characteristics and final cell concentration in the heakers of the tested algae. Algal Strain Mean Diameter (um) Toxicity Final Cell Concentration (Cells/mL ') Thalassiosira weissflogii Alexandrium tamarense (UK) .Si rippsiella troclwidea Alextmdrium tamarense (MOG 835) Alexandrium minutum Halim Gymnodinium mikimotoi Tunduff 87 Gymnodirdum mikimotoi Tunduft 95 16 36 26 31 29 26-37 Nontoxic Nontoxic Nontoxic Gonyautoxins 1.2,3,4; C toxins Gonyautoxins 2,3; saxitoxins Lipidic haemolysins 18:5n3 2.892 ±421 830 ± 127 675 ± 112 850 ± 130 1.200+ 148 586 ± 79 1.100 ±96 measurements, cell concentrations were determined with a fluo- rometer (Turner Designs Co.). based on the measurement of in vivo chlorophyll fluorescence ( Brand et al. 1980). Oyster clearance rates were calculated using the following equation (Coughlan 1969) CR = Ln (Fj/F,) x M/t; where CR is clearance rate, F0 is the initial fluorescence value of ambient seawater. F, is fluorescence after time t. M is the total volume of ambient seawater, and t is the time after the start of the experiments. Strong linear relationships were established between cell concentration and fluorescence of the studied phytoplankton species. Correlation coefficient R ranged from 0.78 to 0.99 with p < .05, which allowed a routine measurement of cell concentrations using a fluorometer apparatus. In each experiment, a series of three oysters were exposed to the tested algae. Each bivalve was placed individually in aerated transparent beakers containing 200 mL of microalgal culture di- luted with 2 Liters of 1-u.m filtered seawater. Control beakers were left without animals to correct for algal cell division during ex- periments. Mean concentrations (Table 1 ) of the algae in the bea- kers fed to the oysters were close to those observed at the moment of a red tide to simulate bloom conditions (Erard-Le Denn 1991 ). No aeration was used during feeding experiments with dinoflagel- lates. When oysters were fed diatoms, the water was gently stirred periodically to avoid sedimentation of the algae. Feces and pseud- ofeces were removed at regular intervals with Pasteur pipettes. All experiments were carried out in triplicate. Data are expressed as means ± SD and statistical analysis were performed with Student's /-test. Another set of experiments gave an equivalent ratio of 10/90 in volume of tested sdga/Thalassiosira for measurement of clear- ance rate. Clearance rate of oysters exposed to T. weissflogii was defined as the control. This diatom is known to be a good diet for C. Cigas (Lassus et al. 1996). Viability Test and Alexandrium minutum Growth Experiment During this study, we used principally fluorescein diacetate (FDA), a vital stain based on the measurement of intracellular esterase, which colors viable cells green under blue light excita- tion. Accumulation of fluorescein in cells happens when nonspe- cific esterases are present in the cell, when membranes are intact i- 90%) swam normally and responded positively to vital staining (Fig. 4F). We also observed numerous empty thecae (Fig. 4D) at the bottom of the well presumably belonging to the cells that recov- ered their vitality and had shed their old thecae. When the egested A. minutum cells contained in disrupted or intact feces were in- oculated in f/2 medium, they showed normal growth and reached concentrations not significantly (p < .05) different from those of noningested control cells (Fig. 5). DISCUSSION Previous workers have shown a variety of responses of bivalves to the presence of dinoflagellates. ranging from total avoidance to normal filtration (Gainey and Shumway 1988, Shumway 1990. Bricelj et al. 1993). In our experiments, clearance rates were gen- erally close to those reported by Bardouil et al. (1993) Bardouil et al. (1996), and Lassus et al. (1996). We observed a significant decrease in C. gigas filtration in all of the tested dinoflagellates. Toxic A. minutum and A. tamarense were the most avoided spe- cies, and the oysters arrested filtration activity by complete shell valve closure within a short time after exposure. Surprisingly, the lack of toxicity of both A. tamarense (Plymouth strain) and S. trochoidea did not enhance their clearance rate by the oysters in comparison to control. Bardouil et al. ( 1993) showed that the in- clusion of even 10% of the toxic strain of A. tamarense in the algal diet of C. gigas induced a significant decrease in clearance rate equivalent to that induced by toxic A. tamarense alone. These results are confirmed here and extended to other toxic dinoflagel- late species: A. minutum and two strains of G. mikimotoi. Our work showed a high sensitivity of C. gigas to PSP-producing dinoflagel- lates but also to other types of toxins produced by species of the genus Gymnodinium. Bardouil et al. ( 1993) observed numerous intact cells in feces when oysters were fed different dinoflagellates. Shumway and Cucci (1987) reported that Protogonyaulax tamarense was filtered and rejected in pseudofeces. Bricelj et al. ( 1993) has demonstrated that cells of toxic Alexandrium fimdyense ingested by the blue mussel Mytilus edulis survived gut passage. In this study, we ob- served that, a few hours after the start of the ingestion, the diatom T. weissflogii and the two naked strains of G. mikimotoi were egested principally as ruptured cells. In contrast, all of the four thecate dinoflagellates (A. minutum. S. trochoidea. and two strains of A. tamarense) were egested intact. Just after their egestion. the cells were immobile and were not colored positively with FDA. This did not imply that these cells were dead. Less than 24 h after their egestion, up to 90%- of the cells recovered their motility. Hence, the egested A. minutum when cultured in favorable condi- tions divides normally. Further experiments must study the fate and growth of the other egested viable dinoflagellates. The nega- tive coloration with FDA of freshly egested cells might imply an arrest of their membrane exchange and/or a drastic reduction of their metabolic activity attributable to gut passage. However, these egested cells did not differ morphologically from vegetative ones: they may correspond to temporary cysts becoming active in pres- ence of favorable conditions. The resistance of the ingested di- noflagellates to gut passage may be responsible for the mainte- nance of blooms (Scarratt et al. 1993). Different oyster species exhibit very low levels of toxicity when exposed to toxic algal blooms (Bricelj et al. 1991). Lassus et al. ( 1989) reported that the level of paralytic phycotoxin accumulation in C. gigas is low as compared to that of other shellfish when fed dense A. tamarense cultures. The present work showed that C. gigas fed the tested toxic thecate dinoflagellates produced feces full of intact cells, process that may contribute to the initial detoxification, as was suggested by Bricelj and Cembella (1991) for other mollusks. ACKNOWLEDGMENTS We thank all the members of the "Proliferations Phytoplancto- niques" team for their help and relevant comments and especially Laabir and Gentien A. Youenou for her technical help. This work was supported by grant of IFREMER-DRCI as a fellowship for Dr. M. Laabir from March to June 1997. Many thanks to Dr. A. Ianora for her valuable criticism during the redaction of this manuscript. LITERATURE CITED Arzul. G.. P. Gentien. G. Bodennec, F. Toularastel. A. Youenou & M. P. Crassous. 1995. Comparison of toxic effects in Gymnodinium cf. na- gasakiense polyunsaturated fatty acids, pp. 395-400. In: Lassus et al. (eds.). Harmful Marine Algal Blooms. Bardouil. M, M. Bohec, M. Cormerais. S. Bougrier & P. Lassus. 1993. Experimental study of the effects of a toxic microalgal diet on feeding of the oyster Crassostrea gigas Thunberg. / Shell. Res. 12:417^122. Bardouil. M., M. Bohec. S. Bougrier. P. Lassus & P. Truquet. 1996. Feed- ing responses of Crassostrea gigas (Thunberg) to inclusion of different proportions of toxic dinoflagellates in their diet. Oceanol Acta. 19:177- 182. Bodennec, G.. P. Gentien. C. C. Parrish. G. Arzul. A. Youenou & M. P. Crassous. 1995. Production of suspected lipid phycotoxins by Gymno- dinium cf. nagasakiense in batch cultures, pp. 407—1 1 2. In: Lassus et al. (eds.). Harmful Marine Algal Blooms. Brand, L. E., R. R. L. Guillard & L. S. Murphy. 1980. A method for rapid and precise determination of acclimated phytoplankton reproduction rates. J. Plankton Res. 3:193-201. Bricelj, V. M.. J. H. Lee & A. D. Cembella. 1991. Influence of dinoflagel- late cell toxicity on uptake and loss of paralytic shellfish toxins in the northern cjuahog Merceneria merceneria. Mar. Eco. Prog. Ser. 74:34- 46. Bricelj, V. M.. M. Greene & A. D. Cembella. 1993. Growth of the blue mussel Mytilus edulis on toxic Alexandrium fundyense and effects of gut passage on dinoflagellate cells, pp. 371-376. In: Smayda and Shimizu (eds.). Toxic Phytoplankton Blooms in the Sea. Elsevier Sci- ence. New York. 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 virginica (Gmelin). J. Shellfish Res. 11:331-347. Bricelj, V. M. & S. E. Shumway. 1998. Paralytic shellfish toxins in bivalve mollusks: occurrence, transfer kinetics, and biotransformation. Rev. Fish. Sci. 6:315-383. Carriker. M. R. 1992. Introductions and transfers of mollusks: risk consid- erations and implications. J. Shell Res. 1 1:507-510. Coughlan, J. 1969. The estimation of filtering rate from the clearance of suspensions. Mar. Biol. 2:356-358. Cucci, T.L., S. E. Shumway. R. C. Newell & CM. Yentsch. 1985. A preliminary study of the effects of Gonyaulax lamarensis on feeding in bivalve mollusks. pp. 395^400. In: Anderson et al. (eds.). Toxic Di- noflagellates. Elsevier Science. New York. Erard-Le Denn, E. 1991. Recent occurrence of red tide dinoflagellate Al- exandrium minutum Halim from the northwestern coasts of France, pp. 85-98. In: Park. J. S. and H. G. Kim (eds.). Recent Approaches on Red Tides, Proceedings of the 1990 Korean French Seminar on Red-Tides- NFRDA. Gainey. L. F. & S. E. Shumway. 1988. A compendium of the response of bivalve molluscs to toxic dinoflagellates. J. Shellfish Res. 7:623-628. Gentien. P. & G. Arzul. 1990. Exotoxin production by Gymnodinium cf. Aureolum (Dinophyceae). J. Mar. Biol. Ass. UK. 70:571-581. Guillard. R. R. L. & Ryther. J. H. 1962. Studies of marine planktonic dia- toms. I. Cyclotella nana Hustedt and Detonula confervaceae (Cleve) Gran. Can. J. Microbiol. 8:229-239. Heinig. C. S. & D. E. Campbell. 1992. The environmental context of a Gyrodinium aureolum bloom and shellfish kill in Maquoit Bay. Maine. September 1988. J. Shellfish Res. 11:11-122. Helm. M. M.. B. T. Hepper. B. E. Spencer & P. R. Walne. 1974. Lugworm mortalities and a bloom of Gyrodinium aureolum Hulbert in the eastern Irish Sea, autumn 1971. J. Mar. Ass. UK. 54:857-869. Lassus, P., D. J. Wildish. M. Bardouil. J. L. Martin, M. Bohec & S. Bou- grier. 1996. Ecophysiological study of toxic Alexandrium spp. effects on the oyster Crassostrea gigas. pp. 409-^112. In: Yasumoto et al. (eds. I. Harmful and toxic algal blooms. IOC of UNESCO. Lassus. P., J. M. Fremy. M. Ledoux. M. Bardouil & Bohec. 1989. Patterns of experimental contamination by Protogonyaulax lamarensis in some French commercial shellfish. Toxicon 27:1313-1321. LeDoux, M., J. M. Fremy, E. Nezan & E. Errard. 1990. Recent occurrence of paralytic shellfish poisoning (PSP) toxins from the northwestern coasts of France. J. Shellfish Res. 2:486. Scarratt, A.M.. D.J. Scarratt & M. G. Scarratt. 1993. Survival of live Alexandrium tamarense cells in mussel and scallop spat under simu- lated transfer conditions. /. Shellfish Res. 12:383-388. Shumway. S. E. 1990. A review of the effects of algal blooms on shellfish and aquaculture. J. World Aquacult. Soc. 21:65-104. Shumway. S. E. & T. L. Cucci. 1987. The effects of the toxic dinoflagellate Protogonyaulax lamarensis on the feeding and behavior of bivalve mollusks. Aquat. Toxicol. 10:9-27. Shumway. S. E., T L. Cucci. L. Gainey & C. M. Yentsch. 1985. A pre- liminary study of the behavioral and physiological effects of Gon- yaulax tamarensis on bivalve mollusks. pp. 389-394. In: Anderson et al. (eds.). Toxic Dinoflagellates. Elsevier Science. Amsterdam. Smolowitz, R. & S. E. Shumway. 1997. Possible cytotoxic effects of the dinoflagellate, Gyrodinium aureolum. on juvenile bivalve mollusks. Aquacult. Int. 5:291-300. Twarog, B. M. & H. Yamaguchi. 1974. Resistance to paralytic shellfish toxins in bivalve mollusks. pp. 382-393. In: V. R. LoCicero (eds.). Proceedings of the First International Conference on Toxic Dinoflagel- late Blooms. Massachusetts Science and Technology Foundation, Wakefield. Massachusetts. Wikfors, G. H. & R. M. Smolowitz. 1995. Experimental and histological studies of four life-history stages of the eastern oyster, Crassostrea virginica. exposed to a cultured strain of the dinoflagellate Prorocen- trum minimum. Biol. Bull. Mar. Biol. Lab. Woods Hole. 3:313-328. Journal of Shellfish Research. Vol. 18. No. I. 223-226. L999. PHENOLOXIDASE ACTIVITY IN THE HEMOLYMPH OF BIVALVE MOLLUSKS LEWIS E. DEATON, PERCY J. JORDAN, AND JOHN R. DANKERT Biology Department, University of Southwestern Louisiana, Lafayette, Louisiana 70504 ABSTRACT The oxidative enzyme phenoloxidase has been proposed as a component of interna] defense in a variety of organisms. We measured the phenoloxidase activity of hemolymph plasma and hemocytes from a selection of marine and freshwater bivalve mollusks. Phenoloxidase activity is present in the hemolymph and hemocytes of Placopecten magellanicus, Argopecten irradians, Geukensia demissa, Mercenaria mercenaria, Lampsilis teres, and Lampsilis claibornensis. The enzyme activity in the freshwater species was 5 to 10 times higher than that in the marine animals. Phenoloxidase activity was not consistently stimulated by pre- incubation of hemolymph plasma or extracts of hemocytes with trypsin, bacterial cell wall lipopolysaccharides and fungal cell wall components (zymosan). The role of this enzyme, if any. in internal defense remains to be elucidated. KEY WORDS: bivalves, phenoloxidase. internal defense, hemolymph INTRODUCTION Mollusks possess a variety of mechanisms that are involved in internal defense against pathogenic organisms. The internal de- fenses consist of both cellular and humoral components (Simina and van der Knaap 1986. Bachere et al. 1995). The hemolymph of all mollusks contains a variety of hemocytes that protect the ani- mal from a wide spectrum of harmful organisms (Fisher 1986. Lopez et al. 1997). In addition, the hemolymph contains a variety of molecules that have been implicated in internal defense. These molecules include cytotoxic factors, lectins, hemagglutinins, and oxidizing enzymes (Renwrantz 1986. Leippe and Renwrantz 1988. Lopez et al. 1997). Phenoloxidase is an enzyme ubiquitous in the hemolymph of many animals; this enzyme has been proposed as a component of the internal defenses of invertebrates (Smith and Soderhall 1991 ). The activation of phenoloxidase by foreign proteins has been documented for a variety of arthropod species (Soderhall 1982. Soderhall et al. 1994). In decapod crustaceans, the phenoloxi- dase is secreted by hemocytes into the hemolymph in an inactive form; activation occurs in the presence of pathogens or pathogenic cellular components and involves the action of a serine protease on the inactive prophenoloxidase (Aspan and Soderhall 1991). In mollusks. the enzyme occurs in mantle tissues and is in- volved in shell growth and repair (Waite and Wilber 1976, Jones and Salueddin 1978). Phenoloxidase activity has also been re- ported in both hemocytes and hemolymph from several bivalves (Coles and Pipe 1994. Renwrantz et al. 1996. Asokan et al. 1997, Carballal et al. 1977, Lopez et al. 1977). The presence of phe- noloxidase in the hemolymph of mollusks and stimulation of the activity of the enzyme by bacterial and fungal cell wall compo- nents have led to interest in the possible role of the protein in internal defense. We have examined the activity of phenoloxidase in the hemolymph of several species of mollusks. Phenoloxidase occurs in both plasma and hemocytes in all of these species. The activity varies widely among the species and among individuals within a species. Neither bacterial lipopolysaccharides nor fungal cell wall extracts consistently stimulated the activity of phenoloxi- dase. MATERIALS AND METHODS Animals Specimens of the Atlantic ribbed mussel. Geukensia demissa were collected from salt marshes either in St. John's County, Florida or Jasper County, Mississippi. Scallops (Placopecten ma- gellanicus and Argopecten irradians) and clams (Mercenaria mer- cenaria) were obtained from the Marine Biological Laboratory. Woods Hole, Massachusetts. The freshwater mussels Lampsilis claibornensis and Lampsilis teres were collected from the Och- lockonee River near Tallahassee, Florida, and from drainage ditches on the USL experimental farm in Cade, Louisiana, respec- tively. The marine species were maintained in recirculating, fil- tered seawater (30 %o) at room temperature (23°C). The freshwater animals were maintained in aquaria containing pondwater at room temperature. The animals were not fed and were used within 2 weeks of collection. Collection of Hemolymph and Hemocyte Extracts Hemolymph was collected from G. demissa by prying the valves apart and inserting a syringe with a 22-gauge needle into the adductor muscle sinus. The hemolymph (usually 1-1.5 mL from each animal) was centrifuged in a microcentrifuge (Beckman mi- crofuge E) for 2 min. The supernatant plasma was removed with a Pasteur pipette, and the cellular pellet was resuspended in buffer (10 mm Tris HC1, pH 7.5) and agitated. This preparation was then centrifuged, and the supernatant (hemocyte extract) removed. Phenoloxidase Assays We compared several assays for detection of phenoloxidase activity. Initially, we used a modification of the method of Horo- witz and Shen ( 1952). This assay uses L-3,4-dihydroxyphenylala- nine (L-DOPA) as a substate. Our assay mixture contained 50 p-L of L-DOPA (3 mg mL"1). 50 p.L Tris HC1 (50 mm. pH 7.5) and 50-100 u.L enzyme preparation in the wells of a flat-bottomed 96-well microplate. We used an automated plate reader set to a wavelength of 492 nm or 6 1 5 nm to measure the absorbance in- crease caused by the formation of melanin. The L-DOPA rapidly oxidized producing a high background absorbance in the blank wells. The low activity of phenoloxidase in molluscan hemolymph necessitated long (15-60 min) incubation times, and it was diffi- cult to quantitate phenoloxidase activity consistently with this as- say. Next, we adapted the method of Pye ( 1974) to our microplate reader protocol. This assay employs 4-methylcatechol as a sub- strate; the oxidized catechol forms a colored product when com- plexed with either the methyl or benzyl ester of hydroxy proline. The 223 ::i Deaton et al. rate of spontaneous oxidation of catechol was much slower than that of L-DOPA. and we used the catechol assay with benzyl- hydroxyproline ester for all data reported here. The assay mixture contained 50 p.L Tris HC1 (50 mm. pH 7.5), 100 p.L ester. 10 p.L 4-methylcatechol. and 50-100 p.L of hemolymph plasma or hemocyte extract in the wells of a 96-well flat-bottom microplate. In blank wells, deionized water was substituted for the plasma or hemocyte extract. For the assays of preparations from marine spe- cies, the Tris buffer contained 900 mm NaCl. All components of the assay mixture except catechol were added and the absorbance at 492 nm of each well was measured with the microplate reader immediately after addition of the catechol. The mixtures were then incubated, and additional measurements to the absorbance were made after 15, 30. 60. and 120 min. The effects of trypsin, li- popolysaccharides from Vibrio cholerae, and the fungal cell wall extract zymosan were investigated by adding 10 u.L of stock so- lutions to achieve the desired concentration in the assay well. Ten u.L of water were added to the blank wells for these assays. The protein content of the hemolymph plasma and hemocyte extracts was determined by a modified Lowry assay (Miller 1950). with bovine serum albumin as a standard. Phenoloxidase activity is reported as the change in absorbance units x 101 min-1 mg pro- teirT1. Statistics Student's r-tests were used to assess whether changes in activ- ity were different from zero; we used a = 0.05. All percentage data were transformed by the arcsin transformation before statis- tical analysis. RESULTS Phenoloxidase activity occurs in the hemolymph plasma and hemocyte extracts of all of the species studied. Phenoloxidase activity in the hemolymph plasma and in the hemocyte extract is shown in Figures 1 and 2. respectively. The variability among species and among individuals of the same species is large. The specific activity is highest in both plasma and cell extract from the two unionid mussels. The concentration of protein in the plasma of the animals examined in this study ranged from 0.4 to 4 mg ml"'. .1400 O 0 M.m G.d P.m A. i Lt L.c Species Figure 1. Phenoloxidase activity in bivalve hemolymph plasma. Each bar is the mean ± standard deviation (n = 8-12). Mm = Mercenaria mercenaria; Gd = Geukensia demissa: Pm = Placopecten magellanicus; Ai = Argopecten irradians; Lc = iMinpsilis claibornensis; Lt = Lampsilis teres. ~ 200 O o Mm G.d P.m A. i L.t L.c Species Figure 2. Phenoloxidase activity in bivalve hemocyte extracts. Each bar is the mean ± standard deviation (n = 8-12). Mm = Mercenaria mercenaria; (Id = Geukensia demissa; Pm = Placopecten magellanicus; Ai = Argopecten irradians; Lc = Lampsilis claibornensis; Lt = Lampsilis teres. The effects of trypsin on phenoloxidase activity are summa- rized in Table 1 . There is no consistent stimulation of activity by trypsin. In fact, many of our plasma or hemocyte extract prepara- tions were inhibited by trypsin. This accounts for the large stan- dard deviations. Of the samples tested, only the hemocyte extract preparations from L. claibornensis were all stimulated by the trypsin treatment. We assayed the phenoloxidase activity of hemolymph plasma and hemocyte extracts from P. magellanicus after incubation with different concentrations of trypsin; the results are shown in Figure 3. There was no consistent stimulation or inhibition of these preparations in a dose-dependent manner by trypsin. The effects of increasing concentrations of the fungal cell wall extract, zymosan, and of lipopolysacharrides from the cell wall of the bacterium Vibrio cholerae on phenoloxidase activity of hemolymph plasma and hemocyte extracts from P. magellanicus are shown in Figures 4 and 5, respectively. Neither agent produced consistent stimulation of enzyme activity. As with trypsin, the activity of some preparations was stimulated and the activity of others inhibited by preincubation with these agents. None of the means of percentage change is significantly different from zero. Additional experiments with hemolymph plasma and hemocyte extracts from A. irradians produced similar results-neither zymo- san (0.01-1.0 mg mr'l nor LPS ( 1 0 3- 1 0 ' m) stimulated phe- noloxidase activity (Figs. 6, 7). Phenoloxidase activity in the two TABLE 1. The percentage change in phenoloxidase activity in bivalve hemolymph plasma and hemocyte extracts preincubated with trypsin (2.5 pg mlr'l for 30 min. Species Plasma Hemocyte Extract n Mercenaria mercenaria Geukensia demissa Lampsilis teres Lampsilis claibornensis 5 1 .9 ± 55.3 17.7 ± 124.1 -17.5 ±22.7 -10.0 ±23.2 38.0 ± 80.2 238.4 ± 243.3 106.7 ±91.1* 9.9 ± 25.8 8 8 9 7 Values are mean ± standard deviation. * Increase significantly different from 0 (p < .05). Phenoloxidase Activity in Hemolymph of Bivalve Mollusks 225 70 60 5 50 40 ? in > 20 o to 10 B 01 o> -1U r ra r O -X 40 ■so Trypsin (mg ml"1) Figure 3. The effect of increasing concentrations of trypsin on phe- noloxidase activity in hemolymph plasma (solid circles) and hemocyte extracts (open circles) from Placopecten magellanicus. Each point is mean ± standard deviation (n = 6). > n a> ° u> c ra .c O -io 107 10J 10"3 102 10' Lipopolysaccharide (M) Figure 5. The effect of increasing concentrations of lipopolysacharride from Vibrio cholerae on phenoloxidase activity in hemolymph plasma (solid circles) and hemocyte extracts (open circles) from Placopecten magellanicus. Each point is mean ± standard deviation (n = 6). species of Lumpsilis is not consistently stimulated by either zy- mosan or LPS (Table 2). DISCUSSION Phenoloxidase activity has been found in hemocytes of the marine mussels Mylilus galloprovincialis, Mytilus edulis, Perna viridis and the clam Ruditapes decussatus (Coles and Pipe 1994. Renwrantz et al. 1996. Asokan et al. 1997, Carballal et al. 1997, Lopez et al. 1997). Phenoloxidase activity is not found in hemocytes from the arcid clam Scapharca inequivalvis (Holden et al. 1994). The oxidation of L-DOPA by phenoloxidase in M. edulis hemolymph is not stimulated by trypsin and zymosan, but 0.25% zymosan supernatant stimulates the enzyme activity 10-fold (Coles and Pipe 1994). We did not find zymosan to be an effective acti- vator of phenoloxidase activity; it is possible that differences in experimental protocol account for the difference in results. It should be noted that the concentrations of zymosan and LPS reported to activate phenoloxidase in mollusks are easily 4 or more orders of magnitude higher than the threshold for activation of mammalian white blood cells. It is possible that components of 20 S? -10 T &■ > o ____- — < __ — - — o II — ra o c ■o 9 at 1 c „„ ra -10 -C O -20 0.1 1.0 Zymosan (mg ml"1) Figure 6. The effect of increasing concentrations of zymosan on phe- noloxidase activity in hemolymph plasma (solid circles) and hemocyte extracts (open circles) from Argopecten irradians. Each point is mean ± standard deviation (n = 9). 5 °s u ra o c o ra -1° I "P. Zymosan (mg ml) Figure 4. The effect of increasing concentrations of zymosan on phe- noloxidase activity in hemolymph plasma (solid circles) and hemocyte extracts (open circles) from Placopecten magellanicus. Each point is mean ± standard deviation (n = 6). 20 — 10 > ~ 0 o ra •E-10 0 U) ra -20 O -40 10J 10' Lipopolysaccharide (M) Figure 7. The effect of increasing concentrations of lipopolysacharride from Vibrio cholerae on phenoloxidase activity in hemolymph plasma (solid circles) and hemocyte extracts (open circles) from Argopecten irradians. Each point is mean ± standard deviation (n = 5-11). 226 Deaton et al. TABLE 2. Percentage change in phenoloxidase activity in hemolymph plasma and hemocjte extracts from two species of iMinpsilis preincubated with either zymosan (10_s m) or bacterial lipopolysacharrides (LPS: 0.1 ugmL"'). Species Treatment Plasma Hemocyte Extract n L. teres LPS -1.2 ±35.6 -13.2 ±43.2 8 Zymo 3.5 ± 27.3 -5.5 ±22.1 7 L. claibornensis LPS 6.9 ± 19.2 -17.8 ±37.1 9 Zymo 39.3 ± 29.2 -17.1 ±29.9 8 Values are mean ± standard deviation. molluskan pathogenic organisms will be more potent stimulators of phenoloxidase activity. Phenoloxidase activity seems to be a common feature of the hemocytes of bivalve mollusks. The enzyme is secreted into the hemolymph, and the activity varies widely among individual ani- mals. This variability may reflect the relative proportions of dif- ferent kinds of blood cells in the hemolymph of individual animals. The granulocytes of Myriltts edulis can be separated into basophilic and eosinophilic cells by cytochemistry (Moore and Lowe 1977). Phenoloxidase activity is present in granulocytes but not in baso- philic cells (Bayne et al. 1979, Coles and Pipe 1994). The propor- tion of hemocytes that exhibit phenoloxidase activity varies among individual animals (Renwrantz et al. 1996), and the proportion of granulocytic cells that contain phenoloxidase is also highly vari- able (Coles and Pipe 1994). The phenoloxidase activity of an individual mollusk may be determined by the environmental and pathological stresses that the animal has been subjected to, and may be a component of internal defense. However, unlike the phenoloxidase system in many arthropods, the activity of the en- zyme in bivalves is not consistently stimulated by trypsin or by bacterial and fungal components. A clearer understanding of the function of phenoloxidase in bivalve hemolymph depends on fur- ther experiments. It is particularly important to demonstrate that the enzyme activity is affected by natural pathogens of mollusks. ACKNOWLEDGMENTS This is contribution 32 1 from the Tallahassee, Sopchoppy, and Gulf Coast Marine Biological Association. LITERATURE CITED Asokan, R.. M. Arumugam & P. Mullainadhan. 1997. Activation of pro- phenoloxidase in the plasma and hemocytes of the marine mussel. Perna viridis Linnaeus. Dev. Comp. Immunol 21:1-12. Aspan, A. & K. Soderhall. 1991. Purification of prophenoloxidase from crayfish blood cells and its activation by an endogenous serine pro- tease. Insect Biochem. 21:363-373. Bachere. E., E. Mialhe. D. Noel. V. Buolo. A. Morvan & J. Rodriguez. 1995. Aquaculture 132:17-32. Bayne, C. J., M. N. Moore. T. H. Carefoot & R. J. Thomson. 1979. Hemolymph functions in Mytilus califomianus: the cytochemistry of hemocytes and their responses to foreign implants and hemolymph factors in phagocytosis. J. Imertebr. Pathol. 34:1-20. Carballal. M. J.. C. Lopez. C. Azevedo & A. Villalba. 1997. Enzymes involved in defence function of hemocytes of mussel Mytilus gallo- provincialis. J. Invert. Pathol. 70:96-105. Coles, J. A. & R. K. Pipe. 1994. Phenoloxidase activity in the haemolymph and haemocytes of the marine mussel Mytilus edulis. Fish Shellfish Immunol. 4:337-352. Fisher. W. S. 1986. Structure and functions of oyster hemocytes. pp. 25- 35. In: Brehelin. M. (ed.l. Immunity in Invertebrates. Springer-Verlag. Berlin. Holden. J. A.. R. K. Pipe & A. Ciani. 1994. Blood cells of the arcid clam. Scapharca inequivalvis. J. Mar. Biol. Assoc. U.K. 74:287-299. Horowitz. N. H. & S.-C. Shen. 1952. Neurospora tyrosinase. J. Biol. Chem. 197:513-520. Jones, G. M. & A. S. M. Saleuddin. 1978. Cellular mechanisms of perios- tracum formation in Physa spp. (Mollusca: pulmonata). Can. J. Zool 56:2299-2311. Leippe. M. & L. Renwrantz. 1988. Release of cytotoxic and agglutinating molecules by Mytilus hemocytes. Dev. Comp. Immunol. 12:297-308. Lopez. C. M. J. Carballal, C. Azevedo & A. Villalba. 1997. Enzyme characterisation of the circulating hemocytes of the carpet shell clam Ruditapes decussates (Mollusca: bivalvia). Fish Shellfish Immunol 7: 595-608. Miller. G. L. 1950. Protein determination for large numbers of samples. Anal. Chem. 31:964. Moore, M. N. & D. M. Lowe. 1977. The cytology and cytochemistry of the hemocytes of Mytilus edulis and their responses to experimentally injected carbon particles. J. Imertebr. Pathol. 29:18-30. Pye, A. E. 1974. Microbial activation of prophenoloxidase from immune insect larvae. Nature 251:610-613. Renwrantz. L.. W. Schmalmack. R. Redel, B. Friebel & H. Schneeweiss. 1996. Conversion of phenoloxidase and peroxidase indicators in indi- vidual haemocytes of Mytilus edulis specimens and isolation of phe- noloxidase from haemocyte extract. J. Comp. Physiol. B 165:647-658. Sminia. T. & W. P. W. van der Knaap. 1986. Immunorecognition in in- vertebrates with special reference to molluscs, pp. 112-124. In: Breh- lin. M. (ed.). Immunity in Invertebrates. Springer-Verlag. Berlin. Smith, V. S. & K. Soderhall. 1991. A comparison of phenoloxidase activity in ihe blood of marine invertebrates. Dev. Comp. Immunol 15:251- 261. Soderhall, K. 1982. Prophenoloxidase activating system and melaniza- tion — a recognition mechanism of arthropods? a review. Dev. Comp. Immunol. 6:601-611. Soderhall. K.. L. Cerenius & M. W. Johansson. 1994. The prophenoloxi- dase activating system and its role in invertebrate defence. Ann. N. Y. Acad. Sci. 712:155-161. Waite. H. J. & K. M. Wilbur. 1976. Phenoloxidase in the periostracum of the marine bivalve Modiolus demissus Dillwyn. J. Exp. Zool. 195:359- 368. Journal of Shellfish Research, Vol. 18. No. 1. 227-233. 1999. PROTEIN CONTENT DETERMINES THE NUTRITIONAL VALUE OF THE SEAWEED ULVA LACTUCA L FOR THE ABALONE HALIOTIS TUBERCULATA L. AND H. DISCUS HANNAI INO MUKI SHPIGEL, NORMAN L. RAGG,* AMIR NEORIt Israel Oceanographic and Limnological, Research, National Center for Mariculture Eilat 88112; Israel INGRID LUPATSCH, AND ABSTRACT The nutritional value to abalone of Ulva lactuca L. with different tissue nitrogen levels was studied. The seaweed was cultured at two levels of ammonia-N enrichment. Cultures receiving 0.5 g ammonia-N m~2 d"1 ("low-N") yielded 164 g m"" d"' of fresh thalli containing 12% crude protein in dry matter and 12 kJ g"1 energy; cultures receiving 10 g ammonia-N m~2 d~' ("high-N") produced 105 g of fresh thalli m-2 d~' containing 44% protein and 16 kJ g"1 energy. High-N and low-N algae and a "standard" mixed diet of 75% U. lactuca and 25% Gracilaria conferta (w/w) containing 33% protein and 15 kJ g"1 energy were fed to juvenile (0.7-2.1 g) and adult (6.9-19.6 g) Haliotis tuberculata and H. discus hamuli in a 16-week feeding trial. Voluntary feed intake of the high-N and standard diets were significantly lower than the low-N diet in all the cases. Clear differences in performance between treatments were found in the juvenile and adult abalone of both species. Juveniles fed high-N and standard diets grew significantly faster (specific- growth rate of H. tuberculata was 1.03% day"1 on high-N algae as compared to 0.72% on low-N algae; H. discus hannai grew 0.63 and 0.3% day-' on high- and low-N algae, respectively) and showed much better food conversion ratios. The nutritional value of Ulva lactuca to abalone is greatly improved by a high protein content, attainable by cultunng the seaweed with high supply rates of ammonia. KEY WORDS: integrated mariculture. abalone. seaweed, protein content, nutrition, FCR INTRODUCTION Reduced fishery landings and increasing global demand for abalone have resulted in major market opportunities for cultured abalone (Oakes and Ponte 1996). The European abalone. Haliotis tuberculata L., and the Japanese H. discus hannai Ino. are valuable shellfish (Mgaya and Mercer 1994. Oakes and Ponte 1996). stimu- lating considerable effort into the development and optimization practices of their culture. The supply of sufficient amounts of adequate food throughout the abalone's long growout phase continues to limit the develop- ment of intensive abalone culture; suitable artificial diets are gen- erally too expensive (Fleming and Hone 1996) unless fed in con- junction with macroalgae (Uki and Watanabe 1992). Traditionally, cultured abalone have been fed wild-gathered macroalgae; how- ever, the increasing intensity of abalone culture, the unreliable and seasonal nature (availability and quality) of the food harvest (Uki and Watanabe 1992. Mai et al. 1995). tougher environmental leg- islation, and the establishment of cultures in areas with no access to wild seaweed populations (Shpigel et al. 1996) have lead to increasing interest in the use of cultured macroalgae as abalone food (Tenore 1976. reviewed in Uki 1989. Marsden and Williams 1996, Shpigel and Neori 1996). The robust chlorophyte Ulva lactuca L. has been successfully adapted to vegetative tank culture (DeBusk et al. 1986). It has been a highly effective biofilter species, removing dissolved nutrients from mariculture effluents and sustaining rapid production of sea- weed biomass (Tenore 1976. Vandermeulen and Gordin 1990. Cohen and Neori 1991, Neori et al. 1991. 1996). Cultured and wild Ulva lactuca have been successfully used as diets for abalone (Tenore 1976. Shpigel et al. 1996); but to obtain commercially acceptable abalone growth rates, it has been necessary to feed other algal species in conjunction with U. lactuca (Mercer et al. ♦Present address; Department of Zoology. University of Canterbury. Pri- vate Bag 4800, Chnstchurch, New Zealand. tCorresponding author. 1993. Stuart and Brown. 1994). Given the worldwide distribution, ease of culture, biofiltration potential, and high productivity of Ulva lactuca, combined with its high palatability to abalone (Stuart and Brown 1994). learning to manipulate the dietary value of this seaweed to the nutritional requirements of abalone and understand- ing these requirements are of considerable value. The availability of suitable quantity and quality of dietary pro- tein are considered to be a prime factor governing the growth of abalone fed natural diets (Mai et al. 1995. Britz 1996). Previous studies have noted that the relative amounts of tissue nitrogen, predominantly in the form of protein and free amino acids (Duke et al. 1989b, Pedersen 1994), can vary in Ulva spp.. depending upon the alga's growth conditions and the availability of inorganic nitrogen in the growth medium (DeBusk et al. 1986. Duke et al. 1989a. Vandermeulen and Gordin 1990. Cohen and Neori 1991, Neori et al. 1991). However, such studies have not considered possible subsequent effects upon the macroalgivores. Attempts to manipulate the chemical composition of a seaweed by adjusting the culture environment have so far been restricted to macroalgae of direct commercial value, such as the agarophyte genus Gracilaria spp. (e.g., Lapointe and Ryther 1979, Lignell and Ped- ersen 1987). Some important progress has, however, been made in enhancing the dietary value of certain phytoplankton species fed to bivalves, by adjusting the nutrient composition of the algal culture medium (Engirt et al. 1986. Herrero et al. 1991 1. The present study was established with two main objectives: ( 1 ) to develop Ulva lactuca cultures of substantially different and stable tissue nitrogen levels; and (2) to determine the nutritional values of these U. lactuca cultures for the abalone Haliotis tuber- culata and H. discus hannai. MATERIALS AND METHODS Abalone Fifty juvenile European Abalone, Haliotis tuberculata (18.2- 24.7 mm shell length, 0.8-2.1 g individual live weight; 76-89 g total biomass) or Japanese Ezo Awabi, H. discus hannai (17.9- 227 228 Shpigel et al. 24.4 mm, 0.7-1.8 g; 56-63 g biomass) were stocked to 28-L aquaria. Each aquarium contained two half-pipe shelters, and re- ceived 50 volume exchanges d~' of 10 (j.m filtered seawater at 21.7-23.1°C. An airline suspended in the center of each aquarium supplied vigorous aeration to circulate food material, which was retained by a 1-mm mesh covering the outlet pipe. Aquaria were also established with adult animals: 10 H. tuberculata (37.0-50.2 mm, 6.9-19.6 g each: 123-125 g total biomass) or 10 H. discus hannai (34.0-50.8 mm, 5.0-16.1 g; 79-81 g biomass) individually identified by Dymo™ tags. All animals had previously received a mixed diet of U. lactuca and G. conferta. Diets and Algal Cultures Three distinct live seaweed diets were produced for the aba- lone. Two were monospecific Ulva lactuca, grown at two different levels of nitrogen enrichment. A "standard" diet consisted of a mixture of medially enriched U. lactuca together with Gracilaria conferta. This mixed diet had been found in our preliminary ob- servations to support good growth in weaned abalone juveniles Macroalgal cultures were established in November 1995 in the land-based facilities of the National Center for Mariculture. Eilat, Israel. The cultures were supplied with 5-6 volume exchanges d_1 of 10-p.m filtered water pumped from the Red Sea at 20-m depth (41 ppt, 19.5-25. 3°C). Vegetative Ulva lactuca thalli. isolated from the Red Sea as in Vandermeulen and Gordin (1990), were grown in lnr. 600-L tanks agitated with vigorous aeration, the technique has been described in Vandermeulen and Gordin ( 1990) and in Cohen and Neori ( 1991 ). Inorganic nutrients were added to the media in a concentrated solution, containing disodium phos- phate ( DSP, at a flux of 0.6 g P m~2 d" ' ). and ammonium sulphate (at fluxes determined by the experimental treatment). The solution was dripped into the cultures over a 4-h period every morning, this being considered ample time for Ulva spp. to take up its daily ammonia-N requirement (Fujita et al. 1988). The ammonia-N was added at two levels. 0.5g N m~- d~' ("low-N" U. lactuca culture) and 10 g N nT; d"1 ("high-N" culture). These levels were chosen based on Cohen and Neori ( 1991 ), who suggest that these N-fluxes would produce U. lactuca of considerably different tissue-nitrogen levels, while sustaining sufficient production to permit harvesting for feeding. Several cultures of high-N and low-N U. lactuca were established. Every week, the vessels were emptied and cleaned, the algae were centrifuged (500 r.p.m. for 3 min) to remove surface water, weighed to assess biomass production, and restocked at the original density. The mixed standard diet was obtained by harvest- ing U. lactuca and Gracilaria conferta grown in simulation bio- filters (Cohen and Neori 1991) receiving 0.6 g DSP m"2 d"1 and 4 g ammonia-N m- d . Feeding and Growth A preliminary test was conducted to determine whether the N content of Ulva lactuca was decreasing after days of immersion in the abalone tanks. High-N and low-N U. lactuca were stocked to separate aquaria without animals. Subsamples were removed after 0. 12. 18, 24, and 48 h, rinsed in deionised water, freeze dried, and the nitrogen content was determined. High-N and low-N ulva lactuca dietary treatments were sup- plied to triplicate aquaria with juveniles and duplicate aquaria with adult abalone of either species; in addition, a single aquarium for each abalone size class and species was fed the mixed diet. Ani- mals were first stocked to the experimental aquaria on 8 February 1996 after individual live weights (following 2 min drying on absorbent paper ± 0.01 g) and shell lengths had been measured (±0.05 mm). An additional 25 juveniles and 10 adults were taken from the source populations of either species, weighed and held without food for 48 h; the soft body and shell were then separated and freeze dried. Food algae were added to the aquaria in excess (equivalent to approximately 20<7r of the resident abalone biomass) at dusk and removed 16 h later, because abalone are assumed to show minimal daytime feeding activity (Barkai and Griffiths 1987. Uki and Wa- tanabe 1992, Mgaya and Mercer 1994). This period was consid- ered representative of total daily feeding. Feed intake rate assess- ments began following a 2-week acclimation to the diets; har- vested algae were centrifuged (as described above), and a known weight (± 0.01 g) was supplied to each aquarium. The mixed diet consisted of U. lactuca and G. conferta offered in a wet weight ratio of 3:1. It was assumed that the abalone showed no preferen- tial feeding behavior. Uneaten algae were collected by siphoning aquarium contents through a 1-mm mesh, allowing feces and de- tritus to be washed out; collected algae were centrifuged and weighed. A control aquarium, identical to the experimental vessels, but without animals, was supplied with the effluent water of a ran- domly selected abalone aquarium from each treatment and was stocked with algae corresponding to the dietary treatment being offered. Change in algal wet weight was assessed, as described above, and the mean percentage weight change was calculated for each treatment (three dietary treatments, juveniles and adults, total n = 6) and used as a correction factor (C) of the initial weight of algae fed (see in Definitions, below). On the occasion of each feed intake rate trial, algal samples were taken from each food-stock culture to produce samples of approximately 20 g wet weight per fortnight of each diet. The samples were weighed, freeze-dried. and reweighed (± 0.01 g) to determine water content. The three dried samples of each treatment collected during each 2-week period were combined and stored at -20°C for subsequent analysis of chemical composition. At the end of each trial, abalone were removed from the ex- perimental vessels, and individual weights and lengths were re- measured; all of the 10 adult abalone and 25 randomly selected juveniles from each vessel were then shucked and freeze dried for subsequent assessment of condition and soft body composition. The juveniles were not tagged. Thus, growth (in weight and length) in each aquarium was estimated by the difference between the average values of the population at the beginning of the ex- periment and the values of 10 randomly selected juveniles at the end of the experiment. Analytical Procedures The freeze-dried samples of algae and bodies of the abalone were homogenized in a mill before being subjected to analyses. Water content was calculated by weight loss after 24-h drying at 105°C. Crude protein was measured using the Kjeldahl technique and multiplying N by 6.25. Crude lipid was measured after ehlo- roform-methanol extraction (Folch et al. 1957). Samples were homogenized with a high-speed homogenizer for 5 min. and lipid was determined gravimetrically after separation and vacuum dry- ing. Crude carbohydrate was determined using the phenol- sulphuric acid method (Dubois et al. 1956) after boiling the sample in IN H2S04 for 1 h. The resulting color was measured by spec- trophotometer against a glucose standard at 490 nm. Ash-free dry weight was calculated from the weight loss after incubation of Ulva Lactuca Nutritional Value to Abalone 229 samples for 24 h at 550°C in a muffle furnace. Heat of combustion was measured in a Parr bomb calorimeter using benzoic acid as a standard. Definitions Net feed intake ( I ) was determined for each aquarium accord- ing to the equation I = (A1N x C) - AOUT where AIN and AOL,T represent, respectively, the measured weights of algae placed in. and removed from an aquarium and C is the treatment-specific correction factor used to compensate for endog- enous changes in fresh weight. Daily feed intake was assessed in this way at 3 to 4 day intervals. Specific growth rate (SGR%, %/d) = 100 ■ OnWrlnW0)A ( W0 is the weight of an animal at the beginning of each monitoring interval, and Wt is the weight after t days of growth at the end of the interval): Shell growth (u.m/day) = (LrL,)/t (L, is the wet length of an animal at the beginning of each moni- toring interval, and L2 is the length at the end of the interval); Condition index (CI) = soft flesh (g wet (/shell (g wet) Feed intake rate (mg algae/g abalone/day) = I/abalone standing stock Food conversion ratio (FCR) = total feed intake (g wet)/total weight gain (g wet) Protein productive value (PPV) = 100 • protein gain (g)/protein consumed (g) Energy productive value (EPV) = 100 ■ energy gain/gross energy consumed Statistical Analyses The responses of each abalone species and varied sizes to the various diets were analyzed separately. Growth and condition pa- rameters examining the response to the treatments of individual animals were compared by analysis of variance ( ANOVA) (Sokal and Rohlf 1995) and Duncan multiple range test. Feed intake and FCR parameters were analyzed for entire aquarium populations using f-test. All analyses were carried out with SPSS software. RESULTS Algal Production and Composition By the second week of culture, Ulva lactuca production had stabilized in all low-N and high-N cultures. From November 1995 to the end of feeding trials in May 1996, low-N cultures yielded a mean of 164 ± 6 g (SE) fresh U. lactuca trT2 d_I (n = 57). High-N cultures yielded 105 ± 7 g rrf2 d"1 (n = 38) and showed evidence of "perforation disease." described by Colorni (1989). High-N thalli were also considerably darker than low-N, but morphologi- cally similar, flat, and sheet-like. Nitrogen (expressed as crude protein), energy, water, and ash content all remained stable in the algal samples collected during the feeding trials. Water content was similar in all three diets, but the ash content of high-N U. lactuca was 30-35% lower than that of the low-N or mixed diet (Table 1 ). The mixed and high-N diets had similar calorific values, but the mixed diet had 25% less crude protein. Both the energy and the nitrogen content of low-N U. lactuca were low as compared to the other two diets. Samples of high-N U. lactuca placed in nonenriched seawater (data not shown) lost approximately 18% of their tissue N in 24 h. falling from 5.8 ± 0.08% N in dry tissue to a stable 4.8 ± 0.04% N (n = 3); whereas. low-N tissue nitrogen remained constant over 48 h. Abalone Feed Intake and Performance Juvenile Abalone After 15.5 weeks, mean growth rate of juvenile Haliotis tuber- culata, expressed in terms of SGR% or shell length increment, was significantly higher in aquaria receiving high-N diet as compared to those fed low-N diet (Table 2). Growth in the single aquarium that continued to receive a standard mixed diet was significantly faster than in the monospecific Ulva lactuca treatments. It was also noted that new shell growth during the feeding trials appeared light green in juvenile H. tuberculata fed low-N U. lactuca, in contrast to the characteristic red-brown shell increments of those animals fed high-N or mixed diets. Voluntary feed intake as apparent daily intake (mg alga per g of abalone biomass) also showed significant differences between the two U. lactuca treatments, with the feed intake rate of high-N U. lactuca being 68% that of low-N. The combined effects of relatively low feed intake rates and fast growth of juvenile H. tubeculata fed high-N U. lactuca resulted in a significantly lower (i.e.. more efficient) FCR in the high-N treat- ment. Although the growth and feed intake rates in juvenile Haliotis discus hannai after 16 weeks (Table 3) were considerably lower than in juvenile H. tuberculata (Table 2), similar patterns were apparent for the two species: overall feed intake rates of high-N algae were significantly lower (p < .001 ) than for low-N, and use of high-N or mixed diets resulted in significantly higher (p < .01) SGR% than with low-N U. lactuca. However, only the mixed diet treatment sustained significantly greater shell length growth, when compared to low-N, in H. discus hannai. Feeding high-N diet to juvenile H. discus hannai also resulted in a significantly more efficient (p < .001) FCR. However, only the mixed diet produced a significant effect on the condition (wet flesh: wet shell weight, p < .001 ) of juveniles of both species (Tables 2. 3), as compared to the high-N and low-N treatments. The loss of condition in juvenile H. discus hannai fed low-N seaweed was clearly apparent during the experiment as a progressive atrophying ("withering") of the foot muscle: two of the most reduced individuals in one replicate eventually died, these representing the only mortalities during the course of the experiment. TABLE 1. Composition of the three experimental diets, low ti-Ulva, high-N Viva, and the control diet intermediate N-(7ra and Gracilaria in relation of 3:1 (w/w). Low N-C/ia High N-Ulva Control Dry matter Crude protein Carbohydrate Lipid Ash Gross energy (kJ/g) 13.41 ±2.37 1.66 ±0.29 3.62 ±0.67 0.14 + 0.03 4.49 ±0.86 1.60 ±0.29 16.86 ± 1.79 7.38 ± 1 .02 4.05 ±0.51 0.13 ±0.03 2.91 ±0.31 2.63 ± 0.40 16.69 ±2.72 5.54 ± 1.09 3.81 ±0.87 0.14 ±0.04 4.23 ±0.91 2.43 ± 0.43 Average values during the whole experimental period are given and com- ponents are expressed as % of fresh weight (±SD). 230 Shpigel et al. TABLE 2. Growth of juvenile Haliotis luberculata fed three algal diets (108 days). TABLE 3. Growth of juvenile Haliotis discus hannai fed three algal diets (112 days I. Low-protein High-protein Low-protein High-protein Viva Viva Control Ulva Viva Control Initial wt (g) 1.74 1.57 1.71 Initial wt (g) 1.205 1.172 1.132 ± 0.258 ±0.176 ± 0.236 ±0.148 ±0.013 ± 0.082 Final wt (g) 3.8 r 4.79" 6.50c Final wt (g) 1.680" 2.367" 2.571" ±0.880 ± 1.678 ± 1.805 ±0.601 ± 0.650 ± 0.785 SGR%' 0.725" 1.028" 1.234c SGR%' 0.296" 0.627" 0.732" ±0.216 ±0.288 ±0.164 ±0.201 ± 0.267 ± 0.262 Shell growth (u.m/day) 80.72" 121.47" 160.30c Shell growth (u.m/day) 31.70" 44.47"" 54.93" ± 20.38 ± 33.9 ±35.8 ± 22.75 ±25.61 ±21.30 Feed intake (nig algae/g 127.53" 86.19" 98.60 Feed intake (mg algae/g 85.0" 32.97" 44.23 abalone/day) ± 11.07 ±0.77 abalone/day) ± 18.22 ±4.6 FCR2 18.17" ± 1.30 7.81" ± 0.794 7.70 FCR2 31.50" ± 5.83 5.54" ± 1.12 5.93 Condition index3 0.586" 0.570" 0.682" Condition index' 0.42a 0.45"" 0.48" ± 0.073 ± 0.077 ± 0.088 ± 0.069 ± 0.070 ±0.08 PPV (p)4 24.81" ±2.69 15.06" ± 1.31 22.71 PPV C, i4 9.15" ±7.08 20 4" ± 4.93 22.32 EPV (%)5 11.17" ± 1.34 13.38" ± 1.11 15.58 EPV r; I5 5.99a ± 1.21 19.59" ± 1.28 17.70 Average value ± SD. 1 Specific growth rate = (InWVlnW, (/days* 100. 2 Food conversion ratio = feed intake (g wetl/weight gain (g wet). 3 Soft flesh (g wet)/shell (g wet). 4 Protein productive value = protein gain (g)/crude protein consumed (g) *100. 5 Energy productive value = energy gain (kJ I/gross energy consumed (kJ) * 1 00. "•"'c Values with the same superscript are not significantly different (p < .05) using ANOVA and Duncan multiple range test (SPSS). Average value ± SD. 1 Specific growth rate = (lnW2-lnW1)/days*100. 2 Food conversion ratio = feed intake (g wetl/weight gain (g wet). 3 Soft flesh (g dry)/shell (g dry). 4 Protein productive value = protein gain (g)/crude protein consumed (g) *100. 5 Energy productive value = energy gain (kJ)/gross energy consumed (kJ) *100. "" Values with the same superscript are not significantly different (p < .05) using ANOVA and Duncan multiple range test (SPSS). Juvenile Haliotis tuberculata fed low-N seaweed made more efficient use of the ingested protein (PPV) compared to individuals receiving high-N treatment (Table 2). The reverse was observed in juvenile H. discus hamuli, where significantly less of the protein ingested as low-N diet was incoporated into abalone tissue, as compared to high-N diet (Table 3). On the other hand, the juve- niles of both species utilized energy (EPV) better in the high-N diet, but only in H. discus hannai was this trend statistically sig- nificant. Adult Abalone Adults of both Haliotis tuberculata and H. discus hannai grew significantly better when fed with the mixed and high-N diets than with the low-N diet (Tables 4. 5, respectively). Adults of both species also voluntarily ate significantly more low-N Ulva lactuca, as compared with the high-N seaweed. As a mathematical conse- quence of these two observations, the FCRs for both species fed high-N U. lactuca were significantly and strikingly better (lower) than in the animals fed the low-N U. lactuca. In H. tuberculata the reduction in FCR by feeding high-N seaweed was by 64% and in H. discus hannai by 77% (Tables 4, 5 ). DISCUSSION Performance in Response to Dietary Treatment Culturing Ulva lactuca at high- and low-ammonia fluxes yielded thalli that were of considerably different nutritional value to both abalone species. The animals fed N-enriched seaweed sub- sequently grew significantly faster, while consuming significantly less seaweed than the animals fed N-deprived seaweed. The above observations agree with the broad principles of herbivorous graz- ing. Feed intake rate is the main compensatory mechanism for diet quality in herbivores (Bowen et al. 1995). including abalone (Koike et al. 1979. Mgaya and Mercer 1994). Nevertheless, when fed low-quality food, a herbivore feeding to capacity may still be undernourished (White 1978). Bowen et al. (1995) and Britz ( 1996) suggested that dietary energy content also regulates abalone feed intake rate. In the present study, differences in feed intake rates of the abalone juveniles correlate numerically more closely with differences in energy content of Ulva lactuca than with its N content. Low-N and high-N feed intake rates are separated by a factor of 1 .5 in juvenile H. tuberculata and 2.6 in H. discus hannai: high-N U. lactuca has 1 .6 x energy and 4.4 x N of the low-N U. lactuca. Therefore, at least the H. tuberculata juveniles eating both diets had about the same energy intakes but very different N in- takes. These results suggest that in addition to crude protein con- tent, different energy contents in the U. lactuca diets may contrib- ute to the observed differences in feed intake. To corroborate this explanation, it would be necessary to assess the gut capacity of the abalone to determine whether the animals were simply feeding to capacity. Abalone somatic growth is considered to depend upon the amount of crude dietary protein (Uki and Watanabe 1992, Fleming 1995b. Mai et al. 1995, Britz 1996). Maximum abalone growth Ulva Lactuca Nutritional Value to Abalone 231 TABLE 4. Growth of adult Haliotis tuberculoid fed three algal diets (106 days). TABLE 5. Growth of adult Haliotis discus hannai fed three algal diets (106 days). Low-protein Viva High-protein Viva Control Low-protein Ulva High-protein Viva Control Initial wt (g) 12.56 ±3.17 12.26 ±4.49 12.34 ± 3.73 Initial wt (g) 7.81 8.06 7.98 Final wt (g) 15.47-' 16.05a 18.27a ± 2.22 ± 3.53 ± 3.94 ± 3.52 ±5.15 ± 5.64 Final wt (g) 9.16a 10.47a 11. 24" SGR%' 0.202a 0.27 1J" 0.371" ± 3.05 ±3.98 ±5.71 ±0.107 ±0.193 ±0.154 SGR%' 0.1 43" 0.264" 0.337" Shell growth (u,m/day) 43.95a 53.77" 59.88a ±0.109 ±0.163 ± 0.227 ± 16.33 ± 23.09 ± 23.06 Shell growth (u.m/day) 17.68" 29.13h 43.18c Feed intake (mg algae/g 77.0" 36.16b 36.15 ± 12.61 ± 18.14 ±24.3 abalone/day) ±2.61 ±0.94 Feed intake (mg algae/g 54.15" 20.82b 19.01 FCR2 39. 1SJ 14.20b 9.71 abalone/day) ±4.17 ± 2.32 ± 1.7 ± 3.96 FCR2 36.2a 8.4" 5.8 Condition index' 0.54J 0.51" 0.52a ±6.76 ± 1.24 ±0.106 ± 0.084 ±0.102 Condition index' 0.43" 0.46a 0.50a PPV (%)4 -2.0" 2.32" 7.79 ±0.086 ±0.15 ±0.104 ±9.87 ±0.24 PPV (%)4 -3.53" 9.58" 24.84 EPV (%)5 2.85" 1.91" 4.59 ±7.21 ±6.66 ±1.91 ± 1.94 EPV (%)5 -0.77a ± 3.34 4.19J ±2.85 14.27 Averape value + sn 1 Specific growth rate = (lnWrlnW,)/days*100. 2 Food conversion ratio = feed intake (g wet)/weight gain (g wet). 3 Soft flesh (g dry)/shell (g dry). 4 Protein productive value = protein gain (g)/crude protein consumed (g) *100. 5 Energy productive value = energy gain (kJ)/gross energy consumed (kJ) *100. a" Values with the same superscript are not significantly different (p < .05) using ANOVA and Duncan multiple range test (SPSS). was typically attained from diets of 35% protein (by dry weight; see in Uki 1989. Uki and Watanabe 1992, Mai et al. 1995, Britz 1996). These findings of previous researchers are corroborated in the present study by the rapid growth and efficient energy use (EPV) of abalone juveniles fed high-N (44% protein) Ulva lactuca, as compared to the apparent N-limitation in low-N (12% protein) fed abalone. The substantially more efficient food conversion ra- tios with high-N diet as compared with low-N diet in juveniles and adults of both species are typical of herbivorous nutrition (Mattson 1980). This has been shown to apply to juvenile Haliotis discus hannai with dietary crude protein up to 28-30% of dry matter (Uki 1989, Uki and Watanabe 1992) and to at least 47% for H. midae (Britz 1996). Britz ( 1996) and Uki and Watanabe (1992) found that the efficiency with which ingested protein was utilized for growth (PPV) in abalone increased with decreasing dietary protein level, as corroborated here by the significantly higher PPV of juvenile H. tuberculata fed low-N U. lactuca. The reverse is seen in juvenile H. discus hannai. It is suggested that the loss of condition of low-N fed H. discus hannai: that is. reduction in the soft tissue fraction of total dry weight, resulting in low (occasionally negative) levels of protein utilization, has caused this anomaly. Growth rates of Haliotis discus hannai in the present study were lower than in H. tuberculata receiving the same diets. Faster H. discus hannai growth has been recorded elsewhere (e.g., Uki 1989 reports shell growth of up to 270 p.m d_1 for young juvenile H. discus hannai grown at similar temperatures to those used here). However, this study concurs with the results of Mercer et al. (1993). who recorded H. discus hannai growth to be consistently Average value ± SD. 1 Specific growth rate = (lnW2-lnW,)/days*100. 2 Food conversion ratio = feed intake (g wet)/weight gain (g wet). 3 Soft flesh (g dry (/shell (g dry). 4 Protein productive value = protein gain (g)/crude protein consumed (g) *100. 5 Energy productive value = energy gain (kJ)/gross energy consumed (kJ) *100. *■ Values with the same superscript are not significantly different (/> < .05) using ANOVA and Duncan multiple range test (SPSS). inferior to that of H. tuberculata in comparative feeding experi- ments. They also found, as we have, that for optimal growth H. discus hannai required higher diet protein content than H. tuber- culata. Value of Ulva lactuca as a Dietary Alga for Abalone The adults of both abalone species showed similar feeding behavior, with mean feed intake of high-N being less than half that of low-N Ulva lactuca. Food conversion ratios of Haliotis discus hannai adults resemble those found for juveniles, becoming more efficient in hign-N fed animals. Spawning activity was occasion- ally observed in adult H. tuberculata vessels, affecting individual feeding rates (as in Mgaya and Mercer 1994). as well as flesh weight and shell deposition (as in Mercer et al. 1993), and causing a negative protein utilization at the low-N treatment. It seems that if high-N and low-N Ulva lactuca were offered as diets to abalone throughout the growout period, the high-N diet would sustain significantly faster growth in both species. Using high-N Ulva, both species can sustain high mean growth rates with less than half the food intake. However, previous studies have implied that the extrapolation of short-term abalone feeding trials may be unrepresentative. Day and Fleming ( 1992) found that aba- lone fed monospecific algal diets stopped growing after 50 to 200 days. They suggested it was unlikely that a single algal species could supply all essential nutrients. The performance of juvenile abalone fed the mixed diet sug- gests that using only gross N and energy measurements to assess 232 Shpigel et al. the value of an algal diet has its limits. The values of gross N and energy in the mixed diet were intermediate between those of low-N and high-N Ulva lactuca, but abalone performance with the mixed diet was usually superior to that produced even by high-N diet. Problems also arise when using only energy and crude protein to assess relative nutritional values of a single algal species. In the current study, the N and energy content of low-N and high-N U. lactuca at the time of stocking to the abalone vessels is used as an indication of dietary value for the animals. These parameters may not be representative of the relative amounts that are ultimately available to the abalone. Ulva sp. in a rich medium where N does not limit growth will carry out luxury N uptake, storing excess N in intracellular pools of organic N, predominantly amino acids, and NH* (Fujita et al. 1988, Lundberg et al. 1989). Pedersen (1994) notes that if Ulva spp. is transferred from an N-rich medium to an N-starved medium, there is a fall in tissue N. as observed in the present study when U. lactuca is stocked to the abalone vessels. The NHj pool is highly soluble and considered physiologically impossible to maintain if external concentrations fall (Fujita et al. 1988). The release of intracellular NHj by U. lactuca has been recorded by Vandermeulen and Gordin ( 1990) and is likely to be the cause of decline in high-N nitrogen content found in the current study. The quality of the digestible protein, expressed in terms of amino acid composition, may also vary between low-N and high-N Ulva lactuca. The amino acid profile of an alga may vary, depend- ing upon the level of N available (Lignell and Pedersen 1987, Miyashita and Miyazaki 1993), hence affecting the dietary value of the alga for abalone (Mai et al. 1995). Other components of the algal composition might affect its nutritional value and vary ac- cording to the seaweed's nutrient status. For example, N-starved Ulva spp. has increased levels of high energy soluble carbohy- drates (cf. DeBusk et al. 1986); whereas. Mercer et al. (1993) noted considerable variation in total lipid and carbohydrate levels in wild gathered U. lactuca used as an experimental abalone diet. Morphologically simple, fast-growing opportunistic seaweeds that lack specific chemical defenses, such as Ulva spp., Entero- morpha spp., and Porphyra spp., are considered to be the most palatable taxa for abalone (Stuart and Brown 1994, Fleming 1995a). Such algae also tend to show the greatest range in tissue-N levels in field-gathered specimens and in culture (Kudoh 1987, Bjomsater and Wheeler 1990, Wheeler and Bjomsater 1992). Con- siderable intraspecific variations in tissue N have also been noted in other macroalgal species commonly fed to abalone, such as Gracilaria spp. (Friedlander et al. 1987, Lignell and Pedersen 1987, Jones et al. 1996). The demonstration that abalone performance when fed a mono- specific diet of Ulva lactuca can vary considerably, depending upon the nutrient status of the alga, explains conflicting conclu- sions reached in other reports considering the nutritional value of Ulva spp. Pickering (1990. reviewed in Stuart and Brown 1994) found differences in abalone growth when fed different ecotypes of Gracilaria sordida. However, other researchers have tended to consider only interspecific differences between macroalgae of- fered as diets for abalone (e.g.. Day and Fleming 1992, Shepherd and Steinberg 1992. Mercer et al. 1993. Marsden and Williams 1996). The documented relative dietary value of U. lactuca, as compared to other seaweeds, provides a useful indication of the importance of considering the alga's nutritional status: Ulva spp. has been considered a good (Uki 1989) and preferred food spcies for Haliotis discus hannai (Shepherd and Steinberg 1992); whereas, Mercer et al. ( 1993) noted H. discus hannai growth rates when fed their Ulva sp. (13% crude protein) to be significantly lower than for any other algae tested; results using H. tuberculata tend to be in closer agreement with the present study, the seaweed being considered a preferred diet (Shepherd and Steinberg 1992) of intermediate nutritional value (Koike et al. 1979, Mercer et al. 1993). In other abalone species. Tenore ( 1976) found good growth performance for juvenile H. discus and H. rufescens fed biofilter grown Ulva sp. (30% protein); whereas. Ulva sp. (13.2% protein) was the only diet tested by Stuart and Brown ( 1994) that produced no significant growth in juvenile H. iris. Implications for Commercial Abalone Production and Future Research In nature, macrophyte development tends to be N-limited (Lignell and Pedersen 1987. Duke et al. 1989a); therefore, pro- tein-N availability is suggested as being the main factor limiting field abalone growth (Fleming 1995b). The protein component of artificial diets represents the most costly ingredient of feeds (Mai et al. 1995) that are often prohibitively expensive (Fleming and Hone 1996). The culture of macrophytes in ammonia-enriched seawater. either by chemical supplement or by mariculture efflu- ents, seems to be a logical procedure for removing N-limitation from the food chain. Our results show that enrichment of N content in U. lactuca in this way significantly improves all indices of growth and feed use in juvenile and adult abalone. The present study suggests a considerable scope for modifica- tion of the nutritional value of the algal species commonly used as fresh diets for abalone and also a need for caution when consid- ering interspecific differences between seaweeds without examin- ing intraspecific variations in composition. Seasonal variability of wild seaweed populations with respect to protein content is pro- posed as critical for farmers harvesting natural stocks, in the se- lection of optimum sites for stock enhancement projects, and as an ecological tool to help elucidate the factors governing the food selection and population dynamics of abalone. There remains a need to investigate the long-term affects of algae cultured at a range of nutrient enrichment levels. It is also necessary to deter- mine the nutritional requirements of developing abalone, particu- larly at the critical stages of weaning, rapid juvenile growth, and sexual development in adults. ACKNOWLEDGMENTS The authors thank R. Friedman, A. Marshall, D. Ben Ezra, O. Dvir and E. Rotem for their advice and technical support. Thanks are also extended to Professor John J. Lee. This study was sup- ported by the Israli Ministries for Energy and Infrastructure and for Science, and by EC Grant No. 4564192 to M.S. and A.N. LITERATURE CITED Barkai, R. & C. L. Griffiths. 1987. Consumption, absorption efficiency, respiration, and excretion in the South African abalone Haliotis midae. S. Afr. J. Mar. Sci. 5:523-529. Bjomsater. B. R. & P. A. Wheeler. 1990. Effect of nitrogen and phospho- rus supply on growth and tissue composition of Ulva fenestrata and Enteromorpha intestinalis (Ulvales. Chlorophyta). J. Phycol. 26:603- 611. Bowen, S. H.. E. V. Lutz & M. O. Ahlgren. 1995. Dietary protein and energy as determinants of food quality: trophic strategies compared. Ecology 76:899-907. Ulva Lactuca Nutritional Value to Abalone 233 Britz. P. J. 1996. Effect of dietary protein level on growth performance of South African abalone, Haliotis midae, fed fishmeal-based semipuri- fied diets. Aquaculture 140:55-61. Cohen. I. & A. Neori. 1991. Ulva lactuca biofilters for marine fishpond effluents: I. ammonia uptake kinetics and nitrogen content. Botanica Manna 34:475-182. Colorni, A. 1989. Perforation disease affecting Ulva sp. cultured in Israel on the Red Sea. Dis. Aquat. Org. 71:71-73. Day. R. W. & A. E. Fleming. 1992. The determinants and measurement of abalone growth, pp. 141-168. In: S. A. Shepherd. M. J. Tegner. and S. A. Guzman del Proo (eds.). Abalone of the World: Biology. Fish- eries, and Culture. Fishing News Books. Oxford. DeBusk. T. A., M. Blakeslee & J. H. Ryther. 1986. Studies on the outdoor cultivation of Ulva lactuca L. Botanica Marina 29:381-386. Dubois. M, K. A. Gilles. J. K. Hamilton. P. A. Rebers & F. Smith. 1956. Colorimetric method for determination of sugars and related sub- stances. Anal. Chem. 28:350-356. Duke, C. S.. W. Litaker & Ramus. 1989a. Effects of temperature, nitrogen supply, and tissue nitrogen on ammonium uptake rates of the chloro- phyte seaweeds Ulva curvata and Codium decortication. J. PhycoL 25:113-120. Duke, C. S.. W. Litaker & J. Ramus. 1989b. Effect of temperature on nitrogen-limited growth rate and chemical composition of Ulva curvata (Ulvales: Chlorophyta). Mar. Biol. 100:143-150. Ennght. C. T.. G. F. Newkirk. J. S. Craigie & J. D. Castell. 1986. Growth of juvenile Ostrea edulis L. fed Chaetoceros gracilis Schtitt of varied chemical composition. / Exp. Mar. Biol. Ecol. 96:15-26. Fleming. A. E. 1995a. Growth, intake, feed conversion efficiency, and chemosensory preference of the Australian abalone, Haliotis rubra. Aquaculture 132:297-311. Fleming, A. E. 1995b. Digestive efficiency of the Australian abalone Hali- otis rubra in relation to growth and feed preference. Aquaculture 134: 279-293. Fleming. A. E. & P. W. Hone. 1996. Abalone aquaculture. Aquaculture 140:1-4. Folch, J.. M. Lees & G. H. Sloane. 1957. Simple method for isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:497- 507. Fujita, R. M.. P. A. Wheeler & R. Edwards. 1988. Metabolic regulation of ammonium uptake by Ulva rigida (chlorophyta): a compartmental analysis of the rate-limiting step for uptake. J. Phycol. 24:560-566. Fnedlander. M.. R. Shalev. T. Ganor. S. Strimling, A. Ben-Amotz. H. Klar & Y. Wax. 1987. Seasonal fluctuations of growth rate and chemical composition of Gracilaria cf. conferta in outdoor culture in Israel. Hydrobiologia 15 1/152:501-507. Herrero, C, A. Cid, J. Fabregas & J. Abalde. 1991. Yields in biomass and chemical constitutents of four commercially important marine microal- gae with different culture media. Aquacul. Eng. 10:99-1 10. Jones, A. B.. W. C. Dennison & G. R. Stewart. 1996. Macroalgal responses to nitrogen source and availability: amino acid metabolic profiling as a bioindicator using Gracilaria edulis (Rhodophyta). J. Phvcol. 32:757- 766. Koike, Y.. J. P. Flassch & J. Mazurier. 1979. Biological and ecological studies on the propagation of the Ormer, Haliotis tuberculata Linneaus. La Mer 17:43-52. Kudoh. S. 1987. The efficacy of dissolved inorganic nitrogen for quality in cultivated laver. Hydrobiologia 151/152:443—446. Lapointe, B. E. & J. H. Ryther. 1979. The effects of nitrogen and seawater flow rate on the growth and biochemical composiition of Gracilaria foliifera var. angustissima in mass outdoor cultures. Botanica Manna 22:529-537. Lignell. A.. Pedersen, M. 1987. Nitrogen metabolism in Gracilaria secun- data Harv. Hydrobiologia 151/152:431-141. Lundberg. P.. R. G. Weich. P. Jensen & H. J. Vogel. 1989. Phosphorus-3 1 and nitrogen- 14 NMR studies of the uptake of phosphorus and nitrogen compounds in the marine macroalgae Ulva lactuca. Plant Physiol. 89:1380-1387. Mai. K., J. P. Mercer & J. Donlon. 1995. Comparative studies on the nutrition of two species of abalone. Haliotis tuberculata L. and Haliotis discus hannui Ino. IV. optimum dietary protein level for growth. Aqua- culture 136:165-180. Marsden. I. D. & P. M. J. Williams. 1996. Factors affecting the grazing rate of the New Zealand abalone Haliotis iris Martyn. J. Shell. Res. 15: 401-406. Mattson W.J. Jr. 1980. Herbivory in relation to plant nitrogen content. Ann. Rev. Ecol. Syst. 11:119-161. Mercer, J. P., K. Mai & Donlon. 1993. Comparative studies on the nutrition of two species of abalone, Haliotis tuberculata L. and Haliotis discus hamuli Ino. I. effects of algal diets on growth and biochemical com- position. Invert. Repr. Dev. 23:75-88. Mgaya, Y. D. & J. P. Mercer. 1994. A review of the biology, ecology, fisheries, and mariculture of the European abalone Haliotis tuberculata Linnaeus 1758 (Gastropoda: Haliotidae). Proc. R. Irish Soc. 94B:285- 304. Miyashita. S. & T. Miyazaki. 1993. Effects of nutrient addition on phy- toplankton in Lake Nakanuma. Japan. Hydrobiologia 254:43-51. Neori, A., I. Cohen & H. Gordin. 1991. Ulva lactuca biofilters for marine fishpond effluents: II. growth rate, yield, and C:N ratio. Botanica Ma- rina 34:483^489. Neon. A., M. D. Krom. S. P. Ellner. C. E. Boyd, D. Popper. R. Rabino- vitch, P. J. Davison. O. Dvir. D. Zuber. M. Ucko. D. Angel & H. Gordin. 1996. Seaweed biofilters as regulators of water quality in in- tegrated fish-seaweed culture units. Aquaculture 141:183-199. Oakes, F. R. & R. D. Ponte. 1996. The abalone market: opportunities for cultured abalone. Aquaculture 140:187-195. Pedersen, M. F. 1994. Transient ammonium uptake in the maeroalga Ulva lactuca (Chlorophyta): nature, regulation, and the consequences for choice of measuring technique. J. Phycol. 30:980-986. Shepherd, S. A. & P. D. Steinberg. 1992. Food preferences of three Aus- tralian abalone species with a review of the algal food of abalone. pp. 169-181. In: S. A. Shepherd. M. J. Tegner. and S. A. Guzman del Proo (eds.). Abalone of the World: Biology, Fisheries, and Culture. Fishing News Books, Oxford. Shpigel. M. & A. Neori. 1996. The integrated culture of seaweed, ahalone. fish, and clams in modular intensive land-based systems: I. proportions of size and projected revenues. Aquacult. Eng. 15:313-326. Shpigel. M„ A. Marshall. I. Lupatsch, J. P. Mercer & A. Neori. 1996. Acclimation and propagation of the abalone Haliotis tuberculata in a land-based culture system in Israel. J. World Aquae. Soc. 27:435^442. Sokal, R. R. & F. J. Rohlf. 1995. Biometry: the principles and practice of statistics in biological research, 3rd ed. W. H. Freeman and Company. New York. Stuart, M. D. & M. T. Brown. 1994. Growth and diet of cultivated black- footed abalone. Haliotis iris (Martyn). Aquaculture 127:329-337. Tenore. K. 1976. Food chain dynamics of abalone in a polyculture system. Aquaculture 8:23-27. Uki, N. 1989. Abalone seedling production and its theory (2). Int. J. Aq. Fish. Technol. 1:125-132. Uki, N. & T. Watanabe. 1992. Review of the nutritional requirements of abalone {Haliotis spp.) and development of more efficient artificial diets, pp. 504-517. In: S. A. Shepherd, M. J. Tegner. and S. A. Guz- man del Proo (eds.). Abalone of the World: Biology, Fisheries, and Culture. Fishing News Books. Oxford. Vandermeulen. H. & H. Gordin. 1990. Ammonium uptake using Ulva (Chlorophyta) in intensive fishpond systems: mass culture and treat- ment of effluent. J. Appl. Phycol. 2:363-374. Wheeler. P. A. & B. R. Bjomsater. 1992. Seasonal fluctuations in tissue nitrogen, phosphorus, and N:P for five macroalgal species common the Pacific Northwest coast. J. Phycol. 28:1-6. White. T. C. R. 1978. The importance of a relative shortage of food in animal ecology. Oecologia 33:71-86. Journal of Shellfish Research. Vol. 18. No. 1. 235-241. 1999. THE UTILITY OF APPARENT DIGESTIBILITY COEFFICIENTS FOR PREDICTING COMPARATIVE DIET GROWTH PERFORMANCE IN JUVENILE GREENLIP ABALONE HALIOTIS LAEVIGATA MEEGAN E. VANDEPEER,1 PATRICK W. HONE,1 ROBERT J. VAN BARNEVELD,2 AND JON N. HAVENHAND1 1 South Australian Research and Development Institute Henley Beach 5022 South Australia, Australia 2Barneveld Nutrition Pty. Ltd. Lyndoch, 5351. South Australia, Australia ~ Flinders University- Adelaide, South Australia, Australia ABSTRACT The utility of using apparent amino acid and energy digestibility coefficients of individual ingredients to formulate diets for juvenile greenlip abalone. Haliotis laevigata, to specified digestible protein and energy levels, and rate of food consumption as a function of body size was studied. Five diets were formulated to have the same total digestible protein and energy contents, based on the apparent amino acid and energy digestibility coefficients of their ingredients. Despite similar nutrient profiles, the diets differed in their level of inclusion and combinations of four ingredients: semolina, soyflour, casein, and fishmeal. Diets were offered at two levels, subsatiation (0.01 g/d/abalone) and excess (0.1 g/d/abalone). Specific growth rates of shell length and wet weight of abalone were not significantly different among diets between feed levels. Thus, it seems that amino acid and energy digestibility coefficients of individual ingredients can be used effectively to formulate diets for abalone to desired levels of digestible protein and energy. The rate of food consumption of an abalone increased allometncally with its size dCu)ldt = aU)Wit)1', where C(t) and W{1) are, respectively, its total food consumption during a certain period of time up to time /, and its wet body weight at time t; a[t)s and b are its allometric parameters to be estimated. If a(t)s were assumed to vary with time. 1% increase in body weight required (0.5368 ± 0.1 183)% increase in food consumption. If a(t)s were assumed to be constant overtime. \°/c increase in body weight required (0.6861 ±0.1404)% increase in food consumption. KEY }\'ORDS: digestibility coefficients, protein, energy, growth, abalone INTRODUCTION Commercial nutritionists formulate animal diets to maximize production efficiency (Fleming et al. in press) by defining nutri- tional requirements, assessing the nutritive value of individual in- gredients to be used in the compound diets, and formulating least- cost diets. The current dataset for Australian feed manufacturers formulating diets for greenlip abalone, Haliotis laevigata consists of the following: (1) specifications for minerals and vitamins (Uki et al. 1985. Coote et al. 1996); (2) lipid inclusion level (van Barneveld et al. in press): (3) soft body tissue amino acid profile and requirement for the assumed limiting amino acid-lysine (Coote 1997); (4) apparent digestible protein and energy values for ingredi- ents commonly available in Australia (Coote 1997); and (5) optimal digestible protein level and optimal digestible pro- teimdigestible energy ratio (Coote 1997). To formulate the diets to match the available information on ingredient specifications with the nutritional requirements for ju- venile greenlip abalone on a least-cost basis, feed manufacturers must be confident that altering ingredient levels while maintaining nutritional specifications will not affect the animals' growth. Flem- ing et al. (in press) found for the greenlip abalone. no significant difference in the digestibility of diets that vary in fiber but contain equal amounts of digestible protein and energy level, based on the individual digestibility coefficients of the component ingredients. No data are available, however, on the additive effects of various ingredients on growth of the greenlip abalone. Growth experiments are essential for determining whether there are interactions among nutrients within a diet that can influence the proportion of nutrients retained by an animal and used for its growth. In evaluating diets, it is equally essential that both food con- sumption and growth rates of experimental abalone be measured to determine whether or not there is diet preference. This is because a higher rate of growth of abalone on one diet may be attributable to its higher rate of consumption but not to its better nutritional quality. The rate of food consumption is. however, a function of the size of an abalone. Unfortunately, the rate of food consumption of individual abalone as a function of their size is rarely calculated, because food consumption is usually observed for a group of aba- lone (mainly because of experimental constraints) rather than for each individual, and because of a lack of appropriate models for analyzing the resulting data. In this study, we examine the utility of using the apparent digestibility coefficients of ingredient nutri- ents in formulating diets for the greenlip abalone and propose a method for calculating the rate of food consumption of individual abalone as a function of their size. MATERIALS AND METHODS Abalone Hatchery-derived juvenile ( 12-month-old) greenlip abalone, Haliotis laevigata, were used in this experiment. When the experi- ment started, the mean size and weight of abalone were 20.43 ± 0.22 mm and 1.17 ± 0.04 g. respectively. All abalone were anes- thetised with a 4% stock solution of benzocaine (Ace Chemicals. Camden Park, South Australia) (Hahn 1989). weighed to the ac- 235 236 Vandepeer et al. curacy of 0.001 g in wet weight, and measured to the accuracy of 0.1 mm in total shell length. Before being assigned to tanks, each animal was tagged following Coote et al. (1996). Diet Formulation Five isonitrogenous and isoenergetic diets of different combi- nations of semolina. Peruvian fishmeal. casein, and defatted soy- flour (Baker's Nutrisoy; Ace Chemicals) were formulated, based on previously determined apparent digestible amino acid and en- ergy coefficients of the ingredients for H. laevigata (Coote 1997) (Table 1 ). An "ideal" protein ratio was used in these formulations based on the amino acid profile of the soft body tissues of the abalone (Fleming et al. 1996. Coote 1987). This required inclusion of the free amino acids L-arginine monohydrochloride, L-threonine, L-lysine monohydrochloride. and DL-methionine (Ace Chemicals). Identical amounts of vitamins E (DL-alpha tocopheryl acetate) and C (ascorbic acid), sodium alginate (binder), abalone vitamin and mineral premixes, and phosphate were added to each diet. Although it was not possible to have identical levels of supplemented oil (Jack Mackerel Oil. Triabunna Fish Oils, Tri- abunna. Tasmania) within all diets (to make the diets isoenergetic). its levels of inclusion varied by s 10 g between any two diets. Diet TABLE 1. Ingredient composition (% inclusion on an air dry weight basis) and proximate components of diets. Diets Ingredients 1 5 Semolina 69.5 75.4 6X.6 74.9 71.5 Fishmeal 0 0 10 5 10 Casein 0 10 0 10 5 Soyflour 23.6 7.6 15.6 3.6 7.6 Na alginate 0.5 0.5 0.5 0.5 0.5 Mineral mix" 0.3 0.3 0.3 0.3 0.3 Vitamin mixa 0.2 0.2 0.2 0.2 0.2 Vitamin E 0.01 0.01 0.01 0.01 0.01 Vitamin C 0.05 0.05 0.05 0.05 0.05 Na phosphate 0.5 0.5 0.5 0.5 0.5 Jack Mackerel oil 4.00 3.85 3.00 3.34 2.y2 L-arginine HCL 0.52 0.74 0.55 0.76 0.664 L-threonine 0.18 0.16 0.16 0.15 0.15 L-lysine HCL 0.09 0.09 0.02 0.06 0.025 DL-methionine 0.11 0.09 0.06 0.06 0.05 Proximate Analysis CP% GEMJ kg"1 DP% DE MJ kg"1 NDF% ADF% CF% ASH% DM% Crude Fat% 20.99 17.8 17.15 10.75 8.88 3.13 1.91 3.04 93.03 4.74 22.91 18.0 17.15 10.75 6.39 2.20 1.79 2 22 92.65 4.44 23.02 17.7 17.15 10.75 11.07 3.32 1 .53 4.04 93.03 4.95 23.84 18.1 17.15 10.75 5.91 2.99 1.00 3.13 93.34 4.54 23.66 18.0 17.15 10.75 10.86 2.71 0.96 3.56 93.81 4.91 CP, crude protein; GE. gross energy; DP. apparent digestible protein; DE. apparent digestible energy; NDF. neutral detergent fiber; ADF. acid deter- gent fibre; CF. crude fiber; DM. dry matter. J Vitamin and mineral mix (Uki et al. 1985). 1 served as a control and contained only semolina and soyflour. All diets were fed to subsatiation or to excess, giving a total of 10 treatments (5 diets x 2 feeding levels). Each treatment had four replicates (40 tanks in total). All diets were stored at -30°C before feeding. Experimental Conditions Four replicates of each diet to be tested were randomly as- signed to an 8-L plastic tank of the dimensions 18x21 x 31 cm. Each tank contained 10 abalone randomly chosen from a group of 480 and had a piece of PVC stormwater pipe as a hide. Tanks were on a flow through water system. The seawater was filtered to 30 p.m by primary sand filters and then to 10 u.m by secondary composite sand filters. Temperature was maintained at I8.0°C. measured on day 29 of the 45-day experimental period, and ranged from 17.5 to 20.0°C. except for 2 days when it dropped to 15.0°C because of a system failure. Salinity was 36.5«>^ - 1 ] where T(t,, t,+1) is the total food consumption of all n abalone in a tank during the ith observational period [t,. t,+ 1], and we have assumed that a(t)s remain constant during that observational pe- riod. Six measurements of T(t,, t,+1) were made in the course of our experiment. The size of all test abalone was, however, mea- sured only at the start and end of the experiment. Fortunately, the values of LU) can be calculated (see above) as L,(t) = L,U0)e GAt-n) = L,U0) L,(t,) L,.(f0) where L,U0) is the size of individual i at the start of the experiment at time f0 = 0 d; L,itt) that at the end of the experiment at time tf = 45 d. Both L,(tu) and Lf$ were calculated very precisely, and their errors are negligible. Now, parameters a(t)s and b can be 238 Vandepeer et al. estimated from Tit,, t,+ 1) and Z.,(f) tor abalone from different tanks, by making appropriate assumptions about the distributions of the errors in F(t,, ti+1). We assumed that the errors in F(t,, t,+ 1) follow independent normal distributions, with a mean of T(t,, t,+ 1) and a constant variance of a2, and estimated all parameters by the standard nonlinear least squares method (SAS. Inc. 1988). The parameters thus estimated are those for shell length. A simple replacement of shell length of an abalone by its body mass yielded estimates of a(t)s and b for body mass. Measurements Both shell length (SL) and wet body weight sured at the start and end of the experiment specific growth rates. Shell length was measured mm using an electronic digital caliper, and wet measured to the nearest 0.001 g. Excess water abalone using a paper towel before weighing lasted 45 days. The specific growth rate (SGR) (BW) were mea- to estimate their to the nearest 0. 1 body weight was was dried off the The experiment was calculated as SGR. 100% = In C1./1- In G(i') It where G{i) is the SL (mm) or BW (g) at the start of the experiment. Gtf) is the SL (mm) or BW (g) at its end, and A? is the experi- mental duration (d) (see above). Chemical Analyses Proximate analyses of the diets (i.e.. dry matter, crude protein, crude, neutral detergent and acid detergent fiber, ash and crude fat) were undertaken using the methods of the Association of Official Analytical Chemists ( 1984). Gross energy was determined using a Parr 1.281 bomb calorimeter. Statistical Analyses Analysis of variance (ANOVA) (SAS Institute 1988) was used to determine whether there are differences in mean size of abalone among tanks at the start of the experiment, in mean SGRs among diets for both BW and SL and in mean feed consumption. RESULTS No significant difference was found in mean length of abalone among tanks at the start of the experiment (F3936] = 1.16, p = .2436, n = 400). Because of a failure of the water circulating system, one tank of abalone fed diet 1 to excess died. Also dead were one abalone fed diet 1 to subsatiation, one fed diet 3 to subsatiation, one fed diet 5 to excess, two fed diet diet 4 to sub- satiation, and three fed diet 2 to excess. Mean and standard errors of SGR for both SL and BW are given in Table 3 and Figure 1. No significant differences occurred among the five diets for either SL (F4364 = 0.78. p = .5371) or BW (F4; 0.42 .7934). In contrast, significant differ- ences existed between feeding levels for both SL (F, ,64 = 1 12.02. p = .0001) and BW (F, 364 = 144.14. p = .0001). The abalone fed to excess had achieved much greater SGRs of both SL and BW than those fed to subsatiation for the same diets (Fig. 1 ). Signifi- cant diet by feeding level interactions were found for BW (F4364 = 3.11. p = .0156) but not for SL (F4 ,64 = 1.85, p = .1184). No significant differences were detected in feed consumption among the five diets for any of the 6 days of measurement (day 1, F4 14 = 1.37. p = .2929; day 2, F4 14 = 2.76. p = .0699; day 3. F4 ,4 = 0.84. p = .5197; day 4. F4 14 = 1.60. p = .2289; day 5. F4 l4 = 1.13. p = 1.13, p = -.3822; day 6, F4 14 = 1.89, p = TABLE 3. Mean and standard error (in parentheses) of the specific growth rates (SGR) of shell length and body weight of abalone for both feeding levels for each diet (A, fed to subsatiation: B, fed to excess). SGR of shell SGR of Body Diet Feeding Level N Length (d1) Weight (d_l) 1 A 4 0.40 (0.04) 0.89(0.10) B 3 0.61 (0.03) 1.54(0.07) 2 A 4 0.43 (0.04) 1.02(0.12) B 4 0.60 (0.04) 1.52(0.12) 3 A 4 0.42 (0.04) 0.98 (0.06) B 4 0.56(0.02) 1.40(0.05) 4 A 4 0.42(0.05) 1.03(0.13) B 4 0.57(0.01) 1.40(0.07) 5 A 4 0.44 (0.07) 1.10(0.16) B 4 0.53 (0.02) 1.34(0.04) .1682). The mean consumption rates varied from 0.024 g/d/ abalone for diets 1 and 3 to 0.027 g/d/abalone for diet 5 (Table 4). In subsatiation treatments, abalone ate all the offered feed, ap- proximately 0.012 g/d/abalone. In comparison, those fed to excess ate approximately 0.025 g/d/abalone. The feed conversion ratio (FCR) ranged from 0.97 for diet 4 to 1.17 for diet 1 in the subsa- tiation treatments, and 1.04 for diet 4 to 1.28 for diet 2 in the excess treatments (Table 4). For diets 2. 4. and 5, the FCRs for the subsatiation treatments were all lower than those for the excess treatments for the same diet. Fitting of the above model to the abalone data by the standard nonlinear least-squares method, assuming that the errors in T (t,. t,+ |) follow independent normal distributions, with a mean of r 0.3 "S g 0.2 0.1 0 i I i I 1.8 1.6 1.4 \ 1.2 1 1 2 0.8 JS 0.6 0.4 0.2 0 I i i ♦ LevelA ■ Level B i Figure 1. n = 4. Diet Specific growth rates of shell length and body weight of experimental abalone. Level A, subsatiation (0.1 g/d); level B (1.0 g/day); difficult, because of a lack of a biologically meaningful interpre- tation of time-varying a(1)s. DISCUSSION Diets with different ingredients or ingredient combinations, but formulated to be both isonitrogenous and isoenergetic. based on the apparent digestibility coefficients of nutrients in the ingredi- ents, should yield equal growth rates. In this instance, differences in growth rate of test animals would suggest that ingredient com- ponents interact to influence their utilization and that the apparent digestibility coefficients are not additive and. hence, are unsuitable for use in diet formulation. The digestibility coefficients of indi- vidual ingredients seem to be additive, and the nutrients from each diet are equally utilized by abalone. This is because similar SGRs were found in abalone fed diets consisting of different ingredients but formulated to have the same levels of digestible protein and energy. Thus, ingredient apparent digestibility coefficients can be used effectively for formulating diets to specific digestible energy and protein levels. Preference of abalone for the diets containing fishmeal was not observed in this study. This is not surprising, considering that abalone are herbivorous and the levels of fishmeal in the three diets were probably too low (a maximum of 10%) for abalone to show significant preferential selection. By comparison, rainbow trout, Oncorhynchus mykiss, fed diets containing 100% replace- ment of fishmeal by plant protein, had a significantly reduced weight gain and specific growth rate, although the diets had the same amount of digestible protein and energy (Gomes et al. 1995). This was attributable to a significant reduction in intake of the 100% plant protein diet, as compared with fishmeal-based diets. Such a reduction of voluntary feed intake in plant meal-based diets by rainbow trout was also observed by Gomes and Kaushik ( 1992), who suggested that this was probably because of their carnivorous nature and hence not being adapted for utilizing plant ingredients in diets. The lower FCR of diets 2, 4, and 5 for abalone fed to subsa- tiation than those fed to excess is readily explained. The growth- ration relationship is characterized by an initial increase in growth rate with ration, followed by a leveling off. In general, the optimal ration is attained when fish are fed at restricted ration levels (Tal- bot 1994). When feed intakes are reduced below the maintenance level, animals, such as limpets (Branch 1992). tend to become more efficient in digesting feed and utilizing its nutrients (May- nard et al. 1969). Similarly, species with an abundant food supply generally have a lower absorption efficiency than those with a TABLE 4. Mean and standard error (in parentheses) of wet body weights at the start (initial) and end (final) of the experiment, total weight gain, feed consumption, and feed conversion ratios (FCR) for both feeding levels of each diet. (A, subsatiation; B, excess). Mean Total Mean Feed Feeding Mean Initial Mean Final Weight (Iain Consumption Diet level Weight (g) Weight (g) (g) (g/d'/abalone-1) FCR 1 A 1.10(0.13) 1.56(0.10) 0.46(0.03) 0.012 1.17 B 1.16(0.07) 2.14(0.15) 0.98 (0.09) 0.024 1.08 2 A 1.00(0.10) 1.53(0.09) 0.53 (0.06) 0.012 1.01 B 1.05(0.15) 2.00(0.14) 0.94 (0.04) 0.027 1.28 3 A 0.94 (0.09) 1.43(0.99) 0.48(0.01) 0.012 1.10 B 1.19(0.12) 2.15(0.21) 0.97 (0.09) 0.024 1.10 4 A 1.09(0.13) 1.66(0.15) 0.57 (0.04) 0.012 0.97 B 1.38(0.14) 2.46(0.14) 1.08(0.04) 0.025 1.04 5 A 0.93(0.13) 1.46(0.15) 0.53 (0.06) 0.012 1.02 B 1.50(0.16) 2.61 (0.25) 1.11 (0.09) 0.027 1.08 240 Vandepeer et al. TABLE 5. Estimates of the mean and standard errors (in parentheses) of parameters in models for food consumption as a function of an individual's size by the nonlinear least squares method. i Independent Variable t, (d) a(t) x 10"3 * F R2 Length (mm) Constant ait) 1 .0450 ( 1 .3600) 2.0103(0.4041) F,,,2 = 518.7876 0.9026 1 Variable a(t) 7 3.0866(3.4626) 1.5148(0.3540) F7I07 = 266.7813 0.9458 2 9 3.9809 (4.4720) 3 11 5.9518(6.7825) 4 21 6.0085 (6.8684) 5 23 5.0358 (5.7764) 6 Weight (mg) 25 1.5148(0.3540) Constant ait) 455.5007 (39.0080) 0.6861 (0.1404) F211, = 514.2863 0.9018 1 Variable ait) 7 606.1055(41.0125) 0.5368(0.1183) F7J07 = 270.7141 0.9466 2 9 295.5456 (33.2396) 3 11 381.7449(35.4795) 4 21 576.1663(47.1054) 5 23 582.8623 (48.6964) 6 25 489.0853(44.3772) p< .0001. n = 114. The unit of constant or variable ait) is g/mm'/d for length-dependent food consumption, and g/g'Vd for weight-dependent food consumption. more limited food supply (Branch 1982). These observations may reflect more metabolism of animals than digestibility alone (May- nard et al. 1969). For species with an abundant supply of food, energy conservation by reducing metabolic losses may not be nec- essary or even desirable: whereas, for species suffering from a shortage of food, metabolic adjustments that reduce energy losses may be critical (Newell and Branch 1980). In this work, we have proposed a novel method for calculating the rate of food consumption of an individual abalone as a function of its size and have demonstrated its utility in data analysis. The relationship thus established between food consumption and aba- lone size allows abalone farmers to adjust feeding levels accord- ingly. Indeed, feeding the "correct" amount of food on a commer- cial farm is economically important. This is because overfeeding of abalone results in food wastage and increases cost; whereas, underfeeding leads to their slower growth and increases time to reach the market size, which in turn increases production cost. In summary, the growth of abalone growth is unaffected by varying ingredient combinations of their diet, provided that the diets have the same DP:DE ratio. Thus, the apparent digestible protein and energy coefficients are sufficient to define the nutri- tional quality of their ingredients and can be used effectively in formulating diets. Future experiments should, therefore, determine the amino acid and energy digestibility of alternate, cheaper in- gredients. These could then be substituted into existing diet for- mulations, provided that the DP:DE ratio is kept constant, equal growth rates are achieved, and the cost of the diet is reduced. These ingredients would, however, have to be assessed for their maximum inclusion levels, because they may contain antinutri- tional factors, which, at above certain levels within a diet, are detrimental to the growth of the abalone. In addition, future ex- periments for greenlip abalone should examine how food con- sumption changes with temperatures that are experienced on com- mercial farms and construct a temperature- and size-dependent feeding table. This would assist abalone farmers in seasonally adjusting their feeding protocols. ACKNOWLEDGMENTS Thanks go to staff at the nutrition laboratory of the Pig and Poultry Production Institute (SARDI). including Jurek Kruk, Janet Hattam, Bronni Davis. Chris Adley, and Steve Szarvas for their advice and assistance with proximate analyses. We also thank Dr. Yongshun Xiao for his statistical advice and comments on the manuscript. This research was funded by the Fisheries Research and Development Corporation. LITERATURE CITED Association of Official Analytical Chemists. 1984. Official methods of analysis of the association of Official Analytical Chemists, 14th ed. Association of Official Analytical Chemists, Washington DC. Branch. G. M. 1982. The biology of limpets: physical factors, energy flow, and ecological interactions. Oceanogr. Mar. Biol. Ann. Rev. 19:235- 380. Coote, T. A., P. W. Hone, R. Kenyon & G. B. Maguire. 1996. The effect of different combinations of dietary calcium and phosphorous on the growth of juvenile Haliotis laevigata. Aquaculture 145:267-279. Coote, T. A. 1997. The protein, energy, and lysine requirements of greenlip abalone iHaliotis laevigata). Ph.D. dissertation. University of Tasma- nia, Australia. Fleming. A. E.. R. J. van Barneveld & P. W. Hone. 1996. The development of artificial diets for abalone: a review and future directions. Aquacul- ture 140:5-53. Fleming. A. E., R. J. van Barneveld, P. W. Hone, M. E. Vandeeper & J. A. Kruk. in press. Complementary additivity of feed ingredients fed to juvenile greenlip abalone iHaliotis laevigata). Gomes, E. & S. Kaushik. 1992. Effect of the replacement dietary inorganic- zinc by zinc/methionine on vegetable and animal protein utilization by Utility of Digestibility Coefficients for Abalone 241 rainbow trout, pp. 897-902. In: S. J. Kaushik and P. Luquet teds.). Fish Nutrition in Practice. INRA Editions, Biarritz, France. Gomes, E., P. Remo & S. Kaushik. 1995. Replacement of fish meal by plant proteins in the diet of rainbow trout {Oncorhynchus mykiss): digestibility and growth performance. Aquaculture 130:177-186. Hahn. K. O. 1989. Handbook of culture of abalone and other marine gas- tropods. CRC Press, Boca Raton, Florida. 348 pp. Maynard, L. A., J. K. Loosli, H. F. Hintz & R. G. Warner. 1969. Animal Nutrition. McGraw-Hill, Inc. Newell. R. C. & G. M. Branch. 1980. The influence of temperature on the maintenance of metabolic energy balance in marine invertebrates. Mar. Biol. 17:329-396. Peterson C. H. & P. E. Renaud. 1989. Analysis of feeding preference ex- periments. Oecologia 80:82-86. Peters, R. H. 1983. The ecological implications of body size. Cambridge University Press, New York. 329 pp. SAS Institute, Inc. 1988. SAS/STAT® Users guide, release 6.03 ed SAS Institute, Inc., Cary. North Carolina. 1028 pp. Talbot. C. 1994. How growth relates to ration size. Fish Farmer January/ February: 45^46. Uki, N.. A. Kemuyama & T. Watanabe. 1985. Development of semi- purified test diets for abalone. Bull. Japan. Soc. Sci. Fish. 51:1825- 1833. van Bameveld, R. J., A. E. Fleming, M. E. Vandepeer. J. A. Kruk & P. W. Hone, in press. Influence of dietary oil type and oil inclusion level in manufactured feeds on the digestibility of nutrients by juvenile greenlip abalone (Haliotis laevigata). Journal of Shellfish Research. Vol. 18, No. 1. 243-250. 1999. GROWTH AND SURVIVAL OF POSTLARVAL ABALONE (HALIOTIS IRIS) IN RELATION TO DEVELOPMENT AND DIATOM DIET RODNEY D. ROBERTS,1 2 TOMOHIKO KAWAMURA,3 AND CHRISTINE M. NICHOLSON1 Cawthron Institute Nelson, New Zealand 'University of Otago Dunedin, New Zealand Tohoku National Fisheries Research Institute Shinhama-cho, 3-27-5 Shiogama, Miyagi 985, Japan ABSTRACT Postlarval abalone (Haliolis iris) were reared on five unialgal diatom diets from 3 to 68 days postsettlement. Diatom strain affected both survival and growth, which were positively correlated (r = 0.88. p = 0.05. n = 5). The digestibility and ingestibility of diatoms were both important. Survival ranged from low (<10% on day 37 postsettlement) on Pleurosigma sp. diet, to high 070% on day 68) on Cocconeis scutellum, Cylindrotheca closterium, and Navicula ramosissima diets. Diet had little effect on growth and survival in the first 16 days after settlement, provided postlarvae were ingesting adequate food. Growth rates during the week to day 10 were relatively high (means of 20-29 |xm shell length per day). Growth rates from days 10 to 16 were lower than in the first week (t = 7.33. /» < 0.001) and again similar among all diets (means 15-20 p.m/day) except Pleurosigma sp. (2 p.m/day). which was not ingested by larvae <1 mm shell length. After day 17, postlarvae grew fastest on the strains that were most efficiently digested (C. scutellum and C. closterium). The digestion efficiency of two diatom strains increased markedly during the experiment, because of changes in diatom condition. Postlarvae were feeding on small diatoms (12x4x3 p.m) by day 2 postsettlement. and digestive gland development became visible on day 3. Fecal volume increased approximately cubically in relation to shell length, indicating rapidly increasing food consumption. KEY WORDS: benthic diatom, abalone development, growth, survival, postlarval abalone, digestion efficiency INTRODUCTION Even after decades of abalone farming, consistent and efficient postlarval culture remains elusive (Leighton 1989, Roberts et al. 1998). Survival rates through the few months postsettlement are generally low and variable (Searcy-Bernal et al. 19921. Many hatchery operators report mortality peaks 3 to 8 weeks after settle- ment (Leighton 1989. Roberts et al. 1998). These events are often thought to be food related (Roberts et al. 1998), but the precise mechanisms involved are unknown. The timing of mortality events suggests that the cause may lie in postlarval developments or changes in the bioftlm on which postlarvae feed. Postlarvae in hatcheries feed on a biofilm that is usually domi- nated by diatoms and their extracellular products. The composition of the biofilm can affect growth and survival of postlarvae. One critical factor is the efficiency with which diatoms are digested. Many diatom strains pass through the abalone gut alive and intact, but efficiently digested strains can produce better growth and sur- vival of postlarvae (Kawamura et al. 1998b). The "digestion effi- ciency" of a diatom strain has been defined as the proportion of live diatoms that are ruptured during ingestion/digestion (Kawa- mura et al. 1995). Digestion efficiency has previously been as- sessed at only one time in an experiment (Kawamura and Takami 1995, Kawamura et al. 1995. Kawamura et al. 1998a). However, digestion efficiency may vary with the condition of the algal cul- ture, or as postlarvae develop. Another important feature controlling the food value of diatoms is the ease with which they are removed from the substrate and swallowed (their "ingestibility"). Postlarval abalone grazing favors certain diatom species (Norman-Boudreau et al. 1986, Matthews and Cook 1995). Passive selection can arise through such diatom characteristics as cell size, stalk length, or attachment strength (Matthews and Cook 1995, Kawamura et al. 1995, Kawamura et al. 1998a). Few such examples have been quantified or related to postlarval abalone development. Most previous studies on the growth of postlarval abalone on specific diatom diets have covered only a few weeks of the post- larval period (Ohgai et al. 1991. Ishida et al. 1995, Kawamura and Takami 1995. Kawamura et al. 1995. Kawamura et al. 1998a). This has limited the ability of those studies to demonstrate changes in growth rates related to postlarval development. For example, previous data suggest that a diatom diet made little difference to the growth rate of small abalone postlarvae (Kawamura and Takami 1995); whereas, larger postlarvae [>800 u.m shell length (SL)] grew more quickly on diatoms with high digestion efficiency (Kawamura et al. 1995, Kawamura et al. 1998a). This suggestion was based on separate short-term experiments with different ages of postlarvae. Studies covering the whole postlarval period would be more conclusive and valuable. Postlarval developments that affect growth and survival are likely to involve the feeding or digestive apparatus. Information about the onset of diatom feeding in postlarval abalone is variable. Seki and Kan-no (1981) reported that postlarval Haliotis discus hamuli Ino 1953 in hatcheries consume diatoms within 2 days of settlement. In similar conditions. Norman-Boudreau et al. (1986) found that feeding on diatoms was initiated 2 to 6 days after settlement. In contrast. Kitting and Morse (1997) reported that postlarval H. rufescens Swainson 1822 on coralline algae did not ingest cellular solids in the first 10 days after settlement, although postlarvae did show feeding movements. This issue is important for hatchery management, because it will affect decisions about the type of food suitable for very young postlarvae and the time at which it should be added. 243 244 Roberts et al. A related issue is the need to balance food supply with the increasing food consumption of growing postlarvae. Abalone farmers report difficulty in re-establishing suitable diatom films in areas cleared by postlarval grazing (Roberts et al. 1998). A better knowledge of food consumption rates would help in balancing food supply to avoid overgrazing. Despite widespread recognition of this difficulty, there are few quantitative data on abalone grazing rates. Martinez-Ponce and Searcy-Bernal (1998) reported that grazing (measured by diatom clearance) was minimal during the first week but increased abruptly at -470 u,m SL (2-3 weeks postsettlement). In this paper, we follow the growth and survival of a single batch of abalone throughout the postlarval period on five diatom diets. This lets us examine ways in which postlarval development and diet interact to influence postlarval growth and survival. Dia- tom diet emerges as a critical factor, although not for very young postlarvae. We show that the food value of some diatoms changes as postlarvae grow or diatom characteristics change. MATERIALS AND METHODS Abalone Rearing Adult abalone (Haliotis iris Gmelin 1791) were induced to spawn by adding hydrogen peroxide (5 mM final concentration) to alkaline (pH 9.1) seawater (Morse et al. 1977). Larvae were hatched and reared in l-|xm filtered. UV treated, flowing, natural seawater (15 ± 1 °C ) at Island Hatcheries Ltd. on Stewart Island, or at the Glenhaven Aquaculture Centre Ltd. in Nelson. Operculate larvae from Stewart Island were transported to Nelson in wet mesh (Tong and Moss 1992) then reared as above. Competent larvae were transferred to tissue culture dishes (Falcon 3046) with 10 mL of 0.2 p.m filtered natural seawater (FSW. 17.5 ± 1°C, dark) con- taining 150 p.g/mL each of Penicillin G sodium (Biochemie) and Streptomycin sulphate BP (Sigma). These larvae were induced to attach and metamorphose by the addition of 2 u.M GAB A (Roberts and Nicholson 1997). Two days after settlement induction, a dia- tom culture [Cylindrotheca closterium (Ehrenberg) Reimann and Lewin 1964] was added as a food supply, and the trays were thereafter incubated in the light (35-70 u.E/m""7s"1 on a 12:12 lightVdark cycle) at 17.5 ± 1°C. These postlarvae were maintained as a source of experimental animals, by adding supplementary diatom (C closterium) as necessary and replacing the water each 3 to 4 days with new FSW without antibiotics. Diatom Cultures Five strains of benthic diatom were used as diets (Table 1 ). Four were isolated from Nelson coastal waters, and the fifth (Pleu- rosigma sp.) was obtained from the Tohoku National Fisheries Research Institute in Japan. Cultures were uni-algal but not axenic. Diatoms were grown in tissue culture dishes (Falcon) in the light (35-70 u.E/m-2/s-' on a 12:12 light/dark) at 17.5 ± 1°C. Growth medium was Jorgensen's (1962) recipe, supplemented with 0.05 p.g/L vitamin Bu. For experiments, growth medium was replaced with FSW. Diatom subcultures were established regularly so that postlarvae could be transferred to a new, healthy culture at weekly intervals. Postlarvae were manipulated onto their foot after all transfers. Diatom cell density was quantified at the beginning and end of each week. An excess of food was maintained at all times, except as discussed in Results. Growth and Survival of Postlarvae Three days after settlement induction, postlarvae were dis- lodged with fine needles and individually pipetted through six FSW washes to remove external diatoms, then placed into diet treatments (three larvae in each of six replicate wells). The C. closterium fed to "spare" postlarvae had consistently high diges- tion efficiency (Fig. 1A). so there was negligible contamination of diets by live diatoms in feces. We measured SL of individual postlarvae at approximately weekly intervals for 9 weeks. Postlarvae less than -2 mm SL were measured on an inverted compound microscope video linked to a high-resolution monitor (accuracy ± 10 u.m). Larger postlarvae were measured with an eyepiece graticule on a dissecting micro- scope (accuracy ± 50 p.m). Survival was measured as the percent- age of postlarvae surviving each week. Postlarvae that died as a result of handling damage, loss, or emersion were excluded from survival calculations. "Specific daily growth rate" was calculated TABLE 1. Details of the diatom cultures used in feeding experiments. Cell Dimensions: Length Median Density Diatom Species x Width x Depth (pm) in Food Cultures Adhesive Growth Collection and Authority and [Volume" (urn')! ( Cells/cm- ) Strengthb Formb Reference Coeconeis scutellum Ehrenberg 1838 Cylindrotheca closterium (Ehrenberg) Reimann and Lewin 1964 Navicula ramosissima (Agardh) Cleve 1895 Navicula britannica Hustedt et Aleem 1951 Pleurosigma sp. 29 x 17x2.8 [1.105] 22 x 3.5 x 3.5 [228a] 12x4.1 x 3.0 [lis] 30x8.0x7.6(1.452] 115 x 16x8.5 [12.193] 1.6 x 105 II x 10s 2.2 x 105 1.0 x 105 2.0 x 104 B CAWZ F A CAWZ 1 A CAWZC A CAWZM A _ J Volumes calculated assuming that all species were ellipsoidal prisms. For C. closterium, the thin distal extensions were ignored, and only the swollen section of the cell was used for volume calculations. b According to Kawamura and Hirano (1992). Growth form A are solitary, mobile diatoms of low-to-moderate adhesive strength. Growth form B are highly adhesive, solitary diatoms. Growth and Survival of Postlarval Abalone 245 A) Digestion efficiency D) Shell length 10 20 30 40 50 60 70 10 20 30 40 50 60 70 0000 -I 3) Faecal volume /*- r~^T7 1000 - r -wT fioo - ( ■ ^r >L Faecal vo. — i — H fe 1 \ 1 1 80 T I 60 £ 40 -- I 20- O E) Growth rate »-f»-"t 10 20 30 40 50 60 70 10 20 30 40 50 60 70 100 T to > £ 03 > 80 -- 60 -- 3 40 -- 20 -- C) Survival t> + -pe — h 10 20 30 40 50 60 70 Days since settlement 6 -r F) Specific growth rate (SGR) 10 20 30 40 50 60 Days since settlement -A — Cocconeis scutellum Navicula britannica - - Cylindrotheca closterium o ■ ■ Pleurosigma sp. •Navicula ramosissima Figure 1. Comparison of diatom digestion efficiency (A) and abalone fecal volumes (B), survival (C), and growth (D-F) for postlarvae fed five diatom diets. Growth rate data (E, F) applv to the period since the last measurement. Other data are instantaneous. Data are mean ± SE; n = 5 for graphs A-B, n = 6 for D-F. Graph C represents survival across all replicates. In graph A, there are no data for Pleurosigma, because postlarvae did not ingest cells. In Graph B. the y-axis is logarithmic. 246 Roberts et al. as the percentage of SL grown per day (using final SL for each week). Digestion Efficiency, Fecal Production, and Feeding Observations In this paper, we use the term "digestion efficiency" to describe the ability of postlarvae to remove the cell contents of diatoms passed through the gut (Kawamura et al. 1995). The term "ingest- ibility" refers to the ease with which a food item is detached and swallowed by a postlarva. To measure digestion efficiency, indi- vidual postlarvae (1 per well, n = 5) were removed from the growth experiment, and pipetted through six wells of FSW (to remove external diatoms) before being placed in a clean tissue culture dish with FSW. Recently released fecal material was ob- served with an inverted microscope, and the proportion of intact versus ruptured diatom cells in the fecal pellet was counted. A parallel count was made of live versus dead cells in the diatom culture from which the fecal material was grazed. Digestion effi- ciency (%) was calculated as: (1 - L/Lu) * 100, where L0 is the proportion of live cells in the source culture, and L is the propor- tion of live cells in the fecal pellet. In most instances, feces were sufficiently translucent that ac- curate counts of live/ruptured cells could be made, either on intact fecal pellets or after live cells had migrated out. However, when Navicula ramosissima developed high digestion efficiency this method was unsatisfactory, because the core of the fecal pellet contained many damaged cells. Such feces were transferred to glass slides by capillary-pipette in a drop of FSW. A cover slip was applied, then gently moved from side to side, spreading the dia- toms in the pellet. Checks confirmed that this technique did not damage intact cells. Fecal volume was quantified by measuring the diameter and length of all fecal pellets released within an hour of the postlarva being removed from its diatom diet. We assumed that feces were cylindrical and that fecal production provided an indication of ingestion rate. Feeding behavior and gut developments were observed directly with an inverted compound microscope. Observations of feeding in the first few days after settlement used a diet of Navicula ra- mosissima (cells 12-(xm long and 5-p.m wide across the valve) added 20 hours postsettlement. Data Analysis Differences between growth rates were tested using analysis of variance (ANOVA) with Tukey's HSD. Homoscedasticity was assessed using Levene"s test, and normality by probability plots of residuals (Kirby 1993). In testing the correlation between survival and growth rate, the growth curve for the Pleurosigma sp. diet was extrapolated assuming that postlarval growth in days 30 to 68 continued at the rate observed from days 10 to 30. In fitting the regression curve to fecal volume data, a power relationship was specified. RESULTS Early Postlarval Development and Feeding By 24 hours after the addition of GABA. many larvae had commenced metamorphosis (Table 2). The velum had been shed, and the heart had begun beating. The mouth had opened and the initial processes had fused anteriorly to form the snout. Peristomal shell growth had commenced in some postlarvae. and feeding TABLE 2. Timing of the physical changes associated with metamorphosis of H. iris larvae. Developmental Age (Days Since Settlement Induction) Milestone 0.25 1 2 3 Heart beating? No Yes Yes Yes Velum shed? No Yes Yes Yes Peristomal shell growth'1 No Yes Yes Yes Mouth formed? No Yes Yes Yes Feeding movements? No Yesa Yesa Yes Ingestion visible? No No Yes Yes Feces visible? No No Yes" Yes Rotation in gut? No Yes Yes Yes Ginger color of digestive gland? No No No Yesc Larvae were -160°C days old when induced to settle with 2 u.M GABA. Data relate to the postlarvae that responded rapidly to the cue. Some larvae showed slower metamorphosis, or failed to metamorphose. J Feeding movements were weak and intermittent in some postlarvae on day 1, but strong and regular on day 2. b Feces contained diatoms (Navicula ramosissima) on day 2. c Ginger color of the future digestive gland began to develop on day 3, lateral to the yolk. movements were visible, although weak and intermittent. No in- gestion or feces were observed, but ciliary rotation was visible in the gut between the heart and yolk. By 2 days postsettlement. feeding movements were strong and regular, material was being ingested and rotated in the gut. and fecal pellets were being pro- duced. The feces were dominated by the diatom food supplied (Navicula ramosissima). By 3 days postsettlement, the ginger color of the future digestive gland had begun to develop, lateral to the yolk. The ginger color spread throughout the yolk by day 7, progressively obscuring and/or replacing the yolk. Diatom Ingestion/Digestion/Egestion The five diatom strains showed four different patterns with respect to digestion efficiency and ingestibility. Cylindrotheca closterium represented the simplest situation. This diatom was common in fecal pellets by day 9 and had high digestion efficiency throughout the experiment (Fig. 1A). For all other strains, the ingestibility or digestion efficiency changed in relation to postlar- val development or diatom culture condition. Cocconeis scutellum was digested very efficiently by all ages of postlarvae (Fig. 1A). However, young postlarvae were ineffi- cient at ingesting C. scutellum, and were observed grazing repeat- edly and smoothly over the same diatoms without detaching or rupturing the cells. Occasional C. scutellum cells were present in some fecal pellets of postlarvae on day 10. but they contributed a very small proportion of the fecal volume. The postlarvae became progressively more efficient at ingesting C. scutellum as they grew to ~1 mm SL at day 30 (Fig. 2). The fecal pellets of postlarvae >1 mm SL were densely packed with C. scutellum frustules. Both Navicula britannica and Navicula ramosissima were common in fecal pellets throughout the experiment. Their diges- tion efficiency was initially low. but increased during the experi- ment. The change in digestion efficiency was accompanied by a change in the integrity of fecal pellets. When digestion efficiency was low, live cells moved rapidly out of fecal pellets, and pellets dispersed within -10 to 30 minutes. When digestion efficiency was Growth and Survival of Postlarval Abalone 247 0.3 0.6 0.9 1.2 Shell length (mm) Figure 2. As postlarvae grew, they became more efficient at ingesting Cocconeis scutellum cells. Ingestion efficiency was measured as the number of diatom cells per unit volume of feces. Data are mean ± SE, n = 5. high, the pellets retained their shape. For N. britarmica, digestion efficiency rose sharply between day 45 and day 61 (Fig. 1A). Navicula ramosissima feces showed a similar pattern, with fecal pellets largely dispersing on days 9. 18. 24. and 31 but persisting in subsequent observations. On the latter occasions, accurate di- gestion efficiency counts were not possible because of the compact nature of the fecal pellets and the small cell size of N. ramosissima. A modified technique was developed (see Methods) and used in a follow-up experiment. This showed that young postlarvae (14-16 days postsettlement, 600-700 u.m SL) had high digestion efficien- cies for both N. britarmica (mean = 89%, SE = 4.5, n = 5) and N. ramosissima (mean = 669c. SE = 6.2. n = 5). This suggests that the change in digestion efficiency was attributable to a change in the diatom culture rather than a postlarval development. The postlarvae fed Plewosigma sp. grazed less actively than those fed on other diets, and -60% did not produce feces when transferred to clean containers. Fecal volumes were low and vari- able (Fig. 1 B ), and Pleurosigma cells were not present in feces on days 9 to 3 1 . Parallel experiments with spare postlarvae transferred weekly to Pleurosigma sp. showed that Pleurosigma cells were rare in feces on days 18 (-700 u.m SL). 24 (-800 u.m SL). and 31 (-1 mm SL). but common on days 45 (-1.55 mm SL) and 61 (-1.6 mm SL). Even these larger postlarvae ingested Pleurosigma sp. inefficiently. The large and loosely attached Pleurosigma cells tended to be deflected by the snout, or lost by the radula before being swallowed. Some cells entered the mouth, but reappeared on the next stroke of the radula, and were dislodged. Most of these cells had been ruptured and lost some of their cell contents. Fecal volume increased rapidly as postlarvae grew, but was highly variable among individuals. The relationship between shell length and fecal volume was approximately cubic (Fig. 3). Postlarval Growth Initial postlarval SL (day 3 postsettlement) was -0.35 mm (Fig. 1 D). Growth rate from days 3 to 10 was similar on all diets (means of 20-29 u.m/day). Growth rate was significantly lower on C. scutellum (Table 3). although the actual difference was small (Fig. IE). Growth rate then dropped for four of the five diets between days 10 and 16 and, with all diets combined, was significantly lower in the second week than the first (paired sample r-test, t = 7.33, p < 001. n = 30). Growth rates from days 10 to 16 were again similar among diets (means 15-20 u.m/day), with the excep- tion of Pleurosigma sp. (2 u,m/day) (Fig. IE. Table 3). After day 16, growth rates varied markedly between diets (Fig. IE), giving rise to diverging growth curves (Fig. ID). The three best diets produced significantly higher growth rates than the two poorest diets by day 23 (Table 3). By day 37. the growth rate on the C. scutellum and C. closterium was significantly higher than that on N. ramosissima (Table 3). Growth on Pleurosigma sp. remained very low until day 30, when data collection was discontinued because of inadequate sur- vival. Growth rate on N. britarmica was low until day 37, then showed a moderate increase in subsequent weeks (Fig. IE). The N. ramosissima diet produced intermediate growth rates, which did not increase markedly after day 30, when the digestion efficiency of this diet increased. C. scutellum produced rapid postlarval growth, with a dip at day 53. A dip in growth rate at day 53 was also observed for N. ramosissima and C. closterium when a longer- than-usual interval between food renewals led to food depletion in these diets. Cylindrotheca closterium produced rapid growth up to day 45. The growth rate then dropped sharply for the following 3 weeks (Fig. IE), when postlarvae repeatedly depleted their food supply. A follow-up experiment with larger experimental containers con- firmed that postlarvae continued to grow very rapidly when C. closterium was provided in excess (average of 68 u.m/day over 19 days from 2.2 mm initial SL. SE = 4.4. n = 6). Specific growth rates were highest from days 3 to 10. then declined during days 10 to 16 (Fig. IF). On the three best diets, specific growth rate then climbed for 1 to 3 weeks and declined thereafter. Postlarval Survival Survival to day 68 correlated positively with growth rate (r = 0.88, p = .05, n = 5). Survival was very low (<10% at day 37) 2.73 j; =0.722.* ' r = 0.84, PO.001 0 12 3 4 Shell length (mm) Figure 3. Relationship between postlarval shell length and fecal vol- ume. Data are from all diets, except Pleurosigma sp. that produced abnormal feeding activity. The regression line and statistics assume a power relationship between x and y. 248 Roberts et al. TABLE 3. Statistical comparison of growth rates (see Fig. IE) of postlarvae fed five diatom diets. Diatom Diet Davs Since Cocconeis Cylindrotheca Navicula Navicula Pleurosigma Settlement scutellum closterium ramosissima britannica sp. 10 B A A A A 16 A A A A B 23 A A A B C 30 A A A B B 37 A AB B C Tukey's groupings from ANOVA are presented for five times in the experiment. For each time, those diets in the same Tukey's group are given the same letter. Tukey's group "A" represents significantly (/> < .05) higher growth rate than "B." which is higher than "C." Low survival precluded the inclusion of Pleurosigma sp. beyond day 30. on Pleurosigma sp.. intermediate (48% at day 68) on N. britan- nica. and high (>70% at day 68) on the three best diets. C. scutel- lum, C. closterium, N. ramosissima (Fig. 1C). Survival to day 37 was near 100% on all diets other than Pleurosigma sp. (Fig. 1C). Mortality in the Pleurosigma sp.. C. closterium. and N. britannica diets occurred steadily: whereas, the mortalities in the C scutellum diet all occurred in a single week (Fig. 1C). DISCUSSION Diatom strain was a major determinant of postlarval survival and growth, which were positively correlated. Both the ingestibil- ity and digestion efficiency of diets were important, and their influence varied over time. This temporal variation was attribut- able to postlarval development and changes in diatom cultures. Dietary Benefits vs. Postlarval Size Diatom diet made little difference to growth rates in the first 2 to 3 weeks after settlement, but older postlarvae grew more rapidly on the efficiently digested diatoms. Growth rates on the four in- gestible diets (excludes Pleurosigma sp.) were similar for the weeks ending day 10 and day 16 (Fig. IE), despite widely varying digestion efficiencies (Fig. 1A). Postlarvae were -0.8-1.0 mm SL when divergence of growth rates among diets became significant (Fig. ID. E, Table 3). Previous studies have suggested that dietary benefits were size dependent in postlarval abalone (Kawamura and Takami 1995, Kawamura et al. 1995. Kawamura et al. 1998a) but this is the first time that this relationship has been quantified with a continuous dataset. There are several possible reasons for this relationship. The first relates to development of the digestive system. By day 2 postsettlement. postlarvae ingested diatoms, rotated them in the stomach, and excreted ruptured diatom frustules. It is possible, however, that small postlarvae lack the digestive capabilities to fully utilize diatom cell contents. We observed progressive devel- opment of the digestive gland from day 3. but we do not know when it became a functional organ. Takami et al. ( 1998) detected polysaccharases in H. discus hannai at -1 mm SL, but not at -0.5 mm SL, suggesting that digestive enzymes may develop at about the same time that postlarvae begin responding to better diets. A second possible explanation relates to nutrition from resid- ual yolk. As essential micronutrients from the yolk become de- pleted, postlarvae may require access to diatom cell contents to obtain these compounds and maintain rapid growth. Diatoms with low digestion efficiency may not provide a nutritionally complete diet. Importance of Ingestibility Pleurosigma sp. was a poor diet primarily because it was not efficiently ingested. The cells were very large, ( 1 15 x 16 x 8.5 u.m) and loosely attached and were not ingested by the small postlarvae in the growth experiment. Even large postlarvae of 1.5-2 mm SL were inefficient at ingesting Pleurosigma sp. The importance of diatom cell size has previously been discussed in relation to size of the postlarval mouth (Seki and Kan-no 1981. Fleming et al. 1996). The mouth is 30 p.m across just 2 days after metamorphosis begins (350 p.m SL) (Seki and Kan-no 1981). Given that most diatom cells are much less than 30-p.m wide, mouth size is unlikely to limit ingestion. Our observations suggest that a more important factor is the superior efficiency with which the radula handles small cells. Large and loosely attached Pleurosigma sp. cells were swept aside, or lost, from the radula during feeding. The efficiency of feeding on various diatoms may be determined by the morphol- ogy of the developing radula (Roberts et al. submitted). The growth pattern seen on Pleurosigma sp. was typical of inadequate diets. The postlarvae fed Pleurosigma sp. grew rapidly to 540 u.m SL by day 10, then very slowly to -600 p.m SL by day 30. Similar "tailing-off ' of growth at 500 to 700 p,m SL has been reported previously for poor diets, including trail mucus (Takami et al. 1997a). coralline algae without diatoms (Takami et al. 1997b). and noningestible Cocconeis spp. (Kawamura and Takami 1995. Takami et al. 1997a). The relatively rapid growth observed in the first week may be sustained by residual yolk and uptake of dissolved organic matter (DOM). Previous research found that postlarval abalone only began ingesting C. scutellum once they reached about 800 p.m SL (Daume et al. 1997. Takami et al. 1997a). Our experiment quan- tified that observation and documented a steady increase in inges- tion efficiency rather than a sudden transition. The abruptness of the change may be influenced by the attachment strength of the diatoms and the number of loose cells in the culture. The reason for increasing ingestion efficiency on Cocconeis diets is not obvious. Increasing postlarval strength is unlikely to explain the change, because the radula of young postlarvae grazed smoothly over C. scutellum cells, rather than gripping cells but failing to remove/ rupture them. The radula undergoes various morphological devel- Growth and Survival of Postlarval Abalone 249 opments during this part of the postlarval period, but the link between these changes and the ingestion of C. scutellum is equiv- ocal (Roberts et al. submitted). Temporal Change in Diatom Digestion Efficiency The digestion efficiency of N. britannica changed from low (0-30%) to high (80-90%) within 2 weeks. These are the first data documenting major changes in the digestion efficiency of a diatom strain. A similar increase was shown with N. ramosissima, but was not quantified at the time because of difficulty in accurately count- ing the small cells in cohesive fecal pellets. The change in digestion efficiency was apparently caused by a change in the diatom cultures rather than postlarval development. After the digestion efficiency increased, we retested the digestion efficiency of small (600-800 u.m SL) postlarvae and found it to be high. The increase in diatom digestion efficiency persisted for at least several weeks (R. Roberts, unpubl. data). Two factors that can affect digestion efficiency are diatom attachment strength and structural strength (Kawamura et al. 1998a). There was no indica- tion that the attachment strength of N. britannica and N. ramosis- sima changed during our experiment. Reduction in structural strength is a possibility but was not measured. The ability to con- trol diatom digestion efficiency would be useful for abalone hatch- eries. Fecal Volume vs. Postlarval Size Regression analysis showed that fecal production increased roughly cubically in relation to shell length (Fig. 3), as expected for a length-to-volume relationship. However, fecal volume varied greatly among replicates, and shell length explained only 70% of variation in fecal volume across four diets. Thus, the relationship was not precise, and the regression equation should not be inter- preted accordingly. However, the data do confirm the rapidly in- creasing food consumption of growing postlarvae, and, with re- finement, such a regression may be useful in predicting food re- quirements in a hatchery. Overlap Between Yolk and Particulate Food Our data suggest that nutrition sources other than particulate food are important in early postlarval life. Growth rates were rela- tively high between days 3 and 10 postsettlement, and dropped significantly for days 10 to 17. Our indirect evidence is consistent with the analyses of Shilling et al. (1996), who demonstrated that yolk and uptake of DOM can contribute substantially to early postlarval development in Haliotis rufescens. The lower growth rates in the second week of our experiment may reflect exhaustion of yolk reserves, reduction of DOM uptake, or a reduction in the relative contribution from these energy sources. The metabolic requirements increase 3 to 5-fold from larva to early postlarva (Shilling et al. 1996) and would continue to increase rapidly as postlarvae grow. Presumably, there must be a point at which yolk and/or DOM uptake become insignificant as supplementary energy sources for feeding postlarvae. Factors Controlling Postlarval Survival Abalone in hatcheries often suffer 90 to 100% mortality by 2 months postsettlement (Searcy-Bernal 1992) and survival of 20 to 40% is regarded as excellent (e.g., Leighton 1989). Our experi- ments recorded >70% survival to 68 days on three out of five diets. Our experiments were conducted in small containers without an- tibiotics, with water exchanged each 3 to 4 days and postlarvae transferred to a new diatom culture weekly. The species of diatom dominating the biofilm was very impor- tant, with both the ingestibility and the digestibility of the diatom affecting growth and survival (discussed above). Our data also suggest that constancy of food supply is important. The few in- stances of food limitation we encountered resulted in rapid slowing of postlarval growth (e.g., day 53 data in Fig. IE). In our experi- ment, food limitation did not result in any immediate mortalities, but longer-term food limitation could be a major source of mor- tality. Although adult abalone can survive protracted periods with- out food (e.g.. Carefoot et al. 1993). the tolerance of postlarval abalone has not been determined. Constant food supply in hatch- eries is complicated by rapidly increasing food demand. This study suggests that fecal production rates (and presumably food con- sumption) increase approximately cubically in relation to shell length (Fig. 3). This study emphasizes the importance of diatom diet in con- trolling the growth and survival of postlarval abalone. The ingest- ibility and digestibility of diatoms affects growth rates. We find that ingestibility of some diatom strains changes as postlarvae grow, and the digestibility of a diatom strain can vary markedly over time. Improved understanding and control of the diatoms used in abalone hatcheries should improve the efficiency and consis- tency of juvenile production. ACKNOWLEDGMENTS We thank Henry Kaspar and Yoh Yamashita for constructive review of drafts. Larval abalone were supplied in part by Island Hatcheries Ltd. This research was supported by the New Zealand Foundation for Research Science and Technology, the Asia 2000 Foundation of New Zealand, the New Zealand Ministry for Re- search Science and Technology, an Alliance Group postgraduate fellowship through the University of Otago, and the Japan Fish- eries Agency. LITERATURE CITED Carefoot, T. H., P. Qian, B. E. Taylor. T. West & J. Osborne. 1993. Effect of starvation on energy reserves and metabolism of the northern aba- lone Haliotis kamtschatkana. Aquaculture 118:315-325. Daume, S., S. Brand & W.J. Woelkerling. 1997. Effects of postlarval abalone [Haliotis rubra) grazing on the epiphytic diatom assemblage of coralline red algal surfaces. Moll. Res. 18:119-130. Fleming, A. E.. R. J. Van Barneveld & P. W. Hone. 1996. The develop- ment of artificial diets for abalone: a review and future directions. Aquaculture 140:5-53. Ishida, T., T. Akutsu & K. Torisawa. 1995. Effects of monocultured benthic diatoms on the metamorphosis of veliger larvae and on the growth of juveniles of abalone, Haliotis gigantea. Bull. Shizuoka Pref. Fish. Exp. Sm. 30:17-21. Jorgensen. E. G. 1962. Antibiotic substances from cells and solution of unicellular algae with special reference to some chlorophyll deriva- tives. Physiol. Plant. 15:530-545. Kawamura. T. & R. Hirano. 1992. Seasonal changes in benthic diatom communities colonizing glass slides in Aburatsubo Bay. Japan. Diatom Res. 7:227-239. Kawamura. T.. T. Saido. H. Takami & Y. Yamashita. 1995. Dietary value 250 Roberts et al. of benthic diatoms for the growth of postlarval abalone Haliolis discus hannai. J. Exp. Mar. Biol. Ecol. 194:189-199. Kawamura. T. & H. Takami. 1995. Analysis of feeding and growth rate of newly metamorphosed abalone Haliotis discus hannai fed on four spe- cies of benthic diatom. Fisheries Sci. 61:357-358. Kawamura. T., R. D. Roberts & C. M. Nicholson. 1998a. Factors affecting the food value of diatom strains for postlarval abalone Haliotis iris. Aquaculture 160:81-88. Kawamura. T.. R. D. Roberts & H. Takami. 1998b. A review of the feeding and growth of postlarval abalone. J. Shellfish Res. 17:615-625. Kirby, K. N. 1993. Advanced data analysis with SYSTAT. Van Nostrand Reinhold, New York. 475 pp. Kitting, C. L. & D. E. Morse. 1997. Feeding effects of postlarval red aba- lone Haliotis rufescens (Mollusca: Gastropoda) on encrusting coralline algae. Moll. Res. 18:183-196. Leighton. D. L. 1989. Abalone (Genus Haliotis) mariculture on the North American Pacific Coast. Fish. Bull. 87:689-702. Martinez-Ponce. D., R. Searcy-Bernal & C. Anguiano-Beltran. 1998. Graz- ing rates of red abalone. Haliotis rufescens. postlarvae. /. Shellfish Res. 17:627-630. Matthews, I. & P. A. Cook. 1995. Diatom diet of abalone postlarvae {Hali- otis midae) and the effect of pregrazing the diatom overstorey. Mar. Freshwater Res. 46:545-548. Morse. D. E., H. Duncan, N. Hooker & A. Morse. 1977. Hydrogen per- oxide induces spawning in mollusks, with activation of prostaglandin endoperoxide synthetase. Science 196:298-300. Norman-Boudreau. K.. D. Bums. C. A. Cooke & A. Austin. 1986. A simple technique for detection of feeding in newly metamorphosed abalone. Aquaculture 51:313-317. Ohgai, M., M. Wakano & S. Nagai. 1991. Effect of attached microalgae on the settlement of juveniles in abalone. Haliotis discus hannai Ino. Su- isanzoshoku 39:263-266. Roberts. R. D. & C. M. Nicholson. 1997. Variable response from abalone larvae (Haliotis iris. H. virginea) to a range of settlement cues. Moll. Res. 18:131-141. Roberts, R. D.. R. Searcy-Bernal & P. A. Cook. 1998. A workshop on the culture of larval and postlarval abalone. Proceedings of the 3rd Inter- national Abalone Symposium. October 26-31. Monterey. California. USA. Cawthron Rep. 477. 5 pp. Roberts. R. D.. T. Kawamura & H. Takami. submitted. Morphological changes in the radula of abalone (Haliotis iris) during postlarval de- velopment. J. Shellfish Res. Searcy-Bernal. R.. A. E. Salas-Garza & R. A. Flores-Aguilar. 1992. Re- search in Mexico on the critical stage of abalone (Haliotis spp.) seed production, pp. 547-560. In: S. A. Shepherd, M. J. Tegner. and S. A. Guzman del Proo (eds.). Abalone of the World: Biology. Fisheries, and Culture. Fishing News Books, Oxford. Seki. T. & H. Kan-no. 1981. Observations on the settlement and metamor- phosis of the veliger of the Japanese abalone. Haliotis discus hannai Ino. Haliotidae. Gastropoda. Bull. Tohoku Reg. Fish. Res. Lab. 42:31- 39. Shilling. F. M„ O. Hoegh-Guldberg & D. T. Manahan. 1996. Sources of energy for increased metabolic demand during metamorphosis of the abalone Haliotis rufescens (Molluscal. Biol. Bull. 191:402—412. Takami. H.. T. Kawamura & Y. Yamashita. 1997a. Survival and growth rates of postlarval abalone Haliotis discus hannai fed conspecific trail mucus and/or benthic diatom Cocconeis scutellum var. parva. Aqua- culture 152:129-138. Takami. H.. T. Kawamura & Y. Yamashita. 1997b. Contribution of dia- toms as food sources for postlarval abalone Haliotis discus hannai on a crustose coralline alga. Moll. Res. 18:143-151. Takami. H.. T. Kawamura & Y. Yamashita. 1998. Development of polysaccharide degradation activity in postlarval abalone Haliotis dis- cus hannai. J. Shellfish Res. 17:723-727. Tong. L.J. & G. A. Moss. 1992. The New Zealand culture system for abalone. pp. 583-591. In: S. A. Shepherd. M.J. Tegner. and S. A. Guzman del Proo (eds.). Abalone of the World: Biology, Fisheries, and Culture. Fishing News Books, Oxford. Journal of Shellfish Research. Vol. 18. No. 1. 2? 1-257. 1999. HISTOLOGICAL ANALYSIS OF MANTLE-CAVITY CILIATES IN DREISSENA POLYMORPHA: THEIR LOCATION, SYMBIOTIC RELATIONSHIP, AND DISTINGUISHING MORPHOLOGICAL CHARACTERISTICS F. LARUELLE,1 D. P. MOLLOY,2 S. I. FOKIN,3 AND M. A. OVCHARENKO4 1 UMR CNRS 6539 Institut Universitaire Europeen de la Mer UBO, Place Nicolas Copernic Technopole Bresl-Iroise 29280 Plouzane, France 2Biological Survey, New York State Museum The State Education Department Cultural Education Center Albany, New York 12230 'Biological Research Institute St. Petersburg State University St. Petersburg 198904, Russia Institute of Hydrobiology Ukrainian Academy of Sciences 12 Prospect Geroyev Stalingrada Kiev-210 254655, Ukraine ABSTRACT Dissection has traditionally been the sole method used in investigations of the parasites and other endosymbionts of zebra mussels. Dreissena polymorpha. This study demonstrates the value of histological analysis as a complementary technique capable of precisely determining the location of ciliates within zebra mussels and characterizing their symbiotic relationships at the cellular level. The photomicrographs included herein represent the first published histological images of mantle-cavity ciliates of zebra mussels, and we have highlighted morphological characteristics useful in distinguishing individual ciliate species in histological sections. Although zebra mussels from both North America and Europe were sampled for this study, only European populations were found to harbor mantle-cavity ciliates. and five species were observed. The host-specific species Conchophthirus acuminatus (Scuticociliatida: Conchophthindae) was frequently recorded from epithelium covering the outer gill surfaces and occasionally from visceral mass epithelium, but also found in four previously unreported regions: frequently within gill water tubes and occasionally on labial palps, mantle epithelium, and within suprabranchial cavities. Although we sometimes observed zebra mussel sperm in food vacuoles of C. acuminatus, epithelial tissues in contact with high densities of these ciliates showed no evidence of pathology, thus confirming this species' commensal nature. The host-specific species Sphenophrya dreissenae (Rhynchodida: Sphenophryidae) was frequently re- corded attached to mantle cavity epithelium and outer gill surfaces, but also found in three previously unreported regions: frequently within the gill water tubes, occasionally on the visceral mass, and rarely within the suprabranchial cavities. High-intensity infections with this parasitic ciliate did induce hyperplasia, cell hypertrophy, and vacuolization of the epithelia. The host-specific species Hypocomagalma dreissenae (Rhynchodida: Ancistrocomidae) was most frequently observed attached to epithelial cells lining outer gill surfaces, but also in five previously unreported regions: occasionally on the visceral mass, the mantle cavity epithelium, and in gill water tubes, and rarely on labial palps and within the suprabranchial cavities. This parasitic ciliate feeds on the contents of epithelial cells using a suctorial tentacle. The intensity of H. dreissenae infection, however, was usually very low, and no adverse effects on parasitized cells or nearby tissues were evident. The ciliate Ancistrumina limnica (Scuticociliatida: Ancistridae), a nonhost-specific commensal of mollusks, was recorded frequently within gill water tubes, occasionally on outer gill epithelia. and rarely within suprabranchial cavities. This species was also observed to have ingested D. polymorpha sperm cells. Commensal Peritrichia ciliates were also occasionally observed within the mantle cavity, but were likely carried there passively by water currents from their typical location on shell surfaces. The presence of "mantle cavity" ciliate species in the gill water tubes and the suprabranchial cavities of zebra mussels suggests that these ciliates probably can exit into surrounding waters to infect other zebra mussels via the exhalant siphon. KEY WORDS: zebra mussels, ciliophora. Conchophthirus. Sphenophrya, Hypocomagalma. Ancistrumina, Peritrichia INTRODUCTION raw- water conduits within infrastructures (O'Neill 1996. O'Neill 1997). they have also caused significant environmental impacts Zebra mussels. Dreissena polymorpha, were likely transported (Maelsaac 1996). Although considerable research has been carried from Europe to North America in the ballast water of transoceanic out to understand the ecological interrelationships of these bi- vessels (Carlton 1993). Within a few years after their discovery in valves with other aquatic organisms, relatively little effort has been Lake St. Clair in 1988 (Hebert et al. 1989), these freshwater, made to investigate the diversity, distribution, and significance of macrofouling. bivalves were found in high densities throughout the endosymbiotic organisms present within these mussels. We con- Great Lakes Basin. Populations have thus far been reported as far ducted this investigation of Dreissena's mantle cavity ciliates to south as Louisiana and as far west as Oklahoma (New York Sea address this information gap. Grant 1998). Besides the economic impact of their fouling of Interest in the prevalence and types of endosymbionts led to the 251 252 Laruelle et al. establishment of the International Research Consortium on Mol- luscan Symbionts (IRCOMS) — a project currently focusing on the endosymbionts of Dreissena spp. both in Europe and North America. Although comprehensive data on the prevalence and intensity of mantle-cavity infections in zebra mussels are currently being prepared for publication by IRCOMS members, we thought it best to issue this current report so that the scientific community might have access to the photographic images of the histological appearance of the mantle-cavity ciliates that we have observed to date, as well as the latest information on the nature of their sym- biotic relationships and their precise locations within zebra mus- sels. The presented photomicrographs represent the first published histological images of D. polymorpha's mantle-cavity ciliates. Pre- vious investigations have relied almost exclusively on observa- tions made during dissections of live zebra mussels to determine where endosymbiotic ciliates were present within D. polymorpha and whether these ciliates had any adverse effect. The present study demonstrates the value of using histological analysis as a complimentary technique. The information and images contained herein will, it is hoped, prove to be useful to parasitologists, as well as to other researchers (e.g., toxicologists, physiologists, etc.) who need to identify ciliates they encounter within zebra mussel tissue sections. Of all symbiotic endofauna in zebra mussels. Dreissena poly- morpha. mantle-cavity ciliates have been the most frequently re- ported (Molloy et al. 1997). In Europe, the following five species of mantle-cavity ciliates are believed to have an obligate and host- specific association with D. polymorpha: Conchophthirus acumi- natus, Conchophthirus klimentinus, Hypocomagalma dreissenae. Sphenophrya dreissenae, and Sphenophrya naumiana. Ancistru- mina limnica, a nonhost-specific commensal of European fresh- water mollusks, has been recorded only once from D. polymorpha (Raabe 1956). Among the five obligate European species, C. kli- mentinus and S. naumiana have been reported from only one lo- cation (Lake Ohrid in Macedonia; Raabe 1966). The three other obligate species seem to have a much broader geographic distri- bution throughout Europe (Molloy et al. 1997). There is growing evidence that none of these European ciliates successfully accom- panied zebra mussels in their trans-Atlantic crossing since these ciliates have not yet been observed from North American D. poly- morpha populations (Molloy et al.. 1997; authors, unpublished data). MATERIALS AND METHODS During 1992 to 1997, D. polymorpha populations were sampled both within North America (authors, unpublished data) and Europe (Table 1 ) and examined for endosymbionts by dissec- tion. At the same time, subsamples were fixed for histology in 10% neutral buffered (sodium phosphate) formalin. After dehydration in increasing alcohol concentrations, the mussels were embedded in paraffin, sectioned at 5 u,m. and stained with hematoxylin and eosin. Photomicrographs of mantle-cavity ciliates were produced, and data on the precise location of ciliates within their hosts and their effect on host tissues were recorded. Species identifications of ciliates in the histological sections were facilitated by records of which ciliates were observed from the same sample during dissec- tions, as well as from knowledge of which mantle-cavity ciliate species had previously been documented from D. polymorpha (re- viewed in Molloy et al. 1997). TABLE 1. Source of Dreissena polymorpha used for photomicrographs. Figure Country Population Collection Date 1A, IC Russia IE, 2A, Russia 2B IF France 2D Russia 2E. H Russia 2F Belarus 3A. B Netherlands 3C Greece Ivankovskoye Reservoir. Dubnya Moscow River, Rublevo Moselle River, Mirgenbach Volkhof River, Novaya Ladoga Volkhof River, Novgorod Dnieper-Bug canal. Pinsk Volkerak Lake. Ooltgensplaal Lake Volvi. Thessaloniki 93/07/10 and 94/08/01 93/07/1 1 93/09/15 93/07/16 93/07/21 97/06/30 92/04/19 95/10/27 RESULTS AND DISCUSSION In this study, the following five species of ciliates were en- countered within the mantle cavity of European D. polymorpha populations: C. acuminatus, S. dreissenae, H. dreissenae, A. lim- nica. and an unidentified species belonging to the subclass Per- itrichia. It is rarely possible to accurately identify ciliates at the species level solely from histological slides, because key taxo- nomic characters, such as the number and pattern of kineties on the body, are not visible. Except for the peritrichs, however, these species have all been previously reported from European zebra mussels and are sufficiently different morphologically. When a ciliate's full length was visible in the section (e.g., approximately medial longitudinal or saggital cut through the ciliate), they could often be distinguished from each other using key morphological characteristics as discussed below and further detailed in Table 2. Conchophthirus acuminatus (Scuticociliatida: Conchophthiridae) C. acuminatus has never been reported in any other species of freshwater bivalve and is thus considered an obligate, host-specific endosymbiont of D. polymorpha. Key characteristics that distin- guish it from other endosymbionts include the cytopharynx, which appears as a groove directed toward the anterior end of the body (Fig. 1A.B), the small micronuclei. the ovoid-oblong 15 to 20 u,m macronucleus located mainly in the anterior half of the body (Fig. IB). Their body shape is also distinctive, because they are ex- tremely laterally compressed. When creeping on or attached to a surface, they lie on their side; that is, not on their ventral surface (Fig. IC). C. acuminatus has been previously reported only from the ep- ithelial outer surface of the gills and visceral mass. We frequently recorded this species at the point of attachment of the gills to the visceral mass. In addition, we also observed this species com- monly throughout the entire length of gill water tubes (Fig. 1 A.C) and occasionally on labial palps, mantle epithelium, and in the suprabranchial cavities. Ciliates were observed either free in the cavities (Fig. 1 A.B.C) or adhering to epithelium (Fig. 1A) through use of their anterior thigmotactic cilia. Our observations supported the concept that C. acuminatus is a commensal, because no patho- logical effects on adjacent host tissues were observed, even when ciliate densities were high. In studies of Conchophthirus and other scuticociliatids, several authors have reported food vacuoles in these ciliates to contain Mantle-Cavity Ciliates in Zebra Mussels 253 TABLE 2. Distinguishing morphological characteristics of the common mantle-cavity ciliates in European Oreissena polymorpha samples. Species Macronucleus Shape Cytoplasm Cell Size (urn I Other Features Conchophthirus acuminatus Ovoid'' to oblong Sphenophrya dreissenae Highly irregular0 (often globular to banana-shaped) Hypocomagalma dreissenae Ancistrumina limnica Ovoid to spindle-shaped1 Ellipsoidal8 to globular Relatively colorless Heterogeneous with prominent inclusions and vacuoles Almost always pink-purpleb Homogeneous without prominent inclusions and vacuoles Often pink1 with lightly stained anterior tip Heterogeneous with prominent inclusions and vacuoles Relatively colorlessb Heterogeneous with prominent inclusions and vacuoles. 50-120J Prominent grooved cytopharynx Micronuclei often visible Presence of cilia 27-37d Lack of cilia Helmet shape Presence of protruding processes Highly basophilic macronucleus High macronucleus/cytoplasm ratio 32-50' Presence of an attachment knob with clear tube Presence of cilia 18-43* Presence of cilia Lateral elongate cilia on only one side of the body 'Raabe 1971. ' Using standard hematoxylin and eosin staining procedures (see Materials and Methods section). Dobrzanska 1961. 1 Dobrzanska 1958. ■' Jarocki and Raabe 1932. 'Raabe 1970. ? Measurements from present study. Raabe ( 1947) indicated a length of 35 u,m in species description, but did not include a range. host tissues (Kidder 1934, Kirby, 1941, Antipa and Small 1971) and a mix of algae or bacteria (Kidder 1933, Kirby 1941 ). Because no study has been conducted on the food vacuoles of C. acumi- natus, the diet of this species has remained unclear. Fenchel ( 1965) considered that Conchophthirus spp. in bivalves were incapable of ingesting suspended particles, because their adoral ciliature was very much reduced. If this is true of C. acuminates, then this species likely feeds on particles present on epithelial surfaces. We sometimes observed C. acuminatus containing sperm cells within food vacuoles (Fig. 1C). A similar phenomenon was observed in a scuticociliatid ciliate species in Mytilus edulis (Kidder 1933). Be- cause some of the zebra mussels containing these sperm-ingesting C. acuminatus were female, the ingested sperm had to be produced by another zebra mussel in the vicinity and taken into the mantle cavity through the female's inhalant siphon. The only other conchophthirid species from D. polymorpha. C. klimentinus, is apparently very rare and has been reported only from one waterbody in Macedonia (Raabe 1966). Because it is similar in size (morphology reviewed in Molloy et al. 1997), it could be difficult to distinguish from C. acuminatus in histological preparations. Its overall shape, however, is different, because C. klimentinus has an ovoid outline and is slightly rounded at both ends (Raabe 1966, Raabe 1971). Sphenophrya dreissenae (Rhynchodida: Sphenophryidae) Ciliates in the family Sphenophryidae live on the gills of bi- valves and are highly specialized morphologically. Adults lack cilia and have lost the streamline body form typical of mantle- cavity ciliates (Fenchel 1965). Relative to other mantle-cavity cili- ates, S. dreissenae has a high macronucleus/cytoplasm ratio (Fig. ID). The macronucleus is distinctive (Fig. IE), because it is rela- tively large, densely basophilic, and highly irregular in shape (al- though often globular to banana-shaped). The cytoplasm is par- ticularly unusual, because it is rather homogeneous and does not contain prominent food vacuoles or cytoplasmic inclusions — a condition probably related to the absence of a mouth and absorp- tion of dissolved nutrients directly from epithelial cells. In fact, all Sphenophryidae have no mouth opening, and in ciliate evolution this is considered a secondary phenomenon relating to parasitism (Fenchel 1965). Although the feeding process of sphenophryid ciliates has not yet been precisely determined, two hypotheses have been offered. Chatton and Lwoff (1921) suggested that Sphe- nophrya spp. used the entire surface in contact with the host epi- thelial cells as a "sucking disk", whereas, Kirby (1941) theorized that they fed osmotically. S. dreissenae are typically attached to the gill epithelium using their flat clinging sole, and thus may appear flat or even concave (Dobrzanska 1958). Previously reported only from the epithelium on outer gill surfaces or lining the mantle cavity, we also observed these ciliates frequently within the gill water tubes, occasionally on the visceral mass, and rarely within the suprabranchial cavities (Fig. IE, 2A,B). S. dreissenae reproduces by budding, and one to several layers of individuals were often observed coating the vis- ceral mass epithelium or the gill inner epithelium (Fig. IE, 2A.B). Foci of high ciliate intensity frequently resulted in tissue damage, including epithelial hyperplasia, cell hypertrophy, and extensive vacuolization (Fig. IE). Epithelial cell hypertrophy and hyperpla- sia resulted in papillary protrusion of the epithelium into the water tubes thus increasing the surface area in contact with the ciliates (Fig. 2A,B). Secretory material, likely mucous, was frequently observed at the apical ends of the epithelial cells; in most of our photomicrographs, this secretory material had a homogeneous tex- ture (see arrows. Fig. 2B) somewhat similar to the cytoplasm of the S. dreissenae cells themselves. Although 5. dreissenae has always 254 Laruelle et al. -3L wt * Figure 1. (A) Oblique section through a gill showing three Conchophthirus acuminates (arrows); note that one of them is in a water tube. (B) Lateral view of C. acuminatus free in the mantle cavity; note the cytopharynx, dense cilia, and macronucleus in anterior of cell. (C) Dorsoventral view of C. acuminatus in a water tube; note Dreissena sperm cells within its vacuoles, dense cilia, macronucleus in anterior of cell, and relatively long, lateral compressed body. (D) Isolated Sphenophrya dreissenae clinging on gill tissue; note the high macronucleus/cytoplasm ratio, lack of cilia, relatively homogeneous cytoplasm, and the helmet shape. (E) Details of an interlamellar bridge infected by S. dreissenae; note the protrusion of the epithelium caused by epithelial hyperplasia and the vacuolization of the cells on the infected side; compare to normal interlamellar bridges in Fig. 1A. Scale bar = 5 urn (CD); 1(1 um (B); 20 um (E); 50 um (Al. age: abnormal gill cells, ci: cilia, cp: cytopharynx, ge: gill epithelium, ib: interlamellar bridge, ma: macronucleus, mc: mantle cavity, mi: micronucleus, ngc: normal gill cells, sc: sperm cell, wt: water tube. been considered as being parasitic (Dobrzanska 1958, Dobrzanska 1961). our histological observations of adverse tissue reaction leading to gill deformities is the first conclusive evidence of pa- thology. The relatively small size (27 x 37 p.m). helmet shape, lack of cilia, occasional presence of protruding processes, large nucleus/ cytoplasm ratio, homogeneous cytoplasm, and dense irregularly shaped, highly basophilic macronucleus are characteristics that distinguish S. dreissenae from other recorded mantle-cavity cili- ates. The only other Sphenophrya documented from D. polymor- pha, S. naumiana, is different in shape (elongate to canoe-shaped). has an elongate nucleus, and is generally larger in size (Raabe 1966). Hypocomagalma dreissenae ( Rhynchodida: Ancistrocomidae) H. dreissenae, as with other Hypocomagalma spp.. have at their anterior tip a suctorial tentacle that acts in part as an attachment knob (Fig. 2E.F). The presence of this structure readily separates this species from all other ciliates reported from D. polymorpha. Within the anterior of this species, we often observed a clear tube leadinc to the attachment knob — similar to the "hollow tube" re- Mantle-Cavity Ciliates in Zebra Mussels 255 A f >»» y- :■ v Wrl: i; v. #1 tffeSKB^ 2P* ■* -. * swBCF j«K iM^efi* ' » . D ** * *-£b*^ wt ^t - 1 i 4< ink / mc • 1 JL?| vme F % ', "I 0*9 wt Figure 2. (A) and (B) Oblique sections through gill demibranch showing proliferations of Sphenophrya dreissenae in water tubes; note secretory material, likely mucous, at the apical ends of the epithelial cells (arrows) and the swelling of the epithelial lining in infected water tubes. (C) Oblique section through gill with a pyriform Hypocomagalma dreissenae (arrow I present within a water tube. (D) Two pyriform H. dreissenae within a gill yyater tube. (E) Banana-shaped H. dreissenae on the yisceral mass epithelium: note globular macronucleus, cilia, and thin clear tube (arrow) in anterior tip. (F) H. dreissenae in a yyater tube; note spindle-shaped macronucleus partially obscured by yacuoles and the clear tube (arrow) at the anterior end. Scale bar = 5 um (F): 10 pm (D,E); 50 um (B,C(: 80 pm (A), gf: gill filament, ib: interlamellar bridge, ma: macronucleus, mc: mantle cavity, vme: yisceral mass epithelium, wt: water tube. ported in other rhynchodid species (Lom and Kozloff 1968). Al- though there is some overlap in size with 5. dreissenae, the pres- ence of cilia can be used to distinguish H. dreissenae (Fig. 2E) from S. dreissenae (Fig. ID). Other characteristics that help to distinguish H. dreissenae from the other mantle-cavity ciliates are the ovoid to spindle shape of its macronucleus and the heteroge- neity of its cytoplasm (e.g.. relatively prominent granules and vacuoles) (Fig. 2D.E.F). H. dreissenae has previously been reported only from the outer gill surfaces. We observed this frequently, but also noted their presence in five other epithelial regions: occasionally on the vis- ceral mass (Fig. 2E). on the mantle cavity epithelium, and in gill water tubes (Fig. 2C.D). and rarely on labial palps and in su- prabranchial cavities. This species is considered as a true parasite of D. polymorpha, because it inserts its tentacle into the cytoplasm of an epithelial cell, and by some as yet unexplained means, nu- trients from the host cell pass into the ciliate. Bradbury (1994) indicated that Hypocomagalma spp. probably damage the cell to which they are attached, but because they are few in number per host, infection usually has little pathological effect. Our observa- tions of H. dreissenae infection of D. polymorpha were similar. Ancistrumina limnica (Scuticoeiliatida: Ancislridae) A. limnica is a nonhost-specific invader of freshwater lamelli- branchs and gastropods (Raabe 1956). Ciliates such as this species typically feed on bacteria, diatoms, and other material extracted from water currents (Kirby 1941) and are. thus, not considered parasitic. As in the commensal C. acuminatus, however, we did sometimes observe zebra mussel sperm in food vacuoles. In the only previous report of A. limnica in zebra mussels, its precise location within the mantle cavity was not indicated (Raabe 1956). We observed A. limnica frequently within the gill water tubes (especially near the food groove) (Fig. 3A). occasionally on the 256 Laruelle et al. " wt ma Figure 3. (A) Oblique section through gill demibranch showing two Ancistrumina limnica in adjacent water tubes. (B) A. limnica in a water tube; note the distinctive, elongate cilia (ci) on only one side of its body, as illustrated in Raabe (1947). (C) Peritrich ciliate attached to the visceral mass epithelium in the mantle cavity; note elongate, coiled macronucleus. Scale bar = 10 pm (B, C); 35 um (At. ci: cilia, gf: gill filament, ma: macronucleus. mc: mantle cavity, mi: micronucleus, wt: water tube. outer gill epithelium, and rarely in the suprabranchial cavities. No apparent signs of pathology were evident. Identifying characteris- tics of this species (Fig. 3B) are: ( 1 ) the presence of an ellipsoidal to globular macronucleus; (2) relatively colorless, highly vacu- olated cytoplasm (similar in appearance to the much larger ciliate C. acuminates); (3) the lack of an appearance of any firm attach- ment to epithelial surfaces (because they are free swimming and have, in contrast to H. dreissenae and S. dreissenae, no highly- specialized attachment structures); and (4) the presence of rows of lateral elongate cilia on only one side of their body (rarely visible in most histological sections). Peritrichia Ciliates in the subclass Peritrichia with elongate coiled macro- nuclei were occasionally observed in the mantle cavity (Fig. 3C). Although peritrichs have been previously reported from bivalve mantle cavities (Fenchel 1965). this is the first report of these commensal ciliates within zebra mussels. It is likely that these ciliates were attached to visceral mass epithelium, and not simply free floating. In any case, no host reaction was evident in the adjacent epithelia. Because peritrich populations were observed externally on D. polymorphs shells, these ciliates were likely car- ried passively by water currents into the mantle cavity where they reattached. CONCLUSIONS Histological analysis proved to be a very useful technique to complement information gained through dissections of living ma- terial. Tissue sections revealed some symbionts to be more abun- dant than realized from dissections (authors, unpublished data), and the smaller in size and the more sessile the symbiont. the more this was true. Our observations of host tissue condition further strengthened current hypotheses regarding the symbiotic relation- ships of these mantle-cavity ciliates; that is. C. acuminatus, A. limnica. and peritrich ciliates are commensals, and both 5. dreis- senae and H. dreissenae are parasites. Histological analysis was particularly valuable, because it revealed the precise location of each ciliate species on or within zebra mussel organs. It was pre- viously unknown, for example, that the "mantle-cavity" ciliates of zebra mussels also inhabited their gill water tubes and suprabran- chial cavities. It would seem, then, that these ciliates normally migrate with water currents into the water tubes through the gill ostia and then are carried with the water flow up into the su- prabranchial cavities. The presence of these ciliates in these latter water cavities suggests a pathway by which these ciliates may exit their hosts to infect other zebra mussels. For example, laboratory transmission of C. acuminatus directly between infected and un- infected groups of zebra mussels has been achieved (Burlakova et al. 1998), but it has remained unknown from which orifice(s) these ciliates departed from their infected hosts. If such ciliates as C. acuminatus were only present in the mantle cavity per se, they could only leave by the inhalant siphon. This could certainly occur during pseudofeces ejection, when valves are quickly closed, rap- idly forcing water, pseudofeces. and possibly, some ciliates out of the mantle cavity. The presence of "mantle-cavity" ciliates in gill waters tubes and suprabranchial cavities, however, also gives these ciliates the ability to exit via their host's exhalant siphon, because it receives water from the suprabranchial cavities. ACKNOWLEDGMENTS Funding in part from the U.S. Army Engineers Waterways Experiment Station Zebra Mussel Research Program (D.P.M.) and the National Science Foundation Division of International Pro- grams (Robert E. Baier and D.P.M.) is gratefully acknowledged. Special thanks for the histological processing of the samples to the laboratory staff of R. F. Morado. including L. Mooney, L. Chere- Mantle-Cavity Ciliates in Zebra Mussels 257 pow, S. Chilson, and K. Brooks (Fisheries Resource Pathobiology Program. Resource Assessment & Conservation Engineering Di- vision, Alaska Fisheries Science Center. National Marine Fisheries Service, National Oceanic and Atmospheric Administration). We gratefully acknowledge the following for their help in screening the slides and their constructive discussions: L. E. Burlakova (Be- larussian State University), A. Y. Karatayev (Belarussian State University), D. P. Kurandina (Ukrainian Institute of Hydrobiol- ogy). J. F. Morado (U. S. National Marine Fisheries Service), V. A. Roitman (Russian Institute of Parasitology). P. A. Mitrakovitch (Belarusian State University), and V. N. Prostokvashin (Belarusian State University). We also thank L. Giamberini (Universite de Metz) for her loan of histological slides and M. Zarfdjian (Uni- versity of Thessaloniki) for providing mussel samples, and Kyle Alderman (CTC Vermont Color Photo Lab) for assistance with illustrations. Contribution number 777 of the New York State Mu- seum and Science Service. LITERATURE CITED Antipa. G. A. & E. B. Small. 1971. The occurrence of thigmotrichous ciliated protozoa inhabiting the mantle cavity of unionid mollusks of Illinois. Trans. Am. Microsc. Soc. 90:463^172. Bradbury, P. C. 1994. Parasitic protozoa of mollusks and Crustacea, pp. 139-263. In: J. P. Kreier (ed.). Parasitic Protozoa, vol. 8. 2nd ed. Aca- demic Press, San Diego. Burlakova. L. E„ Karatayev. A. Y. & Molloy, D. P. 1998. Field and labo- ratory studies of zebra mussel (Dreissena polymorpha ) infection by the ciliate Conchophthirus acuminatus in the Republic of Belarus. ./. In- vert. Pathol. 71:251-257. Carlton. J. T. 1993. Dispersal mechanisms of the zebra mussel (Dreissena polymorpha). pp. 677-697. In: T. F. Nalepa and D. W. Schloesser (eds.). Zebra Mussels: Biology, Impacts, and Control. Lewis Publish- ers. Boca Raton. Chatton. E. & A. Lwoff. 1921. Sur une famille nouvelle d'acinetiens, les sphenophryides, adaptes aux branchies des mollusques acephales. C. R. Hebd. Seances Acad. Sci. 173:1495-1498. Dobrzanska. J. 1958. Sphenophrya dreissenae sp. n. (Ciliata, Holotricha, Thigmotrichida) living on the gill epithelium of Dreissena polymorpha Pall.. 1754. Bull. Acad. Pol. Sci. Ser. Sci. Biol. 6:173-178 and Figs. 6-10 on unnumbered pages. Dobrzanska, J. 1961. Further study on Sphenophrya dreissenae Dobrzan- ska. 1958 (Ciliata, Thigmotricha). Acta Parasitol. Pol. 9:117-139. Fenchel, T. 1965. Ciliates from Scandinavian mollusks. Ophelia 2:71-174. Hebert, P. D. N., B. W. Muncaster & G. L. Mackie. 1989. Ecological and genetic studies on Dreissena polymorpha Pallas a new mollusk in the Great Lakes. Can. J. Fish. Aquat. Sci. 46:1587-1591. Jarocki. A. & Z. Raabe. Z. 1932. Uber drei neue Infusorien-Genera der Familie Hypocomidae (Ciliata: Thigmotricha). Parasiten in SuGwasser- muscheln. Bull. Acad. Pol. Sci. Lett. 1:29-45. Kidder. G. W. 1933. Studies on Conchophthirius mytili De Morgan. I. morphology and division. II. conjugation and nuclear reorganization. Arch. Protistenkd. 79:1-49. Kidder. G. W. 1934. Studies on the ciliates from freshwater mussels. I. the structure and neuromotor system of Conchophthirus anodontae Stein, C. curtus Engl., and C. magna sp. nov. Biol. Bull. 66:69-90. Kirby, H. 1941. Relationships between certain protozoa and other animals. In: G. N. Calkins and F. M. Summers (eds.). Protozoa in Biological Research, pp. 890-1008. Columbia University Press, New York. Lorn. J. & E. N. Kozloff. 1968. Observations on the ultrastructure of the suctorial tube of ancistrocomid ciliates. Folia Parasitol. t Prague) 15: 291-308. Maclsaac. H. J. 1996. Potential abiotic and biotic impacts of zebra mussels on the inland waters of North America. Am. Zool. 36:287-299. Molloy. D. P.. A. Y. Karatayev, L. E. Burlakova. D. P. Kurandina & F. Laruelle. 1997. Natural enemies of zebra mussels: predators, parasites. and ecological competitors. Rev. Fisheries Sci. 5:27-97 '. New York Sea Grant. 1998. North American range of the zebra mussel. Dreissena (Newsletter of the National Aquatic Nuisance Species Clear- inghouse, SUNY College at Brockport, Brockport. NY). 9:8-9. O'Neill, C. R.. Jr. 1996. The zebra mussel: impacts and control. Cornell Coop. Ext. Inf. Bull. 238:62 pp. O'Neill. C. R.. Jr. 1997. Economic impact of zebra mussels — results of the 1995 National Zebra Mussel Information Clearinghouse Study. Gt. Lakes Res. Rev. 3:35-44. Raabe. Z. 1947. Recherches sur les cilies Thigmotriches ( Thigmotricha Ch. Lw.). II. espece nouvelle d'eau douce du genre Ancistrina Cheissin. Ann. Univ. Mariae Curie-Sklodowska Sect. C Biol. 2:111-120. Raabe, Z. 1956. Investigations on the parasitofauna of freshwater mollusks in brackish waters. Acta Parasitol. Pol. 4:375^t06. (in Polish). Raabe. Z. 1966. The parasitic ciliates of Dreissena polymorpha and other bivalvia in the Ohrid Lake. Acta Protozoal. 4:1-14. Raabe. Z. 1970. Ordo Thigmotricha (Ciliata-Holotrichal. III. familiae An- cistrocomidae et Sphenophryidae. Acta Protozoal. 7:385^163. Raabe. Z. 1971. Ordo Thigmotricha (Ciliata-Holotricha). IV. familia Thig- mophryidae. Acta Protozoal. 9:121-170. Journal of Shellfish Research. Vol. 18. No. 1. 259-280. 1999. ABSTRACTS OF TECHNICAL PAPERS Presented at the 19"' Annual Meeting MILFORD AQUACULTURE SEMINAR Milford. Connecticut February 27-March 1. 1999 259 Milford Aquaculture Seminar. Milford. Connecticut Abstracts, February 27-Murch I, 1999 261 CONTENTS Walter J. Blogoslawski Overview, 1 9"' Milford Aquaculture Seminar 263 Joseph Choromanski, Sheila Stiles, Christopher Cooper, Eric Bedan, Sherry W. Lonergan and Paul ./. Trupp Growth and survival of juvenile bay scallops from genetic lines at different densities and depths: Collaborative study between the National Marine Fisheries Service and the Bridgeport Aquaculture School 263 John ./. Curtis, Sherry W. Lonergan, Thomas McGann and Paul J. Trupp The effect of density on growth of Argopecten irradians in Long Island Sound: In partnership with National Marine Fisheries Service Scientists 263 Joseph DeCrescenzo, hike Sunila, John Karolus and John Volk Histopathologic^ survey of the Quahog, Mercenaria mercenaries along the Connecticut coastline 264 Mark S. Dixon, Barry C. Smith and Gary H. Wikfors The Inverted Propeller-Beanie. ..A new way to mix large microalgal tanks 264 Richard A. French, Salvatore Frasca, Jr., Sylvian DeGuise and Hebert J. Van Kruiningen Aquatic animal health and UCONN aquaculture program: New faculty and opportunities 265 Richard A. French Vibrio parahaemolyticus and other shellfish diseases of public health significance: A review 265 David W. Grunden Wampanoag shellfish aquaculture 266 Mining Guo Superior growth as a general feature of triploid shellfish: Evidence and possible causes 266 William Hastback Vibrio parahaemolyticus — A new challenge for state shellfish control agencies 267 Porter Hoagland, Hauke L. Kite-Powell and Di Jin The economics of sea scallop grow-out: Aquaculture at an offshore site 267 Mark L. Homer, Mitchell Tarnowski and Robert Basse II The potential for bivalve aquaculture in Maryland's coastal bays 268 Marina Huber, Eric Moore, Neil Marcaccio, Robin Katersky and David Bengtson Effects of photoperiod on survival, growth and pigmentation of summer flounder (Paralichthys dentatus) larvae in laboratory culture 268 Diane Kapareiko and Richard A. Robohm A comparison of CHROMagar E. coli. Millipore coli-count samplers, and the MPN procedure for enumeration of coliforms in bay scallops 269 Richard C. Karney and John C. Blake Developments in the private aquaculture industry on Martha's Vineyard 269 John Karolus, hike Sunila, Stacey Spear, Joseph DeCrescenzo and John Volk The presence of Haplosporidium nelsoni (MSX) and Perkinsus marinus (DERMO) in Crassostrea virginica along the Connecticut and northern Long Island shoreline in 1998 — An extensive survey 270 Grace Klein-MacPhee and Aimee Keller Early induction of spawning of a captive tautog broodstock by light and photoperiod manipulation 270 Brenda Ijindau and Ximing Guo Growth characteristics in Triploid Pacific oysters — A new dimension 270 Steven Lang A social and economic evaluation of an oyster mariculture training program for Long Island commercial fishermen 271 Richard luingan The transition from commercial fishing to oyster culture: Results of a NMFS fishing industry grants project 271 Kenneth J. La Valley, Thomas L. Howell, Riley Y. Morse, Brian Beal and Bertrand Dubois Experimental testing of field techniques for farming the softshell clam (Mya arenaria) 272 Dale F. Leavitt, Patricia L. Gohring and William P. Burt A tour of upwellers on Cape Cod 272 Robert Link The need for aquaculture in the world today 272 262 Abstracts, February 27-Murch 1, 1999 Milford Aquaculture Seminar, Milford. Connecticut Michael Ludwig Recent streamlining of the aquaculture regulatory process 272 Lindsay Lydon and Grace Klein-MacPhee The effects of stocking density on growth of larval tautog 273 Gisele Magnusson and James Anderson Progress in bioeconomie evaluation of the Milford laboratory scallop nursery recirculating system 273 Harriette L. Phelps Australian/Tasmanian oyster culture 274 Steven Pitchford and Richard A. Robohm A comparison of antidumping solutions used for initial recovery of hemocytes from the bay scallop {Argopecten irradians) 274 Laurel J. Ramseyer Effect of dietary pH on the utilization of semipurified diets by tautog. Tautoga onitis 274 David R. Relyea Vibrio parahaemolyticus — A new problem for the shellfish industry in the northeast 275 Edwin Rhodes NOAA fisheries and aquaculture 275 Michael A. Rice Control of eutrophication by bivalves: Filtration of particulates and removal of nitrogen through harvest of rapidly growing stocks 275 Gregg Rivara and David A. Bengtson Summer flounder culture in the northeast: Update on recent research and industry status 276 Shawn M.C. Robinson An overview of aquaculture research in Atlantic Canada 276 Tessa L. Simlick, Robin S. Katersky, Neil Marcaccio and David A. Bengtson Post-metamorphic growth of summer flounder in laboratory culture: Do early-settling larvae grow faster than late settlers? 277 Barry C. Smith, Sara Barcia, Jennifer H. Alix and Gary H. Wikfors Fertilization rates and procedures using commercial "F/2" nutrient mixes to grow T-ISO (Isochrysis sp.) and PLY429 ( Tetraselmis chui) 277 Ron Smolowitz and Harlyn Halvorson Updating the plans for sea scallop aquaculture in Massachusetts 277 Jeff Southworth, Maronda Brown, Sheila Stiles and Linda Strausbaugh Methodology for the generation of polymorphic molecular tags in the bay scallop, Argopecten irradians 278 Inke Sitnila, John Volk, John Karolus, Terry Backer, Stan Czyzyk, Ed Lang, Matt Mroczka and Karen Rivara Disease-resistant oysters, Crassostrea virginica, in Long Island sound 278 James C. Widman, Jr. Reflections on biofilter selection for shellfish culture 279 Gary H. Wikfors, Jennifer H. Alix, Mark S. Dixon, and Barry C. Smith Feeding rations and regimes for post-set oysters, Crassostrea virginica, fed cultured microalgae in a land-based nursery 279 Steve Yankocy, Grace Klein-MacPhee and Aimee Keller Feeding studies on juvenile tautog, two experiments: Weaning juvenile tautog to an artificial diet and effects of feeding frequency on growth of juvenile Tautog 280 Milford Aquaculture Seminar. Milford, Connecticut Abstracts, February 27-March I, 1999 263 OVERVIEW, 19th MILFORD AQUACULTURE SEMINAR. Walter J. Blogoslawski. U.S. Department of Commerce. National Oceanic & Atmospheric Administration, National Marine Fisher- ies Service. Northeast Fisheries Science Center. Milford Labora- tory, 212 Rogers Avenue. Milford, CT 06460. The 19"' Milford Aquaculture Seminar attracted 41 speakers whose topics included shellfish diseases and their effects on the local industry, new techniques for bay and sea scallop culture, the culture of blackfish (Tautog) and summer flounder, biofiltration in recirculating systems, some Canadian experiences in aquaculture, the use of triploidy to enlarge crop animals, crop insurance avail- ability, new regulations for permitting, and societal aspects of aquafarms. The 150 attendees from the US and Canada met in formal and informal sessions to discuss the recent problems in the aquaculture industry and to share potential solutions. The coop- eration exhibited at the meeting highlighted the ability of persons sharing the same concerns to work together for acceptable resolu- tions of their common problems. Attendees representing 42 public and private shellfish and fin- fish-aquaculture industry ventures were joined by persons from 15 educational institutions and 10 state and federal government agen- cies to exchange ideas and experiences in developing aquaculture technology. The participation of our speakers and exhibitors is greatly ap- preciated as is the financial support from our sponsors, the U.S. Department of Commerce's National Marine Fisheries Service. Mil- ford Laboratory. Milford, CT and the U.S. Department of Agriculture. Northeastern Regional Aquaculture Center in North Dartmouth, MA. GROWTH AND SURVIVAL OF JUVENILE BAY SCAL- LOPS FROM GENETIC LINES AT DIFFERENT DENSI- TIES AND DEPTHS: COLLABORATIVE STUDY BE- TWEEN THE NATIONAL MARINE FISHERIES SERVICE AND THE BRIDGEPORT AQUACULTURE SCHOOL. Jo- seph Choromanski, Sheila Stiles, Christopher Cooper, and Eric Bedan, USDOC, NOAA, National Marine Fisheries Service, Northeast Fisheries Science Center, Milford Laboratory, Milford, CT 06460; Sherry W. Lonergan and Paul J. Trupp, Bridgeport Regional Vocational Aquaculture School. 60 St. Stephens Road, Bridgeport. CT 06605. Hatchery-reared juvenile scallops were field-tested to evaluate growth and survival of genetic lines of bay scallops (Argopecten irradians) in western Long Island Sound. Students and staff from the Bridgeport Regional Vocational Aquaculture School cooper- ated with scientists at the Milford National Marine Fisheries Ser- vice Laboratory in these collaborative grow-out studies. In May 1998, Milford scientists provided a total of five thousand scallops from different genetic lines for the students to use in their project. The lines used were large- and small-selected. The scallops were counted, measured, and their total volume determined. They then were divided into groups containing varying numbers of scallops to determine effects of density. Chinese lantern nets and Japanese lantern and pearl nets were deployed off Bridgeport/Fairfield near Penfield Reef in Long Island Sound. Nets were suspended from a floating longline to a depth of approximately three meters. A sec- ond set of scallops was counted and measured, then placed in lantern nets anchored to the bottom (ten meters at high tide) and buoyed up into the water column to a height of two meters, to serve as comparisons for nets suspended from longlines. Counts and measurements were made approximately midway through the experiment. Nets were heavily fouled with seaweed and tunicates. However, overall survival was high. Growth was not exceptional, which was presumed attributable to rather heavy cur- rent and wave action. A severe storm, which occurred in the area in July, loosened an anchor mooring and resulted in the loss of some of the nets. A videotape and visual inspection was made by NMFS divers to assess conditions and damage which had been caused by the storm. The final phase of this particular experiment ended in November when the nets were retrieved and evaluated. One longline net from the comparative depth study was lost, while both bottom nets were intact. Initial observations showed slightly better growth from the bottom nets, while survival was comparable. This longline project was a unique opportunity for several rea- sons. While the final results of this first year may be preliminary, invaluable logistical experience in longline aquaculture was at- tained for the staff of the lab and the school; this will be used in planning future projects. The mutually beneficial experiment pro- vided both valuable assistance for the NMFS scientists, and a learning experience for the students to become familiar with tools and protocols used in measuring the scallops in particular, as well as "hands-on" exposure to aquaculture in general. THE EFFECT OF DENSITY ON GROWTH OF AR- GOPECTEN IRRADIANS IN LONG ISLAND SOUND: IN PARTNERSHIP WITH NATIONAL MARINE FISHERIES SERVICE SCIENTISTS. John J. Curtis, Sherry W. Lonergan, Thomas MeGann, and Paul J. Trupp. Bridgeport Regional Vo- cational Aquaculture School. 60 Saint Stephens Road. Bridgeport CT 06605. Being consistent with its philosophy of infusing meaningful activities into the instruction at the Bridgeport Regional Voca- tional Aquaculture School, an invitation was accepted to have students work with National Marine Fisheries Service scientists of the Milford. Connecticut laboratory on a project to study "the effect of density on growth of Argopecten irradians in Long Island Sound." The initial study was recently concluded in part, in De- cember 1998 with the harvest of the targeted crop. The school's role was clearly defined with educational goals and objectives established at the onset of the project for the in- volved students. The species, Argopecten irradians. is one familiar to the students at the Aquaculture School since its introduction in a 1994-96 international collaboration with the People's Republic 264 Abstracts, February 27-March I. 1999 Milford Aquaculture Seminar, Milford. Connecticut of China. In that project the students and staff of our school learned not only the biology of the bay scallop but also the methods of spawning, grow-out and harvesting. In addition, procedures for statistical analysis of the collected scientific data were included for follow-up studies. Students from the school's Intensive program were introduced to the project by NMFS scientists in the spring of 1998. They began work on design, construction and deployment of a long-line at the school's farm in Long Island Sound. The project began with NMFS scientists, students and staff transporting juvenile bay scal- lops, reared at the Milford Laboratory, to our test location. Scal- lops were sorted into test groups, measured and transferred to various style culture nets and attached to the long line. In the fall, students and scientists collected data on the growth rate (shell height and width I of the scallops which were then transferred to nets of a larger mesh size and returned to the water. The final phase of this project had two objectives. The first was to gather growth data on one group of test scallops and the second was to implement a separate long-line for another group of scallops for research on the effects of over-wintering. The educational objectives of this project were many and var- ied. By immersing students in real-life scientific study, they were presented the procedures necessary to assist in the design and implementation of a high-level research project from beginning to end. Discussion of scientific methods, proper research techniques, data collection and analysis augmented the standard curriculum of science and technology. This project has offered our students op- portunities to develop the skills of problem-solving in a meaning- ful activity that has already translated into higher academic per- formance and a better scientific understanding. From an educational perspective, much has been learned to date and much more can be extracted from this project through con- tinuation. The difficulties encountered in the initial attempt will be addressed through earlier phases of conditioning, spawning and placement at the farm site. The problems to gear presented by natural conditions are being addressed in the CAD classroom with students redesigning lantern nets and researching better methods of deployment. The students and staff of the Bridgeport Aquaculture School look optimistically to our continued involvement and the accomplishment of the prime goal of the project: to develop better methods to grow bay scallops which can be competitively sold in the market place. HISTOPATHOLOGICAL SURVEY OF THE QUAHOG, MERCENAR1A MERCENARIA, ALONG THE CONNECTI- CUT COASTLINE. Joseph DeCrescenzo, Inke Sunila, John Karolus and John Volk. State of Connecticut, Department of Agriculture, Bureau of Aquaculture, P.O. Box 97, Milford, CT 06460. A histopathological survey was conducted along the Connecti- cut coast line on the hard clam, Mercenaria mercenaria. Quahog Parasite Unknown (QPX). an economically important parasite, phylum Labyrinthomorpha, has been found off the coast of Mas- sachusetts. The purpose of this survey was to detect QPX or other conditions which might possess a threat to Connecticut's hard clam harvest. Eleven different locations were sampled along the Connecticut coast line. Samples of 30 clams each were harvested from either commercial or wild clam beds. A gross pathologic examination was then conducted before they were processed for histopathologic examination. The clams were shucked and placed into Davidson's fixative. Sections were then stained in hematoxylin-eosin. Samples were diagnosed for infectious agents such as viruses. Chlamydia, bacteria, fungi or any protozoan or metazoan parasite. Histopathological lesions were classified as inflammations, degen- erative process, cell or tissue death, or proliferative responses. The results showed no signs of the commercially important parasite QPX. However, some infectious and non-infectious agents were found in the examination. The following conditions appeared at low prevalences: Chlamydiales, ceroidosis, ciliates in the gill re- gion, sloughing of the epithelia in the digestive diverticula, hem- orrhage in the intestine and stomach, and mucus around the foot. In conclusion, no economically important parasites were present in the samples. Recent mortalities in oyster beds due to infection with MSX have shifted more harvesting pressure toward hard clams. This study, based on the low prevalences of histopath- ological conditions and active gametogenesis in the gonads, pre- dicts a positive future for Connecticut's clamming industry. THE INVERTED PROPELLER-BEANIE...A NEW WAY TO MIX LARGE MICROALGAL TANKS. Mark S. Dixon, Barry C. Smith, and Gary H. Wikfors, USDOC, NOAA, National Ma- rine Fisheries Service, Northeast Fisheries Science Center, Milford Laboratory, Milford. CT 06460. Large-scale open microalgal cultures can be viewed as biore- actors; they are defined as "stirred reactors" by chemical engi- neers. Energy must be supplied to the system, nutrients added at a rate equal to their use. wastes removed as they are generated, and cells must be exposed to the reactive surface (the culture/air in- terface). The Greenhouse for Research on Algal Mass Production Systems (GRAMPS) at the National Marine Fisheries Service Laboratory in Milford, Connecticut houses two 20,000 liter oval (5.5m x 3m x 1.2m) fiberglass tanks used for the production of large volumes of microalgae to feed post-set shellfish in a land- based nursery. Three stirring methods were considered for GRAMPS production tanks: 1) air mixing, 2) manual stirring with paddles, 3) mechanical stirring with a motorized device. Air mixing was rejected based upon previous experience showing that bubbles in dense, open microalgal suspensions encourage bacterial growth. The benefit of constant mixing by mechanical means needed to be established before investment in equipment Milford Aquaculture Seminar, Milford, Connecticut Abstracts, February 27-March I, 1999 265 and its operation could be justified. Paddle-wheel mixers have been used in other large-scale microalgal production systems, but all are custom made and too expensive for routine application. In many industrial processes, tanks of reactants are stirred by a foil on a shaft, driven by a motor-essentially an inverted propeller beanie. The recent addition of a propeller style mechanical mixer to one of the large tanks provided an opportunity for comparison with previous tank cultures which required regular manual mixing via a muscle-powered paddle. The reality of occasional hand mixing is a poorly-mixed culture which can only be maintained at relatively shallow depths. Reduced culture volumes, settling cells, and un- equal exposure of cells to light all reduce the productivity of a culture. By contrast, a tank culture which is mixed continually allows all cells to be on the lighted, reactive surface on a calculated regular interval controlled by varying the speed of the propeller, thereby providing all cells access to light for photosynthesis and the culture/air interface for gas exchange many times during the day. A culture of Tetraselmis chui (PLY429) has been maintained in the mechanically-mixed tank for the past 6 months using semi- continuous management. The mechanically-mixed tank was oper- ated at a maximum depth of I.I meters or approximately 18,000 liters. Maximum operation depths for the hand-mixed tanks were less than 0.4 meters or approximately 6.000 liters. Cell densities in the mechanically-mixed tank often exceed 1 x \0h cells/ml. while hand-mixed tanks average approximately I x IO5 cells/ml. Me- chanically-mixed tanks are much longer lived than hand-mixed tanks under a semi-continuous harvest strategy; 6 months versus 6 weeks. Higher productivity, greater culture volumes, a superior algal product, long lived cultures, and reduced maintenance are all ben- efits of mechanically-mixed, large-scale microalgal cultures. Eco- nomic analysis and evaluation of performance of the mixer as it is integrated into the overall automation of GRAMPS are planned for later this year. AQUATIC ANIMAL HEALTH AND UCONN AQUACUL- TURE PROGRAM: NEW FACULTY AND OPPORTUNI- TIES. Richard A. French, Salvatore Frasca, Jr., Sylvain De Guise, and Herbert J. Van Kruiningen, University of Connecti- cut, Department of Pathobiology, Northeastern Research Center for Wildlife Diseases. 61 North Eagleville Road. Storrs. CT 06269. The State of Connecticut has made a substantial investment in aquaculture industries and teaching aquaculture science. In April 1989, the Connecticut State Board of Education approved a request by the Bridgeport Board of Education to create a Regional Voca- tional Aquaculture School. The state appropriated 7.5 million dol- lars for the construction of the first of these schools, the Bridgeport Regional Vocational Aquaculture School. In 1994, the State Board of Education established the Sound School Reaional Vocational Aquaculture Center in New Haven. Both institutions provide high school students with specialized laboratories and classrooms that complement a marine science-related curriculum. Facilities at these centers include pathology laboratories, aquaculture tanks to grow finfish and shellfish, indoor boat shops, marine engine labo- ratories, and computer-assisted drafting laboratories. The University of Connecticut has made a commitment to de- velop an undergraduate teaching program in aquaculture science, which is embodied in the formulation of a multidisciplinary aqua- culture minor, including courses from the College of Agriculture and Natural Resources and the College of Liberal Arts and Sci- ence. The contribution of the Department of Pathobiology to this multidisciplinary aquaculture minor will be didactic and active teaching in the field of aquatic animal health (e.g.. preventive medicine, disease recognition and treatment, mechanisms of dis- ease, health surveillance). To this end. the University has hired three new veterinary research faculty in the Department of Patho- biology; Dr. Sylvain De Guise. Dr. Salvatore Frasca. and Dr. Rich- ard French, and added an undergraduate course entitled, "Systemic Finfish and Shellfish Pathology and Microbiology," to be offered in Spring Semester 2000. The aquaculture science program is affiliated with regional aquaria, fisheries, and professionals (Mystic Aquarium, The Mari- time Aquarium, The Connecticut Department of Agriculture, Bu- reau of Aquaculture and Laboratory, and private industry), pro- viding active instruction and cooperative training opportunities to students. In addition, a state-of-the-art Marine Science and Tech- nology Center facility is under construction at the University of Connecticut Avery Point Campus, which will offer unique educa- tional prospects to undergraduate and graduate students. The Con- necticut Veterinary Diagnostic Laboratory, which provides au- topsy service for state and private concerns, will expand and sup- port the accession of numerous aquatic animal cases directed toward a primary teaching, diagnostics and research initiative. The Department of Pathobiology offers a Bachelor of Science under- graduate degree, and Master of Science and Doctor of Philosophy graduate degrees. Graduate students in Pathobiology may special- ize in Pathology, Microbiology, Virology, Immunology, Clinical Chemistry, Avian and Aquatic Animal Pathology, and Wildlife Diseases. Regarding aquatic animal health, ongoing research in- cludes studies of marine mammals, marine and freshwater finfish. and shellfish. VIBRIO PARAHAEMOLYTICUS AND OTHER SHELLFISH DISEASES OF PUBLIC HEALTH SIGNIFICANCE: A RE- VIEW. Richard A. French, University of Connecticut, Depart- ment of Pathobiology, Northeastern Research Center for Wildlife Diseases, 61 North Eagleville Road, Storrs, CT 06269. The incidence of foodbome illness associated with consump- tion of contaminated seafood products has recently triggered me- dia attention that has helped to increase public awareness of issues 266 Abstracts, Febmary 27-March 1. 1999 Milford Aquaculture Seminar. Milford. Connecticut related to food safety. This media coverage has also generated a number of misconceptions regarding the safety of eating seafood. Microorganisms and other toxic substances commonly ingested by shellfish may accumulate within animal tissues and be passively transmitted to humans when they consume the tainted seafood products. Though generally relatively harmless to the affected shellfish, these microorganisms and chemicals are often patho- genic or toxic to humans. Pathogens of public health significance associated with contaminated seafood include causative agents of hepatitis and gastroenteritis, biotoxins (paralytic shellfish poison- ing) and toxic industrial chemicals (heavy metals, polycyclic aro- matic hydrocarbons, and chlorinated hydrocarbons). One impor- tant microbial pathogen of marine species, including crabs, shrimp, lobster, and oysters is Vibrio parahaemolyticus. Recent foodborne disease outbreaks associated with consumption of oysters in the Pacific Northwest (1997), Galveston Bay. Texas (1998) and Oys- ter Bay. New York ( 1998). have heightened awareness of V. para- haemolyticus. This Vibrio species is a halophilic bacterium that is part of the normal flora of estuarine and other coastal areas world- wide. Human illness associated with V. parahaemolyticus is char- acterized by a self-limiting, mild to moderate gastroenteritis oc- curring within 4-96 hours after consumption of raw or improperly cooked, and/or stored fish and shellfish. Several halophilic Vibrio species associated with mollusks are reported to cause gastroen- teritis in humans. Disease is strain-specific within Vibrio species and correlated with production of various toxins, including entero- toxins, cytotoxins, and hemolysins. In addition to surveillance ef- forts designed to identify the pathogenic strains of V. para- haemolyticus, epidemiologic and pathogenesis studies are cur- rently underway. Such research will help determine the geographic distribution of V. parahaemolyticus and provide a better under- standing of the mechanisms of the disease process. Diagnostic methods for the detection of V. parahaemolyticus and species typ- ing continue to improve. A review of shellfish-associated food- borne diseases and current efforts to improve food safety in the United States will be addressed. WAMPANOAG SHELLFISH AQUACULTURE. David W. Grunden. Wampanoag Aquaculture Director. Island Aquaculture. Oak Bluffs. MA 02557. The Wampanoag Tribe of Gay Head Aquinnah is a Federally Acknowledged Native American Tribe located on Martha's Vine- yard Island. Their tribal offices are in the town of Aquinnah. MA. They have observed the decline of the local shellfishery in the town and have two goals for their aquaculture enterprise. The first, of course, is to make a profit. The second is to return a percentage of the yield from the hatchery back to the wild and to protect their cultural heritage of depending on the local shellfish as a food. They also hope that it will allow the fishery to recover so that many of their tribal members can continue to fish commercially within the local ponds. In working to achieve this second goal the tribe has a Memorandum of Understanding with the Town to assist their shellfish department with any of its propagation and predator con- trol programs. This has evolved into assisting the Town's shellfish department in developing a comprehensive plan to manage the shellstock. A report on the early development of the Native American Wampanoag Tribe of Gay Head Aquinnah' s commercial shellfish aquaculture enterprise will be presented. It will include an intro- duction to where the project is located, what has been done to date and the expected developments for 1999. A hatchery is planned as well as grow-out of the seed to both field plant and market sizes. Additionally, the Tribe is working with the local Town to develop a comprehensive shellfish management plan. SUPERIOR GROWTH AS A GENERAL FEATURE OF TRIPLOID SHELLFISH: EVIDENCE AND POSSIBLE CAUSES. Ximing Guo, Rutgers University. Haskin Shellfish Re- search Laboratory, 6959 Miller Avenue, Port Norris. NJ 08349, USA. Triploids are organisms with three sets of chromosomes instead of the two sets found in normal diploids. Aquacultural interest in triploid shellfish so far has primarily focused on their sterility. The presence of an extra set of chromosomes poses a problem for meiosis and leads to complete or partial sterility in triploids. Be- cause excessive gonadal development negatively affects meat quality of diploid molluscs, sterile triploids provide a high quality product that can be marketed year round. Triploid Pacific oyster is now widely used for aquaculture production. However, another important benefit of triploid molluscs, superior growth, has been largely overlooked by early studies and aquaculturists. During the past decade, triploids have been studied in over 20 species of molluscs. A review of recent data indicates that superior growth may be a general feature of triploid molluscs. Triploids exhibit significantly higher growth rate than diploids in almost all species studied so far. Triploids grow faster than diploids by 12-30% in Crassostrea virginica, 25-51% in Crassostrea gigas, 42-52% in Crassostrea dalienwhanensis, 127c in Mulinia lateralis, 27-58% in Pinctada martensii, 36% in Argopecten irradians, 32-59% in Chlamys nobilis, and 81% in Chlamys farreri. The adductor muscle of triploid scallops is larger than that of diploids, by 73% in A. irradians, 96% in C. farreri. and 167% in Argopecten ven- tricosus. The expression of the triploid advantage in growth may be influenced by genetic and environmental factors. Triploids may not show superior growth in food-limiting environments. Several hypotheses have been proposed to account for the superior growth in triploids. One hypothesis attributes the superior growth to in- creased heterozygosity in triploids. A positive correlation between heterozygosity and growth rate has been found in diploid molluscs. Triploids are theoretically more heterozygous than diploids. The heterozygosity hypothesis is supported by the observation that trip- Milford Aquaculture Seminar. Milford. Connecticut Abstracts, February 27-March 1. 1999 267 loids produced from blocking polar body I and diploid x tetraploid mating, which are more heterozygous, grow faster than triploids produced from blocking polar body II. Another hypothesis views that sterility in triploids distributes more energy to growth rather than sexual reproduction. The energy relocation hypothesis cannot explain growth difference expressed before sexual maturation. Fi- nally, it has also been suggested that triploid cells are larger than diploid cells and may contribute to an overall increase in body size. All these factors may contribute somewhat to the overall growth of triploids. Regardless of causes, triploid molluscs may benefit aqua- culture by offering greatly improved growth. The challenge is that commercial production of triploids is technically difficult in most species. Commercial use of triploids may ultimately depend on the development of tetraploids. which can produce 100% pure trip- loids simply by mating with normal diploids. Tetraploids have been successfully developed for triploid production in the Pacific oyster, and success in other species may soon follow. VIBRIO PARAHEMOLYTICUS— A NEW CHALLENGE FOR STATE SHELLFISH CONTROL AGENCIES. William Hastback. New York State Department of Environmental Con- servation, 203 North Belle Mead Road. Suite 1. East Setauket, New York 11733. In late August 1998. the New York State Department of Envi- ronmental Conservation (NYSDEC) shellfish sanitation pro- gram was advised by the New York State Department of Health (NYSDOH) of a series of five (5) individual illnesses in shellfish consumers. Laboratory analyses of patient samples indicated that the illnesses were caused by the naturally occurring marine bac- terium— Vibrio parahaemolyticus. (VP). The initial information available indicated that the shellfish implicated in the illnesses had been harvested from the area des- ignated as NS-2, including Oyster Bay and Cold Spring Harbors, in northwestern Nassau County. On September 8. we learned that two individuals in New Jersey had become ill after consuming oysters from the same area. On September 9. the NYSDOH ad- vised NYSDEC of their determination of a statistical association between the consumption of shellfish and the illnesses. On Sep- tember 10. the NYSDEC Bureau of Marine Resources designated Oyster Bay and Cold Spring Harbors as uncertified for the harvest of shellfish on an emergency basis. The closure was in effect through October 22, a period of six weeks. The decision to reopen was based on declining water temperatures and the results of DNA probe examinations of oyster samples conducted by two FDA laboratories. In the interim, the federal Centers for Disease Control identi- fied the 03:K6 strain of VP isolated from patient samples. That strain of VP had been identified as the cause of an oyster-related illness outbreak that affected approximately 450 people in several states during June. Galveston Bay, Texas was the source of the oysters in that outbreak. The 03:K6 strain has also been respon- sible for large seafood related illness outbreaks in southeast Asia. from India to Japan. THE ECONOMICS OF SEA SCALLOP GROW-OUT: AQUACULTURE AT AN OFFSHORE SITE. Porter Hoag- land, Hauke L. Kite-Powell, and Di Jin. Marine Policy Center. Woods Hole Oceanographic Institution, Woods Hole, MA 02543. The extent to which offshore sea scallop aquaculture is a com- mercially viable business depends upon both the costs of growing scallops relative to wild harvest operations and conditions in the relevant product market. Here we report on the development of a discounted cash flow model of the grow-out of sea scallops at an offshore farm, such as that represented by the Westport Fishing Corporation's sea scallop experiment off the coast of Martha's Vineyard, Massachusetts. We examine the economic viability of four alternative approaches to scallop farming: seabed seeding and three variations on cage cul- ture: lantern cages: bottom cage trawls; and bottom cage clusters. For each alternative, we estimate capital and operating costs and revenues over a 20 year period. We assume a two-year cycle from collection of juveniles to harvest, and scale the farming operation in every case to produce 100 thousand pounds of scallop meat per two-year cycle (that is, every other year). Under baseline assumptions, the only alternative that is profit- able is seabed seeding. A 100 thousand lbs/cycle seabed seeding operation requires less than $400 thousand in start-up capital and pays back the initial investment in four years. It requires a lease area of about 150 acres and requires the use of a large scallop vessel about 3 months out of the year, on average. The cage op- erations are not profitable because the higher survival rate and growth are not enough to justify the added cost of buying, main- taining, deploying, and harvesting the cages and associated moor- ings. Although they require smaller lease areas, the cage opera- tions demand between $1-2 million in startup funding. Of the three alternatives, bottom cage trawls come closest to break even be- cause gear costs are relatively modest. There are several sources of uncertainty in the model, including the ex-vessel price for sea scallops. In order to help manage this uncertainty, we have estimated a model of supply and demand for New England sea scallops using monthly data during the period 1985-93. The model is a linear representation of both supply and demand for "average size" sea scallops, implying a market equi- librium over the 1985-93 period of $5.42 per pound. It is useful to think of the production of scallops from an offshore farm as an inventory problem. At an offshore site, seed scallops grow over a period of about two years to a size that may command a premium over the average size scallop. We have de- veloped a simple algorithm to help the farmer take advantage of historical monthly variability in sea scallop demand. If this vari- ability persists, we find that when farm output is small relative to the market, the farmer should act as a price taker, harvesting and 268 Abstracts, February 27-March 1. 1999 Milford Aquaculture Seminar, Milford. Connecticut marketing his product only in January. As potential output in- creases, however, the time profile of output shifts. Output of up to 150 thousand pounds should be produced in January and Novem- ber. When output reaches 200 thousand pounds, there should be some level of production in every month except July. It is costly to monitor sea scallop mortality at an offshore site. Because of uncertainty about mortality, the time profile of pro- duction is suggestive of a strategy for harvesting the aquaculture product. It may be sensible to sample the product through partial harvesting, say. in October. This sample will give the fanner an estimate of mortality. If mortality is low, then a production profile that places product on the market in every month might be fol- lowed. If mortality is high, then production should be adjusted accordingly, and product would be placed on the market in No- vember or January. Note also that the production profile can be readjusted during the year as market conditions become revealed and as uncertainty about the quality of the farmed product is re- duced. This research has been supported with funds from the West- port Fishing Corporation and the National Sea Grant College Pro- gram. land DNR re-introduction project, growth and survivorship data are now available. Preliminary results indicate that growth rates may not be sufficient to produce marketable scallops before their second winter. Given the short life-span of this species and the labor involved in battling fouling organisms, bay scallop culture in Maryland has some serious problems to overcome. Although Virginia has established a significant hard calm aqua- culture industry, including production in Chincoteague Bay, few attempts have been made within Maryland"s boundaries. The main impediment to hard clam culture appears to be associated with the permitting process, which includes three state agencies, five fed- eral agencies, public hearings, and. on occasion, an appeals board. This daunting array of agencies, associated regulations, and oppo- sition from waterfront property owners has attracted few individu- als to the process. Environmental conditions in the Maryland coastal bays, however, appear to be sufficient to establish at least a modest hard clam aquaculture industry. There are areas outside Federal jurisdiction that provide clean, hard bottom for either planting bags or netting small beds of seed clams. Hatchery-reared clams are readily available from Virginia and there is a suitable, nearby market for hard clams. THE POTENTIAL FOR BIVALVE AQUACULTURE IN MARYLAND'S COASTAL BAYS. Mark L. Homer. Mitchell Tarnowski, and Robert Bussell, Maryland Department of Natural Resources. Tawes State Office Building. B-2. Annapolis. Mary- land 21401. It has been over a century since the coastal bays of Maryland supported a substantial public oyster fishery, nearly 30 years since hard clam catches peaked and then collapsed, and some 70 years since bay scallops even inhabited this region. Although a success- ful relay industry on private grounds was established for oysters after natural populations almost disappeared, it essentially ended about 50 years ago. Hard clams currently support only a remnant fishery, while, until this year, bay scallops had not been seen in the wild since the early 1930's. There have been sporadic, and ultimately unsuccessful, at- tempts to culture oysters, hard clams, and bay scallops in the Maryland portion of Chincoteague Bay during the past seven de- cades. Oysters are particularly subjected to a hostile environment, related to environmental changes caused by the stabilization of the Ocean City Inlet in 1933. Three oyster parasites. Dermo, MSX. and SSO are active in the coastal bays, as are two species of highly abundant oyster drills. Any hard substrate placed into Chinco- teague Bay is rapidly colonized by a variety of fouling organisms, including serpulid worms, colonial tunicates, hydrozoans, bryozo- ans, and barnacles. These factors tend to diminish the possibility of successful oyster aquaculture ventures in this region, with the pos- sible exception of a rapid turnaround relay fishery. Bay scallop culture has only recently been attempted in Mary- land, although initial results are not encouraging. Through a Mary- EFFECTS OF PHOTOPERIOD ON SURVIVAL. GROWTH AND PIGMENTATION OF SUMMER FLOUNDER (PARALICHTHYS DENTATUS) LARVAE IN LABORA- TORY CULTURE. Marina Huber. Eric Moore. Neil Marcac- cio, Robin Katersky, and David Bengtson. Department of Fish- eries, Animal and Veterinary Science, University of Rhode Island, Kingston. RI 02881. Summer flounder represents a promising species for commer- cial aquaculture in the northeastern United States. In order to op- timize production, the effects of various environmental parameters on biological production parameters must be studied. We investi- gated the effects of photoperiod on three parameters important to hatchery production: survival, growth and abnormal pigmentation. The last parameter involves incomplete pigmentation of the eyed side, including minor non-pigmented blotches to complete albi- nism. Flounder larvae were reared in replicate 75-L aquaria under three light regimes. 24L:0D (constant light). 16L:8D (summer con- ditions), 8L:4D:8L:4D (abnormal conditions to trick the fish into physiologically living two "days" in one). No significant differ- ences in survival or growth were detected in the larvae through metamorphosis; however, after metamorphosis, fish reared in con- stant light had significantly lower levels of abnormal pigmentation. The experiment was repeated with an additional treatment, 8L: 16D (winter conditions); no significant differences in pigmentation were observed among treatments, but fish in the 8:4:8:4 treatment grew significantly more. Milford Aquaculture Seminar. Milford, Connecticut Abstracts, February 27-March 1. 1499 269 A COMPARISON OF CHROMAGAR E. COLI, MILLI- PORE COLI-COUNT SAMPLERS, AND THE MPN PRO- CEDURE FOR ENUMERATION OF COLIFORMS IN BAY SCALLOPS. Diane Kapareiko and Richard A. Robohm. USDOC. NOAA. National Marine Fisheries Service. Northeast Fisheries Science Center. Milford Laboratory. Milford, CT 06460. Little information exists on whether bay scallops are capable of harboring microorganisms of human health significance. To assure a safe product resulting from our bay scallop aquaculture devel- opment activities, we are interested in surveying bay scallops for the presence of fecal coliforms (human health indicator organ- isms). The recommended procedure for coliform detection, the MPN procedure, requires a relatively large expenditure of time, labor and materials. Two recently-developed products. CHROMagar E. coli ,™ and Millipore Coli-Count Samplers.™ ap- pear to be simpler to apply for detection of coliform bacteria. CHROMagar E. coli, a chromogenic plate medium, expedites the identification of Escherichia coli on the basis of contrasting colony colors. Millipore Coli-Count Samplers combine a grid- marked. 0.45 um Millipore membrane filter over a nutrient pad. which facilitates growth. Microorganisms present in the sample being tested affix to the filter pad and are cultured within its plastic- case. Both products contain a chromogenic substance which reacts with genus or species-specific enzymes to produce a blue colony color for fecal coliforms. Other gram-negative bacteria remain colorless. We tested both of these products for their accuracy in enumer- ating E. coli in comparison with the traditional MPN method. Aliquots of bay scallop homogenate were "seeded" in blender jars with three individual doses of a pure culture of E. coli; this con- sisted of a low dose (mean colony count = 1.8667 x 10~/ml), a medium dose (mean colony count = 1 .8667 x 104/ml), and a high dose (mean colony count = 1.8667 x 105/ml). A fourth aliquot of scallop homogenate was not seeded in order to provide colony count information on possible pre-existing levels of E. coli. The number of organisms recovered from each dose was enumerated using the MPN procedure. CHROMagar E. coli and Millipore Coli-Count Samplers simultaneously; the experiment was repeated for a total of three trials. Individual counts were expressed as the ratio of the number of E. coli recovered (minus the background count of E. coli) divided by the dose administered (colonies/ml). Ratios resulting from CHROMagar E. coli and Millipore Coli- Count Sampler counts for each dose were normally distributed. Analysis by two-sample t-test indicated no significant differences between the means of these ratios for each technique at the p = 0.05 level of significance (p = 0.7564 for low dose, p = 0.9542 for medium dose, p = 0.2540 for high dose). Overall ratios (re- gardless of dose) for the MPN, CHROMagar and Millipore procedures were also normally distributed and no significant dif- ferences existed when MPN ratios were compared either to CHROMagar E. coli ratios (p = 0.0957) or to Millipore ratios (p = 0.0959). A practical evaluation indicated that CHROMagar E. coli had a longer refrigerated shelf-life for the powder concen- trate, was more cost-effective than the Millipore Coli-Count Sam- plers ($0.80/test for CHROMagar. $2.50/test for Millipore). and always produced consistent results. ™Trademarks are used to identify products only and are not indicative of endorsement by the USDOC. NMFS. DEVELOPMENTS IN THE PRIVATE AQUACULTURE IN- DUSTRY ON MARTHA'S VINEYARD. Richard C. Karney. Martha's Vineyard Shellfish Group. Inc.. Box 1552, Oak Bluffs, MA 02557; and John C. Blake. Sweet Neck Farm. Box 1468, Edgartown, MA 02539. The progress of private aquaculture ventures spawned by two National Marine Fisheries Service (NMFS) Fishing Industry Grants (FIG) is reported. Five growers have successfully taken thousands of 2 mm oyster seed to a 3 inch legal size in about three years. The single oysters grown in cages off-the-bottom are well shaped, deeply cupped and as good or better quality than most cocktail oysters seen in raw bars. The growers consistently re- ceived fifty to sixty-five cents apiece for their oysters. Despite the loss of up to 25 percent of the three year old oysters due to Seaside Organism (SSO. Haplosporidium costale) in the spring of 1998. production of market size oysters per grower for the past year ranged between 800- 1 1 ,000 oysters with an average of about 3.000 oysters per grower. Over a half million new seed oysters were cultured by twelve growers in 1998. Growth and survival have been excellent. Grow- ers who have been conscientious about thinning and cleaning the seed, report that 2 mm seed set out in July was averaging 2 inches in September. Seed cultured in a tidal upwelling nursery grew to 2 inches (from 2 mm) in eight weeks! Some of these oysters reached 3 inches by December. Island cultured shellfish were promoted under a $4,000 grant from the Southeastern Massachusetts Aquaculture Center (SEMAC). The grant provided for the development of a logo, printing of promotional materials, and introducing the new cul- tured seafood products to Island chefs, retailers, and the general public. Under a grant from the Massachusetts Department of Food and Agriculture, Jack Blake, an Edgartown grower, constructed and operated a floating hatchery/nursery prototype. The first two at- tempts at larval culture of quahogs failed. During the first attempt, fertilized eggs introduced to the flow-through larval culture system escaped when a drain screen dislodged. A second attempt was made to culture quahog larvae. This time the larvae were cultured in a closed system tank where water was changed every other day and cultured phytoplankton was fed daily. This culture succumbed to a Vibrio infection traced to source water which was drawn from a prefilter reservoir contaminated with oyster feces from an adja- 270 Abstracts, February 27-March I. 1999 Milford Aquaculture Seminar. Milford. Connecticut cent nursery culture system. In a third attempt, two million two- week-old oyster larvae introduced into the system were success- fully grown in a closed system mode. Within a week, the veligers progressed to eyed larvae and were set on microcultch in the system. This culture attempt has resulted in over 1 10,000 of 2-5 mm oyster seed. Results from these early trials are promising and the innovative hatchery is scheduled to be run again next year. THE PRESENCE OF HAPLOSPOR1DWM NELSON1 (MSX) AND PERKINSUS MARINUS (DERMOt IN CRASSOSTREA VIRGimCA ALONG THE CONNECTICUT AND NORTH- ERN LONG ISLAND SHORELINE IN 1998— AN EXTEN- SIVE SURVEY. John Karolus, Inke Sunila, Stacey Spear, Jo- seph DeCrescenzo, and John Volk. Connecticut Department of Agriculture, Bureau of Aquaculture. P. O. Box 97, Milford, CT 06460. Previous data generated by this laboratory determined a wide- spread prevalence of Perkinsus marinus (Dermo) starting in 1996 and Haplosporidium nelsoni (MSX) starting in August, 1997 in the Crassostrea virginica (eastern oyster) population along the Con- necticut coast. An extensive survey was conducted in 1998 to include the entire coast line of Connecticut and the northern shore of Long Island. Samples of 30 oysters each were collected at selected sites representing both leased oyster growing areas, seed areas and lo- cations of special interest. For the diagnosis of MSX. oyster tissue was preserved in Davidson's fixative with 20% artificial seawater. Paraffin-sections were stained with hematoxylin — eosin and Ziehl's acid fast stain for detecting spores. For the diagnosis of Dermo, anal rectal tissue were cultured in Ray's Fluid Thiogly- collate Medium. Haplosporidium nelsoni was found at epizootic levels at most sites along the Connecticut shoreline in 1998. Ninety one percent of the Connecticut samples were found positive for MSX. Preva- lence varied from 3 to 77%. Adult oysters at three locations were found to contain sporulating MSX during the autumn. Forty-four percent of the New York samples were positive. The prevalence range was from 3 to 37%. For Connecticut sampling sites, the results for Perkinsus mari- nus in 1998 indicated no significant difference from the intensity of infection between the shallow waters (eight feet or less) and the deeper water samples. Perkinsus marinus was found in 100% of the Connecticut samples. In addition, there was no significant dif- ference between the intensity of infection for 1997 versus 1998. In New York, 69% of the samples were positive for Dermo. How- ever, the Dermo intensity of infection was not significantly differ- ent from that found in Connecticut during 1998. MSX results for 1998 indicated most of the CT shoreline was experiencing a post-epizootic period. It appeared that Dermo had established an enzootic prevalence in Long Island Sound. EARLY INDUCTION OF SPAWNING OF A CAPTIVE TAUTOG BROODSTOCK BY LIGHT AND PHOTOPE- RIOD MANIPULATION. Grace Klein-MacPhee and Aimee Keller. University of Rhode Island. Graduate School of Oceanog- raphy. Narragansett Bay Campus, Narragansett, RI 02882. Broodstock collected by hook and line from the east passage of Narragansett Bay. RI was maintained in the laboratory in an 8 foot, black fiberglass tank with running seawater, an airstone. and sev- eral PVC tubes for shelter. The tank was tented with black plastic, and a fluorescent light set on a timer maintained a photoperiod of 10 hours light and 14 hours dark (approximate ambient photo- period for winter). Seawater was at ambient temperature and sa- linity from November 25, 1997 to March 17. 1998. The fish were fed chopped quahogs, live crabs and whole mussels daily through- out December then every other day through March 17. It was a mild winter and water temperatures averaged 7.4°C (range 5.5-9) for November-December and 6°C (range 3.8-7.8) for January- March. During this time the fish were relatively inactive, spending most of the time in their shelters with an occasional sortie around the tanks to feed. On March 17, the water temperature was raised to 14°C and maintained at an average temperature of 13.4°C (range 10.7-16.6) through March 3 1 . The photoperiod was changed over a period of a week from 10L / 14D to 13L / 1 ID. On March 31 the fish began to spawn. The first large batch of eggs was collected on April 2 and the progeny from this spawning were raised through juvenile stage. The juveniles are now 10 months old and are healthy and active fish. The broodstock continued to spawn with periodic resting states through November 1998. Tautog spawn in Narragansett Bay from May through August with a peak in June and July. Ichthyoplankton samples collected in the bay in 1998 contained tautog larvae in June-August. We successfully succeeded in advancing the spawning date to late March using temperature and photoperiod manipulation, and obtained viable eggs and larvae. We intend to repeat the experi- ment this year with the addition of a control tank of fish which will be maintained at ambient temperature and photoperiod, and we will begin inducing spawning in February. GROWTH CHARACTERISTICS IN TRIPLOID PACIFIC OYSTERS— A NEW DIMENSION. Brenda Landau and Xim- ing Guo. Rutgers University. Haskin Shellfish Research Labora- tory. 6959 Miller Avenue, Port Norris. NJ 08349. Standard practice has been to measure length and whole weight of randomly sampled oysters as a means of documenting perfor- mance. During routine random sampling earlier, the observation was made that the height or thickness of triploid oysters compared to their diploid controls was noticeably larger. To test this hypoth- esis, an allometric study was done to compare four measurements, length, width, thickness, and whole weight of triploid oysters to those of their diploid controls for two different year classes, 1994 Milford Aquaculture Seminar, Milford, Connecticut Abstracts, February 27-March 1, 1999 271 and 1995. The triploids were produced from a previous study by diploid x tetraploid matings. It has been previously documented that triploid oysters grow faster than diploids and that polyploid gigantism may. in part, account for the larger overall size, though thickness measurements were not used. A comparison of means shows triploids to be 17.0%, 19.7%, 42.8%, and 93.2% larger than their diploid counterparts for length, width, thickness, and whole weight, respectively, for the '94 year class; and, 5.1%, 12.2%, 25.2%. and 46.3% larger for the '95 year class. An analysis of variance for the general linear model in which group (triploid vs. diploid) and replicate (k = 2) are main factors shows the group effect to be significant (p = 0.000 to 0.002) in both year classes for each of the four measurements. Results of this study show that thickness, rather than length and width, is the primary dimension for the increased growth in triploid oysters. Consequently, triploid oysters are more deeply cupped than diploids. This study used only two replicates, so a follow-up study with more replicates is needed to confirm these results. A SOCIAL AND ECONOMIC EVALUATION OF AN OYS- TER MARICULTURE TRAINING PROGRAM FOR LONG ISLAND COMMERCIAL FISHERMEN. Steven Lang, York College. The City University of New York. 94-20 Guy Brewer Blvd., Jamaica. N.Y. 11451. Despite a long history of shellfish mariculture, numerous public shellfish enhancement programs, and large tracts of available and potentially productive underwater land, New York's mariculture industry remains stagnant. For the most part, the constraints on shellfish mariculture are social, political and economic rather than environmental or technological. By far. one of the major obstacles hindering the development of mariculture has been the antagonis- tic attitude of commercial fishermen who have a long history of being opposed to the private use of public underwater land. The history of shellfishing in Long Island has been filled with struggles and conflicts between small-scale commercial fishermen who har- vest wild shellfish from the public resources and shellfish farmers who cultivate shellfish and are dependent upon a system of exclu- sive property rights. At the present time, amidst a steady decline in the natural stocks and shrinking opportunities caused by several factors, some fishermen are beginning to reconsider their negative attitude to- wards mariculture. While a few fishermen place great hope in mariculture' s potential to create new opportunities, most are skep- tical. For the potential of mariculture to be realized, attitudes will have to change so that it is viewed as a legitimate marine activity by members of the commercial fishing community. In 1995. through the East End Institute, funds were made avail- able from New York State to establish an oyster mariculture train- ing program for Long Island commercial fishermen to learn simple off-bottom culture techniques. In 1996. additional funds from the Fishing Industry Grants Program of the National Marine Fisheries Service were made available. Approximately 40 fisherman were given seed, culture gear, and informal training in oyster culture with the hope of them starting mariculture "cottage industries" that would supplement their incomes. A major objective of the oyster mariculture training program was to encourage small-scale mariculture by changing attitudes on the part of fishermen who have traditionally been opposed to it. The operating logic of the training program was to create oppor- tunities for a few fishermen to become successful so that other fishermen would become interested and pursue mariculture on their own. At the heart of the oyster mariculture training program is the notion that top-down development and management schemes initiated from distant external authorities are counterproductive and will not change negative attitudes on the part of fishermen. Change has to emerge from within the fishing community and must be facilitated in non-threatening ways which encourage ac- tive participation on the part of fishermen. The long and difficult process of institutionalizing mariculture as a way of life has to be based on some type of co-management arrangement between fish- ermen and government agencies. Development schemes have to incorporate attitudes and concerns of fishermen and include them in project designs and management arrangements. If developed rationally and in ways that are socially acceptable, mariculture could help to preserve traditional fishermen and enable them to follow the water in their customary ways. For the most part, the oyster mariculture training program has been successful. This paper will explore some of the reasons for that success as well as implications for the future of small-scale mariculture in the reaion. THE TRANSITION FROM COMMERCIAL FISHING TO OYSTER CULTURE: RESULTS OF A NMFS FISHING IN- DUSTRY GRANTS PROJECT. Richard Langan, Jackson Es- tuarine Laboratory, University of New Hampshire, 85 Adams Point Rd.. Durham, NH 03824. With support from the NOAA National Marine Fisheries Ser- vice Fishing Industry Grants Program, three New Hampshire com- mercial fishermen participated in a comprehensive oyster culture training program designed to give them an opportunity to evaluate shellfish culture as a part-time alternative to wild harvest fisheries. The fishermen were provided with guidance and assistance with site selection, permitting, evaluation of oyster culture methodolo- gies, three year classes of oyster seed, and the supplies and equip- ment needed to continue in aquaculture after project completion. Culture methodologies included remote setting of hatchery- reared eyed larvae on natural and artificial cultch. suspension nurs- ery culture, and bottom grow-out. Permitting woes, shortages of larvae, extreme weather events, an oil spill, the specter of MSX, and predation of oyster drills and green crabs were balanced by some excellent sets, good growth, and a positive production out- look and provided the fishermen with the opportunity to experi- 272 Abstracts. February 27-March I. L999 Milford Aquaculture Seminar. Milford. Connecticut ence first hand the risks and opportunities of shellfish culture. Of the three fishermen that participated in the project, two will very likely continue with oyster culture. EXPERIMENTAL TESTING OF FIELD TECHNIQUES FOR FARMING THE SOFT-SHELL CLAM (MYA ARENARIA). Kenneth J. La Valley, Thomas L. Howell, and Riley Y. Morse. Spinney Creek Shellfish. Inc.. Eliot. ME 03903; Brian Beal. and Bertrand Dubois, University of Maine. Machias. ME 04654. The purpose of this USDA/SBIR Phase I research project was to determine the feasibility of commercially farming the soft-shell clam. Mya arenaria. To accomplish this objective the research proposed to: 1 ) Produce a high volume of high quality hatchery- reared seed; 2) Optimize Floating Up-weller techniques for soft- shell clam culture; 3) Investigate the added benefits of condition- ing seed beds by harrowing; and 4) Investigate several seed plant- ing/grow-out techniques. To optimize hatchery production, larval stocking densities of 10K. 20K. 40K. and 100K larvae/gallon were evaluated for sur- vival and maturity to settlement. Pediveliger metamorphosis oc- curred predictably from days 18 to 21 at 22CC. Highest larval survival was observed at stocking densities between 20K and 30K larvae/gallon. The Floating Upweller System (FLUPSY) design exceeded our expectations, delivering 189 1/min. of upwelled water flow through each silo. The FLUPSY produced 20-22 mm animals in 2.5 months with near 100% survival. This reduced the grow-out time by up to three seasons compared to natural stocks. Soft-shell clams were stocked at high densities in upweller silos without a com- promise in growth or survival, which reinforced the commercial application of the FLUPSY design. A design capacity of 100K ( 10 mm) seed clams per silo was determined, for a total capacity of 1 million (10 mm) seed clams per ten silo Floating Upweller System. August field experiments were conducted at two sites in Ken- nebunkport. and one site in Portland. Maine to determine the added benefits of conditioning seed beds by harrowing. Soft bags similar to those used in the Florida quahog fishery were considered as a potential grow-out technique. Survival was a problem at all three sites. Netting, especially in the harrowed treatments at the Portland site, provided the best survival, and soft bags were found to be inappropriate, except, possibly in the softest muds. Recognizing the potential and regional importance of soft shell clam farming. Spinney Creek Shellfish (SCSI has begun to address the administrative framework necessary for fostering this new commercial activity. SCS has drafted proposals for enhancing the existing framework along with the Maine Department of Marine Resources so that the best possible revised structure is in place at the point in time that this activity comes to full commercial po- tential. In conclusion, the specific objectives were met. establishing this species as a candidate for commercial farming. The FLUPSY rapidly produced plantable seed clams with minimal maintenance, favoring scaling to commercial capacity. The bottleneck for this species was grow-out. Future research will include determining optimal planting size and sediment type, tidal height, tidal stage, and time of year for enhanced survival and growth. A TOUR OF UPWELLERS ON CAPE COD. Dale F. Leavitt. Patricia L. Gohring, and William P. Burt, Southeastern Massa- chusetts Aquaculture Center, c/o Hurley Library — Mass. Maritime Academy. 101 Academy Drive. Buzzards Bay. MA 02532. The use of upwelling culture systems for nursery grow-out of commercially important bivalve mollusks has become an impor- tant component of community and private shellfish aquaculture on Cape Cod. The nursery phase of bivalve culture is frequently a limiting step for bivalve seed production due to limitations in space and food availability in commercial hatcheries. A concerted effort has been made on Cape Cod to increase shellfish seed pro- duction by expanding the region"s capability to raise seed through the nursery phase. During September, the Southeastern Massachu- setts Aquaculture Center (SEMAC) conducted a tour of ten dif- ferent upweller systems to investigate the design and operation of a variety of approaches to upwelling. The technical information compiled from these systems will be presented along with a pic- torial display of various approaches to upweller design. THE NEED FOR AQUACULTURE IN THE WORLD TO- DAY. Robert Link, Liquid Life Technologies. Inc.. 1727 Veter- ans Memorial Highway, Islandia, NY 1 1722. This paper will describe components of the finfish and shellfish aquaculture industries while explaining the challenges and oppor- tunities that exist in those industries. Although the aquaculture industry is growing at a rapid pace, there are some impediments that affect all parts which include: regulations, marketing, and financing. All of these components are necessary for healthy arowth and will be discussed in detail. RECENT STREAMLINING OF THE AQUACULTURE REGULATORY PROCESS. Michael Ludwig. USDOC, NOAA, National Marine Fisheries Service, Habitat Conservation Division, Milford, CT 06460. While aquaculture operations have been permitted and pres- ently occur in the Gulf of Mexico. Atlantic and Pacific waters, initial permitting efforts for a given geographical area, a new cul- ture species or innovative technology can be hampered by a lack of Milford Aquaculture Seminar, Milford. Connecticut Abstracts. February 27-March 1. 1999 273 understanding among the applicant, regulators and other involved parties. In addition, the public is often confused by what they perceive as conflicting positions taken by the regulatory agencies. Their confusion often arises from a perception that the agencies are a single entity rather than consortiums of representatives from a variety of programs, each with different (and occasionally con- flicting) mandates and responsibilities. The differing expectations of all involved parties about the amount or type of information needed to describe an aquaculture proposal before it is deemed ready for evaluation can result in costly and protracted reviews. The National Marine Fisheries Service (NMFS) Northeast Re- gion's Habitat Conservation and Protected Resources Divisions has consolidated much of the guidance offered to aquaculturists across the US and Canada. The document was created to establish a standard level of information quality for use in seeking federal authorization of aquaculture projects. The document is intended to be all encompassing and not the minimum required for permit consideration. The all-species package contains guidance from which parties in a regulatory action can select elements for use in application forums. It is our expectation that the document will become the standard for submissions of environmental compat- ibility and regulatory acceptability in the Northeast and provide a semblance of uniformity for regional evaluation processes. It is our intention that compliance with the entire guidance package would be required only on occasions when the ecological sensitivity of a proposed culturing site or other project details are so complex as to warrant that degree of thoroughness. By seeking out those in- volved in the regulatory process as early as possible in the devel- opment of a proposal and using tools such as pre-applications meetings, an applicant will be able to identify the project assess- ment requirements to which one will be held. This will greatly facilitate and even expedite a project's evaluation. However, we recommend that before undertaking any data gathering or committing to any physical site evaluation efforts, the applicant seeks guidance and thoroughly coordinate planning and site evaluation efforts with the appropriate regulatory agencies. The Guidance package will be described and made available at the meeting. THE EFFECTS OF STOCKING DENSITY ON GROWTH OF LARVAL TAUTOG. Lindsay Lydon and Grace Klein- MacPhee, University of Rhode Island, Graduate School of Ocean- ography. Narragansett Bay Campus. Narragansett RI 02882. A preliminary study was conducted on larval tautog to deter- mine if two different 15 gallon tank shapes, conical and rectangu- lar, affected tautog survival. Although larval survival did not ap- pear to be related to tank form, the stocking density influenced tautog growth. Tautog eggs were collected from broodstock that spawned 7/29/98 and 7/30/98. After 8 hours of incubation. 10 mis of eggs were added to the four tanks with a static system. Green alga. Tetraselmis suecica was added daily to each tank for the first 19 days. Larvae were fed a combination of dry and live food, rotifers, artemia. and a Kyowa™ diet. The tanks were transferred to flow- through systems after larvae were 19 days old. Survival was good in all tanks except for one conical which had 100% mortality by day 5. Although additional tautog were supplemented from a rect- angular tank with a high density (848) of fish, these larvae in the conical tank developed abnormally. The majority of fish (92%) in this tank exhibited an unusual jaw development which prevented mouth closure. This jaw anomaly most likely interfered with their ability to eat dry food and to make a complete transition from live food, Artemia nauplii, to dry food. As a result, poor growth oc- curred in this tank. After 14 weeks there was a statistically significant difference in size between fish groups. Fish growth was best in the tank with an initial stocking density of 2.8 fish/liter with a total of about 148 tautog. These fish had a mean length of 36 mm and weight of .90 grams compared to 29.28 mm and .50 grams for the tank with the highest larval density of 848 tautog at a stocking density of 15 fish/liter. The results of this study indicated that 15 gallon experi- mental tanks are suitable for rearing tautog larvae at low stocking densities. Larger systems would be required to raise tautog at higher stocking densities. Further studies should be conducted on larval tautog to determine stocking density in relation to optimum growth and survival. "'Trademarks are used to identify products only and are not indicative of endorsement by URL PROGRESS IN BIOECONOMIC EVALUATION OF THE MILFORD LABORATORY SCALLOP NURSERY RECIR- CULATING SYSTEM. Gisele Magnusson and James Ander- son, Environmental and Natural Resource Economics, University of Rhode Island. Kingston, RI 02881. Costs and returns for a land-based recirculation nursery system for bay scallops (Argopecten irradians) were calculated and key economic factors affecting the financial viability of the system identified. The system under consideration was developed by the NMFS Milford Laboratory and incorporated a greenhouse for al- gae production with a re-circulating nursery to test several differ- ent filtration systems. A bioeconomic simulation incorporating the stochastic nature of key variables and the dynamic nature of an integrated production system was developed. Rudimentary hatch- ery and grow-out systems were incorporated to track costs and revenue impacts of various changes to the nursery system. Pre- liminary results of the model suggest that the average cost of algae from the greenhouse was within the range of published results. However, both the average cost of algae and the growth rate of juvenile scallops will have to change significantly to allow such a system to be financially viable. For the greenhouse system, labor, capital costs and nutrient costs are significant, while to the nursery, algae costs, labor and capital costs were most important. A value 274 Abstracts. February 27-March 1. 1999 Milford Aquaeulture Seminar, Milford, Connecticut for automation systems can be calculated based on the potential savings in labor and nutrient costs, over the base cost. AUSTRALIAN/TASMANIAN OYSTER CULTURE. Harri- ette L. Phelps, University of the District of Columbia. Biology Department. 4200 Connecticut Ave.. N.W.. Washington. DC 20008. In Australia, the two main cultured oyster species are the Syd- ney Rock Oyster (Saccostrea commercialis) and the Japanese Oys- ter (Crassostrea gigas). The Sydney Rock Oyster is presently cul- tured mostly in a few estuaries near Sydney and although the Japanese Oyster is found in some of those estuaries, the oyster farmers consider C. gigas highly detrimental because of competi- tion for space from its earlier settlement pattern and faster growth. The majority of commercial oysters are C. gigas spawned and grown in Southern Australia and the island of Tasmania. Tasmania is lightly settled and has numerous shallow estuaries with clean water mostly on the east coast and used for aquaculture. Presently. the oyster farmers send seed oysters for spawning to the oyster hatchery at Bicheno. The seed oysters are conditioned, spawned together, and the larvae raised in large tanks with some cultured algae addition. The oyster larvae set and transform on finely ground scallop shell added to the tanks. The young cultchless spat are returned to the farmers and reared in upwellers until transferred to bags set on trays in the shallow estuaries. The oysters are trans- ferred to larger mesh bags until ready for sale at two years, at which time they are marketed mostly in coastal Australian cities as the live-shell product. I saw several varieties of C. gigas being raised at one facility I toured: black, golden, striped-shell, etc. What 1 found interesting was that the central Bicheno hatchery made no attempt (unless requested) to separate the spawning oyster stocks, yet said they could tell which estuary where an oyster was raised in by its shape or other physical characteristics. Sometimes it was impossible to tell live 5. commercialis from C. gigas except by the inner shell teeth of the Rock Oyster. However, the flavor was quite different. A COMPARISON OF ANTICLUMPING SOLUTIONS USED FOR INITIAL RECOVERY OF HEMOCYTES FROM THE BAY SCALLOP (ARGOPECTEN IRRADIANSl Steven Pitchford and Richard A. Robohm. USDOC, NOAA. National Marine Fisheries Service, Northeast Fisheries Science Center, Milford Laboratory. Milford, CT 06460. Efforts to study the immune capabilities of bay scallop hemocytes in vitro were hampered by excessive clumping of cells upon withdrawal of hemolymph from adductor muscles. To re- solve this problem, seven solutions used by others to prevent clumping of blood cells in vitro (five used in other invertebrate species and two in human medicine) were modified by adjusting osmolality and pH to match that of scallop blood and examined for their ability to prevent scallop cell aggregation. In brief, the protocol consisted of withdrawing hemolymph from scallop adductor muscles, dispensing it into each of the an- ticoagulant solutions in multi-chambered, glass, microscope slides and allowing the hemocytes to attach. This was followed by cell fixation. The number of single attached cells, small clumps (2—4 cells), and large clumps (> 4 cells) were counted in at least 15 fields for each of the anticoagulants. In addition, observations were made on the appearance and relative degree of attachment of the hemocytes to the substrate. Results were collected using weighted values from three ex- periments; each experiment used hemolymph from three scal- lops— each exposed to all seven solutions. Dunnett's test for pair- wise multiple comparisons showed that a modified solution. Ade- ma's solution, previously used to prevent clumping in cells of a freshwater snail, was statistically superior to all but one of the other solutions. Three solutions developed by others for use with various molluscs were very poor in their ability to prevent clump- ing of scallop cells. Use of the best solution will be essential in many of our subsequent studies of bay scallop immunity. EFFECT OF DIETARY PH ON THE UTILIZATION OF SEMIPURIFIED DIETS BY TAUTOG, TAUTOGA ONITIS. Laurel J. Ramseyer, National Academy of Sciences, USDOC, NOAA, National Marine Fisheries Service, Northeast Fisheries Science Center, Milford Laboratory. Milford, CT 06460. Experiments were conducted to determine whether semipuri- fied diets would support growth in the tautog. Tautoga onitis, a stomachless fish. Semipurified diets contained 50% protein and either 7% or 12% lipid on a dry matter basis. A commercial salmo- nid feed containing 547c protein and 15% lipid on a dry matter basis was used as a reference feed. Semipurified diets were either as-is. with intrinsic pHs of 5.5-5.8. or with pH adjusted during preparation to 7.2-7.6. Experiments were conducted with 6-12 g tautog at 20°C for 28-31 d. Tautog were fed the reference and the 7% lipid semipurified diets at rates providing 10. 12. or 14 g protein • kg-1 body weight • d"1. Fish fed the semipurified diets required at least 14 g protein • kg-1 body weight • d"1 for weight gain, whereas fish fed the reference feed gained weight at all three feeding rates. Fish fed the pH 7.2 diet gained more weight than fish fed the pH 5.5 diet. However, when dietary lipid was increased to 12%. weight gain of fish fed the lower pH diet was not significantly different from weight gain of fish fed the pH 7.6 diet or the reference feed. The pH of digesta was 8.7-9. 1 throughout the gut regardless of dietary treatment. The results indicated that the alkalinization of digesta in tautog is an energy-dependent process. Semipurified diets sup- ported growth in tautog. but should be alkalinized to a standard pH before use in tautog nutrition studies. Milford Aquaculture Seminar. Milford. Connecticut Abstracts, February 27-March 1, 1999 275 VIBRIO PARAHAEMOLYTICUS— A NEW PROBLEM FOR THE SHELLFISH INDUSTRY IN THE NORTHEAST. David R. Relyea, Frank M. Flower and Sons Inc., P.O. Box 88. Oyster Bay. NY 11771. During the time period between 8/10/98 and 8/29/98. eight cases of gastroenteritis occurred and were eventually reported to the New York State Department of Environmental Conservation (NYSDEC). The illnesses occurred in Nassau and Suffolk counties in New York (61 and (2) cases were from New Jersey. Stool samples from patients indicated that the illnesses were caused by a naturally occurring marine bacterium. Vibrio parahaemolyticus. Tagging information seemed to indicate that the source of the bacteria was oysters and clams from area NS2 which includes Oyster Bay and Cold Spring Harbor. However, most patients had also eaten other seafoods (crabs, shrimp, etc.) that are known to be sources of Vibrio parahaemolyticus. Health Department officials claimed that the only food common to all patients was shellfish from NS2. New York State Department of Health (NYSDOH) and USFDA notified NYSDEC that NS2 had to be closed and NS2 was closed to shellfishing on 9/10/98. Due to confusion and insufficient Federal guidelines the area was not able to be reopened until 10/22/98. During that time Frank M. Flower and Sons with 40 employees and about 50 individual baymen had no source of income. This presentation gives the industry perspective of this perplexing problem. NOAA FISHERIES AND AQUACULTURE. Edwin Rhodes. USDOC. NOAA. National Marine Fisheries Service. 1315 East- West Highway. Silver Spring, MD 20910. Aquaculture has played a significant role in NOAA Fisheries and its predecessor agencies since their origins in the 19th century. The continuing efforts by the agency in its 127 year history con- tributed some of the key science in the field of aquaculture. in- cluding research that contributed to the commercial development of salmon, shrimp and shellfish culture. Since the 1980's, agency priorities in the area of fisheries man- agement, coupled with budget limitations, have restricted the par- ticipation of NOAA Fisheries in aquaculture. Recently, aquacul- ture has reemerged as an important consideration as NOAA Fish- eries plans for the new century. This planning and policy development stage is critical because it is through this process that agency priorities are set and budgets are driven. The NOAA Fish- eries strategic plan has as one of its objectives to "promote the development of robust and environmentally sound aquaculture" and outlines specific goals in the areas of technology development, siting, permitting and financial assistance. Partly based on this plan, the Northeast and Northwest Centers of NOAA Fisheries have reorganized to include aquaculture divisions, and new aqua- culture industry financing programs are being developed. At the NOAA level. NOAA Fisheries, the National Ocean Ser- vice and the Office of Oceanic and Atmospheric Research have collaborated to put a new aquaculture policy in place that recog- nizes the significant role that environmentally sound aquaculture will play in meeting future demand for seafood, as well as the potential to contribute to wild stocks through enhancement. The NOAA policy also foresees a major aquaculture effort for the production of non-food products such as bait, aquaria species, chemicals and pharmaceuticals. Finally, an active task force is developing a Commerce-wide policy for aquaculture, and its formulating plans to facilitate aqua- culture permitting in the U.S. exclusive economic zone. This policy and planning activity has helped to generate a new interest in aquaculture in the Department and the recognition of the po- tential for aquaculture speaks to an optimistic future. CONTROL OF EUTROPHICATION BY BIVALVES: FIL- TRATION OF PARTICULATES AND REMOVAL OF NI- TROGEN THROUGH HARVEST OF RAPIDLY GROWING STOCKS. Michael A. Rice, Department of Fisheries, Animal and Veterinary Science. University of Rhode Island. Kingston. RI 02881. Filter feeding by populations of bivalves has been suggested as a means of reducing eutrophication in coastal estuaries by exerting control of phytoplankton populations in the water column. In some estuaries, programs have been established for the purpose of im- proving water quality and, frequently, large populations of mature shellfish that reside behind pollution closure lines in estuaries rep- resent a large filter feeding biomass. The rate of filter feeding by bivalves is size dependent and allometrically related to shell di- mensions, so the largest and oldest individuals filter the greatest volumes of water. In most areas closed to shellfishing. bivalve populations are composed of mostly older adults. These large ani- mals are slow growing, have a low rate of new tissue production in relation to standing crop biomass. and have a neutral nitrogen balance (organic-N assimilated = NH,-N excreted). These large adults increase sedimentation through filter feeding, but since they are neither harvested nor growing rapidly, they do not directly remove much nitrogen from the estuary. Although it is possible that increased sedimentation can lead to greater denitrification rates in the sediments. The only way filter feeding can directly remove nitrogen from the environment is through tissue growth. The dry weight of the soft tissues of most bivalves is typically around 307c protein, so for each kilogram of shucked shellfish meats harvested there are 16.8 grams of organic nitrogen removed from the estuary. Nutrient removal from estuaries can be maxi- mized through management of shellfisheries for maximum bio- mass production and harvest, and the development of aquaculture projects in which rapidly growing shellfish are harvested regularly. This is publication 3681 of the College of the Environment and Life Sciences. University of Rhode Island. 276 Abstracts. February 27-March 1. 1999 Milford Aquaculture Seminar. Milford. Connecticut SUMMER FLOUNDER CULTURE IN THE NORTHEAST: UPDATE ON RECENT RESEARCH AND INDUSTRY STA- TUS. Gregg Rivara, Cornell Cooperative Extension-Suffolk County. 3690 Cedar Beach Road, Southold. NY 1 1971; David A. Bengtson, Department of Fisheries. Animal and Veterinary Sci- ence. University of Rhode Island. Kingston. RI 02881. The Northeastern Regional Aquaculture Center and Sea Grant have been sponsoring research on summer flounder as an emerging aquaculture species in the Northeast. As part of outreach and ex- tension efforts, a workshop is being held for industry just prior to the 19th Milford Aquaculture Seminar. Researchers are presenting their results to industry at that workshop and also describing the future research for which they have received funding. A roundtable discussion involving researchers and industry is also part of the workshop so that industry can describe their research needs. The industry is still small, but is making progress and starting to sell product. A summary of the workshop activities, major research findings and industry status will be provided to the participants at the 19th Milford Aquaculture Seminar. AN OVERVIEW OF AQUACULTURE RESEARCH IN AT- LANTIC CANADA. Shawn M.C. Robinson. Dept. Fisheries and Oceans. Biological Station, St. Andrews, New Brunswick, Canada. EOG 2X0. The aquaculture industry in Canada is in a growth phase. Since 1986. shellfish culture has grown in production volume at an an- nual rate of 109c and finfish culture has grown at an annual rate of 28%. This rapid growth in the industry has fueled a push in re- search to support the development of existing species in culture as well as to bring new species on-line. There is also active research on factors that affect the industry such as disease and environmen- tal interactions. In Atlantic Canada, the lead federal agency for aquaculture research is the Dept. of Fisheries and Oceans. Its role is to provide scientific knowledge for the sustainable development of aquacul- ture including the development of an economically competitive and environmentally sustainable industry. The research laborato- ries are located at the Biological Station in St. Andrews, New Brunswick, the Gulf Fisheries Centre in Moncton, New Bruns- wick, the Bedford Institute of Oceanography in Dartmouth. Nova Scotia and the Northwest Atlantic Fisheries Centre in St. John's. Newfoundland. However, there are other major research organi- zations in Atlantic Canada as well such as: 1 ) provincial aquacul- ture agencies in Newfoundland. Nova Scotia. Prince Edward Is- land. New Brunswick and Quebec 2) the National Research Coun- cil-Institute for Marine Biosciences in Halifax Nova Scotia 3) the universities (UNB. Moncton, UPEI-AVC, NSAC, Acadia, Dal- housie. Laval. Quebec and Memorial) 4) the Canadian Centre for Fisheries Innovation and 5) the industry itself including the pro- vincial aquaculture associations. The majority of the research is done in collaboration with industry partners. Research on finfish has been strongly emphasized to date. At- lantic salmon is the most commercialized species so far and much of the early developmental work has been done. Research on this species is ongoing in the Bay of Fundy and Newfoundland and is concentrating on broodstock (new strains, transgenics), fish health (record of performance, husbandry practices, therapeutants). nu- trition (area and species specific diets), grow-out technology and environmental linkages (freshwater discharge, waste manage- ment). There are also a number of new finfish species that are being studied such as: halibut, haddock, winter flounder, striped bass, steelhead and American eels. These studies are going on primarily in Newfoundland. Nova Scotia and New Brunswick. The scope of research on these new species falls into four categories: 1 ) Broodstock/seedstock (environmental influences on maturation, influence of diets, prediction of maturation, capture techniques) 2) Fish health (identification and life histories of diseases and para- sites, baseline data on normal fish) 3) Nutrition (nutritional require- ments, culture techniques for native plankton, micro-encapsulated larval diets, larval feeding behaviour) and 4) Grow-out (early rear- ing techniques, tank and grow-out designs, refinement of auto- matic feeding techniques, in situ estimation of fish size in cages). Shellfish research is active in all provinces. The industry is presently mostly located in Newfoundland, the Atlantic coast of Nova Scotia and the Gulf of St. Lawrence although it is starting to grow in the Bay of Fundy. In general there are four research areas being targeted: 1 ) Broodstock/seedstock (hatchery development, natural spat collection) 2) Shellfish health (identification of dis- eases and parasites, diagnostic tools and treatments) 3) Grow-out (optimize rearing of juveniles, seeding densities, predator control, roe enhancement) 4) Environment (environmental effects on growth and survival, site selection, carrying capacity, effects of winter ice). Species being studied are: blue mussels, sea scallops. American oysters, European oysters, hard-shell clams, soft-shell clams and sea urchins. There is also some work being done on bio-fouling species such as tunicates. There is a small program on algal research. Programs are un- derway on the dynamics of phytoplankton blooms and some shell fish sites are being monitored for toxic algal effects. Grow-out trials are being done on some macro-algal species such as dulse (Pabnaria palmata), nori (Porphyra spp. ) and kelp (Laminaria longicruris). As the marine culture industry develops, there is an increasing research effort being directed toward the linkages between the commercial culture of various species and the environment. Oceanographic modeling techniques are being developed for area management strategies, site assessment and remediation tech- niques are being studied, and practical methods for monitoring by the industry are being developed. Past research has contributed substantially to the development of the Atlantic Provinces aquaculture industry and there is strong sup- port from industry for work in the future. The major research impedi- ment to-date has been securing reliable long-term research funds. Milford Aquaculture Seminar, Milford. Connecticut Abstracts, February 27-March 1. 1999 277 POST-METAMORPHIC GROWTH OF SUMMER FLOUN- DER IN LABORATORY CULTURE: DO EARLY- SETTLING LARVAE GROW FASTER THAN LATE SET- TLERS? Tessa L. Simlick, Robin S. Katersky, Neil Marcaccio. and David A. Bengtson. Department of Fisheries. Animal and Veterinary Science, University of Rhode Island. Kingston, RI 02881. Laboratory-reared summer flounder larvae begin to settle to a benthic existence 30 to 35 days after hatching but settlement can continue for about a 30-day period, because completion of meta- morphosis among individuals does not occur simultaneously. We perform weekly gradings (i.e., removal of settled flounder) until all fish have settled in order to prevent cannibalism and stress, be- cause newly settled juveniles tend to be larger than swimming larvae. Although we know there is a strong correlation between larval growth and time of settlement (fastest growers settle first), no data exist on post-settlement growth variability. We wanted to know whether fast-growing larvae become fast-growing juveniles or whether slow-growing larvae can 'catch up' in growth rate. Experiments were designed and conducted at the Narragansett Bay Campus Research Facility to explore these inquiries. Settled fish were graded from the larval tank at 32 days after hatch (DAH) (Grade 1 ), 39 DAH (Grade 2), and 46 DAH (Grade 3). Graded fish were randomly placed in three replicate 75-L aquaria per grade, at a density of 30 fish per aquarium. Flounder were fed Artemia for 30 days after removal from the larval tank and were then weaned onto a commercial diet. All fish were measured by Image Analysis at bi-weekly intervals until the fish were 95 DAH. No significant differences in post-settlement growth rate were seen among the three grades. In the final set of measurements, the fish exhibited an increase in size variation within replicates and cannibalistic attacks were again causing mortality. Future experiments will continue to investigate specific growth rate variation in all stages of juvenile crowth. FERTILIZATION RATES AND PROCEDURES USING COMMERCIALF/2" NUTRIENT MIXES TO GROW T- ISO (ISOCHRYSIS SP.) AND PLY429 (TETRASELMIS CHUI). Barry C. Smith, Sara Barcia, Jennifer H. Alix, and Gary H. Wikfors, USDOC. NOAA. National Marine Fisheries Service. Northeast Fisheries Science Center, Milford Laboratory. Milford. CT 06460. Two developments intended to make microalgal feed culture more convenient have collided. First, pre-mixed concentrated nu- trient products, formulated according to Guillard's "f/2" recipe, are entering wide use; these products are sold in two solutions that are kept separate to avoid chemical complex formation until used to make culture media. The second development is the use of auto- mated methods of adding nutrients to culture water, e.g.. metering pumps and venturi eductors. The two-part nutrient mixes will re- quire duplicate apparatus for their automated dispensing, increas- ing technical complication, chances for malfunction, and cost. A simple solution (pun intended) to this dilemma would be a dilution with water of the combined two-part product that would remain stable. Several dilutions of combined f/2 concentrates were ob- served over two weeks for visible precipitates. Then, culture media were prepared with the concentrate dilutions, and their ability to support growth of T-ISO and PLY429 was compared to freshly- prepared f/2 of two brands. In a nested Analysis of Variance de- sign, four concentrate dilutions (Part A:Part B:water, 50:50:0. 40: 40:20, 30:30:40, 20:20:60) were used to prepare four final nutrient media (f/4. f/2, f. 2f) each. Algal division rates were calculated from optical density readings of triplicate test-tube cultures, and final population densities were determined by microscope cell- counts. All nutrient dilutions precipitated but were re-dissolved easily. For Tetraselmis chui, PLY429, nutrient pre-dilution and final con- centration had significant effects upon division rate and final popu- lation. Maximal division rates were higher when nutrient solutions were pre-diluted. and significantly higher in the 2f media as com- pared with lower enrichments. Final cell densities tended to be higher when nutrients were not pre-diluted. especially at lower fertilization rates. Predictably, higher fertilization rates led to higher final cell yields, but a nitrogen budget analysis showed nearly half of the added nitrate was not taken up by PLY429 at the 2f enrichment. For Isochrysis sp., T-ISO, division rate was highest in the most pre-diluted nutrients and lowest at the highest fertil- ization rate. Cell yield of T-ISO tended to be higher in nutrient mixes less pre-diluted and was significantly lower in the f/4 en- richment compared with higher fertilization rates. Nitrogen budget analysis of T-ISO cultures showed that nitrate remained unassimi- lated at all concentrations above the f/4 enrichment. Indications from these experiments are: I ) enrichments above the "f level for PLY429 and above "f/4" for T-ISO result in wasted nutrients, suggesting that some other nutrient (perhaps vi- tamins) limited T-ISO in these experiments: 2) pre-dilution of two-part commercial algal fertilizers increases maximal division rate, but not final cell yield, suggesting that chemical complexation is reversed during algal growth, making nutrients available to algae over time. Pre-dilution of combined two-part algal fertilizer prod- ucts can affect performance of cultures: the decision to pre-dilute will depend upon whether cultures are being optimized for rate or yield. UPDATING THE PLANS FOR SEA SCALLOP AQUACUL- TURE IN MASSACHUSETTS. Ron Smolowitz. Coonamessett Farm. 277 Hatchville Rd„ East Falmouth, MA 02536: and Harlyn Halvorson. UMASS-Boston. ECOS, College of Arts and Sci- ences, 100 Morrissey Blvd.. Boston. MA 02125. The Sea Scallop Working Group (SSWG) was started five years ago to provide a forum for discussion of interests in aqua- culture among various stakeholders. These discussions led to the 278 Abstracts. February 27-March 1. 1999 Milford Aquaculture Seminar. Milford, Conneetieut development of a Blueprint for Sea Scallop Aquaculture in 1995. A SSWG summit meeting was held February 8-9 at the Massa- chusetts Maritime Academy to evaluate our progress to date on various sea scallop projects and to review the problems encoun- tered. Through the use of breakout groups, the recommendations of the 1995 Blueprint were reexamined and priorities set for SSWG activities in the coming year. The results of this summit meeting will be reviewed. METHODOLOGY FOR THE GENERATION OF POLY- MORPHIC MOLECULAR TAGS IN THE BAY SCALLOP. ARGOPECTEN IRRADIANS. Jeff Southworth, Maronda Brown. Department of Molecular & Cellular Biology. University of Connecticut. Storrs, CT 06269; Sheila Stiles, USDOC. NOAA, National Marine Fisheries Service, Northeast Fisheries Science Center. Milford Laboratory. Milford. CT 06460; and Linda Strausbaugh. Department of Molecular & Cellular Biology. Uni- versity of Connecticut. Storrs. CT 06269. Several promising species for aquaculture lack genetic and morphological markers. Consequently, there is a critical need for the development of physical and genetic tags for monitoring and identifying brood stocks. We are examining methods to differen- tiate bay scallop populations (Argopecten irradians) at the geno- typic level using Type I and Type II markers. We have chosen to investigate the potential of a Polymerase Chain Reaction (PCR). and Random Amplification of Polymorphic DNA (RAPD) [similar to DNA Fingerprinting (DAF)] or Arbitrarily Primed PCR (AP- PCR) techniques. We have examined levels of polymorphism within and among 5 populations of bay scallops collected from the United States northeast coast. Approximately twenty PCR primers were ana- lyzed for their effectiveness at revealing polymorphisms among individuals and populations of this marine organism. A signifi- cant amount of time was spent optimizing the primers and the PCR conditions to obtain clearer, more consistent results. We have determined that there is a high degree of polymorphism at the level of individual organism genotype. In addition. RAPD analysis is reliable and reproducible, as well as extremely sensi- tive. In a complementary approach we have selected coding regions that might provide species and/or strain-specific markers as well as promoters for genetic engineering applications. Addi- tional investigations include the design of core histone gene primers from both Drosophila melanogaster and Strongylocentro- tus purpuratus to screen a genomic library of both Placopecten magellanicus (sea scallop) and of the bay scallop Argopecten irradians. Development of molecular tags can provide a screen for genetic diversity, ultimately circumventing inbreeding depres- sion. DISEASE-RESISTANT OYSTERS, CRASSOSTREA VIR- GINICA, IN LONG ISLAND SOUND. Inke Sunila, John Volk and John Karolus. State of Connecticut. Department of Agricul- ture. Bureau of Aquaculture. P.O. Box 97. Milford. CT 06460; Terry Backer. Long Island Soundkeeper Fund. Inc.. P.O. Box 4058. East Norwalk. CT 06855; Stan Czyzyk. Bluepoints Co.. Inc.. Atlantic Avenue. P.O. Box 8, West Sayville. NY 1 1796: Ed Lang. P.O. Box 314. Clinton. CT 06413: Matt Mroczka. Cedar Island Marina Research Laboratory. P.O. Box 181. Clinton. CT 06413; Karen Rivara. AEROS Cultured Oyster Company. 41 Heathcote Court. Shirley. NY 1 1967. Under heavy infection pressure, oysters develop resistance to parasitic diseases such as MSX (Haplosporidium nelsoni). Resis- tant oysters still get infected, but their mortality rate is lower than that of susceptible oysters. Genetic resistance can be developed against other economically important oyster diseases such as Dermo-disease (Perkinsus marinus) or JOD (Juvenile Oyster Dis- ease). A MSX-epizootic, associated with high mortalities in some areas, raged in Long Island Sound (LIS) starting in 1997. Hatch- ery-raised, highly susceptible seed experienced 99% mortality. Connecticut's commercial oyster companies were advised to in- crease the prevalence of resistant oysters in two different ways: 1. Establishing brood stock sanctuaries in heavily infected sites to permit survivors to produce resistant seed. This could be done by not harvesting part of the infected lot (10% area) for a period of three years, and 2. Selecting disease-resistant strains when using hatchery-raised seed. The prevalence of potentially MSX-resistant oysters, based on histological characteristics, in the field increased eightfold from 1997 to 1998 on previously exposed sites. A cooperative program was initiated with production hatcheries to produce a commercially available, disease-resistant oyster seed especially bred for Long Island Sound conditions. Present management strategies respect traditional oyster culture methods, which include deployment of hatchery-raised seed concurrently with natural set. This gave rise to the need of developing a hatchery stock with a spawning cycle compatible with wild oysters. Spawning time is an inherited characteristic, which is maintained upon transplanta- tion. The new oyster strain would have the characteristics for spawning time, growth, temperature and salinity tolerance and hardiness of the parent population in LIS. In addition, it would have been selected for disease resistance. Brood stock was created from survivors of 90% mortality (83% MSX prevalence. 100% Dermo prevalence) from Clinton. Connecticut. A commercially available, local seed was used as a control. "Clinton" strains have been tested since the spring 1998 in an infected location for growth, mortality and infection rates. During the first sea- son. "Clinton" had a 16% higher growth rate than the control. Both strains acquired MSX and Dermo infection. "Clinton" experienced a 1% mortality, control 11%. The first generation Milford Aquaculture Seminar, Milford. Connecticut Abstracts, February 27-March I. 1999 279 "Clinton" showed superior survival and growth characteristics. This seed has been deployed in commercial scale at several sites in LIS. Grow-out facilities have been established to grow hatchery-raised seed prior to dispersing it to the field. Brood stocks are exposed in different locations to other oyster diseases such as Dermo, JOD and SSO (Seaside Organism or related spe- cies). REFLECTIONS ON BIOFILTER SELECTION FOR SHELLFISH CULTURE. James C. Widman Jr.. USDOC. NOAA. National Marine Fisheries Service. Northeast Fisheries Science Center. Milford Laboratory. Milford. CT 06460. Selecting biofilters for shellfish recirculating systems can be a perplexing process. Many of the biofilters used in finfish recircu- lating systems have numerous disadvantages when the require- ments of shellfish culture are considered. One of the main require- ments is live phytoplankton as a food source. Many commercially available biofiltration systems are capable of trapping and remov- ing phytoplankton-size particles. Finfish systems, on the other hand, are designed to remove fecal material and unused food from the water stream. Many of them also use high flow rates essential to many finfish species but usually detrimental to shellfish. These systems must be modified before being considered for shellfish recirculating systems. Bead filters are effective at removing and trapping phytoplank- ton size particles and sand-bed filters have similar drawbacks. Protein skimmers/foam fractionators not only remove organics, but also phytoplankton-size particles. Numerous drum filters capable of removing fish wastes do not appear to be appropriate for the fragile waste products of shellfish; some mesh sizes also remove phytoplankton. Many of these systems are still undergoing modi- fications and may eventually evolve into effective shellfish biofil- ters in the future. There are systems that appear to satisfy the requirements of shellfish culture today. Many of these systems act as contact fil- ters, basically providing large surface areas for nitrifying bacteria to grow. Rotating biological contact filters (RBC's). Aquacube. moving bed biofilters. and submerged panels are examples of this technology. Trickle filters and fluidized sand bed filters may be effective, but may be subject to fouling, or damage to the phytoplankton. Caution must be exercised when biofilter specifications mention the removal of solids. Pads. mats, foams and types of trickle filter media can become fouled with both waste material and phy- toplankton. making them less efficient at nitrogen removal. FEEDING RATIONS AND REGIMES FOR POST-SET OYS- TERS, CRASSOSTREA VIRGINICA, FED CULTURED MI- CROALGAE IN A LAND-BASED NURSERY. Gary H. Wik- fors, Jennifer H. Alix, Mark S. Dixon, and Barry C. Smith. USDOC. NOAA. National Marine Fisheries Service. Northeast Fisheries Science Center. Milford Laboratory, Milford. CT 06460. Filter-feeding bivalves, such as the eastern oyster, have evolved feeding behaviors that respond to changes in both quantity and quality of suspended particles. In near-shore waters, particle load- ings are highly variable and beyond the control of the shellfish farmer relying upon natural primary production to feed his or her seed oysters. The risks associated with "raw water" nursery culture of oysters (poor nutrition; exposure to pollutants, disease, and predators; vandalism), and seasonal limitations, eventually will exceed the costs of controlled, land-based nursery culture. As this occurs, information needs about oyster feeding will shift from describing responses to varying environmental conditions to pro- viding feeds in a way that optimizes their use by the spat. Feeding standards developed for animal agriculture that list daily nutri- tional input, in biochemical terms, and growth obtained on these specific diets, represent a useful model for aquacultured animals. For filter feeders, however, the "daily allowance" concept is com- plicated somewhat by the behavioral responses to particle loadings mentioned above. To begin the process of identifying practical feeding rations and regimes for post-set oysters, we conducted an experiment in which daily rations were varied (1. 2. 5. or 10% of oyster live weight in dry matter of feed) and each ration was provided in 2. 4, or 16 feedings each day; a 50:50 mix of two high-lipid Tetraselmis strains, PLY429 and PLAT-P, was used for all experimental treat- ments. Oyster growth was determined weekly — in terms of live weight, volume displacement, and shell size — and at the end of the seven-week experiment in terms of dry weight. Feed conversion efficiency was calculated from change in oyster dry weights and the sum of algal dry weight feed provided during the experiment. Oysters (65 mg live weight initially) grew progressively faster on 1. 2. and 5% rations, regardless of regime, but there was no significant increase in growth when the ration was increased to 10%. Oysters grew fastest when fed most often, but the statistical significance of this effect was dependent upon ration, e.g., at the highest ration, effect of regime was not significant. Feed conver- sion efficiency was inversely related to growth, and was in the range of 2-15% from high to low rations. Results of this experi- ment indicate that the optimal feeding ration for oysters on a qualitatively suitable diet will lie within the range of 2-5% of live weight in dry matter per day. and that providing the daily ration in multiple daily feedings becomes more important at lower rations. These findings confirm results obtained previously with bay scal- lops. Argopecten irradians, in that maximal growth is obtained on daily rations between 2 and 5% and multiple daily feedings im- prove growth. Oysters, however, with a maximal conversion effi- 280 Abstracts. February 27-March 1, 1999 Milford Aquaculture Seminar, Milford. Connecticut 6tyciency of less than 20% appear to be less efficient at converting feed to growth than scallops which can achieve a conversion ef- ficiency of 24% with identical nutritional input. FEEDING STUDIES ON JUVENILE TAUTOG, TWO EX- PERIMENTS: WEANING JUVENILE TAUTOG TO AN AR- TIFICIAL DIET AND EFFECTS OF FEEDING FRE- QUENCY ON GROWTH OF JUVENILE TAUTOG. Steve Yankocy, Grace Klein-MacPhee, and Aimee Keller. University of Rhode Island. Graduate School of Oceanography. Narragansett Bay Campus. Narragansett RI 02882-1 197. Focus was placed on reducing costs and labor by enhancing growth through selection of a good commercial diet for juveniles and determining the best feeding schedule. Two feeding experi- ments were conducted on juvenile tautog with the goal of finding an optimum feeding regime. The first experiment dealt with the type of food which would be consumed by the fish. The experi- ment utilized two types of food: Kyowa™ brand dry food and live brine shrimp. Three different feeding regimes were used. One group was fed Kyowa. one live brine shrimp and the third a com- bination of brine shrimp and Kyowa. There were three replicates of each treatment. The fish were weighed and measured prior to the experiment with each tank receiving an equal weight of fish. The experiment lasted two weeks. The fish fed the Kyowa diet were larger than those fed brine shrimp or a combination of both, how- ever the results were not statistically significant. Based on the cost of brine shrimp cysts and the extra effort to hatch and enrich them, the kyowa diet was more economical. The second experiment dealt with feeding frequencies. Feeding schedule has been shown to influence growth patterns or food conversion rates significantly in a number of species. Three groups were set up all of which would receive the same amount of Kyowa brand dry food on different feeding schedules. One group received two feedings a day, the second four feedings a day and the third six feedings a day. The fish were weighed and measured at the start, and there were three replicates of each treatment. Feedings were done by hand and with the help of mechanical feeders. Fish fed four and six times a day were significantly larger than fish fed twice a day, but were not significantly different from each other. ™Trademarks are used to identify products only and are not in- dicative of endorsement by URI. Journal of Shellfish Research. Vol. 18. No. 1. 281-335. 1999. ABSTRACTS OF TECHNICAL PAPERS Presented at the 91st Annual Meeting NATIONAL SHELLFISHERIES ASSOCIATION Halifax, Nova Scotia, Canada April 18-22, 1999 281 National Shellt'isheries Association. Halifax. Nova Scotia, Canada Abstracts, 1999 Annual Meeting. April 18-22, 1999 283 CONTENTS BIOLOGY AND MORPHOLOGY OF SHELLFISH Nicole T. Bruit, Andrew D. Boghen and Jacques Allard The presence of the turbellarian Urastoma cyprinae from different areas of the gills of the eastern oyster Crassostrea virginica 291 Suzanne C. Dufour and Peter G. Bellinger Cilia and mucocytes on the abfrontal surface of bivalve gills 291 Juli M. Harding and Roger Mann Habitat and prey preferences of veined rapa whelks (Rapana venosa) in the Chesapeake Bay: direct and indirect trophic consequences 29 1 Dan C. Marelli and William S. Arnold Dead scallops do tell tales: archaeological bay scallop morphologies and composition of ancient metapopulations 291 Laurie A. McDuffee, T. Jeffrey Davidson and William P. Ireland Breaking strength and failure energy of M. edulis shells 292 Suzane M. C. Schreiber, Susan A. Krull, Steven H. Jury and Winsor H. Watson, III Effects of temperature on the heart and ventilation rates of the American lobster (Homarus americanus) 292 Grant D. Stentiford, Douglas M. Neil and Graham H. Coombs Haemolymph free amino acids in the Norway lobster (Nephrops norvegicus L. ) and changes associated with infection by the dinoflagellate parasite Hematodinium 292 BIVALVE GENETICS AND MOLECULAR BIOLOGY Standish K. Allen, Jr., Aimee Howe, Tom Gallivan, Xiining Guo and Greg DeBrosse Genotype and environmental variation in reversion of triploid Crassostrea gigas to the heteroploid mosaics state 293 Whitney Chandler, Aimee Howe and Standish K. Allen, Jr. Mosaicism of somatic and gametic tissues in Crassostrea gigas and C. ariakensis 293 Tachih Cheng, John T. Buchanan, Jerome F. La Peyre, Terrence R. Tiersch and Richard K. Cooper Optimization of reverse transcription polymerase chain reaction (RT-PCR) for use with the eastern oyster Crassostrea virginica 293 Maureen K. Krause Molecular evolution of the Gpi locus in bay scallops. Argopecten irradians 294 Bruno Myrand, Rejean Tremblay and Jean-Marie Sevigny Impact of culture practices on the heterozygosity of suspension-cultured blue mussels 294 Huiping Yang, Xiining Guo and Fusui Zhang Tetraploid zhikong scallop ( Chlamys farreri) produced by inhibiting polar body I 294 COLD WATER AQUACULTURE HEALTH (Invited Session: Bruce Barber and Sharon McGladdery) Bassem Allam, Kathryn A. Ashton-Alcox and Susan E. Ford Resistance to brown ring disease in clams: potential cellular mechanisms 294 Gregory S. Bacon, Sharon E. McGladdery and Bruce A. MacDonald Quahog parasite X ("QPX") of hard-shell clams, Mercenaria mercenaria and M. mercenaria var. notata in Atlantic Canada: observations from wild and cultured clams 295 Bruce J. Barber and Gregory S. Bacon Geographic distribution of gonadal neoplasms in softshell clams, Mya arenaria, from Maine and Atlantic Canada 295 Kathy J. Boettcher, J. T. Singer and Bruce J. Barber A novel species of alpha-proteobacterium is associated with signs of juvenile oyster disease (JOD) in Crassostrea virginica 295 Andrew D. Boghen, Nicole T. Brun and Erick Bataller The association between the turbellarian Urastoma cyprinae and the eastern oyster Crassostrea virginica 296 Susan M. Bower and Gaiy R. Meyer Effect of cold water on limiting or exacerbating some oyster diseases 296 Ryan B. Carnegie and Bruce J. Barber Impact of Bonamia ostreae on cultured Ostrea edulis at two sites on the Damariscotta River. Maine 296 T. Jeffrey Davidson, Claude Morris and David Groman Mycotic periostracal sloughing 297 284 Abstracts, 1999 Annual Meeting. April 18-22, 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada Carolyn S. Friedman, Gary N. Cherr, James S. Clegg, A. H. Hamdoun, J. L. Jacobsen, Susan A. Jackson and K. R. Uhlinger Investigation of the stress response, summer mortality and disease resistance of oysters, Crassostrea spp 297 Sharon E. McGladdery, Mary F. Stephenson and Fiona McArthur Prosorhynchus squamatus (Digenea: Platyhelminthes) infection of blue mussels. Mytilus edulis, in Atlantic Canada — 297 Kelly Moret, Cyr Couturier, G. Jay Parsons and Kate Williams Monitoring shellfish health in Newfoundland: a preventative approach 297 Christine Paillard, Bassem Allam, Radouane Oubella and Susan E. Ford Temperature effects on brown ring disease susceptibility and defense-related activities in the manila clam. Ruditapes philippinarum 298 FEEDING PHYSIOLOGY AND ECOLOGY OF BIVALVES (Invited Session: Bruce A. MacDonald) Shirley M. Baker, Jeffrey S. Levinton and J. Evan Ward Ctenidia as the site of particle selection in bivalves: a comparison between simple and complex ctenidial systems 298 Peter G. Beninger and Anne Yeniot The end of the particle processing line: mantle pseudofaeces rejection mechanisms in suspension-feeding bivalves 298 Martha G. S. Brillant and Bruce A. MacDonald Challenges of examining postingestive selection in bivalves 299 Peter J. Cranford Seasonal variation in food utilization by sea scallops and blue mussels 299 Jon Grant and Michael Nickerson Particle aggregates in seston and their role in bivalve particle selection 300 Melissa Mooney, G. Jay Parsons and Cyr Couturier A comparison of feeding physiology in different sizes of cultured and wild Mytilus edulis and M. trossulus 300 Carter R. Newell The effects of current speed and particle concentration on mussel (Mytilus edulis) filtration rate: a recirculating flume study 300 Roger I. E. Newell, JeffC. Cornwell. Mike Owens and Jon Tuttle Role of oysters in maintaining estuarine water quality 300 Anne Yeniot and Peter G. Beninger Composite cilia: description of a new type of cilium used in particle processing in bivalves 301 J. Evan Ward. Jeffrey S. Levinton, Sandra E. Shumway, and Lisa Milke Mediation of feeding and selection by secondary metabolites of detrital particles 301 GROWTH AND CULTURE OF SHELLFISH George R. Abbe and Brian W. Albright Effect of fishing pressure on size of adult male blue crabs in Maryland: Calvert cliffs and the Patuxent river — 1998... 301 John W. Brake, Jeffrey Davidson and Jonathan Davis Triploid production of Mytilus edulis in Prince Edward Island — an industrial initiative 302 V. Monica Bricelj, Scott MacQuarrie and Roxanna M. Smolowitz Differential effects of two isolates of Aureococcus anophagefferens. in unialgal and mixed suspensions, on feeding and growth of bivalves 302 Gustavo W. Calvo, Mark W. Luckenbach and Eugene M. Burreson Evaluating the performance of non-native oyster species in Virginia 303 Christopher V. Davis Juvenile growth of cage-reared Stimpson's surfclams (Maaromeris polynyma) in Maine. USA 303 Nils T. Hagen Survival and growth of juvenile green sea urchins on different macroalgal settlement substrates 303 Eddy J. Kennedy, Shawn M. C. Robinson and G. Jay Parsons Increasing the somatic growth rate of juvenile green sea urchins (Strongylocentrotus droebachiensis) using prepared diets 304 Dorothy L. Leonard Shellfish restoration: have we been successful? 304 National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22, 1999 285 Clyde L. Mackenzie, Jr. Effects of sea lettuce. Ulva lactuca, mats on abundances of softshell clams. Mya arenaria, and associated invertebrates in New Jersey 304 Gina McNeil and Cyr Couturier Spatio-temporal variation in seston flux, growth and production of the blue mussel. Mytilus edulis, held in suspended culture, in a subarctic environment 304 Linda E. Waite, Thomas Landry and Jeff Davidson The effect of location and time of year on mussel productivity in an aquaculture estuary 305 William C. Walton, Gregory M. Ruiz, and Bethany A. Starr Mitigating predation by the European green crab, Carcinus maenas, upon publicly maricultured quahogs, Mercenaria mercenaria 305 LOBSTER ECOLOGY AND FISHERIES (Invited Session: M. John Tremblav) Michel Comeau, Marc iMiiteigne, Guy Rohichaud and Fernand Savoie Lobster (Homarus americanus) movement in the southern Gulf of St. Lawrence 305 Gareth C. Harding and A. J. Fraser Development of a lipid condition index in lobsters (Homarus americanus) and its application in the interpretation of larval distribution in close proximity to Georges Bank, Gulf of Maine 306 Steve Jury, W. Hunt Howell and Winsor Watson Behavioral thermoregulation and its effects on the movement of lobsters in the field 306 Marc Lanteigne Lobster {Homarus americanus) commercial catch composition fluctuations based on a tight temporal and geographical sea sampling program 306 R. J. Miller and K. F. Drinkwater Egg per recruit as a management target for American lobster fisheries 307 Robert W. Rangeley and Peter Lawton Spatial scaling of habitat distributions in the American lobster 307 M. John Tremblay, R. Duggan, Ron O'Dor, C. Curtis, D. Webber and Y. Andrade Daily movements of lobsters from ultrasonic tracking 307 MODELLING SHELLFISH ECOSYSTEMS (Invited Session: Eileen E. Hofmann) E. A. Bochenek, E. N. Powell, E. Hofmann and J. Klinck A physiologically-based model of the grow th and development of Crassostrea gigas larvae 307 Michael Dowd, Renate Meyer and W. Carlisle Thacker A bayesian approach to shellfish ecosystem modelling 308 Jean-Francois Dumais, Xavier Boespflug, Dominique Baudinet and Marcel Frechette Effect of spawner density and distribution on fertilization success in the sea scallop. Placopecten magellanicus Gmelin 308 Jon Grant and Cedric Bacher Modeling resuspension and its effect on bivalve food supplies 308 Kyung-Hoon Hyun, Ig-Chan Pang, Kwang-Sik Choi, Eric N. Powell, John M. Klinck and Eileen E. Hofmann Modelling population dynamics of the pacific oyster. Crassostrea gigas in Korea 309 Roger Mann, Juli Harding and Stephanie L. Haywood Rapana venosa in the Chesapeake Bay: current status and prospects for range extension based on salinity tolerance of early life history stages 309 Mark B. Meyers, Dominic M. Di Toro and James J. Fitzpatrick Linking water quality and living resources: a coupled suspension feeder-eutrophication model 309 Eric N. Powell, Susan E. Ford, Eileen E. Hofmann and John M. Klinck Modeling the MSX parasite in eastern oyster (Crassostrea virginica) populations: model development. implementation and verification 309 Thomas M. Soniat, Enrique V. Kortright and Sammy M. Ray A simple model for estimating time to critical levels of Perkinsus marinus in eastern oysters, Crassostrea virginica ... 310 286 Abstracts. 1999 Annual Meeting. April 18-22, 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada Melissa Southworth and Roger Mann Quantitative aspects of oyster reef broodstock enhancement in the Great Wicomico River. Virginia 310 REPRODUCTION & RECRUITMENT Linda A. MacLean, Neil G. MacNair, T. Jeffrey Davidson, and Gerald G. Johnson Two year comparison of spawning patterns in soft-shell clams (Afya arenaria) 310 Nature A. McGinn, Michael P. Lesser and Charles W. Walker The influence of estradiol on vitellogenesis in the green sea urchin. Strongylocentrotus droebachiensis 311 Miranda Pryor, G. Jay Parsons and Cyr Couturier Temporal patterns of larval and post-set distributions of the blue mussel (Mytilus edulis I M. trossulus) and the starfish (Asterias vulgaris) on Newfoundland mussel culture sites 311 Stephen T. Tettelbach, Roxanna Smolowitz, Christopher F. Smith, Kim Tetrault and Sandra Dumais Evidence for fall spawning of northern bay scallops. Argopecten irradians irradians (Lamarck. 1819). in New York .. 311 Tracy Vassiliev, William Congleton, Brian Beat and Stephen Fegley An investigation of Mya arenaria (soft-shell clam) recruitment in Maine 311 SCALLOP FISHERIES: ECOLOGY AND APPLIED BIOLOGY (Invited Session: Ellen Kenchington) Shelley L. Armsworthy, Peter J. Cranford and Kenneth Lee Effects of a new bitumen fuel source on the growth and energetics of sea scallops 312 Peter J. Cranford, Donald C. Gordon, Jr., Charles G. Hannah, John W. Loder, Timothy G. Milligan and Dwight K. Muschenheim Modelling potential effects of drilling wastes on George"s Bank scallop stocks 312 Leslie-Anne Davidson and Yves Poussart Management advice of giant scallop. Placopecten magellanicus, based on gonad maturation 312 M. Edwin DeMont Jet-propelled swimming in scallops 313 William D. DuPaul, James E. Kirkley and David B. Rudders Scallop gear selectivity and scallop biology: a mismatch in resource management 313 Ellen Kenchington, Carolyn J. Bird and Elefterios Zouros Genetic variation in Placopecten magellanicus w ith implications for fisheries management 313 Lorelei A. Levy, G. Jay Parsons and Patrick Dabinett Effect of deployment date on sea scallop growth and survival 313 Joan L. Manual Retention of scallop veligers and consequences for stock enchancement programs, aquaculture and stock management 314 Shawn M. C. Robinson, James D. Martin, Ross A. Chandler and G. Jay Parsons An examination of the linkage between the early life history processes of the sea scallop and local hydrographic characteristics 314 Dale Roddick, Ellen Kenchington, Stephen Smith and Jon Grant The use of RNA/DNA ratios as an index of health for the sea scallop {Placopecten magellanicus) 314 Kevin Stokesbury Physical and biological variables influencing the spatial distribution of the giant scallop Placopecten magellanicus — 315 Mitchell L. Tarnowski and Mark L. Homer Re-introducing the bay scallop Argopecten irradians into Chincoteague Bay. MD 315 Ami E. Wilbur, William S. Arnold and Theresa M. Bert Evaluating bay scallop stock enhancement efforts with molecular genetic markers 315 SHELLFISH BIOCHEMISTRY Yank Marty, Philippe Soudant, Sebastien Perrotte, Jeanne Moal, Jean-Francois Samain and Jacques Dussauze Identification and occurrence of a novel fatty acid in pectinids 316 Jean-Francois Samain, Philippe Soudant, Yanic Marty and Jeanne Moal Fatty acids for reproduction and larval development in two bivalves molluscs: polar lipid approach 316 National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting, April 18-22. 1999 287 Philippe Soudant, Karla Van Ryckeghem, Jeanne Modi, Yanic Marty, Jean-Francois Samain and Patrick Sorgeloos Comparison of essential fatty acid accumulation between a reproductive cycle in nature and a hatchery conditioning of Crassostrea gigas 316 SHELLFISH DISEASE Brian IV. Albright and George R. Abbe Recent trends in infection of the eastern oyster Crassostrea virginica by the parasite Perkinsus marinus in the Patuxent River estuary 317 Gwynne D. Brown, Shabon Kotob, and Mohamed Faisal Diversity among Perkinsus marinus isolates from the Chesapeake Bay 317 David Bushek, A. J. Erskine, Richard F. Dame, Loren D. Coen and Nancy Hadley Transmission of Perkinsus marinus to intertidal oysters 317 Fu-Lin E. Chit, Philippe Soudant, Yongqin Huang, Aswani K. Volety and Georgeta Constantin Uptake, distribution, and byconversion of fluorescent lipid analogs in the oyster protozoan parasite. Perkinsus marinus 318 Cathleen A. Coss, Jose A. F. Robledo, Gerardo R. Vasta and Gregory M. Ruiz Identification of a new Perkinsus species isolated from Macoma balthica by characterization of the ribosomal RNA locus. Evidence of its presence, simultaneous with P. marinus, in Crassostrea virginica, Macoma mitchelli and Mercenaria mercenaria 318 William S. Fisher and Benjamin H. Sherman Integrated monitoring of marine disease and mortality 318 James D. Moore, Thea T. Robbins and Carolyn S. Friedman Withering Syndrome in farmed red abalone. Haliotis rufescens 319 Jacques L. Oliver, Mohamed Faisal and Stephen L. Kaattari Plasma of Crassostrea spp. possess a low molecular weight inhibitor of Perkinsus marinus protease 319 Soledad Penna, Richard A. French, John Yolk, John Karolus, Irike Sunila and Roxanna Smolowitz Diagnostic screening of oyster pathogens: preliminary field trials of multiplex PCR 319 Kimberly S. Reece, David Bushek and Karen L. Hudson Analysis of the geographic distribution of Perkinsus marinus strains 320 Adel A. Shaheen Long-term survival of Perkinsus marinus cells outside its host 320 SHELLFISH IMMUNOLOGY: ADAPTATION AND MODULATION (Invited Session: Robert Anderson and Cal Baier-Anderson) Michel Auffret and Radouane Oubella Xenobiotic-induced immunotoxicity in the pacific oyster. Crassostrea gigas: field and laboratory experiments 320 Lisa H. Bramble and Robert S. Anderson Effect of the NADPH oxidase inhibitor diphenyleneiodonium on the bactericidal activity of Crassostrea virginica hemocytes 321 Louis E. Burnett, John Boyd, Chris Milardo, Tina Mikulski, Libby Wilson and Karen Burnett The effects of hypoxia and hypercapnia on cellular defenses of oysters, shrimp, and fish 321 Fu-Lin E. Chu Effects of temperature, salinity, and environmental pollutants on cellular and humoral responses in oysters (Crassostrea virginica) 321 Mohamed Faisal The role of protease-antiprotease interactions in Perkinsus marinus infection in Crassostrea spp 322 Carolyn S. Friedman, Thea Robbins, Jacqueline L. Jacobsen and Jeffrey D. Shields Examination of the cellular immune response of black abalone. Haliotis cracherodii, with and without w ithering syndrome 322 Jerome F. La Peyre and Aswani K. Volety Modulation of eastern oyster neurocyte activities by Perkinsus marinus extracellular proteins 322 Leah M. Oliver, Aswani K. Volety and William S. Fisher Chemical effects on oyster (Crassostrea virginica) hemocyte microbicidal activity 323 288 Abstracts, 1999 Annual Meeting, April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada Kennedy T. Paynter Phagosomal mechanisms in eastern oyster ( Crassostrea virginica ) blood cells 323 Aswani K. Volety, James T. Winstead and William S. Fisher Influence of seasonal factors on oyster hemocyte killing of Vibrio parahemolyticus 323 SHELLFISH-MICROBIAL INTERACTIONS: ECOLOGICAL AND HUMAN HEALTH PERSPECTIVES (Invited Session: Fred Genthner and Aswani Volety) John T. Buchanan, Ta Chi Cheng, Jerome F. La Peyre, Richard K. Cooper and Terrence R. Tiersch In vivo transfection of adult oysters 324 Fred J. Genthner, Leah M. Oliver, William S. Fisher and Aswani K. Volety Factors influencing in vitro killing of bacteria by hemocytes of the eastern oyster {Crassostrea virginica) 324 Jerome F. La Peyre, Richard A. Cooper, John E. Supan and Aswani K. Volety Total bacteria and Vibrio vulnificus load in diploid and triploid eastern oysters in Louisiana 324 Paul G. Olin and Gregg iMiiglois Regulation and management of water quality to preserve shellfish harvesting and human health in Tomales Bay. California 324 Jeffrey D. Shields and Christopher M. Squyars Hematology of blue crabs. Callinectes sapidus, infected with the parasitic dinoflagellate Hematodinium perezi 325 Kim M. Stowell, Stephen D. Torosian and Aaron B. Margolin Detection of protozoa pathogens in the eastern oyster taken from the Great Bay estuary 325 Ben D. Tall, Maya Crosby, Deanna Prince, James Becker, Gaskov Clerge, Donald Lightner, Leone Mohney, Manashi Dey, Farukh Khambaty, Keith Lampel, Jeffrey \\. Bier, Brodrick E. Eribo and Robert Bayer Vibrio fluvialis implicated in recent outbreaks among American lobsters 325 Aswani K. Volety, Fred J. Genthner, William S. Fisher, Susan A. McCarthy and Kirk Wiles Differential effects of oyster {Crassostrea virginica) defenses on clinical and environmental isolates of Vibrio parahemolyticus 326 POSTER SESSION Amy E. Beaven and Kennedy T. Paynter Bafilomycin A, inhibits acidification of granular and agranular oyster hemocyte phagosomes 326 Jodi Brewster, David Bushek and Richard F. Dame An ecosystem model of Perkinsus marinus 326 Shelley Burton, Allan MacKenzie, T. Jeffrey Davidson and Audrey Eraser Evaluation of a glucose oxidase/peroxidase method for indirect measurement of glycogen content in oysters ( Crassostrea virginica) 327 Debbie Cayer, Marli Mac Neil and Andrew G. Bagnall Tunicate fouling in Nova Scotia aquaculture: a new development 327 Whitney Chandler, Aimee Howe and Standish K. Allen, Jr. Use of flow cytometry and histology to assess gametogenesis in triploid Crassostrea ariakensis 327 Gregory M. Coates, Richard K. Cooper and Jerome F. La Peyre Improvement of the whole-oyster procedure for enumerating Perkinsus marinus in oyster tissues 328 Rebecca C. Ellin and David Bushek Potential use of Ray's Fluid Thioglycollate Medium to detect and quantify Perkinsus marinus in environmental water samples 328 Ehab Elsayed and Mohamed Faisal Correlation between the level of protease inhibitors and intensity of Perkinsus marinus infection in eastern oyster ( Crassostrea virginica) 328 Scott C. Feindel, Ray Thompson, Pat Dabinett and Christopher Parrish Effects of broodstock and larval diets on lipid and fatty acid composition of sea scallop {Placopecten magellanicus) eggs and larvae in relation to culture optimization 329 Zaul Garcia-Esquivel, Marco A. Gonzalez-Gomez and Dahen L. Gomez-Togo Growth, mortality and biochemical content of the Pacific oyster. Crassostrea gigas, during spat-adult development — 329 National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts. 1999 Annual Meeting. April 18-22. 1999 Eileen E. Hofmann, John M. Klinck, Susan E. Ford and Eric N. Powell Disease dynamics: modeling the effect of climate change on oyster disease 329 Barbara S. Homey, Allan L. Mackenzie, Richard J. Cawthorn, Claude C. Morris, Ixtrry R. Hammett and Robert MacMillan Reference ranges for chemical and cellular constituents of hemolymph from "healthy" lobsters ( Homarus americanus) 329 William P. Ireland, T. Jeffrey Davidson and iMurie McDuffee Prediction of blue mussel (Mytilus edulis) failure load 330 Stephen Jones, Margo Chase, John Sowles, Peter Hennigar and Peter Wells Spatial trends for toxic contaminants in Mytilus edulis from the Gulf of Maine 330 Timothy Roles and Rennedy T. Paynter Oyster restoration in Maryland: measuring progress and productivity 330 Ren Leonard III, Marta Gomez-Chiarri and Arthur Ganz Detecting the presence of Perkinsus marinus in the eastern oyster, Crassostrea virginica, in Rhode Island waters 331 Neil G. MacNair and Matt Smith Investigations into treatments to control fouling organisms affecting oyster production 331 Antonio Mazzola, Tiziana La Rosa, Benedetto Savona and Gianluca Sara Seston dynamics and food availability in a mussel system (Gulf of Gaeta, Southern Tyrrhenian Sea. Italy) 331 Lisa Milke, J. Evan Ward and Sandra E. Shumway Effects of food quality on the particle handling time in bivalves 331 J. Moal, C. Seguineau, J. F. Sainain, P. Soudant, M. Cansell, J. R. LeCoz, H. Migaud, M. Sanies, B. Ponce and C. iMiigdon How to provide essential nutriments to bivalves in hatchery 332 Madeleine Nadeau, Georges Cliche and Denyse Hebert Experimental dredging of starfishes and crabs before commercial seeding of sea scallops in Magdalen Islands ( Quebec. Canada) 332 Fernando Ribeiro, Fernando Simoes, iMiiren Swenarchuk Mussel culture potential in southern Mozambique 332 Michael A. Rice, April Valliere, Mark Gibson and Arthur Ganz Eutrophication control by bivalves: population filtration, sedimentation and nutrient removal through secondary production 333 Jose A. F. Robledo, Cathleen A, Coss and Gerardo R. Vasta Development of a PCR-based diagnostic assay for a novel Perkinus species isolated from Macoma baltica 333 Gianluca Sara, Chiara Romano and Antonio Mazzola The new western Mediterranean entry Brachidontes pharaonis (Fischer p.. 1870) (Bivalvia. Mytilidae): changes in filtration rate under varying natural food conditions 333 Jeffrey D. Shields Partial culture and cryopreservation of the parasitic dinoflagellate Hematodinium perezi from the blue crab 334 Grant D. Stentiford, Douglas M. Neil, and R. J. A. Atkinson Infection by the dinoflagellate parasite Hematodinium in the Norway lobster (Nephrops norvegicus L.) on the west coast of Scotland. United Kingdom 334 Nancy A. Stokes, Brenda Sandy Flares Rraus, Eugene M. Burreson, Rathryn A. Ashton-Alcox and Susan E. Ford Searching for the putative MSX intermediate host using molecular diagnostics 334 Stewart Tweed and Mining Guo Preliminary evaluation of triploid eastern oysters, Crassostrea virginica. on a mid-Atlantic oyster farm 335 Gregory Ziegler and Rennedy T. Paynter Proteolytic activity from blood cells of the eastern oyster. Crassostrea virginica 335 National Shellfisheries Association, Halifax. Nova Scotia, Canada Abstracts. 1999 Annual Meeting, April 18-22. 1999 291 BIOLOGY AND MORPHOLOGY OF SHELLFISH THE PRESENCE OF THE TURBELLARIAN URASTOMA CYPRINAE FROM DIFFERENT AREAS OF THE GILLS OF THE EASTERN OYSTER CRASSOSTREA VIRGINICA. Nicole T. Brun and Andrew D. Boghen, Department of Biology. Universite de Moncton. Moncton, NB. Canada E1A 3E9; Jacques Allard, Department of Mathematics and Statistics, Universite de Moncton. Moncton, NB, Canada E1A 3E9. Urastoma cyprinae is a turbellarian that occurs on the gills of various bivalve species. Recent investigations indicate that U. cyp- rinae can induce pathology to the gill tissue of mussels. In Atlantic Canada, this worm has been reported from the gills of the Eastern Oyster Crassostrea virginica. Past work has demonstrated that U. cyprinae is attracted to oyster mucus on which it may be actively feeding. The present study focuses on the distribution of U. cyp- rinae in oysters in relation to certain properties of mucus associ- ated with different areas of the gills. The worms occur in higher numbers along the basal food tract compared to the medial and ventral regions of the gills. U. cyprinae 's preference for this site may be explained by a number of factors, including the fluidity and accessibility of the mucus, as well as the protection provided com- pared to more exposed parts of the gills when the oysters are actively feeding. HABITAT AND PREY PREFERENCES OF VEINED RAPA WHELKS {RAPANA VENOSA) IN THE CHESAPEAKE BAY: DIRECT AND INDIRECT TROPHIC CONSE- QUENCES. Juli M. Harding and Roger Mann, Virginia Insti- tute of Marine Science, Gloucester Point, VA 23062. The recent discovery of Veined Rapa Whelks (Rapana venosa) in the lower Chesapeake Bay has ecological consequences beyond the obvious potential for predation on commercially valuable shellfish prey species (e.g., Crassostrea virginica, Mercenaria mercenaria). In the Black Sea and in their native Sea of Japan, Rapana have been reported primarily from hard bottom habitats. Adult Chesapeake Bay Rapana have been collected from both hard and soft bottom habitat. Laboratory observations indicate that adult Rapana prefer sand bottom and will burrow almost completely into the sand at water temperatures >20 C (i.e.. not overwintering behavior). Burrowing behavior by these large apex predators ex- pands the potential suite of vulnerable prey items to include in- faunal shellfish (e.g., Mya arenaria, Ensis directus, Cyrtopleura costata). The presence of large (>100 mm) empty Rapana shells in Chesapeake Bay may enhance growth of the local hermit crab (Clibanarius vittatus). Recent collections of Clibanarius vittatus from the Hampton Roads area indicate they use empty Rapana shells as shelters and are reaching previously unrecorded sizes. The implications of abnormally large crustacean scavengers on Chesapeake Bay benthic epifauna (e.g., oyster spat) are discussed. CILIA AND MUCOCYTES ON THE ABFRONTAL SUR- FACE OF BIVALVE GILLS. Suzanne C. Dufour, Scripps In- stitution of Oceanography. University of California. San Diego, La Jolla, CA 92093-0202; Peter G. Beninger, Laboratoire de biolo- gie marine, Universite de Nantes, Nantes cedex 3. France. The lack of fundamental data on the abfrontal surface of bi- valve gills has prompted a comparative study of cilia and muco- cytes on this surface. These features have been studied by scanning electron microscopy and histology on eight species of bivalves, representing seven families and the four major gill types (Mytihts edulis, Modiolus modiolus. Area zebra, Placopecten magellanicus, Crassostrea virginica, Spisula solidissima, Mya arenaria and Mer- cenaria mercenaria). Inter-species variations were found: gradi- ents in the numbers and diversity of cilia and mucocytes were observed for each gill type. These results seem to indicate that the abfrontal surface had a primitive role in mucociliary cleaning (prior to filament folding), and that the cilia and mucocytes here observed are vestigial. The loss of this primitive function brought forth two possibilities: 1. selective pressures led to the reduction in numbers and types of abfrontal mucocytes and cilia; 2. the abfron- tal cilia anoVor mucocytes were retained as they assumed new functions. In general, the degree of loss of abfrontal cilia and mucocytes follows the degree of evolution of the gill: eulamelli- branchs have less abfrontal cilia and mucocytes than homorhabdic filibranchs. DEAD SCALLOPS DO TELL TALES: ARCHAEOLOGI- CAL BAY SCALLOP MORPHOLOGIES AND COMPOSI- TION OF ANCIENT METAPOPULATIONS. Dan C. Marelli and William S. Arnold, Florida Department of Environmental Protection. Florida Marine Research Institute. 100 8th Avenue SE. St. Petersburg, FL 33701-5095. Shell middens are conspicuous features of some sites along the Florida Gulf coast, representing thousands of years of human ex- ploitation of marine resources. The bay scallop (Argopecten irra- dians) was extensively harvested by prehistoric Floridians, but many populations of Floridan bay scallops have declined precipi- tously during the past 4 decades probably because of the effects of overharvest, habitat alteration, water quality degradation, and toxic- algal blooms. Traditional metapopulations may lose stability in Florida because of the collapse of large local populations that traditionally acted as sources for less temporally consistent local populations. Unfortunately, because of our late start on the prob- lem, much of the useful information from local populations that would allow us to reconstruct the traditional metapopulations may already be lost. Some backcasting of past populations structure is possible using biochemical techniques, but the only uncorrupted data we have are the scallop shells left behind by the original 292 Abstracts. 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada Floridians. Examining morphologies of archaeological scallops, comparing these morphologies with those of modern scallops, and referencing allozyme data will allow us to speculate on what an- cient metapopulations may have looked like and how we can use this information in attempting to enhance or restore local popula- tions of bay scallops in Florida. BREAKING STRENGTH AND FAILURE ENERGY OF M. EDULIS SHELLS. Laurie A. McDuffee and T. Jeffrey David- son, Department of Health Management, and William P. Ireland, Department of Anatomy and Physiology. Atlantic Veterinary Col- lege. University of Prince Edward Island. Charlottetown. PEI C1A4P3. Mussel processors on Prince Edward Island (PEI) have esti- mated that approximately 3% of marketable mussels entering the processing line have to be culled due to cracked or broken shells. The mussel processing sector also observed that mussels from certain estuaries have a greater percentage of broken/cracked shells during processing compared to other areas. With PEFs mus- sel industry worth $24 million a year, this problem represents an annual loss to the industry of $720,000. We hypothesized that one reason for differences in shell crack- ing may be differences in mechanical properties of mussel shells between estuaries. Because mechanical properties are influenced by the geometry, we also hypothesized that measurements of shape and size would effect mussel breaking strength. Mussels (sp. M epulis) were obtained from PEI Department of Fisheries and Tourism who collected mussel samples from seven mussel growing areas of PEI. Mussels were cleaned, measured with an electronic caliper, and weighed. Measurements included length, width, height, weight, and thickness of the shells at the highest point. Shells were separated and meats removed. The half shells underwent biomechanical testing in a materials testing sys- tem under a single compressive load at a rate of 0.07 cm/sec. Mechanical properties and failure configurations were determined. Mechanical variables of interest included the maximum failure load, (also considered breaking strength), failure energy, and stiff- ness. These variables were obtained from the load deformation curves generated for each specimen. Maximum failure, or breaking strength, was defined as the highest load that the shell could with- stand before breaking. Maximum failure energy was determined from the area under the load deformation curve to the maximum failure point. Stiffness was calculated as the slope of the linear region of the load deformation curve. Mechanical properties be- tween estuaries were compared statistically. There were significant differences in breaking strength and failure energy between the various mussel growing areas of PEI. For instance, it required almost twice the force to break a shell from March Water than it did from the Brudenell River. EFFECTS OF TEMPERATURE ON THE HEART AND VENTILATION RATES OF THE AMERICAN LOBSTER (HOMARUS AMERICANUS). S. M. C. Schreiber, S. Krull, S. H. Jury, and W. H. Watson III, Dept. of Zoology, University of New Hampshire. Durham. NH 03824. The behavior of lobsters is strongly influenced by the tempera- ture of their surroundings. However, the underlying physiological responses to temperature are not well understood. In this study, we examined the effects of a series of 5°C temperature increases on the heart and ventilation rates of intact animals and isolated hearts. Heart rates of intact animals increased continuously over the range of 5-25°C. However, this increase was non-linear, with Ql(1 values decreasing from near 2.5 at the low end to near 1.5 at the high end of the range. Ventilation rates were strongly correlated with heart rates. In contrast, the isolated hearts showed very little response to temperature. Qlo values for isolated hearts were markedly lower than those in intact animals for all temperature ranges. Their maxi- mum rates were exhibited between 14-19°C. At higher tempera- tures, beat rate decreased dramatically and, in many instances, the heart failed. Basal heart rates of isolated hearts were significantly lower than those of intact animals at all but the lowest tempera- tures, suggesting that there is modulation occurring in the whole animal that is absent in the exposed preparation. Currently, we are carrying out experiments to determine if the effects of temperature on heart activity are mediated by the cardioregulatory nerves or by neurohormones. HAEMOLYMPH FREE AMINO ACIDS IN THE NORWAY LOBSTER [NEPHROPS NORVEGICUS L.) AND CHANGES ASSOCIATED WITH INFECTION BY THE DINOFLAGEL- LATE PARASITE HEMATODINIUM. Grant D. Stentiford and Douglas M. Neil, Division of Environmental & Evolutionary Biology, Institute of Biomedical & Life Sciences, University of Glasgow. G12 8QQ. Scotland. UK.; Graham H. Coombs, Divi- sion of Infection and Immunity. Institute of Biomedical & Life Sciences, University of Glasgow. G12 8QQ. Scotland. UK. Changes in plasma free amino acid (FAA) concentrations have been used to assess various stresses, including disease, in a number of animal groups. However, with the exception of osmotic stress studies, this measure has received relatively little attention in Crus- tacea, despite their large tissue FAA pools. Norway Lobsters (Nephrops norvegicus L.) on the west coast of Scotland undergo a seasonal infection by a parasitic dinoflagellate (Hematodinium sp.) and this systemic disease causes gross histopathological changes in host tissue, eventually leading to death. We therefore set up a study to investigate the profile of FAA compounds in normal and He- matodinium-mfected Nephrops. Results show an almost doubling of FAA compounds during severe infection, with most of this increase being due to the sulfonic amino acid taurine. In heavily infected animals, plasma taurine increased both absolutely (from 0.22 u-mol • ml"' to 2.56 pimol • ml-1) and relatively (from 5.8% to National Shellfisheries Association, Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting, April 18-22. 1999 293 40% of total identified FAA) when compared with healthy ani- mals, and also made the dominant contribution to the observed rise in the non-essential/essential plasma FAA ratio. We also found that plasma taurine increased proportionally to infection stage, as estimated by parasite accumulation in the pleopods of Nephrops. Increased plasma taurine may originate from host tissue or from the parasite and this is currently under investigation. Such elevated taurine levels may have implications in predator attraction to in- fected Nephrops and on the general functioning of the host nervous system. We believe that plasma taurine concentration has the po- tential for use as an indicator of disease and possibly other stresses in captive and wild Crustacea. BIVALVE GENETICS AND MOLECULAR BIOLOGY adductor muscle, digestive gland and hemolymph) as well as the frequency of haploid gamete formation was examined by flow cytometry. Gametic tissue was examined via histology. Of the six tissue types examined, hemolymph contained the highest propor- tion of diploid cells and gill and gametic tissue contained the lowest, in both species. No haploid gametes were found. In C. ariakensis, 66% were male. 17% female, and 17% hermaphrodite. In C. gigas 25% were male, 62% females and 13% hermaphrodite. Gametogenesis in mosaic oysters was deemed nearly identical to that in triploids. Follicles were few and incompletely formed and gametes were in various stages of maturation within the same animal. We conclude that hemolymph is a more sensitive indicator of mosaicism than other tissue types and should be used for routine certification. Also, there is no evidence of normal (diploid) ga- metic activity in two-year-old male mosaics, although the eventual formation of haploid cells cannot be ruled out. GENOTYPE AND ENVIRONMENTAL VARIATION IN RE- VERSION OF TRIPLOID CRASSOSTREA GIGAS TO THE HETEROPLOID MOSAICS STATE. Standish K. Allen, Jr., Aimee Howe, and Tom Gallivan, Aquaculture Genetics and Breeding Technology Center, Virginia Institute of Marine Science. Gloucester Point. VA 23062; Ximing Guo and Greg DeBrosse, Haskin Shellfish Research Laboratory. Rutgers University, Port Norris, NJ 08349. In 1993, we were surprised by the occurrence of heteroploid mosaics among supposedly ""certified triploid" individuals. At the time, their origin was not clear although reversion — loss of chro- mosomes from the triploid yielding heteroploid mosaicism — was suspected. In 1995 we began an experiment to unequivocally dem- onstrate reversion among triploid Crassostrea gigas. Triploids pro- duced using cytochalasin (CB) treatments and from tetraploid x diploid matings were deployed in a quarantine system adjacent to Delaware Bay and in the Chesapeake Bay. Observations on labeled individuals were carried out for two years. The frequency of re- version in CB induced triploids was about 2-3 times higher than in tetraploid x diploid ones. The frequency of "reversion" also varies between grow out sites. Harsher environments may exacerbate the problem of reversion. Reversion is progressive with more and more diploid cells accumulating with time. The gametes produced in mosaics seldom have haploid cells. Cytogenetic data suggests that chromosome elimination may be caused by unusual chromo- some clumping, as evidenced by mitotic metaphase spreads. MOSAICISM OF SOMATIC AND GAMETIC TISSUES IN CRASSOSTREA GIGAS AND C. ARIAKENSIS. Whitney Chandler, Aimee Howe, and S. K. Allen, Jr., Aquaculture Ge- netics and Breeding Technology Center. Virginia Institute of Ma- rine Science. Gloucester Point. VA 23062. Putative two-year old mosaic Crassostrea ariakensis and C. gigas were conditioned for two months, sacrificed and biopsied. Percentage mosaicism among the tissue types (gill, gonad, heart. OPTIMIZATION OF REVERSE TRANSCRIPTION POLY- MERASE CHAIN REACTION (RT-PCR) FOR USE WITH THE EASTERN OYSTER CRASSOSTREA VIRGINICA. Taehih Cheng, Department of Veterinary Science, John T. Buchanan, Department of Oceanography and Coastal Sciences. Jerome F. La Peyre, Department of Veterinary Science. Ter- rence R. Tiersch. School of Forestry. Wildlife, and Fisheries. Richard K. Cooper, Department of Veterinary, Louisiana State University, Baton Rouge. LA 70803. In order to monitor gene expression and library construction of complement DNA (cDNA) in oysters, RT-PCR condition was op- timized to amplify actin cDNA. To determine the importance of DNase I digestion before reverse transcription, standard PCR and RT-PCR were used to amplify actin gene and products were se- quenced to identify whether introns existed. A specific primer set was designed to amplify a 770 base pair segment of a published actin cDNA sequence. The genomic DNA (gDNA) and total RNA from hemocytes were isolated. The gDNA was used in PCR as templates; total RNA was used for RT-PCR after digestion with RNase-free DNase I and transcribed to single strand cDNA (sscDNA) by reverse transcriptase and a universal poly oligo- deoxythymidylate (pol dT) primer. Single strand cDNA was used as the template. In 2% agarose gel electrophoresis, an -770 base pair band was found in PCR and RT-PCR amplifications. The bands were recovered and sequenced and showed >93% similarity with each other. These results indicate that RT-PCR with the spe- cific primer set not only amplified the expected size actin cDNA. but also an actin-like gDNA of similar size and sequence. This data suggests that there is no large intron in the oyster actin gene and that it was necessary to digest total RNA with RNase-free DNase I to eliminate the gDNA before reverse transcription to avoid generating false positive RT-PCR. 294 Abstracts, 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia, Canada MOLECULAR EVOLUTION OF THE GPI LOCUS IN BAY SCALLOPS, ARGOPECTEN IRRAD1ANS. Maureen K. Krause, Natural Science Division. Southampton College of Long Island University. Southampton. NY I 1968. Previous research demonstrated that allozyme variation at the polymorphic enzyme-coding locus, glucose phosphate isomerase, explains a large and significant proportion of variation in produc- tion-related traits in the bay scallop Argopecten irradians. Addi- tionally, apparent in vivo biochemical functional differences among allozyme genotypes are correlated with higher-order phe- notypic effects. The collective evidence suggests that natural se- lection may be maintaining genetic variation for Gpi, but whether the observed association between Gpi genotype and production- related traits is due to selection at the Gpi locus itself or to loci in linkage disequilibrium is unclear. The hypothesis for selective maintenance of genetic polymorphism can be tested using DNA sequence data and recently developed statistical methods. This paper presents the initial results from a molecular evolutionary study of nucleotide variation at Gpi. We have isolated and char- acterized a large portion of the coding region of bay scallop Gpi. The 917 bp cDNA shows approximately 66% nucleotide homol- ogy with Calanus finmarchicus Gpi. 73% homology to the human gene, and 69% homology with Drosophila melanogaster. Initial analyses of polymorphisms and correspondence to allozyme varia- tion will be presented. IMPACT OF CULTURE PRACTICES ON THE HET- EROZYGOSITY OF SUSPENSION-CULTURED BLUE MUSSELS. Bruno Myrand, Station Technologique Maricole des Iles-de-la-Madeleine. Cap-aux-Meules, Que., GOB 1B0; Rejean Tremblay, Groupe Interuniversitaire de Recherche en Oceanog- raphie du Quebec. Que., G1K 7P4; and Jean-Marie Sevigny, Institut Maurice-Lamontagne. Mont-Joli, Que., G5H 3Z4. The mean heterozygosity of suspension-cultured mussels (Mxtilus edulis L. ) from the Magdalen Islands, southern Gulf of St. Lawrence, is lower than that of the spat used for sleeving. This may have important impacts on the productivity of the local mus- sel culture industry because of the apparent relationship between heterozygosity and some fitness components such as resistance to stress and. probably, to summer mortality. To understand the causes of such decrease in heterozygosity, spat was sampled from collectors in October 1996 and then at various steps of the com- mercial sleeving operations. Graded spat was sleeved at usual (-850 per m) and high density (-1 950 per m) or kept in pearl-nets to impede fall-offs. The only significant decrease in heterozygosity (measured at 7 enzymatic loci) occurred once the sleeves were attached to longlines in the lagoon. The major decrease occurred before early June 1997. In September 1997, the mean heterozy- gosity of sleeved mussels was 2.44 ± 0.14 (usual density) and 2.14 ± 0.21 (high density) while that of the graded spat used for sleev- ing was 3.13 ± 0.20. Heterozygosity of the graded spat in pearl- nets was 2.83 ±0.14 and did not decrease significantly. We hy- pothesize that more heterozygous spat goes out of the sleeves' meshes more rapidly and/or in greater numbers than do more ho- mozygous individuals. If so, they could be more prone to fall-offs. mainly those resulting from turbulence created by heavy winds over the shallow lagoons in fall and spring. TETRAPLOID ZHIKONG SCALLOP [CHLAMYS FAR- RERI) PRODUCED BY INHIBITING POLAR BODY I. Huiping Yang.1 Ximing Guo.1" and Fusui Zhang.1 'institute of Oceanology, Chinese Academy of Sciences. Qingdao. Shandong 266071. China: 2Haskin Shellfish Research Laboratory. Institute of Marine and Coastal Sciences. Rutgers University, 6959 Miller Avenue. Port Norris. NJ 08349. Triploid shellfish are useful for aquaculture because of their sterility, superior growth and improved meat quality. Triploid scal- lops have greatly enlarged adductor muscle (by 73-167% over normal diploids), making them extremely valuable for aquaculture. Although triploids can be produced by blocking polar body II. the ideal way for triploid production is through diploid x tetraploid mating. In this study, we tested tetraploid induction in the zhikong scallop. Chlamys farreri, by inhibiting the release of polar body I (PB1) in newly fertilized eggs. Cytochalasin B (0.75 mg/1) was applied at 7-10 min post-fertilization (PF) and terminated when PB1 was released in about 60% of the untreated eggs. The treat- ment and its control were repeated 12 times using the different pairs of parents. Blocking PB 1 greatly altered chromosome seg- regation and the ploidy of resultant embryos. Variable proportions of aneuploid. triploid. tetraploid and pentaploid larvae were pro- duced among the replicates. Most of the tetraploid. pentaploid and aneuploid larvae died with the first two weeks PF. At three months PF, five tetraploids (or 1.8%) were found in a sample of 271 spat (2—1 mm) from one of the replicates. Although the tetraploid per- centage is low. this result suggests that tetraploid zhikong scallop can survive to at least the juvenile stage. COLD WATER AQUACULTURE HEALTH RESISTANCE TO BROWN RING DISEASE IN CLAMS: POTENTIAL CELLULAR MECHANISMS. Bassem Allam, Kathryn A. Ashton-Alcox, and Susan E. Ford, Haskin Shellfish Research Laboratory. Rutgers University, Port Norris. NJ 08349. Brown Ring Disease (BRD) is a temperature-controlled disease that has affected the clam Ruditapes philippinarum in Western Europe since the late 80s. The etiological agent. Vibrio tapetis. colonizes the periostracal lamina, disturbing the normal calcifica- tion process. In this study, flow cytometry was used to analyze hemocyte parameters in experimentally infected and control clams National Shellfisheries Association. Halifax. Nova Scotia, Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 295 from France (FR) and the US west coast (Oregon). Four weeks following challenge with V. tapetis, BRD prevalence was 52 to 57% in US R. philippinarwn and 93 to 100% in FR specimens, suggesting some resistance in the US clam. Only 21-37% of the native European species, R. decussatus, developed BRD. In the hemolymph and the extrapallial fluid of animals that did not de- velop BRD. total hemocyte counts, the percentage of granulocytes, the percentage of viable cells, and the phagocytic activity against latex beads by the hemocytes were significantly higher in US than in FR R. philippinarwn. In diseased animals, phagocytosis of V. tapetis was elevated in R. decussatus, remained unchanged in US R. philippinarwn, and decreased significantly in FR R. philippi- narwn. Phagocytosis of V. tapetis by hemocytes in extrapallial fluid could be an important defense mechanism in BRD. Our re- sults suggest that the relative resistance of R. decussatus and the US population of R. philippinarwn to BRD could be related to differences in cellular parameters of the defense system and par- ticularly the number and the phagocytic activity of hemocytes. GEOGRAPHIC DISTRIBUTION OF GONADAL NEO- PLASMS IN SOFTSHELL CLAMS, MYA ARENARIA, FROM MAINE AND ATLANTIC CANADA. Bruce J. Barber, School of Marine Sciences. University of Maine, Orono. ME 04469; Gregory S. Bacon, Gulf Fisheries Centre. Moncton. NB. E1C 9B6. Gonadal neoplasms (germinomas) in softshell clams. Mya arenaria, have only been reported from locations in Maine, USA despite the fact that M. arenaria occurs from Labrador to North Carolina on the east coast of North America. To more accurately determine the geographic distribution of this disease, adult clams (n = 18-60 per sample) obtained between 1989 and 1997 from sites along the entire coast of Maine and from Atlantic Canada (New Brunswick, Nova Scotia, and Prince Edward Island) were examined histologically for the presence of neoplasms. Gonadal neoplasms were found at six of the 28 locations sampled, with prevalences ranging from 3.3% to 50%. All sites positive for neo- plasms were located between Penobscot Bay. Maine and Passa- maquoddy Bay. New Brunswick. Definition of the geographic dis- tribution of this disease may provide insight into its presently undetermined etiology. QUAHOG PARASITE X ("QPX") OF HARD-SHELL CLAMS, MERCENARIA MERCENAR1A AND M. MERCE- NARY VAR. NOTATA IN ATLANTIC CANADA- OBSERVATIONS FROM WILD AND CULTURED CLAMS. Gregory S. Bacon and Sharon E. McGladdery, Gulf Fisheries Centre, Moncton. NB. E1C 9B6; Bruce A. MacDonald, Depart- ment of Biology, University of New Brunswick. Saint John, NB, E2L 4L5. Quahog Parasite Unknown (QPX), has been associated with mortalities of hatchery broodstock in PEL Atlantic Canada, since the early 1990's. however, no similar mortalities have been de- tected in wild populations since the 1960"s. Investigations were, therefore, started in 1996, to determine if uninfected populations are present which could be used for hatchery broodstock. Seven hundred and twenty (720) adult clams (48-80 mm) were collected from four sites in three provinces (St. Andrews. NB (Bay of Fundy); Shediac Bridge, NB, Wallace, NS, and West River, PEI (Gulf of St. Lawrence)). Only Shediac Bridge clams were negative for QPX using histological screening. Quahogs from St. Andrews. Wallace and West River had prevalences ranging from 3.3-6.7%, 6.7-20.0% and 3.3%. respectively. The highest prevalence (20.0%) occurred in clams collected from St. Andrews in summer 1997, with a mean intensity of 86.8 and range of 1—465 organisms/ tissue section. Culture results indicate that QPX infections in healthy clams and in smaller size-groups may be difficult to detect using histology. Smaller clams (13-18 mm) which have not yet been found to be infected in the wild, or in hatcheries, were in- fected experimentally with QPX from infected clams from St. Andrews. Investigations of culture vs. histology results are on- going. A NOVEL SPECIES OF ALPHA-PROTEOBACTERIUM IS ASSOCIATED WITH SIGNS OF JUVENILE OYSTER DISEASE (JOD) IN CRASSOSTREA VIRGINICA. K.J. Boettcher, J. T. Singer, and B. J. Barber, Univ. of Maine. Orono. ME. In 1997. we deployed juvenile C. virginica on Maine's Dama- riscotta River as part of a study designed to elucidate the etiology of JOD. Cumulative losses in those oysters ranged from 55% in groups receiving weekly treatment with the antibacterial agent norfloxacin to over 84% in untreated groups. Bacteriological analysis of JOD-affected animals yielded counts > 10-fold higher than those obtained from healthy animals, and one specific colony type represented 40-95% of total recoverable colony-forming units (CFUs). These colonies achieved 1-mm diameter after 4-5 d and produced a non-diffusible pigment that resulted in a pinkish- beige appearance after 7 d. Over 200 representative isolates were phenotypically characterized; all were Gram-negative motile rods, oxidase positive, weakly catalase positive, and produced no reac- tion in oxidative/fermentative media. The I6S rRNA genes from two isolates were amplified, sequenced, and found to be identical. Alignments produced using the Ribosomal Database Project indi- cate that these bacteria are a previously undescribed species of the marine alpha-proteobacteria. In 1998. the impact of JOD on ani- mals we deployed on the Damariscotta River was minimal (less than 4% cumulative mortality); however, bacteriological analysis of affected oysters revealed the presence of CFUs morphologically identical to those isolated from 1997 JOD-animals. These bacteria 296 Abstracts, 1999 Annual Meeting. April 18-22. 1999 National Shelltlsheries Association. Halifax. Nova Scotia. Canada were not detected in healthy oysters cultured nearby at that site. nor in C. virginica maintained simultaneously in a non-JOD en- zootic area. These results provide further evidence that a bacte- rium, specifically a novel species of alpha-proteobacteria, is in- volved in JOD mortalities. THE ASSOCIATION BETWEEN THE TURBELLARIAN URASTOMA CYPR1NAE AND THE EASTERN OYSTER CRASSOSTREA VIRGINICA. Andrew D. Boghen, Nicole T. Brun, and Erick Bataller. Department of Biology, Universite de Moncton. Moncton. NB. Canada, El A 3E9. Turbellarian flatworms are primarily free-living. Many are also facultative commensals and occasionally parasitic. Such is the case for Urastoma cyprinae, a common gill-worm of bivalves that in- duces serious damage to gill tissue in mussels. In Atlantic Canada the worm is widespread in the Eastern Oyster, Crassostrea vir- ginica. Given the importance of the oyster industry on Canada's east coast, studies were initiated to shed light on the nature of the host-symbiont relationship between this worm and the oyster. While initial histological and biochemical studies provide incon- clusive evidence of pathology, our work confirms that U. cyprinae is strongly attracted to the mucus secreted by the gills. The distri- bution of the worms in different areas of the gills, coupled with specific behavioral activities as revealed by in-vivo endoscopic observations, support the likelihood that the worms may be ac- tively feeding on mucus. Such findings are further reinforced by studies employing zymographic techniques that demonstrate higher levels of protease activity in mucus for parasitized com- pared to non-parasitized oysters. Because of the visible presence of worms on the gills, investigations are currently underway to elimi- nate the worms prior to marketing using short-term depuration methods based on salinity tolerance. EFFECT OF COLD WATER ON LIMITING OR EXACER- BATING SOME OYSTER DISEASES. Susan M. Bower and Gary R. Meyer, Department of Fisheries and Oceans, Pacific Biological Station. Nanaimo, British Columbia. V9R 5K6. Canada. Like most other pathogens, agents of oyster diseases are influ- enced by environmental conditions. Information accumulated to date indicates that some parasites are only pathogenic when water temperatures are cold for extended periods while others do not produce disease when temperatures are low. For example, con- trolled laboratory studies revealed that Mikrocytos mackini, the agent of Denman Island disease in Crassostrea gigas, did not develop in oysters held at >15°C and temperatures <10°C were required for at least 3 months for the development of disease and associated mortalities. This temperature requirement explains the occurrence of the disease only in the spring of the year and its absence in oysters south of British Columbia despite the extensive historic relocation of potentially infected oysters to the Pacific United States. Conversely, other oyster diseases seem to be cur- tailed by cold temperatures. For example. Nocardia sp. has been detected in C. gigas in British Columbia throughout the year. However, nocardiosis occurs only towards the end of exceptionally warm summers. The literature also indicates that Ostracoblabe implexa, the agent of oyster shell disease, requires >22°C for more than two weeks to proliferate and optimum growth occurs at 30°C (Alderman & Jones 1971. Trans. Br. Mycol. Soc. 57:213-225). This requirement for warm temperatures probably explains why this fungus is not a problem in British Columbia although we have detected it in a few oysters. Diseases caused by Perkinstis marinus and Bonamia ostreae have not been detected in Canada possibly because these parasites require warm temperatures to replicate. The question that research must now address is how long can these pathogens survive in hosts held at temperatures not suitable for their pathogenic expression before potential carriers can be certi- fied as free of infection. IMPACT OF BONAMIA OSTREAE ON CULTURED OS- TREA EDULIS AT TWO SITES ON THE DAMARISCOTTA RIVER. MAINE. Ryan B. Carnegie and Bruce J. Barber, School of Marine Sciences. University of Maine, Orono. ME 04469. The oyster parasite. Bonamia ostreae. persists in Ostrea edulis populations at several locations in Maine. However, B. ostreae prevalence is always low, and heavy infections are rare. To assess the potential role of temperature on growth and prevalence of B. ostreae, we deployed 2000 O. edulis seed in July 1997 among eight surface trays at each of two sites on the Damariscotta River. One location (Little Point) was in a warm (>21°C in summer) area in the upper river, and the other (Lowes Cove) was a cooler (<18°C). mid-river site. Shell height, mortality, and prevalence of B. ostreae were measured bimonthly beginning in May 1998. Temperature was measured 4x daily and salinity weekly. Growth was significantly greater at the warmer Little Point site, where the oysters had reached 70. 1 ± 2.8 mm in shell height by September 1998. At Lowes Cove, average shell height was 48.8 ± 3.0 mm. However, cumulative mortality was also greater at Little Point: 41.4 ± 1 L.6% in September 1998. vs. 21.8 + 6.8% at Lowes Cove. Instantaneous mortality was 31.8% at Little Point in July 1998 following a significant mid-June rain event that reduced salinity below 22%o. B. ostreae was observed only in animals reared at Little Point. Prevalence of B. ostreae was 3.1% in May 1998. and 9.4% in July 1998. and infection intensity in all cases was low. Thus, it is more likely that the mortality was due to a sharp de- crease in salinity than to B. ostreae. National Shellfisheries Association, Halifax. Nova Scotia. Canada Abstracts, 1994 Annual Meeting. April 18-22. 1999 297 MYCOTIC PERIOSTRACAL SLOUGHING. T. Jeffrey Davidson, Claude Morris, and David Groman, Atlantic Veteri- nary College. University of Prince Edward Island. 550 University Ave., Charottetown. PEL Canada CIA 4P3. A brown discolouration has been observed on the shells of market-sized blue mussels {Mytilus edulis) for the past number of years. Histopathology and electron microscopy have identified the pathology as loss of the shell's periostracum due to a marine fungus, hence the name of the new disorder. Mycotic Periostracal Sloughing. This fungus appears to deposit on the shell in late summer/early fall in the first growing season in socks. It has not been observed in spat or mussels in their first growing season before October. The pathogenesis begins with single foci on the shell. The lesions extend horizontally causing coalescing of the foci and vertically, causing deterioration and sloughing of the pe- riostracum and to a lesser degree the prismatic layer. The brown discolouration seen on the shell could be due to the loss of the periostracal layer causing a decrease in refractivity. Studies are ongoing to determine if this disorder affects the health of the meat or growth parameters. All mussel growing areas on P.E.I, are affected. It has also been observed in Mytilus trossulus and M. galloprovincialis in Washington State. INVESTIGATION OF THE STRESS RESPONSE, SUMMER MORTALITY AND DISEASE RESISTANCE OF OYSTERS, CRASSOSTREA SPP. C. S. Friedman, G. N. Cherr, J. S. Clegg, A. H. Hamdoun, J. L. Jacobsen. S. A. Jackson, and K. R. Uhlinger, Bodega Marine Laboratory, P.O. Box 247, Bodega Bay. C.A. 94923. The ability to mount a stress response is often essential for an organism's survival, especially for oysters inhabiting dynamic and stressful environments. We have characterized the response of Pacific oysters. Crassostrea gigas, to heat shock (HSl. hypo- osmotic acclimation and disease. Pacific oysters synthesized heat shock proteins (HSP) in the 70 kD family and exhibited prolonged induced thermotolerance (ITT. at least 3 wk) after exposure to heat shock. The temperature needed to induce a stress response appears to be related to the upper thermal limit of the oyster (43.5-44 C) not on the magnitude of thermal shock. Despite an increase in thermal tolerance after HS. we observed a significant reduction in chemotaxis and percent phagocytosis in oysters after HS relative to those without a heat treatment. Hypo-osmotic acclimation delayed HSP production and tolerance to lethal temperatures. Oysters chal- lenged with Nocardia synthesized HSPs in a pattern similar to control animals. However, the degree of ITT was reduced in oys- ters with nocardiosis. Although reproductive stress does not alter the stress response of Pacific oysters, the ehemotactic ability of hemocytes from gravid oysters was significantly reduced as com- pared with nonreproductive animals (p < 0.001). The Eastern oyster, C. virginica, produced HSPs after heat shock a similar manner as the Pacific oyster. However, the mag- nitude of ITT was reduced relative to that of Pacific oysters held under the same culture and experimental conditions. PROSORHYNCHUS SQUAMATUS (DIGENEA: PLATY- HELMINTHES) INFECTION OF BLUE MUSSELS, MYTI- LUS EDULIS, IN ATLANTIC CANADA. Sharon E. McGlad- dery, Mary F. Stephenson, and Fiona McArthur, Gulf Fisheries Centre. Moncton. NB. E1C 9B6. In May, 1997, a sample of mussels (Mytilus edulis) from the Atlantic coast of Nova Scotia, was found to contain the sporocysts of the digenean Prosorhynchus squamatus at a prevalence (%P) of 13.3% (n = 60). A single infected specimen (<7rP = 3.3%) was found at a neighbouring site but. with the exception of a single mussel from the Magdalen Islands and another possible specimen from northern New Brunswick, no other infections had previously been recorded. Mussels at the affected site demonstrated reduced tolerance of processing handling and mortalities were observed in the samples collected. Mantles showed abnormal colouration (patchy yellow-white) in heavily infected individuals, however. early/light infections could not be detected by gross observation. Four species of fish (Gadus morhua, Hippoglossoides plates- soides, Melanogrammus aeglefinus and Myoxocephalus scorpius) have been recorded as definitive hosts of P. squamatus, in Atlantic Canada, but this is the first record of the sporocyst stage in blue mussels. Metacercarial host(s) has/have yet to be identified. Due to concern about mussel health and possible spread of the parasite, along with routine mussel transfers, a survey was initiated in 1997/ 98 to determine what size-groups were infected, the life-cycle, seasonal dynamics, pathogenic effects and current distribution of infection. Results from this survey were compared with records from 2300 mussels, collected throughout Atlantic Canada between 1990-1997 and confirmed an apparent focal distribution. Work on the other hosts involved in the life-cycle is ongoing. MONITORING SHELLFISH HEALTH IN NEWFOUND- LAND: A PREVENTATIVE APPROACH. Kelly Moret, Cyr Couturier, G. Jay Parsons, and Kate Williams, Fisheries and Marine Institute of Memorial University of Newfoundland. St. John's. NF. Canada. A1C 5R3. The rapid expansion of the mussel aquaculture industry in Newfoundland prompted the University. Industry Association and funding agencies to initiate a research project aimed at establishing a health/disease profile of cultured mussels throughout the prov- ince. Prior to the establishment of this program, farmers wishing to assess the health status of their stock were required to submit samples to either the Department of Fisheries and Oceans. Monc- 298 Abstracts. 1999 Annual Meeting. April 18-22, 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada ton. NB. or the Atlantic Veterinary College. UPEI. Charlottetown. PEL The status of the local industry is ideal for compiling a health baseline survey as imports, exports, and transfers to the island have been limited, thereby allowing us to take a preventative approach to health care. In 1997, a total of 13 farms participated in the initial survey (792 mussels). Shell examination revealed that 2 of the 13 farms had a very low incidence of the fungus identified as Mycotic Periostracal Sloughing Disease ("MPS"). Histological examina- tion revealed that 76. 92% of the farms had mussels which con- tained parasites. The most prevalent parasites were the gill ciliates. Ancistrum mytili and Sphenopliiya-tike ciliates. The total parasite prevalence for the 792 mussels was low (average 7.8%). For 1998. sampling was expanded to 23 sites. In addition to routine histo- logical examination, other research included studies on the fungus "MPS" and stress protein analysis. Conclusions to date include: parasites found in Newfoundland mussels are similar to those found in mussels throughout Atlantic Canada and the parasite load in Newfoundland cultured mussels is apparently lower than that in mussels in the rest of Atlantic Canada. TEMPERATURE EFFECTS ON BROWN RING DISEASE SUSCEPTIBILITY AND DEFENSE-RELATED ACTIVI- TIES IN THE MANILA CLAM, RUD1TAPES PHILIPPI- NARUM. Christine Paillard. Bassem Allam, and Radouane Oubella, UMR CNRS 6539. Institut Universitaire Europeen de la Mer. UBO. Brest. France; Susan E. Ford, Haskin Shellfish Re- search Laboratory. Rutgers University. Port Norris. NJ 08349. Brown Ring Disease (BRD) in the manila clam. Ruditapes philippinarum is a bacterial disease, caused by Vibrio tapetis. which perturbs the calcification process. Field and experimental studies confirm that BRD can be classified as a cold-water disease. It occurs in wild and cultured clam populations along the European Atlantic coasts, but it is practically absent in areas that have high summer temperatures. It has never been reported in manila clams in the USA or Japan. Epizoological surveys of BRD along the French Atlantic coast have shown inter-site prevalence variations that could be explained by temperature extremes. Experimental Vibrio tapetis infections carried out at different temperatures (8. 14 and 21°C) have documented variations in BRD prevalence as a function of temperature, i.e. 20% at 21°C versus 100% at 8 and 14°C, 4-weeks after inoculation. Bacterial growth rate is maximal between 14 and 21°C and declines at higher temperatures. Total hemocyte counts, phagocytosis, and leucine-aminopeptidase ac- tivities were significantly higher in control or inoculated clams at high than at low temperature. Higher lysozyme activity was mea- sured at 21C compared to the lower temperatures, in clams in- oculated with V. tapetis. Greater activity of potential cellular de- fense mechanisms at high temperature, as well as reduced bacterial growth, could be associated with the recovery process occurring at temperatures of 21°C or higher. FEEDING PHYSIOLOGY & ECOLOGY OF BIVALVES CTENIDIA AS THE SITE OF PARTICLE SELECTION IN BIVALVES: A COMPARISON BETWEEN SIMPLE AND COMPLEX CTENIDIAL SYSTEMS. Shirley M. Baker and Jeffrey S. Levinton, Dept. of Ecology and Evolution. State Univ. of New York. Stony Brook, 1 1794: J. Evan Ward, Dept. of Ma- rine Sciences. Univ. Connecticut, Groton. 06340. Suspension feeding bivalves compensate for fluctuations in the quality and quantity of their food supply by changing pumping and ingestion rates, and by rejecting non-nutritive particles as pseud- ofeces. Sorting and rejection of particles have been thought to occur on the labial palps. Recently however. Ward et al. (1998) report that in oysters (Crassostrea virginica and C. gigas) particle selection takes place on the ctenidia, while in mussels (Mytilus trossulus) the ctenidia play little role in particle selection. They suggest that the complex structure of the oyster ctenidia (plicate, heterorhabdic) contributes to the ability to sort particles. Using video endoscopy, we have found that other bivalves are also able to sort particles on the ctenidia. These include the zebra mussel, Dreissena polymorpha, and the freshwater unionid mussels Amblema plicata, Pyganodon cataracta, and EUiptio complanata, which have non-plicate, homorhabdic ctenidia. In D. polymorpha, material transported in the dorsal ciliated tract was primarily pre- ferred particles and underwent little further processing by the la- bial palps. Particles were carried in two ways at the ventral food groove; ultimately rejected particles were transported in a mucus string above the food groove, while ultimately accepted panicles were often transported deep within the groove. This type of two- layer transport in the ventral food groove was also observed in the unionids. At high particle concentrations, the unionid species ap- pressed the ventral areas of the two inner demibranchs. While ultimately accepted particles continued to be transported in the ventral food grooves, rejected particles were carried in a mucus string in the "groove" formed by the two demibranchs. Where the two inner demibranchs parted, near the pedal gape, the mucus string was transferred to the mantle and expelled. These examples provide evidence for some degree of particle sorting on ctenidia that are less complex than those of oysters. THE END OF THE PARTICLE PROCESSING LINE: MANTLE PSEUDOFECES REJECTION MECHANISMS IN SUSPENSION-FEEDING BIVALVES. Peter G. Beninger, Laboratoire de Biologie Marine. Universite de Nantes, Nantes, 44322 France, and Anne Veniot, Departement de Biologie, Uni- versite de Moncton. Moncton N.B. E1A 3E9 Canada. Characteristics of pseudofeces voidance from the mantle were investigated in the four major bivalve particle processing systems using scanning electron microscopy, to determine what mecha- National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 299 nisms underlie the final step in the rejection route. The entire mantle surfaces of Mytilus edulis (homorhabdic filibranch), Mya arenaria and Spisula solidissima (eulamellibranch). Crassostrea virginica and C. gigas (pseudolamellibranch). and Placopeclen magellanicus (heterorhabdic filibranch) were surveyed and photo- graphed. In both the homorhabdic filibranchs and the eulamelli- branchs. the mantle rejection tracts previously located using video endoscopy were characterized by cilia which were extraordinarily long compared to the cilia of the general pallial surface. These long cilia were grouped into closely-adhering tufts, herein termed com- posite cilia. In Mytilus edulis and Mya arenaria, the general pallial surface presented shorter simple cilia, whereas in Spisula solidis- sima the general pallial surface presented short simple cilia dor- sally, long composite cilia ventrally, and an intermediate band of composite cilia. Effective mucociliary transport using long cilia can only be accomplished if they are grouped, as in all three species with mantle rejection tracts. In C. virginica and C. gigas, pseudofeces are carried atop short simple cilia, which are them- selves elevated above the general pallial surface by prominent, specialized radial ridges. These four species all possess a gill ven- tral particle groove, hence the site of pseudofeces production is the labial palp, and a discrete mantle rejection tract is necessary to transport pseudofeces to the inhalent siphon for expulsion. The long composite cilia and the specialized rejection ridges within this tract provide vertical isolation of pseudofeces from the general pallial surface. In the heterorhabdic filibranch Placopecten magellanicus, no specialized cilia were observed on the mantle: this corresponds to the absence of a mantle rejection tract in this system, which does not possess a gill ventral particle groove and relies on valve ad- duction to expel pseudofeces. These results suggest that elevation above the surrounding mantle cilia is the rule in species relying on mantle rejection tracts for the voidance of pseudofeces. Cilia map- ping is thus seen to be a valuable technique in the ongoing study of particle processing mechanisms in suspension-feeding bivalves. CHALLENGES OF EXAMINING POSTINGESTIVE SE- LECTION IN BIVALVES. Martha G. S. Brillant and Bruce A. MacDonald, Biology Department and Centre for Coastal Studies and Aquaculture. University of New Brunswick. Saint John. NB, Canada E2L 4L5. Postingestive selection is known to occur in several species of bivalves, however the factors responsible for selection have not been established. Studies on postingestive selection have not usu- ally attempted to isolate which particle variables selection is based on or test particles were not presented to the bivalves simulta- neously so that no choice was given. We have shown that size and density of particles play a role in postingestive selection in the sea scallop (Placopecten magellanicus). Determining the role of par- ticle chemistry or quality is more challenging. Factors which must be considered when choosing test particles to study postingestive selection by chemical properties are 1 ) test particles should appear physically identical but chemically distinct to the bivalve, 2) test particles must be easily distinguishable from one another by the researcher. 3 1 test particles must be traceable and quantifiable after passage through the bivalve. 4) preferably the test particles should be natural food particles and. 5) the integrity of the test particles in the bivalve stomach should be similar. We have addressed most of these factors by developing a protocol using two colours of fluo- rescent carboxylate-modified microspheres, one coated with pro- tein. Scallops will be fed coated and uncoated beads simulta- neously with the feces being collected and analyzed by flow cy- tometry. Gut retention times of the two bead colours will be compared to determine if postingestive selection has occurred. This method should provide a means of determining whether scal- lops can distinguish particles within the stomach on the basis of chemical properties alone. SEASONAL VARIATION IN FOOD UTILIZATION BY SEA SCALLOPS AND BLUE MUSSELS. Peter J. Cranford, Fish eries and Oceans Canada. Bedford Institute of Oceanography, P.O. Box 1006, Dartmouth, NS, B2Y 4A2. Seston utilization by adult Placopecten magellanicus and Myti- lus edulis cohorts was measured using an in situ method during the spring, summer and fall of 1995. Large seasonal changes in the rates and efficiencies of feeding and absorption were observed, but only 28% of the variance in daily ingestion rates of both species could be explained by empirical models developed using a wide range of potential environmental influences (temperature, seston abundance and composition and vertical particle flux). Ingestion and absorption rates of scallops and mussels showed similar sea- sonal patterns with highest rates observed during the spring, when diet quantity and quality were high, and during late-autumn, when quantity and quality were both low. These data indicate that seston utilization and related growth was not caused solely by seasonal food and temperature fluctuations, but imply physiological regu- lation of feeding and digestion. Both species displayed a large capacity for controlling clearance and absorption rates. Clearance rates during October and November were at least twice as high as observed at other times of the year, and absorption efficiency gradually decreased at high diet quality and increased when quality was low. The accuracy of available bivalve clearance (filtration) rate models was assessed by comparing predicted responses with average in situ clearance rates. Only those models based on natural seston rations provided adequate predictions of observed clearance rates. Clearance rate predictions based on algal cell rations greatly overestimated in situ clearance at all times of the year and do not appear to apply to animals in nature. Future directions in ecophysi- ology need to focus on the in situ responses of bivalve filter feed- ers. 300 Abstracts. 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association. Halifax, Nova Scotia, Canada PARTICLE AGGREGATES IN SESTON AND THEIR ROLE IN BIVALVE PARTICLE SELECTION. Jon Grant and Mike Nickerson, Department of Oceanography. Dalhousie University, Halifax, NS B3H 4J 1 . Natural suspended particles consist of a complex mixture of phytoplankton, detritus, and inorganic material. A persistent fea- ture of marine particles in that they exist as aggregates which make the characterization of food quality for suspension feeders particu- larly troublesome. Bulk measures of food including organic con- tent do not contain information about particle size or aggregation state which determine aspects of suspension feeder handling such as selection efficiency. There are few options for measuring ag- gregation state, since particles are fragile and subject to sampling artifacts. Moreover, although there has been theoretical and em- pirical progress regarding particle aggregation (e.g., fractal dimen- sion!, the organic components of aggregates are poorly known. Two approaches to these problems include novel ways of sampling and analyzing particles from the field, and manipulation of aggre- gates in the laboratory for use in bivalve feeding experiments. Newly developed small volume particle samplers were used to examine the size and composition of natural aggregates collected in the field, with particular reference to organic components. In addition, feeding experiments were initiated using clay-protein ag- gregates to examine selection efficiency in blue mussels. Initial results from these experiments are discussed in terms of assess- ment of the food supplies available to suspension feeding bivalves and their implications for understanding the energetics of feeding. A COMPARISON OF FEEDING PHYSIOLOGY IN DIF- FERENT SIZES OF CULTURED AND WILD MYTILUS EDULIS AND M. TROSSULUS. Melissa Mooney, G. Jay Par- sons, and Cyr Couturier, Fisheries and Marine Institute of Me- morial University of Newfoundland, St. John's, NF, Canada A1C 5R3. Seasonal patterns in food demand were examined in two spe- cies of cultured and wild blue mussels from a single site in New- foundland to contribute towards an improved model of estimating production capacity. Small and large (shell length <20 or >50 mm. respectively) cultured and wild blue mussels were obtained monthly. Mussels were acclimated to laboratory conditions, main- tained at ambient temperature, and fed low. medium or high food rations (<3500, 5000-7500 or >9000 cells/mL. respectively). Diets consisted of 50/50 mixtures of Isochrysis galbana and Chaetoc- eros muelleri. Weight specific rates of filtration (g/h/g), clearance (L/h/g) and metabolism (mg 0-,/h/g) were measured. Overall, clear- ance and filtration rates were significantly higher at higher tem- peratures (ANOVA, P < 0.05). lower during spawning (ANOVA, P < 0.01 ) and higher in smaller mussels compared to larger mus- sels (ANOVA. P < 0.01). Metabolic intensity was significantly higher at higher temperatures (ANOVA, P < 0.01). Mytilus tros- sulus demonstrated a significantly higher clearance rate and meta- bolic intensity than M. edulis (ANOVA. P < 0.01 ). No significant differences were observed between cultured and wild mussels for any of the factors tested. The proportion of different size mussels and species on a given site has direct implications for stocking density, and will impact the overall production capacity of a given culture site. The results validate the importance of grading, and suggest practices such as stocking smaller mussels near the outer boundary of a site (where food availability is maximal) may be more effective. THE EFFECTS OF CURRENT SPEED AND PARTICLE CONCENTRATION ON MUSSEL [MYTILUS EDULIS) FIL- TRATION RATE: A RECIRCULATING FLUME STUDY. Carter R. Newell, Great Eastern Mussel Farms, Inc., P.O. Box 141, Tenants Harbor. ME 04860. Bivalve filtration rates are one of the most important yet un- certain terms in models of shellfish carrying capacity. Previous workers (Newell et al., 1998) investigated mussel shell gape in the field using a time lapse video over two tidal cycles. They found that shell gape (distance between valves), an indication of pumping rate, was positively correlated with filtration rate in experimental chambers. Due to the difficulty of separating covarying environ- mental variables, namely current speed and particle concentration, in the field, a laboratory flume approach was chosen to calibrate the shell gape assay. After acclimation to experimental diets, individual mussels were attached to plexiglass stands and subjected to various con- centrations of food with current speed constant and various current speeds with food concentration constant while a time-lapse video system recorded shell gape and exhalent siphon area. Each run included a 15-20 minute acclimation period followed by data cap- ture over 40—15 minutes. Video tapes were analyzed with an Opta- mus system. Mussel exhalent siphon area showed a significant decrease with an increase in current speed from 5 to 30 cm s~ . Below about 5 million particles P1, mussels significantly reduced their pumping rates. The negative effects of high current speeds on mussel pumping and filtration rates have important implications for the suspension culture of mussels. ROLE OF OYSTERS IN MAINTAINING ESTUARINE WA- TER QUALITY. Roger I. E. Newell, Jeff C. Cornwell, and Mike Owens, Horn Point Laboratory, UMCES. PO Box 775. Cambridge, MD 21631; Jon Tuttle, Chesapeake Biological Labo- ratory, UMCES, PO Box 38, Solomons. MD 20688. Environmental changes in Chesapeake Bay, such as elevated phytoplankton biomass and loss of benthic plants, are often thought to be largely a function of nutrient-driven eutrophication. We propose, however, that populations of the eastern oyster, Cras- sostrea virginica, which have been reduced to <1 ck of their historic levels, may have exerted "top-down" control on phytoplankton National She sheries Association, Halifax, Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 301 stocks and also reduced turbidity, thereby increasing light avail- able to benthic plants. In laboratory incubations under oxic and anoxic conditions we measured changes in sediment geochemistry, nutrient fluxes, and denitrification in response to loading by dif- ferent amounts of algal paste, an experimental analog of oyster biodeposits. Increased organic loading to the sediment under oxi- dized conditions resulted both in higher rates of coupled nitrifica- tion/denitrification and denitrification in the presence of water col- umn nitrate. In contrast, coupled nitrification/denitrification was suppressed under anoxic conditions. Similar incubations in the presence of benthic microalgae showed negligible ammonium fluxes from sediments, with the algal/microbial community effi- ciently retaining ammonium and fixing nitrogen. Because no DIN was recycled to the water column under oxic conditions we con- clude that rehabilitation of natural oyster stocks will have the beneficial effect of removing phytoplankton from the water col- umn without stimulating further phytoplankton production. Fur- thermore, nitrogen will be removed from the Bay via increased denitrification. These data also suggest that private-sector oyster aquaculture should be encouraged not only for the obvious eco- nomic value but also for the broader ecological benefits to the Bay. COMPOSITE CILIA: DESCRIPTION OF A NEW TYPE OF CILIUM USED IN PARTICLE PROCESSING IN BI- VALVES. Anne Veniot. Departement de Biologie, Universite de Moncton, Moncton N.B.. E1C 3E9. Canada; Peter G. Beninger, Laboratoire de Biologie Marine. Universite de Nantes, Nantes, 44322 France. Cilia are the elementary mechanism responsible for the creation of water currents in the pallial cavity, as well as for particle capture and subsequent processing. Recent scanning electron microscopy observations of several pallial surfaces involved in particle capture and processing in bivalves have shown the presence of unusually long, grouped cilia. These cilia were named composite cilia to distinguish them from the known compound cilia (cirri) and simple cilia. However, high-resolution ultrastructural observations and determination of the characteristics of composite cilia are neces- sary in order to accurately determine whether these cilia represent a new ciliary type. We used scanning electron microscopy, light microscopy and transmission electron microscopy to examine the structure and ultrastructure of composite cilia in four Bivalve spe- cies: Mytiltts edulis. Mya arenaria, Spisula solidissima and Cras- sostrea virginica. Data show that composite cilia are made up of a group of simple cilia originating from a single cell. In contrast to compound cilia (cirri), the component simple cilia present no fu- sion at any point along their length. The component cilia are densely packed and uniformly spaced, with one microvillus be- tween adjacent cilia, a striking contrast with both compound cilia, which have no space or microvilli between adjacent cilia, and simple cilia which are unevenly spaced with several intervening microvilli. These characteristics suggest that composite cilia are a new category of cilium, with a distinct arrangement of non-fused component cilia. The coarse frontal cilia of frontal filaments of the gill of C. virginica are shown to belong to this category of cilium. We suggest that mechanisms of adhesion in compound cilia may consist of weak molecular interactions, such as hydrogen bonds or attraction between protein groups of opposing polarity. The struc- ture of composite cilia impart properties which enable them to perform specialized tasks in particle processing. MEDIATION OF FEEDING AND SELECTION BY SEC- ONDARY METABOLITES OF DETRITAL PARTICLES. J. Evan Ward, Department of Marine Sciences. University of Connecticut, Groton, CT 06340; Jeffrey S. Levinton, Ecology and Evolution, SUNY at Stony Brook, Stony Brook, NY 1 1794; San- dra E. Shumway, Natural Sciences Division, Southampton Col- lege, LIU. Southampton. NY 1 1968; Lisa Milke, Department of Marine Sciences, University of Connecticut, Groton, CT 06340. In environments dominated by kelp forests, a large amount of the suspended detrital material is derived from the seaweeds. These kelp-derived particles can support growth of suspension feeding bivalves. We employed flow cytometry and video endos- copy to study the effects of kelp detritus on feeding and particle selection in the oyster, Crassoslrea gigas. Oysters were fed mix- tures of phytoplankton and similar size detrital particles prepared from ground kelp. Fresh particles derived from the kelp Agarwn fimbriatum had higher concentrations of phenolic compounds than those from Costaria costata. Aging for 4 days reduced phenolic concentrations of particles from both species. Clearance rates of oysters were significantly inhibited by high phenolic concentra- tions of the kelp detritus, and clearance increased as Agarwn par- ticles aged. Selection by the gill of the oyster was also affected by phenolic concentration of the kelp detritus. Particles with low (aged Costaria) and high (fresh Agarwn) phenolic concentration produced low selection, whereas particles with intermediate phe- nolic concentration produced high selection. In all assays, kelp particles were significantly rejected in the pseudofeces. Our data suggest that phenolics are inhibitory to feeding and selection mechanisms in oysters. GROWTH AND CULTURE OF SHELLFISH EFFECT OF FISHING PRESSURE ON SIZE OF ADULT MALE BLUE CRABS IN MARYLAND: CALVERT CLIFFS AND THE PATUXENT RIVER— 1998. George R. Abbe and Brian W. Albright, Academy of Natural Sciences Esturaine Re- search Center. St. Leonard, MD 20685. The Maryland blue crab fishery has experienced increasing fishing pressure since the 1940s. Regulations enacted in 1994 to limit or reduce effort and improve the fishery have had little posi- 302 Abstracts, 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada tive effect. Data from near Calvert Cliffs in Chesapeake Bay have shown a declining trend in mean size for total male crabs and legal size males since 1968. To determine if this decline was due to fishing pressure we compared results from Calvert Cliffs, where fishing pressure is intense, with an area in the Patuxent River where fishing pressure is low. Crab pots are the gear of choice on the Bay. but are restricted for commercial use in the river. The two areas examined were about 8 km apart by air and had similar temperatures and salinities. Peeler pots of 1-inch mesh baited with menhaden were fished at both areas at similar depths from June to November 1998. During this time 41 1 and 670 pots were fished on the river and Bay. respectively, yielding 974 and 1888 crabs. Total catches per unit effort were similar at 2.4 and 2.8. but river crabs were 89% male compared to 41% on the Bay. Legal males (s5- inch carapace width) made up 69% of the river catch compared to 10% of the Bay catch, and males >6 inches accounted for 28% of the river catch compared to only 2% in the Bay. Mean size of all males was 5.5 and 4.6 inches in the river and Bay. respectively: legal males were 5.8 and 5.6 inches, respectively. Reduced fishing pressure in the river allows many male crabs the opportunity to molt again after reaching minimum legal size as evidenced by the percentage of 6-inch males, whereas most males in the Bay are caught shortly after reaching legal size. Further regulations to re- duce effort in the Chesapeake or an increase in minimum legal size may by necessary to improve the quality of the fishery. TRIPLOID PRODUCTION OF MYTILUS EDULIS IN PRINCE EDWARD ISLAND— AN INDUSTRIAL INITIA- TIVE. John W. Brake, Jeffrey Davidson, Atlantic Veterinary College, University of Prince Edward Island, and Dr. Jonathan Davis, Taylor Resources Inc.. Quilcene. WA. The mussel aquaculture industry in Prince Edward Island (PEI), Canada is a well established major contributor to the island economy. In 1997 farm gate value exceeded $12 M (Can) and export value exceeded $24 M (Can). The industry currently sup- ports over 1250 full and part time jobs. Harvesting and marketing during the spawning season is cur- rently an area of concern for the industry. Mussels that have re- cently spawned have a low (less appealing) meat yield, while those close to spawning can spawn out en route to the market from processors. In both cases the potential exists to increase consumer dissatisfaction with the product. Triploids have very poorly devel- oped gonads, thus more energy can be used for meat production instead of gonad production, allowing for the possibility of larger meat yields than normal diploids during the spawning season. The production of triploid mussels might therefore alleviate these prob- lems, allowing the marketing of a high quality product year round. The production of triploid Pacific oysters (Crassostrea gigas) is currently extensively practiced in the Pacific Northwest. Triploid clams (Ruditapes philipparum), scallops {Placopecten magellani- cus), and mussels (A/, galloprovincialis) have all been produced as well as others. Identified methods of triploid induction (used at different levels or in combinations) in shellfish include tempera- ture and/or pressure shocking and the use of chemicals such as caffeine, cytochalasin B, or 6-dimethylaminopurine. The mussel industry has recognized the potential of harvesting triploid mussels during the spawning season. The objective of this study is to elucidate the best triploid induction methods for com- mercial use in PEI by the use of a matrix of previously identified triploidy induction methods. These combinations of methods will be ranked by % induction and % survivorship, as well as feasibility in order to determine the best method. Hatchery techniques will be fine tuned for the species, then the performance of triploids in the field will be evaluated to find when it might be more or less advantageous to utilize them. DIFFERENTIAL EFFECTS OF TWO ISOLATES OF AU- REOCOCCUS ANOPHAGEFFERENS, IN UNIALGAL AND MIXED SUSPENSIONS, ON FEEDING AND GROWTH OF BIVALVES. V. Monica Bricelj, Seott MacQuarrie, Institute for Marine Biosciences. National Research Council. 1411 Oxford Street. Halifax. NS B3H 3Z1, Canada: Roxanna M. Smolowitz, Laboratory for Aquatic Animal Medicine and Pathology. Univer- sity of Pennsylvania. Marine Biological Laboratory. Woods Hole. MA 02543. Previous work examined only short-term effects of Aureococ- cus anophagefferens, the causative agent of brown tides in mid- Atlantic estuaries, on feeding of bivalves. We conducted three- week laboratory trials to determine the effects of two cultured isolates of A. anophagefferens on survival, growth (in shell length, biovolume and total organics) and histopathology of juvenile (6 mm) quahogs. Mercenaria mercenaria. The algal strains were iso- lated in 1986 and 1995 from Long Island, New York, bays. Three unialgal treatments were used: Isocluysis galbana (t-Iso) at 60 x 106 cells L"1. and A. anophagefferens at a high cell density ( 1 x 10g cells L"1) simulating natural blooms, and moderate density (400 x 106 cells L~'); two were mixed suspensions of t-Iso (60 x 10" cells L_1 ) spiked with either 400 or 20 x 106 Aureococcus cells L"1. A non-fed control was included. Survival was not signifi- cantly affected in any of the treatments. No growth was observed at high or moderate densities of Aureococcus. Marked growth suppression relative to the t-Iso control occurred in the mixture spiked with 400 x 10h Aureococcus cells L*1, supporting our find- ings with adult mussels that Aureococcus inhibits capture of other non-toxic algae co-occurring in a mixed assemblage. Less severe growth inhibition of clams occurred in the mixture containing 20 x 10n Aureococcus cells L~\ and growth performance in this treatment increased gradually over the course of three weeks. In contrast, growth trials with the 1986 isolate showed positive growth of M. mercenaria juveniles in both high and moderate concentrations of this strain, and comparable growth in the t-Iso control and mixture spiked with 400 x 10 Aureococcus cells L~ . National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 303 These results indicate that cell densities (20 x 106 cells L"'l an order of magnitude lower than previously documented, can have deleterious effects on bivalve growth, but that acclimation to such low levels of Aureococcus can occur over time. They also support our bioassay results with adult mussels which show no evidence of toxicity with the 1986 brown tide isolate, whereas two 1995 iso- lates elicit strong feeding inhibition. EVALUATING THE PERFORMANCE OF NON-NATIVE OYSTER SPECIES IN VIRGINIA. Gustavo W. Calvo, Mark W. Luckenbach, and Eugene M. Burreson. School of Marine Science. Virginia Institute of Marine Science. College of William and Mary. Gloucester Point. VA 23062. The Pacfic oyster, Crassostrea gigas, and the Suminoe oyster. Crassostrea ariakensis. have been suggested as potential substi- tutes for and/or enhance native oyster. Crassostrea virginica stocks which have been devastated by disease and other factors in Virginia. Field experiments with triploid progeny of quarantined broodstock have been established as a means to evaluate the per- formance of non-native oysters, while preventing undesired intro- duction of disease agents and/or uncontrolled proliferation of ex- otic species in Chesapeake Bay and the Atlantic Coast of Virginia. During a one year experiment using juvenile oysters at 9 sites covering a wide range of salinity. C. gigas has demonstrated over- all higher disease resistance and superior growth and survival at high salinity sites (>25 ppt), as compared to C. virginica. Con- versely, at low salinity sites (<15 ppt). C. virginica has had higher survival and growth than C. gigas. Performance at intermediate salinity sites (15-25 ppt). was similar for both species. In an on- going study using 2 year old oysters at a subset of 6 of the study sites described above. C. ariakensis is exhibiting overall higher disease resistance and equal or superior survival and growth than C. virginica. after six months of deployment. The results of these field studies provide only a partial basis for evaluating potentials and pitfalls of using non-native oyster species to enhance the oyster fishery and aquaculture industry in Virginia. Other factors not addressed in these studies, including reproductive capabilities of the exotic species and its ecological interactions with other species in the new environment, must also be under- stood to more fully evaluate the environmental impact of potential introductions. JUVENILE GROWTH OF CAGE-REARED STIMPSON'S SURFCLAMS (MACTROMERIS POLYNYMA) IN MAINE, USA. Christopher V. Davis, Pemaquid Oyster Company. Inc., PO Box 302, Waldoboro. ME 04572. Stimpson's surfclam (Mactromeris polynyma, Stimpson 1860) is a circumboreal species ranging throughout northwestern Atlan- tic and northeastern Pacific inshore waters. Distinguished from other Mactrids by its purple foot, siphon, and mantle edge which turn brilliant orange-red when cooked, this species is highly sought after for the Japanese sushi market. A preliminary investigation of this species' mariculture potential in Gulf of Maine waters was conducted in 1997 from biological, technical, and commercial per- spectives. Laboratory and field growout studies investigated the effects of substrate type and growing site on juvenile growth and survival during one growing season. Cage-reared surfclams set out in the low intertidal zone exhib- ited both poor growth and survival. Surfclams grew an average of 1.6 mm in shell length (6.4% increase in SL) and had a 30% cumulative mortality over a 3.5 month period. In contrast, subtid- ally-reared juvenile surfclams grew from an initial size of 23 mm SL to 33, 35 and 39 mm SL at the Horseshoe Cove, Damariscotta River and Beals Island sites, respectively. Mean size among co- horts at the three growing sites were significantly different from one other after two months growth and continued to diverge in subsequent months. Mortality was nil at all three growing sites. Although the overall time to rear this species to the target size of 50 mm SL in Maine waters is still unknown, this growth informa- tion, when combined with data from other year/size classes, will assist in predicting the growing time required to reach market size. SURVIVAL AND GROWTH OF JUVENILE GREEN SEA URCHINS ON DIFFERENT MACROALGAL SETTLE- MENT SUBSTRATES. Nils T. Hagen, Department of Fisheries and Natural Science. Bodo College. N-8002 Bod0, Norway. The effect of settlement substrate on the survival and growth of juvenile green sea urchins {Strongylocentrotus droebachiensis) was investigated in a laboratory experiment which was indepen- dently replicated 3 times. Five substrates were investigated: the articulate coralline alga Corallina officinalis; the crustose coralline alga Lithothamnion glaciale: the foliose red alga Palmaria pal- mata; the minute green alga Ulvella lens; and the natural biofilm which developed in empty experimental containers. A total of 4750 competent urchin larvae gave rise to 2064 juveniles during Phase 1 of the experiment, which lasted between 6 and 1 1 weeks. During Phase 2 of the experiment, which lasted between 8 and 1 1 weeks, all surviving juveniles were fed P. palmata. The total num- ber of surviving juveniles was reduced to 1 1 53 by the end of Phase 2. In Phase 1 the control treatments with natural biofilm had the best survival rates (=60-70%), closely followed by Corallina and Lithothamnion. whereas survival on Palmaria was 25-50%. and survival on Ulvella was £10%. In Phase 2 the pattern of survival changed substantially, and juveniles which had settled on the cor- alline algae clearly had the best overall survival. The growth of juveniles settled on the coralline algae was consistently better than controls during both phases of the experiment, although juveniles settled on Palmaria had similar growth in the third replicate of the experiment and in Phase 1 of the first replicate. In conclusion, the best overall growth and survival was obtained in treatments with coralline algae. 304 Abstracts, 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia, Canada INCREASING THE SOMATIC GROWTH RATE OF JUVE- NILE GREEN SEA URCHINS (STRONGYLOCENTROTUS DROEBACHIENSIS) USING PREPARED DIETS. Eddy J. Kennedy and Shawn M. C. Robinson, DFO, Biological Station. St. Andrews, NB. Canada EOG 2X0; G. Jay Parsons, Marine Institute, Memorial University, St. John"s, NF, Canada A1C 5R3. As the world market demand for sea urchins continues to rise, the wild fishery is reaching a limit for providing good quality sea urchins thus other means, such as aquaculture. must supply world demand. In order to provide a constant seed supply for a sea urchin aquaculture industry, a hatchery must become a realization. Part of such an operation is maximizing juvenile somatic growth rates to shorten the overall production cycle, increase quality and increase production revenue. One approach is to use prepared diets as these have a positive effect on gonad growth rates of adult sea urchins. Components of the prepared diets must be investigated to deter- mine the effects on juvenile growth so that a diet yielding maximal growth rates can be developed. This study investigated the effect of three factors: protein source, protein concentration and juvenile size on somatic growth rates. Protein sources used were (a) soy- bean protein, (b) 95% soybean protein and 5% fish protein, and (c) 50% soybean and 50% fish protein. For each set of diets, 4 dif- ferent protein concentrations were used: 20%, 30%. 40%, and 50%, for a total of 12 diets. The two cohorts of sea urchins had a size range of 1-8 mm and 12-20 mm initial test diameter. Each treatment consisted of 4 replicates in a randomized block design. Twenty-four tanks, with 4 treatment baskets per tank, were subject to the prepared diets while 2 "control" tanks were fed a natural diet of seaweeds for comparison. Results of the study will be further discussed with regard to establishing diets for future hatchery/ nursery production of the green sea urchin. SHELLFISH RESTORATION: HAVE WE BEEN SUCCESS- FUL? Dorothy L. Leonard, Habitat Conservation Office, Na- tional Marine Fisheries Service, Silver Spring, MD 20910. The climate of diminished public funding and the desire to maintain and improve shellfish and its habitat has motivated many in the scientific, educational, public, nonprofit and private sectors to join forces and share resources to achieve shellfish restoration goals. Also concerned with declines in environmental quality and fisheries are numerous local volunteer organizations who have developed successful programs to identify problem areas, recom- mend improvements and monitor progress. As a result the last five years have seen an acceleration in the number of shellfish resto- ration activities undertaken by individuals and partnerships formed to facilitate funding and implementation. The question raised in this presentation is how successful these efforts have been. Is there an increase in shellfish waters approved for harvest? Have shell- fish landings increased? Are the enhancement and aquaculture efforts improving the health of the ecosystem? And, what have the costs been? Examples of successful shellfish restoration projects are presented with an evaluation of the impacts. EFFECTS OF SEA LETTUCE, ULVA LACTUCA, MATS ON ABUNDANCES OF SOFTSHELL CLAMS, MYA ARENARIA, AND ASSOCIATED INVERTEBRATES IN NEW JERSEY. Clyde L. MacKenzie, Jr., James J. Howard Marine Sciences Laboratory. Northeast Fisheries Science Center, 74 Magruder Road. Highlands, NJ 07732. Eutrophication of estuarine waters throughout the world has resulted in a proliferation of Ulva sp. and other algas. In the Nave- sink River in northeastern New Jersey, sea lettuce, Ulva lactuca, mats grew over beds of softshells, Mya arenaria. The softshells initially extended their siphons several cm out of the substrates and within a few weeks their entire bodies emerged from the substrates and the softshells died. The mats also sharply reduced the abun- dances of associated invertebrates, such as polychaetes. other mol- lusks, and crustaceans. Few invertebrates lived on the surfaces of the mats. In southern New Jersey, other investigators showed that sea lettuce mats provide habitat for some small fishes. We need studies to determine effects of the presence of sea lettuce on a broad spectrum of animals before we can remove it to increase abundances of softshells for the benefit of commercial and recre- ational fishermen in this area. SPATIO-TEMPORAL VARIATION IN SESTON FLUX, GROWTH AND PRODUCTION OF THE BLUE MUSSEL, MYTILUS EDULIS, HELD IN SUSPENDED CULTURE, IN A SUB-ARCTIC ENVIRONMENT. Gina McNeil and Cyr Couturier, Fisheries and Marine Institute of Memorial University of Newfoundland. St. John's. NF. Canada A1C 5R3. The Newfoundland mussel culture industry has experienced significant growth over the last five years and growers are begin- ning to fully utilise the available space. This has raised questions as to the extent/level of maximum/optimum stocking levels. A reciprocal transplant experiment of three mussel populations was undertaken at three commercial aquaculture sites with different hydrographic and environmental regimes. Variations in growth, survival and production were assessed bimonthly in relation to seston flux, temperature and salinity which were measured every 2-3 weeks at several stations on each site. CaS04 cylinders ( 15 cm long x 3.75 cm in diameter) were calibrated with S4 current probes to assess relative current speeds. A strong positive relationship was established between cylinder dissolution and actual current speeds, providing a useful index for calculating seston flux. Mussel growth and production varied according to season, populations, site and location on the site (ANOVA, p < 0.05). There were no observable differences in natural mortality amongst populations. Differences in mussel performance were related to the relative seston flux, generally showing higher growth and production in areas of higher flux. The importance of relative food flux measurements is dis- cussed in relation to site evaluation and production capacity esti- mates. National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 305 THE EFFECT OF LOCATION AND TIME OF YEAR ON MUSSEL PRODUCTIVITY IN AN AQUACULTURE ESTU- ARY. Linda E. Waite, Thomas Landry,1 and Jeff Davidson, Atlantic Veterinary College. PEL 'Department of Fisheries and Oceans, Gulf Fisheries Centre, Moncton, NB. The PEI aquaculture shellfish industry, worth more than $25 million annually, is beginning to see signs of production limita- tions. Production differences among bays and within bays have been questioned by the industry and this uncertainty is leading to concerns on the capacity of grow-out sites to withstand further development. A DFO (Department of Fisheries and Oceans) Stra- tegic Science Fund initiative research program called COSAD (Coastal Oceanography and Sustainable Aquaculture Develop- ment) has been formed to address this issue. It recognizes the importance to understand the relationship between marine habitats and aquaculture production. As part of the COSAD initiative, stud- ies have been conducted to determine if geographic location and time of year have an effect on mussel productivity. Mussel pro- ductivity, measured as the incremental increase in shell length and physiological condition factor over a specified time period, was sampled from five sites over two years from May to November and one year from January to March. These sites are in Tracadie Bay. a productive aquaculture estuary on the north side of Prince Ed- ward Island. They were chosen for their diverse geographic loca- tions and cover the entire bay where aquaculture is located. Analy- sis of mussel productivity depending on the location within the bay showed differences with higher productivity near the mouth of the bay. Analysis of mussel productivity depending on the time of year showed differences with the highest productivity in June with a second peak in early October. The effect of physical and biological interaction on mussel productivity will be discussed. MITIGATING PREDATION BY THE EUROPEAN GREEN CRAB, CARCINUS MAENAS, UPON PUBLICLY MARI- CULTURED QUAHOGS, MERCENARIA MERCENARIA. William C. Walton and Gregory M. Ruiz, Smithsonian Environ- mental Research Center. PO Box 28. Edgewater. MD 21037; Be- thany A. Starr, Martha's Vineyard Shellfish Group. PO Box 1552. Oak Bluffs. MA 02557. To supplement local clam, M. mercenaria, fisheries, munici- palities on Martha's Vineyard, MA (USA) seed local mud flats with juvenile clams. Despite the apparent modest success of this program, managers would like to improve seed survival. Observed seedings indicate a rapid (<72 hrs) decline in seed numbers coin- cident with evidence of predation (e.g.. shell fragments). Field per capita feeding experiments and the overwhelming predominance of green crabs relative to other predators indicate that the European green crab. Carcinus maenas, is the most damaging predator. To mitigate this predation. we tested several mechanistic seeding fac- tors: timing, size and density. Monthly outplants of seed from April through October did not indicate any 'window of decreased predation'. In cage experiments, however, seed size significantly influenced predation rates of green crabs. Further laboratory ex- periments indicated that large seed (18+ mm shell length) were relatively safe from green crabs (by far the most common crab) and lady crabs Ovalipes ocellatus, but to a lesser degree from blue crabs, Callinectes sapidus. Lastly, per capita predation rates by green crabs enclosed in field cages increased with increasing seed density (8 to 144 per sq. m). Open field plots of different seed densities, however, exhibited no such increase; rather, predation was approximately 50% across the range tested. In conclusion, we present a preliminary model integrating these factors, as well as efforts to directly control green crabs, to provide managers with an optimum seeding strategy. LOBSTER ECOLOGY AND FISHERIES LOBSTER {HOMARUS AMERICANUS) MOVEMENT IN THE SOUTHERN GULF OF ST. LAWRENCE. Michel Comeau, Mare Lanteigne, Guy Rohichaud, and Fernand Savoie, Science Branch, Department of Fisheries and Oceans, Maritimes Region. P.O. Box 5030. Moncton. N.B., Canada. E1C 9B6. Lobster (Homarus americanus) tagging studies were conducted in 34 sites throughout the southwestern Gulf of St. Lawrence be- tween 1980 and 1997. Results show relatively small traveled dis- tances between the release and recapture position for animals rang- ing from 48 to 152 mm of carapace length. The average traveled distance ranged from 2 km in parts of Baie des Chaleurs and Cape Breton to 15 km in central Northumberland strait. No relationship was observed between the traveled distance and the number of days at liberty or the size of the animal. Furthermore, except for 6 out of 22 sites, no significant difference was observed for the average traveled distance between males, females and berried fe- males. For the six sites where a significant difference was ob- served, the average traveled distance was farther for female lob- sters. Geographical differences observed in the traveled distances seem to be related to the bottom topography. Long average trav- eled distances were observed in sites characterized by flat bottom with relatively smooth transition between shallow and deep areas (1 to 20 m). Short average traveled distances were observed in near shore lobster habitat characterized by a change in bathymetry ( to depths >20 m) over a relatively short distance. 306 Abstracts, 1999 Annual Meeting. April 18-22, 1999 National Shellfisheries Association. Halifax. Nova Scotia, Canada DEVELOPMENT OF A LIPID CONDITION INDEX IN LOBSTERS (HOMARUS AMERICANUS H. MILNE ED- WARDS, 1837) AND ITS APPLICATION IN THE INTER- PRETATION OF LARVAL DISTRIBUTION IN CLOSE PROXIMITY TO GEORGES BANK, GULF OF MAINE. Gareth C. Harding, Department of Fisheries and Oceans. Bed- ford Institute of Oceanography, P.O. Box 1006. Dartmouth. N.S.. B2Y 4A2. Canada; A.J. Fraser, 33 Battery Drive. Halifax. N.S.. B3P 2G9, Canada. The triacylglycerol/sterol condition index was applied to larval lobster populations in the vicinity of Georges Bank in the Gulf of Maine. This index is related to larval size by an increasing power function which explains around 40% of the variation. This poor fit can be explained by the uneven increase in triacylglycerol levels during development within each moult stage. Increased pigmenta- tion is not related to larval condition, as measured by lipid storage, and masks the increased yellowish hue of lipids as development proceeds. The larval triacylglycerol/sterol index appears to un- dergo a diurnal cycle in stage IV and possibly stage III lobster, with lowest values at midday and highest values after dark. This pattern can not be explained completely by either nocturnal feed- ing or vertical diffusion by diurnal migrations, which leaves the possibility that healthy larvae might also be more likely to see the trawl and escape during daylight. Few lobster larvae were found in the vicinity of Georges Bank with a condition index less than 0.1, which is the level laboratory studies indicate approaches the "point-of-no-return". The condition of all developmental stages was found to be better in individuals located off Georges Bank. This is not ecologically significant in the case of the first two stages because such a small proportion of the population was ac- tually located off the bank. It is not resolved how the third and fourth stages arrive off Georges Bank, since shoal water hatching is the norm, but their lipid reserves are significantly greater than identical developmental stages on the bank. Finally, the abundance of stage IVs per m3 in the surface waters over Gulf of Maine is twice that found over Georges Bank which suggests that the lob- ster has evolved a life cycle in offshore waters in which the last two planktonic/pelagic stages either seek or are transported to. and by stage IV thrive in, the warmer stratified waters over greater depths. BEHAVIORAL THERMOREGULATION AND ITS EF- FECTS ON THE MOVEMENT OF LOBSTERS IN THE FIELD. Steve Jury, W. Hunt Howell, and Winsor Watson, Center for Marine Biology & Zoology Dept., University of New Hampshire. Durham, N.H. 03824. We have shown that lobsters, Homarus americanus, are ca- pable of behaviorally thermoregulating in laboratory thermal gra- dients where males consistently preferred warmer temperatures than females, showing a final preferred temperature of 16.3°C (i.e., the temperature where acclimation temperature equals selected temperature), whereas females had a final preferred temperature of 14.3°C. Lobsters preferred significantly wanner temperatures in colder months (2.8 ± 0.66 warmer than ambient temperatures of 8.2 ± 0.3°C) than in warmer months (0.2 ± 0.39 warmer than ambient temperatures of 15.5 ± 0.4°C). Activity and avoidance assays also show that lobsters avoid water warmer than 23.5 ± 0.4CC and at <5°C they are relatively immobile and thus cannot behaviorally thermoregulate. Catch per unit effort (CPUE) data from 1989-1991 in the Great Bay Estuary. New Hampshire show significant non-linear relation- ships with these laboratory data such that the highest catches tended to correlate with the preferred temperatures. In addition, a mechanistic model of the behavioral response to temperature based upon laboratory data results in movements comparable to those estimated from estuarine tag-recapture data from 1989-1991 in terms of direction and relative distance moved. These results in- dicate that laboratory and field derived estimates of thermal pref- erence of lobsters are comparable in several ways and that models predicting lobster movements, distribution and abundance using temperature as a parameter may benefit by incorporating sex and acclimation effects. LOBSTER (HOMARUS AMERICANUS) COMMERCIAL CATCH COMPOSITION FLUCTUATIONS BASED ON A TIGHT TEMPORAL AND GEOGRAPHICAL SEA SAM- PLING PROGRAM. Mare Lanteigne, Science Branch. Depart- ment of Fisheries and Oceans, Maritimes Region. P.O. Box 5030. Moncton, N.B.. Canada, E1C 9B6. A weekly lobster (Homarus americanus) commercial catch sampling program was conducted in 35 landing sites along the coast of Prince Edward Island (PEI). Canada, in 1998. By sam- pling onboard commercial fishing vessels, information was col- lected on the prevalence of lobsters under the minimum legal size and of egg bearing females, which are both returned at sea during regular fishing activities. Results show important temporal and geographical size com- position variability. The removal of commercial size lobsters, and the accumulation of sub-legal size and egg bearing females as they are returned at sea, explains the temporal changes observed in the size compositions. The prevalence of egg bearing females in the commercial catches shows areas of abundance along the coast of PEI. These observations are suggesting that some areas may have higher levels of lobster egg production compared to neighbouring areas. A distinct size composition gradient is observed in the Northumberland Strait. The central section of the Strait is charac- terized by the presence of larger lobsters and an overall multiple mode wide size range compared to the single mode narrow size range in both extremities of the Strait. These observations corrobo- rate with the information obtained from fisheries statistics and suggest a phenomenon that may be explained by a particular lob- ster movement pattern within the Northumberland Strait. National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 307 EGG PER RECRUIT AS A MANAGEMENT TARGET FOR AMERICAN LOBSTER FISHERIES. R. J. Miller and K. F. Drinkwater, Bedford Institute of Oceanography. Dartmouth. Nova Scotia B2Y 4A2. Management targets for American and spiny lobsters have been defined using the ratio of eggs per recruit in fished to unfished populations. Chosen values not to be exceeded range from 0.05- 0.2. For 13 areas in eastern Canada we found no correlation be- tween eggs/recruit (e/r) and fishery yield/km2 of fishing ground. Adjusting for possible larval drift between adjacent fishing areas did not improve the correlation. This result suggests that manage- ment targets using egg production are not useful, or should be area specific. Using the e/r ratio for fished and unfished populations presents other problems. Data for calculating e/r for unfished (vir- gin) populations seldom exist. The concept of e/r in fished and unfished populations is difficult to communicate to stakeholders. There is little or no theoretical basis for choosing a particular ratio. The calculation is based on eggs per lobster rather than eggs per geographic area; we usually manage lobster fisheries by area. In- puts to calculating e/r (growth, fishing mortality, size at maturity) can change temporally. SPATIAL SCALING OF HABITAT DISTRIBUTIONS IN THE AMERICAN LOBSTER. Robert W. Rangelev and Peter Lawton, Biological Station. Department of Fisheries and Oceans. St. Andrews, New Brunswick, Canada EOG 2X0. American lobster (Homarus americanus) preferences for near- shore cobble habitats are well known yet little is understood about the effects of spatial scaling of habitat patches on lobster distribu- tion patterns. We studied patterns of lobster distributions for a range of spatial scales in two regions in Atlantic Canada immedi- ately following heavy fisheries exploitation. In the southern Gulf of St. Lawrence habitat patches were extremely complex and frag- mented at large spatial scales yet the flat cobble and bedrock substrates yielded relatively little crevice space at the scale of individual lobster shelters. Densities in the nearshore were low and heavily skewed towards juveniles. Exceptions were in relic stream beds which were small in area but contained the most productive habitats. In contrast, at our Gulf of Maine sites patch sizes were very large and fragmentation was low. Small scale structural com- plexity was high and supported greater densities of all sizes classes of lobsters. In addition, adult lobsters were also distributed on soft sediments and on large patches of horse mussels far from shelter patches. Our study demonstrates the value of analysing distribution patterns at multiple spatial scales. DAILY MOVEMENTS OF LOBSTERS FROM ULTRA- SONIC TRACKING. M. John Tremblay and R. Duggan, In- vertebrate Fisheries Division. Bedford Institute of Oceanography. P.O. Box 1006. Dartmouth. N.S. B2Y 4A2: R. O'Dor, C. Curtis. D. Webber, and Y. Andrade, Department of Biology, Dalhousie University, Halifax. N.S. The daily range and periodicity of lobster (Homarus america- nus) movement is of interest from several perspectives, including the likelihood of entry to traps, and the size of marine protected areas. Data are reviewed from studies that have located lobsters intermittently, and from 2 studies on the east coast of Nova Scotia where lobsters were tracked continuously. In Jeddore Harbour (early summers of 1989 and 1997), lobsters were tracked with ultrasonic transmitters fixed to the carapace. An array of 3 acoustic- buoys received pulses from the transmitters and a base station calculated position based on pulse arrival time. Five ovigerous females (CL > 100 mm) were tracked in 1989; four immature females (72-88 mm CL) in 1997. For several reasons the 1989 and 1997 studies are not directly comparable but the results suggest differences in activity and movement related to size or maturity. Compared to the ovigerous females, the immature females tended to move less, particularly during daylight. Between the hours of 0800 and 2000 h immature females moved 32^18% as far as during the night (2100-0700 h). Ovigerous female movement dur- ing the day was 58-98% of that during the night. The greater daylight activity of large female lobsters may result in increased catchability which has been observed elsewhere at this time of year. The generality of size-related movement differences needs further study. MODELLING SHELLFISH ECOSYSTEMS A PHYSIOLOGICALLY-BASED MODEL OF THE GROWTH AND DEVELOPMENT OF CRASSOSTREA GI- GAS LARVAE. E. A. Bochenek, New Jersey Sea Grant Exten- sion Program. Rutgers Cooperative Extension, 1623 Whitesville Rd.. Toms River, NJ 08755; E. N. Powell, Haskin Shellfish Re- search Laboratory, Rutgers University, 6959 Miller Ave., Port Norris, NJ 08349; E. Hofmann and J. Kiinck, Center for Coastal Physical Oceanography. Crittendon Hall. Old Dominion Univer- sity. Norfolk. VA 23529. A physiologically-based model was used to investigate the pro- cesses that control the growth and development of larvae of the Japanese oyster. Crassostrea gigas. The model is structured around the nonlinear relationship between larval caloric content and larval length; larvae reach a minimum in caloric content at 130-160 urn. Formulations used to model larval filtration, inges- tion and respiration rates were based upon laboratory and field studies. Simulations of larval growth show that under environmen- tal conditions typical of most temperate estuaries, larval develop- 308 Abstracts, 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia, Canada ment is completed in 20-30 days, depending upon ambient water temperature and food concentration. Periods of reduced food con- centration, especially as the larvae are approaching their minimum caloric value, result in the larvae reaching and exceeding their metabolic point-of-no-return. In these simulations, providing food later in the larval development does not enhance larval survivor- ship. Thus, predictions of C. gigas larval survival from only adult egg condition will be in error if environmental conditions during the early larval life history are not considered. A BAYESIAN APPROACH TO SHELLFISH ECOSYSTEM MODELLING. Michael Dowd, Satlantic Inc.. Halifax. Canada; Renate Meyer, Dept. of Statistics, University of Auckland. Auck- land. New Zealand; W. Carlisle Thacker, Atlantic Oceanographic and Meteorological Laboratory. Miami. FL. Limited ecosystem models provide a useful framework for pre- dicting shellfish growth and for addressing aquaculture concerns related to stocking density and carrying capacity. Such models predict the time evolution of that portion of the ecosystem directly important to shellfish growth and are coupled to circulation or particle dispersion models. For practical or management purposes, it is imperative that the uncertainty in the predictions of these intrinsically nonlinear models be quantified. Towards this end. we argue for a stochastic framework to predicting shellfish growth which optimally combines observations and models. Specifically, we offer a Bayesian approach to data assimilation which is appli- cable to time-dependent, nonlinear, and non-Gaussian ecosystem models. We focus on the practical aspects of implementing the methodology and provide an illustration based on a simple model describing bivalve growth in a coastal inlet. Extensive use is made of available software that takes advantage of recent advances in Monte Carlo integration and provides a computational environ- ment suitable for the non-statistician. Our overall goal is to assess the promise (and pitfalls) of this Bayesian approach to unifying statistical and mechanistic models in order to better describe and predict the dynamics of shellfish ecosystems. EFFECT OF SPAWNER DENSITY AND DISTRIBUTION ON FERTILIZATION SUCCESS IN THE SEA SCALLOP, PLACOPECTEN MAGELLANICUS GMELIN. Jean-Francois Dumais and Xavier Boespflug, Universite du Quebec a Rimouski. 310 allee des Ursulines, Rimouski. Quebec. G5L 3A1 Canada; Dominique Baudinet and Marcel Frechette, Institut Maurice- Lamontagne, Ministere des Peehes et des Oceans. C.P. 1000. Mont-Joli. Quebec, G5H 3Z4 Canada. We model critical spawner concentration for preventing re- cruitment overfishing (RO) in sea scallops, Placopecten magel- lanicus. We assume that fertilization rates are dependent on ga- mete concentration as measured in the laboratory. Fertilization rate in situ varies along gamete diffusion plumes downstream of spawners. The criterion for RO is that reproductive effort of an individual female is insufficient to balance total mortality during larval, juvenile and adult life. Natural mortality estimates are taken from the literature. Preliminary runs of the model suggest that for the particular settings used (regular spatial distribution, size of scallop groups = 4, 36 individuals. 10 cm shell height females, current speed = 0.2 m/s. etc.). results are sensitive to group size, which implies that small scale spatial patterns are important in fertilization success. Populations with the highest natural mortality rates tested do not persist. At median natural mortality rates and annual exploitation rate = 0.1. 0.2 and 0.3. RO occurs at spawner population density lower than ca. 0.004/m2, 0.45/irr and 5/nr. respectively. In groups with low natural mortality rate. RO does not occur at annual exploitation rates lower than 0.5. MODELLING RESUSPENSION AND ITS EFFECTS ON BI- VALVE FOOD SUPPLIES. Jon Grant, Department of Ocean- ography. Dalhousie University. Halifax, NS B3H 4J1; Cedric Bacher, Centre de Recherche en Ecologie et Aquaculture (CREMA). BP 5. PHoumeau. 17300. France. Many models of shellfish growth in culture are based on food supplies being a primary determinant of growth. The representa- tion of suspended food in these models is difficult because seston is a mixture of phytoplankton. detritus, and other material with complex temporal dynamics. In shallow coastal systems, resuspen- sion contributes significantly to the particle load, although the material injected into the water column is of variable quality, depending on the abundance of benthic microalgae, macrophyte detritus, etc. Rates of resuspension are regulated by substrate type (sand and mud), shear stress due to waves and currents, and bio- logical processes such as bioturbation. For cohesive muds, a model of resuspension and its effect on water column turbidity is super- ficially straightforward. However, there are a variety of pitfalls including critical erosion threshold, depth of erosion, deposition, food quality, and groundtruth data all of which complicate model formulation. A model of tidal resuspension is applied to a mussel aquaculture site in a shallow muddy bay (Upper South Cove, Nova Scotia). Previous studies in the cove have examined erosion rate and threshold as well as near-bottom suspended particulate matter, providing an unusually complete data set for model verification. Considerations of erosion formulation and its assumptions are pre- sented for this model. The implications for applying resuspension to overall models of bivalve culture carrying capacity are then explored. National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts. 1999 Annual Meeting. April 18-22. 1999 309 MODELLING POPULATION DYNAMICS OF PACIFIC OYSTER CRASSOSTREA GIGAS IN KOREA. Kyung-Hoon Hyun, Ig-Chan Pang, and Kwang-Sik Choi, College of Ocean Science. Cheju University Korea, Cheju-Do 690-756. Korea; Eric N. Powell, Haskin Shellfish Research Laboratory, Rutgers Uni- versity, Port Norris, NJ 08349: John M. Klinck and Eileen E. Hnfmann, Center for Coastal Physical Oceanography, Old Do- minion University, Norfolk. VA 23529. To determine the growth relationships of Pacific oyster, Cras- sostrea gigas, shell length, weight and gonadal condition of post- settlement oysters in mariculture fields in the Kamakman Bay. Choun-Nam Province. Korea, were measured from June 1997 to May 1998. Environmental factors, including food levels, were also measured. Dry weights were used to determine the rate of oyster growth in the field, and these were compared to the simulated growth rates obtained from a model of oyster population dynamics. The Kamakman Bay oysters with a shell length of 17 mm and a dry weight of 0.3 g that settled in June had increased to 73 mm and 3.2 g by May of the next year. This growth rate is slow compared to the growth rate of the same species grown in Japanese maricul- ture fields. No significant spatial variations in oyster growth were observed within one oyster farm sampled in detail, suggesting that flow speed through the farm is sufficient to prevent a local reduc- tion in food supply produced by oyster filtration. The oyster model predicts lower growth curves than observed in the field when chlorophyll-a is used as a measure of oyster food supply. This is not unusual. Adding a correction factor from Soniat et al. (in press) based on measures of non-chlorophyll explained food produces a much better fit between simulation and observation. Using labile carbohydrates+protein as a food source results in higher growth rates than observed. Reducing filtration efficiency by assuming that most of the food is present in small particles results in simu- lated growth rates that are similar to field conditions. tory. Estimates of the salinity tolerance of the larval stages of R. venasa are described as a precursor to using current 3-D modeling of Chesapeake Bay circulation to estimate a potential range of distribution of the species within the Chesapeake Bay and its sub- estuaries. Such estimates are crucial to establishing which shellfish resources are potentially susceptible to predation by local R. venosa populations. LINKING WATER QUALITY AND LIVING RESOURCES: A COUPLED SUSPENSION FEEDER-EUTROPHICATION MODEL. Mark B. Meyers, HydroQual, Inc., 1 Lethbridge Plaza, Mahwah, NJ 07430; Dominic M. Di Toro, HydroQual. Inc. and Dept. of Environmental Engineering. Manhattan College. River- dale, NY; James J. Fitzpatrick, HydroQual. Inc. Modern eutrophication models explicitly simulate many as- pects of the carbon, nutrient, and oxygen dynamics of complex estuaries, including the interactions of settling organic matter, sediment diagenesis, and fluxes across the sediment-water inter- face. At the same time resource managers are seeking insight regarding connections between traditional eutrophication problems and living resource issues. We have coupled a model of suspen- sion-feeding bivalve production to the Chesapeake Bay Water Quality Model, providing a dynamic feedback in the eutrophica- tion-production-anoxia loop and a link between traditional water quality concerns and the food resources of higher trophic levels. Incorporation of suspension-feeding bivalves into the model im- proved calibration in regions where extensive bivalve populations are known to exist. Predicted bivalve biomass compared well with an extensive 12-year benthic monitoring program data set, in re- sponse to temporal and spatial variations in food and bottom water dissolved oxygen conditions. RAPANA VENOSA IN THE CHESAPEAKE BAY: CUR- RENT STATUS AND PROSPECTS FOR RANGE EXTEN- SION BASED ON SALINITY TOLERANCE OF EARLY LIFE HISTORY STAGES. Roger Mann, Juli Harding, and Stephanie L. Haywood, Virginia Institute of Marine Science. Gloucester Point, VA 23062. The Veined Rapa Whelk. Rapana venosa, has recently been identified as present in the Hampton Roads region of the Chesa- peake Bay. The species is native to the Sea of Japan, but was introduced to the Black Sea in the 1940's. and has since spread to the Aegean and Adriatic Seas. There is strong evidence that range extension is mediated by transport of early life history stages in ballast water. The current status of knowledge of distribution of R. venosa in the Chesapeake Bay is described. There is concern over the potential impact of R. venosa on local shellfish populations and the industry that they support. Egg cases of R. venosa have been collected from the field, and larval forms cultured in the labora- MODELING THE MSX PARASITE IN EASTERN OYSTER (CRASSOSTREA VIRGINICA) POPULATIONS: MODEL DEVELOPMENT, IMPLEMENTATION AND VERIFICA- TION. Eric N. Powell and Susan E. Ford, Haskin Shellfish Re- search Laboratory. Rutgers University. Port Norris. NJ 08349; Eileen E. Hofmann and John M. Klinck, Center for Coastal Physical Oceanography. Crittenton Hall. Old Dominion Univer- sity, Norfolk. VA 23529. A model simulating the host-parasite-environmental interac- tions of Eastern oysters (Crassostrea virginica) and the pathogen. Haplosporidium nelsoni. which causes the disease MSX. has been developed. The model is physiologically-based and is structured around proliferation and death rates of H. nelsoni under different environmental conditions. Equations describing these rates were constructed using data from long-term field observations and field and laboratory experiments. Simulations that use environmental conditions characteristic of Delaware Bay reproduce the observed seasonal H. nelsoni cycles and consequent oyster mortality. These 310 Abstracts. 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia, Canada simulations show the effect of environmental factors, such as sa- linity and cold temperatures, on controlling the intensity and prevalence of H. nelsoni infections. However, biological controls from density-dependent feedback on H. nelsoni proliferation and H. nelsoni sporulation events also greatly affect disease prevalence and intensity. The oyster-//, nelsoni model provides a quantitative framework for guiding future laboratory and field studies, as well as management efforts. A SIMPLE MODEL FOR ESTIMATING TIME TO CRITI- CAL LEVELS OF PERKINSUS MARINUS IN EASTERN OYSTERS, CRASSOSTREA VIRGINICA. Thomas M. Soniat. Department of Biological Sciences. Nicholls State University, Thi- bodaux. LA 70310; Enrique V. Kortright, Department of Com- puter Science, Nicholls State University, Thibodaux, LA 70310; Sammy M. Ray, Department of Marine Biology, Texas A&M University at Galveston, Galveston, TX 77553. A simple mathematical model is presented which uses an Excel spreadsheet to estimate a time to a critical level (tCnl) of Perkinsus marinus, in eastern oysters Crassostrea virginica. The estimate assumes that a weighted incidence (WI) of disease of 1.5 is critical. It converts measured WI values and the critical WI to parasite number, calculates a rate of change (r) of the parasite population using measured values of water temperature (T) and salinity (S), and solves for tCn[ by simulation. Estimates of tCm and r using a long-term data set of T. S. and WI from the Terrebonne estuary of Louisiana are provided. The model does not predict future values of WI since it cannot predict future trends in T and S; however, regularly determining T and S, considering their interaction in a model, measuring WI at reasonable intervals, and iteratively esti- mating tCnl, should be useful to oyster management. Users will have access to the model through the worldwide web. Estimates of tCnt would support decisions concerning transplanting infected oysters to lower salinity areas, harvesting heavily-infected popu- lations early, and diverting freshwater into high-salinity estuaries. QUANTITATIVE ASPECTS OF OYSTER REEF BROOD- STOCK ENHANCEMENT IN THE GREAT WICOMICO RIVER, VIRGINIA. Melissa Southworth and Roger Mann. Virginia Institute of Marine Science, Gloucester Point VA 23062. The Great Wicomico River is a small, trap type estaury on the western shore of the Chesapeake Bay. Resident oyster populations were eliminated by the combined effects of Tropical Storm Agnes in 1972, and subsequent disease mortalities related to Perkinsus marinus and MSX. Oyster broodstock enhancement was initiated in June 1996 by the construction of a three dimensional intertidal reef with oyster shell, followed by the "seeding", in December 1996, of that reef with high densities of large oysters from disease challenged populations in Pocomoke and Tangier Sound. Esti- mated egg production of the reef population is within an order of magnitude of total egg production in the Great Wicomico River prior to Tropical Storm Agnes. Field studies in 1997 indicate spawning by reef oysters from July through September. MSX was absent. Perkinsus prevalence increased from 32% in June to 100% in July, while intensity increased from June to September. Plank- ton tows recorded oyster larval concentrations as high of 37,362 + 4,380 m-3 on June 23 — orders of magnitude higher than typically recorded in Virginia subestuaries of the Chesapeake Bay in the past three decades. Drifter studies suggest strong local retention of larvae, a suggestion reinforced by marked increases in local oyster spatfall on both shellstring collectors and bottom substrate in com- parison to years prior to 1997. REPRODUCTION AND RECRUITMENT TWO-YEAR COMPARISON OF SPAWNING PATTERNS IN SOFT-SHELL CLAMS (MYA ARENAR1A). Linda A. MacLean, Atlantic Veterinary College. UPE1, 550 University Av- enue, Charlottetown, PE CIA 4P3; Neil G. MacNair, Prince Ed- ward Island Department of Fisheries and Tourism. P.O. Box 2000. Charlottetown, PE CIA 7N8; T. Jeffrey Davidson and Gerald G. Johnson, Atlantic Veterinary College. UPEI. 550 University Av- enue. Charlottetown, PE CIA 4P3. On Prince Edward Island. Canada, soft-shell clam (Mya arenaria) fishing is characterized by licensed seasonal commercial fishers harvesting between 125 and 400 tonnes on an annual basis in the last few years. Recent interest in aquaculture development for Mya arenaria requires understanding of the characteristics of spawning since the supply of soft-shell clams is entirely dependent on natural reproduction. The spawning of Mya arenaria was moni- tored and the environmental conditions under which this process occurred were recorded for the field seasons of 1997 (May-Nov.) and 1998 (Apr.-Sept.). Samples of 30 clams greater than 50 mm (2") and samples of 30 clams between 35-50 mm (1.4"-2") in overall length were collected on a weekly basis from three sites. Sites were chosen from both the north and south shores based on different water temperature profiles, tidal flushing, and sediment type. The two north shore sites are located in the same estuary, one in the sheltered upper reaches and the second in an exposed loca- tion near the mouth of the estuary. At the exposed site, clams greater than 50 mm in length were not consistently available Steamed meat yields were performed and provided indicators of physiological changes associated with spawning in Mya arenaria. Temperature probes were left in situ at each site over the field season (both years) to regularly record water temperature. The spawning performance at the three sites indicate differences in the spawning behavior between locations, and the pattern varied be- tween years. In July of 1997. a large single spawning event oc- curred in each site whereas, in 1998. there were continuous small spawning events at most sites over the entire field season. This paper will show some of the natural variation in the spawning cycle of Mya arenaria. National Shellt'isheries Association. Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 31 1 THE INFLUENCE OF ESTRADIOL ON VITELLOGEN- ESIS IN THE GREEN SEA URCHIN, STRONGYLOCEN- TROTUS DROEBACH1ENSIS. Nature A. McGinn, Michael P. Lesser, and Charles W. Walker, Department of Zoology, Uni- versity of New Hampshire, Durham, NH 03824. Sea urchins store nutrients in specialized gonadal cells called nutritive phagocytes. The size and contents of these cells fluctuate predictably during the reproductive cycle. A major element of the large nutritive phagocytes at the onset of gametogenesis is a yolk related protein. Little is known about the origin and accumulation of this protein in sea urchins. Yolk protein synthesis, part of vi- tellogenesis. is initiated by estradiol in some animals. Treatment with estradiol may increase yolk related protein production in S. droebachiensis. Green sea urchins were collected in June 1 998 and maintained in flowing seawater. They were fed a prepared diet either untreated ( = control) or treated with 17B-estradiol (1 p.g/g feed). Test and gonad measurements were made and gonad index was determined for monthly urchin collections through October 1998. Histological sections from fixed and resin-embedded gonads were examined by light microscopy. We are determining the amount and location of yolk related protein in developing urchin gonads using Western blot analysis and immunohistochemistry with rabbit polyclonal anti-urchin major yolk protein antibody. Implications for sea urchin aquaculture will be discussed. Sup- ported by CMB grant to NAM and Sea Grant to CWW and MPL. TEMPORAL PATTERNS OF LARVAL AND POST-SET DISTRIBUTIONS OF THE BLUE MUSSEL (MYTILUS EDULIS/M. TROSSULUS) AND THE STARFISH (ASTERIAS VULGARIS) ON NEWFOUNDLAND MUSSEL CULTURE SITES. Miranda Pryor, G. Jay Parsons, and Cyr Couturier, Fisheries and Marine Institute of Memorial University of New- foundland. St. John's. NF. Canada AIC 5R3. As the blue mussel culture industry in Newfoundland grows, farmers are experiencing problems with the predatory starfish spe- cies (Asterias vulgaris). Temporal and spatial patterns of plank- tonic larvae and post-set stages of both species have been studied in-depth to determine if a consistency in the timing of appearance and abundance exists between these organisms. Four geographi- cally distinct sites were chosen throughout the province, with weekly larval samples and spat/juvenile collector retrievals taken from May through November. 1998. Mussel larvae were abundant at three sites, located on the North coast, from mid-June through late August; starfish larvae were present from late July through late August at these sites. Mussel spat and starfish juvenile settlement subsequently occurred at varying rates for all three sites, with peak starfish set occurring about 2-3 weeks after peak mussel set. At the fourth site, located on the southern shore of the island, mussel spawning was sporadic resulting in low settlement on collectors. As well, no planktonic starfish larvae or settling juveniles were observed on this site. Environmental (temperature, salinity, chlo- rophyll) and hydrographic conditions were examined in relation to timing of appearance and larval and post-set abundances. Predic- tion of starfish juvenile settlement, in relation to mussel spat settle- ment, will provide mussel growers with the ability to make deci- sions regarding collector deployment in an effort to reduce the potential impact of this predatory species. EVIDENCE FOR FALL SPAWNING OF NORTHERN BAY SCALLOPS, ARGOPECTEN IRRADIANS IRRADIANS (LAMARCK, 1819), IN NEW YORK. Stephen T. Tettelbach, Southampton College, Long Island University, Southampton. NY 11968: Roxanna Smolowitz, Laboratory for Aquatic Animal Medicine and Pathology. University of Pennsylvania, Marine Bio- logical Laboratory. Woods Hole. MA 02543; Christopher F. Smith, Kim Tetrault, and Sandra Dumais, Marine Program, Cornell Cooperative Extension, Riverhead, NY 1 1901. Spawning of Argopecten irradians irradians is generally be- lieved to occur between late May-August; however, some litera- ture reports and our personal observations have suggested that ripe individuals may be present well into the fall. In this study, we examined samples that we had archived from different bay scallop populations in eastern Long Island. New York waters, from dif- ferent years, to determine if there was any histological evidence of fall spawning. At two sites, a spawning peak in September fol- lowed a discrete spawning peak in early summer (late June/early July). Scallops at four different sites were conclusively shown to spawn well into the fall (late September, October or early Novem- ber) during three different years (1993, 1994, and 1995). Fall spawning of A. i. irradians in New York waters does not appear to be uncommon and may be very important during certain years. AN INVESTIGATION OF MYA ARENARIA (SOFT-SHELL CLAM) RECRUITMENT IN MAINE. Tracy Vassiliev and William Congleton, Department of Bio-systems Science and En- gineering. University of Maine. Orono, ME 04469; Brian Beal, University of Maine Machias. Machias. ME 04654; Stephen Fegley, Maine Maritime Academy. Castine, ME 04420. The declining Mya arenaria recruitment in Eastern Maine con- trasts with increasing clam harvests on Maine's Southwestern coast. Core samples (0.0133 m2) were collected in the fall/winter from mudflats in Eastern Maine and Southwestern Maine for two years. Densities of juveniles ( 1.8-6.0 mm) averaged 16.95 clams/ m2 in 120 samples in Eastern Maine and 204.46 clams/m2 in 120 samples in Southwestern Maine. In May through October 1998 312 Abstracts, 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association, Halifax. Nova Scotia. Canada spat bag systems were deployed in Mason Bay (Eastern Maine) and Jones Creek (Southwestern Maine) to determine the density of Mxa larvae in near-shore waters. Spat bags were replaced and the contents sieved with a 125 p. mesh each month. Twenty, four-inch flowerpots were filled with mason sand and placed on the intertidal mudflats near the spat bag systems. Pots were replaced monthly and Mya settlement was determined by sieving. Preliminary ob- servations indicate a larger number of bivalves collected at the Southwestern site in both spat bags and intertidal pots versus the Eastern sites. These results indicate the reduced number of clam spat, on the mudflats in Eastern Maine, is due to reduced densities of larvae in near-shore waters rather than location rejection by larvae for settlement on the mudflats. SCALLOP FISHERIES: ECOLOGY AND APPLIED BIOLOGY EFFECTS OF A NEW BITUMEN FUEL SOURCE ON THE GROWTH AND ENERGETICS OF SEA SCALLOPS. Shel- ley L. Armsworthy and Peter J. Cranford, Fisheries and Oceans Canada, Bedford Institute of Oceanography, P.O. Box 1006, Dart- mouth, NS, B2Y 4A2; Kenneth Lee, Fisheries and Oceans Canada, Maurice Lamontagne Institute. P.O. Box 1000 Mont-Joli. PQ, G5H 3Z4. A newly developed bitumen-based fuel is presently being used by thermal power plants as a cost-effective alternative to tradi- tional heavy fuels. Based on what is known of the physical prop- erties (relatively high density), chemical composition (large pro- portion of polyaromatic hydrocarbons), and behaviour of this fuel in seawater (formation of fine bitumen droplets), an accidental spill could pose a threat benthic suspensions-feeding organisms. Laboratory experiments were conducted to determine the lethal and sublethal effects of this fuel to a commercially important benthic suspension feeder, the sea scallop, Placopecten magellani- cus. Toxicity was determined from acute and chronic mortalities, effects on tissue growth, and changes in the animals energy status (scope for growth: SFG). Mortalities during 43-day chronic expo- sures increased with increasing fuel concentration (0. 0.01. 0.1,1 and 10 mg 1_1). but were generally low with 82% survivorship at the highest concentration tested. Scallop energy budget measure- ments provided insight into the fuel's impact on growth and the physiological mechanism! s) responsible for the impact. Scallop SFG was reduced by 80% at 0.01 mg l"1 and was negative for animals exposed to concentrations greater than 1 mg l"1. Growth reductions resulted mainly from effects on clearance rate, which was reduced by 90% at fuel concentrations greater than 1 mg 1"' (EC5n value was between 0.01 and 0.1 mg I-1). Respiration rate and absorption efficiency of scallops exposed to dispersed bitumen increased significantly relative to controls. Research on a new formulation of this fuel source is underway to assess manufacturer claims of reduced toxicity. MODELLING POTENTIAL EFFECTS OF DRILLING WASTES ON GEORGES BANK SCALLOP STOCKS. Peter J. Cranford, Donald C. Gordon. Jr.. Charles G. Hannah, John W. Loder, Timothy G. Milligan, and Dwight K. Muschenheim, Fisheries and Oceans Canada. Bedford Institute of Oceanography. P.O. Box 1006, Dartmouth. NS, B2Y 4A2. Moratoria on oil and gas activities on Georges Bank are cur- rently in place until 2000 (Canada) and 2012 (USA). If not ex- tended, exploration drilling could take place with the attendant risks to the marine ecosystem and aquatic resources. A numerical benthic boundary layer transport model {bblt) was developed to provide estimates of the suspension, drift, dispersion and concen- tration of water-based drilling mud which could be released from a hypothetical 92-day exploration well at different sites on Georges Bank. Simulations predict that highest near-bottom con- centrations of drilling mud would occur in the relatively deep Side region (>100 m) as a result of relatively low suspension, dispersion and drift. Lowest concentrations would occur in the central Mixed region (<65 m) because of high dispersion, while intermediate concentrations would occur in the Frontal region. Laboratory ex- periments show that adult scallops are highly sensitive to drilling mud, and the near-bottom concentration time series from bblt simulations provide a basis for estimating impacts. The region of greatest potential impact on scallop growth is the Side region where mud concentrations from the hypothetical release scenario are predicted to prevent scallop growth for 2—40 days depending upon the settling velocity used and area over which results are averaged. Scallop stocks in this region are relatively small but dense aggregations are found in some areas. Growth losses in the Frontal region, which has the densest scallop stocks, are predicted to be more localized and confined to a range of 0-15 days. Pre- dicted growth loss in the central Mixed region is predicted to be negligible (<2 days). MANAGEMENT ADVICE OF GIANT SCALLOP PLA- COPECTEN MAGELLANICUS BASED ON GONAD MATU- RATION. Leslie-Anne Davidson, GFC. 343 Archibald St. Monc- ton. N.B. E1C 9B6; Yves Poussart, Universite de Moncton. Moncton, N.B. El A 3E9. The giant scallop. Placopecten magellanicus, which is com- mercially fished off the east coast of Canada and the United State is also a cultured specie. Management strategies to optimize har- vest of both wild and cultured scallop, must not adversely affect stock replenishment. The reproduction cycle of the commercial size scallops (>80 mm) has been extensively studied, however, gametogenesis of pre-recruit size groups (<80 mm) has not been studied in detail until the present study. Weekly samples during the reproduction season were retained from the following shell height size intervals: 5-20 mm, 21-25 mm, 26-30 mm, 31—10 mm, 41- 50 mm, 51-60 mm. 61-70 mm and 71-80 mm. Histological go- nadal section of every scallop sampled were obtained and ob- National Shellfisheries Association. Halifax, Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 313 served. Scallops in size classes 61-70 mm and 71-80 mm filled their follicles with mature sex cells and released the majority of these cells during the spawning period. Scallops in the size classes between 2 1 mm and 60 mm accumulated the sexual cells however did not release the majority of these cells during the spawning period. Scallops are fished with dredges made up of rings and washers and the selectivity of the dredge is dependent on the ring size. If the rings are too small, it is possible to harvest a scallop before it effectively contributes to future generations. In addition. if an aquaculturist collects his own spat and also markets half-shell or princess size scallop (60-70 mm), he/she must retain some scallops >60 mm to maintain a viable broodstock on the culture site to assure a successful spat collection. JET-PROPELLED SWIMMING IN SCALLOPS. M. Edwin DeMont, Biology Department. St. Francis Xavier University. An- tigonish. N.S. Canada B2G 2W5. A comprehensive study on the mechanics of jet-propelled swimming in the scallop Placopecten magellanicus has been com- pleted. The work integrated unsteady and steady-state internal and external fluid forces, the physical properties of the hinge, and the whole dynamic system. Jet-propelled swimming is initiated by contraction of the adductor muscle. The two shells are rapidly pulled together, which generates fluid pressure in the mantle cav- ity. Jets of water are generated, which propel the animal. Refilling of the mantle cavity is powered by both fluid forces, and release of energy stored in the hinge during the contraction of the muscle. The results showed that the cost of keeping the shells oscillating is very low. so that most of the work done by the muscle is used to generate useful mantle cavity pressure. Scale effects were exam- ined, and the results showed that inferior hydrodynamic charac- teristics of the shells are enhanced as the animals grow. Potential environmental effects on swimming ability were examined, and the results showed that the physical properties of the hinge are independent of temperature. Riblets on the external surface of the shells may reduce friction drag. SCALLOP GEAR SELECTIVITY AND SCALLOP BIOL- OGY: A MISMATCH IN RESOURCE MANAGEMENT. William D. DuPaul. James E. Kirkley, and David B. Rudders, Virginia Institute of Marine Science. Gloucester Point. VA 23062. Regulated changes on sea scallop [Placopecten magellanicus) harvesting gear configuration since 1994 have resulted in signifi- cant changes in size selectivity patterns. Sequential increases in dredge ring size and trawl mesh size have shifted length frequency modes that reflect the differential selectivity of the gear to larger animals but not to an equal extent. Dredges with 88.9 mm rings inexplicably capture a greater number of animals >120 mm which are larger than the internal diameter of the rings and greater than the inter-ring spaces. In this presentation, we compare the selec- tivity patterns of various ring and mesh sizes with various size and age related biological and stock parameters such as yield per re- cruit (YPR), age and size at first capture, fecundity, and number and biomass at age when fishing mortality (F| equals 0.0. We also examine whether or not the present gear requirements would be beneficial in terms of enhanced growth and reproduction which would likely occur with closed area management strategies. Re- sults of the comparison indicate that previous and present gear sizes and configurations are inadequate for optimizing the poten- tial for gains in growth and reproduction, nor would they be suf- ficient to take full advantage of gains resulting from area closures. We conclude that area management strategies should include the use of a more size selective harvesting strategy in order to maxi- mize benefits related to increases in YPR. Recognizing the broad selectivity pattern and inherent inefficiencies of harvesting gear, additional regulatory measures may be needed to maximize the biological potential of the resource. GENETIC VARIATION IN PLACOPECTEN MAGELLANI- CUS WITH IMPLICATIONS FOR FISHERIES MANAGE- MENT. Ellen Kenchington, Fisheries and Oceans Canada. Bed- ford Institute of Oceanography, P.O. Box 1006. Dartmouth. NS. B2Y 4A2; Carolyn J. Bird, Institute for Marine Biosciences. National Research Council of Canada. 141 1 Oxford St., Halifax, NS, CANADA B3H 3Z1; Elefterios Zouros, Institute of Marine Biology of Crete, P.O. Box 2214. 710 03 Iraklio Crete. Greece. Six microsatellite variable nucleotide regions (VNTRs) have been developed for the sea scallop Placopecten magellanicus. These markers have been used in a survey of ten sea scallops beds ranging from New Jersey (U.S.A.) to St. Pierre Bank (Nfld.) and including the Gulf of St. Lawrence. We present data on allele frequencies which identify significant differences between the scallop beds. Additionally, four year classes from the Digby scal- lop grounds in the Bay of Fundy were analyzed for differences in allele frequency by locus, and a significant year class effect was observed. These results are discussed with respect to sea ranching, transfers of scallop spat for aquaculture purposes, forensic identi- fication and fisheries management. EFFECT OF DEPLOYMENT DATE ON SEA SCALLOP GROWTH AND SURVIVAL. Lorelei A. Levy and G. Jay Par- sons, Fisheries and Marine Institute of Memorial University of Newfoundland, St. John's, NF, Canada A1C 5R3; Patrick Dabi- nett, Ocean Sciences Centre, Memorial University of Newfound- land, St. John's. NF. Canada A1B 3X9. Growth and survival rates of hatchery-reared sea scallop spat (Placopecten magellanicus) (1.7 mm shell height) deployed on a farm-based nursery are variable. For a commercial operation the 314 Abstracts. 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada problem arises of when intermediate-sized scallops (7 mm SH) are available. Commencing in July, deployment of scallops over five consecutive 18-day intervals on a farm-based nursery provided the opportunity to determine the effects of deployment time on scallop growth and survival. Shell height was measured at the end of each interval, and again in November and the following June. Water quality and sea star settlement were monitored also. Scallops de- ployed in August 1997 reached intermediate size by November 8. 1997, while scallops deployed in September 1997 reached inter- mediate size by June 24. 1998. Scallops deployed on October 19. 1997. however, had not reached intermediate size by June 24, 1998. Growth rates were significantly different between intervals (ANOVA. F = 95.162: d.f. 4.1 1. P < 0.001 1 and correlated with temperature (r = 0.840: P < 0.001 ) and total phytoplankton levels (r = 0.994. P < 0.001). Survival declined over the study period. Survival was significantly different between intervals (ANOVA. F = 47.129. d.f. 4, 1 1, P < 0.001 ) and negatively correlated with sea star settlement (r = -0.796. P < 0.001). Reduced growth and survival of scallops over different time intervals at a farm-based nursery illustrates the need to deploy scallops in July and August when temperature and food conditions are high. Management of spat with respect to the settlement of sea stars in September is also important. RETENTION OF SCALLOP VELIGERS AND CONSE- QUENCES FOR STOCK ENHANCEMENT PROGRAMS, AQUACULTURE AND STOCK MANAGEMENT. Joan L. Manual, Department of Biology. Dalhousie University, Halifax, Nova Scotia. Canada. B3H 4J 1 . The vertical migration behavior of the veliger larvae of the giant scallop (Placopecten magellanicus) is examined as a model for the retention of small organisms in off shore habitat. Veliger responses to thermoclines similar to natural conditions (as low as 1.2°C) varied with time of day and size of the veliger. Veligers from different populations (Georges Bank, Passamaquoddy Bay and Mahone Bay) exhibited different vertical distributions. I pro- pose that vertical migration behavior of the veligers is in part determined by the horizontal transport consequences of the migra- tion, and that population differences are the result of the different hydrographic properties in home regions. Veligers of P. magel- lanicus appear to migrate in response to both light and tidal stimuli. Moon rise early in veliger life may set an internal clock that controls migration at a tidal period, while the diel migration results from responses to changes in light at dawn and dusk. Such a mechanism could be widespread among marine populations, al- lowing planktonic organisms a means of utilising tidal cycles for horizontal transport in offshore regions. I discuss the consequences of such population differences for management of commercially important stocks, aquaculture, and stock enhancement programs. AN EXAMINATION OF THE LINKAGE BETWEEN THE EARLY LIFE HISTORY PROCESSES OF THE SEA SCAL- LOP AND LOCAL HYDROGRAPHIC CHARACTERIS- TICS. Shawn M. C. Robinson, James D. Martin, and Ross A. Chandler, St. Andrews Biological Station. Dept. Fisheries and Oceans, St. Andrews, New Brunswick, Canada, E0G 2X0: G. Jay Parsons, Marine Institute of Memorial University, P.O. Box 4920. St. John's, Newfoundland. A1C. 5R3. The sea scallop, Placopecten magellanicus, is similar to many other scallop species around the world in that large recruitment fluctuations can occur on an annual basis. There have been specu- lations in the literature on why these recruitment pulses occur, but there have been few long-term studies set up to examine this phenomenon. Our team has been studying temporal and spatial scallop recruitment patterns in Passamaquoddy Bay. an enclosed body of water at the mouth of the Bay of Fundy. since 1989. Annual spawning patterns for local populations were monitored through weekly sampling of the gonadosomatic index (GSI ) during the spawning season and biomass surveys using Digby scallop drags were done annually. The resulting spat settlement was moni- tored through the use of standard Japanese onion bags deployed in a uniform grid pattern for 25 sampling stations. These stations were also sampled monthly for temperature, salinity and chloro- phyll a using a CTD. Results indicated that the spatial patterns of settlement were highly consistent from year to year and that the areas with the highest settlement also had the largest animals. The number and size of animals correlated well with wanner areas of the bay as well as with higher chlorophyll a levels. The distribution of larvae (estimated by spat settlement) was different than the distribution of adults and appeared to be more related to physical circulation patterns. Inter-annual variability in the GSI did not appear to explain inter-annual differences in spat settlement. Im- plications to the fishery and culture will be discussed in the pre- sentation. THE USE OF RNA/DNA RATIOS AS AN INDEX OF HEALTH FOR THE SEA SCALLOP {PLACOPECTEN MA- GELLANICUS). Dale Roddick, Ellen Kenchington, and Stephen Smith, Fisheries and Oceans Canada. Bedford Institute of Oceanography. P.O. Box 1006. Dartmouth. NS. B2Y 4A2; Jon Grant, Oceanography Department. Dalhousie University. Halifax. Nova Scotia, Canada. B3H 4J1. The use of RNA/DNA ratios as an index of health has been an active field of research in fish larvae. It has been proposed as a means of monitoring the health of wild Placopecten magellanicus stocks after a die-off occurred in the Digby Nova Scotia stock in 1989. Previous studies have shown both an inter-annual and spatial variation in the RNA/DNA ratios for this stock. This study used a combination of field measurements and laboratory experiments to National Shellfisheries Association, Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 315 examine 3 key pieces of information necessary to determine the suitability of the RNA/DNA ratio in such a monitoring program: 1 ) the seasonal variation present in the scallop population; 2) the response time before a change due to nutritional stress can be detected; and 3 1 the influence of temperature on the RNA/DNA ratio. Seasonal variations were found in the RNA/DNA ratios and could be modeled with a sine function. This model could only account for 22% of the total variation over a 37 month period. Inter-annual variations were also significant, and deviations from seasonal RNA/DNA ratios were significantly correlated with sub- sequent deviations from seasonal growth. Under laboratory con- ditions the RNA/DNA ratio of the adductor muscle provided a detectable response to nutritional stress in 2 to 3 weeks, and growth rates were significantly correlated with RNA/DNA ratios. The RNA/DNA ratio of the mantle tissue did not provide a suitable index of nutritional stress. The RNA/DNA ratio of the adductor muscle does provide an index of health for Placopecten magel- lanicus, but the precision of this index may not be fine enough to discriminate between populations that are under critical stress from those that are stressed but not in any danger of increased mortality rates. PHYSICAL AND BIOLOGICAL VARIABLES INFLUENC- ING THE SPATIAL DISTRIBUTION OF THE GIANT SCALLOP PLACOPECTEN MA GELLAN1CUS. Kevin Stokes- bury. Center for Marine Science and Technology. University of Massachusetts Dartmouth. 706 South Rodney French Boulevard. New Bedford. MA. 02744-1221. The purpose of this study was to determine the physical and biological variables influencing the spatial distribution of the giant scallop. Placopecten magellanicus. Scallops were aggregated on both a large (km) and a small (cm) scale. Large scale aggregations were strongly associated with gravel substrates while small scale aggregations (clumps) were not. The short distance between scal- lops within clumps, the high proportion of clumps with both sexes present, and an average of 3 scallops per clump suggested high fertilization success within clumps. Comparisons of the physical and biological conditions within scallop beds and in adjacent areas with low scallop densities indicated that gravel substratum, low decapod predation. and presence of filamentous flora and fauna were critical factors determining scallop aggregation location. Contrary to previous experimental results, scallops were not safe from decapod predation once they attained a large size. Scallop movement reduced predation rates. Scallop movement was ran- dom, and scallops did not appear to migrate from unsuitable to suitable habitats. However, scallops may move to form clumps resulting in increased fertilization success. If this is true, disturbing this small scale distribution prior to a spawning event may de- crease reproductive success. RE-INTRODUCING THE BAY SCALLOP ARGOPECTEN IRRADIANS INTO CHINCOTEAGUE BAY, MD. Mitchell L. Tarnowski and Mark L. Homer, Maryland Department of Natu- ral Resources. Tawes State Office Building. B-2. Annapolis. MD 21401. Nearly 70 years ago. the bay scallop disappeared from Chin- coteague Bay, coincident with a destructive disease that wiped out the region's eelgrass beds. During the past decade, however, sea- grasses have made a remarkable recovery in this area. With thou- sands of acres of seagrass meadows now in existence and stable, relatively high year-round salinities afforded through the stabili- zation of the Ocean City (Md.) Inlet foremost among several posi- tive parameters, conditions appear to be optimal for the return of the bay scallop. In October 1997, 533,000 seed scallops (8 mm mean length) were introduced into Chincoteague Bay and placed in predator exclusion pens. By the end of the 1997 growing season, the scal- lops had tripled in size and survivorship was about 85%. Over- wintering mortality was extremely low. less than 10%. and by May 1998, the scallops had grown to an average shell length of 30 mm. Survivorship was compromised in August 1998 due to a severe thermal event in the shallower water pen. Overall survival was estimated to be 45% in September and 20% in late November. The surviving scallops attained a size of about 50 mm by November. Two distinct spawning events occurred in 1998, one in May-June, followed by another in August-September. Water column sam- pling revealed the presence of scallop larvae during the summer of 1998. In late October 1998, an additional 610.000 seed scallops were placed into pens in Chincoteague Bay. Initially 20 mm in size, these scallops attained mean shell lengths of 26 mm by early December. Initial survivorship was better than 95%. Samples were taken in December 1998 to determine recruitment success. "Wild" scallops were collected in a number of areas some 8-1 1 km south of the exclosure pens, although at this time it is not known if these were progeny from scallops released in Virginia or from the penned scallops held in Maryland. EVALUATING BAY SCALLOP STOCK ENHANCEMENT EFFORTS WITH MOLECULAR GENETIC MARKERS. Ami E. Wilbur, William S. Arnold, and Theresa M. Bert, Florida Marine Research Institute, 100 8th Ave S.E.. St. Peters- burg. FL 33701. Overfishing, habitat degradation, and toxic algal blooms have all contributed to the collapse of bay scallop (Argopecten irradi- ans) populations in nearshore waters off west-central Florida. Management efforts to halt this decline (regional closures, reduced bag limits) have not resulted in any significant increase in scallop abundance. This lack of natural recovery has led to heightened interest in the enhancement of decimated areas with hatcherv- 316 Abstracts, 1999 Annual Meeting, April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada produced seed. Although such enhancements are becoming a well- recognized component of environmental remediation, few have been objectively evaluated for their efficacy. We have imple- mented an extensive genetic monitoring program to assess the impact of hatchery-seed outplantings on local recruitment in the nearshore waters off west-central Florida. We have developed an array of genetic markers (mtDNA. microsatellites and introns) that can function as a "genetic tag" for the hatchery stocks, and will allow unambiguous identification of the progeny of the outplanted scallops. We are currently monitoring recruitment in and around two sites where approximately 25.000 scallops were planted in the spring of 1998. The new recruits will be typed genetically and the results will allow an assessment of the efficacy of this enhance- ment strategy for facilitating the recovery of bay scallop popula- tions. SHELLFISH BIOCHEMISTRY IDENTIFICATION AND OCCURENCE OF A NOVEL FATTY ACID IN PECTINIDS. Yanic Marty,1 Philippe Soudant,2 Sebastien Perrotte.1 Jeanne Moal,2 Jean Francois Samain,2 Jacques Dussauze'; 'UMR/CNRS 6521. Universite de Bretagne Occidentale, BP 809, 29285 Brest; 2Laboratoire de Physiologie des Invertebres. Ifremer Centre de Brest BP 70. 29280 Plouzane, France: 'Laboratoire d'analyse Brest-Ocean. Pole Ana- lytique des Eaux. 29280 Plouzane, France. A new fatty acid discovered from the scallop Pecten maximus was isolated as a methyl ester using silver nitrate liquid chroma- tography and silicic high performance liquid chromatography. This fatty acid was determined to be a 22:4 geometrical isomer [A4?,7?,10?,13?] (X) using gas chromatography-mass spectrom- etry analysis of their 2-alkenyl-4-4-dimethyloxazoline derivatives (GC/MS DMOX). The number of trans double bounds of this fatty acid was determined on silver ion high-performance liquid chro- matography by comparison with the mono-trans geometrical iso- mers of 22:4 [A4cw,7ds,10cw,13cw] (S). obtained through partial hydrazine reduction of all-cis-22:6(n-3) followed by a p- toluenesulfinic isomerization of the derivatives. Comparing the four trienes (22:3) obtained through partial hydrazine reduction of the two compounds (X) and (S) determined that the trans double bound was located in A 13. This fatty acid identified as m-4,7,10,frare*-13-docosa- tetraenoic has not been reported previously in any organisms. It is believed to be specific for the pectinid family. Mainly in polar lipids, it was found in high proportion in gills and mantle. The new identified fatty acid showed an apparent association with the phos- phatidylserine. FATTY ACIDS FOR REPRODUCTION AND LARVAL DE- VELOPMENT IN TWO BIVALVES MOLLUSCS: POLAR LIPID APPROACH. Jean-Francois Samain, Philippe Soudant, Yanic Marty, and Jeanne Moal, Laboratoire de Physi- ologie des Invertebres, Ifremer Centre de Brest BP 70, 29280 Plouzane, France. Laboratoire de Chimie Marine, URA CNRS 322. UBO 29200 Brest. The aim of this study is to develop an approach of lipid re- quirements for aquacultured marine bivalves during reproduction and larval development. Specificity in PUFA composition of polar lipid classes of gonads, eggs and larvae, was studied during the reproductive cycle in two bivalve species, the scallop Pecten maxi- mus and the oyster Crassostrea gigas in different lipid food com- positions. All the six separated phospholipid classes, demonstrated a specificity in their PUFA composition relatively similar in the two species. Some identical phospholipid classes resisted more to food composition changes of the diet than others. So, compensatory mechanisms exist in peculiar phospholipids, leading to a selective retention of specific essential fatty acids, suggesting their biologi- cal importance for the two species. Specific events occured in the fatty acid composition of the PL classes during gametogenesis and larval development, opening a possible way to characterize essential fatty acid requirements for these two important biological phases. Differences between the two species were also observed and discussed. COMPARISON OF ESSENTIAL FATTY ACID ACCUMU- LATION BETWEEN A REPRODUCTIVE CYCLE IN NA- TURE AND A HATCHERY CONDITIONING OF CRASSOS- TREA GIGAS. P. Soudant," K. Van Ryckeghem," J. Moal,h Y. Marty,c J. F. Samain,b and P. Sorgeloos." 'Laboratory of Aqua- culture & Anemia Reference Centre, University of Gent, Rozier 44, B-9000 Gent, Belgium: bDRV/A, Laboratoire de physiologie des mollusques, IFREMER centre de Brest. BP 70. 29280 Plou- zane, France; CUMR/CNRS 6521. Universite de Bretagne Occi- dentale, BP 809, 29285 Brest, France. Hatchery techniques have been developed based on empirical trials and a basic knowledge of key aspects of artificial reproduc- tion, including broodstock nutrition and timing of gametogenetic cycles, are lacking even for the leading commercial species C. gigas. Lipids deposited in the eggs during broodstock conditioning play a major role as source of energy and essential fatty acid for embryonic and early larval development. In the present study, the lipid content increased and accumu- lated in the gonads during the reproductive phase of the oysters from natural and artificial conditioning but to a higher extend in the naturally-conditioned animals. During that period the neutral lipid percentage of total lipids in the gonad plus mantle was stable National Shellfisheries Association, Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 317 and high (>70%) and equal under both conditions, underscoring that the lipid reserves were preferentially located in that organ. The fatty acid (FA) composition of the neutral lipids in all organs was influenced by changes in the diet but differently ac- cording to organs: high dietary impact occured in the digestive gland whereas the muscle was less affected. The total polyunsatu- rated fatty acids (PUFA) content of the neutral and polar lipids in the gonads changed little whatever the conditions but the respec- tive proportions of (n-3) and (n-6) PUFA differed drastically as a result of diet composition. In all organs, there was a clear evidence for a specific accumulation of 22:6(n-3) and 20:5(n-3) in the polar lipids for both conditionings. Nevertheless the percentages of 22: 6(n-3) and 20:5(n-3) from neutral and polar lipids in the artificial conditioning were significantly lower than those in nature. SHELLFISH DISEASE RECENT TRENDS IN INFECTION OF THE EASTERN OYSTER CRASSOSTREA V1RG1NICA BY THE PARASITE PERKINSUS MARINUS IN THE PATUXENT RIVER ESTU- ARY. Brian W. Albright and George R. Abbe, Academy of Natural Sciences Esturaine Research Center, St. Leonard. MD 20685. Perkinsus marinus is currently distributed throughout oyster populations in Chesapeake Bay and its tributaries, having spread into most low salinity areas of Maryland by the fall of 1992. Since the late 1980s. P. marinus has been the most virulent pathogen of the eastern oyster (Crassostrea virginica) in the Patuxent River owing to its widespread distribution and persistence in low salinity areas. Oysters from ten natural oyster beds in low- to high-salinity areas of the Patuxent River were analyzed annually since 1995 for P. marinus using rectal tissue assays incubated in Rays Fluid Thioglycolate Medium. All sampling was conducted in September yielding maximum values for both prevalence and intensity which are known to occur immediately following maximal summer tem- peratures. No natural oyster beds presently exist in the Patuxent River that are completely free of infection, although salinity con- tinues as the primary environmental factor controlling local distri- bution and intensity. To compare individual beds within the Patuxent the develop- ment of an infection index was needed. The infection index se- lected is the percent prevalence times the weighted intensity (weighted intensity is defined as the mean intensity of the infected individuals in a sample) which yields a number from zero (no infection) to 700 (all individuals maximally infected). Although the overall river wide infection index decreased from 336 in 1995 to 239 in 1996 to 150 in 1997. this decrease was not uniform throughout the river. Beds with initially low levels of infection (principally the most upriver locations) were unable to rid them- selves further of the parasite. While summer salinities and tem- peratures have remained relatively similar over the study period, the abnormally warm winter of 1997 allowed P. marinus to pro- liferate resulting in a river wide mean infection of 333 in 1998. DIVERSITY AMONG PERKINSUS MARINUS ISOLATES FROM THE CHESAPEAKE BAY. Gwynne D. Brown, S. Ko- tob, and M. Faisal, Virginia Institute of Marine Science, College of William and Mary, Gloucester Point. VA 23062. Perkinsus marinus, the causative agent of Dermo, is the pri- mary oyster pathogen of the Eastern oyster (Crassostrea virginica) in the Chesapeake Bay. Recent developments in P. marinus cell culture have enabled scientists to produce several isolates, all be- lieved to be the same species. Variations in virulence among P. marinus isolates from the Chesapeake Bay remain to be elucidated. In this study, seven isolates from the Chesapeake Bay were ex- amined for genetic variations within the internal transcribed spacer regions (ITS1 and ITS2) and 5.8S region of the ribosomal gene unit. Extracellular proteins (ECPs) collected from the seven iso- lates were also compared by SDS-PAGE and analyzed for proteo- lytic activity. Sequence analysis of the ITS-5.8S region showed little variation when compared to the published sequences of P. marinus, confirming that the isolates are P. marinus. Protein pro- files of ECP, revealed by silver staining, showed subtle differences between isolates. Variations in protease activity were detected uti- lizing hide powder azure assay. In addition, differences in the average size of protozoal cells and variations in growth rate were observed. We are currently investigating other phenotypic varia- tions among isolates. TRANSMISSION OF PERKINSUS MARINUS TO INTER- TIDAL OYSTERS. David Bushek and A. J. Erskine, Baruch Marine Field Laboratory. University of South Carolina, George- town, SC 29442; Richard F. Dame, Department of Marine Sci- ence, Coastal Carolina University, Conway. SC 29528; Loren D. Coen and Nancy Hadley, Marine Resources Research Institute, South Carolina Department of Natural Resources, Charleston, SC 29442-2559. Our basic understanding of the processes that control dispersal and transmission of the oyster pathogen Perkinsus marinus is lim- ited. Some studies show that transmission rates decrease rapidly with distance (within 15 m) from infected oysters while others imply that transmissible stages are widely dispersed over long distances (km). A recent model by Powell and co-workers predicts that heavy oyster recruitment can stall an epizootic by diluting the per capita infective 'dose'. Conversely, do reductions in the size of 318 Abstracts, 1999 Annual Meeting. April 18-22, 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada oyster populations enhance infection rates and stimulate epizootics by concentrating the per capita infective 'dose' among remaining oysters'? We examined the roles of dispersal distance and local host population density on infection rate. Specific pathogen free (SPF) oysters were deployed seasonally at distances of 0. 50, 100 and 150 m from the mouths of eight small intertidal creeks, and then monitored for the onset of P. marinus. All native oysters were removed from four of the eight creeks to determine the effect of local host density. Parasites were detected in SPF's in as few as two weeks regardless of season, the presence of native oysters, or distance from the creek mouth. These results indicate that trans- mission occurs throughout the year in South Carolina and that P. marinus is ubiquitously dispersed in this system. UPTAKE, DISTRIBUTION, AND BYCONVERSION OF FLUORESCENT LIPID ANALOGS IN THE OYSTER PRO- TOZOAN PARASITE, PERK1NSUS MARINUS. Fu-Lin E. Chu, Philippe Soudant, Yongqin Huang, Aswani K. Volety, and Georgeta Constantin, Virginia Institute of Marine Science. School of Marine Science, College of William and Mary, Glouc- ester Point, VA 23062. It has been established that host lipids play a unique role for long term survival and life cycle completion in endogenous para- sites. Parasites exploit fatty acids and lipids from the host, not only for membrane synthesis, but also for modification of their surface integrity to avoid host defense. To study lipid utilization by this parasite, we followed the uptake of 'fluorescent-labeled' fatty acid and phospholipid analogs in the merozoite/meront and prezoospo- rangia stages of P. marinus. After 24 hr incubation at 28°C with fluorescent labeled phosphatidylethanolamine (PE). phosphatidyl- choline (PC), and free fatty acid (FFA, C12:0) analogs, yellow- gold fluorescence was present primarily in the discrete lipid drop- lets throughout the parasite. Analysis of lipid class composition using thin layer chromatography showed that the merozoite/ meront stage incorporated PC and transformed it into PE, FFA. triacylglycerol (TAG) and one component tentatively identified as phosphatidylserine (PS) and PE was also metabolized to PS, PC and TAG in meront/merozoite. Incorporation of PC and FFA (CI 2: 0) was observed in prezoosporangia, but there was no conversion of PC to PE or PS, instead PC in prezoosporangia was metabolized into two components presumptively to be FFAs. The fluorescent- labeled CI 2:0 in prezoosporangia was incorporated in TAG. These results suggest that P. marinus incorporated and modified lipids from exogenous sources and that the metabolic modes of meront differed from prezoosporangia. The uptake and bioconversion of fluorescent lipid analogs in the parasite are currently being quan- tified. IDENTIFICATION OF A NEW PERKINSUS SPECIES ISO- LATED FROM MACOMA BALTHICA BY CHARACTER- IZATION OF THE RIBOSOMAL RNA LOCUS. EVIDENCE OF ITS PRESENCE, SIMULTANEOUS WITH P. MARINUS, IN CRASSOSTREA VIRGINICA, MACOMA MITCHELLI AND MERCENARIA MERCENARIA^. Cathleen A. Coss, Jose A. F. Robledo, and Gerardo R. Vasta, Center of Marine Bio- technology. University of Maryland Biotechnology Institute, 701 East Pratt Street. Baltimore, MD 21202; Gregory M. Ruiz, Smith- sonian Environmental Research Center, Edgewater, MD 21037. A Perkinsus species from the Baltic clam Macoma baltluca was isolated and an in vitro culture established under culture con- ditions previously optimized for P. marinus (Gauthier and Vasta. 1995). Examination of the cell morphology and proliferation be- havior of the cultured isolate revealed differences with P. marinus. Regions of the rRNA locus (SSU. ITS1 and ITS2) of this isolate were cloned, sequenced, and compared with those available for other Perkinsus species and isolates. Sequence data from the rRNA locus indicates not only that Perkinsus sp. from M. baltluca is not P. marinus. but also that it is different from P. atlanticus and P. olseni. The degree of difference is comparable or even greater than differences between all three accepted Perkinsus species, P. mari- nus. P. atlanticus. and P. olseni. Therefore, we formally designate the Perkinsus sp. from M. baltluca as a separate species, Perkinsus balthicae. after the bivalve host from which it was isolated and cultured. Using a PCR-based assay specific for P. balthicae and the P. marinus-specific PCR assay, we found that P. balthicae can be present in the eastern oyster Crassostrea virginica. the clams M. mitchelli and Mercenaria mercenaria. and can coexist with P. marinus in all four bivalve species. tThis study was supported by DOC Cooperative Agreement No. NA46RG0091 awarded by NOAA through the Maryland Sea Grant to G.R.V. and a Smith- sonian predoctoral fellowship to C.A.C. INTEGRATED MONITORING OF MARINE DISEASE AND MORTALITY. William S. Fisher, U S Environmental Protection Agency. National Health and Environmental Effects Research Laboratory, Gulf Ecology Division. Gulf Breeze, FL 32561: Benjamin H. Sherman, University of New Hampshire, 211 Nesmith Hall. Durham. NH 03824. There have been apparent increases over the last several de- cades in disease and mortality of marine and estuarine organisms, including shellfish, presumably due to greater anthropogenic stress generated both in watersheds and coastal areas. These events are investigated from a local perspective even though they may have been equally driven, or at least influenced, by regional or global conditions. An ability to link the events with co-occurring physical and chemical disturbances, biophysical characteristics (water qual- ity, harmful algal blooms), hydrographic characteristics, and ma- National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 319 rine-related human diseases can promote understanding of the ef- fects of larger scale and interactive factors, and has potential value for indicating cause and/or environmental condition. Establishing these linkages will require a comprehensive program for collect- ing, documenting, compiling and integrating data collected from a variety of local sources over long periods of time. Two existing programs illustrate the potential to overcome challenges related to data collection, data quality, and database development, as well as establishing and maintaining continuity over time: ( 1 ) The Gulf of Mexico Aquatic Mortality Network, a program of state resource managers, demonstrates the ability to develop consistent investi- gative approaches, standard protocols, and standard data formats that are required to composite data, all the while maintaining state autonomy and ownership of the data. (2) The Health. Ecological and Economic Dimensions of Global Change Program is a retro- spective compilation of disease and mortality events. This rela- tional data framework offers the potential to link events with mul- tiple environmental characteristics (water quality, temperature anomalies, etc.) within the same or external databases, and should serve as a template for future entries. WITHERING SYNDROME IN FARMED RED ABALONE. HALIOTIS RUFESCENS. James D. Moore, Thea T. Robbins. and Carolyn S. Friedman, Bodega Marine Laboratory. P.O. Box 247. Bodega Bay, CA 94923. Withering Syndrome (WS) is a chronic wasting disease respon- sible for mass mortality in wild populations of black abalone [Haliotis cracherodii). The etiology of WS is uncertain with lim- ited evidence for the role of a gastrointestinal rickettsia-like pro- caryote (RLP). Signs of WS and accompanying mortality have been reported in other abalone species, including wild and cultured red abalone (H. rufescens). In the current study, 60 juvenile red abalone (8 cm) were randomly selected from a farmed population raised at 14°C and having low intensity RLP infections. The aba- lone were held in triplicate containers receiving either relatively cool ( 14.7°C. CW) or warm ( 18.5°C. WW) flowing seawater and fed equal amounts for 220 days. Survival was 1007r (30/30) for the CW group and 66<7r (20/30) for the WW group. WW animals had higher RLP infection intensities, showed more clinical signs of WS. and fed at less than half the rate of CW animals. For data from both groups combined. RLP infection intensity was negatively correlated with WS signs including total weight, condition index and digestive gland condition. During 1997-98. several abalone farms in California noticed a dramatic increase in the number of shrunken animals, in conjunction with ENSO-elevated seawater temperatures. Examination of 70 abalone from four farms revealed highly significant correlations between RLP infection intensity and WS clinical signs, validating the laboratory study and strengthen- ing the hypothesis that temperature-enhanced RLP infection plays a direct role in the etiology of Withering Syndrome. Supported in part by Saltonstall-Kennedy grant NA76FD0046 and the Califor- nia Department of Fish and Game. PLASMA OF CRASSOSTREA SPP POSSESS A LOW MO- LECULAR WEIGHT INHIBITOR OF PERKINSUS MARI- NUS PROTEASE. Jacques L. Oliver. Mohamed Faisal, and Stephen L. Kaattari, Department of Environmental Sciences. Virginia Institute of Marine Science. College of William and Mary. Gloucester Point. Virginia 23062. Perkinsus marinus is a protozoan pathogen of the eastern oys- ter. Crassostrea virginica, and is the cause of severe mortalities in eastern oysters throughout the Chesapeake Bay. The eastern oyster is known to be susceptible to P. marinus. however, it has been demonstrated that the Pacific oyster. Crassostrea gigas, is tolerant to this infection. The mechanism(s) of this differential susceptibil- ity to P. marinus is not known. Recent research has implicated serine proteases of the pathogen as likely virulence factors in the progression of the disease. The ability of oysters to produce in- hibitors to pathogen proteases might alter disease progression and thus, be important in oyster defenses against P. marinus. Recently, we have detected the presence of low molecular weight protease inhibitory activity in the plasma of both C. virginica and C. gigas. The role of this inhibitory activity in oyster humoral defense is not presently known. However, we have observed high anti- proteolytic activity in Pacific oysters as well as eastern oysters that have exhibited high survival following natural challenges with P. marinus. These results suggest that protease inhibitors might play a role in oyster host defense mechanisms. Work in our laboratory has focused on the effects of this inhibitory activity on P. marinus proliferation in vitro and the characterization of these putative inhibitors. This research has been supported by NOAA-Sea Grant NA56RG0141. DIAGNOSTIC SCREENING OF OYSTER PATHOGENS: PRELIMINARY FIELD TRIALS OF MULTIPLEX PCR. Soledad Penna and Richard A. French, University of Connecti- cut. Department of Pathobiology. 61 N. Eagleville Rd.. Storrs, CT 06269; John Volk, John Karolus, and Inke Sunila, Connecticut Department of Agriculture. Bureau of Aquaculture and Labora- tory, Milford. CT 06278; Roxanna Smolowitz, University of Pennsylvania. Coastal Research Laboratory, Woods Hole Oceano- graphic Institution. 193 Oyster Pond Rd., Woods Hole. MA 02543. Parasitic diseases are one of the greatest obstacles threatening the success of the oyster {Crassostrea virginica) industry. Two protozoal infections, particularly Haplosporidium nelsoni (MSX) and Perkinsus marinus (Dermo) are the major causes of oyster mortalities and declining oyster production along the Atlantic coast. This situation has necessitated heightened diagnostic testing and initiated the development of disease-resistant oyster strains. 320 Abstracts, 1999 Annual Meeting, April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada However, oyster populations are infrequently monitored for the presence of pathogens. The lack of routine surveillance is attrib- utable to reliance on histopathology for H. nelsoni, Haplospo- ridium costale (SSO) and the Ray/Mackin assay for P. marinus diagnosis. The continued success of the oyster industry requires a practical strategy for diagnosing, monitoring and screening oyster diseases. This strategy would rely on diagnostic aids which are sensitive, rapid, cost effective and convenient. This study presents the development of the multiplex PCR for application in the screening and surveillance of disease agents of the eastern oyster. Multiplex PCR allows for the simultaneous testing of two or more pathogens in a single test reaction. Preliminary results have iden- tified three PCR primer sets which amplify 564-bp. 150-bp. and 304-bp products of H. nelsoni, H. costale, and P. marinus. respec- tively, under equivalent conditions. In multiplex reactions, prod- ucts of control cloned small subunit rRNA of MSX and SSO. and Dermo DNA extracts can be differentiated. The multiplex PCR test has recently been applied to field samples and preliminary findings indicate a high specificity and sensitivity for H. nelsoni, H. costale and P. marinus. ANALYSIS OF THE GEOGRAPHIC DISTRIBUTION OF PERKINSUS MARINUS STRAINS. Kimberly S. Reece, Vir ginia Institute of Marine Science, The College of William and Mary. Gloucester Point. VA 23062: David Bushek and Karen L. Hudson, Baruch Marine Field Laboratory, University of South Carolina, Georgetown. SC 29442. Recently developed /'/; vitro culture and cloning methods have facilitated molecular genetic analyses of the oyster pathogen Per- kinsus marinus. We have examined the genetic relatedness of P. marinus in vitro cultures and their clonal composition using iso- lates collected from throughout the U.S. Atlantic and Gulf coasts. Genetic relatedness of 63 primary cultures and 88 clonal cultures derived from these isolates was examined at six polymorphic loci by restriction fragment length polymorphism analysis. Compari- son of clonal culture and primary culture genotypes indicated that primary isolate cultures from a single oyster can be polyclonal, evidence that oysters can be infected with multiple strains. Het- erozygous genotypes for several clonal cultures suggested that cultured P. marinus cells were diploid, a fact confirmed by DNA sequence analysis of alleles at two anonymous single copy loci in heterozygous clonal cultures. Allele sequences were identical in isolates from geographically distant sites, but allelic and genotypic frequencies differed significantly among the Northeast. Southeast and Gulf coast regions. Overall, there were fourteen different com- posite genotypes detected with more than 80% of the isolates possessing one of three predominant genotypes. One of the major composite genotypes was unique to Gulf coast isolates. Genetic distance analysis indicated three major genotypic clades. each con- taining one of the three major genotypes. LONG-TERM SURVIVAL OF PERKINSUS MARINUS CELLS OUTSIDE ITS HOST. Adel A. Shaheen, Faculty of Vet Med. Zagazig University (Benha branch) — Moshtohor — Egypt. Although Perkinsus marinus has been associated with severe mortalities among eastern oyster (Crassostrea virginica) popula- tions of the Chesapeake bay. little is known about its survival in vegetative form outside host tissues. In the present study, effects of different salinities and temperatures on long-term survival of Per- kinsus marinus cells in seawater are investigated. Perkinsus mari- nus cells were cultured in artificial sea water (ASW) at three temperatures (4. 18 and 27°C at 22 ppt salinity) and four salinities (10. 22. 28 and 35 ppt at 27°C temp). Viability, growth rate. zoosporulation. ability to regrow on growth culture medium, and to form hypnospores in Ray's fluid thioglycolate medium (RFTM) are investigated. Our results suggest that P. marinus culture kept in ASW in 22 ppt salinity at 27°C can survive without exogenous supply of nutrients for more than six weeks. During this time, the parasite divided by both binary fission and successive bipartition. Viability was decreased by 25%. The average cell diameter was increased 30% in ASW than the original cell size. Low percent of sporulation has occurred (0.5-1.5%) with failure of zoospores to be released. Currently, effects of exposure to other different tem- peratures and salinities are investigated. SHELLFISH IMMUNOLOGY: ADAPTATION AND MODULATION XENOBIOTIC-INDUCED IMMUNOTOXICITY IN THE PACIFIC OYSTER, CRASSOSTREA GIGAS: FIELD AND LABORATORY EXPERIMENTS. Michel Auffret and Rad- ouane Oubella, Laboratoire BioFlux UMR 6539. Institut Univer- sitaire Europeen de la Mer, Universite de Bretagne Occidentale. F-29280 PLOUZANE. France. As a part of a monitoring programme of a coastal ecosystem submitted to chronic anthropic influxes, immunomodulation has been studied in a potential sentinel species, the Pacific oyster C. gigas. A double approach has been conducted, including a study of immunotoxic effects of selected environmental contaminants by controlled exposure in the laboratory, and investigations on pos- sible alterations of hemolymph parameters in caged individuals. Standardized biological parameters have been measured in hemolymph samples, including immunopathological parameters (total and differential counts, cell viability, cellular toxicity, serum proteins) and functional parameters of hemocytes (aggregation, phagocytosis, intracellular bactericidal activities). Exposure of oysters in the laboratory to contaminants identified in estuarine sites (trace metals, pesticides) resulted in heavy changes in hemolymph. as hemocytosis and inverted subpopulation propor- tions. Several hemocyte functions related to internal defense were altered. These experiments indicate that most of the xenobiotics tested have potential immunotoxic effects depending on the con- National Shellfisheries Association. Halifax, Nova Scotia, Canada Abstracts. 1999 Annual Meeting, April 18-22. 1999 321 centrations applied. In caged individuals, responses identified in laboratory experiments were also observed. However, most of these responses did not indicate an homogenous trend towards depression since cases of stimulation were found. When examining all the parameters together, an altered overall status of the internal defense system could be established in the sites with the higher contaminant levels. In conclusion, several proofs of immunomodulation revealed in this study indicate that many components and functions of hemolymph in the Pacific oyster are targets for environmental xenobiotics, depressing defense capacities of the individuals. However, as already mentioned by other authors, any study on immunotoxicity in such organisms should included a panel of mea- sures and assays to established if immunosuppression really occurs. EFFECT OF THE NADPH OXIDASE INHIBITOR DIPHE- NYLENEIODONIUM ON THE BACTERICIDAL ACTIV- ITY OF CRASSOSTREA VIRGINICA HEMOCYTES. Lisa H. Bramble and Robert S. Anderson, University of Maryland, Cen- ter for Environmental Science, Chesapeake Biological Laboratory, Solomons, MD 20688. Reactive oxygen species (ROS) produced by eastern oyster hemocytes are hypothesized to serve as bactericidal agents. Inhi- bition by diphenyleneiodonium (DPI) of the putative NADPH oxi- dase responsible for initiating the hemocyte ROS-generating path- way allowed testing of this hypothesis. DPI caused a dose- dependent inhibition of hemocyte ROS produced in response to zymosan; 0.5 p-M DPI abrogated ROS stimulation. Hemocyte phagocytic capability was unaffected by this DPI concentration, suggesting that the effect of DPI on ROS production was specific to NADPH oxidase function. Bactericidal assays were conducted in the presence of 0.5 u.M DPI, using Bacillus megaterium and Pseudomonas fluorescens as target species. DPI had no effect on the ability of hemocytes to kill B. megaterium and reduced bacte- ricidal activity towards P. fluorescens by only 4%. These data suggest that ROS produced by eastern oyster hemocytes do not function as effective bactericidal agents. An alternative hypothesis for the role of ROS in eastern oyster immunology — with interest- ing evolutionary implications — is as immunoregulatory molecules, as has recently been demonstrated in mammalian models. THE EFFECTS OF HYPOXIA AND HYPERCAPNIA ON CELLULAR DEFENSES OF OYSTERS, SHRIMP, AND FISH. Louis E. Burnett, John Boyd, Chris Milardo, Tina Mikulski, and Libby Wilson, Grice Marine Laboratory, Univer- sity of Charleston, 205 Fort Johnson, Charleston, SC 29412; Karen Burnett, Medical University of South Carolina, 217 Fort Johnson, Charleston, SC 29412. Organisms living in estuaries have long been known to have mechanisms that enable them to exist in water containing low amounts of oxygen. The consumption of 02 that generates hypoxia is also responsible for producing significant amounts of C02 (hy- percapnia). An elevation of water CO, causes an acidosis in aquatic organisms. Thus, the combination of low 02 and elevated CO, represents a significant environmental challenge to organisms living in estuaries. The effects of hypercapnic hypoxia on cellular defenses of estuarine organisms has been investigated. Hemocytes of the oyster Crassostrea virginica produce significantly less Re- active Oxygen Intermediates in response to stimulation when held under physiological conditions simulating hypoxia compared to well-aerated normoxia. This response is due separately to specific effects of 02 and the accompanying low pH. Similar results have been obtained using the macrophages of the fish, Fundulus het- eroclitus. These results suggest that hypercapnic hypoxia de- presses the cellular defense mechanisms. To extend these obser- vations to whole organisms, a pathogen challenge model was used to test the hypothesis that hypercapnic hypoxia makes shrimp more susceptible to pathogenic infections of Vibrio parahaemolyticus. The shrimp Palaemonetes pugio and Penaeus vannamei showed lower survival when injected with Vibrio and held under 30% air saturation compared with controls held in well-aerated water. These studies suggest that the generalized innate immune response is depressed in hypoxia and that this contributes to animal mor- tality. (Supported by SC Sea Grant R/ER-14.) EFFECTS OF TEMPERATURE, SALINITY, AND ENVI- RONMENTAL POLLUTANTS ON CELLULAR AND HU- MORAL RESPONSES IN OYSTERS (CRASSOSTREA VIR- GINICA ). Fu-Lin E. Chu, Virginia Institute of Marine Science, School of Marine Science, College of William and Mary, Glouc- ester Point, VA 23062. This paper reviews our recent studies about the effects of tem- perature, salinity, field contaminated sediments (CS) and related soluble water soluble fraction (WSF) on the cellular and humoral responses in relation to Perkinsus marinus infection and progres- sion in oysters. Results suggest that oyster hemocytes, which are believed to be the primary line in the defense system, are not active in defense against the oyster protozoan parasite. P. marinus. Higher numbers of circulating hemocytes. 9c of granulocytes, and phagocytic activity in oysters maintained at high temperature oc- curred concomitantly with higher P. marinus prevalence. Salinity- did not produce significant effects on the above cellular param- eters. Neither temperature nor salinity affected the plasma protein level in oysters. Oysters maintained at different temperatures had relatively similar hemagglutination titers. Plasma lysozyme activ- ity was negatively and significantly correlated with salinity, tem- perature, and P. marinus prevalence. /;; vitro exposure ot hemocytes to water soluble fractions derived from field contami- nated sediments (CS) reduced hemocytes' chemotaxic, phago- cytic, and chemiluminescent responses. CS exposure stimulated neutral red uptake, mitochondrial dehydrogenase production, and 322 Abstracts, 1999 Annual Meeting. April 18-22, 1999 National Shellt'isheries Association, Halifax, Nova Scotia. Canada 3H-leucine incorporation in hemocytes. but did not affect the con- centrations of total circulating hemocytes. plasma protein and lipid, plasma lipid class composition, and plasma lactate dehydro- genase level. Exposure of oysters to 15, and 30% dilutions of WSF for 33 days or 1.0, 1.5, or 2.0 g CS for 30 days significantly elevated the expression/progression of latent P. marinus infection in oysters in a dose-dependent manner. No direct relationship was observed between cellular responses and P. marinus expression in oysters. THE ROLE OF PROTEASE-ANTIPROTEASE INTERAC- TIONS IN PERKINSUS MARINUS INFECTION IN CRAS- SOSTREA SPP. Mohamed Faisal, School of Marine Science, Virginia Institute of Marine Science. The College of William and Mary. Gloucester Point. VA 23062. Perkinsus marinus causes devastating losses in populations of the eastern oyster (Crassostrea virginica). Of particular impor- tance to the invasiveness of protozoan parasites is their elaboration of a spectrum of tissue-disruptive proteases. In this vein, our pre- vious studies have demonstrated that P. marinus secretes extracel- lular serine proteases which enhance parasite propagation and compromise host defenses. Crassostrea virginica, however, has been found to possess several inhibitors of these proteases. Among these, a serine protease inhibitor (serpin) has been identified which specifically blocks proteolytic digestion of oyster protiens. The Pacific oyster (C. gigas) also possesses protease inhibitors with higher specific activities. Interestingly. Crassostrea spp. themselves, elaborate metallo- protease activities which can be detected in their plasma, and are increased during P. marinus infections. Together our work suggest that there may be a broad spectrum of humoral host defenses that is brought to bear on P. marinus infections by these two Crassos- trea species. EXAMINATION OF THE CELLULAR IMMUNE RE- SPONSE OF BLACK ABALONE. HALIOTIS CRACHERO- DIE WITH AND WITHOUT WITHERING SYNDROME. Carolyn S. Friedman, California Department of Fish and Game and Department of Medicine and Epidemiology, Bodega Marine Laboratory. P.O. Box 247. Bodega Bay. CA 94923; Thea Robbins and Jacqueline L. Jacobsen, Bodega Marine Laboratory, P.O. Box 247. Bodega Bay. CA 94923; Jeffrey D. Shields, Virginia Institute of Marine Science. Gloucester Pt., VA 23062. Withering syndrome (WS) is a chronic disease that has resulted in dramatic declines in black abalone abundances along the south- ern and central California coast. A rickettsiales-like procaryote has recently been identified as the likely etiological agent. We hypoth- esized that differences in hemocyte cellular functions may be af- fected by WS and may serve as indicators of early disease. We examined the chemotactic. phagocytic and chemiluminescent abilities of hemocytes from abalone with and without WS. Hemocytes from abalone with WS were more chemotactically ac- tive than those from asymptomatic abalone (n = 35, p < 0.01). However, hemocytes from diseased abalone were less able to en- gulf foreign particles (n = 59, p < 0.01), engulfed fewer particles (n = 52. p = 0.00). and produced a reduced respiratory burst (n = 26, p = 0.00) relative to those from asymptomatic abalone. The immune capability of the hemocytes correlated with degree of WS. Thus, hemocytes from abalone with WS may be more chemotac- tically active as a result of degeneration of the digestive gland and utilization of the foot muscle as an energy source. However, the capability of these stimulated cells to engulf and destroy foreign particles appears to be compromised and may contribute to mor- tality associated with this disease. MODULATION OF EASTERN OYSTER HEMOCYTE AC- TIVITIES BY PERKINSUS MARINUS EXTRACELLULAR PROTEINS. Jerome F. La Peyre, Department of Veterinary Sci- ence. Louisiana State University. Baton Rouge. LA 70808; and Aswani K. Volety, EPA. Gulf Breeze. FL 32561. The oyster pathogen Perkinsus marinus produces many extra- cellular proteins (ECP) in vitro. Analysis of this ECP revealed a battery of hydrolytic enzymes. Some of these enzymes are known to modulate the activity of host defense cells. Although informa- tion on the effects of P. marinus ECP on oyster hemocytes is limited, it has been shown that ECP can inhibit hemocyte motility and hemocyte chemiluminescence response. Moreover, ECP ef- fects on Gulf coast oyster hemocytes are of special concern since these oysters harbor human pathogens such as Vibrio vulnificus and Vibrio parahaemolyticus. Perkinsus marinus infections may thus be associated with higher numbers of Vibrio bacteria in oyster tissues. The objective of this initial study was to better characterize the effects of P. marinus ECP on oyster hemocyte activities in vitro. Increasing concentrations of ECP caused increase degranula- tion and vacuolization of granulocytes (granular hemocytes). The normal spreading of the granulocytes and large hyalinocytes (agranular hemocytes) was also reduced and hemocyte clumping was increased. Phagocytosis of zymosan particles was signifi- cantly decreased at the highest concentration of ECP tested. More important, the capacity of hemocytes to kill Vibrio parahaemolyti- cus was decreased in a dose dependent manner within 30 min of exposure to ECP. P. marinus can clearly be detrimental to oyster hemocytes in vitro. The potential effect of ECP on oyster hemocytes in vivo and the possible relationship between P. mari- nus infection intensities and oyster tissue Vibrio load in Gulf coast oysters remain to be investigated. National Shellfisheries Association, Halifax. Nova Scotia, Canada Abstracts, 1999 Annual Meeting, April 18-22. 1999 323 CHEMICAL EFFECTS ON OYSTER (CRASSOSTREA VIR- G1N1CA) HEMOCYTE MICROBICIDAL ACTIVITY. Leah M. Oliver, Aswani K. Volety,2 and William S. Fisher, U.S. Environmental Protection Agency, National Health and Environ- mental Effects Research Laboratory, Gulf Ecology Division, 1 Sabine Island Drive, Gulf Breeze. FL 32561-5299. ^National Re- search Council Post-Doctoral Associate. Oyster (Crassostrea virginica) hemocytes, or blood cells, per- form important internal defense functions such as phagocytosis and intracellular destruction of pathogens and bacteria. Using tech- niques such as phagocytosis and chemiluminescence assays, po- tential impairment of hemocyte immunocompetence resulting from in vitro and in vivo exposure to anthropogenic chemicals has been demonstrated. A new microbicidal assay recently optimized for oyster hemocytes shows promise for this type of investigation, and may better measure the integrative cidal function of hemocytes compared to measuring discrete portions of the phagocytic pro- cess. Hemocytes were exposed in vitro for 3-18 h to various chemicals including metals, organics, and biotoxins. Tributylin (TBT), previously shown to exacerbate Perkinsus marinus infec- tions when administered to oysters in vivo, inhibited killing of Vibrio parahaemolyticus and cultured P. marinus, at in vitro con- centrations exceeding 32 ppb. The lowest TBT concentration caused a slight elevation, or hormesis, of hemocyte killing activity. Although in vitro results suggest immunosuppression by chemical exposure, previous assessment of defense activities of indigenous oysters collected from Tampa Bay, FL, suggested that these ac- tivities were elevated in oysters with high tissue burdens of certain metals. In a separate study, oysters were deployed at different sites in Pensacola Bay. FL. to test the effect of exposure to chemical mixtures on hemocyte microbicidal activity. Oysters deployed at contaminated habitats tended to have higher hemocyte bactericidal activity (21%) compared to oysters from relatively clean areas (0%). PHAGOSOMAL MECHANISMS IN EASTERN OYSTER (CRASSOSTREA VIRGINICA) BLOOD CELLS. Kennedy T. Paynter, Chesapeake Biological Laboratory and Department of Biology, University of Maryland. College Park, MD 20742. Oyster blood cells appear to have much in common with ver- tebrate macrophages. Like their vertebrate brethren, oyster blood cells are motile cells which locate, engulf and destroy foreign entities including particles, cells and viruses. Studies in our labo- ratory have demonstrated that 1 ) chemiluminescence produced by blood cells is likely caused by the production of hypochlorous acid (HOCll, 2) oyster blood cells contain a myeloperoxidase-like ac- tivity which has an acidic pH optimum and may be responsible for the production of HOC1, 3) phagosomal pH becomes extremely acidic (pH < 4.0) shortly after engulfment of a target particle or cell for extended periods of time, and 4) acidification of the pha- gosome lumen can be blocked by bafilomycin A,, a specific in- hibitor of vacuolar (V-type) proton-pump. While some of these observations are consistent with the mechanisms described in vertebrate macrophage cells, others ap- pear to be more protozoan in character. For instance, the high degree of acidity we have observed in oyster blood cells is matched only by the pH of protozoan acidosomes. By comparing and contrasting mechanisms observed in oyster blood cells with those reported in other organisms, we may gain insight into the nature of immune cell function in general and perhaps trace the evolution of phagocytosis from a purely digestive function into one of defense. INFLUENCE OF SEASONAL FACTORS ON OYSTER HEMOCYTE KILLING OF VIBRIO PARAHEMOLYTICUS. Aswani K. Volety, National Research Council, c/o US EPA; James T. Winstead and William S. Fisher, US Environmental Protection Agency, Gulf Ecology Division. 1 Sabine Island. Gulf Breeze. FL 32561. Seasonal variation of cellular defenses of oyster Crassostrea virginica against Vibrio parahaemolyticus were examined from June 1997 to December 1998 using a recently developed bacteri- cidal assay that utilizes a tetrazolium dye. Mean hemocyte num- bers, plasma lysozyme. and P. marinus infection in oysters were also examined. Hemolymph was sampled from oysters collected at Bayou Texar, Pensacola, Florida once every month. To determine if gametogenic cycle and/or energy reserves affect hemocyte ac- tivity, gonadal stage, tissue total lipids and lipid classes in diges- tive diverticulum tissue were also determined. Hemocyte bacteri- cidal activity and mean hemocyte numbers in oyster hemolymph was higher in warmer months (40-80%) compared to winter months (0-30%). In contrast, total lipid and triglyceride concen- tration in digestive diverticulum tissue was higher in winter months and decreased with spawning in summer months. Lysozyme activity was also higher in winter than in summer months. The role of temperature and salinity on hemocyte killing capacity and plasma lysozyme in oysters is being investigated with the recognition that a variety of influences, including seasonal reproductive changes and potential environmental stimulation, are possible. 324 Abstracts, 1999 Annual Meeting, April 18-22. 1999 National Shellfisheries Association, Halifax. Nova Scotia, Canada SHELLFISH-MICROBIAL INTERACTIONS: ECOLOGICAL AND HUMAN HEALTH PERSPECTIVES IN VIVO TRANSFECTION OF ADULT OYSTERS. John T. Buchanan, Department of Oceanography and Coastal Sciences, Louisiana State University, Baton Rouge. LA 70803; Ta Chi Cheng, Jerome F. La Peyre, and Richard K. Cooper, Depart- ment of Veterinary Science. LSU Agricultural Center, Baton Rouge, LA 70803; Terrence R. Tiersch, Aquaculture Research Station, LSU Agricultural Center, Baton Rouge, LA 70820. We are developing techniques for gene transfer in the eastern oyster, Crassostrea virginica to enhance disease resistance. We transfected oysters with two genes, red-shifted green fluorescent protein (rsGFP). commonly used as a reporter gene, and the lytic peptide cecropin, known to have antimicrobial properties. In a preliminary study, oysters were assigned to three groups: the first was injected with the rsGFP gene mixed with superfect transfect- ing reagent (Qiagen); the second was injected with the cecropin gene mixed with superfect transfecting agent. The third group was injected with saline (control group). Hemolymph was collected at 4 and 10 days after injection. DNA was extracted for polymerase chain reaction (PCR). and hemocytes were examined by flow cy- tometry and fluorescence microscopy for detection of green fluo- rescence due to rsGFP expression. The rsGFP gene was detected by PCR in hemocytes of 13 of 15 oysters injected with this gene at day 4, and 15 of 15 at day 10. The cecropin gene was detected by PCR in 12 of 15 oysters at day 4. and 13 of 15 at day 10. No oysters from the control group were positive for either gene at day 4 or 10. Green fluorescence was detected by flow cytometry in oysters injected with rsGFP at significantly higher levels (P < 0.5) than other oysters. This report indicates the ability to introduce DNA into adult oysters with subsequent gene expression. FACTORS INFLUENCING IN VITRO KILLING OF BAC- TERIA BY HEMOCYTES OF THE EASTERN OYSTER (CRASSOSTREA VIRGINICA). Fred J. Genthner, Leah M. Oliver, and William S. Fisher, U.S. Environmental Protection Agency; Aswani K. Volety, National Research Council. National Health and Environmental Effects Research Laboratory and Gulf Ecology Division. 1 Sabine Island Drive. Gulf Breeze. FL 32561. Vibrio parahaemolyticus strains altered in motility or colonial morphology (opaque versus translucent). Listeria monocytogenes mutants lacking catalase. superoxide dismutase. hemolysin, or phospholipase activities, and Vibrio vulnificus strains, possessing and lacking capsules were exposed to oyster hemocytes. Tetrazoli- zum dye reduction was used to quantify bacterial killing by the hemocytes. Higher killing by hemocytes was observed in summer than winter. Listeria monocytogenes was more resistant to the bactericidal activity of hemocytes than V. parahaemolyticus or V. vulnificus. No differences in hemocyte killing were observed be- tween the different L. monocytogenes mutants. Translucent strains of V. parahaemolyticus showed higher susceptibility to killing by hemocytes than the parental opaque strain. No significant differ- ences in killing by hemocytes were observed between encapsu- lated and nonencapsulated pairs of V. vulnificus. No seasonal dif- ferences (winter versus summer) were observed in activities of 19 hydrolytic enzymes associated with hemocytes. TOTAL BACTERIA AND VIBRIO VULNIFICUS LOAD IN DIPLOID AND TRIPLOID EASTERN OYSTERS IN LOUI- SIANA. Jerome F. La Peyre, Richard K. Cooper, Department of Veterinary Science. John E. Supan, Office of Sea Grant De- velopment. Louisiana State University, Baton Rouge. LA 70808; Aswani K. Volety, EPA. Gulf Breeze. FL 32561. Advantages of triploidy in oysters include greater growth rate and better meat quality. It has also been postulated that triploid oysters have better host defenses; energy allocated to reproduction in diploid oysters may be allocated to host defenses in triploid oysters with impaired gonad development. The extended spawning season of Gulf coast oysters and the occurrence of the human pathogen Vibrio vulnificus (V.v.) make triploidy attractive for this region. In this preliminary study we compared the total bacteria and V.v. load between diploid and triploid oysters. The capacity of hemocytes to kill V. v. and Vibrio parahaemolyticus ( V.p. ) was also measured in these oysters. Oysters were obtained from the Grand Isle Oyster Hatchery, LA, and divided into three groups, each composed of 10 diploid and 15 putative triploid oysters. The first group was kept overnight on ice, a second group was left outside of the water for 42 h at 25°C and a third group was maintained for one week at 20°C in U.V. -treated seawater. Each oyster was then homogenized and the number of colony forming units (cfu) of total bacteria and V.v. was determined. In all groups, triploid oysters had significantly lower total bacteria and V.v. cfu than diploid oysters. Hemocytes from triploid oysters had significantly greater bactericidal activity against V.r. and V.p. than hemocytes from diploid oysters. This data must be interpreted with caution since diploid and triploid oysters did not originate from the same broodstock and were not raised at the same site for most of their grow-out period. REGULATION AND MANAGEMENT OF WATER QUAL- ITY TO PRESERVE SHELLFISH HARVESTING AND HU- MAN HEALTH IN TOMALES BAY, CALIFORNIA. Paul G. Olin, University of California Sea Grant Extension. Santa Rosa. CA 95403; Gregg Langlois. California Department of Health Ser- vices. Berkeley. CA 94704. Shellfish growers in Tomales Bay, California experience sig- nificant economic losses annually as a result of rainfall triggered harvest closures. As a result, in 1993 the California State Legis- National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 325 lature passed the Shellfish Protection Act to protect water quality in shellfish growing areas. The act requires formation of a tech- nical advisory committee to identify and remediate pollution sources. In 1994. the Tomales Bay Shellfish Technical Advisory Committee was formed and a subsequent study identified exces- sive coliform loading from animal agriculture as the primary source of pollution. The committee has developed research and management objectives designed to improve water quality in the Bay that include fencing, manure pond treatments, increased hold- ing capacity and adjusted wastewater irrigation schedules. The focus on water quality broadened to include human fecal contamination in 1998 when a viral outbreak involving over 150 people was linked to consumption of raw oysters contaminated with a human pathogen. This resulted in a closure of Bay waters to shellfish harvest for 26 to 81 days. The outbreak triggered an intense examination of water quality and potential pollution sources. These activities and other steps were undertaken to insure consumer safety prior to resuming shellfish harvesting. This illus- trates the need to be proactive in preserving water quality in our heavily used coastal areas. HEMATOLOGY OF BLUE CRABS, CALLINECTES SAPI- DUS, INFECTED WITH THE PARASITIC DINOFLAGEL- LATE HEMATODINWM PEREZ1. Jeffrey D. Shields and Christopher M. Squyars, Department of Environmental Sci- ences. Virginia Institute of Marine Science. The College of Will- iam & Mary. P.O. Box 1346. Gloucester Point, VA 23602. Along the eastern seaboard, the blue crab, Callinectes sapidus, experiences recurring epizootics of the parasite. Hematodinium perezi- We investigated host mortality due to the disease, assessed differential hematological changes in infected crabs, and examined proliferation of the parasite. Healthy, mature female crabs were injected with 103 or 105 cells of//, perezi. Mortalities began 14 d after infection, with a median time to death of 30.3 ± 1.5 d (se). and a mortality rate of 86% in infected crabs. Hemocyte densities declined by 60-80% within 3 d of infection and exhibited differ- ential changes in subpopulations of granulocytes and hyalinocytes that lasted throughout the infection. Injected crabs that did not present infections exhibited significant long-term (21-27 d) granu- locytemia. Patency increased from approximately 30% after 14 d to 60% after 21 d to 100% after 35 d. Plasmodial stages were, however, detectable in histological preparations of the heart within 3 d of infection with significant increases over 5 and 7 d. The mortality studies indicate that H. perezi represents a significant threat to the blue crab fisheries in high salinity estuaries, and may have a greater effect on mature females that move to higher sa- linities to breed. DETECTION OF PROTOZOA PATHOGENS IN THE EASTERN OYSTER TAKEN FROM THE GREAT BAY ES- TUARY. Kim M. Stowell, Stephen D. Torosian, and Aaron B. Margolin, University of New Hampshire. Durham. NH 03824. Many outbreaks of human gastroenteritis are caused by the consumption of raw or undercooked shellfish in which protozoa pathogens such as Giardia lamblia and Cryptosporidium parvum are present. These pathogens are introduced into the waterways via direct human fecal contamination, animal reservoirs and farm run- off. Adult oysters filter approximately 100 gallons of water per day and concentrate protozoa on gills, in hemocytes and in the intes- tinal track. These pathogens remain in the oyster for at least a month after exposure and remain infectious in mice one week after ingestion. The water in Great Bay has the disadvantage of being the drop off site for many rivers that carry wastewater from treatment plants. There are several oyster beds located in the vicinity of these rivers as well as a major site abutting a wildlife refuge. The current method of detection is the immunofluorescence assay (IFA), which requires microscopy to distinguish between the two organisms and may also antigenically cross-react with various species of algae. This research utilizes a molecular approach to detection by the development of a multiplex polymerase chain reaction (PCR), which detects Giardia and Cryptosporidium si- multaneously. Inhibitors to PCR are overcome by the use of streptavidin-coated paramagnetic beads and biotinylated primers. Results derived from seeded oyster extract have shown that using multiplex PCR is an efficient, and extremely specific, means of protozoa detection. Eventually a multiplex can be designed which will detect both viral and protozoa pathogens simultaneously. This data and methodology will be used to improve shellfish testing and combat a public health threat. VIBRIO FLUVIALIS IMPLICATED IN RECENT OUT- BREAKS AMONG AMERICAN LOBSTERS. B. D. Tall,1 M. Crosby,3 D. Prince,' J. Becker.5 G. Clerge,1 4 D. Lightner,2 L. Mohney,2 M. Dey,1 F. M. Khambatv,1 K. A. Lampel.1 J. VV. Bier,1 B. E. Eribo,4 and R. Bayer,3 'JIFSAN. U.S.FDA, Wash- ington. DC 20204. 2Univ. of Arizona. Tucson, AZ 85721, 'The Lobster Inst., Univ. of Maine. Orono. ME 04469, 4Howard Univ., Washington, DC 20059, 5Maine Lobster Technologies, Hancock, ME 04640. An unexplained, highly invasive disease has emerged in Homarus americanus (American lobsters) harvested from Atlantic coastal waters. Economic losses exceeding $2.5 million threaten the $136 million-a-year industry. Gram-negative bacilli were ob- served in hemolymph samples from diseased lobsters; results from antibiotic therapy studies showed that enrofloxicin was highly ef- fective (100% survival) in abating illness in naturally diseased lobsters and lobsters experimentally infected with hemolymph from diseased animals. Culture of hemolymph samples from 5 of 326 Abstracts, 1999 Annual Meeting. April 18-22, 1999 National Shellfisheries Association. Halifax. Nova Scotia, Canada 6 diseased lobsters yielded bacteria, of which Vibrio fluvialis (Vf) was the predominant microorganism. The isolates were highly susceptible to a variety of antibiotics tested. PFGE analysis showed that most of the isolates either shared a common DNA fingerprint or possessed minor variants thereof; two could not be typed. Seven isolates possessed plasmids. A sheep hemagglutinin was found to be expressed by 93% of the isolates. This suggests the presence of cell-associated proteases or adherence factors. In- vasion studies using Atlantic menhaden liver cells demonstrated that the Vf strains were capable of invasion, irrespective of plasmid presence. Our results indicate that this illness was likely caused by a cohort of highly related, invasive Vf strains that expressed a sheep hemagglutinin. The emergence of this pathogen, capable of infecting fish and humans, and reported now for the first time in Crustacea poses a significant health and economic threat that mer- its additional studies. DIFFERENTIAL EFFECTS OF OYSTER (CRASSOSTREA VIRGINICA) DEFENSES ON CLINICAL AND ENVIRON- MENTAL ISOLATES OF VIBRIO PARAHAEMOLYTICUS. Aswani K. Volety, National Research Council, c/o US EPA: Fred J. Genthner and William S. Fisher, US Environmental Protection Agency, Gulf Ecology Division, 1 Sabine Island, Gulf Breeze. FL 32561; Susan A. McCarthy, US Food and Drug Administration, PO Box 158, Dauphin Island. AL 36528; Kirk Wiles, Texas De- partment of Public Health. 1100 West 49th Street. Austin, TX 78756. Three clinical (2030. 2062. and 2107) and three environmental (1094, 1 163. and ATCC 17802) isolates of Vibrio parahaemolyti- cus were exposed to hemocytes and plasma collected from oysters (Crassostrea virginica) to determine their susceptibility to putative oyster defenses. Clinical strains were isolated from patients who became ill from the June 1998 food poisoning outbreak of V. parahaemolyticus associated with oysters harvested from Galveston Bay, Texas. Detection of thermolabile direct hemolysin (tlh) and thermostable direct hemolysin (tdh) genes in isolates was conducted using alkaline phosphatase- and digoxigenin-Iabeled probes. All isolates of V. parahaemolyticus possessed the till gene while only the clinical isolates had the tdh gene. Although pulse- field gel electrophoresis revealed that all clinical strains were iden- tical, isolate 2062 was more susceptible to killing by oyster hemocytes than the other clinical isolates (2030, 2107). Overall, environmental isolates were more susceptible to hemocyte killing than clinical isolates. However, environmental isolates 1163 and ATCC 17802 had higher susceptibility to hemocyte killing than isolate 1094. When grown on nutrient agar plates, all strains dis- played different colonial morphologies. Examination of cells of clinical isolates using electron microscopy did not reveal differ- ences in degree of encapsulation. Bacterial susceptibility to hen egg white lysozyme and oyster plasma was investigated. Results indicate that while clinical strains are genetically identical, they may differentially express putative factors responsible for protec- tion against killing by oyster hemocytes. POSTER SESSION BAFILOMYCIN A, INHIBITS ACIDIFICATION OF GRANULAR AND AGRANULAR OYSTER HEMOCYTE PHAGOSOMES. Amy E. Beaven, University of Maryland. Col- lege Park. MD 20742; and Kennedy T. Paynter, Chesapeake Biological Laboratory. Solomons. MD 20688. Recently, we have shown that the phagosomes of eastern oyster (Crassostrea virginica) hemocytes, like vertebrate macrophages, become acidified after engulfment of foreign particles. Acidifica- tion of intracellular compartments in a variety of cell types and organisms is typically accomplished by proton-pumping mecha- nisms. The vacuolar (V-type) ATPase, which is found in macro- phage phagosomes and lysosomes. generates a proton gradient w ith the concomitant hydrolysis of ATP. In order to determine the mechanism of phagosomal acidification in oyster hemocytes, we exposed hemocytes to bafilomycin A, (Baf) a specific inhibitor of the V-type ATPase. Oyster hemocytes were incubated with Baf and then challenged with zymosan labeled with both rhodamine and fluorescein. The pH of internalized zymosan was determined using the ratio of the emission intensities of rhodamine and fluo- rescein. Zymosan internalized within phagosomes of both granular and agranular hemocytes exposed to Baf became much less acidi- fied than the control hemocytes. Maximal inhibition of both granu- lar and agranular hemocyte phagosomal acidification occurred at 5 u,M. and partial inhibition (40%) occurred at doses as low as 50 nM. Inhibition of hemocyte phagosomal pH reduction suggests a V-type ATPase may be responsible for acidification of the phago- some lumen in both granular and agranular oyster blood cells. AN ECOSYSTEM MODEL OF PERKINSUS MARINUS. Jodi Brewster and David Bushek, Baruch Marine Field Laboratory. Baruch Institute for Marine Biology and Coastal Research. Uni- versity of South Carolina, Georgetown, SC 29442; Richard F. Dame, Department of Marine Science, Coastal Carolina Univer- sity. Conway, SC 29528. The mechanisms that drive and maintain Perkinsiis marinus epizootics are obscure. Relationships to temperature and salinity have been well documented and predictive models using these parameters exist. Still, much uncertainty remains. Modeling pro- vides a tool to develop a better understanding of the forces that control epizootics. An important use of models is to highlight areas that lack solid information. Most studies have considered the in- National Shellfisheries Association, Halifax, Nova Scotia, Canada Abstracts, 1999 Annual Meeting. April 18-22, 1999 327 teractions between the eastern oyster and P. marinus outside of the framework of the ecosystem in which they co-exist. We con- structed a box model with Stella 5.0 software to show the flow of P. marinus through a salt marsh oyster dominated ecosystem — the North Inlet Estuary, SC. Our goal was to identify major processes that control parasite transmission and the development of epizoot- ics in order to identify those processes where management strat- egies can be designed to minimize the risk or impact of an epi- zootic. The model reiterates the importance of temperature and salinity, but also highlights the role of estuarine flushing in main- taining a balanced equilibrium between host and parasite. The model also indicates that information is lacking or considerably limited on the seasonal concentration of transmissible stages in the water column and, perhaps more importantly, measures of the relative importance of the various potential fates of water borne P. marinus. EVALUATION OF A GLUCOSE OXIDASE/PEROXIDASE METHOD FOR INDIRECT MEASUREMENT OF GLYCO- GEN CONTENT IN OYSTERS (CRASSOSTREA VIR- GIN1CA >. Shelley Burton, Allan Mackenzie, T. Jeffrey David- son, and Audrey Fraser, Atlantic Veterinary College. University of Prince Edward Island, 550 University Ave., Charlottetown, PEI, Canada. CIA4P3. A colorimetric method for indirect measurement of glycogen (utilizing glycogen conversion to glucose by amyloglucosidase) in tissue homogenates of oysters {Crassostrea virginica) was evalu- ated. The procedure was optimized as to extracting buffer pH (5.0) and amyloglucosidase concentration (5 mg/ml). Coefficients of variation (n = 10) for oyster homogenates with mean glycogen concentrations of 84 and 242 mg/dL had within-run values of 3.28 and 3.65%, and between-run values of 4.49 and 3.15%. respec- tively. When mean glycogen concentrations of thawed oyster ho- mogenates were compared to those of initial fresh homogenates. no significant (P < 0.05) differences were detected in samples thawed after 1 hour. 1 day, 1 week or 1 month. Glycogen recovery percentages of 104.1, 103.7 and 104.5% were obtained with mixed solutions containing 111, 94 and 19 mg/dL. As dilutions of an oyster homogenate with a high glycogen concentration (430 mg/ dL) gave observed results within 5% of expected, the assay was considered to be linear to 430 mg/dL. The lower limit of sensitivity was approximately 14 mg/dL. An initial laboratory range (mean +/-2 SD = 3-50 mg/g) for tissue glycogen based on wet weights (1.6-3.9 g) was determined for 49 two-year-old oysters obtained during July from the Covehead region of Prince Edward Island, Canada. It was concluded that the colorimetric assay offered a reliable indication of tissue concentrations of glycogen in Cras- sostrea virginica. TUNICATE FOULING IN NOVA SCOTIA AQUACUL- TURE: A NEW DEVELOPMENT. Debbie Cayer and Marli MacNeil, Aquaculture Association of Nova Scotia, Halifax, Nova Scotia; Andrew G. Bagnall, Nova Scotia Department of Fisheries and Aquaculture, Halifax, Nova Scotia. A significant fouling problem with the tunicate Ciona intesti- nalis developed on a suspended blue mussel (Mytilus edulis) aqua- culture site during the 1997 growing season. Ciona intestinalis has subsequently colonized several aquaculture sites at varying levels of intensity. The infestation has had a serious, negative impact on at least one farm and could potentially affect the economic viabil- ity of other operations. The Aquaculture Association of Nova Scotia and the Nova Scotia Department of Fisheries and Aquaculture have been net- working internationally, documenting the impacts of C. ibtestinalis on aquaculture production and searching for strategies to manage its effects. Information has been collected and contacts have been made around the world — a solution has not been found. Methods to manage this fouling problem will be species and site specific. Attempts to remove the tunicates have been cost prohibitive on a large scale. Farm management strategies must be developed as an economically viable solution. Part of this process involves studying the behaviour of C. intestinalis and its interac- tions with Nova Scotia operations. Cliona intestinalis biology and ecology are also being examined in an effort to prevent the further spread of these tunicates to previously unaffected aquaculture sites and to assist operations with implementing mitigative measures. USE OF FLOW CYTOMETRY AND HISTOLOGY TO AS- SESS GAMETOGENESIS IN TRIPLOID CRASSOSTREA ARIAKENSIS. Whitney Chandler, Aimee Howe, and S. K. Allen. Jr., Aquaculture Genetics and Breeding Technology Cen- ter, Virginia Institute of Marine Science, Gloucester Point, VA 23062. Triploid Crassostrea ariakensis were produced by treating fer- tilized diploid eggs with cytochalsisn B. Seventy-seven 5-year-old adults were conditioned and sacrificed at peak gametogenesis. Flow cytometry of gill samples confirmed that all animals were triploid. Eighty four percent (n = 65) were male and the rest hermaphrodites. Almost all-male follicles still had spermatogonia, but the majority of cells were spermatids. Spermatozoa were in all "'ripe'" follicles. Flow cytometry confirmed that the majority of cells in 63 of 65 male triploids were fully reduced. 1.5 n. No haploid gametes were detected. Most of the hermaphrodites had female like follicle structure, characterized by small undeveloped follicles with undifferentiated gametic cells. In these "female-like" individuals, hermaphroditism was detected only by the presence of completely reduced (1.5 n) cells via flow cytometry. Four her- maphrodites had typical female follicle structure, but some of the follicles had "erupted into maleness." One hermaphrodite was 328 Abstracts, 1999 Annual Meeting. April 18-22. 1999 National Shellfisheries Association. Halifax. Nova Scotia. Canada clearly a male, with a number of follicles containing one to a few ripe ova. Flow cytometry and histology in combination were pow- erful tools in assessing gonadal and gametic maturation, in par- ticular, the detection of hermaphrodites. IMPROVEMENT OF THE WHOLE-OYSTER PROCE- DURE FOR ENUMERATING PERKINSUS MARINUS IN OYSTER TISSUES. Gregory M. Coates, School of Forestry. Wildlife and Fisheries. Richard K. Cooper and Jerome F. La Peyre, Department of Veterinary Science, Louisiana State Univer- sity, Baton Rouge. LA. The whole-oyster procedure for enumerating P. marinus in oyster tissues is the most reliable technique to date to determine infection intensity. This procedure depends on the enlargement of the parasites in Ray's fluid thioglycollate medium (RFTM), their subsequent isolation from tissue debris and finally their quantifi- cation after staining with Lugol's solution. Each one of these steps however has yet to be optimized adequately. Insufficient parasite enlargement, incomplete tissue digestion during isolation, parasite clumping and adherence to the walls of centrifuge tubes can lower parasite counts. In this initial study, we report on the use of various chemicals to alleviate some of these problems and increase both precision and accuracy of the whole-oyster procedure. Parasite enlargement was significantly enhanced by addition of pronase or a lipid mixture (i.e., cod liver oil) to RFTM. The size range of parasites also decreased with these two treatments. Nei- ther pronase nor the lipid mixture caused proliferation of cultured parasites in RFTM. Tissue debris remaining after sodium hydrox- ide digestion of oyster tissues could be further eliminated by rins- ing the parasite suspension in hydrochloric acid solution. More- over, the acid treatment intensified staining of the parasite with Lugol's solution and eliminated fading. Finally, mucin was effec- tive in preventing parasite clumping. The use of these chemicals made counting of the parasite easier and increased the number of parasites detected. Future work should focus on developing a tech- nique to expedite parasite counting. POTENTIAL USE OF RAYS FLUID THIOGLYCOLLATE MEDIUM TO DETECT AND QUANTIFY PERKINSUS MARINUS IN ENVIRONMENTAL WATER SAMPLES. Rebecca C. Ellin and David Bushek, Belle W. Baruch Institute for Marine Biology and Coastal Research. University of South Carolina. P.O. Box 1630, Georgetown, South Carolina 29442. Perkinsus marinus is a major pathogen of the Eastern oyster. Crassostrea virginica. Direct transmission of P. marinus occurs via the water column, however, processes that influence planktonic transmission dynamics remain poorly understood. Accurate meth- ods for quantification of planktonic P. marinus are necessary to study transmission dynamics. Available methods include quanti- tative competitive PCR and immunoassay flow cytometry. These methods are expensive and technologically complex. Moreover, cross-reactivity of the antibody probe with other estuarine organ- isms has recently been confirmed. We examined Ray's Fluid Thioglycolate Medium (RFTM) assay as an alternative. The RFTM assay is an inexpensive, accurate methodology to assess P. marinus infection intensities within oyster tissue, hemolymph. feces, and pseudofeces. We detected planktonic stages of P. marinus in environmental water samples using RFTM. Recovery of in vitro cultured P. marinus from spiked water samples was low. but significantly higher using filtration (743,667 ± 106.444) rather than centrifugation methods (37,339 ± 9.385; p < 0.05) to capture parasites. The low cell recovery could not be attributed to various stages of the RFTM method but may be explained by the inability of cultured P. marinus to enlarge in RFTM. This is supported by a significantly higher cell recovery rate with the addition of lipids to RFTM (p < 0.05). Differences between cultured and environmental P. marinus limit the applica- tion of the RFTM method recovery efficiency to natural popula- tions. Therefore, the recovery of naturally occurring planktonic P. marinus in environmental water samples using the RFTM meth- odology will be compared with existing methodologies to deter- mine efficiency and to investigate other sources of cell loss. CORRELATION BETWEEN THE LEVEL OF PROTEASE INHIBITORS AND INTENSITY OF PERKINSUS MARINUS INFECTION IN EASTERN OYSTER (CRASSOSTREA VIR- GINICA ). Ehab Elsayed and Mohamed Faisal, School of Marine Science, Virginia Institute of Marine Science, The College of Wil- liam and Mary, Gloucester Point, VA 23062. Perkinsus marinus is a protozoan parasite that causes a severe loss in eastern oyster (Crassostrea virginica) populations. Pro- teases of P. marinus are believed to play an important role in its pathogenicity. Recent studies in our laboratory demonstrated the presence of PI in the plasma of eastern oysters. The role these inhibitors play in P. marinus infection remains to be elucidated. In this study, the correlation between P. marinus infection intensity and protease inhibitors (PI) level in oyster plasma has been inves- tigated. Crassostrea virginica naturally infected with P. marinus were collected from Points of Shoal. James River, VA. Parasite burden was determined by Ray's fluid thioglycolate medium (RFTM). PI activities against P. marinus extracellular proteases were measured. A statistically significant increase in levels of PI was observed in oysters with low infection burden. Heavily in- fected oyster showed no measurable PI activities. In another set of experiments 10 Crassostrea virginica families were compared for levels of PI. Our results suggest a correlation between these oyster plasma PI levels, parasite burden and mortality rates. National Shellfisheries Association. Halifax. Nova Scotia. Canada Abstracts. 1999 Annual Meeting. April 18-22. 1999 329 EFFECTS OF BROODSTOCK AND LARVAL DIETS ON LIPID AND FATTY ACID COMPOSITION OF SEA SCAL- LOP {PLACOPECTEN MAGELLANICUS) EGGS AND LAR- VAE IN RELATION TO CULTURE OPTIMIZATION. Scott C. Feindel, Ray Thompson, Pat Dabinett, and Christopher Par- rish. Ocean Sciences Center. Memorial University of Newfound- land. St. John's. NF A1C 5S7. Previous studies have indicated the potential importance of lipids and essential fatty acids (EFA) in the diet of some marine bivalves. To investigate this link with regard to egg quality and larval performance in sea scallops (Placopecten magellanicus), feeding trials using live algal cultures with broadly differing EFA composition, specifically eicosapentaenoic acid (20:5n3). docosa- hexaenoic acid (22:6n3). and arachiodonic acid (20:4n6), were carried out on broodstock and larvae. Preliminary results indicate that lipid and EFA composition of eggs and larvae was influenced by diet. However, egg quality in terms of hatchability to D-stage did not correlate with the treatments. This could be due to the transfer of previously stored nutrients into the developing gonad and/or to the conservation and preferential incorporation of EFA in the eggs in the required amounts for normal early development. Differences in larval growth and survivability are partly related to lipid and EFA composition of the diet. Although the treatments consisted of only four species of algae and the biochemical com- position of algae is known to vary with culture conditions, the results may be indicative of a diet with an appropriate EFA ratio and lipid composition that can be used to enhance culture success. Optimal broodstock conditioning may depend on other factors. GROWTH, MORTALITY AND BIOCHEMICAL CON- TENT OF THE PACIFIC OYSTER, CRASSOSTREA GIGAS, DURING SPAT-ADULT DEVELOPMENT. Zaul Garcia- Esquivel, Marco A. Gonzalez-Gomez, and Dahen L. Gomez- Togo, Instituto de Investigaciones Oceanologicas. Universidad Autonoma de Baja California. Ensenada. Baja California. 22800. Mexico. Changes in shell height, tissue dry weight (TDW). mortality and biochemical content were evaluated in Crassostrea gigas dur- ing the spat to adult development in San Quintin Bay (SQB), NW Mexico. Shell growth rate was 1.5x greater at the mouth (0.36 tnm/d) than the head of SQB and tissue growth rate was also 5-fold higher at the mouth (8.6 mg TDW/d). Oysters located at the head of SQB consistently suffered longer emersion periods (17%) than the ones at the mouth. The market size (9 cm) was reached after 8 mo. (mouth) and 13 mo (head) post-setting. Cumulative mortality during the study ranged between 63 and 87% but the highest mortality (50 to 60%) took place within the first mo. post-setting, during the rearing period. Proteins, lipids and carbohydrates were present in proportions of 48-64%; 2-9% and 1-22% respectively, with lower values associated to winter months and longer aerial exposure. Proteins and lipids were the most abundant biochemical components during early spat development, but lipids were re- placed by carbohydrates as the main energy source in the Fall, when oysters reached 50 to 64 mm. Based on the growth rates and biochemical patterns, it is concluded that site-specific differences in developmental rates were mainly due to longer emersion times at the head of SQB. DISEASE DYNAMICS: MODELING THE EFFECT OF CLI- MATE CHANGE ON OYSTER DISEASE. Eileen E. Hof- mann and John M. Klinck, Center for Coastal Physical Ocean- ography. Crittenton Hall, Old Dominion University, Norfolk. VA 23529: Susan E. Ford and Eric N. Powell, Haskin Shellfish Re- search Laboratory. Rutgers University, Port Norris. NJ 08349. Models that simulate the host-parasite-environmental interac- tions of Eastern oysters (Crassostrea virginica) and the pathogens. Haplosporidium nelsoni, which causes the disease MSX, and Per- kinsus marinus. which causes the disease Dermo, have been de- veloped. Both diseases have been epizootic in Delaware and Chesapeake Bays during the 1990s. The models are physiologi- cally-based and are structured around the transmission, prolifera- tion and death rates of the two parasites. Environmental conditions of temperature, salinity and oyster food supply provide the external forcing that results in variations in these biological rates. Simula- tions that use time series characteristic of warm and cool periods show the advantage given both parasites by warm conditions and demonstrate the importance of a sequence of warm and dry years in triggering wide spread epizootic conditions of the two diseases. Results of these simulations can be used to understand the causes underlying the northward spread of these two oyster diseases along the east coast of the United States, from Chesapeake Bay to Maine, in the decade of the 1990s. The models can also be used to re- structure the practices of the oyster industry to maximize produc- tion under conditions where the life span of the commercial spe- cies is controlled by disease. REFERENCE RANGES FOR CHEMICAL AND CELLU- LAR CONSTITUENTS OF HEMOLYMPH FROM "HEALTHY" LOBSTERS (HOMARUS AMERICANUS). Bar- bara S. Horney, Allan L. Mackenzie, Richard J. Cawthorn, Claude C. Morris, Larry K. Hammell, and Robert MacMillan. Atlantic Veterinary College, University of Prince Edward Island. Charlottetown. PEI CIA 4P3. Reference ranges for blood chemical and cellular components in the "healthy" population of interest are required to identify alterations in individuals in association with disease and disorders. 330 Abstracts, 1999 Annual Meeting, April 18-22. 1999 National Shellfisheries Association. Halifax, Nova Scotia. Canada The purpose of this study was to determine reference ranges for various constituents of hemolymph from visually healthy lobsters {Homarus americanus). Hemolymph was collected from the ven- tral sinus from 5 different groups of freshly caught, pound held or laboratory maintained animals. The mean weight and carapace length of these groups ranged from 298 to 614 grams and 72 to 90 cm respectively. Chemical assays included measurement of the concentrations of sodium, potassium, chloride, calcium, phospho- rus, magnesium, glucose, cholesterol, lactate and total protein and the activities of amylase, alanine amino transferase, aspartate amino transferase, alkaline phosphatase, lactate dehydrogenase and arginine kinase on an automated chemistry analyzer. Total solids protein was measured by refractometer and freezing point osmolality was also measured. Hemocyte numbers were evaluated manually and by an automated cell counter. The results are re- ported as mean +/- SD of each parameter for each group of lob- sters. Comparisons of assay results are made between male and female animals, between groups, and between test methods. This collected data of hemolymph parameters in health is valuable in the identification of hemolymph alterations associated with disease or poor survivability of lobsters. SPATIAL TRENDS FOR TOXIC CONTAMINANTS IN MYTILUS EDULIS FROM THE GULF OF MAINE. Stephen Jones, Jackson Estuarine Laboratory. University of New Hamp- shire. Durham, NH 03824; Margo Chase, Department of Zoology, Miami University. Oxford, Ohio, 45056; John Sowles, Depart- ment of Environmental Protection, Augusta. ME 04333; Peter Hennigar and Peter Wells, Environment Canada, 45 Alderney Drive. Dartmouth. NS B2Y 2N6. Gulfwatch is a program which has used Mytilus edulis for 7 years as the sentinel species for habitat exposure to toxic contami- nants in the Gulf of Maine. Mussels were collected at 59 sites in all five jurisdictions for spatial analysis. Tissue was analyzed for trace metals, pesticides. PAHs and PCBs. and growth and condi- tion index were measured. Results showed a southward trend of increasing concentrations for organic contaminants and both silver and lead, reflecting major local and regional pollution sources. Other trace metals showed relatively uniform distribution. Con- centrations were elevated at sites previously assumed to be uncon- taminated. while other areas appear to be hot spots. Gulfwatch has been useful to managers in a variety of ways, and it provides unique. Gulf-wide information that helps focus efforts to reduce contaminant loadina. PREDICTION OF BLUE MUSSEL (MYTILUS EDULIS) FAILURE LOAD. William P. Ireland, Department of Anatomy and Physiology. Atlantic Veterinary College. University of Prince Edward Island. Charlottetown, PEI CI A 4P3; T. Jeffrey Davidson and Laurie McDuffee, Department of Health Management, At- lantic Veterinary College. University of Prince Edward Island. Charlottetown. PEI CIA 4P3. Ten blue mussels {Mytilus edulis) were collected from each of seven mussel growing areas in Prince Edward Island, measured (length, width, height, thickness and weight) and subjected to a compressive load applied at a constant (0.07 cm/sec) rate to failure by a materials testing machine. Elements of the machine include an hydraulic cylinder capable of exerting 5000 lbs force (>2.000.000 Newtons) mounted above a load cell connected via an AID interface to a 386 computer. Half shells were placed between the cylinder and loaded to failure by the cylinder. At 50 msec intervals the computer registered the compressive load and saved time step, distance traveled and load to a file. Correlations were calculated between shell dimension and fail- ure load because we predicted that shell geometry was a factor in shell strength. Regression analysis was done to derive an equation to predict breaking strength from shell measurements. Correlations of strength with shell dimension and weight were significant. Two measures, weight and length taken together significantly predicted shell failure load. This predictive ability was consistent between regions sampled. OYSTER RESTORATION IN MARYLAND: MEASURING PROGRESS AND PRODUCTIVITY. Timothy Koles and Kennedy T. Paynter, Department of Biology. University of Maryland. College Park. MD 20742. Shellfish restoration in coastal waters has become increasingly recognized as an important part of estuarine ecosystem manage- ment. In 1997, in cooperation with the Maryland Department of Natural Resources and the Army Corps of Engineers, five sites in both the Choptank and Patuxent Rivers, extending from the mouth of each river to approximately eight miles upstream, were identi- fied for restoration. At each site, fossil oyster shells were deposited in a configuration of two .5 acre flat areas and one mound ap- proximately three to four meters high. Some of these areas were then planted with hatchery reared spat (1 million/acre; 247/nr) while the rest were left unplanted. Monitoring of these sites was accomplished using divers to obtain quadrat samples from each of the fiats and mounds and YSI 6000 continuous water quality moni- tors to measure ambient water temperature, salinity, pH, and dis- solved oxygen. The samples were then analyzed for oyster size, abundance, mortality, fouling community, and parasite (Perkinsits marinus) prevalence and intensity. Preliminary analyses have shown similarities among oysters from the two rivers as well as some differences. In both rivers, the oysters appeared to be growing vigorously. Parasite prevalence was very low with only a few oysters in each river infected with P. marinus. Mortality was also low and was mostly associated with small oysters that appeared to be overcrowded on the shells. In National Shellfisheries Association. Halifax. Nova Scotia, Canada Abstracts, 1999 Annual Meeting. April 18-22. 1999 331 addition, the unplanted mounds in both rivers recruited higher numbers of natural spat set than the unplanted flat areas nearby. These results will be used to evaluate the impact of differing bottom morphology (mound vs. flat) and differing water charac- teristics (i.e. salinity) on oyster recruitment, growth, mortality, and disease pressures. DETECTING THE PRESENCE OF PERKINSUS MARINUS IN THE EASTERN OYSTER, CRASSOSTREA VIRGINICA, IN RHODE ISLAND WATERS. Ken Leonard HI and Marta Gomez-Chiarri, Fisheries. Animal, and Veterinary Sciences. Uni- versity of Rhode Island. Kingston RI 02881; Arthur Ganz, Rhode Island Department of Environmental Management. Coastal Fish- eries Lab. Wakefield. RI 02879. Dermo and MSX, caused by the parasites Perkinsus minimis and Haplosporidium nelsoni respectively, have been responsible for oyster mortalities throughout the east coast of the United States. The goal of our research is to survey the prevalence and intensity of Dermo and MSX in Eastern oysters from Rhode Island waters. Oysters (30 per site) were collected from 8 locations in Rhode Island from May to November 1998. including 2 aquacul- ture sites. The prevalence and intensity of Dermo infection were evaluated using the Ray's Fluid Thioglycolate Medium (RTFM) method and by examining histological sections. MSX infections were evaluated using histological examination. We have detected the presence of Perkinsus marimis and Haplosporidium nelsoni in Rhode Island oysters. The intensity and prevalence of Dermo dis- ease significantly varied between locations and sampling season. Highest Perkinsus marinus infection levels were seen in the Au- gust sampling session, and may be responsible for the oyster mor- talities observed at several sites. The information obtained through this study will be useful to regulatory agencies in their manage- ment of the oyster populations in Rhode Island waters. INVESTIGATIONS INTO TREATMENTS TO CONTROL FOULING ORGANISMS AFFECTING OYSTER PRODUC- TION. Neil MacNair and Matt Smith, Prince Edward Island Department of Fisheries and Tourism, Fisheries and Aquaculture Division, P.O. Box 2000. Charlottetown, Prince Edward Island, Canada. CIA 7N8. Fouling organisms are an increasing challenge for oyster pro- duction on Prince Edward Island (PEI). Canada. A recently intro- duced green algae species. Codium fragile (the oyster thief), has the potential to interfere with the bottom culture and wild harvest of oysters. In an effort to slow down the transfer of the organism into Codium free areas, the Department investigated the effects of pre-transfer immersion treatments, on affected shellfish, using saturated brine and 4% hydrated lime in an attempt to kill Codium. Results indicate that the plant is very difficult to eradicate using these treatments and that it takes a long period of time to assess their efficacy. As well, the summer of 1997 proved to be an ex- ceptional year for set of the sea grape, Molgula sp., which fouled oyster spat collectors and resulted in spat mortalities. The Depart- ment investigated the use of saturated brine and 4% hydrated lime to control Molgula sp. The study concluded that a three minute dip in saturated brine or a one minute dip in 4c/c hydrated lime effec- tively killed Molgula sp. on collectors without causing mortalities of the oyster spat. SESTON DYNAMICS AND FOOD AVAILABILITY IN A MUSSEL SYSTEM (GULF OF GAETA, SOUTHERN TYR- RHENIAN SEA, ITALY). Antonio Mazzola, Tiziana La Rosa, Benedetto Savona. and Gianluca Sara, Marine Biology and Re- source Laboratory, Dept. Animal Biology, University of Palermo, Via Archirafi, 18, 1-90123 Palermo. Italy. Spatial and temporal changes in the biochemical composition of suspended organic matter in the Gulf of Gaeta (southern Tyr- rhenian Sea. Italy), were investigated during a one-year period in order to assess the origin and nutritional value of POM for cultured suspension feeders [Mytilus galloprovincialis). Water samples were collected monthly from January to December 1997 at the surface and near the bottom (-12 m) and analyzed for total sus- pended matter, suspended pigments, particulate carbohydrate, pro- tein and lipid concentrations. The biopolymeric fraction of par- ticulate organic carbon was defined as the sum of carbohydrate, protein and lipid carbon and used as an index of the particulate organic matter readily available to benthic consumers. Mean total suspended matter concentration was 4.8 ± 3.9 mg l"1, while inor- ganic suspended material concentrations were high, indicating a possible dilution effect of the particulate organic matter. Phytopig- ment concentrations of 1.2 ± 0.91 u.g 1~' lead us to define the waters of Gulf as meso-trophical. Particulate biopolymeric organic- carbon concentrations were quite low (on average 120 ± 85 p-g C l-1) and no significant difference was detected in comparison with a control station outside the Gulf. Proteins were the dominant class in the particulate matter (40%). followed by carbohydrates (34%) and lipids (26%); POM represented on average only 7% of total suspended matter. These results do not substantiate the fact that cultivation of mussels is a normal commercial activity in the Gulf (yielding about 200-300 t per year). The role of physiological compensation in the maintenance of a relatively constant food absorption rate in environments which are characterized by time- varying resources could be invoked. EFFECTS OF FOOD QUALITY ON THE PARTICLE HAN- DLING TIME IN BIVALVES. Lisa M. Milke and J. Evan Ward, Department of Marine Science, University of Connecticut. Groton, CT, 06340; Sandra E. Shumway, Southampton College. LIU. Southampton. NY 1 1968. Suspension-feeding bivalves may assume a large ecological role by linking benthic and pelagic systems. Therefore, a knowl- edge of feeding processes is necessary to fully understand bivalve 332 Abstracts. 1999 Annual Meeting, April 18-22. 1999 National Shellfisheries Association, Halifax, Nova Scotia, Canada dominated environments. In this study, we examined the pallial cavity residence time for Mytilus edulis and Crassostrea virginica. By measuring residence times and subtracting previously calcu- lated particle transport rates found on the gills of mussels and oysters, handling time on the labial palps can be determined. Bivalves were offered one of three food types: Rhodomonas lens cells, Spartina alterniflora particles, or a 50/50 mixture of both. Once actively feeding, bivalves were delivered 10 u.m poly- styrene beads as a tracer and feeding continued. Bivalves were sampled at 30 s intervals for 0 to 150 s and placed in liquid nitrogen, ensuring the cessation of particle transport. Digestive systems were then isolated and examined for the presence of tracer beads. For mussels, it appears that food quality has little effect on handling time. At the 30 s interval, 10-30% of the animals had tracer beads in their gut, increasing to 90% by the 150 s time interval. Assuming a residence time of 60s (25-50% of animals) for mussels with average gill lengths of 2.8 cm (used in our study), handling time on the palps would be 12 s. Therefore, estimated labial palp transport rates are indicative of slurry as opposed to mucus transport. Results of this and parallel studies with oysters will be utilized to create a numeric model of pallial processes using the modeling software Stella. HOW TO PROVIDE ESSENTIAL NUTRIMENTS TO BI- VALVES IN HATCHERV. J. Moal,1 C. Seguineau.1 J. F. Sa- main,1 P. Soudant,2 M. Cansell,3 J. R. LeCoz,' H. Migaud,1 M. Sanies,1 B. Ponce,' C. Langdon.4 'ifremer. Centre de Brest, BP 70, 29280 Plouzane. "Virginia Institute of Marine Science. 1208 Greate Road, Gloucester Point. VA 23062. 31STAB, Universite Bordeaux I, av des Facultes. 33405 Talence. 4Hatfield Marine Science Center, 2030 S. Marine Science Drive. Newport, OR 97365. A large variability is observed in essential nutriment composi- tion between eggs from different spawnings and from different females of Crassostrea gigas oysters. It depends on the food re- gime of broodstock. We have tested different ways to modify the egg and the larval composition, either by supplementation of the food regime of adults and larvae by emulsions, liposomes or spray beads, or by direct modification of egg composition by lipofection technique. Emulsions (INVE Aquaculture: ICES 30/0.6/C) were used for lipophylic supplementation. Marine liposomes of a size range adapted to bivalve ingestion capacities (enrichment possibility: lipids from their membrane and an aqueous internal volume (Cansell in prep.)). The similarity of their membrane lipid com- position with that of mollusc membranes is a possible way to increase the lipofection process. Spray beads for food supplemen- tation of aqueous and lipophylic molecules (Buchal and Langdon (in press)). First experiments were conducted to determine if these differ- ent particules were ingested and digested by C. gigas at different developmental stages. Empty liposomes were radiolabeled on the membrane and mixed to an oocyte preparation to study the lipo- fection possibilities. Results are presented and discussed. EXPERIMENTAL DREDGING OF STARFISHES AND CRABS BEFORE COMMERCIAL SEEDING OF SEA SCALLOPS IN MAGDALEN ISLANDS (QUEBEC, CANADA). Madeleine Nadeau and Georges Cliche, Ministere de l'Agriculture, des Pecheries et de l'Alimentation du Quebec, Direction de 1' Innovation et des Technologies, C.P. 658. Cap-aux- Meules, Quebec. GOB 1 B0: Denyse Hebert, Association des pecheurs de petoncles des Iles-de-la-Madeleine, C.P. 516, Etang- du-Nord, GOB I E0. Commercial bottom seeding of juvenile sea scallops (Pla- copecten magellanicus) are performed annually in the Magdalen Islands (southern Gulf of St-Lawrence) since 1993. In 1997, the areas seeded in 1993. 1994 and 1995 were re-opened to commer- cial fishing. About 6% of the scallops seeded in 1993 and less than 1% of those seeded in 1994 and 1995 seeding have been recap- tured. Predation by starfishes and crabs appeared to be a major factor in the loss of scallops. Recapure rates in Japan were in- creased from 5.5% to more than 30% by removing with dredges predators before seeding. Therefore we tried such an operation of predators removal by dredging just before seeding in spring 1998. After 230 hours of dredging, more than 200,000 predators (92% starfishes and 8% crabs) were removed from a 2.25 km" during this operation. Starfishes were mostly represented by Solaster en- deca (50%), Leptasterias polaris (25%) and Crossaster papposus (20%). Hyas sp. composed 70%- of the crabs. An inventory realised before dredging with a video camera suggested that more than 25% of the predators were taken off during the operation. This capture level was comparable to Japanese results. The impact on scallops survival and the commercial interest of such operation will be evaluated more precisely at the reopening of this seeded area, in 4 years. MUSSEL CULTURE POTENTIAL IN SOUTHERN MO- ZAMBIQUE. F. Ribeiro, F. Simoes, and L. Swenarchuk, Insti- tuto de Investigacao Pesqueira. P.O. Box 4603, Maputo, Mozam- bique. From 1976 to 1989. the Mozambican government carried out a study into mussel (Perna perna) aquaculture. with support from CUSO-SUCO. a Canadian NGO. Natural coastal settlements, found south of Inhambane. served as seed for transfer to culture ropes and growout on long-lines or existing fixed structures in Maputo and Inhambane Bays. Growth was excellent in both sites. National Shellfisheries Association. Halifax, Nova Scotia, Canada Abstracts, 1999 Annual Meeting, April 18-22, 1999 333 Repeated failure of natural spatfall, however, proved to be a criti- cal obstacle blocking development of industrial-scale culture. This problem was subsequently resolved using hatchery tech- niques. Spawning of wild mature adults was artificially induced by temperature shifts in the Maputo Laboratory, and veliger larvae (5 larvae/ml) were fed on Isochrysis galbana and Tetraselmis suecica. growing to 300 p, as metamorphosed spat within 19-23 days, before settling on collectors made from discarded multifila- ment fishing nets. After incubation in outdoor lm3 tanks for 4-6 weeks, seed mussels had reached 2-3 mm. and were transferred to growout ropes, reaching 9-10 cm in 11 months. Inhambane and Maputo bays are suitable sites for integrated enterprises including hatcheries and growout zones, with estimated minimum potential production of 0.56 x 106MT/Year and 10 x 103MT/Year (wet weight), respectively. Moreover, culture in the interior regions of Inhambane Bay may ultimately reduce the need for hatchery-supplied seed, as reproduction from the cultured ani- mals becomes self-sustaining. Major target markets should be the coastal populations, already accustomed to consuming mussels, and the rapidly growing shrimp culture industry. EUTROPHICATION CONTROL BY BIVALVES: POPULA- TION FILTRATION, SEDIMENTATION AND NUTRIENT REMOVAL THROUGH SECONDARY PRODUCTION. Michael A. Rice, Department of Fisheries. Animal and Veterinary Science, University of Rhode Island. Kingston, RI 02881; April Valliere, Mark Gibson, and Arthur Ganz, Rhode Island Divi- sion of Fish and Wildlife. Coastal Fisheries Laboratory. 1231 Suc- cotash Rd., RR#1. Wakefield, RI, 02897. Filter feeding by populations of bivalves has been suggested as a means of reducing eutrophication in coastal estuaries by exerting control of phytoplankton populations in the water column. Fre- quently large populations of mature shellfish reside behind pollu- tion closure lines in estuaries represent a large filter feeding bio- mass. The standing crop of quahogs. Mercenaria mercenaria, in the Providence River averages 9.1 clams/m2 or about 26.400 tonnes, filtering about 1 .5 x 107 m3 of water daily or a rate equiva- lent to 24% of the rate of water exchange during a tide cycle. The population of quahogs, however, is composed of mostly older adults with valve lengths in excess of 60 mm. These large animals are slow growing, have a low rate of secondary production in relation to standing crop biomass. and have a neutral nitrogen balance (organic-N assimilated = NH,-N excreted). These large adults increase sedimentation through filter feeding, but since they are neither harvested nor growing they do not directly remove much nitrogen from the system, although the increased sedimen- tation rates may result in increased sediment denitrification. Smaller more rapidly growing quahogs have the capability of in- corporating organic nitrogen into growing tissues and if harvested regularly provide a mechanism for direct removal of nitrogen from the estuary. As part of a Narragansett Bay wide shellfisheries management plan, 10% of the standing crop of quahogs in the Providence River is recommended for relay to management beds down bay for later harvest. At this level of relay effort, 2,640 tonnes of shellfish would be moved to harvest beds annually, representing about 25 tonnes of nitrogen removal from the estuary if these were eventually harvested. The removal of quahogs from the dense assemblages in the Providence River reduces the popu- lation filtration by only 10%, but it culls the population making room for faster growing juveniles and small adults. This is publi- cation 3682 of the Rhode Island Agricultural Experiment Station. University of Rhode Island. DEVELOPMENT OF A PCR-BASED DIAGNOSTIC ASSAY FOR A NOVEL PERK1NUS SPECIES ISOLATED FROM MACOMA BALTIC AA Jose A. F. Robledo, Cathleen A. Coss, and Gerardo R. Vasta, Center of Marine Biotechnology. Univer- sity of Maryland Biotechnology Institute, 701 East Pratt Street, Baltimore, MD 21202. The non-transcribed spacer (NTS) at the rRNA locus of a new Perkimiis species isolated from the baltic clam Macoma balthica was amplified, cloned, sequenced, and compared to that of the sympatric species P. marinus. Length and sequence of the NTS are dramatically different from that of P. marinus, supporting the con- clusion based on the analysis of the SSU, ITS1 and ITS2, that the clam isolate constitutes a distinct species. Based on the NTS se- quence a set of primers was designed for a PCR-based diagnostic assay that specifically amplifies Perkinsus sp. from M. balthica. Because the host specificity of P. marinus and Perkinsus sp. has not been determined, this assay, together with the one previously developed and validated for P. marinus, should contribute to ad- dressing not only this question, but issues related to the role of other invertebrate or vertebrate species as putative reservoirs, in- termediate hosts or vectors for both parasite species. fThis study was supported by DOC Cooperative Agreement No. NA46RG0091 awarded by NOAA through the Maryland Sea Grant to G.R.V. THE NEW WESTERN MEDITERRANEAN ENTRY BRACHIDONTES PHARAONIS (FISCHER P., 1870) (BI- VALVIA, MYTILIDAE): CHANGES IN FILTRATION RATE UNDER VARYING NATURAL FOOD CONDI- TIONS. Gianluca Sara, Chiara Romano, and Antonio Mazzola, Marine Biology and Resource Laboratory. Dept. Animal Biology. University of Palermo. Via Archirafi. 18. 1-90123 Palermo. Italy. The hvperhaline bivalve mollusc Brachidontes pharaonis (B. pharaonis) is a common Indo-Pacific species that has recently reached the western Mediterranean throueh the Suez Channel. The 334 Abstracts. 1999 Annual Meeting, April 18-22. 1999 National Shellfisheries Association, Halifax, Nova Scotia. Canada most western MED finding of B. pharaonis has been documented in a saltworks in western Sicily. Here. B. pharaonis has colonized hard substrates and now represents the most important suspension filter feeder of the area. There is very little information on the population dynamics, feeding habits and behaviour of B. phara- onis. Seasonal changes in food availability (May-December 1998) were investigated in order to assess the relationship with somatic growth and filtration rate of B. pharaonis. The filtration rate was measured seasonally in a flowing open system and calculated by taking the difference between the suspended organic matter con- centrations in inflow and outflow water samples. During the study period, water temperature and salinity ranged from 8°C to 27°C in winter and summer respectively, and 44 psu and 58 psu in spring and summer respectively. Food concentration expressed as total suspended organic matter was found to be highly variable (form 2 to 200 mg I-1). Filtration rate was significantly and negatively correlated to food concentration, varying between 1.4 and 7.6 1 g mussel-1 h~'. Lastly, the role of the quality and quantity of organic matter on the performance of B. pharaonis and the reasons for of its absence in the adjacent basin (Stagnone di Marsala) are dis- cussed. PARTIAL CULTURE AND CRYOPRESERVATION OF THE PARASITIC DINOFLAGELLATE HEMATODINIUM PEREZ! FROM THE BLUE CRAB. Jeffrey D. Shields, De partment of Environmental Sciences, Virginia Institute of Marine Science. The College of William & Mary, P.O. Box 1346, Glouc- ester Point. VA 23602. Hematodinium perezi is a parasitic dinoflagellate that infects blue crabs along the eastern seaboard of the United States. Cur- rently, the parasite can only be maintained in the laboratory via serial injection. In vitro culture and cryopreservation of the patho- gen were attempted. Isolates were held in a balanced salts buffer at temperatures of 4°. 15° and 20°C. The main stages in the life cycle of H. perezi were observed, but only partial completion of the life cycle was attained. Under culture conditions, plasmodia developed into schizonts but plasmodial budding and schizogony were not completed. Trophonts (vegetative stages) progressed to highly vacuolated presporonts which rarely sporulated to became dino- spores. Primary cultures of the parasite lived for up to 16 d at 4°C. 28 d at 15°C and 7 d at 20°C. Naive crabs acquired infections when inoculated with cultures that had been maintained in vitro for 14 d. but those inoculated with reconstituted cryopreserved samples did not acquire infections. Parasites reconstituted from cryopreservation were alive, but did not grow in culture, nor were they infectious. Recovery of live parasites was significantly higher in glycerol than in dimethyl sulfoxide. Successful culture and re- constitution of cryopreserved H. perezi from the blue crab will require a better support medium. INFECTION BY THE DINOFLAGELLATE PARASITE HE- MATODINIUM IN THE NORWAY LOBSTER (NEPHROPS NORVEGICUS L.) ON THE WEST COAST OF SCOTLAND, UNITED KINGDOM. Grant D. Stentiford and D. M. Neil, Di- vision of Environmental & Evolutionary Biology, IBLS. Univer- sity of Glasgow, Glasgow. G12 8QQ. Scotland, UK.: R. J. A. Atkinson, University Marine Biology Station Millport. Isle of Cumbrae. Scotland UK. The Norway Lobster (Nephrops non'egicus L.) is the most commercially important shellfish in the UK, with a first sale value of -£63 (~$105m US) in 1997. the major fishery being in Scot- land. Stocks of Nephrops off the west coast of Scotland are known to harbour infection by the parasitic dinoflagellate Hematodinium. The infection shows a clear seasonal pattern, with peak prevalence in the spring and may also show an overlying long-term epidemi- ology. Prevalence levels reach 80% in some years and seem to be greatest in post-moult animals. Infection staging is possible by observing the degree of parasite accumulation in a pleopod. and by immuno-cytoehemistry using a polyclonal antibody raised against Hematodinium. Recently, certain biochemical parameters of the host haemolymph have been identified as infection indicators and work is continuing in this area. The parasite itself causes gross histopathological changes to most host tissues and organs, and it is suggested that severe infections lead to a considerable physiologi- cal compromise in Nephrops. Such effects on host metabolism may be expected to cause alterations in the behaviour of Nephrops in the field, and may reduce its ability to evade trawl capture or predation. Understanding of the life cycle of Hematodinium has increased following its successful culturing in the laboratory, and such knowledge may help to elucidate the as-yet unknown mode of infection in Nephrops. It may also help to identify possible sec- ondary or alternative hosts of this parasite on Nephrops grounds and lead to measures for its control in both Scottish Nephrops and other commercially important crustacean fisheries around the world. SEARCHING FOR THE PUTATIVE MSX INTERMEDI- ATE HOST USING MOLECULAR DIAGNOSTICS. Nancy A. Stokes, Brenda Sandy Flores Kraus, and Eugene M. Burre- son, Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, VA 23062; Kathryn A. Ashton-Alcox and Susan E. Ford, Haskin Shellfish Research Laboratory, Rut- gers University, Port Norris, NJ 08349. Forty years after its introduction to the eastern United States, the complete life cycle of the oyster pathogen Haplosporidium nelsoni. the agent of MSX disease, still has not been elucidated. Attempts to infect oysters directly with H. nelsoni spores have been unsuccessful, leading to speculation that parasite transmis- sion between oysters occurs via an obligate intermediate host. Our National Shellfisheries Association. Halifax. Nova Scotia, Canada Abstracts, 1999 Annual Meeting. April 18-22, 1999 335 H. Hf/so/ii'-specific polymerase chain reaction (PCR) and in situ hybridization (ISH) diagnostic assays have been optimized for use with environmental samples and are being used in the search for the putative intermediate host(s). Samples of water and sediment fractions and of macroinvertebrates were taken from MSX- endemic areas of York River. VA and Delaware Bay for three years. Total genomic DNA was extracted from each sample and subjected to PCR amplification. Samples that yielded H. nelsoni PCR product were more frequent from the York River than from Delaware Bay, consistent with MSX disease prevalence in oysters from these locations. PCR-positive samples were subjected to ISH to allow visualization of parasite infections to discriminate be- tween true infections and those where H. nelsoni simply adhered to the external surface or passed through the gut. PRELIMINARY EVALUATION OF TRIPLOID AMERI- CAN OYSTERS, CRASSOSTREA VIRGINICA, ON A MID- ATLANTIC OYSTER FARM. Stewart M. Tweed, Rutgers Co- operative Extension of Cape May County. Cape May Court House. N.J. 08210: Ximing Guo, Haskin Shellfish Research Laboratory, Rutgers University, Port Norris, NJ 08349. In 1996. a commercial oyster farm was established in New Jersey with oysters from Rutgers Highly Selected Resistant Lines (HSRL). The success of these lines prompted interest in develop- ing HSRL triploids in order to produce faster growing better qual- ity market oysters. Triploid and Diploid American Oysters. Crassostrea virginica, were produced from a HSRL stock at the Haskin Shellfish Re- search Laboratory in the summer of 1998. In July. 1 to 2 millimeter seed oysters were placed in upwellers at the Atlantic Capes Fishery Oyster Farm. These seed were maintained on the farm and moved to nursery areas in the Fall and growout sites in the Spring. Growth and survival of Triploid and Diploid cohorts were com- pared when they were moved from the upweller, and at three different nursery and growout sites at 5 and 8 month intervals. Results are being used by the Oyster Farm to evaluate growing areas and triploid potential for improved production and market- ing. PROTEOLYTIC ACTIVITY FROM BLOOD CELLS OF THE EASTERN OYSTER, CRASSOSTREA VIRGIMCA. Gregory Ziegler, Marine. Estuarine, Environmental Sciences, University of Maryland. College Park. MD 20742; and Kennedy T. Paynter, Chesapeake Biological Laboratory and Department of Biology, University of Maryland. College Park. MD 20742. A protease was isolated from the eastern oyster, Crassostrea virginica, hemocytes and characterized using substrate impreg- nated sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and spectrophotometric assays. Hemocytes were iso- lated by centrifugation and homogenized in the presence of SDS at pH values ranging from 4.4 to 7.4. Crude homogenates were loaded into vertical slab gels and run at 30 niA for approximately 1 hour. Proteolytic activity was revealed as clear bands on Coo- massie Blue stained, substrate impregnated gels. Proteolytic activ- ity was found in acrylamide gels impregnated with gelatin, but activity was absent in gels impregnated with other substrates in- cluding casein, fibrin, fibronectin and laminin. Molecular weight of the enzyme responsible for the single proteolytic activity band was estimated to be 87 kDa when analyzed by SDS-PAGE. Ac- tivity was maximal at a pH of 7.4 on gelatin impregnated gels. A suite of inhibitors was employed to determine the mechanistic- class of the isolated protease and indicated that the proteolytic activity was likely a cysteine protease. Spectrophotometric assays were performed to confirm protease characteristics and to ascertain a Vmax and Km for the enzyme-catalyzed reaction. The lack of additional proteolytic activity, even at lower pH extractions and incubations, is surprising given that most cells are thought to have a suite of proteases that serve various and different functions. 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 care- fully prepared in the style followed in prior issues of the Journal of Shellfish Research (1991) before submission 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 or in italics 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 Publi- cations 16 and 17: Common and Scientific Names of Aquatic Invertebrates from the United States and Canada: Mollusks and CSNAWSC: Decapod Crustaceans, or relevant publi- cations for other geographic regions. Abbreviations, Style, Numbers: Authors should follow the style recommended by the sixth edition (1994) of the Council of Biology Editors [CBE] Style Manual, distributed 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 or laser print and planned so that important details will be clear after reduction 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 prepared 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 repro- duced without reduction. Originals of graphic materials (i.e.. line drawings) are preferred and will be returned to the author. Each illustration 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 alpha- betically at the end of the paper. Abbreviations in this sec- tion should be those recommended in the American Stan- dard 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 a recent issue of the Jour- nal 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. Students (only if first author and a member of NSA) will not be assessed page charges. 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: Appropriate photographs may be submitted for consideration for use on the cover of the Jour- nal of Shellfish Research. Black and white photographs and color illustrations will 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@southampton.liunet.edu Membership information may be obtained from the Edi- tor or the Treasurer using the form in the Journal. Institu- tional subscribers should send requests to: Journal of Shell- fish Research, P.O. Box 465, Hanover, PA 17331. 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: Mailimi 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 Nancy Lewis, Bookkeeper, PO Box 350, V.I.M.S. Eastern Shore Lab, Wachapreague. VA 23480, USA. MBI. WHOl LIBRARY ,AAK 4 Cesar J. Lodeiros, Jose Jesus Rengel, and John H. Himmelman Growth of Pteria colymbus (Roding. 1798) in suspended culture in Golfo de Cariaco, Venezuela 1 55 Kim J. Friedman and Paul C. Southgate Growout of blacklip pearl oysters, Pinctada margaritifera collected as wild spat in the Solomon Islands 159 Inke Sunila, John Karolus, and John Volk A new epizootic of Haplosporidium nelsoni (MSX), a Haplosporidian oyster parasite, in Long Island Sound. Connecticut 1 69 Nicole T. Brun, Andrew D. Boghen, and Jacques Allard Distribution of the turbellarian Urastoma cyprinae on the gills of the eastern oyster Crassostrea virginica 1 75 Marcial Villalejo-Fuerte, Bertha Patricia Ceballos-Vdzquez, Marcial Arellano-Martinez, and Federico Garcia-Dominguez Fecundity of the velvet spidercrab Stenocionops ovata (Bell, 1835) (Brachyura: Majidae) in the Gulf of California. Mexico) 181 Yimin Ye, J. M. Bishop, H. Mohammed, and A. H. Alsaffar Development of a recruitment index for forecasting seasonal landings of the Kuwait shrimp fisheries 185 W. Huntting Howell. Winsor H. Watson, HI, and Steven H. Jury Skewed sex ratio in an estuarine lobster (Homarus americanus) population 193 Dario Andrinolo, Norma Santinelli, Silvia Otaiio, Viviana Sastre, and Nestor Lagos Paralytic shellfish toxins in mussels and Alexandrium tamarense at Valdes Peninsula. Chubut. Patagonia. Argentina: Kinetics of a natural depuration 203 Patrick Lassus, Michele Bardouil, Benoit Beliaeff, Pierre Masselin, Magali Naviner, and Philippe Truquet Effect of a continuous supply of the toxic dinoflagellate Alexandrium minutum Halim on the feeding behavior of the Pacific oyster (Crassostrea gigas Thunberg) 211 Mohamed luiabir and Patrick Gentien Survival of toxic dinoflagellates after gut passage in the Pacific oyster Crassostrea gigas Thunberg 217 Lewis E. Deaton, Percy J. Jordan, and John R. Dankert Phenoloxidase activity in the hemolymph of bivalve mollusks 223 Muki Shpigel, Norman L. Ragg, Ingrid Lupatsch, and Amir Neori Protein content determines the nutritional value of the seaweed Ulva lactuca L for the abalone Haliotis tuberculata L. and H. discus hannai Ino 227 Meegan E. Vandepeer, Patrick W. Hone, Robert J. van Barneveld, and Jon N. Havenhand The utility of apparent digestibility coefficients for predicting comparative diet growth performance in juvenile greenlip abalone Haliotis laevigata 235 Rodney D. Roberts, Tomohiko Kawamura, and Christine M. Nicholson Growth and survival of postlarval abalone (Haliotis iris) in relation to development and diatom diet 243 F . Laruelle, D. P. Molloy, S. I. Fokin, and M. A. Ovcharenko Histological analysis of mantle-cavity ciliates in Dreissena polymorpha: Their location, symbiotic relationship, and distinguishing morphological characteristics 25 1 Abstracts of technical papers presented at the 19th annual meeting of the Milford Aquaculture Seminar. Milford, Connecticut, February 27-March 1 , 1999 259 Abstracts of technical papers presented at the 91st annual meeting of the National Shellfisheries Association, Halifax, Nova Scotia, Canada April 18-22, 1999 28 1 COVER PHOTO: Veined rapa whelk, Rapana venosa, courtesy of Juliana M. Harding 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. 18, No. 1 JUNE 1999 CONTENTS IN MEMORIAM L. Eugene Cronin ( 1 9 1 7- 1 998) 1 IN MEMORIAM Terrance Henry Butler ( 1923-1998) 5 Juliana M. Harding and Roger Mann Observations on the biology of the veined rapa whelk. Rapana venosa (Valenciennes, 1846) in the Chesapeake Bay ... 9 Ximing Guo, Susan E. Ford, and Fusui Zhang Molluscan aquaculture in China 19 Sergio Curiel Ramirez and Jorge Cdceres-Martinez Settlement of the blue mussel Mytilus galloprovincialis Lamarck on artificial substrates in Bahi'a de Todos Santos B.C.. Mexico 33 G. Martinez, C. Aguilera, and E. O. Campos Induction of settlement and metamorphosis of the scallop Argopecten purpuratus Lamarck by excess K+ and epinephrine: Energetic costs 41 Stephen T. Tettelback, Christopher F. Smith, Roxanna Smolowitz, Kim Tetrault, and Sandra Dumais Evidence for fall spawning of northern bay scallops Argopecten Irradians irradians (Lamarck 1819) in New York 47 Julie A. Maguire, Pierre G. Fleury, and Gavin M. Burnell Some methods for quantifying quality in the scallop Pecten maximus (L.) 59 Mohsin U. Patwary, Michael Reith, and Ellen L. Kenchington Cloning and characterization of tropomyosin cDNAs from the sea scallop Placopecten magellanicus (Gmelin. 1791 ). . 67 Hongsheng Yang, Tao Zhang, Jian Wang, Ping Wang, Yichao He, and Fusui Zhang Growth characteristics of Chlamys farreri and its relation with environmental factors in intensive raft-culture areas of Sishiliwan Bay. Yantai 71 Eva M. Fernandez, Junda Lin, and John Scarpa Culture of Mercenaria mercenaria (Linnaeus): Effects of density, predator exclusion device, and bag inversion 77 Jorge Caceres-Martinez, Gissel Dalila Tinoco, Marco Linne Unzueta Bustamante, and Ignacio Mendez Gomez-Humaran Relationship between the burrowing worm Polydora sp. and the black clam Chione fluctifraga Showerby 85 Lourdes Lopez-Cortes, Antonio Luque, Eduardo Martinez-Manzanares, Dolores Castro, and Juan J. Borrego Adhesion of Vibrio tapetis to clam cells 91 Carrie J. Deming and Michael P. Russell Assessing manipulations of larval density and culling in hatchery production of the hard clam, Mercenaria mercenaria 99 Erick E. Bataller, Andrew D. Boghen, and Michael D. B. Burt Comparative growth of the eastern oyster Crassostrea virginica (Gmelin) reared at low and high salinities in New Brunswick, Canada 107 Q. Zhang, G. Yu, R. K. Cooper, and T. R. Tiersch High-resolution analysis of karyotypes prepared from different tissues of the eastern oyster Crassostrea virginica 115 Beth M. Hirschfeld, Arun K. Dftar, Karl Rask, and Acacia Alcivar-Warren Genetic diversity in the eastern oyster (Crassostrea virginica) from Massachusetts using the RAPD technique 121 J. Cigarria Effects of age, size, and season on growth of soft tissue in the oyster Crassostrea gigas (Thunberg, 1793) 127 Marianne Walch, Ronald M. Weiner, Rita R. Colwell, and Steven L. Coon Use of L-DOPA and soluble bacterial products to improve set of Crassostrea virginica (Gmelin, 1791) and C. gigas (Thunberg, 1793) 133 Maria Jose Almeida, Jorge Machado, and Joao Coimbra Growth and biochemical composition of Crassostrea gigas (Thunberg) and Ostrea edulis (Linne) in two estuaries from the north of Portugal 1 39 Laureana Rebordinos, Pedro Garcia, and Jesus M. Cantoral Founder effect, genetic variability, and weight in the cultivated Portuguese oyster Crassostrea angulata 147 CONTENTS CONTINUED ON INSIDE BACK COVER JOURNAL OF SHELLFISH RESEARCH VOLUME 18, NUMBER 2 DECEMBER 1999 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. Long Island University Southampton. NY 11968 Dr. Standish K. Allen. Jr. (2000) School of Marine Science Virginia Institute of Marine Science Gloucester Point, VA 23062-1 1346 Dr. Peter Beninger (1999) Laboratoire de Biologie Marine Faculte des Sciences Universite de Nantes BP 92208 44322 Nantes Cedex 3 France Dr. Andrew Boghen (1999) Department of Biology University of Moncton Moncton, New Brunswick Canada El A 3E9 Dr. Neil Bourne (1999) Fisheries and Oceans Pacific Biological Station Nanaimo. British Columbia Canada V9R 5K6 Dr. Andrew Brand (1999) University of Liverpool Marine Biological Station Port Erin, Isle of Man Dr. Eugene Burreson (1999) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Peter Cook (2000) Department of Zoology University of Cape Town Rondebosch 7700 Cape Town, South Africa EDITORIAL BOARD Dr. Simon Cragg (2000) Institute of Marine Sciences University of Portsmouth Ferry Road Portsmouth P04 9LY United Kingdom Dr. Leroy Creswell (1999) Harbor Branch Oceanographic Institute US Highway 1 North Fort Pierce, Florida 34946 Dr. Lou D'Abramo (2000) Mississippi State University Dept of Wildlife and Fisheries Box 9690 Mississippi State, Mississippi 39762 Dr. Ralph Elston (1999) Battelle Northwest Marine Sciences Laboratory 439 West Sequim Bay Road Sequim, Washington 98382 Dr. Susan Ford (2000) Rutgers University Haskin Laboratory for Shellfish Research P.O. Box 687 Port Norris, New Jersey 08349 Dr. Raymond Grizzle (1999) Randall Environmental Studies Center Taylor University Upland, Indiana 46989 Dr. Mark Luckenbach (1999) Virginia Institute of Marine Science Wachapreague. Virginia 23480 Dr. Bruce MacDonald (2000) Department of Biology University of New Brunswick P.O. Box 5050 Saint John, New Brunswick Canada E2L 4L5 Dr. Roger Mann (2000) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Islay D. Marsden (2000) Department of Zoology Canterbury University Christchurch, New Zealand Dr. Tom Soniat (2000) Biology Department Nicholls State University Thibodaux, Louisiana 70310 Dr. J. Evan Ward (2001) Dept. of Marine Sciences University of Connecticut Groton, CT 06340-6097 Dr. Gary Wikfors (2000) NOAA/NMFS Rogers Avenue Milford, Connecticut 06460 Journal of Shellfish Research Volume 18, Number 1 ISSN: 0077571 1 June 1999 Journal of Shellfish Research. Vol. 18, No. 2. 337-339. 1999. 2 0 2000 Harold Haley Haskin Honored Life Member Harold Haskin is affectionately known as "Doc" by his many students; by members of the shellfish industry who, over the years, came to respect and trust him in a way that few academics have ever been; and by state and federal government officials, who regularly sought his advice on matters pertaining to marine and estuarine environments. During his many years of service on the faculty of Rutgers University and to the shellfish industry of the US. he gained a reputation'for bringing sound scientific reasoning to the often contentious issues concerning the use and protection of shellfish resources, and for teaching others to do the same. Hal was bom Harold Haley in Niagara Falls, NY on January 3, 1915, and was orphaned 3 years later during the influenza pandemic of 1918-1919. By his own admission, Hal was a bit of a "hellion" at this early age, and his future was in some doubt after his parents died. Fortunately, he was adopted by Frederick J. Haskin a family acquaintance who was a bachelor pipefitter. His adoptive father moved with Hal to southern New Jersey, where they lodged with retired farmers. This fortuitous arrangement introduced Hal to Delaware Bay, a focus of much of his professional life. The lifestyle instilled in Hal a lifetime love for the outdoors and an ethic of hard work. His father also strongly encouraged young Hal to advance himself, and Hal's high school principal, sensing an unusual intellectual ability in the young student, helped him prepare for college. Because of his fascination with the water, Hal hoped to attend Annapolis, but he was also interested in science and finally entered Rutgers College (class of 1936) because of the reputation of its biology program. It was a professor of English, however, who encouraged Zoology Department Chairman Thurlow (T.C.) Nelson to offer Hal the scholarship and job that allowed him to remain in school during the lean Depression years. Hal was an outstanding student and the first graduate of Rutgers College to attain a straight A average. At the same time, he was a championship boxer and became the cadet colonel of the Rutgers ROTC. But it was his association with T.C, and particularly a summer job studying the biology of oyster drills in Barnegat Bay. that directed Hal toward his lifelong love affair with oysters. On the basis of his outstanding academic record and his keen observations during field work, T.C. recommended Hal for graduate school to colleagues at Harvard, where he was awarded a fellowship to study marine biochemistry. There, he completed a master's degree on chemoattractants. a PhD degree in algal physiology, and met his wife Peg (Smith College 1939). One month after graduating in 1941, Hal was inducted into the Army. He first commanded units guarding the coast from Long Island to the Virginia Capes and later was in charge of training infantry recruits — a job he did so well that the Army consistently refused his requests for overseas transfer. Major Haskin was discharged in early 1946 and spent the next 6 months as a research associate in coastal oceanography at WHOI. working 337 338 Honored Life Member: Harold Haley Haskin under his PhD advisor, Alfred C. Redfield. By the fall semester, however, he was back at Rutgers College where he began teaching general biology, limology, malacology, and oceanography, and resumed his research on shellfish problems. His early work on hard clam biology led to an offer to become director of the Virginia Fisheries Laboratory (now the Virginia Institute of Marine Science), but he decided to remain at Rutgers and in 1950 he succeeded L.C. as Chairman of the Department of Oyster Culture in Rutgers College of Agriculture and Director of the Oyster Investigation Laboratory. From 1951-1953. he took two years off from teaching and moved his growing family (eventually to include four daughters and a son) to southern New Jersey, near the "Oyster Lab" at Bivalve, where he directed a Navy-funded hydrography project within and offshore of Delaware Bay. There were few days during those years when Hal was not aboard the RV Julius Nelson sampling around the estuary. It was at this time that he met and formed lasting friendships with many of the local oyster planters, a feat made possible by his own "South Jersey" upbringing, his first-hand knowledge of the Bay, and his unpretentious manners. In the early 1950s, while Hal was engaged in the "Navy Project." the Delaware Bay oyster industry was facing serious problems. Planters relied on seed oysters from natural setting areas, but overharvesting and poor recruitment had severely depleted that supply. Hal not only recognized the gravity of the problem, but was willing to argue, essentially alone, for drastic action to turn the situation around. He convinced the state management agency that data needed to be gathered regularly and scientifically to document the status and trends of the seed oyster population and the factors affecting it. It was more difficult, however, to convince the oystermen that additional restrictions were needed on the quantity of seed being harvest and that the restrictions should be based on rigorously collected and analyzed data. Hal later recalled rancorous meetings and threats against the Laboratory, but his steadfastness, the high regard that the oyster-producing community had for him. and the volumes of data he presented eventually led to the establishment of the desired controls and improving conditions on the seed beds. More importantly, his efforts led to the establishment of an unusual tripartite management system, which is still in effect, consisting of the state management agency, the oyster industry, and the University researchers, each with defined roles and shared responsibility for managing the oyster resource. The concept of the University as an independent and unbiased supplier of data to industry and management agencies was later adopted, with Hal playing a leading role, in the management of the Atlantic surf clam fishery. The MSX epizootic, which began in Delaware Bay in 1957, led Hal in new directions: oyster pathology and genetics. Working with numerous students and colleagues. Hal pursued the search for answers about this mysterious new disease, which was devastating oyster populations in Delaware and Chesapeake Bays. One of his most valued associates was Leslie A. Stauber. a parasitologist and fellow member of the Rutger's Zoology Department. The Haskin-Stauber team spawned generations of teachers and researchers (some now in the fourth generation) dedicated to studying disease processes and defense mechanisms in oysters and other marine bivalves. One of Hal's major contributions to science and to the oyster industry has been the development of oyster strains resistant to MSX disease. Beginning shortly after the MSX outbreak, he began breeding the survivors of the epizootic — to determine whether these individuals possessed a heritable trait that could be improved by selective breeding, and if so, to provide broodstock to the industry. The breeding program, which did produce resistant strains, is ongoing today and now includes selective breeding for resistance to Dermo disease. It is one of the longest sustained breeding programs for an aquatic species. Hal's work with oysters has sometimes overshadowed his contributions in other fields, which include research on hard clam depuration, oil and sewage pollution, oyster drill biology, and the effects of dredging, damming, and development on the estuarine environment. Hal has always insisted that basic and applied science are integral to each other. He pursued both, to their mutual benefit, even when this clashed with the prevailing philosophy of University administrators. Hal is a dedicated educator, researcher and public servant. In addition to the large cadre of Haskin-trained students (including some Haskin children) who teach in colleges and universities in this country and abroad, many serve in state and federal agencies where they are involved in marine research and policy making. He is at his best when he combines his research interest with education, both formal and informal. He is famous for his graduate courses that involved field trips most every weekend, regardless of weather; hence another affectionate title "Hurricane Hal." Students who took Hal's courses in estuarine ecology and coastal oceanography can attest to being in the field on blustery days with Hal standing in icy water, seemingly oblivious to the cold, lecturing on one of his favorite subjects — oyster biology. Years later, former students would approach him at meetings or other functions and say that these classes were the most valuable they had taken in their graduate career. Hal's classroom lectures were fact filled and precise, reflecting his exhaustive preparation. A crew assigned to film lectures for General Biology at Rutgers found that most faculty lectures could easily be edited and considerably reduced, and many had to be refilmed. Hal's, however, took the full time allotted and required virtually no editing. Hal's teaching was by no means restricted to the classroom. Students were to be found everywhere: visitors to his office or lab who were entranced by his discourses on oysters, or clams, or estuarine processes; businessmen he met at social functions, who learned how the oyster industry operates; students and employees who were treated to impromptu dialogues in front of a shucked oyster; and the legions of undergraduates who spent a summer rearing oysters at the Cape Shore Laboratory or field sampling in Delaware Bay or off the New Jersey coast. In recognition of his contributions to Rutgers, to state and federal governments, and to science and industry, Hal has received numerous awards. Until his retirement in 1984. he was the Julius Nelson Professor of Zoology at Rutgers, where he won the Lindback Award for Distinguished Faculty Research in 1980 and the University Medal for Distinguished Service in 1986. In 1980. he was the recipient of the Governor's Earth Day Anniversary Award for outstanding contributions to the protection of New Jersey's environment and in 1988. he was given an EPA Environmental Quality Award in recognition of commitment to environmental protection. He has been a member of the National Shellfisheries Association for his entire professional life and served as president from 1967 to 1969. He became an Honored Life Member in 1979 and received the David Wallace Award in 1984. Hal received a particular honor when, by a vote of the Board of Trustees of Rutgers University, the new, modern laboratory he worked so hard to build in Bivalve was named after him in 1990. This was the first time that a Rutgers building was named for a living person — a second "first" for Hal at the University. Honored Life Member: Harold Haley Haskin 339 Hal's many contributions to maintaining a safe and clean environment cannot be fully appreciated without acknowledging the major role played by his wife. Peg. The Haskins are known throughout New Jersey for their efforts to preserve water quality and to allow natural systems to operate while trying to integrate the many conflicting uses. It has been in these endeavors that the pair's access to state government served both science and the estuary well. In 1994, the "Hal and Peggy" team was honored by the Water Resources Association of the Delaware River Basin with the Dr. Ruth Patrick Excellence in Education Award for their "long-standing efforts to inform and create an awareness of the water resources of the Delaware River Basin." Their most recent award, given in 1999 on behalf of the people of New Jersey by the Governor, was in gratitude for their many efforts to "benefit the natural resources of the Delaware River Basin." Hal Haskin's singular dedication and commitment to his mission has been extremely effective in the difficult task of integrating scientific information with political reality. Too often, science focuses on a single problem and loses perspective. We can all learn from Hal's dedication, his broad and balanced outlook, and his insistence on facts and good science. He said it best while speaking of the many interests that compete with oysters for use of the Delaware Estuary: "I am not one of those who believes in keeping everything just the way it is in the name of conservation. I think we can arrange things to serve many different needs, but we have to recognize we have a very valuable, renewable resource here, and the pressures on it are already great. We should be very watchful when we consider changes that could add new stresses." John Kraeuter and Susan Ford Port Norris, New Jersey September 1999 Journal of Shellfish Research. Vol. 18, No. 2, 341-342. 1999. Carl James Sindermann Honored Life Member Among marine biologists and parasitologists around the world, the name Carl Sindermann is almost synonymous with books and articles on diseases of marine organisms. A background in fish and shellfish pathology, coupled with a talent for writing, propelled him into a successful career as a scientific author. Carl is a prolific and thoughtful writer, whose output includes not only scientific treatises, but books on the discipline of science and the scientists who practice it. Volumes that he has authored or co-authored are indispensable components of marine biology libraries worldwide. Carl grew up in North Adams, Massachusetts, where he was born on August 28, 1922. After graduating from high school and working for two years for Pratt and Whitney, an airplane engine manufacturer in Hartford. Connecticut, he joined the US Army at the outbreak of World War II. Carl served as a medic in an infantry reconnaissance platoon, landing in Normandy a month after D-Day and moving with Patton's army through France. Germany, Austria, and Czechoslovakia before the war ended. In 1946. Carl enrolled at the University of Massachusetts on the GI Bill. It was not until his senior year, however, that his interest in science was galvanized. A female faculty member, who. Carl recalls, was still an assistant professor after 20 years in the Zoology Department, assigned him a research project that became his senior honor's thesis: the study of an invasion of western Massachusetts by a large, predatory land planarian. Carl worked out the life cycle of this flat worm, which had been imported in soil from the tropics and was destroying natural earthworms in greenhouses — and launched into a career as a parasitologist. He had already been accepted at Purdue University when the head of the Zoology Department at the University of Massachusetts took Carl to visit colleagues at Harvard University. Carl was accepted on the spot to pursue a graduate program in parasitology. Carl studied with the protozoologist. L. R. Cleveland, working on life cycles of parasites of wood-eating cockroaches. He shifted to the marine field for his PhD research, working with mycologist W. H. Weston. Carl's Dissertation was based on summer research at the US Fish and Wildlife Service Laboratory at Boothbay Harbor, Maine, where he studied a fungus disease of herring. While at Harvard. Carl became a teaching assistant in Parasitology and Tropical Public Health at the Harvard Medical School, and also an instructor at nearby Brandeis University where he taught undergraduate coarses in biology and invertebrate zoology. He also continued his association with the US Fish and Wildlife Service, serving as the Chief of the North Atlantic Herring Investigations. After obtaining his PhD from Harvard in 1953. Carl remained on the Brandeis faculty until 1956. when he elected to return to the Boothbay Harbor Laboratory and become a research biologist. Carl remained at Boothbay Harbor until 1962. by which time the Laboratory, along with all US fisheries programs, had been transferred to the newly created Bureau of Commercial Fisheries (BCF). Carl's administrative skills were recognized within the Bureau, and in 1963 he moved to Maryland's Eastern Shore to become Director of the new laboratory at Oxford (now the Sarbanes Cooperative Oxford Laboratory |. The Oxford Laboratory was built as a consequence of the epizootic mortalities of eastern oysters, caused by MSX disease, that had begun in the Delaware and Chesapeake Bays a few years earlier. The new laboratory specialized in disease studies of commercially important fish and shellfish, and under Carl's direction, its scientists played important roles in the early days of oyster disease research and the laboratory's reputation became known world-wide. In 1968, Carl left the mid- Atlantic for a new post as director of the BCF's Tropical Atlantic Biology Laboratory in Miami, Florida, a job that he held for the next 4 years. In 1971, Carl returned to the mid-Atlantic, this time New Jersey, where he became director of the Middle Atlantic Coastal Fisheries Center with headquarters at the Sandy Hook Laboratory of the National Marine Fisheries Service (NMFS-the old BCF). While at the Sandy Hook Laboratory, Carl added to his administrative duties by becoming Assistant Director for Environmental Managenint of NMFS's Northeast Fisheries Center in 1976. While serving in these posts, Carl had written not only numerous articles and reports, but had also become a renowned book 341 342 Honored Life Member: Carl James Sindermann author. In 1985, he withdrew from administration to devote full time to writing. From 1985 to 1990. he was an Intergovernmental Personnel Act Appointee, first at the University of Miami and later at the Maryland Department of Natural Resources, after which he returned to the Oxford Laboratory. Throughout his career as a Federal employee, Carl retained close ties to academic institutions near his various postings. He held visiting or adjunct professorships at Georgetown. Florida Atlantic. Lehigh, and Cornell Universities, and the Universities of Miami. Guelph. and Rhode Island, where he taught invertebrate zoology, marine biology, fish pathology, and marine parasite ecology. He has served on the editorial boards of Aquaculture, Cheasapeake Science, the Journal of Fish Biology, the Journal of Invertebrate Pathology, and the Proceedings of the National Shellfisheries Association. He was the Scientific Editor of the Fishery- Bulletin. The honors and awards that Carl has received are too numerous to list, but a sampling shows the breadth of activities and interests that have occupied him during the past half century: member. Bureau of Commercial Fisheries advisory group to NASA on back contamination from lunar exploration, 1967: recipient of the Department of Commerce Silver Medal for administrative and research activities, 1975: chairman, New Jersey Sea Grant Advisory Board, 1981-1985; keynote speaker for the Sixth Symposium on Pollution and Physiology of Marine Organisms. Charleston, SC. 1983. He served as the President of the World Mariculture Society in 1980-1981. and was chosen as an Honored Life Member of the National Shellfisheries Association in 1991. Although Carl has been a member of various international fisheries bodies, his work with the International Council for the Exploration of the Sea (ICES) is perhaps the most important. His affiliation with that organization began in 1959 when he attended his first meeting in Copenhagen. In the 1970s and 1980s he served on a number of ICES working groups, including those for Fisheries Improvement, Marine Aquaculture, Marine Pathology, and Introduced Species (of which he was chairman for a decade). His ability to synthesize and analyze great quantities of material was critical to the preparation of numerous reports for these groups, some of which served as the basis for later publications. An important contribution of these working groups was the issuance of the ICES "Code of Practice," which lists steps to be taken during the transfer of aquatic species to reduce the risks of disease spread when aquatic organisms are moved to new locations. The guidelines are used at present by most European countries and many US states. Although Carl devoted much of his career to laboratory administration, he is best known as a book author. His scientific writing began as papers describing his research on marine parasites and pathology. His first publication (1953) was on "clam digger's itch." a human problem, but caused by a trematode parasite with a marine snail intermediate host. Carl's interests subsequently turned to parasites of the marine organisms themselves. Because he was in charge of the Atlantic herring project, his studies focused on this species, with a number of publications in the 1950s describing parasites and diseases of herring. Several of Carl's early papers showed that parasites could be used as tags to trace the movement of fish stocks. At the Boothbay Harbor Lab, Carl's work on serology of fishes resulted in a series of papers ranging from comparative serotyping of different fish species to the effects of disease on blood characteristics. In the early 1960's Carl's genius for synthesizing material became evident in an article entitled "Disease in marine populations" ( 1963). Not long afterward, he teamed up with Oxford Lab colleague Aaron Rosenfield. whom he had met while both were on the faculty of Brandeis University, to produce the now classic paper "Principal diseases of commercially important marine bivalve Mollusca and Crustacea," published in the Fishery Bulletin in 1967. In 1970. Carl expanded his earlier work in a volume entitled "Principal Diseases of Marine Fish and Shellfish" (Academic Press), which won the Wildlife Society of America award for best scientific publication in fisheries for 1970. Carl later updated this important work, which was re-issued in a 1990 two-volume set. These publications are acclaimed not only for the breadth of material included and the depth of analysis, but for the clarity of language and illustrations. His growing interest in aquaculture led to another indispensible book for the aquatic pathologist: "Disease Diagnosis and Control in North American Marine Aquaculture." edited by Carl and published in 1977 by Elsevier. This volume was also updated, in 1988, and in collaboration with Don Lightner. While director of the Sandy Hook Laboratory, which is situated on the shore of the New York Bight, it was natural, perhaps inevitable, that Carl's attention would be drawn to the effects of coastal pollution on marine organisms. Once again, he meshed his surehanded grasp of disease processes in the marine environment with what he was learning about pollution in a series of publications showing links between environmental contaminants and disease in marine fish. One of his most recent books. "Ocean Pollution — Effects on Living Resources and Humans" ( 1996. CRC Press) is an outgrowth of these concerns. Carl's enthusiasm for writing has led him into areas not often entered by scientists: writing about the scientists themselves. His first foray, entitled "Winning the Games Scientists Play" ( 1982, Plenum) elicited enthusiastic reviews, and some consternation among a few colleagues who recognized themselves in the vignettes he used as illustrations. "The Joy of Science" (Plenum) was published in 1985. followed in 1987 by "Survival Strategies for New Scientists" and in 1992 by "The Woman Scientist" (co-authored by Clarice Yentsch). Carl's most recent offering, written in collaboration with Tom Sawyer, is entitled "The Scientist as Consultant: Building New Career Opportunities" (1997). The books show Carl to be a keen observer of scientists and an accurate reporter of their behavior. They have an underlying theme: to analyze, often with a lighthearted touch, what makes a person successful in the scientific profession. They are realistic, discussing both pros and cons of certain career paths, and contain a wealth of practical advice — valuable not only for those considering or just embarking on a scientific career, but with admonitions that individuals well along in their professions would do well to follow. Carl and his wife Joan are the parents of two daughters (both social scientists) and three sons (all in construction). When Carl retired in 1991, he and Joan decided to remain on the Eastern Shore near the Sarbanes Cooperative Laboratory. Carl continues to come into his office at the Laboratory, to use the library, to chat with Aaron Rosenfield over lunch, and to work on yet another addition to the long list of publications under the Sindermann name. His current work in progress is tentatively titled "Rhyme of an Ancient Scientist: the Aging Scientist in Today's Society." Susan Ford Port Norris, New Jersey September 1999 Journal of Shellfish Research, Vol. 18, No. 2. 343-345. 1999. Aaron Rosenfleld Honored Life Member His unassuming and modest demeanor belies the fact that Aaron Rosenfield has been one of the most important movers in the field of shellfish pathobiology over the past 40 years. At the Oxford, Maryland Laboratory, he lead a team of researchers that made outstanding contributions in this field. Studies on the distribution, causes, and effects of diseases on aquatic biota have been his main research specialty. Information and technical transfer have also been a particular strength. He has played a major role over the past three decades in bringing together talented individuals with creative minds for symposia, workshops, and conferences. These assemblies have resulted in the synthesis of scientific information into reports and books for use by resource managers, industry regulators, and other decision makers, particularly in relation to the control of biological and anthropogenic pollution, fisheries conservation, and aquaculture. Aaron was born in Boston Massachusetts on October 14, 1924. He grew up in nearby Cambridge, where he attended public schools, graduating from Cambridge Latin School in 1942. He then enlisted in the Navy and served for the remainder of World War II as a quartermaster on P.T. boats in the South Pacific, where he saw action in the Solomon and Philippine Islands, New Guinea, and Palau. After his discharge from the Navy in 1 946, Aaron entered the University of Massachusetts. His interest in science had been stimulated in high school by the Sinclair Lewis novel "Arrowsmith," about a microbiologist, and he pursued a degree in bacteriology and public health which he received in 1950. He remained at U. Mass to study for a Master's degree in bacteriology and food science which he received in 1951. Diploma in hand, Aaron entered the world of industry where he worked as an analytical chemist for a company that produced emergency rations for the military. He quickly decided that industry was not for him and in the fall of 1952, he was awarded a Sara Hays teaching fellowship at Brandeis University where he taught laboratory courses in general biology, botany, microbiology, and plant physiology. At the same time, he began graduate work at Boston University. During this period he worked summers with Carl Sindermann. whom he had met at Brandeis, at the Maine Department of Sea and Shore Fisheries Laboratory at Boothbay Harbor where they collaborated on a herring parasite project. Aaron became a full-time graduate student in 1953 when he enrolled at the University of Texas at Austin. From 1954 to 1956. Aaron returned to teach at Brandeis with his new bride Clarice, whom he had married in 1953. In 1956. he returned to Texas to complete his doctoral degree while Clarice worked at the University's Chemistry Department to provide living expenses for the family, which soon included a daughter, Sandy. Despite his early interest in bacteriology, Aaron switched focus during his PhD research to a project examining the physiological consequences of heterosis in corn. The quantification of DNA to standardize other measures was a newly emerging technique, which Aaron adopted and which inspired his life-long fascination with cell and molecular biology. After obtaining his PhD in 1960, Aaron remained for several months at the University of Texas as a postdoc studying mechanisms of bioluminescence in marine organisms. At the end of 1960. he returned to the Boothbay Harbor Laboratory where he became Microbiology Program Leader with the US Bureau of Commercial Fisheries (BCF). His research projects included in vitro tissue culture of molluscs, mechanisms of disease transmission and resistance, and cytogenetics. In 1962, Aaron was invited to take charge of the Shellfish Mortality Program at BCF's new Oxford Laboratory on Maryland's Eastern Shore. At Oxford. Aaron became deeply involved in research on the newly emerging oyster diseases. MSX and SSO. as well as Dermo. He and his research team discovered and described the elusive spore stage of the MSX parasite {Haplosporidium [then Minchinia] nelsoni). During this period. Aaron also made occasional forays across 343 344 Honored Life Member: Aaron Rosenfield Chesapeake Bay to the Washington headquarters of BCF where he held temporary duty assignments, including Chief of the Bureau of Shellfisheries and Chief of the Division of Resource Research and Management. Aaron's roots in New England showed themselves forcefully when he first came to the Oxford Lab. He had a pronounced accent. So did this technician, only she came from South Carolina and had a heavy southern accent. The two often had difficulty understanding each other and colleagues recall having to act as interpreters in the laboratory. Aaron's accent has diminished over the years, but one is never in doubt about his origin. In the early 1970s, the BCF was transferred to the National Oceanic and Atmospheric Administration (NOAA) and renamed the National Marine Fisheries Service (NMFS). The Oxford Laboratory was then assigned to the Middle Atlantic Coastal Fisheries Center, with headquarters at Sandy Hook, New Jersey. Aaron became its Officer-in-Charge of the Oxford Laboratory and its Director of Pathobiology Investigations. In this position he had the responsibility for establishing and supervising major national and international research programs in coastal aquatic animal health, biomedicine, and comparative pathobiology. His work put him in close contact with the shellfish industry, university and government researchers, and resource managers. These contacts became increasingly important as diseases and die-offs of fish and shellfish, and pollution associated problems, began to increase not only in North America, but in other parts of the world as well. Meanwhile, the Oxford Laboratory's pathology program expanded to include projects and personnel at the Milford, Connecticut and Sandy Hook, New Jersey Laboratories. Pathobiology research also expanded to encompass studies on other invertebrates, especially crustaceans, and finfish, as well as studies on microbial pathogens and tumors in "lower animals." In the mid 1970s, with still another NMFS reorganization, Aaron continued to head the Oxford Laboratory as its Director and as Chief of the Northeast Fisheries Research and Science Center's Division of Pathobiology. One of the most important contributions of the Pathobiology Program under Aaron's direction was the establishment of cooperative alliances with other federal agencies, with state agencies and commissions along all U.S. coasts, with foreign governments, and with scientific and professional organizations around the world. The Oxford Laboratory's reputation grew until it was recognized world-wide as one of the foremost institutions in the detection and control of marine fish diseases. Students and established researchers from all over the U.S. and from dozens of foreign countries traveled to Oxford to study pathology of marine fish and shellfish, or to engage in joint studies with the Laboratory's expert staff. Because the Laboratory's activities had such a wide geographic scope, Aaron's outlook on the problems of diseases in marine organisms also became global. The potential transmission of disease organisms among marine ecosystems developed into a major concern and focus of efforts. Aaron and his staff were instrumental in the design of strategies and programs for aquatic animal health, including quarantine and inspection to prevent the spread of pathogens, pests, predators, and competitors in previously unaffected areas. Early in his career, he urged all of the state fishery commissions (Atlantic. Gulf. Pacific) and the International Council for the Exploration of the Sea (ICES) to consider the consequences of introduced genes (both hosts and parasites) on indigenous populations. These efforts have resulted in guidelines for the movement of fish and shellfish, which are now used by many states and foreign countries. In another move to foster global communications and data analysis among fish and shellfish pathologists, and to house and catalogue the growing collection of pathology specimens found by the Oxford Laboratory's own investigators or sent to them by distant collaborators, Aaron conceived of the Registry of Marine Pathology, now incorporated into the Registry of Tumors of Lower Animals at the George Washington University Medical Center. Submissions to the Registry are compared with archived samples, thus providing a world-wide data base of marine parasites and pathogens and a mechanism whereby new discoveries can be evaluated in the light of existing information. During his career as a government scientist and administrator, Aaron maintained unusually close ties with academic institutions. He has been a research associate at Georgetown University's Biology Department, an associate faculty member at the Johns Hopkins School of Public Health and Hygiene, and an adjunct professor at the University of Maryland, Center for Estuarine and Environmental Studies. He has been highly supportive of university researchers, enlisting funds from NOAA, the Army Corps of Engineers, the Environmental Protection Agency, the National Institutes of Health, the National Science Foundation, and other agencies for equipment, publications, and especially for meetings and workshops on special topics in marine pathology. Among the most important and influential of these were the "Shellfish Mortality Conferences." stimulated by the outbreaks of MSX disease in the late 1 950s and early 1960s, which brought together most of the founding generation of molluscan pathologists in the US. Again in the 1980s. Aaron was instrumental in supporting a second series of these conferences, where a new generation of researchers, as well as many of the original participants, gathered to discuss the current status and future direction of molluscan pathology. In addition to the Mortality Conferences. Aaron helped organize and support workshops in this country and abroad on pathology and in vitro biology for the American Fisheries Society, the Society for Invertebrate Pathology (of which Aaron was the permanent program chair), the American Institute for Biological Sciences, the National Shellfisheries Association, and the Society for in Vitro Biology. He helped found the Society for Invertebrate Pathology and was elected treasurer for 1983-1984. He has been an NSA member since 1962, becoming vice president in 1978 and president in 1979. Aaron has been very active on the international scene also, serving on oversight committees for collaborative programs in fisheries, aquaculture. and pathology between the US and Asia, including Japan. Indonesia, Taiwan, South Korea, and the People's Republic of China. In recognition of his many achievements in government service. Aaron has received numerous awards from the LIS Department of the Interior and the National Oceanic and Atmospheric Administration. He was recognized by the Chesapeake Bay Foundation with an award for his conservation efforts, was awarded certificates of recognition by Maryland members of the US Congress, and received a certificate of achievement from the Governor of Maryland for his work on state resource issues. He was made an Honorary Life Member of the National Shellfisheries Association in 1991. In 1987, when the Oxford Laboratory was turned over to the State of Maryland. Aaron took an Interagency Personnel Assignment at the University of Maryland's Center for Estuarine and Environmental Studies. There he continued his mission to organize workshops Honored Life Member: Aaron Rosenfield 345 and symposia, and to publish the resulting output. One of the most important of these was a session at the 1989 NSA meeting in Los Angeles on the subject of introduced species, which resulted in the seminal work "Dispersal of Living Organisms into Aquatic Ecosystems" (Maryland Sea Grant), published in 1992 with Aaron as senior editor. Aaron officially retired from NMFS in 1993, but he and Clarice continue to live in Easton. near the Oxford Laboratory (now the Sarbanes Cooperative Oxford Laboratory), where he can be found working in his office each day as a visiting senior scientist. In his "retirement." he has retained his global perspective, conceiving of and organizing symposia that attract participants from around the world. One of these, also held in conjunction with an NSA meeting (Orlando. FL in 1992). on the history and status of molluscan shellfisheries, resulted in a massive, three-volume set entitled "The History. Present Condition, and Future of the Molluscan Fisheries of North and Central America and Europe" (NOAA Technical Publications, 1997), of which Aaron was a co-editor. In 1996, he organized another major symposium, "The Blue Crab Fisheries of North America." Results of this symposium occupy the entire September 1998 issue of the Journal of Shellfish Research. Aaron is currently exploring the possibilities of organizing a similar meeting in conjunction with the Organization of American States on "Crustacean Fisheries in the Americas." Aaron's ever active mind is constantly coming up with new ideas, which are focusing more and more on helping promote and invigorate the Eastern Shore of Maryland. Among these ideas are the establishment of aquaria on routes traveled by vacationers to the Atlantic beaches, which would have both educational and heritage components and would encourage participation by watermen; a summer teacher's institute where local teachers could learn about biotechnology; and a natural products laboratory that would investigate potential uses of estuarine organisms. Aaron's mind is sure to be stimulated even more by the Internet, which he has recently discovered, and which astounds and delights him. In a typically self-effacing manner, Aaron measures much of his achievements by those of the Oxford Laboratory. He cites its impressive qualitative and quantitative publication record and its contributions to the scientific community through joint participation in professional and fishery conservation activities. He is most proud of the Laboratory's outstanding contributions toward advancing the fields of marine fish health research and comparative pathobiology, both of which were barely recognized by most fishery scientists and resource managers as integral parts of the marine fisheries ecosystems, or marine science in general, before he joined the Laboratory. Susan Ford Port Norris, New Jersey September 1999 Journal of Shellfish Research, Vol. 18. No. 2. 347-360, 1999. OVERVIEW AND BIBLIOGRAPHY OF RESEARCH ON THE GREENSHELL MUSSEL, PERNA CANALICULUS, FROM NEW ZEALAND WATERS A.G. JEFFS, R.C. HOLLAND, S.H. HOOKER, AND B. J. HAYDEN National Institute of Water and Atmospheric Research P. O. Box 109-695 Auckland, New Zealand ABSTRACT The greenshell mussel. Perna canaliculus (Gmelin 1791 ). from New Zealand, is the most intensively cultivated member of the genus Pema within the mussel family Mytilidae. The aquaculture production of this species has grown extremely rapidly over the past 20 years and is now facing some constraints on further expansion. Overcoming these constraints will require a more thorough understanding of the biology of this species. Although Perna canaliculus has been the subject of extensive research, much of this information remains obscured in unpublished literature. Therefore, an overview of the biology and research on the greenshell mussel, Pema canaliculus (Gmelin 17911 from New Zealand waters is presented along with a comprehensive bibliography of over 500 references. KEY WORDS: Pema canaliculus, greenshell mussel, green-lipped mussel, bibliography, aquaculture INTRODUCTION In terms of world aquaculture production, volumes of the Greenshell®1 mussel Perna canaliculus is second only to the blue mussel, Mxtilus edulis (Linnaeus 1758) (FAO 1999). The fanning of M. edulis, which began in Europe over 700 years ago, now takes place in over 15 countries worldwide, and the aquaculture of P. canaliculus is confined to New Zealand and only began around 20 years ago (Greenway 1969a, Vakily 1989). Perna canaliculus is now internationally recognized as a premium eating mussel and, consequently, the New Zealand mussel aquaculture industry has continued to expand rapidly, with production almost doubling in the last 4 years to reach 70.000 tonnes/ y_1 (Fig. 1 ) (New Zealand Mussel Industry Council unpublished data). The continuing rapid expansion of this industry will increasingly rely on an improved understanding of the biology of this species, particularly in relation to other members of the family Mytilidae, which are also cultured on a large scale. Although a great deal of research has been done on P. canaliculus, much of it is in the form of unpublished reports, theses, and local publications. Therefore, the purpose of this paper is to draw together a bibliography of this information and provide a synthesis of the current state of our knowledge for this species and suggest suitable avenues for future research. TAXONOMY Pema canaliculus is endemic to New Zealand, where it is commonly known as the green-lipped mussel because of a char- acteristic emerald green interior shell margin. It is also known by the tradename Greenshell® mussel. This species was first described from New Zealand as Mxtilus canaliculus by Martyn in 1784. but was not recognized until some- time later (Gmelin 1791. Siddall 1980). Other names such as Mxti- lus latus (Chemnitz, 1785) and Mxtilus smaragdinus (Hutton 1873) have been proposed, but failed to be recognized. Fleming (1959) proposed the incorporation of the species into the genus Perna (Retzius 1788) after the initial work of Soot-Ryen ( 1955), who used hinge and ligament structures and muscle scars as a basis for establishing taxonomic differences between Perna and Mxtilus. The differentiation of Perna from Mxtilus is deter- mined by the presence of postmetamorphic lateral hinge teeth on the larval shells (Siddall 1980); whereas, adult Perna lack an an- terior adductor muscle and have a discontinuous posterior retractor muscle. Other members of the genus Perna are found in coastal South America and Africa {Perna perna Linnaeus 1758) and the Indo- Pacific (Perna viridis Linnaeus 1758). These species are distin- guished by their geographic origin, coloration, soft part morphol- ogy, and most reliably, by differences in the patterns of shell muscle scars (Siddall 1980). ECOLOGY Fossilised P. canaliculus is known from a number of shallow water marine deposits around New Zealand from the lower Pliocene, 13 million years ago (Powell 1979). Extant P. canalicu- lus is widely distributed throughout much of New Zealand, but is most common in central and northern New Zealand, where it fre- quently forms dense beds of up to 100 m"2 (Fig. 2) (Stead 1971, Flaws 1975, Hickman 1991). It is found from the midlittoral to depths of over 50 m (Powell 1979. Buchanan 1994a). The physi- ological inability of small mussels to survive aerial exposure re- stricts the occurrence of these mussels on the upper shore (Paine 1971. Kennedy 1976. Marsden and Weatherhead 1998); whereas, lower depth limits are thought to be regulated by predation rather than competition (Paine 1971). Mussels in the littoral zone main- tain lower condition indices than subtidal individuals. (Hickman and lllingworth 1980), because aerial exposure reduces the oxygen uptake by up to 87% (Weatherhead 1993); decreasing feeding time; and, therefore, the energy available to maintain basal me- tabolism and growth (Vakily 1989). Perna canaliculus are found in a variety of habitats, attached to rock faces, wharf piles, and among algal holdfasts in the intertidal. and in deeper water they are often found living over mud or sand (Morton and Miller 1973). Consequently. P. canaliculus is known to tolerate a wide range of water salinities and temperatures. Nor- mal coastal salinities for this mussel are in the range of 30 to 35 ppt. but laboratory experiments have shown that P. canaliculus can survive at 25 ppt and that lower salinities can also be tolerated for short periods (Flaws 1975). This mussel is known to inhabit waters with water temperatures as low as 5.3 °C in the south of New Zealand and 27 °C in northern areas (MacDonald 1963, Hickman 1991). Through genetic research, populations at either end of the country have been shown to be partially differentiated stocks. 347 348 Jeffs et al. 70000 - 60000 - Hauraki Gulf Fishery Tasman Bay/Marlborough Sounds Fishery Aquaculture Aquaculture Forecast / i i W M. trossulus for all variables). Intraspecific variation in morphometric phenotype in both species was significantly related to variation in multilocus genotype. The impact of this morphometric variation on commercial mussel farm production in mixed M.edulis/trossulus areas is discussed. KEY WORDS: allozymes. genetics; F statistics, morphometry, Mytilus edulis, Mytilus trossulus INTRODUCTION Electrophoretic studies of allozymes in the marine mussel com- plex. Mytilus spp. (Koehn et al. 1984, Varvio et al. 1988, Mc- Donald et al. 1991) have demonstrated the existence of two elec- trophoretically distinct species. Mytilus edulis, and M. trossulus in eastern North America. The allozyme-based separation of the two species has since been validated by nuclear and mtDNA marker techniques (Zouros et al. 1992. Heath et al. 1995). Allele frequen- cies at the Mpi locus were found particularly useful in separating individuals of the two species (McDonald and Koehn 1988, Mc- Donald et al. 1991 ) and have been considered virtually diagnostic for M. edulis and M. trossulus in North American populations (Varvio et al. 1988). Along the Atlantic coast of Canada, Mytilus edulis and M. trossulus are widely distributed, extending throughout the Gulf of St. Lawrence, the Quebec north shore, the Gaspe Peninsula, New Brunswick, Cape Breton Island south through Nova Scotia (Mallet and Carver 1992; McDonald et al. 1991). Bates and Innes (1995) reported both species to be widespread along a section of the northeastern coast of Newfoundland. In Atlantic Canada, M. edulis is typically the numerically dominant species (McDonald et al. 1991, Mallet and Carver 1992. Bates and Innes 1995), although some populations (e.g.. in the Bras D'Or Lakes of Cape Breton Island and northern Gulf of St. Lawrence) are predominately M. trossulus. This distributional pattern overlaps the major growth centers of the developing Atlantic Canadian commercial mussel aquaculture industry. Mxtilus spp. in wild populations have long been recognized for their considerable morphological diversity (see Gosling 1992 for review). Using samples from wild populations representing the very extensive geographic range of the various species. McDonald et al. (1991) described the interspecific morphological relation- ships between allopatric populations of wild M. edulis and M. trossulus, and developed a discriminant function that accurately classified the two species based on a suite of morphometric and morphological characters. In Europe, interspecific morphological differences between M. edulis and M. galloprovincialis, which are evident in allopatric populations, become eroded in areas where hybridization between the two species occurs (Gardner 1996). Whether similar morphological erosion occurs in areas of contact between M. edulis and M. trossulus has not been determined. However, because mixed species stocks of M. edulis and M. tros- sulus are the norm rather than the exception within the distribu- tional area encompassed by the developing Canadian mussel cul- ture industry, inter- and intraspecific relationships among commer- cially important morphological, morphometric, or physiologic parameters and stock genetics may have considerable commercial impact. The distribution and allozyme frequencies of Mytilus spp. from the northeast coast of Newfoundland have previously been re- ported by Bates and Innes (1995). This paper extends knowledge concerning the distribution and population structure of Mytilus spp. to include the entire coast of Newfoundland, an area including nearly 10,000 km of coastline. Inter- and intraspecific relationships among morphometric and allelic variation are explored within sev- eral mixed species cultured stocks. Evidence is presented support- ing the theory that morphometric variation in shell characters is significantly associated with genotypic variation in Mytilus spp. METHODS Electrophoresis During the period 1994 to 96, a total of 25 samples from wild mussel beds and 15 samples of rope-cultured mussels from mussel farms were collected. Wild samples were taken from beaches, with the exception of the samples from Shag Rocks, Comfort Cove, and Stock Cove, which were collected by diver at depths of 3-10 meters. Rope-cultured samples were collected by stripping com- mercial mussel socks directly on site or at the processing plant before grading. Samples were returned to the laboratory' and held in tanks at ambient water temperatures until used. Individuals used for subsequent analysis ranged from 46-88 mm (mean = 65 mm) 367 368 Penney and Hart in shell length. Hepatopancreas tissue was excised, lyophilized. and stored at 5 °C for later analysis. A small amount of freeze - dried material was chopped to a fine powder and ground with Tris HCL pH 8.0 buffer with 20% glycerol. Four polymorphic loci were investigated: mannose phosphate isomerase (Mpi, EC 5.3.1.8), aminopeptidase-I (Lap, EC 3.4.11.-). phosphoglucomu- tase (Pgm, EC 2.7.5.1), and glucose-6-phosphate isomerase (Gpi, EC 5.3.1.9). Cellulose acetate electrophoresis and staining were carried out according to procedures described by Hebert and Beaton ( 1989), with the single modification that Lap was run with Tris glycine pH 8.6 buffer. We attempted a fifth locus. esterase-D (EST, EC 3.1.1.1), but were unable to obtain consistent band pat- terns with freeze-dried tissue on the cellulose acetate system. Al- lele nomenclature is similar to that employed by previous authors (Koehn et al. 1984. McDonald and Koehn 1988). Statistical Analyses Analysis of allele frequencies, Wright's F statistics with jack- knifed estimators (Weir and Cockerham 1984) and associated probability testing were carried out using Fstat version 1 .2 (Goudet 1995). Heterozygote deficiencies (1-H0/He) and associated \2 sig- nificance testing were calculated using G-Stat, version 3.1 (Siegismund 1995) with genetic distance output files processed into UPGMA cluster dendrograms using PHYLIP, version 3.57c (Felsenstein 1995). A multilocus probabilistic estimation function, the Campden and Utter (CU) Index (Campden and Utter 1985) was used to consolidate multilocus allelic variation into a single linear vector for use in further data analyses. In general terms, the CU index uses multilocus allelic frequency data to estimate and com- pare the probability that an individual's multilocus genotype be- longs to one or the other of two reference groups, given a set of allelic frequency data for both reference groups. Thus, all resulting CU scores range from 0 to 1, with individuals with multilocus allelic frequencies most similar to one or the other of the reference groups having calculated CU scores approaching 0 or 1. respec- tively, and individuals with intermediate allelic frequencies ap- proaching 0.5. Campden and Utter (1985), Gardner (1996), and others have used these scores to infer the existence of hybrids among mixed species populations. When applied to a set of multilocus frequency data for two species with allelic frequencies at one or more loci being diagnos- tic (mutually exclusive frequencies), the resulting CU scores be- come bimodal and nonoverlapping between species. Individuals of either species, but possessing alleles that are more typical of the opposite species at one or more loci, will have CU scores ap- proaching 0.5. Individuals with CU scores within the range gen- erated by the respective reference population may be considered "pure" individuals of that species (if the reference samples repre- sent pure, allopatric populations); whereas, individuals outside that range and approaching 0.5 may be considered atypical of the ref- erence population (species) and intermediate between it and the opposite species. Thus, composite genotypic allelic groupings may be created in a manner similar in concept to the derivation of compound genotypic allele classes by Skibinski (1983) for Euro- pean mixed populations of M. galloprovincialis, M. edulis, and their hybrids. A site on Prince Edward Island and another in Bras D'Or Lake were the geographically nearest known allopatric populations of M. edulis and M. trossulus, respectively (Mallet and Carver 1992). Selection of allopatric populations as reference standards for com- putation of the CU index is consistent with similar recent usage on other Mytilus spp. populations (see Gardner 1996). We re-sampled these populations and obtained the same result, confirming Mallet and Carver's (1995) findings. The Prince Edward Island M. edulis and Bras D'Or Lake M. trossulus populations yielded CU scores > 0.63 and < 0.32. respectively. In the present analysis, allele fre- quencies derived from the re-sampling of these allopatric popula- tions were used as reference standards to classify individuals in the Newfoundland population samples into four discrete composite allelic groups, denoted as E/E, ("pure" M. edulis, CU scores within the same range as the allopatric M. edulis; CU score > 0.63); E/T, (M. edulis, CU scores 0.5 < E/T < 0.63); T/E, (M. trossulus, CU scores 0.5 > T/E > 0.32); and T/T, ("pure" M. trossulus, CU scores within the same range as the allopatric M. trossulus; CU score < 0.32). Because no hybrid Mpi genotypes between the two species were observed, the two composite genotypic classes. E/T and T/E. contain individuals of M. edulis or M. trossulus, respectively, but with allelic frequencies at one or more of the Lap, Pgm, or Gpi loci, which are probabilistically, according to their CU scores, more typically found in individuals of the alternate species. The derived CU index scores were used in two ways: ( 1 ) as a single linear continuous vector to consolidate multilocus allelic variation for use as a covariate in regression analyses; and (b) to construct composite-allele genotypic classes consolidating mul- tilocus allelic variation for use in analysis of variance tests for relationships between multilocus allelic variation and morphomet- ric variation among Mytilus spp. All associated statistical analyses were accomplished by SAS statistical procedures (SAS Institute, Inc.. Cary, North Carolina 1985). RESULTS Species Identity Classification of each individual to either taxon was inferred from its Mpi genotype in accordance with interspecific genotypic patterns previously reported for North American Mytilus spp. (Mc- Donald and Koehn 1988, Varvio et al. 1988, McDonald et al. 1991 ). The genotypes Mpi100-100, Mp9090, or Mpi100'90 distinguish M. edulis from M. trossulus, which, in turn, is represented by genotypes Mpi Mpi" or Mpi " . Comparative allele fre- quencies of both species at the remaining three partially diagnostic loci (Table 1 ) are consistent with those reported previously using other allozyme-based or multivariate statistical protocols (Mc- Donald and Koehn 1988, Varvio et al. 1988; McDonald et al. 1991, Bates and Innes 1995). This consistency with previous pub- lished reports is taken as corroborative evidence of the validity of our iWp/-based classification of species at Newfoundland sites. Allele frequency distributions at the Lap, Pgm, and Gpi loci are significantly different (\2, P < .001) but overlapping between the two mytilid species. With the possible exception of the very rare alleles Lap'02, Gpi19, and Gpi15, which are too rare to be practi- cally useful, no species-specific diagnostic alleles were found at these three loci. The alleles Lap98, and Pgm'01' are strongly asso- ciated with M. edulis; whereas. Lap94 and Pgm ' " are associated with M. trossulus. No genotypic combinations between the four diagnostic Mpi alleles, which would have indicated the presence of F, hybrids, were observed. However, several instances of dilocus combinations of Mpi100 with the partially diagnostic Pgm'1' (n = 5.8%), or Mpi94 with Pgm100 (n = 7.7%) were noted. These frequencies are considerably greater than those observed in the two Mytilus Within a Mixed Species Zone 369 TABLE 1. Comparative allele frequencies of Mytilus edutis and M. trossulus in Newfoundland, using the Mpi locus (McDonald and Koehn 1988, Varvio et al. 1988: McDonald et al. 1991 1 as the distinguishing criterion, with \2 test for interspecific differences in frequency distributions in = 2031). Locus Allele M. edit lis M. trossulus x2 Mpi 100 0.915 — 94 — 0.976 90 0.085 — 84 — 0.024 Lap 100J 0.016 0.007 221.9** 98 0.536 0. 1 76 (df = 5) 96 0.351 0.398 94 0.088 0.324 92 0.008 0.081 90b 0.001 0.014 Pgm 118 0.004 0.011 435.5** 114 0.037 0.096 (df = 5) 111 0.049 0.464 106 0.159 0.220 100 0.670 0.186 93c 0.081 0.022 Gpi 110 0.013 0.009 67.8** 107 0.123 0.025 (df = 7) 102 0.151 0.058 100 0.213 0.147 98 0.206 0.359 96 0.134 0.232 93 0.088 0.094 89 0.055 0.060 86 0.012 0.011 83d 0.005 0.004 a' Includes rare Lap and Lapss alleles, respectively. c'd Includes rare Pgm*9, Gpi19, and Gpi15 alleles respectively. ** Denotes P < .01. Where necessary, frequencies of rare alleles were pooled to ensure statis- tical validity. allopatric Prince Edward Island M. edulis and Bras D'Or Lake M. trossulus populations (3.0 and 1.9%, respectively). Species Distribution All populations, except at one site, Shag Rocks (site #27), were found to be mixtures of the two species (Fig. 1 ). At Shag Rocks, only M. edulis was found. Their relative frequencies varied con- siderably between individual sites. M. edulis was, typically, the dominant species in terms of relative abundance, but M. trossulus was found at relatively higher frequency at four sites, reaching its greatest relative frequency (84%) at Prince Edward Bay (site #2). Sites with high proportions of M. trossulus were widely dispersed, often in close proximity to other sites with high proportions of M. edulis. However, relatively low numbers of M. trossulus were found all along the southwest and west coasts (site # 1,31, 34—39). This area is also the least heavily indented section (perhaps coin- cidentally) of the island's coastline, with a high proportion of sites relatively unsheltered from macroscale wind and wave action. Samples from cultured populations had higher relative frequencies of M. trossulus as compared to wild beds (R x C x" test of asso- ciation, P < .01 ). As a group, the mean relative proportion of M. trossulus in cultured samples was 26% versus 14% at wild beds. Genetic Structure Within and among site variation in genetic structure was ex- plored using Wright's F statistics and x2 analysis. Significant het- erozygote deficiencies (\2. P < .05) at individual loci were found to be common in both species. For both M. edulis and M. trossulus, heterozygote deficiencies ( 1-HL,/He) occurred at one or more loci at more than 70% of all sites. Analysis of within site and intersite allelic variation using jack-knifed estimation of Wright's F statis- tics (Weir and Cockerham 1984) revealed significant F1S, Fit, and Fst values for most species x locus combinations (P < .01) (Table 2). Estimators of the means pooled over all four loci were also significant for both species, indicating the existence of significant intraspecific genetic heterogeneity among sample sites. For M. trossulus, significant intersite genetic heterogeneity was evident at only the Pgm locus. To explore the cause or causes of this heterogeneity, we cal- culated matrices of pairwise Fst values for each site x site combi- nation separately for both species. We then determined whether these were associated with between-site species structure, geo- graphic distance, and sample type (e.g.. cultured or wild) differ- ences. For M, edulis, pairwise FSI values were significantly corre- lated with the magnitude of intersite difference in species frequen- cies (analysis of variance [ANOVA. P <.0001] ). This indicates intersite variation in genetic structure within M. edulis is signifi- cantly associated with intersite variation in the presence of M. trossulus. However, the reverse was not true for intersite genetic variation in M. trossulus (ANOVA. P > .05). Intersite variation in genetic structure was graphically illustrated by clustering dendro- grams (UPGMA method) using Nei's genetic distance matrices (Fig. 2). For M. edulis (Fig. 2a). the first three clusters from the root include five sites with relatively high M. trossulus frequencies TABLE 2. Summary of Wright's F statistics for individual loci and jack-knifed estimators of the means over all loci for M. edulis and M. trossulus. M. edulis M. trossulus Locus FIS FIT Fst Fis rIT Fst Mpi 0.180** 0.196** 0.019** 0.170** 0.164 -0.008 Lap 0.081** 0.085** 0.004 0.166** 0.180** 0.017 Pgm 0.145** 0.167** 0.026** 0.115** 0.137** 0.026** Gpi 0.138** 0.145** 0.008** 0.203** 0.217** 0.017 Mean 0.124** 0.134** 0.011** 0.165** 0.181** 0.019** P < .01, Bonferroni adjusted. 370 Penney and Hart Figure 1. Proportions of Mytilus ednlis (white) and M. trossulus (black) in both wild and cultured populations in Newfoundland. Cultured populations are denoted by an asterisk. (sites 2. 4, 15, 33, 40). However, no identifiable clustering patterns are found for M. trossulus (Fig. 2b). This pattern of relationship between genetic distance and occurrence of the alternate species is consistent with the expectation based on the analysis of variance test results above. Geographic distance between sites was not a significant factor in explaining the variation in pairwise Vsl values for either species (ANOVA, P > .05). Likewise, sample type (wild or cultured) was not a significant contributing factor to variation in intersite Fst values for either species (ANOVA, P > .05). Genetic Identity and Distribution of Composite Genotypic Classes Differences in allele frequencies among the four composite genotypic classes are summarized in Figure 3. E/E composite ge- notype mussels had high loadings of Mpi100, Pgm'00. Lap9S. and to a lesser degree, Gpi102 and Gpi'07. The conspecific E/T genotype mussels had relatively higher loadings of Mpi90, Pgm'". Pgm"4. Lap94, Lap92, and Gpi9S. T/T composite genotype mussels had Morphometrie Relationships in Cultured Mussels Individuals with composite genotypes E/T and T/E were dis- tributed widely throughout Newfoundland. E/T genotype mussels were found at 26 out of 40 sampled locations, ranging in frequency from 3-22% of the M. edulis population. T/E genotype mussels occurred at 35 of 40 sites, ranging from 9-85% of the M. trossulus population. Proportionally. E/T genotypes were present, on aver- age, much less frequently than T/E genotypes (mean = 7% versus 327c over-all populations combined; P < .01 ). Intraspecifically, the relative frequencies of E/T and T/E genotypes varied with sample type. Wild samples of M. trossulus contained T/E genotype mus- sels at significantly higher frequencies than did cultured samples (X;, P < -001; 53% vs. 18%). Conversely, M. edulis E/T genotype mussels were more abundant in cultured populations ( 10% vs. 4%) than in wild populations, but the difference in frequencies was not significant (\2, P > 05). high loadings on Mpi94. Pgninl. Up9h. Lap94. Gpi9*, Gpi9" as compared to its conspecific T/E genotype mussels, which had high loadings of Mpi*4, Pgm100, Lap9S. Gpi100, Gpi'02, and Gpi107. When allele frequencies at individual loci for E/T and T/E com- posite genotype individuals are compared to their conspecific E/E and T/T genotypes, the allele frequencies for both species at all loci are significantly different (\2. P< -001 ), with the exception of the Mpi locus for M. trossulus (Fig. 3). The allele frequency load- ings at Lap. Pgm. and Gpi in composite genotypes E/T and T/E mussels demonstrate the intermediacy of the multilocus allele fre- quencies in these individuals between those of "pure" individuals of either species. Relationships among shell width, shell depth, cavity volume, and shell weight, with shell length and genotype were determined by analysis of covariance (ANCOVA) analyses. Individuals of both species were pooled over all sites for these analyses. Using shell length as a covariate. significant relationships were found with CU score for shell width, shell depth, and shell weight (Table 3). Using composite genotype classes to group CU scores, ANCOVAs were done to compute least-square means (LSMs) of shell width, shell depth, cavity volume, and shell weight for each of the four genotype classes (Table 4). For all four morphometrie char- acters. LSMs were greater in M. edulis than in M. trossulus. The Mytilus Within a Mixed Species Zone 371 0.20 0.15 0.10 Nei's D 0.05 rdE (b) - HARBRE (32) - FORHAR"(9) - PRIBAV (2) - GREHAW (33) - BURARM-(11) - CRAISL (40) - TRABRI (20) - FARHAR-(10) - THWISL"(15) - MOUCAR (26) - NORPEN- (4) 0.20 0.15 0.05 0.10 Nei's D Figure 2. UPGMA clustering dendrograms of the 40 sample sites in Newfoundland for (a) M. edulis, and (b) M. trossulus. Cultured sample sites are denoted by an asterisk. Numbers in brackets refer to site numbers in Figure 2. composite genotype classes, E/T and T/E, also had LSMs for shell width and shell weight intermediate between those of the other two genotype classes. Summarized for a typical harvest size, cultured mussel (55-65 mm shell length), the differences in shell width and weight trans- late into approximately 8-9% wider shell aspect and a 25-27% heavier shell weight for E/E genotype M. edulis as compared to T/T genotype M. trossulus. E/T genotype M. edulis mussels were approximately 3-4% narrower in shell aspect and 12-14% lighter in shell weight than a comparable E/E genotype M. edulis. Con- versely, T/E genotype M. trossulus mussels were about 10% wider in shell aspect and had about 1 1-12% heavier shells than compa- rable T/T genotype M. trossulus mussels. LSMs for shell depth and cavity volume were not intraspecifically significantly different be- tween genotypes, although the interspecific differences in LSMs between the two species were significant. Composite Genotype Composite Genotype Composite Genotype Composite Genotype Figure 3. Comparison of allele frequencies among the four composite genotype classes at each of four loci (a) Mpi, (b) Pgm, (c) Lap, and (d) Gpi. Vertical bars denote individual alleles numbered as per the side legend. Alleles at very low frequency are pooled with the nearest ad- jacent allele at each locus. Probability values, P„ and P0 are for \2 tests for intraspecific differences between allele frequencies [Pe - E/E vs. E/T; P, = T/T vs. T/E). DISCUSSION The present study uses allozyme variation at the Mpi locus as the distinguishing criterion to separate individuals of M. edulis from M. trossulus in Newfoundland. This technique results in re- ported allele frequencies at the Pgm, Lap, and Gpi loci consistent with those determined previously for the two species using other allozyme protocols (McDonald et al. 1991. Bates and Innes 1995). Before the present work, Koehn et al. ( 1984) and Bates and Innes ( 1 995 ) reported the presence of M. trossulus from several sites on the northeast coast of Newfoundland. Bates and Innes (1995) used a statistical approach based on observed allelic variation at the Est and Pgm loci to distinguish between species. Our study has con- firmed the presence of both species in Newfoundland using the diagnostic Mpi locus (Varvio et al. 1988). We extended both spe- cies' respective known ranges to include the entire coast of New- foundland, an area in excess of 10,000 km of coastline. All but one site contained mixtures of both species with M. edulis typically predominating in relative frequency, a finding consistent with Bates and Innes (1995) study of the northeast coast area. The seemingly haphazard nature of the two species intermingling throughout the island lends weight to the idea the overall popula- tion structure of Mytilus spp. in Newfoundland may be termed a species "mosaic." 372 Penney and Hart TABLE 3. Statistical results of the covariance analyses on shell morphometries. Model (R2) Mean Square Prob. >F Shell width (0.73) CU score 102.2 0.0001 Shell length 2457.1 0.0001 CU score x shell length 16.3 n.s. Shell depth (0.66) CU score 40.1 0.01 Shell length 1696.0 0.0001 CU score x shell length 47.5 0.01 Log cavity volume (0.88) CU score 0.1 n.s. Shell length 21.3 0.0001 CU score x log cavity volume 0.1 n.s. Log shell weight (0.59) CU score 1.73 0.005 Shell length 45.8 0.0001 CU score x shell length 1.1 0.01 Within each species, analyses of Wright's F statistics found significant intersite geographic heterogeneity in species' genetic structure. This heterogeneity can best be described as 'patchy,' because no identifiable macrogeographic pattern was evident. However, site x site pairwise F SI values for M. edulis were greater between sites with the largest differences in relative proportion of M. trossulus. Sites in close proximity to each other displayed ge- netic variation on a scale comparable to sites large distances apart. Heterozygote deficiencies at one or more loci were common. Ray- mond (1997) proposed that the significant heterozygote deficien- cies observed in mussel populations in such areas as Newfound- land, where two species co-exist sympatrically. may be explained simply by the Wahlund effect. However, the contribution of other such factors, as genotype-dependent variation in survival related to environmental or other ecological or physiological factors cannot be ruled out. Such environmental influences on population allele frequencies at the Lap locus, in particular, have been well docu- mented in Mytilus spp. (Koehn et al. 1980. Hilbish and Koehn 1984, Gardner and Palmer 1998). In sympatric European popula- tions of M. edulis and M. galloprovincialis, genotype-dependent mortality patterns are known to occur (Gardner and Skibinski TABLE 4. Least-square means (LSMs) of composite genotype classes for shell morphometric characters in M. edulis and M. trossulus. Shell Shell Cavity Shell Genotype Width Depth Volume Weight M. edulis E/E 3 1 .55" 23.25J 2.73" 1.95" EfT 30.591, 23.70J 2.75a 1.86b M. trossulus T/E 30.55b 22.46b 2.65b 1.85" T/T 27.85c 22.48'1 2.64h 1.66c Values with different column superscripts are significant at P < .05. Values for cavity and shell weight are log transformed. 1991. Gardner et al.. 1993. Wilhelm and Hilbish 1998). Further studies on this aspect in sympatric M. edulis and M. trossulus populations are in progress in Newfoundland. We have determined that the shell morphometry of cultured M. edulis differs significantly from that of M. trossulus. At harvest size. M. edulis are larger in shell width, depth, cavity volume, and weight as compared to M. trossulus. Furthermore, we have shown that intraspecific variation in shell morphometry was not random. Intraspecific morphometric variation was significantly associated with variation in multilocus allele frequencies, represented by CU index scores and related composite genotype classes. Morphomet- ric phenotype (shell width and weight) of individuals with inter- mediate composite genotypes (E/T and T/E genotypes) was also intermediate between those with composite E/E and T/T geno- types. This is consistent with previously reported significant rela- tionships between commercially relevant production characteris- tics and variation at various enzyme loci in other bivalve species (Gaffney and Scott 1984. Diehl and Koehn 1985. Krause and Bricelj 1995). Instances of erosion in interspecific phenotypic differences, comparable to that found in our work, have been noted in wild populations of M. galloprovincialis, where they overlap distribu- tionally and hybridize readily with either of M. edulis (Gardner 1996) or M. trossulus (Sarver and Foltz 1993). However, such linkages between morphometric phenotype and genotype have not previously been reported between M. edulis and M. trossulus. In the case of Sarver and Foltz ( 1 993 ) and Gardner ( 1 996 ). the pattern of erosion of interspecific differences in morphometry were inter- preted as evidence of introgressive hybridization between M. gal- loprovincialis and M. trossulus or M. edulis. respectively. Gardner (1996) theorized that M. edulis and M. galloprovincialis maintain their genetic identities in areas of geographic overlap, despite high dispersal potential, widespread hybridization, and high levels of introgression, as a result of adaptation to different environments. Localized genotype-dependent adaptation of indigenous mussel stocks to environmental conditions may be a contributory cause of the extensive intersite variation in Fst values found among the Newfoundland sites. The observed pattern of genotype-linked erosion in interspe- cific phenotypic characters (shell morphometry) between sympa- tric M. edulis and M. trossulus in Newfoundland is comparable to that reported by Sarver and Foltz (1993) and Gardner ( 1 996) for M. galloprovincialis and its congeners. In England, the extensive hy- bridization between sympatric M. edulis and M. galloprovincialis has resulted in a continuum of individuals that are most M. edulis- like to individuals that are most M. galloprovincialis-\\ke in ge- netic structure (Skibinski 1983, Gardner and Skibinski, 1988). These hybridized populations also exhibit genotype-dependent breakdown of species-specific morphometric differences that are evident between allopatric populations of the two species (Gardner 1996). Mallet and Carver (1992) reported hybrid Mpi genotypes between M. edulis and M. trossulus at frequencies ranging from 2.5 to 19.5% in Nova Scotia and New Brunswick. However, we found no Mpi genotypes expected of F, hybrids in Newfoundland; nor did Bates and Innes (1995). Therefore, our data do not support the conclusion the linkage we observed between shell morphom- etry and genotype in mixed edulis/trossulus populations in New- foundland results from hybridization. However, the possibility individuals with E/T and T/E com- posite genotypes could be backcross hybrids of mixed ancestry Mytilus Within a Mixed Species Zone 373 cannot be ruled out. This possibility is worth noting, because we know from DNA marker research that true interspecific hybrids do exist in Newfoundland (Comesana et al. 1999). although F, hy- brids are apparently quite rare. From samples taken at two sites in Trinity Bay. Newfoundland. Comesana, et al. (1999) noted < 2% of all individuals present could be classified as F, hybrids either by allozymes alone or in combination with nuclear DNA markers. Backcrossed hybrids were much more abundant (26^ overall) but concentrated in smaller size classes < 35 mm in shell length (80% of all backcross hybrids). This implies that edulis/trossulus hy- brids, particularly F, individuals, may suffer relatively high mor- tality rates as compared to nonhybrids. If this were indeed the case, it offers an explanation as to how the geographically widespread E/T and T/E genotype individuals could be back-cross hybrids of mixed ancestry, although F, Mpi genotypes were not found. It also may explain the apparent disparity concerning the occurrence of hybrids detected by Comesana et al. ( 1999) with DNA markers and that of allozyme-based studies (this study. Koehn et al. 1984. Bates and Innes 1995), because these allozyme studies also used adult specimens. Further research on this aspect is clearly required to confirm the correct taxonomic identity of E/T and T/E genotype mussels from Newfoundland and to define the basis for the ob- served genotype-dependent erosion of interspecific morphometric characters. Such research may also elucidate whether the consid- erable intersite genetic heterogeneity observed in Newfoundland mussel populations is indicative of environmentally induced local- ized adaptive responses or simply to intersite variability in rates of interspecific hybridization. In Nova Scotia, mussel farms using indigenous M. trossulus stocks are considered to be at some economic disadvantage as compared to others using M. edulis stocks (Mallet and Carver 1995). Using cultured mussel data from a site in Lunenberg Bay in Nova Scotia, Mallet and Carver (1995) estimated, based on com- parative weight and survival data, and commercial grading trials during product processing, that the over-all economic value of M. edidis was 1.7 times higher than M. trossulus. However, whether these findings are valid for other areas outside Mallet and Carver's study area remain to be convincingly demonstrated. Because most cultured stocks are likely to be mixtures of the two taxa, what proportion of M. trossulus within a farm stock is commercially significant in terms of any realized production or marketing dis- advantage becomes of critical importance. Moreover, the eco- nomic implications of M. trossulus culture reported by Mallet and Carver ( 1995) need to be re-assessed to account for the influence of culturing stocks with intermediate morphometric characteristics due to presence of E/T and T/E genotype mussels. Differences in shell shape are of commercial concern to farm- ers because shape differences indirectly influence buyer perception of over-all product quality. Such perceptions of quality differences may limit a farmer's ability to market product successfully in a competitive industry. Further research needs to be done to deter- mine whether observed shell shape and cavity volume differences among mussels of different genotypes translate into differences in meat weight. Differences in weight have a more direct impact on farm profitability, because they directly affect economic returns from production. The difference in shell weight between E/E ge- notype M. edulis and T/T genotype M. trossulus is commercially highly relevant, on the order of 25% loss of weight for M. tros- sulus. Other genotypes produce intermediate values of shell weight. Becausee intermediate genotypic mussels are both wide- spread and, at some sites, locally abundant, this has obvious im- portant implications for commercial industry. The findings of the present work suggest that cultured stock attributes of shape and production weight may be substantially improved at some sites by substitution of the indigenous stocks with another derived from an external source with higher over-all proportions of E/E genotype M. edulis. However, in recognition of the possibility that indigenous stocks may be better adapted to local environmental conditions, such stock substitutions should be approached on a pilot scale initially and carefully monitored. In consideration of the widespread occurrence of E/T and T/E geno- types throughout Newfoundland, potential substitution candidate stocks should be carefully screened genetically. The best alterna- tive may be to import stock from outside the Newfoundland area. Such work should be a priority for further applied research efforts. Another approach to improve stock performance may be spat grading. Coincidently or not, the Prince Edward Island stock used in the present work as the reference E/E genotype M. edulis popu- lation was a stock subjected to grading at the spat stage before commercial stocking. Spat grading has not normally been applied on mussel farms in Newfoundland. The possibility that interspe- cific differences in shell shape found here for market-size mussels also extends to spat-size individuals (-10 mm) should be deter- mined. If it does, then the possibility for rigorous spat grading to decrease the frequency of T/T and T/E genotype M. trossulus and E/T genotype M. edulis among cultured stocks may exist. Further work on this subject and the related question of whether such stocks also exhibit genotype-dependent growth and survival are in progress. ACKNOWLEDGMENTS The authors thank the many Newfoundland mussel farmers who gave their time and effort to provide farm samples, and the staff of the former Department of Fisheries and Oceans Inspection Services Branch (now the Canadian Food Inspection Agency) who collected the many samples from wild mussel beds. We also thank Dr. John Brattey. Dr. Robin Anderson, and two anonymous re- viewers for their comments on an earlier draft. LITERATURE CITED Bates. J. A. & D. J. Innes. 1995. Genetic variation among populations of Mytilus spp. in eastern Newfoundland. Mar. Biol. 124:417^124. Campden. D. E. & F. M. Utter. 1985. Natural hybridization between steel- head trout (Salmo gairdneri) and coastal cuttroat trout {Salmo clarki clarki) in two Puget Sound streams. Can. J. Fish. Aquat. Sci. 42:1 10— 119. Comesana. A. S. J. E. Toro. D. J. Innes & R. J. Thompson. 1999. A molecular approach to the ecology of a mussel {Mytilus edulis — M. trossulus) hybrid zone on the east coast of Newfoundland. Canada. Mar. Biol. 133:213-221. Diehl, W. J. & R. K. Koehn. 1985. Multiple-locus heterozygosity, mor- tality, and growth in a cohort of Mytilus edulis. Mar. Biol. 88:265- 271. Felsenstein, J. 1995. PHYLIP (Phylogeny Inference Package) Version 3.57c. Department of Genetics. University of Washington. Gaffney. P. M. & T. M. Scott. 1984. Genetic heterozygosity and production 374 Penney and Hart traits in natural and hatchery populations of bivalves. Aquaculture 42: 289-302. Gardner. J. P. A. 1996. The My til us edulis species complex in southwest England: effects of hybridization and introgression upon interlocus associations and morphometric variation. Mar. Biol. 125:385-399. Gardner. J. P. A.. D. O. F. Skibinski & C. D. Bajdik. 1993. Shell growth and viability differences between the marine mussels Mytilus edulis (L.). Mytilus galloprovincialis (Lmk.). and their hybrids from two sym- patric populations in S. W. England. Biol. Bull. 185:405—416. Gardner, J. P. A. & N. L. Palmer. 1998. Size-dependent, spatial and tem- poral genetic variation at a leucine aminopeptidase (LAP) locus among blue mussel (Mytilus galloprovincialis) populations along a salinity gradient. Mar. Biol. 132:275-281. Gardner. J. P. A. & D. O. F. Skibinski. 1988. Historical and size-dependent genetic variation in hybrid mussel populations. Heredity. Land. 61:93- 105. Gardner. J. P. A. & D. O. F. Skibinski. 1991. Biological and physical factors influencing genotype-dependent mortality in hybrid mussel populations. Mar. Ecol. Prog. Ser. 71:235-243. Gosling. E. 1992. The mussel mytilus: ecology, physiology, genetics, and culture, developments in aquatic fisheries sciences, vol. 25. Elsevier. London. 589 pp. Goudet. J. 1995. Fstat v-1.2: a computer program to calculate F-statistics. /. Hered. 86 485^86 Heath. D. D.. P. W. Rawson & T. J. Hilbish. 1995. PCR-based nuclear markers identify alien blue mussel (Mytilus spp.) genotypes on the west coast of Canada. Can. J. Fish. Aquat. Sci. 52: 2621-2627 Hebert. P. D. N. & M. J. Beaton. 1989. Methodologies for allozyme analy- sis using cellulose acetate electrophoresis. Helena Laboratories. Beau- mont. Texas. 32 pp. Hilbish. T. J. & R. K. Koehn. 1984. Genetic variation in nitrogen metabo- lism in Mytilus edulis: contribution of the Lap locus, pp. 497-504. In: Proceedings of the Nineteenth European Marine Biology Symposium. Plymouth, Devon. U. K.. 16-21 September. 1984. Cambridge Univer- sity Press. Cambridge UK. Koehn, R. K., R. I. E. Newell & F. Immerman. 1980. Maintenance of an aminopeptidase allele frequency dine by natural selection. Proc. Natl. Acad. Sci. USA 77:5385-5389. Koehn. R. K.. J. G. Hall. D. J. Innes & A. J. Zera. 1984. Genetic differ- entiation of Mytilus edulis in eastern North America. Mar. Biol. 79: 117-126. Krause. M. K. & V. M. Bricelj. 1995. Gpi genotypic effect on quantitative traits in the northern bay scallop. Argopecten irradians irradians. Mar. Biol. 123:511-522. Mallet. A. L. & C. E. Carver. 1992. The ecological and commercial im- plications of genetically distinct mytilids: completion of the geographi- cal survey. Department of Fisheries and Oceans Aquaculture Resource Development Branch. Biological Sciences Directorate. Ottawa. On- tario. Canada. 53 pp. Mallet. A. L. & C. E. Carver. 1995. Comparative growth and survival patterns of Mytilus trossulus and Mytilus edulis in Atlantic Canada. Can. J. Fish. Aquat. Sci. 52:1873-1880. McDonald, J. H. & R. K. Koehn. 1988. The mussels Mytilus trossulus and M. galloprovincialis on the Pacific coast of North America. Mar. Biol. 99:111-118 McDonald, J. H.. R. Seed & R. K. Koehn. 1991. Allozymes and morpho- metric characters of three species of Mytilus in the northern and south- ern hemispheres. Mar. Biol. 1 1 1 :323-333. Raymond. M. 1997. Heterozygote deficiency in the mussel Mytilus edulis species complex revisited. Mar. Ecol. Prog. Ser. 156:225-237. Sarver, S. K. & D. W. Foltz. 1993. Genetic population structure of a species' complex of blue mussels (Mytilus spp.). Mar. Biol. 117:105- 112. SAS Institute. Inc. 1985. SAS user's guide: statistics. SAS Institute. Cary. NC. 957 pp. Siegismund. H. R. 1995. G-Stat. version 3.1. Genetical statistical programs for the analysis of population data. Arboretum. Royal Veterinary and Agricultural University, Horsholm. Denmark. Skibinski. D. O. F. 1983. Natural selection in hybrid mussel populations, pp. 283-298. In: G. S. Oxford and D. Rollinson (eds.). Protein Poly- morphism: Adaptive and Taxonomic Significance. Systematica Asso- ciation Special Volume 24. Academic Press, London. Varvio. S-L., R. K. Koehn & R. Vainola. 1988. Evolutionary genetics of the Mytilus edulis complex in the North Atlantic region. Mar. Biol. 98:51-60 Weir. B. S. & C. C. Cockerham. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38:1358-1370. Wilhelm R. & T. J. Hilbish. 1998. Assessment of natural selection in a hybrid population of mussels: evaluation of exogenous versus endog- enous selection models. Mar. Biol. 131:505-514. Zouros. E., K. R. Freeman. A. Ball & G. H. Pogson. 1992. Direct evidence for extensive paternal mitochondrial DNA inheritance in the marine mussel Mytilus. Nature. Lond .359:412— 414 Journal of Shellfish Research, Vol. 18. No. 2. 375-384, 1999. SALINITY AND SEDIMENT-MEDIATED BYSSAL THREAD PRODUCTION BY MYTILUS EDULIS LINNAEUS AND GEUKENSIA DEMISSA DILLWYN FROM NEW JERSEY SALT MARSHES MICHELE PELC AND RICHARD R. ALEXANDER Department of Geological and Marine Sciences Rider University Lawrenceville, New Jersey 08648 ABSTRACT Sediment grain size and salinity influenced byssal thread production by epibyssate Mytilus edulis and endobyssate Geukensia demissa in week-long experiments in closed aquaria. Experimental substrata included sieved sediment ranging from fine sand (0.13 mm) to gravel (3 mm) in seawater with a salinity of 30 ppt. Experimental salinities over a constant, course-grained substratum ranged from 15 to 45 ppt in increments of 5 ppt. Mean byssal thread production by the much larger Geukenisa demissa significantly (ANOVA) exceeds that of Mytilus edulis on fine gravel over the entire experimental range of salinities. However, mass-normalized mean thread production for M. edulis exceeds that of G. demissa at the optimum salinity ( 30 ppt. ) water, even though the ribbed mussels experienced a slightly higher mean temperature during the experiments (23.5 vs. 21.5 °C). This disparity in thread production in favor of the blue mussel reflects its adaptation for secure attachment against acceleration forces on its shell surface area exposed in Spanina-free low intertidal habitats. Acceleration forces against shells of endobyssate adult ribbed mussels are reduced by minimum exposure of the valve posterior, coupled with current baffling by surrounding grasses. Mean thread production at progres- sively lower salinities decreases at a greater rate for M. edulis than for G. demissa. This distinction reflects increased sensitivity to osmotic stress by the lower intertidal blue mussels in comparison to ribbed mussels acclimatized to perched tide pools wherein occur seasonal dilutions through snow melt and elevation through summertime evaporation. Mass-normalized and unnormalized mean thread production by M. edulis increases substantially from medium sand to gravel at 30 ppt at 15-16 °C, a contrast with insignificant changes by G. demissa over the same range of substrata in water 17-20 °C. Substratum texture may have more influence on thread production by M. edulis than G. demissa possibly because critical masses of thread-bound, agglomerated sediment grains provide the anchor for epibyssate blue mussel against dislodgment. Ribbed mussels mostly buried in cohesive mud rarely have to copy with erosional forces destabilizing their substrata. KEY WORDS: Mytilus edulis. Geukensia demissa, byssus, salinity, substrata INTRODUCTION Byssal thread production by the blue mussel Mytilus edulis Linnaeus and ribbed mussel Geukensia demissa Dillwyn are af- fected by several abiotic factors. Thread production by M. edulis decreases above or below optimum temperature (Allen et al. 1976, Glaus 1968) and salinity (Glaus 1968. Van Winkle 1970, Allen et al. 1976). Water velocity (Maheo 1970. Van Winkle 1970, Price 1981 ), agitation (Young 1985), and oxygen tension (Widdows and Bayne 1971). in addition to circadian and tidal rhythms (Martella 1974), significant affect secretion of byssal threads. Thread pro- duction by M. edulis is also influenced by sediment grain size, increasing from very fine sand to gravel (Meadows and Shand 1989). However, documented influences on thread production by G. demissa (Van Winkle 1970. Shumway et al. 1987) do no evalu- ate substrata. Blue and ribbed mussels occupy overlapping distributions in New Jersey salt marshes, where two of the above-mentioned abi- otic variables, salinity and substrata, vary appreciably from peaty, gravelly embankments abutting tidal fiats to microbial mat- carpeted, perched tide pools in the high salt marsh along the south- ern New Jersey inlets and bays. Salinities near the blue mussel- veneered embankment bathed by open ocean waters is uniformly 33 ppt throughout the year, but may reach 45 ppt during the sum- mer in evaporative, perched tide pools in the high salt marsh inhabited exclusively by G. demissa (Alexander, pers. comm.). In the winter, occasional ice melt temporarily dilutes salinities to 25 ppt in perched tide pools in the high marsh (McAloon and Browne 1998). Salinity may vary by as much as eight ppt during one tidal cycle in evaporated high marsh tide pools in late summer, while concurrently lower salt marsh tide pools, which are disconnected from the open ocean for a short portion of the tidal cycle, fluctuate by less than 3 ppt. The extent to which this naturally occurring magnitude of salinity fluctuations may differentiate byssal thread production between these two mussels is central to this investiga- tion. Epibyssate blue mussels and endobyssate ribbed mussels (Fig. 1 ) are commonly attached to a wide variety of substrata that in- cludes conspecific shells, valves of other clams, quartz pebbles and coarse sand, stranded drift wood, roots of the marsh grass Spartina alterniflora. and fucoid algal holdfasts. Byssal attachment of G. demissa to substrata need not be as secure as that for M. edulis according to Stanley (1970. Stanley 1972) because adult ribbed mussels (Fig. 1) are often semi-encased in cohesive muddy sedi- ments that require higher current velocities to erode than sand (Hjulstrom 1939). Acceleration forces against their minimally pro- truding shell posterior, which intercepts little of the flow, may be reduced further by flanking stalks of Spartina alterniflora that baffle the current (Smith and Frey 1985). In contrast, M. edulis (Fig. 1 ) has the entire surface of one valve, or cross-sectional area of both valves, exposed to acceleration forces of waves and cur- rents (Bell and Gosline 1997, Denny 1995), necessitating a secure byssal attachment against dislodgment (Stanley 1970). particularly at the Spartina-free embankments, where wave shock is greatest. Accordingly, substrata composition, texture, and mass are pre- dicted to influence significantly byssal thread production to ag- glomerate a stable anchor for epibyssate blue mussels. Conversely, thread production by ribbed mussels sheltered in marsh grass patches should be less influenced by the substrata in which they are mostly buried. 375 376 Pelc and Alexander Figure 1. Mytiloid adductor (stipled), byssal, and pedal retractor musculature, usual life position relative to the sediment surface (represented by horizontal line) and cross-sectional shape (shaded to show position of maximum width). Specimens not to scale. A, Brachidontes citrinus; B. Geukensia demissa; C, Modiolus modiolus; D, Mytilus edulis (modified from Stanley, 1972). A series of laboratory experiments were designed to test the influence of sediment substrata and salinities on thread production by each mussel species. Unlike the pattern of increasing thread production by M. edulis from fine to coarse sand (Meadows and Shand 1989), we predicted a priori that thread production by en- dobyssate G. demissa should not vary substantially with increasing sediment grain size. Similarly, we predicted that thread production by M. edulis, restricted to more stenohaline lower intertidal habi- tats, should decrease significantly in salinities progressively lesser or greater than the optimum (Glaus 1968); whereas, thread pro- duction by G. demissa, acclimatized to seasonal salinity fluctua- tions, should not decrease appreciably in salinities divergent from the optimum. MATERIALS AND METHODS Ribbed mussels were collected from tide pools above the midtide line in salt marshes at Tuckerton, NJ during January through March of 1996. Sampled tide pools were monitored for diurnal variation in temperature, salinity, dissolved oxygen, pH. turbidity, and C02 during each season. Mussels were submerged in their native seawater in an aerated, 55 gallon holding aquaria for one day while seawater equilibrated with air temperature in the experimental room. Foot-probing by mussels was monitored to select the most active specimens for experimentation. Selected specimens had their byssus trimmed even with the ventral shell margin, without stimulating expulsion of the byssal stem (Price 1981 ). the same procedure utilized by Van Winkle ( 1970). Mead- ows and Shand (1989) and Lee et al. (1990) on blue and horse mussels. Number of threads in the severed byssus were counted as a base line for comparison with experimental production. Wet mass of tissue plus shell of individuals was recorded. Specimens were apportioned among seven, hooded, aerated, 15 gallon aquaria, each filled with eight gallons of seawater from native tide pools wherein salinity was 29 ppt (Table 1 ). Equal numbers of small, intermediate, and large size individuals were distributed to each tank. Experiments were repeated in successive weeks. TABLE 1. Range of values for temperature, dissolved oxygen, and current velocity in aquaria experiments and sampled tide pools from Tuckerton, NJ Mean/Range of Daily Concurrent Maximum Range of Daily Concurrent Maximum Tested Tested Temperature Diurnal Range Dissolved Oxygen Diurnal Range in Species Variable Duration <°C) in Native Tide Pools Values (ppm) Native Tide Pools G. demissa Substratum 2/7-2/16/97 17/13-19 5-16 7.2-7.8 5.0-10.0 G. demissa Substratum 2/22-2/29/96 20/17-21 5-16 7.0-7.5 5.0-10.0 G. demissa Substratum 1/3-7/3/96 19/19-20 5-16 7.1-7.5 5.0-10.0 G. demissa Substratum 3/25^1/1/96 20/19-21 5-16 7.0-7.5 5.0-10.0 M. edulis Substratum 1/8-1/15/97 15/15-16 5-16 7.3-8.3 5.0-10.0 M. edulis Substratum 1/15-1/22/97 16/15-17 5-16 6.0-8.3 5.0-10.0 G, demissa Salinity 6/26-7/3/96 23/22-25 20-29 6.0-8.0 4.1-8.9 G. demissa Salinity 7/10-7/18/96 24/21-26 20-29 6.0-8.6 4.1-8.9 G. demissa Salinity 7/18-7/25/96 24/22-26 20-29 5.9-8.2 4.1-8.9 M. edulis Salinity 9/17-9/24/96 21/19-22 18-23 5.3-7.0 4.9-6.6 M. edulis Salinity 9/26-10/3/96 22/21-23 18-23 5.6-7.0 4.9-6.6 Source lor diurnal fluctuations, Browne and McAloon 1998 and personal measurements. Thrkad Production by Mussels 377 Specimens were placed ventral side down on sediment surfaces in plastic trays each filled with a different mean grain size. Sieved, well-sorted sediments included: ( 1 ) very find sand, (0.06 to 0.125 mm); (2) fine sand (0.125 mm-0.25 mm); medium sand (0.25 mm-0.50 mm); coarse sand (0.50-1.00 mm); very coarse sand (1.00-2.00 mm), and granules or gravel (2.00— 1. 00 mm), a range of grain sizes used by Meadows and Shand (1989) on Mytilus edulis and Modiolus modiolus. Trays were filled to a minimum depth of five cm, which prevented a fully extended, probing mus- sel foot from accessing the plastic bottom of any tray. Specimens were positioned so that the foot could not access the tray sides without "crawling" across the sediment surface, a capability dis- played mostly by juveniles. Although specimens were not fed, each tank was monitored daily for foot-probing activity by all specimen. After 1 week, specimens were removed, and the number of byssal threads attached to sediment grains were counted. Speci- mens that did not secret threads, but actively probed sediment throughout the experiment, were recorded as a zero in the produc- tion tally. However, specimens that showed no foot-probing ac- tivity during the experiment were deleted from the tally. Temperature, dissolved oxygen, and pH were recorded daily for each tank (Table 1). These variables fluctuated slightly throughout the week in closed aquaria, although all tanks fluctu- ated in unison. Variations in these factors were within ranges of diurnal fluctuations experienced in native tide pools during the same time frame. Thread production has been shown to be surface area related (Lee et al. 1990), which may be approximated by the square of the anterior-posterior length. Specimen mass (g) is highly correlated with length (r = 0.92 and 0.91 forM. edulis and G. demissa, respectively, with n = 100), and this readily obtain- able measurement is used to normalize thread production for dif- ferent size specimens (Table 1). Freshly acquired seawater and mussel specimens from the previously sampled tide pools were used in each succeeding series of experiments. After experiments on G demissa were completed in January-March 1996, the pre- viously described procedure was repeated for M. edulis in Janu- ary-March 97, except that blue mussels were collected from below the midtide line near the embankment. Diurnal range of abiotic variables in sampled tide pools were comparable for both years, although winter 1997 aquaria experiments with M. edulis experi- enced slightly cooler mean temperature (15-16 °C) than winter 1996 trials with G. demissa (17-20 :'C) (Table 1 ). Following completion of experiments with different sediment grain sizes on G demissa. samples of ribbed mussels and blue mussels were again collected form the same tide pools/ embankment area in June-July, 1996 and August-September. 1996, respectively. Samples were subjected to the same protocol as previously described, except that in these summertime experi- ments, flattened plastic beads 3-mm diameter, filled each tray to ensure uniformity in particle roundness, texture, and size. Mead- ows and Shand ( 1989) indicated that fine gravel size particles are optimum for thread production by blue mussels. Distilled water or instant ocean was added to the initial 8 gallons of tide pool sea- water, with a salinity of 29 ppt in each aquarium to produce tanks of different salinities ranging from 15 to 45 ppt in increments of 5 ppt. Subsequently, water was withdrawn to ensure equal volumes in each tank. Salinities ranging from 21-40 ppts were recorded for perched, high marsh tide pools during successive seasons (McA- loon and Browne 1998), a spectrum that influenced delimitation of the range of experimental salinities. Ten specimens were placed in each aquarium during successive weekly experiments. No foot-probing activity was observed for some specimens at extreme salinities. Consequently, slightly more observations are recorded for intermediate salinities than extremes. Evaporation from hooded tanks did require that the salinity in each tank be reduced during the experimental period, but never by more than 1 ppt. on any given day (Table 1 ). All tanks experienced the same range of dissolved oxygen, temperature, and pH for the week (Table I ), which were, again, within diurnal ranges in native tide pool for the same time period. Number of byssal threads per speci- men were mass-normalized. Experiments were repeated with fresh seawater and mussels. Both mass-normalized and unnormalized thread production values were regressed separately on sediment grain size and sa- linity for each species. Because a range of size (mass) classes of mussels were utilized for each species, thread production was re- gressed on specimen mass for each species in each salinity and on each sediment grain size to determine if ontogenetic (growth) stage of the mussel influenced thread production. Both linear regression and second-order polynomial regression (curvilinear) were ex- ecuted to determine the best fit line or curve for the distribution of thread production by mussels of each species across the spectrum of experimental salinities or sediment grain size. The second-order polynomial invariably produced the better fit, or the greater cor- relation value (r referred to as the eta coefficient for curvilinearly related, ratio-scale data). Subsequently, analysis of variance ( ANOVA) was performed on data for each species for each abiotic variable to determine if mean thread production for each species varied significantly across the spectrum of sediment grain sizes or salinities. In addition, a pair of two-way ANOVA tests were ex- ecuted that compared thread production between species (also dif- ferent temperature regimes) for varying salinities and between spe- cies (temperature regimes) on varying subtrata. RESULTS AND DISCUSSION Mean number of threads in the pre-experimental byssus trimmed off of a sample of blue and ribbed mussels was 45 Ui = 23, avg. mass = 4.8 g) and 295, respectively (/i = 12; avg. mass = 14.3 g), respectively. Some individuals produced as many bys- sal threads in 1 week at optimum salinity (25-30 ppt.) or on optimum grain size (fine gravel) (see Figs. 3. 5. 7, 9) as the mean cited above for their respective species. However, the predicted effect of ontogeny (growth stage) was not apparent in all experi- ments (Fig. 2). Only G. demissa in the salinity experiments dis- played the predictable, significant increase in thread production with increasing specimen size (Fig. 2A). This pattern was reversed for M. edulis in salinity experiments, however, with smaller speci- mens producing more threads than larger specimens, resulting in an inverse correlation. (Fig. 2B). Neither species displayed a sta- tistically verified ontogenetic effect in experiments dealing with sediment grain sizes (Fig. 2C, D). Not surprisingly, larger ribbed mussels produced a greater mean number of threads than smaller blue mussels bathed in slightly cooler water than ribbed mussels (Table 1 ) at each experi- mental salinity and sediment grain size (Table 2; see Fig. 3 vs. 5; and Fig. 7 vs. 9). However, when thread production is mass- normalized, distinction between species disappears despite the temperature difference in the experiments (Table 1 ). Indeed, blue mussels produced more threads, normalized for specimen mass, in slighter cooler mean water temperature (15-16 °C) than ribbed mussels in slightly warmer water temperature (17-20 °C) experi- 378 Pelc and Alexander B Geukensia demissa vs. Salinity 50-1 ■a a 30" Mytilus edulis vs. Salinity ,20 ppt 30 ppfv a ^all salinities) 7.0- 33 — a 25 ppT^- J2 _3S ppt H 3 10" ~~~~-15ppt ^ 40 ppt "i 1 r 10 20 30 40 Specimen Mass (grams) 1 50 D 1 1 1 T- 3 6 9 12 Specimen Mass (grams) Mytilus edulis vs. Grain Size 30-k Geukensia demissa 1001 vs. Grain Size 25 20 •a a - 15" 35 It E 3 z Specimen Mass (grams) 2.83 mm 2.00 mm (all sediments) 0.18 mm 0 mm 25 mm T 12 3 4 5 6 Specimen Mass (grams) Figure 2. Regression lines for number of byssal threads secreted regressed on specimen mass at various experimental salinities (A, B) and sediment grain sizes (C, Dl for each mussel pecies. Significant {P < .05) r values for G. demissa (A) at combined salinities, r = 0.654 (n = 118); for 35 ppt, r = 0.871 (n = 16); for 30 ppt., r = 0.76 (n = 14); for 25 ppt., r = 0.695 (n = 16); for 20 ppt., r = 0.720 (n = 15); for 15 ppt., r = 0.783 (n = 15). Significant r values for M. edulis (B) at combined salinities r = 0.244 in = 135); for 30 ppt.. r = 0.546 in = 20); for 20 ppt., r = 0.52 in = 20); For G. demissa (C) and for M. edulis (D) with sediments combined, r = 0.140 in = 117) and r = 0.053 (;i = 165), respectively. Both r values statistically insignificant, as are all r values for data on each substrata. Thread Production by Mussels 379 TABLE 2. Mean number of byssal threads secreted by blue (at lower temperature, see Table 1) and ribbed mussels (at higher temperature, see Table 1) on each sediment grain size and in each salinity during week-long aquaria experiments. Data analyzed by two-way (species, temperature treatment) ANOVA below. Sediment Grain Size (mm) (Salinity Constant at 30 ppt.) Species 0.18 mm 0.25 0.35 0.50 0.71 1.00 1.41 2.00 2.83 G. demissa 32 threads 21 13 63 42 44 25 44 30 M. eduli.s 4 threads 10 7 6 Salinity (ppt.) 7 13 6 4 9 Species 15 ppt. 20 25 30 35 40 45 G. demissa 70 threads 57 69 72 83 39 38 M. edulis 5 threads 26 14 26 15 18 4 Variable df F-ratio f-value Salinity Mussel species/temp (A) 1 88.31 .001 Salinity (B) 6 2.71 .015 Factor interaction (AB) 6 1.75 .073 Sediment texture Mussel species/temp. (A) 1 63.57 .001 Sediment texture (B) 8 2.43 .015 Factor Interaction (AB) 8 2.45 .014 ments on coarse sands and gravel (see Fig. 4 vs. 6). Mean thread production by M. edulis was maximum on medium sand (13 threads/week: Fig. 3 ) and decreased significantly on progressively finer grained sediments (Table 2). In fact, blue mussels did not attach threads to very fine sand grains (<0.18 mm) but instead attached threads to themselves after the foot probed the sediment. This significantly correlated relationship between thread produc- tion and grain size (r = 0.625; Fig. 3) persisted even after thread production was mass-normalized (r = 0.525; Fig. 4; Table 3). Similar to M. edulis. specimens of C. demissa also produced the highest average of threads (44) on course sand (Table 2, Fig. 5), but unnormalized thread production by the ribbed mussel was not significantly correlated with sediment grain size (r = 0.198; Table 3). and mean thread production values were not significantly different from fine- to coarse-grained sediments (Table 3). How- ever, correlation between thread production and sediment grain size was significant (Table 3) when thread production values were mass-normalized for ribbed mussels (r = 0.316; Fig. 6). Likewise, mean thread production by ribbed mussels on each sediment grain size were significantly different when these values are mass- normalized (Table 3). Statistical comparison of correlation coeffi- cients for thread production regressed on sediment grain size for M. edulis (r = 0.627) versus G. demissa (r = 0.198) were sig- nificantly different (Z value of 3.52. P = < .002). Thread production by individuals of M. edulis regressed on salinities from 15 to 45 ppt. showed a significantly correlated (r = 0.298; Table 3) hyperbolic (inverted U-shaped) distribution of mean values across the salinity spectrum (Fig. 7). This pattern indicates that production decreased significantly (Table 3) in sa- linities below and above the optimum salinity interval, of 25-30 ppt. The hyperbolic relationship of thread production regressed on salinity persists (Fig. 8), and significant correlation between salin- ity and thread production remained (r = 0.297; Table 3), when values were mass-normalized. Interestingly, the greatest mass- normalized mean value, 22 threads/g/week. occurred at 30 ppt. (Fig. 8), which is also the salinity in the tide pools to which the blue mussels are native. Although blue mussels in waters of 22-23 °C; Table 1 ) secreted fewer mean number of threads than ribbed mussels in waters of 25-26 °C (Table 1 ) at every salinity (Table 2), mass normalized values for M. edulis exceeded those for G. demissa at intermediate (20-35 ppt.) salinities (Fig. 8 vs. 10). In contrast to M. edulis, thread production of G. demissa was not significantly correlated with salinity (r = 0.183: Table 3), despite the hyperbolic pattern of thread production values re- gressed on salinity values (Fig. 9). The greatest mean value, 83 threads/week (Table 2), occurred at a salinity of 35 ppt.. but this mean was not significantly different than the mean value at 15 ppt.. namely. 70 threads/week (Table 3). Furthermore, mass normaliza- tion of the data (Fig. 10) did not significantly increase either the correlation between mean thread production and salinity (r = 0.127), or the differences in mean thread production values from lowest to highest tested salinities (Table 3). Comparable maximum values. 12 threads/g/week, occurred at salinities of 15 and 35 ppt. DISCUSSION Admittedly, dynamics of tide pool setting of ribbed and blue mussels cannot be reconstructed in closed aquaria experiments. Laboratory experiments on byssal thread production did not dem- onstrate possible synergistic effect of the combined influence of substratum and salinity concurrently, because each variable, salin- ity and substrata, was treated separately in different seasons. Com- parisons of thread production between mussel species must be restricted to same-season tests of the same abiotic variable because of the disparity in the temperature regimes under which the sub- strata versus salinity experiments were conducted (Table 1 ). Fur- thermore, other potentially very influential variables, such as cur- rent velocity or agitation (Van Winkle 1970, Young 1985). cannot be evaluated in a closed aquaria system. Current velocities, how- ever, in Spartina-fringed. microbial mat-veneered high marsh tide pools at Tuckerton are between 5 and 10 cm/s for most of the tidal cycle, exceeding 10 cm/s value for only 15 min of the 12-h tidal 380 Pelc and Alexander = -0.190x2 + 10.983x - 2.729 r = 0.627 Grain Size (mm) Figure 3. Byssal thread production by Mytilus edulis in aquaria dur- ing 1 week on very well sorted sieved sediment, fine sand to fine gravel, at 30 ppt. Successive experiments conducted Jan. 8 to Jan. 22, 1997. Range of mean daily temperature for successive tests 15-16 °C; Range for daily dissolved oxygen 6.00 to 8.3 ppm. Significance of correlation coefficient and differences in successive means on each sediment indi- cated in Table 3. Significance of correlation coefficients and differ- ences in successive means on each sediment indicated in Table 2. Best fit curve is second order polynomial. cycle (Browne, pers. comm.). Current velocities in isolated perched tide pools is negligible for much of the tidal cycle when pools are disconnected. In addition, daily temperature ranges and dissolved oxygen readings in individual tide pools for any one season are comparable to the range of temperature and DO values experienced in the aquaria over a week during the same season (Browne and McAloon 1998) (Table 1). A potentially important factor precluded in the aquarium experiments is diurnal tidal sub- mergence and emergence. However, submergence-emergence of blue mussels and the tidal cycle fluctuation have been reported to show little or no effect on byssus production (Martella 1974, Price 1982, Young 1983). Bubbling aerators in the half-filled 15-gallon aquaria provided as much water agitation as mussels experienced in emergent, isolated tide pools. Nevertheless, fluctuations in the aforementioned abiotic factors in the aquaria experiments did not differ appreciably from fluctuations occurring concurrently in tide pools, except for variables that were deliberately controlled; namely substratum and salinity (Table 1 ). Despite certain debatable limitations, closed system experi- ments facilitated evaluation of the influence of incremental in- creases in salinity on thread production, as per other experimental investigations (Glaus 1968, Van Winkle 1970. Allen et al. 1976). Results of this investigation on M. edulis are congruent with those of Glaus (1968), who showed that thread production was optimum for the blue mussel at 31 ppt and decreased substantially in salini- ties of 15 and 46 ppt. because of increasing osmotic stress for mussels in ambient hyposaline or hypersaline water. Results of experiments on G. demissa paralleled results of Van Winkle (1970); greater thread production occurred in ribbed and horse mussels species at 30 and 32 ppt, respectively, versus 15 and 16 ppt. respectively. Similar to published results, decreasing thread production is attributed to increasing stress associated with a hy- posaline medium bathing ribbed and horse mussels. Results on the influence of substratum are also congruent with other investigations that compared thread production of the epibys- sate blue mussel with an endobyssate mussel; namely. Modiolus modiolus (Meadows and Shand 1989). The horse mussel produced more byssal threads per unit time at all experimental sediment grain sizes, 0.05 mm to 16 mm, versus the blue mussel. Similarly. the endobyssate G. demissa produced more threads on all sediment grain sizes, 0.12 to 2.0 mm. and at slightly higher temperatures (17-20 vs. 15-16 °C) relative to M. edulis (Table 2). Although thread production by M. edulis increased progressively from finer to coarser textured sediments in both experiments (this investiga- tion and Meadows and Shand 1989), the substratum that induced production of the most threads among either endobyssate clam G. demissa (this investigation) or M. modiolus (Meadows and Shand 1989) was not the coarsest used in either experiment. Maximum thread production by M. modiolus and G. demissa occurred on coarse sand (Meadows and Shand 1989) and very coarse sand (this investigation), respectively, but declined on gravel in both experi- ments. Ribbed mussels (Figs. 5, 9) secreted more threads in unit time relative to the blue mussel (Table 2; Figs. 3 vs, 5, and 7 vs. 9), an observation consistent with the temperature-dependent experi- ments of Van Winkle ( 1970) on M. edulis versus G. demissa. Part of this difference in thread production between the two mussel species may reflect the slightly elevated water temperatures for the substratum experiments on G. demissa (17-20 °C) versus M. edu- y = -0.377x2 + 6.060x - 1.764 r = 0.525 S 42- | 40- n = 165 B ,38- **- - ^ O 36^ g 34H * 32- a □ a> 3 a = (0 rj 30- □ a *- (3 u 28- T3 £ 26- , 24^ C a D 2 22- 2 201 ■o ■o c CO S flJ a d 3 18- 50 16- c H CO ■D C co a> o o D •s i4- 0) a CO in w (0 > „ D S 12- c o a a/ J= 101 il E u H 8- 5 j" a S D a *s u 4- ■S I' — n B D, B : n E W-| i 3 - t—H o z s *-H Grain Size (mm) Figure 4. Byssal thread production by Mytilus edulis, normalized for (divided by) mussel mass (g), in aquaria during 1 week on very well sorted sieved sediment, fine sand to fine gravel, at 30 ppt. Successive experiments conducted Jan. 8 to Jan. 22, 1997. Range of mean daily temperature for successive tests 15-16 C; Range for daily dissolved oxygen values 6.00 to 8.3 ppm. Significance of correlation coefficient and differences in successive means on each sediment indicated in Table 3. Best fit curve is second-order polynomial. Thread Production by Mussels 381 TABLE 3. Statistical significance of eta correlation coefficients, r. and compared means for unnormalized and mass-normalized thread production by Geukensia demissa and Mytilus edulis versus salinity (15—45 ppt.) and sediment grain size (0.09 to 2.8 mm). Compared or Regressed Variables Calculated Value Probability and Decision Regression: G. demissa Thread production vs. salinity Regression: G. demissa Threads/gram of mussel vs. salinity Regression: G. demissa Thread production vs. grain size Regression: G. demissa Threads/gram of mussel vs. grain size Regression: At. edulis Thread production vs. salinity Regression: M. edulis Threads/gram of mussel vs. salinity Regression: At. edulis Thread production vs. grain size Regression: M. edulis Threads/gram of mussel vs. grain size ANOVA: G demissa Thread production vs. salinity ANOVA: G demissa Threads/gram of mussel vs. salinity ANOVA: G. demissa Threads vs. grain size ANOVA: G. demissa Threads/gram of mussel vs. grain size ANOVA: At. edulis Thread production vs. salinity ANOVA: M. edulis Threads/gram of mussel vs. salinity ANOVA: At. edulis Thread production vs. grain size ANOVA: At. edulis Threads/gram of mussel vs. grain size 2nd-order polynomial r = 0.183 2nd-order polynomial r = 0.126 2nd-order polynomial r = 0.198 2nd-order polynomial r = 0.316 2nd-order polynomial r = 0.298 2nd-order polynomial r = 0.267 2nd-order polynomial r = 0.627 2nd-order polynomial r = 0.525 F ratio = 1 .37 with 6/134 df F ratio = 0.60 with 6/101 df F ratio = 1.90 with 10/106 df F ratio = 2.06 with 10/106 df F ratio = 5.09 with 6/156 df F ratio = 4.27 with 6/156 df F ratio = 8.16 with 8/90 df F ratio = 7.38 with 8/89 df P = 0.0905; accept null hypothesis of no correlation P = 0.4306; accept null hypothesis of no correlation P = 0.1018: accept null hypothesis of no correlation P = 0.0025; reject null hypothesis of no correlation P = 0.0006; reject null hypothesis of no correlation P = 0.00076; reject null hypothesis of no correlation P = 0.001; reject null hypothesis of no correlation P = 0.001; reject null hypothesis of no correlation P = 0.231; accept null hypothesis that means are similar P = 0.731; accept null hypothesis that means are similar P = 0.052; accept null hypothesis that means are similar P = 0.034; reject null hypothesis that means are similar P = 0.0001; reject null hypothesis that means are similar P = 0.0006; reject null hypothesis that means are similar P = .0001; reject null hypothesis that means are similar P = .0001; reject null hypothesis that means are similar lis (15-16 °C) (Table 1). Thread productivity increases with in- creasing temperature (Allen et al. 1976. Glaus 1968). Neverthe- less. Meadows and Shand (1989) demonstrated that M. modiolus secreted more threads than M. edulis on a spectrum of sediment grains sizes under the same temperature regime, further substan- tiating the size dependency of thread production (Lee et al. 1990). When thread production was normalized for mass, however, the epibyssate M. edulis shows greater thread production per gram of mussel than for G. demissa (Table 2). even given the slightly lower aquaria temperatures for experiments on the blue mussels (Table 1). This greater mass-normalized value for blue mussels may reflect the need to counter the greater acceleration forces experienced by the greater amount of shell surface area of M. edulis intercepting the current relative to G. demissa. Attachment strength is directly related to number of byssal threads secreted, and dislodgment force of blue mussels is directly proportional to shell area (Lee et al. 1990, Witman and Suchanek 1984) or shell height (Willis and Skibinski 1992). Both measurements are pro- portional to the shell cross-sectional area intercepting the current. Whereas mostly buried adult specimens of G. demissa may expose less than a one cm squared surface area to Spartina grass-baffled currents, the smaller M. edulis often exposes several square cm of shell surface area to intercept the current. Even under the stress of suboptimum salinities, blue mussels produce more threads per specimen gram than ribbed mussels (Table 2). Threads of M. edu- lis are also thicker than those of G. demissa. Thicker and longer threads may have greater tensile strength (Price 1982). Mytilus califomianus has greater adherence "tenacity" than M. trossulus because of among other variables, greater thread thickness (Bell and Gosline 1997). Although it secreted fewer threads, attachment strength of M. edulis may be greater than G. demissa given the greater collective strength of individually thicker threads. The absence of the expected ontogenetic effect of increased thread production with increasing size (mass) of blue mussels for varying salinities (Fig. 2B) may indicate a possible inhibitory ef- fect on thread production induced by suboptimum salinities. In- deed, the inverse relationship between thread production and specimen size suggests that thread secretion by smaller blue mus- sels is less stressed and inhibited by the associated with subopti- mum salinities. The predicted direct relationship between thread production and specimen mass (Lee et al. 1990) did materialize for G. demissa for all experimental salinities (Fig. 2). an indication that any size-dependency of thread production by ribbed mussels was not obfuscated by osmotic stress. This mussel occupies perched tide pools where salinities fluctuate by as much as 20 ppt between seasons. The lack of correlation between thread produc- tion and mussel specimen mass on each tested sediment's texture (Fig. 2) is enigmatic, but may indicate that a threshold of current 382 Pelc and Alexander 250 -, 200- 0* c o ■a — u u (0 ■O 3 ' 2 0) c :=, ra « U o 0 w Jj O (0 flj I D D o iJ O 3 ■o ii. a C c >. « a (0 □ a a 1- k- > o c a B D il a S >» a 0 > n a □ □ □ 1 □ n a □ a ImJ D 3 So = 0.198 a a - an a n o ©' © © © o © Grain Size (mm) Figure 5. Byssal thread production by Geukensia demissa in aquaria during 1 week on very well sorted sieved sediment, fine sand to fine gravel, at 30 ppt. Successive experiments conducted Feb. 7 to April 1, 1996. Range of mean daily temperature for successive tests: 17-20 °C. Range for daily dissolved oxygen values, 7.0-7.8 ppm. Significance of correlation coefficient and differences in successive means on each sediment indicated in Table 3. Best fit curve is second-order polynomial. agitation is needed to stimulate thread production commensurate with specimen size, as noted by Lee et al. ( 1990). Current agitation may merely reinforce the existing pattern established in less agi- tated closed aquaria experiments wherein blue mussels secrete more threads per gram than ribbed mussels. Comparison of correlation coefficients for mass-normalized thread production regressed on sediment grain size for M. edulis (r = 0.627) versus G. demissa (r = 0.198) were significantly dif- ferent (Z value of 3.52, P < .002), a disparity that suggests greater sensitivity to substrata texture on the part of blue mussels. Specu- lation on how the substratum may mediate thread production fo- cuses on tactile cueing of the probing foot. The tapered end of the probing foot may possibly sense the diameter of the sediment particle. Sites to which threads are attached are first cleaned before secretion of the thread and the adhesive disk by a "plungered" foot (Amato 1981 ). If particle sizes are sensed as too small in diameter. the foot may continue probing for alternative, larger surfaces. Fail- ing to locate suitable sized grains, many mussels placed on fine sand (0.13 mm) secreted threads attached to their own shell ex- clusively. Whatever the stimulus, thread production by blue mus- sels did vary significantly on different grain size sediments in contrast to ribbed mussels. Agglomeration of a critical mass of sediment particles would be essential to stabilize the epibyssate blue mussel in place at wave-affected marsh embankment. CONCLUSIONS Mean byssal thread production per week by the larger Geuken- sia demissa exceeded that of the smaller Mytihts edulis on a fine gravel-sized substratum over the entire experimental range of sa- linities. However, mass-normalized mean thread production forM. edulis exceeded that of G. demissa at the optimum salinity; namely, 30 ppt. Furthermore, mean thread production at progres- Thread Production by Mussels 383 3 — o s eg u C y 14-, 12 10 8- 6- 4 2H 0 -0.421x2 + 1.920x + 0.603 r = 0.316 n = 117 a a a a a a <^ .01 ) of IR and CR was obtained between 19 and 22 C. Irrigation efficiency (liters of water filtered per ml 02 consumed) followed the same trend as the scope for activity, with a maximum at 19 °C. The highest growth rate was at 19 °C in a 54-day experiment. All these results indicate the temperature optimum for the juvenile catarina scallop population studied was between 19 and 22 °C. KEY WORDS: Argopecten ventricosus-circularis, catarina scallop, clearance rate, ingestion rate, irrigation efficiency, scope for activity, temperature optimum INTRODUCTION Catarina scallop {Argopecten ventricosus-circularis) has been cultured commercially in Bahia Magdalena. Mexico (Maeda- Marti'nez et al. unpubl. data) using a bottom-culture technology (Maeda-Martinez and Ormart-Castro, 1995). The average tempera- ture there is several degrees colder than in other places where this scallop has been exploited, such as Bahia de La Paz and Bahia Concepcion in Mexico, or elsewhere in Panama and Ecuador. This lower temperature is caused by the cold waters of the California Current (Alvarez-Borrego et al. 1975: Lynn and Sympson 1987). From these experiences, differences in growth between growout zones were related to temperature, harvesting larger scallops in zones with lower temperature. There were no differences in growth and survival of wild spat versus hatchery-produced spat, making the hatchery an alternative for spat production when spat collection is insufficient or out of season. To determine the possibility of expanding the culture techniques to other areas and to optimize hatchery production by culturing at or close to optimum tempera- ture, we needed to know the effect of temperature on the growth of this species. Temperature is considered the most important determinant of the level of activity in poikilotherms (Bayne 1976). Temperature affects, directly or indirectly, larval and adult survival, and its effects on reproduction (i.e.. gonad maturation, spawning) and development, larval life span, and settlement are known (Kinne 1970). The temperature optimum is commonly recognized as that temperature at which the organism has the highest energy for maintenance, growth, reproduction, movement, and so forth. Fry (1947) proposed scope for activity as a concept to determine tem- perature optimum in animals, defined as the arithmetic difference between active and standard rates of oxygen uptake. Thompson and Bayne (1972). Bayne (1973). Bayne et al. (1973). and Wid- dows (1973) have considered the scope for activity as a practical way to find the energy available to the organism for growth, re- production, and so forth. Active 02 uptake rate is obtained from an animal that has been fed continuously to satiation; whereas, the standard rate comes from an animal kept under inanition for a long period (Bayne 1976). Here the Oz uptake rate declines to an stable minimum value for several days before death. Ingestion (the amount of particles cleared from a volume of water per unit time) and clearance (the volume of water cleared from particles by the animal per unit time) rates have also been used to measure the influence of environmental variables on bi- valve mollusks ( Widdows 1973. Gonzalez et al. 1990, Iglesias and Navarro 1991. Albentosa et al. 1994. Espina and Buckle-Ramirez 1994, Navarro and Iglesias 1995, Urrutia et al. 1996). These vari- ables vary with age, size, and reproductive stage, but also are affected by such environmental factors as particle concentration, salinity. 02 concentration, and, most importantly, temperature (Bayne 1976. Shumway 1991 ). Thus, ingestion and filtration rates could be used as temperature optimum indicators. Irrigation efficiency (the amount of water pumped by the ani- mal per ml of 02 consumed) has been measured in several mol- lusks (j0rgensen"l96O. Vahl 1972, McLusky 1973. Bayne 1976, Newell et al. 1977). but this has not been used to find the tem- perature optimum. In the present work, the temperature optimum of a catarina scallop population has been investigated, studying the scope for activity, ingestion and clearance rates, and the irrigation efficiency between 12 and 28 °C. The results from these studies were con- trasted with a growth experiment at the same temperatures. MATERIALS AND METHODS Experimental Animals Juvenile catarina scallop were collected in onion bags sus- pended from a longline in the tidal channel at Rancho Bueno. B.C.S.. Mexico (Fig. 1). Spat were detached from collectors and transported to the laboratory in La Paz during a 3-hour trip in a plastic container receiving constant aeration. The spat were placed in 40-1 plastic tanks containing well seawater at 19 °C, 30%c sa- linity, and constant aeration. Well seawater was filtered through 5-p.m filter cartridges and then sterilized with UV radiation. The juveniles were fed 1 .2 x 108 cells/scallop/day of a mixture of 6:3: 1 Isochrysis galbana, Monochrysis lutheri, and Chaetoceros gracil- is 386 SlCARD ET AL. MEXICO 24u20' 24" 17' 111c 27' 111021' IK RANCHO BUENO vCX^s vi Mangrove 1I& BAJA O&l CALIFORNIA ^i PENINSULA Sand bamef \\Vl PACIFIC OCEAN 1 Km \V) Figure. 1. Spat collection site for catarina scallop {Argopecten ventricous-circularis). lis. Half of the water volume was changed every day, and the mieroalgae concentration was then re-established. Dry Tissue Weights In the present work, results were referred to the dry tissue weight of the animals. Nearly 50 scallops between 6 and 57 mm shell height were dried for 36 h in an oven at 65 °C. Constant weights of the animals were obtained. The tissues were then re- moved from the shells with a dissection needle and were weighed on an electronic microbalance with 0.1 mg resolution. With these data, regression analyses were made, looking for the best fit to describe the dry tissue weight-shell height relation. Scope for Activity Experiments Twelve groups of 30 juveniles (10 mm shell height) were placed in 19-L buckets containing 15 L of seawater from a coastal well at 19 °C and 309ct salinity. The temperature of each bucket (with one replicate), was gradually set at six experimental tem- peratures (12. 16, 19. 22, 25, and 28 °C), varying 1 ± 0.5 °C every 3 days, by using calibrated immersion heaters. In treatments lower than ambient temperature (25 °C), the buckets were dipped into a 1,100 1 water bath at 12 °C, equipped with a 1.5 HP water chiller. Under these circumstances, all the animals were at their various temperature treatments in 27 days. The scallops were maintained at constant temperature for 10 days before starting the experiments. One group of each treatment was used for active rate and the other for the standard rate. The six active-rate groups were fed three times in a day with 2.85 x 10s cells/scallop/day of a mixture of 6:3:1 Isochrysis galbana, Monochrysis lutheri, and Chaetoceros gracillis. The six standard-rate groups were kept in filtered (3 u.m) well seawater with no food additions. In this work, the rates of oxygen uptake (V02) were measured periodically over 45 days with a microWinkler method (Maeda- Martinez 1985). Ten scallops of each group were incubated for 1 h in BOD bottles of 300 ml, containing 02-saturated well seawater at 30%c and at the experimental temperatures. Over this incubation time, V02 was not affected by the drop in O, concentration (P02) in the bottles (Sicard 1999). The water was previously filtered through a GF/F (0.75 (xm) membrane. One bottle without scallops served as a blank. After incubation, the water of each bottle was siphoned into three ground-neck borosilicate tubes of approximately 7 ml. The PO, in the tubes was measured with a miniaturized version ( Maeda-Martinez 1985) (Fig. 2) of a titrator developed by Bryan et al. ( 1976). V02 was calculated by sum of the P02 in the blank bottle minus that in the animal bottles. Ingestion and Clearance Rates Ingestion (IR) and clearance (CR) rates were studied following the principle of the method by Winter (1973). which consists in keeping the algal concentration stable in the incubation chamber. Thus, the number of algae added to the chamber is equivalent to the number of algae consumed. For this work a microprocessor- controlled system (Fig. 3) was constructed to determine IR and CR. based on the design of Gallager and Mann ( 1980). Ten juve- nile scallops (11.8 ± 0.6 mm shell height), acclimated to the ex- perimental temperatures, were transferred to 1-L glass beakers containing GF/F-filtered well seawater with 1.5 x 105 cells/ml of Isochrysis galbana at 309tc salinity and at the experimental tem- peratures. The experiment was done in triplicate. A closed system was made to detect the cell concentration in the chambers. The Optimum Temperature for Catarina Scallop Growth 387 Light detector i=i Digital-to-analog converter to Fluorometer signal 2 seconds lapse Magnetic stirrer Figure. 2. Minaturized version of the titrator employed for oxygen uptake determinations (after Maeda-Martinez 1985). water in the beakers was continuously pumped with a peristaltic pump to a Turner 1 12 fluorometer at 87 mL/min, and the chloro- phyll concentration in the water was read. The fluorometer was equipped with a flow-through cell and a 47B filter specific for chlorophyll a detection at 420-500 nm. The effluent of the cell was then returned to the beaker. A analog-to-digital converter digi- talized the signal from the fluorometer. which was compared with the desired value stored in the memory of the computer (Fig. 4). If the value was equal or higher than the stored value, the computer waited for 2 min before taking a second reading. However, when the value was lower, the system operated a second peristaltic pump ( 1 .025 mL/min) to replace the algae consumed by the animals from a known algal suspension kept in a conical flask. The flask re- Air Peristaltic pump 1 I — <2> Solid state relay on/off Computer Digit al-to-analog converter Reposition cells Mesh Peristaltic pump 2 — CD C/l in L. Q J.O Y = 8.7 x 10~6x X3 2.5 r = 0.9858 • 2.0 •/ 1.5 - •y /• 1.0 • y • 0.5 JBi 0.0< iii i 10 20 30 40 50 Shell height (mm) 60 Figure. 3. Apparatus developed for measuring ingestion and clearance Figure. 5. Relation between shell height and dry tissue weight in the rates on juvenile catarina scallop (Argopecten ventricous-circularis). catarina scallop {Argopecten ventricosus-circularis). 388 SlCARD ET AL. 2.4 x 108 cells/day of a mixture of 6:3:1 Isochrysis galbana. Monochrxsis lutheri. and Chaetoceros gracillis. A 509c water ex- change was done everyday, replacing the algae consumed. The shell height of the scallops from all treatments was measured pe- riodically, with a plastic caliper, during the 54-day duration of the experiment. RESULTS Dry Tissue Weight The relation between dry tissue weight (DTW) and shell height (SH) (Fig. 5) was closely described (r = 0.98: n = 48) by the equation; DTW = 8.7 x 1(T6 x SH3 Scope for Activity The VO, of fed and starved juvenile catarina scallops at dif- ferent temperatures during 45 days is shown in Figure 6. From this. the V02 of fed animals reflecting the active rate remained steady during the 45 days, except in treatment at 12 °C. where a decline from 1 to 0.7 ml OJ°/h was measured. In contrast, the VO, of O) CT _i E u c o a. 1.5 r 1.0 0.5 0.0 2.0 1.5 1.0 0.5 0.0 2.5 2.0 1.5 1.0 0.5 0.0 3.0 2.5. 2.0 1.5 1.0 3.0 2.5 2.0 1.5 3.5 3.0 2.5 2.0 12*C 16*C 19*C 22°C 25'C 28*C 10 20 30 40 50 Time (days) TABLE 1. Active and standard respiration rates in juvenile catarina scallop {Argopeclen ventricosus-circularis) at different temperatures (n = 9(. Figure. 6. Respiration rate of fed (-) and starved (- -) juveniles of catarina scallop {Argopecten ventricosus-circularis) during a 45-day experiment at different temperatures. Values are the mean ± s of the mean (it = 9). Active Standard Respiration Respiration Rate Rate Temperature (mLO,/g/h) (mLO,/g/h) (°C) jr s x s 12 1.05 0.02 0.34 0.02 16 1.45 0.02 0.44 0.03 19 1.95 0.02 0.55 0.02 22 2.46 0.01 1.24 0.02 25 2.65 0.02 1.75 0.03 28 3.05 0.03 2.43 0.02 starved animals declined in all treatments until a minimum value was reached, reflecting the standard rate. The standard rates were reached at 23, 32. 40, 35, 15. and 6 days from the beginning of the experiment at 12. 16, 19. 22. 25, and 28 °C. respectively. The active and standard rates are in Table 1 and Figure 7a. From these, the scope for activity is shown in Figure 7b, with the highest (1.4 ml 02/g/h) at 19 °C. Ingestion and Clearance Rates Variations of IR and CR in juvenile catarina scallops at differ- ent temperatures are shown in Figures 8a, 8b, and 9. In Figure 8a, ingestion of /. galbana was highest at 22 °C, removing 2.8 x 10" cells/g/h. with 2.6 x 10" cells/g/h removed at 19 °C. At 25. 28. and 15 °C. IR declined to 1.9. 1.7. and 1.3 x 109 cells/g/h. These values in terms of algal biomass fluctuated between 15 and 32 mg/g/h 3.5 CD 3 ?'£ 75 c u> 0~n 2 ■■=(-) «_l 1.5 nE 1 CD tr 0.5 10 15 20 25 30 1.5 2? > . 1.25 '*•£ o "a cs o i. CN 1 OO --J CD c O 0.75 u (/) 0.5 10 15 20 Temperature 25 30 Figure. 7. Active (• I and standard ( 0 ) respiration rates (a), and the scope for activity (b) of juvenile catarina scallop {Argopecten ventrico- sus-circularis). Optimum Temperature for Catarina Scallop Growth 389 60 CO o •" X 50 -C \ O) V, (0 40 CD o ^-y 15 c 10 16 18 20 22 24 26 28 30 Temperature (°C) 16 30 18 20 22 24 26 28 Temperature (°C) Figure. 8. Ingestion rates of microalgae cells (al and biomass (b) at different temperatures of catarina scallop (Argopecten ventricosus- circularis) juveniles. Values are the mean ± s (b= 3). 30 25 20 O 15 10 CD O c o k_ o * 5 16 18 20 22 24 26 28 30 Temperature (°C) Figure. 9. Clearance rates of catarina scallop {Argopecten ventricosus- circularis) juveniles at different temperatures. Values are the mean + s (n= 3). Irrigation Efficiency The ratio between liters of water cleared from particles against mL of oxygen consumed (L/mL 02) at different temperatures (Fig. 10) varied in the same manner as the scope for activity. Irrigation efficiency increased from 6 L/mL 02 at 16 °C to 8.75 L/mL 02 at 19 °C. Then this gradually declined to the minimum 3.65 L/mL 0\ at 28 °C. Scallop Growth Growth in scallop juveniles at different temperatures is shown in Figure 11. Growth was at a maximum at 19 °C, followed by 22. 16, 25, and 28 °C. A covariance analysis (Table 4) indicates there were significant differences between treatments at P> .01. How- ever, the Tukey's multiple range analysis (Table 5) showed simi- larities for treatments at 19 and 22 C and among 16, 25, and 28 °C treatments. DISCUSSION The scope for activity in catarina scallop juveniles showed the optimum temperature was between 19 and 22 °C. The concept of scope for activity introduced by Fry (1947) allowed us to deter- (Fig. 8b). No pseudofeces were produced by the animals during the experiments. An analysis of variance (ANOVA) indicated a sig- nificant difference between treatments at P > .01. However, a Tukey's multiple range analysis (Table 2) showed that similar IRs were found among 16. 25, and 28 °C treatments and among 19, 22, and 25 °C treatments. Because CR was calculated from IR data, the same trend was obtained (Fig. 9). Values varied between 8.7 and 17.8 1/g/h. An ANOVA indicated that differences between treatments were sig- nificant at P > .01. The Tukey's multiple range analysis (Table 3) showed similarities among 16, 25, and 28 °C treatments, for 22 and 25 °C, and for 19 and 22 °C. TABLE 2. Tukey's multiple range test for ingestion rates in juveniles of catarina scallop (Argopecten ventricosus-circularis) at different temperatures. Temperature It X Homogeneous Groups 16 40 15.17 X 19 39 29.61 X ii 16 3 1 .63 X 25 29 21.37 X x 28 39 19.38 x 390 SlCARD ET AL. TABLE 3. Tukey's multiple range test for clearance rates in juveniles of catarina scallop {Argopecten ventricosus-circularis) at different temperatures. 14 Temperature n X Homogeneous Groups E 16 40 8.71 X 12 19 39 17.05 X -C 22 16 17.84 X X D) 25 29 12.14 X X « o c - o c o o 10 9 8 7 6 5 4 3 16 18 20 22 24 26 28 Temperature (°C) 30 10 -I 10 20 30 40 50 60 Time (days) Figure. 11. Growth of catarina scallop [Argopecteii ventricosus- circularis) juveniles during 55-day experiment at different tempera- tures. Values are the mean ± s (n= 50). rect measurements, we need to know the V02-P02 relation in this species at different temperatures. Tang ( 1933). Mangum and Van- Winkle (1973). and Sassaman and Mangum (1972) have proposed hyperbolic, quadratic, and semilogarithmic equations to describe the VOo-POn relation applicable to all aquatic organisms. Prosser and Brown ( 1961 ) defined the critical tension as the inflexion point of V02 in relation to P02 at which the organism loses its regula- tory capacity to remain independent from P02. Critical tensions are known in several molluscan species (Van Dam 1938. Garder and Eliassen 1954. Van Dam 1954, Rotthauwe 1958, Brand and Roberts 1973, Shumway 1983. Shumway and Scott 1983). Man- gum and VanWinkle (1973) showed the difference between oxy- conformer and oxyregulator species are only the extremes of a continuum. In our work, corrections for a drop in P02 were not necessary, because V02 was measured 1 hour after the start of incubation, before the critical tension was reached. In the catarina scallop, the critical tension was 76% 02 saturation (Sicard 1999), which is similar to the critical tension (75%) reported in Mytilus edulis (Bayne 1976). The effect of temperature on IR and CR of catarina scallop juveniles was similar to the scope for activity, increasing from 16 to 19. and 22 °C. and decreasing to 25 and 28 °C. This indicates that IR and CR also serve as physiological indicators for ecophysi- ology studies. However, in A. irradians, CR remains independent TABLE 4. Covariance analysis between growth and temperature of juvenile catarina scallop [Argopecteii ventricosus-circularis) with time as the covariate factor. Figure. 10. Irrigation efficiency of catarina scallop {Argopecteii ven- tricosus-circularis) juveniles at different temperatures. Source of Sum of Mean Variation Squares D.F. Square f P Time 592.1 1 592.1 281.3 0.000 Temp 332.6 4 83.0 39.4 0.000 Residual 2109.1 1002 2.1 Optimum Temperature for Catarina Scallop Growth 391 TABLE 5. Tukey's multiple range test for growth of juvenile catarina scallop (Argopecten veiitricosus-circularis) at different temperatures. Temperature n X Homogeneous Groups 16 166 10.50 X 19 39 11.65 X 22 16 11.52 X 25 29 10.23 X 28 39 10.44 X of temperature between 10 and 26 °C and is reduced considerably at 5 °C (Kirby-Smith 1970). No explanation could be drawn from the differences of CR versus temperature between these closely related species. In the present work, CR was calculated assuming 100% reten- tion efficiency. In filter feeders, the amount of food available to the organism varies directly with the volume of water pumped through the mantle cavity and the retention efficiency by the gill (Bricelj and Shumway 1991). The majority of filter- feeding bivalves are able to retain particles of 3^1 p.m diameter or larger with 100% efficiency. This declines at lower particle size (25-90% for 2-p.m particles) (M0hlenberg and Riisgard 1978. Riisgard 1988). In five members of the Pectinidae family, 100% retention efficiency has only been obtained when scallops were fed larger particles (5-7 p.m). In the present work, IR and CR were studied using /. galbana cells of 3^4 |j.m diameter and 46-74 p,m3 volume (Enright et al. 1986). It would be convenient to determine retention efficiency versus particle size in the catarina scallop. Another physiological indicator used in the present work was the irrigation effciency (IE), also called convection requirement. In the catarina scallop. IE was a maximum at 19 °C and varied between 3.7 and 8.8 1/ml 02 within the range of temperatures studied. This is lower than that reported for other bivalve species. Mollusks living in coastal waters normally filter 15 I or more water per ml CK equivalent (Jorgensen 1975). A mean irrigation effi- ciency of 17 (8-25) was reported in Pecten latiauratus (Jorgensen 1960) and between 15 (at 5 °C) and 39 (at 20 °C) in O.lg dry tissue weight (DTW) Chimin's operculahs (McLusky 1973). The low IE values found in the catarina scallop could be caused by the small size of the juveniles employed (0.0087 mg DTW), because irriga- tion efficiency varies inversely proportional to the weight of the organism (McLusky 1973). In his work, irrigation efficiency of a 1.0 g DTW scallop was 1 1.5 1/ml 02 as compared to 7.5 1/ml 02 in a 0.1 g animal. The higher IE at 19 °C found in the catarina scallop indicates a maximum physiological efficiency at this tem- perature. The results of scope for activity, ingestion rate (IR). clearance rate (CR). and IE were confirmed with the growth experiment. In this work, growth was a maximum at 19 and 22 °C and lower at lower (16 °C) and higher temperatures (25-28 °C). These results are in agreement with those of Monsalvo-Spencer (1998) in the same species, who reported a growth rate of 70 p,m/day at 20 °C and 44 p.m/day at 28 °C in 2.2 mm shell height juveniles. A similar correlation between growth rate and scope for activity has been reported in adult Corbiculafluminea (Foe and Knight 1986), Cras- sostrea gigcis (Le-Gall and Raillard 1988), Concholepas conchole- pas (Gonzalez et al. 1990), and in juvenile Venerupis pullastra (Albentosa et al. 1994) and Ostrea edulis (Beiras et al. 1994). ACKNOWLEDGMENTS The authors thank Pedro Cruz and Jesus Bautista for their as- sistance during the experimental part of this work and Salvador Lluch for help with the figures. We also thank Dr. Ellis Glazier for correcting this English language manuscript. LITERATURE CITED Albentosa. M.. R. Beiras. A. P. Camacho, 1994. Determination of optimal thermal conditions for growth of clam {Venerupis pullastra) seed. Aquaculture 126:31-328. Alvarez-Borrego, S., L. A. Galindo-Bect, A. Chee-Barragan. 1975. Carac- terfsticas hidroquimicas de Bahi'a Magdalena, B.C.S. Ciencias Mari- nas, Mexico 2:94-110. Bayne, B. L. 1971. Oxygen consumption by three species of lamellibranch mollusc in declining ambient oxygen tension. Comp. Bioch. Phys. 40A: 955-970. Bayne, B. L. 1973. Aspects of the metabolism of Mytilus edulis during starvation. Nether. J. Sea Res. 7:399^110. Bayne, B. L. 1976. Marine Mussels: Their Ecology and Physiology. Cam- bridge University Press. Cambridge. UK. Bayne, B. L.. R. J. Thompson, J. Widdows. 1973. Some effects of tem- perature and food on the rate of oxygen consumption by Mytilus edulis L. pp 181-193. In: W. Wieser (ed.) Effects of Temperature on Ecto- thermic Organisms. Springer-Verlag. Berlin. Beiras. R., A. Perez-Camacho, M. Albentosa. 1994. Comparison of the scope for growth with the growth performance of Ostrea edulis seed reared at different food concentrations in an open-flow system. Mar. Biol. 119:227-233. Brand. A. R.. D. Roberts. 1973. The cardiac responses of the scallop Pecten maximus (L.) to respiratory stress. J. Exp. Mar. Biol. Ecol. 13:29^43. Bricelj. V. M.. S. Shumway. 1991. Physiology: energy acquisition and utilization, pp. 305-346. In: S. E. Shumway (ed.). Scallops: Biology. Ecology, and Aquaculture. Elservier, New York. Bryan. J. R, J. P. Riley, P.J.LeB. Williams. 1976. A Winkler procedure for making precise measurements of oxygen concentration for productivity and related studies. J. Exp. Mar. Biol. Ecol. 21:191-197. Enright. C. T.. G. F. Newkirk. J. S. Craigie, J. D. Castell. 1986. Evaluation of phytoplankton as diets for juvenile Ostrea edulis L. J. Exp. Mar. Biol. Ecol. 96:1-13. Espina, S.. L. F. Buckle-Ramirez . 1994. Scope for growth as function of temperature, salinity, and body weight in Tivela stultorum (Mollusca. Lamellibranchia). J. Applied Aquacult. 4(4): 91-100. Foe. C. A. Knight. 1986. A thermal budget for juvenile Corbicula flu- minea. Malacol. Bull. Special Edition 2:35-38. Fry, F. E. J., 1947. The aquatic respiration of fish. pp. 1-63. In.M. E. Brown (ed.). The Physiology of Fishes. 1. Academic Press. San Diego. Gallager. S. M.. R. Mann. 1980. An apparatus for the measurement of grazing activity of filter feeders at constant food concentrations. Mar. Biol. Lett. 1:341-349. Garder. T.. E. Eliassen. 1954. The energy metabolism of Ostrea edulis Univ. Bergen Arbok. Nalurv. 3:1-7. Gonzalez. M. L.. M. C. Perez. D. A. Lopez. M. S. Bultano. 1990. Effect of temperature in the energy availability for growth of Concholepas con- cholepas (Bruguiere). Biol. Mar. 25(2):71-81. Iglesias, J. I. P., E. Navarro. 1991. Energetics of growth and reproduction in cockles (Cerastodenna edule): seasonal and age-dependent varia- tions. Mar. Biol. 111:359-368. Jorgensen, C. B. 1960. Efficiency of particle retention and rate of water transport in undisturbed lamellibranchs. J. du Conseil. Conseil Perma- nent International pour I'Exploration de la Mer 26:94-1 16. J0rgensen. C. B. 1975. Comparative physiology of suspension feeding. Ann. Rev. Physiol. 37:57-79. 392 SlCARD ET AL. Kinne. O. 1970. Temperature, invertebrate, pp. 407-514. In: O. Kinne (ed.). Marine Ecology. Wiley-Interscience, London. Kirby-Smith. W. W. 1970. Growth of the scallops. Argopecten irradians concenlricus (Say) and Argopecten gibbus (Linne), as influeced by food and temperature. Ph.D. thesis, Duke University. Durham. North Carolina. 126 pp. Le-Gall. J. L.. O. Raillard. 1988. Influence of temperature on the physiol- ogy of the oyster Crassostrea gigas. Oceanis. Doc. Oceanogr. 14: 5-7. Lynn. R. J., J. J. Simpson. 1987. The California current system: the sea- sonal variability of its physical characteristics. / Goeophys. Res. 92: 12947-12966. Maeda-Martinez. A. N. 1985. Studies on the physiology of shell formation in molluscan larvae, with special reference to Crepidula fomicata. Ph.D. thesis. University of Southampton. U.K. 155 pp. Maeda-Martinez, A. N., P. Ormart-Castro. 1995. Sistema marino para el crecimiento y engorda hasta la fase adulta de almeja catarina. Patent 180211 I.M.P.I. Mexico. Mangum. C, W. Van Winkle. 1973. Responses of aquatic invertebrates, pp. 147-163. In: F. J. Vemberg. (ed.). Physiological ecology of estua- rine organisms. University of South Carolina Press. Columbia. South Carolina. McLusky. D. S. 1973. The effect of temperature on the oxygen consump- tion and filtration rate of Chlamys [Aequipecten) operations (L.) (Bil- valvia). Ophelia 10:141-154. Mohlenberg, F., H. U. Riisgard. 1978. Efficiency of particle retention in 13 species of suspension feeding bivalves. Ophelia 17(2):239-246. Monsalvo-Spencer, P. 1998. Estudios sobre el cultivo de larvas y juveniles de almeja catarina Argopecten ventricosus (-circularis) (Sowerby II. 1842) en el laboratorio. Msci. thesis CICIMAR. IPN. Mexico, 90 pp. Navarro, E., J. I. P. Iglesias. 1995. Energetics of reproduction related to enviromental variability in bivalve mollusks. Haliotis 24:43-55. Navarro. E.. J. I. P. Iglesias, A. Perez-Camacho, U. Labarta. R. Beiras. 1991. The physiological energetics of mussels {Mytilus galloprovin- cialis Lmk) from different cultivation rafts in the Rio de Arosa (Galicia. N.W. Spain). Aquaculture 94:197-212. Newell. R. I. E. L. G. Johnson. L. H. Kofoed. 1977. Adjustment of energy balance in reponse to temperature change in Ostrea edulis. Oecologia 30:97-110. Read, K. R. H. 1962. Respiration of the bivalve mollusks Mytilus edulis L. and Brachiodontes demissus plicatulus Lamarck as function of size and temperarure. Comp. Biochem. Physiol. 7:89-101. Riisgard, H. V. 1988. Efficiency of particle retention and filtration rate in six species of Northeast American bivalves. Mar. Ecol. Prog. Ser. 45: 217-223. Rotthauwe, H. W. 1958. Untersuchungen zur Atmungsphysiologie und Osmoregulation bei Mytilus edulis mit einem kurzen Anhang Uber die Blutkonzentration von Dreissena polymorpha in Abhangigkeit vom Elektrolytgehalt des Ausssenmediums. Veroffentlichungen des Insti- tuts fiir Meeresforschung in Bremerhaven 5:143-159. Sassaman, C. C. P. Mangum. 1972. Adaptations to enviromental oxygen levels in infaunal and epifaunal anemones. Biol. Bull. 143:657-678. Shumway, S. E. 1983. Factors affecting oxygen consumption in the coot clam Mulinia laterialis (Say). Ophelia 22(21:143-171. Shumway. S. E. 1991. Scallops: Biology. Ecology, and Aquaculture. Elsevier Science. New York. Shumway, S. E.. T. M. Scott. 1983. The effects of anoxia and hydrogen sulphide on survival, activity, and metabolic rate in the coot clam, Mulinia laterialis (Say). /. Exp. Mar. Biol. Ecol. 71:135-146. Sicard, M. T. 1999. Temperatura letal superior y temperatura optima en una poblacion de almeja catarina (Argopecten ventricosus Sowerby II. 1842). M.S. thesis CICIMAR. IPN, Mexico. 96 pp. Silva-Loera. H. A. 1986. Efecto del tamano corporal, tension de oxfgeno y temperatura sobre la tasa de consumo de oxfgeno en la escalopa Ar- gopecten circularis (SowerbyHMollusca: Lamellibranchia). M.S. the- sis, ITESM. Mexico. 93 pp. Tang. P. S. 1933. On the rate of oxygen consumption by tissues and lower organisms as a function of oxygen tension. Quart. Rev. Biol. 8: 260- 274. Thompson, R. J., B. L. Bayne. 1972. Active metabolism associated with feeding in the mussel Mytilus edulis L. J. Exp. Mar. Biol. Ecol. 8:191— 212. Urrutia. M. B . J. 1. P. Iglesias, E. Navarro. J. Prou. 1996. Feeding and absorption in Cerastodenna edule under environmental conditions in the Bay of Marennesoleron (Western France). J. Mar. Biol. Ass. U.K. 76:431-450. Vahl. O. 1972. Efficiency of particle retention in Mytilus edulis L. Ophelia 10:17-25. Van Dam, K. 1938. On the utilization of oxygen and regulation of breath- ing in some aquatic animals. Dissertation, Drukkerij, Volharding, Groningen. Van Dam. L. 1954. On the respiration in scallops (Lamellibranchiata). Biol. Bull. 107:192-202. Walne. P. R. 1970. Studies on the food value of nineteen genera of algae to juvenile bivalves of the genera Ostrea. Crassostrea. Mercenaria. and Mytilus. Ministry of Agriculture Fisheries, Series 2. 26(5): 62 pp. Widdows, J. 1973. The effects of temperature on the metabolism and activity of Mytilus edulis. Neth. J. Sea Res. 7:387-398. Winter. J. E. 1973. The filtration rate of Mytilus edulis and its dependence on algal concentration, measured by a continuous automatic apparatus. Mar. Biol. 22:317-328. Journal of Shellfish Research, Vol. 18. No. 2, 393-399, 1999. LEVELS OF RECRUITMENT AND ADULT ABUNDANCE IN A COLLAPSED POPULATION OF BAY SCALLOPS (ARGOPECTEN IRRADIANS) IN FLORIDA DAN C. MARELLI, WILLIAM S. ARNOLD, AND CATHERINE BRAY Florida Department of Environmental Protection Florida Marine Research Institute 100 8th Avenue SE, St., Petersburg Florida 33701-5095 ABSTRACT We monitored the recruitment and density of adult bay scallops in a Gulf Coast of Florida population that had collapsed and not recovered. Recruitment and adult stock density were not closely related, and, although interannual variations in the levels of both were observed, no significant populational responses, including recovery of the population, were observed. Such natural stochastic events as toxic algal blooms, severe weather, and freshwater discharge, as well as human influences, act to depress populations. At low densities, recruitment failure because of low fertilization success is likely, and recovery of a population may be constrained by depensatory effects when stock density in a local population is low. Scallops were distributed patchily. and aggregative spawning as well as self fertilization may enhance spawning events generated from low adult densities. Natural recovery of populations may take a decade or more, and stock enhancement efforts may be necessary to restore harvestable stocks more quickly. A'£V WORDS: bay scallop, Argopecien, recruitment, stock-recruitment relationships INTRODUCTION Fisheries biologists and ecologists are interested in the relation- ship between adult abundance and recruitment in broadcast- spawning invertebrates, albeit for different reasons. Biologists and managers concerned with invertebrate fisheries resources are in- terested in achieving some desirable and quantifiable goal, such as increasing the number of target organisms caught or preventing the overharvest of finite fisheries resources (Dredge 1988, Fogarty 1989, Murawski and Serchuk 1989, Quinn et al. 1993, Shepherd and Partington 1995). In some cases, managers need to make short-term predictions of the status of fisheries stocks and adjust allowable harvests based on those predictions. Fisheries models based on estimates of recruitment, growth, and mortality are de- sirable but problematic for short-lived invertebrates (Stearns 1976). The Australian abalone fishery is one example of how invertebrate fisheries can be managed based on stock assessment estimates (Prince and Shepherd 1992). Ecological researchers have attempted to quantify relationships between stock and recruits to understand the population dynamics of broadcast-spawning invertebrates more completely. Although this information also would be useful to fisheries managers, the efforts at quantification have largely been unsuccessful for a va- riety of reasons (Loosanoff 1964, Loosanoff 1966. Hancock 1973, Muus 1973, Wolff 1988, Caddy 1989, Peterson and Summerson 1992, Rodriguez et al. 1993. Olafsson et al. 1994, McShane 1995). Much of the basic research necessary to discern these relationships is beyond the scope or mission of fisheries scientists and begs for interdisciplinary cooperation with ecologists. We have begun to focus on stock-recruitment relationships in Floridian populations of the bay scallop, Argopecten irradians (Lamarck 1819). Many bay scallop populations in Florida have experienced severe reductions in abundance during the past three decades (Arnold 1990). ostensibly because of overharvest and hu- man-induced deterioration of water quality and scallop habitat. Concern has been expressed by citizens and biologists (Arnold et al. 1995) about the failure of many local scallop populations to recover following major declines. Stock densities in some popu- lations may be so low that recruitment failure resulting from low fertilization rates is acting against recovery. Research on fertiliza- tion dynamics in echinoderms suggests that this is a major concern (Pennington 1985. Levitan et al. 1992, Babcock et al. 1994, Levi- tan 1995). We examined and describe here the relationship between adult abundance and recruitment in a local (sensu Hanski and Gilpin 1991 ) Floridian population of A. irradians that was very abundant before 1990 but has since declined and failed to recover. We discuss possible management strategies to enhance local bay scal- lop populations that rely on basic ecological research. We also discuss the dynamics of the local scallop population and speculate on why the population has failed to recover. MATERIALS AND METHODS We began surveying adult scallop populations in June 1992; recruitment monitoring began during the late summer of 1992. We selected Homosassa Bay (Fig. 1 ) as a study site because of the relatively high abundance of bay scallops there. Unfortunately, the scallop population there declined substantially before we began the monitoring effort (Arnold et al. 1993). Initially, we surveyed scallops within an area bounded by the latitudes 28°50'N and 28°40'N and depth contours from 1.2 m (4 ft.) to 3.7 m ( 12 ft.) below mean lower low water (MLLW). The latitudinal boundaries generally represent the geographic limits of Homosassa Bay, and the vertical (depth) boundaries were chosen based upon bay scallop distribution data gathered by Blake et al. ( 1991 ). We realigned the study area in 1993 to focus on habitats between 0.6 m (2 ft.) and 1.8 m (6 ft.) below MLLW. and we extended the northern latitudinal boundary to approximately 28°53'N. These adjustments were made to include what we had determined to be the limits of the majority of the historical Ho- mosassa population. Adult Survey We surveyed the Homosassa Bay scallop population each June from 1992 through 1997. In 1992. we sampled 37 randomly lo- cated survey sites within the study area. At each site a 300-m. 393 394 Marelli et al. PENSACOLA BAY ST. JOSEPH BAY * GULF OF MEXICO HOMOSASSA BAY MARTINS KEYS CRYSTAL RIVER SPRINGS HOMOSASSA SPRINGS 0 1 Kilometer Figure 1. Location of the study area along the Florida Gulf coast. Inset shows detail of Homosassa Bay. weighted transect line was deployed in a roughly equilateral tri- angle. A pair of SCUBA divers swam along the transect and counted all scallops within 1 m of each side of the transect. Fol- lowing the 1992 survey and revision of the study area, 20 new stations were randomly chosen, and these stations were surveyed throughout the remainder of the study. Stations were assigned a posteriori to latitudinal groups: north (n = 6). central (n = 6). and TABLE 1. First deployment and final retrieval dates of Argopecten irradians recruit collectors in Homosassa Bay, Florida, 1992 to 1997. Year First Deployment Final Retrieval ID £ 3 1992 08/11/92 12/16/92 Z 3 E 1993 08/13/93 11/16/93 1994 08/17/94 02/12/95 Fi 1995 07/27/95 02/08/96 1996 08/07/96 01/29/97 re 2 9; < O en 2 E 1992 1993 1994 1995 YEAR 1997 Figure 2. Levels of recruitment and adult abundance of bay scallops {Argopecten irradians) in Homosassa Bay, 1992 to 1997. Figures rep- resent mean and standard error for each year. Bay Scallop Recruitment and Adult Abundance 395 TABLE 2. Analysis of variance for effects of year and latitudinal position on density of adult Argopecten irradians in Homosassa Bay, Florida. 1992 to 1997. TABLE 3. Analysis of variance for effects of year and latitudinal position on recruitment of Argopecten irradians to recruit collectors in Homosassa Bav. Florida, 1992 to 1997. Source d.f. SS MS F P Source d.f. SS MS F P Year 5 11.99 2.40 8.89 <.0001 Year 5 2.99 x 10"4 5.97 x 10~5 25.67 <.0001 Latitude 2 0.52 0.26 0.96 .386 Latitude 2 1.40 x I0~6 1.70 x 10"7 0.30 .739 Yr. x lat. 10 3.89 0.39 1.44 .161 Yr. x Lat. 10 3.19 x lO"5 3.19 x 10"" 1.37 .190 Residual 256 69.10 0.27 Residual 612 1.42 x 10-3 2.41 x 10"6 Total 275 85.33 Total 629 1.75 x 10-3 south (n = 8). Survey data were natural log-transformed [y' = Inly + 1 )] and were analyzed using a two-way factorial analysis of variance (ANOVA) with factors of year and latitude. Hochberg's GT2 method was used to compare means when F-ratios for main effects were significant (P < .05) (Hochberg 1974); this method is useful when variances are equal, but sample sizes are unequal (Day and Quinn 1989). Contouring software (Surfer version 6, Golden Software, Inc.) was also used to plot the adult survey data so that the relative spatial distribution and density patterns within the Homosassa population could be identified. Bay scallops are generally conta- giously distributed, and distribution patterns may be important in casting the distribution of future populations. Recruitment Monitoring In late August 1992, we deployed three recruit collectors at each of 20 randomly located sites in order to monitor scallop recruitment within the study area. A recruit collector consists of a '/^-bushel citrus bag containing a 0.135-m2 rectangle of 4-mm polypropylene mesh (Ambrose et al. 1992, Arnold et al. 1998). Collectors were suspended approximately 0.5 m above the sub- strate with an external doughnut float and held in position with a concrete anchor. Three weeks after the initial deployment, we placed a second set of three collectors at each site. Collectors were retrieved and replaced with new ones every 6 weeks until mid- December. The sampling period was chosen to maximize sampling JUNE 1992 JUNE 1993 St. Martins Keys ^ Land I I Survey Boundary Adult Scallop Densities Figure 3. Distribution and abundance of adult bay scallops [Argopecten irradians) in Homosassa Bay, June 1992 to 1993. Densities are number of scallops • 100 m~2. Survey stations are indicated by (•). 396 Marelli et al. of recruits based on the timing of scallop reproduction near these latitudes (Barber and Blake 1983) and on data from Sastry ( 1965) concerning the length of larval life. Missing collectors were re- placed on the redeployment date. All identifiable bay scallop re- cruits were counted in the lab. and a recruitment rate (number of recruits per collector per day) was calculated for each trap recov- ered. In 1993, we reduced the number of recruitment monitoring sites to nine located on three east-west transects in arbitrarily chosen north, central, and southern locations within the survey area. Each transect included three stations located at depths of 0.6 m. 1.2 m, and 1.8 m below MLLW. Collectors were deployed as in 1992: three per station beginning in early August, and three more deployed after 3 weeks. Collectors were retrieved every 6 weeks and replaced until mid-November 1993. The 1993 experi- mental design was continued from 1994 through 1997 with only slight variations in deployment and final retrieval dates (Table 1). We organized 1992 recruitment data a posteriori by latitudinal position into three groups — north (n = 4). central (n = 10), and south (n = 6) — so that these data could be integrated with the 1993 to 1997 data. The effects of latitude and year on recruitment rate were analyzed using a two-way factorial ANOVA following an arcsin-square-root transformation of the data. Because missing collectors created an unbalanced design, the analysis was con- ducted using the SAS GLM procedure (SAS Institute 1985). Where F-ratios were significant, Hochberg's GT2 method for com- paring means was applied. Finally, the relationship between recruitment and subsequent adult densities in Homosassa was examined with a simple linear regression. We also examined the relationship between adult den- sity and subsequent levels of recruitment. RESULTS Adult densities were very low in Homosassa from 1992 through 1997 and were not much greater than 1 scallop -100 m~2 until 1997 (Fig. 2). Latitude had no significant influence on scallop density (P = .39). but the effect of year on density was significant (Table 2). Adult scallop density was greatest in 1997, the only year in which scallop density differed significantly from that in any other year. Rates of recruitment to the collectors in the Homosassa area were also very low throughout the period we examined (Fig. 2). Recruitment in 3 of these years (1993, 1995, and 1996) was vir- tually undetectable. Analysis of these data suggested that no sig- nificant differences existed between latitudes within the bay (P = .74), but that year had a significant effect on recruitment (Table 3). Mean recruitment was significantly greater in 1994 than in any other year examined. Contour plots of Homosassa scallop density suggest that the population is distributed in patches that have recently had little geographic consistency within the region and that densities have been low throughout the area (Figs. 3-5). Despite interannual variations in density and position, aggregations seem to be a con- stantly identifiable feature in the population. JUNE 1994 JUNE 1995 j. \ Land I I Survey Boundary Adult Scallop Densities Figure 4. Distribution and abundance of adult bay scallops iArgopecten irradians) in Homosassa Bay, June 1994 to 1995. Densities are number of scallops ■ 100 m~2. Survey stations are indicated by (•). Bay Scallop Recruitment and Adult Abundance 397 JUNE 1996 JUNE 1997 Land I I Survey Boundary Adult Scallop Densities Figure 5. Distribution and abundance of adult bay scallops. {Argopecten irradians) in Homosassa Bay, June 1996 to 1997. Densities are number of scallops • 100 nr~. Survey stations are indicated by (•). Bay scallops in Florida are essentially annual animals (Barber and Blake 1983), and therefore, both recruitment and adult abun- dance are temporally related. However, our data demonstrated no significant relationship between recruitment and subsequent year- class density (r2 = 0.003) nor any relationship between adult density and subsequent recruitment (r = 0.001) (Fig. 2). DISCUSSION Low recruitment levels and low densities of adults in the Ho- mosassa bay scallop population, with minor, discordant increases and decreases, may suggest a population collapse. Recruitment levels and adult densities in Homosassa have been lower than those of populations in all other areas we have monitored in Florida (Arnold et al. 1998). In addition, levels of bay scallop recruitment and adult abundance in Homosassa were far exceeded by the values of recruitment and abundance seen in the robust Floridian populations of Steinhatchee and St. Joseph Bay (Arnold et al. 1998) as well as some North Carolina populations (Summer- son and Peterson 1990. Peterson and Summerson 1992) in which basin-scale coherence of levels of recruitment and adult density have been identified. Stock-recruitment relationships in annual broadcast-spawning animals such as scallops are generally assumed to be very weak and difficult to demonstrate, and effects of harvest are assumed to be inconsequential (Garcia and LaReste 1981. Sale 1990. Mc- Shane 1995). However, there is clearly a relationship between recruitment and adult density when harvest pressure severely re- duces population abundances, a situation typically described as "recruitment overfishing" (Pauly 1980, Dredge 1988, Breen 1992. Jamieson 1993, Shepherd and Partington 1995). Low adult densities in the Homosassa population could result in poor recruitment and subsequent year-class strength for such reasons as reproductive failure, fertilization constraints (Clavier 1992, Yund and McCartney 1994), and postsettlement effects, in- cluding harvest pressure (Peterson and Summerson 1992). Conse- quently, it may be extremely difficult to identify stock-recruitment relationships at low scallop densities because of stochastic influ- ences. The most parsimonious explanation for persistent low stock levels is that low densities reduce the probability of fertilization success. Levitan and Petersen (1995) reviewed experiments that have demonstrated this empirically in broadcast-spawning inver- tebrates, and they suggested that distances in excess of 1 meter may preclude the fertilization rates conducive to successful re- population. Low stock abundance in depauperate populations can be per- petuated by unsuccessful fertilization because of the difficulty of locating a mate (Allee effect; Allee 1938, Meyers et al. 1995). Adding the effect of predator saturation (typical of lower latitudes, see Vermeij 1978) can lead to depensation (Allee 1938). Bay scallops may be adapted for dealing with constraints to fertiliza- tion, because they are simultaneous hermaphrodites and capable of self-fertilization (Kellogg 1892. Wilbur 1995). Self-fertilization, although probably not adaptive in the long term, may allow a depauperate local population to persist until it is able to overcome 398 Marelli et al. the constraints or until larvae from adjacent local populations re- supply the local population and increase fertilization probabilities. Aggregative spawning by scallops can also mitigate popula- tion-wide low spawner densities. Recruits in Homosassa may be the product of very few spawners. because aggregations may have an unduly large influence on fertilization in depressed populations, and because bay scallops are undoubtedly dependent upon larval retention mechanisms in local areas to resupply subsequent year classes (Peterson and Summerson 1992). Recovery of collapsed populations to "normal" levels may take a decade or more (Peter- son and Summerson 1992). Management strategies that seek to enhance or restore populations by increasing fertilization success (e.g., Tegner 1992, Quinn et al. 1993, Tettelbach and Wenczel 1993. Blake 1998) or overcoming postsettlement effects (see Olaf- sson et al. 1994) may help to restore such depleted populations if ecosystem-wide conditions are otherwise amenable. ACKNOWLEDGMENTS We thank Melissa Harrison, Kate Hagner, Philip Hoffman. Melanie Parker, Justin Styer, and Hutch Craig for their assistance with data collection. Winnie White provided assistance with con- tour plotting and figures. Robert Glazer and Dr. Michael Murphy provided suggestions on the manuscript. This research was sup- ported by revenues collected under the Florida Saltwater Fishing License. LITERATURE CITED Allee. W. C. 1938. The Social Life of Animals. Norton, New York. 239 pp. Ambrose, W. G„ Jr., C. H. Peterson. H. C. Summerson & J. Lin. 1992. Experimental tests of factors affecting recruitment of bay scallops (Ar- gopecten irradians) to spat collectors. Aquaculture 108:67-86. Arnold, W. S. 1990. A review of the biology of the bay scallop, Argopecten irradians, in Florida waters. Report to the Florida Marine Fisheries Commission. 59 pp. Arnold, W. S.. D. C. Marelli & C. Bray. 1993. Annual report of the bay scallop project 1991-1992. Report to the Florida Marine Fisheries Commission. 32 pp. Arnold. W. S„ D. C. Marelli. C. Bray & M. Harrison. 1995. Annual report of the bay scallop project 1994. Report to the Florida Marine Fisheries Commission. 71 pp. Arnold, W. S.. D. C. Marelli. C. P. Bray & M. M. Harrison. 1998. Recruit- ment of bay scallops (Argopecten irradians) in Floridan Gulf of Mexico waters: scales of coherence. Mar. Ecol. Prog. Ser. 170:143- 157. Babcock. R.C., C.N. Mundy & D. Whitehead. 1994. Sperm diffusion models and in situ confirmation of long-distance fertilization in the free-spawning asteroid Acanthaster planet. Biol. Bull. 186:17-28. Barber. B. J. & N. J. Blake. 1983. Growth and reproduction of the bay scallop. Argopecten irradians (Lamarck) at its southern distributional limit. J. Exp Mar. Biol. Ecol. 66:247-256. Blake, N.J. 1998. The potential for re-establishing bay scallops to the estuaries of the west coast of Florida. Transactions of the 63rd North American Wildlife and Natural Resources Conference. 184-189. Blake, N. J., R. L. Darden & M. A. Moyer. 1991. Population survey data for the bay scallop, (Argopecten irradians). off Crystal River, Florida. 1991. Report to the Florida Department of Natural Resources, Florida Marine Research Institute, St. Petersburg, FL. 60 pp. Breen, P. A. 1992. A review of models used for stock assessment in aba- lone fisheries, pp. 253-275. In: S. A. Shepherd. M. J. Tegner and S. A. Guzzman del Proo (eds.). Abalone of the World: Biology, Fisheries, and Culture. Blackwell Scientific Publications. Oxford. U.K. Caddy. J. F. 1989. A perspective on the population dynamics and assess- ment of scallop fisheries, with special reference to the sea scallop. Placopecten magellanicus Gmelin. pp. 559-589. In: J. F. Caddy led.). Marine Invertebrate Fisheries: Their Assessment and Management. John Wiley & Sons. New York. Clavier, J. 1992. Fecundity and optimal sperm density for fertilization in the ormer (Haliotis tuberculata L.). pp. 86-92. In: S. A. Shepherd. M. J. Tegner and S. A. Guzman del Proo (eds.). Abalone of the World: Biology. Fisheries, and Culture. Blackwell Scientific Publications, Ox- ford, U.K. Day. R. W. & G P. Quinn. 1989. Comparisons of treatments after an analysis of variance in ecology. Ecol. Monogr. 59:433^163. Dredge, M. C. L. 1988. Recruitment overfishing in a tropical scallop fish- ery? J. Shellfish Res. 7:233-239. Fogarty, M. J. 1989. Forecasting yield and abundance of exploited inver- tebrates, pp. 701-724. In: J. F. Caddy (ed.). Marine Invertebrate Fish- eries: Their Assessment and Management. John Wiley & Sons. New York. Garcia, S. & L. LaReste. 1981. Life cycles, dynamics, exploitation, and management of coastal penaeid shrimp stocks. FAO Fisheries Tech. Paper 203. 215 pp. Hancock. D. A. 1973. The relationship between stock and recruitment in exploited invertebrates. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 164: 113-131. Hanksi, I. & M. Gilpin. 1991. Metapopulation dynamics: brief history and conceptual domain. Biol. J. Linn. Soc. 42:3-16. Hochberg, Y. 1974. Some conservative generalizations of the T-method in simultaneous inference. J. Multivar. Anal. 4:224-234. Jamieson. G S. 1993. Marine invertebrate conservation: evaluation of fish- eries over-exploitation concerns. Am. Zool. 33:551-567. Kellogg. J. L. 1892. A contribution to our knowledge of the morphology of lamellibranchiate mollusks. Bull. U.S. Fish. Comm. X:389-136. Levitan. D. R. 1995. The ecology of fertilization in free-spawning inver- tebrates, pp. 123-156. In: L. McEdward (ed.). Ecology of Marine In- vertebrate Larvae. CRC Press, New York. Levitan, D. R. & C. Petersen. 1995. Sperm limitation in the sea. Trends Ecol. Evol. 10:228-231. Levitan. D. R.. M. A. Sewell & M. S. Chia. 1992. How distribution and abundance influence fertilization success in the sea urchin Strongylo- centrotus franciscanus. Ecology 73:248-254. Loosanoff, V. L. 1964. Variations in time and intensity of setting of the starfish. Asterias forbesi. in Long Island Sound during a 25-year period. Biol. Bull. 126:423-439. Loosanoff, V. L. 1966. Time and intensity of setting of the oyster. Cras- sostrea virginica. in Long Island Sound. Biol. Bull. 130:211-227. McShane. P. E. 1995. Recruitment variation in abalone: its importance to fisheries management. Mar. Freshwater Res. 46:555-570. Meyers. R. A.. N. J. Barrowman. J. A. Hutchings & A. A. Rosenberg. 1995. Population dynamics of exploited fish stocks at low population levels. Science 269:1 106-1 108. Murawski. S. M. & F. M. Serchuk. 1989. Mechanized shellfish harvesting and its management, pp. 479-506. In: J. F. Caddy (ed.). Marine Inver- tebrate Fisheries: Their Assessment and Management. John Wiley & Sons. New York. Muus. K. 1973. Settling, growth, and mortality of young bivalves in the Oresund. Ophelia 12:79-116. Olafsson. E. B.. C. H. Peterson & W. G Ambrose. Jr. 1994. Does recruit- ment limitation structure populations and communities of macro- invertebrates in marine soft sediments: the relative significance of pre- and post-settlement processes. Oceanogr. Mar. Biol. Annu. Rev. 32: 65-109. Pauly, D. 1980. A selection of simple methods for the assessment of tropical fish stocks. FAO Fisheries Circular 729. 55 pp. Pennington. J. T. 1985. The ecology of fertilization of echinoid eggs: the Bay Scallop Recruitment and Adult Abundance 399 consequences of sperm dilution, adult aggregation, and synchronous spawning. Biol. Bull. 164:417-430. Peterson. C. H. & H. C. Summerson. 1992. Basin-scale coherence of popu- lation dynamics of an exploited marine invertebrate, the hay scallop: implications of recruitment limitation. Mar. Ecol. Prog. Ser. 90:257- 272. Prince. J. D. & S. A. Shepherd. 1992. Australian abalone fisheries and their management, pp. 407^426. In: S. A. Shepherd. M. J. Tegner and S. A. Guzman del Proo (eds.). Abalone of the World: Biology. Fisheries, and Culture. Blackwell Scientific Publications, Oxford, U.K. Quinn. J. F., S. R. Wing & L. W. Botsford. 1993. Harvest refugia in marine invertebrate fisheries: models and applications to the red sea urchin. Strongylocentrotus franciscamts. Am. Zool. 33:537-550. Rodriguez, S. R.. F. P. Ojeda & N. C. Inestrosa. 1993. Settlement of benthic marine invertebrates. Mar. Ecol. Prog. Ser. 97:193-207. Sale. P. F. 1990. Recruitment of marine species. Is the bandwagon rolling in the right direction'.' Trends Ecol. Evol. 5:25-27. SAS Institute. 1985. SAS User's Guide: Statistics. Version 5. SAS Insti- tute. Cary. NC. 956 pp. Sastry, A. N. 1965. The development and external morphology of pelagic- larval and postlarval stages of the bay scallop, Aequipecten irradians concentricus Say, reared in the laboratory. Bull. Mar. Sci. 15:417—435. Shepherd, S. A. & D. Partington. 1995. Studies on southern Australian abalone (Genus Haliotis). XVI. recruitment, habitat, and stock rela- tions. Mar. Freshw. Res. 46:669-680. Stearns, S. C. 1976. Life history tactics: a review of the ideas. Q. Rev. Biol. 51:3-47. Summerson. H. C. & C. H. Peterson. 1990. Recruitment failure of the bay scallop. Argopecten irradians concentricus. during the first red tide, Ptychodiscus brevis, outbreak recorded in North Carolina. Estuaries 13:322-331. Tegner, M. J. 1992. Brood-stock transplants as an approach to abalone stock enhancement, pp. 461^473. In: S. A. Shepherd, M. J. Tegner and S. A. Guzman del Proo (eds.). Abalone of the World: Biology. Fish- eries, and Culture. Blackwell Scientific Publications, Oxford. U.K. Tettelbach, S. T. & P. Wenczel. 1993. Reseeding efforts and the status of bay scallop Argopecten irradians (Lamarck. 1819) populations in New York following the occurrence of '"brown tide" algal blooms. J. Shell- fish Res. 12:423-431. Vermeij, G. R. 1978. Biogeography and adaptation. Harvard University Press, Cambridge. MA. 333 pp. Wilbur. A. E. 1995. Population genetics of the bay scallop, Argopecten irradians (Lamarck): an analysis of geographic variation and the con- sequences of self-fertilization. Ph.D. dissertation. University of Dela- ware, 128 pp. Wolff, M. 1988. Spawning and recruitment in the Peruvian scallop Ar- gopecten purpuratus. Mar. Ecol. Prog. Ser. 42:213-217. Yund. P.O. & M. A. McCartney. 1994. Male reproductive success in sessile invertebrates: competition for fertilizations. Ecology 75:2151- 2167. Journal of Shellfish Research, Vol. 18. No. 2. 401-404. 1999. THE TELOMERES OF THE BAY SCALLOP, ARGOPECTEN IRRADIANS (LAMARCK) STEPHEN L. ESTABROOKS Nantucket Marine Laboratory 0 Easton Street Nantucket, Massachusetts 02554 ABSTRACT The telomeres of the bay scallop, Argopecten irradians (L.) were isolated and identified using Southern blotting and chemiluminescent hybridization techniques. Four probes, (TTAGGGh. (TTGGGG),, (TTAGGC),, and (TTAGG), were used to determine the degree of crossreactivity with telomere sequences of other phyla. The bay scallop seems to have the same telomeric repeat sequence found in all vertebrates studied to date: namely, (TTAGGG)n. The possible role of telomeres in the short lifespan of this species is discussed. KEY WORDS: telomeres, scallop, Argopecten, aging, lifespan INTRODUCTION The bay scallop, Argopecten irradians (L.) is a commercially important species found as three subspecies along the east coast of the United States and into the Gulf of Mexico (Belding 1910, Clarke 1965). This species lives less than 2 years and generally reproduces only once in its lifetime (Belding 1910); whereas, other species of scallops may live to 5, 10, and even 20 years or more (Vahl 1981, Shafee and Lucas 1982. MacDonald 1986). Many factors, including fluctuations in eelgrass populations, water tem- perature, salinity, currents, food sources, predation, siltation, over- fishing, coupled with the short lifespan of the scallop, have been blamed for the wide fluctuations seen in annual bay scallop har- vests over the years (Belding 1910, Tettelbach et al. 1985, Peterson et al. 1989). The lifespan of the northern subspecies of the bay scallop, A. irradians irradians (L.) has been genetically determined to be between 18-26 months (Belding 1910, Gutzell 1931). with death usually occuring within a window of approximately 4 to 5 months in late winter and early spring in the northern subspecies, with most scallops dead by the end of May. A few, estimated to be between 5-10%. may survive to reproduce again (Belding 1910). Recent studies have shown that telomeres, tandem repeats of usually five to six bases that are found at the end of chromosomes, may play a key role in determining just how long a species may survive (Harley 1991). With each cell replication, a number of telomeric repeats are lost, until a critically few remain, setting the stage for the beginning of cell senescence and death (Harley 1991, Allsopp et al. 1992, Vaziri et al. 1993). Previous studies involving telomeres from yeast to humans have found sequences similar or identical to the tandem repeat (TTAGGG),, found in all vertebrates studied to date (Moyzis et al. 1988. Meyne et al. 1991 ). Variations of this sequence include that of the single celled organism Tetrahymena thermaphila, which has the telomeric sequence (TTGGGG)n (Blackburn and Chiou 1981). A few insects studied to date have the sequence (TTAGG )n (Oka- zaki et al. 1993, Meyne and Imai, 1995); whereas, two marine polychaete worms have also been shown to have the vertebrate sequence (An et al. 1995). The sequence (TTAGGC)n has been identified in the terrestial worm, Ascaris lumbricoides (Muller et al.. 19911. To date, only one study has been done on the phylum Mollusca, which tentatively identified the telomeric sequence of the oyster. Crassostrea gigas (Thunberg) to be (TTAGGG )n using fluorescent in situ hybridization (FISHl techniques (Guo and Allen 1997), who also felt that Southern blotting and hybridization with a (TTAGGG) probe would be necessary to confirm the telomere sequence. It is hypothesized that the bay scallop may begin its life with fewer telomeres, and subsequently, may run out of telomeres sooner than longer-living species. Before performing quantifying studies on the telomeres of long- and short-lived bivalve mollusks, it was first necessary to determine the telomere sequence of the bay scallop. EXPERIMENTAL PROCEDURES Bay scallops were collected from the waters surrounding Nan- tucket Island off the coast of Massachusetts. Approximately 250 mg of digestive gland (excluding digestive tract), kidney, or gonad tissue were homogenized in 5.0 ml of DNAzol obtained from Molecular Research, Inc. (Cincinnati. OH). After remaining at room temperature for 5 minutes, the homogenate was spun at 10,000 x g for 10 minutes, the supernate was poured off into new tubes, and 0.5 ml of 100% ethanol was added. The precipitated DNA was spun at 5,000 x g for 2 minutes and washed 2 times in 95% ethanol. This fraction still contained mucopolysaccharides that were removed by extraction with 200 u.1 of chloroform. The cleaned-up DNA was reprecipitated with 100% ethanol, washed once in 95% ethanol, and dissolved in TE buffer. pH 8.0. DNA purity was determined to have a 260/280 ratio of 1.8 to 1.9, and the integrity of the genomic DNA checked by electrophoresis in a 0.8% gel at 105 V for 30 minutes. Restriction Enzyme Digestion Aliquots of scallop DNA were digested with the restriction enzymes Rsal and Hinf I obtained from Life Technologies, Inc. (Grand Island. NY), according to the manufacturer's protocol. These enzymes cut genomic DNA into pieces that leave the telo- mere restriction fragments (TRF) at the end of the chromosomes intact. These fragments contain the tandemly repeated telomeres along with some subtelomeric DNA. The digested DNA was elec- trophoresed in a 0.8% agarose gel at 105 V for 60 minutes, then stained in 0.5% ethidium bromide for 15 minutes and photo- graphed. Blotting and Hybridization The gel was alkaline blotted onto a nylon membrane in 0.4M NaOH overnight, rinsed once in 2 x SSC buffer, and dried between sheets of blotting paper. The membrane was rinsed briefly in 401 402 ESTABROOKS MW 23.1 9.4 6.6 4.4 2.3 2.0 0.5 1234567 89 10 Figure 1. Increasing concentrations of DNA from the bay scallop A. irradians digested with Rsal and Hinfl and probed with (TTAGGG),, lanes 2-9; lane 1, undigested DNA; lane 10, lambda molecular weight marker (lkb). MW 23.1 9.4 6.6 4.4 2.3 2.0 0.5 1234567 89 Figure 3. Increasing concentrations of DNA from the bay scallop A. irradians digested with Rsal and Hinfl and probed with (TTAGG),, lanes 1-3; (TTAGGC),, lanes 5-8: lanes 4.9 lambda molecular weight marker (lkb). 0.25M Na-,HP04, and prehybridized for 1 hour in 25 mL contain- ing ImM EDTA, 7% SDS. (sodium dodecylsulphate) and 0.25M disodium phosphate. pH 7.2 on a hybridizing rotator at 37 °C. After the addition of 4-5 pmol/mL of biotinylated probe, obtained from Genesys. Inc. (The Woodlands, TX), the membranes were hybridized overnight at 37 °C. Four telomeric probes were used, (TTAGGG),, (TTGGGG),, (TTAGGC),. and (TTAGG),. The membranes were then washed in increasingly stringent washes as follows: 2 x minutes in 2 x SSC; 1% SDS at room temperature; 2 x 15 minutes in 1 x SSC; 1% SDS at 37 °C. according to the manufacturer, with the exception that the highest stringency wash was increased from 2x5 minutes to 2x15 minutes in 1 x SSC at room temperature. The membranes were developed according to the chemiluminescent procedure of Tropix. Inc. (Bedford, MA), which conjugates the enzyme, alkaline phosphatase attached to streptaviden to the biotinylated DNA probe. Alkaline phosphatase MW 23.1 9.4 6.6 4.4 2.3 2.0 12 3 4 5 Figure 2. Increasing concentrations of DNA from the bay scallop A. irradians digested with Rsal and Hinfl and probed with (TTGGGG),, lanes 1-3; lane 4, undigested DNA; lane 5, lambda molecular weight marker (lkb). then produces light as a byproduct of the reaction between alkaline phosphatase and the reagent 1,2 dioxatane. The membrane was inserted into a plastic sleeve and placed in an x-ray cassette and exposed from 2 minutes to 2 hours. The films were automatically developed in a Konica SRX-501A film processor. BalM Digestion Examples of the DNA were digested with the exonuclease. Bal3 1 to confirm that the DNA identified as telomeres were actu- ally located at the end of the chromosomes. Bal3 1 removes bases from the ends of DNA one base at a time. A digest consisting of 200 u.g of scallop DNA was incubated at 30 °C with fifty units Bal31 in 1.2 mL final volume of Bal31 buffer (Life Technologies. Inc.). At varying intervals, 150 p.L of digest were removed, and the digestion was halted with the addition of 10 p.L of 0.5M EDTA. DNA was then precipitated with 50 p.L of 100% ethanol, recon- stituted in TE buffer, ph 8.0. digested with the restriction enzymes Rsal and Hinfl. and hybridized with the probe (TTAGGG),. MW 23.1 9.4 6.6 4.4 23 2.0 0.5 12 3 4 5 Figure 4. Bal31 digestion of DNA from the bay scallop A. irradians probed with (TTAGGG),. Lane 1, T = 0 h; lane 2. T = 1 h; lane 3, T = 3 h; lane 4, T = 4 h; lane 5, lambda molecular weight marker ( lkb). Telomeres of the Bay Scallop 403 RESULTS The telomeric sequence of the bay scallop, Argopecten irradi- ans was found to be (TTAGGG),, (Fig. 1 ). Several bands, contain- ing telomere restriction fragments (TRF)s were found to be repro- ducible from scallop to scallop. These telomere restriction frag- ments, which contain the terminal telomere arrays plus a subtelomeric fraction to the point where the chromosome is cut by restriction enzymes, seem to be specific in length and are repro- ducible from scallop to scallop in the same age group (data not shown). Figure 1 also demonstrates the presence of telomeres of varying length found at any one time. Hybridization with the probes (TTAGGC),, (TTGGGG),, and (TTAGG), demonstrated only slight binding, which was essentially lost with high stringency washes (Figs. 2. 3). Membranes were then stripped and rehybrid- ized with (TTAGGG), to confirm that the telomeres were present initially and available for potential binding with the three other probes. Figure 4 demonstrates digestion of the ends of the chro- mosomes by the exonuclease Bal31, which removes nucleic acids from the terminal end of chromosomes one by one. confirming that the DNA hybridized by the probe actually were telomeres. After 4 hours, all of the telomeres had essentially been removed, leaving only some internal repeat sequences that were not available for digestion by Bal31 . DISCUSSION The vertebrate telomeric probe (TTAGGG), was used as a starting point to determine the telomeric sequence of the bay scal- lop. Guo and Allen ( 1997) found this to be the telomere sequence in the only mollusk. studied to date, the oyster. Crassostrea gigas. Three additional probes, representing three common nonvertebrate telomere sequences; namely, ( TTGGGG )n, the telomere sequence in Tetrahymena thermophila (Blackburn and Chiou 1981), (TTAGG),,. the pentanucleotide form found in several insects; for example, the silkworm Bombyx mori, and (TTAGGC),,. the telo- mere sequence in the nematode, Ascaris lumbricoides (Muller et al. 1991 ) were all found not to bind with the bay scallop telomeres. The individual bands, seen in Figure 1. contain the most numerous of the TRF lengths, and many of these telomeres are lost as the scallop ages (see Fig. 4, which is a Bal3 1 digest of a year 2 scallop, nearing the end of its lifespan). Compare to Fig. 1, which is the DNA of a much younger year 1 scallop. The bay scallop seems to share the same telomeric sequence as found in vertebrates and many other species studied to date; namely, (TTAGGG),, (Meyne et al. 1989). This highly conserved structure may be the mitotic clock that determines the lifespan of a particular species (Harley 1991 ). People suffering from progeria, a syndrome that involves a rapid aging process, have fewer telo- meres than healthy humans (Allsopp et al. 1992) as do those with Down's Syndrome (Vaziri et al. 1993). Both groups fail to live out a normal lifespan. Recently, Bodnar et al. (1998) were able to extend the lifespan of human cells in tissue culture by increasing the number of telo- mere repeats through the introduction of the gene that codes for the enzyme telomerase into the cells. It is hypothesized that the bay scallop may have fewer telomeres initially than, for example, the deep-sea scallop Placopecten magellanicus, (Gmelin), which can live to 20 years of age. with the bay scallop running out of telo- meres sooner. Alternatively, bay scallops may lose telomeres at a greater rate than longer-lived species. Studies are currently under- way to quantify the telomeres from both species at different ages. ACKNOWLEDGMENTS This research was supported in part by a PADI Foundation grant. LITERATURE CITED Allsopp. R. C. H. Vaziri. C. Patterson. S. Goldstein, E. V. Younglai, A. B. Futcher, C. W. Greider & C. B. Harley. (1992). Telomere length pre- dicts replicative capacity of human fibroblasts. Proc. Natl. Acad. Sci. USA 89:10114-10118. An. J H. A.. I. Dominquez, A. S. Balajee, T. H Hutchinson. D. R. Dixon & A. T. Natarajan. 1995. Localization of a vertebrate telomeric se- quence in the chromosomes of two marine worms (phylum Annelida), class Polychaeta). Chroma. Res. 3:507-508. Belding. D. L. (1910). A report upon the scallop fishery of Massachusetts. The Commonwealth of Massachusetts, Boston. 150 pp. Blackburn, E. H. & S. Chiou. (1981). Non-nucleosomal packaging of a tandemly repeated DNA sequence at termini of extrachromosomal DNA coding for rRNA in Tetrahymena. Proc. Natl. Acad. Sci. USA 78:2263-2267. Bodnar, A. G. M. Ouellette. M. Frolkis, S. E. Holt. C. Chiu, G B. Morin. C. B. Harley. J. W. Shay, S. Lichtsteiner & W. E. Wright (1998). Ex- tension of life-span by introduction of telomerase into normal human cells. Science 279:349-352. Clarke, A. H. (1965). The scallop subspecies Aequipecten irradians (La- marck). Malacologia 2:161-188. Guo, X. & S.J. Allen. Jr. (1997). Fluorescence in situ hybridization of vertebrate telomere sequence to chromosome ends of the Pacific oyster. Crassostrea gigas Thunberg. J. Shellfish Res. 16:87-89. Gutzell. J. S. (1931). Natural history of the bay scallop. Bull. U.S. Bureau Fish. 45:569-632. Harley, C. B. (1991). Telomere loss: mitotic clock or genetic time bomb? Mutation Res. 256:271-282. MacDonald. B. A. ( 1986). Production and resource partitioning in the giant scallop Placopecten magellanicus grown on the bottom and in sus- pended culture. Mar. Ecol. Prog. Ser. 34:79-86. Meyne, J., R. L. Ratliff & R. K. Moyzis. (19S9). Conservation of the human telomere sequence (TTAGGG),, among vertebrates. Proc. Natl. Acad. Sci. USA 86:7049-7053. Meyne, J. H. & H. T. Imai. (1995). FISH analysis of the telomere se- quences of bulldog ants (Myrmecia:Formicidae). Chromosoma 104: 14-18. Moyzis, R. K., J. M. Buckingham. L. Scott Cram, M. Dam, L. L. Deaven. M. D. Jones, J. Meyne, R. L. Ratliff & J.-R. Wu. (1988). A highly conserved repetitive DNA sequence (TTAGGG)n, present at the telo- meres of human chromosomes. Proc. Natl. Acad. Sci. USA 85:6622- 6626. Muller, D.. C. Wicky. A. Spicher & H. Tobler. (1991). New telomere formation after developmentally regulated chromosomal breakage dur- ing the process of chromatin diminution in Ascaris lumbricoides. Cell 67:815-822. Okazaki, S., K. Tsuchida, H. Mackawa & H. Fujiwara. (1993). Identifica- tion of a pentanucleotide telomere sequence (TTAGG),, in the silk- worm Bombyx mori and in other insects. Mol. Cell Biol. 13:142+ 1432. 404 ESTABROOKS Peterson, C. H., H. C. Summerson. S. R. Fegley & R. C. Prescott. ( 1989). mass mortality of northern hay scallops. Argopecten irradians irradians. Timing, intensity and sources of autumn mortality of adult bay scallops following a severe spring rainstorm. Veliger 27:381-385. Argopecten irradians concentricus Say. J. Exp. Mar. Biol. Ecol. 127: Vahl, O. (1981). Energy transformation by the Island scallop, Chlamys 121-140. islandica (O.F. Miller), from 70 ° N. I. The age-specific energy budget Shafee, M. S. & A. Lucas. (1982). Variations saisonnieres du bilan ener- and net growth efficiency. J. Exp. Mar. Biol. Ecol. 53:281-296. getique chez les individus d'un population de Chlamys varia (L.KBi- Vaziri, H., F. Schachter. I. Uchida. L. Wei. X. Zhu. R. Effros. D. Cohen & valvia. Pectenidae. Ocenaol. Acta 5:331-338. C. B. Harley. (1993). Loss of telomeric DNA during aging of normal Tettelbach, S. T., P. J. Auster, E. W. Rhodes & J. C. Widman. (1985). A and Trisomy 21 human lymphocytes. Am. J. Hum. Genet. 52:661-667. Journal of Shellfish Research, Vol. 18, No. 2. 405-413. 1999. TEMPORAL VARIATION IN SEA SCALLOP (PLACOPECTEN MAGELLANICUS) ADDUCTOR MUSCLE RNA/DNA RATIOS IN RELATION TO GONOSOMATIC CYCLES, OFF DIGBY, NOVA SCOTIA D. RODDICK,' E. KENCHINGTON,1 J. GRANT,2 AND S. SMITH1 'Department of Fisheries and Oceans Invertebrate Fisheries Division Bedford Institute of Oceanography Dartmouth. Nova Scotia Canada B2Y 4A2 'Department of Biological Oceanography Dalhousie University Halifax, Nova Scotia Canada B3H 4JI ABSTRACT This study examines the usefulness of RNA/DNA ratios as an index of nutritional and other stress in wild populations of the sea scallop (Placopecten magellanicus). The seasonal variation in the RNA/DNA ratios of the adductor muscle of sea scallops, off Digby, Nova Scotia, were determined, and compared to the seasonal variations in adductor muscle and gonad weights. Results show that the RNA/DNA ratio of the adductor muscle for this scallop population varies with the size or age of the scallop and that there are temporal variations on both interannual and seasonal scales. The seasonal variation in RNA/DNA ratios can be modeled and approxi- mates a sine function. This model explains a maximum of 45% of the variation in RNA/DNA ratios, and the residuals from the model are significantly correlated with subsequent growth. Temporal variations in RNA/DNA ratios are related to growth rate differences and reflect variations in food supply and temperature. RNA/DNA ratios seem to be a useful index of the health of a sea scallop stock, but without prior knowledge of what constitutes a level indicating increased mortality rates, it should not be the only health index used. KEY WORDS: Placopecten magellanicus, sea scallop. RNA/DNA. growth, seasonal variation INTRODUCTION RNA/DNA ratios have been used as an index of growth rate and indirectly, health for a variety of organisms (microorganisms Leick 1968: plankton Falkowski and Owens 1982, Dortch et al. 1983; higher invertebrates Sutcliffe 1969; including bivalves Pease 1976. Wright and Hetzel 1985; fish Buckley 1984, Hoven- kamp and Witte 1991, Canino 1997; and mammals Monroe and Gray 1969). The use of the RNA/DNA ratio as an index of growth is based on the fact that the amount of DNA per cell is remarkably constant for all normal somatic cells of a given species; whereas, RNA (ribonucleic acid) varies with the rate of protein synthesis. Because the amount of DNA per cell is essentially constant, and the amount of RNA varies with the level of protein synthesis, the ratio of RNA to DNA (RNA per unit DNA) is a self-calibrating index of the amount of protein synthesis taking place per cell. Kenchington ( 1994) proposed that the method could be used to monitor the health of scallops in areas closed to fishing activity. The purpose of such a monitoring program would be to provide a warning if the health of the scallops in the closed areas deteriorated to the point that an increased mortality rate was possible. In the first study to examine the potential of RNA/DNA ratios as a gen- eral index of health in a wild bivalve stock. Kenchington (1994) reported both spatial and interannual variation in the RNA/DNA ratios of the adductor muscle in the sea scallop {Placopecten ma- gellanicus) population off Digby, Nova Scotia, Canada. She con- cluded that the RNA/DNA ratio method would be useful in moni- toring the health of the scallops, but that the seasonal variation would have to be determined before the method could be applied. The purpose of this study was to examine the temporal variation in the RNA/DNA ratio in the adductor muscles of the adult Pla- copecten magellanicus population off Digby, Nova Scotia in terms of a seasonal and an interannual component. This study compares the seasonal variation in RNA/DNA ratios to the seasonal varia- tions in adductor muscle and gonad weights. The results provide an important piece of information as to the usefulness of this tech- nique as an index of nutritional and other stresses in wild popu- lations. MATERIALS AND METHODS The sampling of scallops for RNA/DNA analysis was attached to a program that had been initiated to examine temporal and spatial variations in adductor muscle weights. The study took ad- vantage of a commercial scallop vessel fishing in the Digby area, whose owner had volunteered to conduct sampling (Vance Hazel- ton. Hazelton Fisheries Ltd., Digby N.S.). This had the advantage of no cost for chartering a vessel (depths and currents in the area excluded the use of small boats and scuba diving for the collec- tions). It had the disadvantage of an irregular sampling schedule, because the vessel was not always fishing in the area (Table 1 ). In June of each year, samples were taken with the DFO research vessel, the J. L. Hart, during the annual scallop stock assessment survey on the Digby bed. The RNA/DNA data from the 1990 to 1992 June surveys was that used by Kenchington ( 1994) to look at interannual and spatial variations, with the addition of the 1993 survey data. A preliminary analysis of the seasonal adductor muscle and gonad weight data up to January of 1993 was pub- lished by Kenchington et al. (1994). The area sampled was in the commercial scallop bed directly off Digby Gut. This area was divided into two zones, an inside zone (2-6 miles from shore), and an outside zone (6-10 miles from shore). Four random locations were sampled in each zone. At each 405 406 Roddick et al. TABLE 1. Sampling dates for RNA/DNA collections. Month Year Jan. Feb. March April May June July Aug. Sept. Oct. Nov. Dec. 1990 1991 1992 1993 X X X X z X X X X o X X Z = RNA/DNA samples only: X. both RNA/DNA and meat weight samples: O. meat weight samples only. location, a tow of approximately 1 nautical mile was conducted using commercial scallop gear. The average depth (corrected for tidal height) was recorded at all stations, and the bottom tempera- ture was recorded with a digital thermometer for the June tows for 1990 to 1993. A sample of up to 72 scallops were randomly collected from each tow for meat weights (except for August 1993. when only 30 were taken). The scallops were shucked at sea. the top shell was retained and the soft parts refrigerated and brought back for weighing onshore. Meat weights for all scallops and gonad weights for the first 30 of each sample were recorded to the nearest 0.01 g. For the RNA/DNA sampling, six scallops in each of three size groups were taken from each tow location. The three size groups, 80-95 mm. 96-1 10 mm, and 1 1 1-125 mm, were selected to cor- respond roughly to ages 3, 4, and 5. These sizes are all commer- cially fished in this area. The scallops were dissected live, and the adductor muscle and top shell were retained. The muscle samples were held and transported in liquid nitrogen and then stored in a -85 °C freezer until analyzed to prevent the degradation of nucleic acids. Shell height was recorded, and the sex of each animal was determined visually when possible. Sex could not be visually de- termined for some animals in the postspawning period. RNA/DNA Analysis The tissue sample consisted of a 1-mm transverse section through the center of the phasic portion of the adductor muscle. (The large phasic portion of the adductor will be referred to simply as the adductor muscle and the smaller, tonic portion as the catch muscle). The nucleic acid concentrations were measured fluoro- metrically following the technique of Karsten and Wollenberger ( 1972, 1977) as modified by Kenchington ( 1994). For processing, the sample of adductor muscle was thawed and homogenized with a Brinkman Polytron homogeniser in 7 mL of ice-cold heparin solution (3.75 u.g mL"' ) and centrifuged at 2000 rpm for 5 minutes at 5 °C. This produced three layers, a bottom phase of cellular debris and a top foam layer, neither of which contained appreciable amounts of nucleic acids, and a middle clear phase in which the nucleic acids were concentrated. This middle phase was split into 5-200 u-L samples for analysis. Replicate total nucleic acid samples had 400 pL heparin added, replicate DNA samples had 200 (jlL heparin and 200 u.L RNAase. The fifth sample was a tissue blank and received 200 |j,L heparin and 400 pL distilled water. All five tubes were then incubated in a water bath at 36-38 °C for 30 minutes. This incubation time allowed the RNAase reaction to go to completion without any degradation of DNA. After the incuba- tion time, 200 pL of ethidium bromide (EtBr) was added to the total nucleic acid and DNA samples. EtBr is a fluorophor that reacts with nucleic acids to produce an increase in fluorescence intensity. It has an excitation wavelength of 365 nm and an emis- sion wavelength of 590 nm. Fluorescence was measured with a Turner 1 12nm fiuorometer using an excitation filter for 320-390 nm and an emission filter of >590 nm. Calf thymus DNA and yeast RNA were used for the calibration of standard curves. Fluores- cence readings were converted to nucleic acid concentrations using the calibration curves. The DNA concentration was calculated di- rectly from the RNAase-treated samples, the RNA concentration was calculated as the difference between the concentrations of the total nucleic acid and DNA samples. RNA/DNA was calculated as a mass ratio. A reagent blank, prepared fresh each run, was made with 200 u-L EtBr, 400 u.L distilled water, and 200 pL heparin. Analytical grade reagents were used for all procedures. DATA ANALYSIS RNA/DNA A plot of the RNA/DNA data indicated that the mean ratio and the associated variance of scallop adductor muscle RNA/DNA ratios were related (Fig. la; r = 0.60. P < .001). Natural log transformation of the data did not resolve this (Fig. lb; r = -0.37, P = .023). and there were still an abundance of outliers (Fig. 1). These factors would compromise the use of a normal probability model for the analysis, and so a generalized linear model (GLM) was used. Generalized linear models differ from such classic linear models as linear regression and analysis of variance (ANOVA) in that the assumptions of normality and constant variance are not required for the error component. Any distribution in the expo- nential family (i.e.. Poisson, binomial, gamma) can be used. These models retain the idea of a predictor based on a linear combination of explanatory variables. In particular, the ideas underlying such factorial models as additivity. interaction, polynomial contrasts, and aliasing are retained in the GLM. Without the assumption of normality for the error component, they estimate parameters by maximum likelihood using an iteratively reweighted least-squares approach. The GLM uses separate functions to model any nonlin- earity between the mean and the predictor (the link function), and the relationship between the mean and variance (the variance func- tion). They are closer to a reparamaterization of the model than to a re-expression of the response (McCullagh and Nelder 1989. Hastie and Pregibon 1992). The analysis was carried out using an analysis of deviance with a gamma distribution and an identity link. The gamma distribution is very useful for positive random variables, because it can assume a wide range of shapes from symmetric to highly skewed (McCul- Temporal Variation in Sea Scallop RNA/DNA 407 0 15 - A - Untransformed Data r2 = 033 n = 30 • 01 - •• • • • 0.05 - %^-~~~~~ • • • • • 0- •* • • 1 — 1 1 0.6 07 OS Sample Mean -0.6 -0.4 Sample Mean Figure 1. Sample variance versus mean for A raw and B Ln trans- formed RNA/DNA ratios of scallop adductor muscles. lagh and Nelder 1989). A constant linear relationship between the mean and variance, as observed here, is a feature of this distribu- tion. The analysis was performed with the S-PLUS statistical pack- age (Statistical Sciences Inc. 1995). Because the same months were not sampled each year, effects were first tested using sampling date (15 samples), zone (2-6 miles, and 6-10 miles from shore), size (3 classes), and tow (four tows, nested within each combination of sampling date and zone). The full model with all effects and interactions was tested first, and then submodels sequentially deleting nonsignificant terms were tested until all remaining terms were significant (Lawless et al. 1978). At each step, differences between nested models were tested for significance using a Chi-squared test. The factor tow was included in the analysis to examine the significance of differences between tows, but would not be in- cluded in a model to examine the general health of a stock. If differences between tows on a single sampling date were larger than seasonal or stress effects, the method would not be as useful for monitoring the health of the stock. As a practical tool, tow effects would be left out of the model, thus incorporating tow-to- tow differences in the error term. The irregular sampling schedule resulted in 6 of the 9 months sampled in only 1 year, and June the only month sampled in every year. This means that modeling seasonal variation by breaking sampling date down into year and month effects would produce a model in which month and year are too highly aliased to produce true month and \ear coefficients. To examine the seasonal and interannual components of the variation in RNA/DNA ratios, the model with sampling date as a factor (15 sampling dates over the 3.25-year period) was used to predict RNA/DNA ratios for a shell height of 100 mm. and a seasonal model (described below) was then fit to the entire sampling period (see seasonal model below). Interannual effects were examined using the May and June samples from each year in a model with the sampling date factor replaced with year and month effects. The seasonal model was fit to the gonad, adductor, and catch muscle weights to compare the fit of the seasonal model to the weight data versus the fit to the RNA/DNA data. Adductor Muscle, Catch Muscle, and Gonad Weights The adductor, gonad, and catch muscle weights were also ana- lyzed with a GLM. Because distinct size classes had not been used for these samples, the log of the shell height was used as a co- variate in the analysis with a log link and gamma error functions. The relationship with the covariate was first tested by fitting a common slope and intercept to all data, a common slope with separate intercepts, separate intercepts with a common slope, and separate intercepts and slopes. The best relationship (most signifi- cant, minimum deviance) was used for the rest of the model. Factors examined were sampling date (14 samples), zone (2-6 miles and 6-10 miles from shore), and tow (four tows, nested within each combination of sampling date and zone). The number of scallops in the adductor and catch muscle datasets varied between samples, because it was not always pos- sible to obtain 72 scallops at each site. To investigate the effects of a highly unbalanced design on the analysis, a random subset of the data taking 30 scallops per tow and three tows were drawn. This subset resulted in a fully balanced design, and this was analyzed and compared with the results from the full dataset. The subset of the data for which both temperature and depth were available was used to compare the effect of these to the factor zone, and the subset of the data for which sex was available was used to examine the effect of this factor. Seasonal Model Adductor and gonad weights have been shown to follow an annual cycle in Placopecten magellanicus (Robinson et al. 1981, Couturier and Newkirk 1991, Kenchington et al. 1994). A sine- wave function was chosen as the model to describe this cycle. It has a well-defined pattern, a small number of parameters to fit to the data, and has been used to incorporate seasonal fluctuations into von Bertalanffy growth functions (Pitcher and MacDonald 1973, Pauly and Gaschultz 1979, Antoine et al. 1979, Hanumara and Hoenig 1987, Allison 1994). A sine function is symmetric and will not model the abrupt decline upon spawning of an individual scallop gonad, but as a model for the population, which has a more protracted spawning event and interannual variation in spawning time, the sine curve should be an adequate model of the annual cycle. The seasonal patterns in the weight and ratio data were, therefore, modeled by fitting the sine function {A + B[Sine(Juli- an_Date/365*2iT + C)]|, where A, B. and C are constants for the vertical position of the function, the amplitude of the sine wave, and the horizontal offset. Because significant differences between years were found in the RNA/DNA ratio data, the data were standardized by dividing each year's data by the mean for the June sample in that year. Recall that June was the only month sampled in every year. The 408 Roddick et al. sine wave model was then fit to the standardized data, and the significance of the different components of the model were esti- mated with partial F tests for parameters added in order (Klein- baum et al. 1988). The model was fit to the data with the nonlinear regression (NLR) procedure in the SPSS statistical package (SPSS Inc. Chi- cago, IL). This procedure fits user-defined models iteratively using the Levenberg-Marquardt algorithm. If there is a relationship between RNA/DNA ratios and growth rates for scallops, then the residuals from the seasonal RNA/DNA ratio model should be correlated with the residuals from the sea- sonal adductor muscle growth model in the subsequent time pe- riod. That is to say, RNA/DNA ratios higher than the predicted seasonal values should correspond to growth rates that are also higher than predicted levels, and. conversely, low RNA/DNA lev- els should occur with low growth rates. Residuals for each time period in the seasonal growth model were calculated as the differ- ence in the residuals from the fit of the sine curve to the GLM sample means in successive samples. The residuals for the RNA/DNA ratios from the fit to the sine curve (Mean GLM Ratio,, - Sine Predicted Ratio,,) were com- pared to the residuals from the adductor muscle weight fit to the sine curve for the subsequent time period [(Mean GLM Weight, , - Sine Predicted Weight,,) - (Mean GLM Weight,, - Sine Pre- dicted Weight, ,)], where mean GLM ratios and weights are the standardized means from the GLM analysis, sine predicted values are those predicted by the sine functions, and the subscripts tl and t2 indicate successive sampling dates. These data were analyzed to determine if there was any correlation between the residuals from seasonal RNA/DNA ratios and residuals from seasonal growth in the subsequent time period. RESULTS RNA/DNA Data The GLM model with all effects and interactions explained a significant portion of the variation in the RNA/DNA ratios. Zone, tow nested within sampling date by zone, and the interaction terms were all nonsignificant. The best model includes only the main factors: sampling date and size (Table 2). The predicted RNA/ DNA ratios for the three size classes (Table 3) show a decrease with size. Using the May and June samples to look at interannual variations showed that year was significant (P = .0004), month and the month by year interaction were not. Comparisons of years showed that mean RNA/DNA ratios for 1991 were significantly higher (P < .01) than those for 1990, 1992 and 1993, which were not significantly different. TABLE 3. RNA/DNA ratios predicted from final model for sampling dates and size classes. Sampling Date 80-95 mm 96-110 mm 111-125 mm June 22, 1990 0.6522 0.5735 0.5 1 59 Oct. 9, 1990 1 .0406 0.9618 0.9043 April 9, 1991 0.9084 0.8297 0.7722 May 6, 1991 0.7665 0.6878 0.6302 June 19. 1991 1.0024 0.9237 0.8661 Nov. 7. 1991 0.6991 0.6204 0.5629 March 31. 1992 0.5314 0.4527 0.3952 May 4. 1992 0.6656 0.5869 0.5294 June 6. 1992 0.6394 0.5607 0.503 1 Aua. 11. 1992 0.8988 0.8201 0.7625 Sept. 9. 1992 0.7437 0.6649 0.6074 Nov. 10. 1992 0.8302 0.7515 0.6939 Jan. 7. 1993 0.6690 0.5903 0.5328 May 10, 1993 0.5878 0.5091 0.4516 June 6. 1993 0.7324 0.6537 0.5962 Adductor Muscle, Catch Muscle, and Gonad Weight Data The analysis of deviance for the three tissue components are shown in Tables 4-6. For all three tissues, the effect of fitting the log of shell height as a covariate was highly significant, and sepa- rate fits for each sample were significantly better than a common equation fit to all samples. A common slope with separate inter- cepts fit best for the meat weights. There was no significant dif- ference between models for the gonad weights, and the minimum deviance model of a common intercept with different slopes was chosen. The model with different intercepts and slopes was sig- nificantly better for the catch muscle weights (P = .04) than a model with a common intercept and different slopes, although not significantly better than a model with a common slope and differ- ent intercepts (P = .08). The model with different intercepts and slopes was chosen, on the basis of minimum deviance, for the rest of the analysis. Sampling date and zone were significant factors in the weight of all three tissues; whereas, tow, nested within sam- pling date and zone, and the interaction of sampling date and zone were not. Interannual comparisons using the May and June samples showed significant differences in the catch muscle (P = .039), but not for the adductor or gonad samples. There was no difference in the results using the full dataset for the weight data or the subset used for the fully balanced design. Shell height, sampling date, and zone were significant (P < .05). and tow nested within sampling date and zone, and the interaction of sampling date and zone were not (P > .8). The final GLM TABLE 2. Analysis of deviance table for the final model for RNA/DNA ratio data using sampling date and size factors. Response: Ratio; Distribution: Gamma: Link: Identity. TABLE 4. Analysis of deviance table for the final model for meat weight data. Distribution: Gamma; Link: Log. Residual Df Deviance Residual P (Chi) Terms Df Deviance ADf ADeviance P(Chi) ADf ADeviance Terms NULL Ln( Height) Date Zone 7366 1 13 1 1492.520 1241.508 81.303 36.336 7365 7352 7351 251.012 169.709 133.372 NULL Date Size 1437 14 2 213.2474 64.5941 9.6559 1423 1421 148.6533 138.9974 <.0001 .0080 <.0001 <.0001 ■c.OOOl Temporal Variation in Sea Scallop RNA/DNA 409 TABLE 5. Analysis of deviance table for the final model for gonad weight data. Distribution: Gamma; Link: Log. TABLE 6. Analysis of deviance table for the final model for catch muscle weight data. Distribution: Gamma: Link: Log. Df Deviance Residual P (Chi) Terms Df Deviance Residual Terms ADf ADeviance xm ADeviance /•(Chi) NULL Zone Date in Ln( Height) 3331 1 14 1693.302 208.998 1229.323 3330 3316 14S 1.304 254.981 <.0001 <.0001 NULL Date Zone Date in Ln( Height) 7366 13 1 14 1399.552 73.482 280.931 7 1 1 .058 7353 7352 7338 1326.070 1045.140 334.082 <.0001 <.0001 <.0001 models with the full dataset were used to predict means for a 100 mm shell height scallop from each zone (Table 7). To examine what made the factor zone significant in all the tissue weights examined, temperature and depth were compared to zone, using the subset of data for which they were both available (June samples. 1990 to 1993). When individually entered into the model after shell height and sample date, the factor zone was the most significant (P = .004). followed by depth [P = .007) and temperature {P = .0160). The factors are, however, all correlated to the extent that after any one of the three is entered, the others will be nonsignificant, and there is no significant difference be- tween models using any one of the three factors. Taking the scal- lops for which sex could be determined (5,947 of 7,367 scallops), there were no significant differences between males and females for meat weight (P = .4548), gonad weight (P = .3071 ), or catch muscle weight (P = .5225). Seasonal Model The predicted means from the final models of the RNA/DNA ratios (Table 3) and the tissue weights (Table 7) were modeled by fitting a sine function to give a seasonal cycle. Figure 2 shows the fit to the small size class (80-95 mm) for the ratio data and the predicted means for the tissue weights for a 1 00 mm shell height scallop from the inside zone (Table 7). The r values for the model fit to the data for the inside zone and small size class were: RNA/ DNA ratio 0.22: meat weight 0.56: gonad weight 0.69: and catch muscle weight 0.26. The cycles in the meat weight and gonad weight are nearly in opposite phase; when one is dropping the other is rising. The predicted values for the adductor muscle weight go from a low in June-August to a high in November- January (Fig. 2). The percentage increase from the low to the high for 1991 to 1993, calculated from the standardized means for the inside zone in Table 7. are 31. 37. and 15%. respectively. There is some indication of interannual differences in the timing of the seasonal cycle. In June of 1991. the adductor muscle weight was still high on the 19th of the month, dropping only 5% from the May value of 14.79. The drops from May to June in 1992 and 1993 were 11 and 14%. Because the annual ranges are similar, this indicates an interannual variation in the timing of the cycle. The curve for the RNA/DNA ratio is slightly lagging the gonad cycle, with a peak during the period when the adductor muscle weight is increasing. The fit of the sine function to the RNA/DNA data (Fig. 2) shows that, although there may be a seasonal component, there is a high level of variation, with the RNA/DNA ratios from Oc- tober 1990 to June 1991 well above the predicted seasonal values. This variation is not a true interannual component, as shown when the data are standardized to the June values within each year (Fig. 2). The fit for the April to June 1991 ratios improves, moving these datapoints closer to the model prediction, but still leaves the Oct. 1990 and November 1991 values as larger outliers. The partial F test on the ratio data standardized to the June samples each year (Table 8) shows that fitting parameter A, es- sentially a straight line fit to the mean RNA/DNA ratio value, is highly significant (P < .0001). Adding B*Sine(Rdate), which in- cludes the sine wave in the function (Rdate is the Julian date TABLE 7. Predicted means for adductor muscle, gonad, and catch muscle weights for a scallop with a shell height of 100 mm. Means for Inside Zone Means for Outside Zone Date Adductor Gonad Catch Adductor Gonad Catch April 9. 19991 14.791 4.041 1.122 12.694 3.379 0.997 May 6. 1991 14.791 6.152 1.187 12.694 5.143 1.055 June 19. 1991 13.980 9.127 1.084 1 1 .998 7.631 0.963 Nov. 7. 1991 18.263 2.799 1.237 15.674 2.340 1.100 March 31. 1992 15.359 3.039 1.167 13.182 2.541 1.037 Mav 4. 1992 15.359 4.907 1.196 13.181 4.102 1.063 June 6. 1992 13.652 6.956 1.062 11.716 5.816 0.944 Aug. 11. 1992 11.916 9.175 0.907 10.226 7.671 0.806 Sept. 9, 1992 15.556 2.685 1.277 13.350 2.245 1.135 Nov. 10. 1992 16.290 1.965 1.212 13.980 1 .643 1.077 Jan. 7, 1993 15.958 1.892 1.153 13.696 1.582 1.025 May 10. 1993 15.388 5.358 1.230 13.206 4.480 1.093 June 6, 1993 13.180 7.115 0.707 11.312 5.949 0.628 Aug. 24. 1993 11.784 4.316 0.935 10.113 3.608 0.831 410 Roddick et al. 45 40 35- 30 3 ■I 25 i 20 15 10 5. Standardized RNA/DNA Ratio RNA/DNA Ratio Adductor Muscle Weight Gonad Weight Catch Muscle -- . 111111111111111111111111111 June Sep Jan May Sep Jan May Sep 1990 1991 1992 I I I I I I I! Jan May 1993 0.5 1 o 0.75 £ < z 0.5 § 0.25 Sep Figure 2. Annual cycle in RNA/DNA ratios, adductor muscle, gonad, and catch muscle weights, as modeled with a sine function for the full time period sampled. RNA/DNA ratios are for adductor muscle tissue from a 80-90 mm shell height scallop. Weights are standardized to a 100 mm shell height scallop from the inside zone. Standardized RNA/ DNA ratios are standardized to the June ratio within each year. converted to 2tt radians per year), results in a significant (P = .01 ) improvement in the model. This means that the addition of the sine wave has explained significantly more of the variance in the model than the fit of a straight line (model with just parameter A). In the case of the ratio data, parameter C does not significantly improve the model. This is because the sine wave is not shifted horizontally in the case of the ratio data, so parameter C is not significantly different than 0 (actually parameter C will converge to a value, depending on the initial starting estimate, of 0. 2tt, 4-tt. etc. I. Standardizing the RNA/DNA ratios to eliminate interannual vari- ability increases the r value from 0.22 to 0.45 for the fit of the seasonal model. The residuals from the predicted seasonal cycle in RNA/DNA ratios (calculated as observed values with GLM size standardiza- tion - values predicted by sine function) are correlated with re- siduals in subsequent growth for the adductor muscle (correlation = 0.484, P = .047). This relationship can be seen in Figure 2, where the high ratios in April to June 1991 are followed by a November muscle weight that is well above the seasonal predic- TABLE 8. Partial F test for parameters added in order for the sine wave model parameters fit to the RNA/DNA ratio data standardized to the June samples. Source DF SS MS A + B*Sine(Rdate) A + B*Sine (Rdate + C) Residual Total 15 15.89001 .34165 .04249 .47091 16.74506 15.89001 .34165 .04249 0.3924 404.94419 8.70668 1.08282 <.0001 .0115 .3186 tion. Low ratios in the spring of 1992 are followed by a low adductor weight in August. DISCUSSION This study confirms that the RNA/DNA ratio of the adductor varies with size; the smaller, faster growing scallops having higher ratios than the older ones (Kenchington 1994). This agrees with the link between RNA/DNA ratio and growth rate shown in other studies (Haines 1973. Wright and Hetzel 1985). It is an expected result, because growth rate, and hence, protein synthesis, declines with age in this species (Kenchington et al. 1995). The significant difference between the RNA/DNA ratios for the inside and outside sampling zones is not in agreement with a high RNA/DNA ratio reflecting a higher growth rate. The inside zone has a higher growth rate than the outside (Robert et al. 1986). The present study shows that the inside zone also has significantly higher meat and gonad weights for comparable shell heights than the outside zone. This greater growth is being achieved at a lower RNA/DNA ratio than in the outside zone. The explanation for this is the influence of temperature on the RNA/DNA ratio. Tempera- ture has been noted to affect the relationship between RNA/DNA ratio and growth in winter flounder by Buckley (1982). He ob- served that at higher temperatures, an increase in growth rate oc- curred without an increase in the RNA/DNA ratio, indicating that increased temperatures may result in a higher activity rate for the RNA rather than an increase in the amount present. In the June scallop surveys for Digby. bottom temperature was recorded for each tow. There was a significant (P < .001) difference between the bottom temperatures of the inside and outside zone, with the inside zone having a consistently higher temperature (Table 9). The higher growth rate of the inside zone is being achieved with a lower RNA/DNA ratio, but at a higher temperature. When looking at the interannual variation in the RNA/DNA ratios, 1991 stands out as being higher than the other years. The temperature range for the June samples was broken down into 0.1 °C cells and the mean ratio for each cell plotted against the temperature (Fig. 3). The 1990. 1992. and 1993 samples all show similar ratios over a 3 °C temperature range. The 1991 sample stands out as having a higher RNA/DNA ratio at a similar tem- perature to 1990. The increased amount of RNA present, and an increased activity of the RNA because of the higher temperatures, should have resulted in an increased growth rate in 1991 as com- pared to 1992 or 1993. This is supported by the June values for the adductor, gonad, and catch muscle weights, which are all higher in 1991 (Table 7). Because the temperatures are similar in 1990 and 1991, why are the RNA/DNA ratios so different? On the Digby side of the Bay of Fundy, readings for chlorophyll fluorescence were taken at 1 meter TABLE 9. Means, standard deviations and number of tows for bottom temperatures taken during the June Digby area scallop surveys in each year. Year Inside Zone Outside Zone 1990 1991 1992 1993 7.93(0.336)53 7.59(0.245)65 5.08 (0.384) 43 6.39(0.078)39 7.32(0.376)51 7.08(0.110)72 4.67(0.126)54 6.14(0.191)53 Temporal Variation in Sea Scallop RNA/DNA 411 Temperature (°C) Figure 3. Average RNA/DNA ratio for 0.1 C cells by temperature. Numbers are sample size for that cell, markers are by year sampled. from the surface. 1 meter from the bottom, and midwater off the Digby wharf for a sampling program in that area (Fig. 4, Paul Keizer, unpublished data). The readings for April to July, had mean values of chlorophyll a (mg m"3) of 2.99 (±1 .21 ) for 1990, and 4.15 (±2.01) for 1991. The higher RNA/DNA ratios in 1991 may be in response to a higher food supply. With the irregular sampling schedule, an ANOVA using year and month factors is too highly aliased to produce realistic annual and seasonal components. Treating each sampling date as a sepa- rate level and then fitting the sine function to the data would indicate if there is a seasonal component to the variation. It would be possible to improve the fit by adding additional parameters; that is, to improve the fit to the gonad and adductor weight data during the spawning period where Figure 2 shows it underestimating the peak gonad weight and overestimating the minimum meat weight. This would make for a less parsimonious model but may improve its predictive capability. As applied here, to examine the question of a significant seasonal component in the different datasets. the most parsimonious model has proved capable of showing a sea- sonal component. A linear fit to the RNA/DNA ratio data explains 16% of the variation in the data, but with a nonsignificant slope (t = -1.613. P = .1306). showing that there is no over-all trend with time running through the data. The partial F test shows that the fit of the sine function is significant, indicating that there is season- able variation in the RNA/DNA ratio in the scallop population off Figure 4. Chlorophyll a levels for water samples taken off Digby wharf, April to July in 1990 and 1991 (Paul Keizer, unpublished data I. Digby, Nova Scotia. The seasonal component, as represented by the sine function, explains a maximum of 45% of the variation in the RNA/DNA ratio, and Figure 2 does not indicate the presence of any consistent pattern in the residuals. The general shape of the sine function models the seasonal component, but there are other factors that have a large influence on the variations in the RNA/ DNA ratio, and the data have large outliers, even when standard- ized within each year. The variation left unexplained by the model, and the large outliers, indicate that the RNA/DNA ratio has varia- tions on a time scale less than seasonal, and that some residuals are as large as the range of seasonal variation predicted by the model. Seasonal variations in the RNA/DNA ratio of the adductor muscle in oysters was noted by Pease (1976) with the highest values occurring in September to October for the populations stud- ied. Paon and Kenchington ( 1 995 ) found that the RNA/DNA in the adductor muscles of Placopecten magellanicus peaked just before spawning during laboratory conditioning. This agrees with the annual cycle found in this study, with the ratio cycle lagging that of the gonad weight and peaking as the adductor muscle increases after the spawning period in the Digby area. The adductor muscle is an important site for energy storage in pectinids (Ansell 1974, Taylor and Venn 1979. Idler et al. 1964). Energy, mainly in the form of glycogen and protein, is stored to support gamete maturation and to meet metabolic requirements during winter, when food availability is low. The amount of sea- sonal variation in adductor muscle weights reported in the litera- ture, and seen in this study, is large. During the annual cycle in adductor weight. Taylor and Venn ( 1979) report a doubling of the adductor dry weight for Chlamys opercularis; whereas, increases of 77% have been reported for Chlamys islandica (Sundet and Vahl 1981), and 39% for Pecten maximus (Comely 1974). Rob- inson et al. (1981). working with Placopecten magellanicus in Boothbay Harbor. Maine, show an increase of approximately 60%. Comely (1974) showed less than a 5% variation in water content of the adductor muscle for Pecten maximus, so the 30-40% varia- tion in adductor wet weight found in this study is not at the high end of the range in pectinids. This seasonal variation in weight has greater implications for management of this species, because most pectinid fisheries outside North America use both the adductor muscle and the gonad; whereas, the majority of the Placopecten magellanicus fisheries, including Digby one, land only the adduc- tor muscle. With a quota-based management system, the seasonal pattern of landings should be taken into account when determining the quota. The timing of the cycle of adductor muscle weight predicted by the sine function (Fig. 2). is unusual, in that it predicts a peak in January. Although there is often a rapid postspawning increase in the adductor weight, it usually occurs during the summer or fall. The adductor weight of Chlamys opercularis peaks in September- October following a June-July spawning (Taylor and Venn 1979). C. islandica in August-September following June-July spawning (Sundet and Vahl 1981 ). and Pecten maximus in November after protracted spawning starting in the early summer (Comely 1974). In studies of Placopecten magellanicus, which spawn August to September in the Digby area. Robinson et al. (1981) reported a spring peak in adductor weight in Maine, with a winter low. Thompson (1977) found a low in total somatic tissue weight from September to March for southeast Newfoundland. The January peak in adductor dry weight predicted from the sine function does not agree with the sample means (Table 7). The peak seems to occur earlier than the sine function predicts, at least in November 412 Roddick et al. from the sample means, and perhaps earlier, because there are no data for October. A January peak would be difficult to achieve with the known pattern of plankton in the Bay of Fundy, which has a fall diatom bloom but low plankton concentrations during the late fall and early winter (Martin et al. 1995). This timing differs from the spring peak reported for the Gulf of Maine and New- foundland areas (Robinson et al. 1981. Thompson 1977). The catch muscle exhibits little seasonal variation in weight. This is expected, because it is not used as an energy storage site, as the large phasic portion of the adductor muscle is. Robinson et al. (1981) found the concentrations of glycogen in the catch and phasic portions of the adductor to be the same in March and April for Placopecten magellanicus, and DeZwaan et al. (1980) found that in July, the glycogen concentration of the quick adductor muscle was over twice that of the catch muscle. The predicted gonad weights produced the best fit to the sine curve, with a peak in July, before the August-September spawning period for the Digby area. The peak in the adductor RNA/DNA ratio occurs as the adduc- tor muscle weight is increasing. Part of this increase in weight is attributable to increased protein synthesis, which requires a higher RNA/DNA ratio. Pectinids have been shown to use the protein content of the adductor as an energy storage mechanism along with its glycogen content and the lipids in the digestive gland. Thomp- son (1977) showed a seasonal variation in protein for total somatic tissue of Placopecten magellanicus, and Taylor and Venn (1979) showed that the seasonal variation in protein weight is larger than that of glycogen for the adductor muscle of Chlamys opercularis. Sundet and Vahl (1981) showed that in Chlamys islandica, both mature and immature scallops catabolize adductor muscle protein during the winter, but that it is a much more important energy source for the juveniles. They postulate that the greater use of protein in the juveniles is a growth strategy, allowing for more growth in times of abundant food. On reaching sexual maturity, the production of gametes requires more energy, and the use of gly- cogen increases, but seasonal variations in protein content con- tinue. The variations seen in the RNA/DNA ratio of the adductor muscle are not tightly bound to the gametogenic cycle. The sea- sonal/sine wave model explains 69% of the variation in gonad weight, but a maximum of 45% of the variation in the RNA/DNA ratio of the adductor muscle. Part of the difference in the fit is that the interannual and subannual variations in the RNA/DNA ratio seem to be larger than those for the gonad or adductor weight. This may be because that the RNA/DNA ratio of the adductor muscle is affected not only by the variable growth rate of the adductor tissue, but also by its use as an energy substrate in support of gameto- genesis, and by protein synthesis related to metabolic activities other than tissue protein growth. The relationship between the RNA/DNA ratio and growth rate is the basis for postulating that RNA/DNA ratios may be a useful index of health for scallop stocks. The fact that residuals from seasonal RNA/DNA ratios are correlated with residual from sub- sequent adductor growth support the existence of this relationship. This correlation is low under natural conditions as seen in the Digby scallop population, however, and the RNA/DNA ratios show a high level of variation on a short time scale. CONCLUSIONS The RNA/DNA ratio of the adductor muscle for the scallop population off Digby Nova Scotia varies with the size or age of the scallop, and there are temporal variations on interannual and sea- sonal scales. Temporal variations are related to growth rate differ- ences and reflect variations in food supply and temperature. The seasonal variation in RNA/DNA ratios can be modeled and ap- proximates a sine function. This model explains a maximum of 45% of the variation in RNA/DNA ratio, but the residual from the model are significantly correlated with subsequent growth. The usefulness of RNA/DNA ratios for monitoring the health of a scallop stock will depend on whether or not the large variation seen in this healthy stock during the sampling period is a true reflection of variances in growth rate or would remain high as the health and average RNA/DNA ratio declined. A large variation in RNA/DNA ratios as levels declined would make it more difficult to discriminate healthy stocks that were under normal levels of nutritional stress during periods of low food availability from those that were in danger of increased mortality rates. The method seems to have the potential to indicate when a population is approaching a critical level of stress, but if a monitoring program were estab- lished, it could not be recommended as the only method used to indicate the health of the stock. LITERATURE CITED Allison. E. H. 1994. Seasonal growth models for Great Scallops (Pecten maximus (L.)) and Queen Scallops [Aequipecten opercularis (L.)]. J. Shellfish Res. 13:555-564. Ansell, A. D. 1974. Seasonal changes in biochemical composition of the bivalve Chlamys septemradiata from the Clyde Sea area. Mar. Biol. 25:85-99. Antoine, L.. P. Arzel, A. Laurec & E. Morize. 1979. La croissance de la coquille Saint-Jacques [Pecten maximus (L.)] dans les divers gisements Francais. Rapp. P-v. Reun. Cons. Int. Explor. Mer. 175:8-19. Buckley. L. J. 1982. Effects of temperature on growth and biochemical composition of larval winter flounder (Pseudopleuronectes america- nus). Mar. Ecol. Prog. Ser. 8:181-186. Buckley. L. J. 1984. RNA-DNA ratio: an index of larval fish growth in the sea. Mar. Biol. 80:291-298. Canino. M. F. 1997. Nucleic acid contents and growth of first-feeding walleye pollock larvae in response to prey densities typical of sub- Arctic ecosystems. /. Fish Biol. 51:41-52. Comely. C. A. 1974. Seasonal variation in the flesh weights and biochemi- cal content of the scallop Pecten maximus L. in the Clyde Sea area. Cons. Int. Explor. Mer. 35:281-295. Couturier. C. Y. & G. F. Newkirk. 1991. Biochemical and gametogenic cycles in scallops, Placopecten magellanicus (Gmelin, 1791). held in suspended culture, pp. 107-1 17. In: S. E. Shumway and P. A. Sandifer (eds.). An International Compendium of Scallop Biology and Culture. World Aquaculture Society. Baton Rouge, LA. DeZwaan. A.. R. J. Thompson & D. R. Livingston. 1980. Physiological and biochemical aspects of the valve snap and closure responses in the giant scallop Placopecten magellanicus. J. Comp. Physiol. 137(Ser. B): 105- 1 14. Dortch. Q. F.. T. L. Roberts, J. R. Clayton. Jr. & S. 1. Ahmed. 1983. RNA/ DNA ratios and DNA concentrations as indicators of growth rate and biomass in planktonic marine organisms. Mar. Ecol. Prog. Ser. 13:67- 71. Falkowski, P. G. & T. G. Owens. 1982. A technique for estimating phy- Temporal Variation in Sea Scallop RNA/DNA 413 toplankton division rates by using a DNA-binding fluorescent dye. LimnoL Oceangr. 27:776-782. Haines, T. A. 1973. An evaluation of RNA-DNA ratio as a measure of long-term growth in fish populations. J. Fish. Res. Board Can. 30: 195- 199. Hanumara. R. C. & N. A. Hoenig. 1987. An empirical comparison of fit of linear and nonlinear models for seasonal growth in fish. Fish Res. 5:359-381. Hastie, T. J. & D. Pregibon. 1992. Generalized linear models, pp. 195-247. In: J. M. Chambers and T. J. Hastie (eds.). Statistical models. S. Wads- worth & Brooks/Cole, Advanced Books & Software, Pacific Grove, CA. Hovenkamp, F. & J. I. Witte. 1991. Growth, otolith growth, and RNA/ DNA ratios of larval plaice Pleuronectes platessa in the North Sea 1987 to 1989. Mar. Ecol. Prog Ser. 70:105-1 16. Idler, D. R.. T. Tamura & T. Wainai. 1964. Seasonal variations in the sterol, fat, and unsaponifiable components of scallop muscle. J. Fish. Res. Board Can. 21:1035-1042. Karsten. U. & A. Wollenberger. 1972. Determination of DNA and RNA in homogenized cells and tissues by surface fluorometry. Anal. Biochem. 46:135-148. Karsten, U. & A. Wollenberger. 1977. Improvements in the ethidium bro- mide method for direct fluorometric estimation of DNA and RNA in cell and tissue homogenates. Anal. Biochem. 77:465^470. Kenchmgton, E. L. R. 1994. Spatial and temporal variation in adductor muscle RNA/DNA ratio in sea scallops (Placopecten magellanicus) in the bay of Fundy. J. Shellfish Res. 13:19-24. Kenchington, E. L. R.. M. Lundy & V. Hazelton. 1994. Seasonal changes in somatic and reproductive tissue weights in wild populations of Pla- copecten magellanicus in the Bay of Fundy, Canada, pp. 154-162. In: N. F. Bourne, B. L. Bunting and L. D. Townsend (eds.). Proceedings of the 9th International Pectinid Workshop, Nanaimo, BC, Canada. April 22-27, 1993. vol. 2. Can. Tech. Repl. Fish. Aquat. Sci. 1994:217 p. Kenchington, E., D. L. Roddick & M. J. Lundy. 1995. Bay of Fundy scal- lop analytical stock assessment and data review 1981-1994: Digby grounds. DFO Atlantic Fisheries Res. Doc. 95/10, 70 pp. Kleinbaum. D. G., L. L. Kupper & K. E. Muller. 1988. Applied regression analysis and other multivariable methods, 2nd ed. Kent. Boston, MA. 718 pp. Lawless, J. F. & K. Singhal. 1978. Efficient screening of non-normal re- gression models. Biometrics 34:318-328. Leick, V. 1968. Ratios between contents of DNA, RNA, and protein in different micro-organisms as a function of maximal growth rate. Nature 217:1153-1155. Martin, J. L„ D. J. Wildish, M. M. LeGresley & M. M. Ringuette. 1995. Phytoplankton monitoring in the southwestern Bay of Fundy during 1990-92. Can. Manusc. Rept. Fish. Aquat. Sci. 2277:iii + 154 pp. McCullagh, P. & J. A. Nelder (eds.). 1989. Generalized linear models. 2nd ed. Monographs on statistics and applied probability 37. Chapman and Hall. New York. 5 1 1 pp. Munro, H. N. & J. A. M. Gray. 1969. The nucleic acid content of skeletal muscle and liver in mammals of different body size. Comp. Biochem. Physiol. 28:879-905. Paon, L. A. & E. L. R. Kenchington. 1995. Changes in somatic and repro- ductive tissues during artificial conditioning of the sea scallop. Pla- copecten magellanicus (Gmellin. 1741). J. Shellfish Res. 14:53-58. Pauly, D. & G. Gaschultz. 1979. A simple method for fitting oscillating length growth data, with a program for pocket calculators. International Council for Exploration of the Sea (ICES) CM 1979/G24, 26 pp. Pease, A. K. 1976. Studies of the relationships of RNA:DNA ratios and the rate of protein synthesis to growth in the oyster Crassostrea virginica. Fisheries Marine Services (Canada) Tech. Rept. 622:1-77. Pntcher. T. J. & P. D. M. MacDonald. 1973. Two models for seasonal growth in fishes. J. Appl. Ecol. 10:599-606. Robert, G., M. A. E. Butler-Connolly & M. J. Lundy. 1986. Bay of Fundy stocks assessment. 1985. Canadian Atlantic Fisheries Scientific Advi- sory Committee Research Document 86/40, 25 pp. Robinson, W.E., W. E. Wehling, M. P. Morse & G. C. McLeod. 1981. Seasonal changes in soft-body component indices and energy reserves in the Atlantic deep sea scallop, Placopecten magellanicus. Fish. Bull. 79:449-458. Statistical Sciences Inc., 1995. S-PLUS user's manual, version 3.3 for Windows. Statistical Sciences Inc. Seattle, Washington, U.S.A. Sundet, J. H. & O. Vahl. 1981. Seasonal changes in dry weight and bio- chemical composition of the tissues of sexually mature and immature Iceland scallops. Chlamys islandica. J. Mar. Biol. Assoc. U.K. 61: 1001-1010. Sutcliffe. W. H.. Jr. 1969. Relationship between growth rate and ribo- nucleic acid concentration in some invertebrates. J. Fish. Res. Bd. Can. 27:606-609. Taylor. A. C. & T. J. Venn. 1979. Seasonal variation in weight and bio- chemical composition of the tissues of the queen scallop, Chlamys opercularis, from the Clyde sea area. J. Mar. Biol. Assoc. U.K. 59: 605-621. Thompson. R. J. 1977. Blood chemistry, biochemical composition, and the annual reproductive cycle in the giant scallop. Placopecten magellani- cus. from southeast Newfoundland. J. Fish. Res. Bd. Can. 34:2104- 2166. Wright, D. A. & E. W. Hetzel. 1985. Use of RNA:DNA ratios as an indi- cator of nutritional stress in the American oyster Crassostrea virginica. Mar. Ecol. Prog. Ser. 25:199-206. Journal of Shellfish Research, Vol. 18. No. 2. 415-418, 1999. TEMPORAL COINCIDENCE OF THE ANNUAL EELGRASS ZOSTERA MARINA AND JUVENILE SCALLOPS ARGOPECTEN VENTR1COSUS (SOWERBY II, 1842) IN BAHIA CONCEPCION, MEXICO NOE A. SANTAMARIA,' ESTEBAN F. FELIX-PICO,2 JOSE LUIS SANCHEZ-LIZASO,1 J. RICARDO PALOMARES-GARCIA,2 AND MANUEL MAZON-SUASTEGUI2 lUnidad de Biologia Marina. Universidad de Alicante. Ap. 99 E-03080, Alicante. Spain 2Centro Interdisciplinario de Ciencias Marinas-IPN. Apdo. Postal 592. C.P. 23000. La Paz. B.C.S.. Mexico. ABSTRACT The possibility that meadows of annual eelgrass Zostera marina L. in Bahi'a Concepcion serve as a nursery habitat for Pacific calico scallop Argopecten ventricosus has been evaluated. Scallop reproduction occurs mainly in the coldest season of the year, just when annual eelgrass is present. In this study, it was observed that the highest abundance of pectinid larvae was in early March, and scallop spat on artificial collectors was higher in late March. The highest above-ground biomass of eelgrass occurred between February and April and declined in May, just when scallops detach themselves from substrata and settle to the bottom. Our results indicate a high coincidence in timing of the analyzed processes and suggest the possibility that eelgrass beds could be used as nursery grounds by Pacific calico scallop. KEY WORDS: annual eelgrass, scallop spat, nursery habitat, Zostera marina, Argopecten ventricosus INTRODUCTION The Pacific calico scallop Argopecten ventricosus ( = circu- laris) (Sowerby II. 1842) supports an important fishery in the state of Baja California Sur. Mexico. Between 1986 and 1991, scallop harvest in Bahi'a Concepcion, the most productive bay for this fishery, had a very high production (up to 5531 t fresh weight in 1989), but there have been large fluctuations in landings since 1991, moving toward critical levels in 1993 ( 100 t) and 1994 when the fishery was closed. That depletion has prompted increased interest in the factors that affect scallop recruitment and survival (Felix-Pico et al. 1997). During their early life history. Pacific calico scallops, as well as other pectinids. settle and attach to substrates that elevate them above the bottom (Felix-Pico et al. 1989). These substrates are usually submerged vegetation or even artificial materials, but eel- grass Zostera marina L. appears to be the main natural substratum for bay scallop, Argopecten irradians (L.) (Thayer and Stuart 1974). Eelgrass beds not only provide juvenile bay scallops with a settlement substrate but also help scallops avoid benthic predators, which results in higher scallop recruitment and survival (Ekman 1987, Peterson et al. 1989, Pohle et al. 1991, Garcfa-Esquivel and Bricelj 1993). Juvenile bay scallops undergo an ontogenic shift in habitat. Scallops often attach themselves in the eelgrass canopy in their earlier life stage until they reach a size of 20-30 mm. after which juvenile scallops settle to the bottom. At this size they have achieved a refuge from some of their major crustaceans predators (Garcfa-Esquivel and Bricelj 1993). Little is known about the association of Pacific calico scallops and eelgrass beds in spite of the fact there are extensive eelgrass meadows in almost all the lagoons of Baja California Sur where scallops occur (Felix-Pico et al. 1989). Z. marina is a widely distributed, temperate seagrass species that occurs in some subtropical regions, such as the Gulf of Cali- fornia (Phillips and Menez 1988). Although this species commonly forms perennial beds, in the Gulf of California eelgrass forms only annual beds, which appear in late autumn as seeds, and disappear in mid-spring when all the plants die (Phillips and Backman 1983, Santamaria 1996). As Pacific calico scallops reproduce mainly in the coolest part of the year. January to March (Villalejo-Fuerte and Ochoa-Baez 1993, Felix-Pico et al. 1997), it is possible to suppose that annual eelgrass beds serve as nursery habitat for juvenile Pacific calico scallops. For this, it is first necessary that spat settlement coincide with the maximum eelgrass standing crop and that scallops reach a sufficient size to be a free living form before eelgrass beds disappear. Direct evaluation of the settlement of juvenile scallops on eelgrass in Bahfa Concepcion was not possible because scallops were very scarce throughout the bay during the year of study. Thus, the objective of the present study was to determine the timing of Z. marina bed presence, pectinid larval abundance, and juvenile scallop settlement on artificial collectors and to determine if eelgrass meadows in Bahi'a Concepcion appear in the time when they could potentially provide scallops with a mechanism to in- crease survival. MATERIALS AND METHODS Bahia Concepcion is located in the Gulf of California on the east coast of the Baja California Peninsula at approximately lat 26°45'N, long 1 1 1°45'W (Fig. 1). Usually, the surface water tem- perature in the Bay varies from 18 °C to 32 °C. but a surface temperature of 16 °C was observed in February 1989 (Villalejo- Fuerte and Ochoa-Baez 1993). To evaluate the Z. marina bed as potential habitat for scallops, data on eelgrass abundance were taken at Punta Arena, Bahia Concepcion, during one season of eelgrass development, from De- cember 1994 to July 1995. Samples were taken every 3 weeks via scuba diving. Eelgrass shoot density was measured in 25 x 25 cm quadrats, with 4 to 10 random replicates. Shoot abundance was 415 416 Santamaria et al. El Coloradito BAJA CALIFORNIA SUR 26"30' 112°00' 26°30' 11)'4S RESULTS The Z. marina bed at Punta Arena appeared in early December as seedlings, then all eelgrass plants suddenly died in late May and the meadow disappeared. In July, there were no eelgrass plants, only seeds in the sediment. Shoot density in the bed was about 1 .400 shoots per m2 from December to April when shoot density started to decrease. Highest values of above-ground biomass oc- curred between February and March 1995 (Fig. 2a). particularly from 15 to 30 March (218 and 235 g DW/nr, respectively). There was a decrease in biomass by early March, but this coincided with the highest mean plant height. 71 cm (Fig. 2a). and we suppose that it could be caused by variation in plant morphology. However, shoot density and height of eelgrass bed showed consistently high values from February to May, suggesting that canopy habitat was high during this period. There were some pectinid larvae in the water column at El Coloradito, on 16 February and 2 March, 95 and 40 larvae/m3. respectively, but the maximum concentrations were observed on 9 and 16 March: 400 and 290 larvae/m3, respectively (Fig. 2b). Thereafter, larvae were almost absent. Spat caught on collectors were most abundant in collectors installed from 16 to 23 March, averaging 35.5 and 20 individuals per collector, respectively (Fig. Figure. 1. Study area: Bahia Cnncepcion. Sampling sites, Punta Arena and EI Coloradito, are indicated by arrows. extrapolated per m2. Percentage cover was examined in 20-m transects, with 4 replicates. Thirty shoots were sampled to deter- mine above-ground biomass and the total height of plants. Plant biomass evaluation was measured by drying individual shoots at 60 °C to constant weight. To evaluate the potential habitat pro- vided by the eelgrass bed. height of the canopy was estimated as the mean height of plants. Simultaneously, to evaluate abundance of D-veliger (2 to 9 days old), umbonated ( 10 to 13 days old), and pediveliger ( 14 to 18 days old) scallop larvae (Aviles-Quevedo 1990. Monsalvo- Spencer 1998), we sampled zooplankton at Punta El Coloradito weekly, using a water-pump and two sieves of 63 and 132 p.m mesh. For zooplankton analysis, subsamples were taken by a Folsom Divider. All D-veliger. umbonated. and pediveliger larvae found in both sieves were counted, and the data were standardized per m\ To test the timing of spat settlement in Bahia Concepcion. artificial collectors were used due to the low levels of scallop abundance here during this study. Collectors were built with onion sacks filled with 300 g of polypropylene nets. They were hung from a longline, held by buoys and anchors, at 1 - and 3-m depth. Ten collectors were installed weekly. 5 at each depth, to monitor spat settlement during the time when scallop spawning was antic- ipated to occur, from 19 January to 30 March. Each set of collec- tors was removed from water 2 months after it was installed. Collectors were dried before juvenile scallops were separated through a 5-mm sieve and Pacific calico scallops counted. Mean spat size and number of scallops settled per collector were re- corded. Previous data showed that no significant differences were found in collector efficiency between Punta Arena and El Colora- dito in 1991 and 1992 (Felix-Pico et al. 1997). Surface seawater temperature was measured at each sampling date. 400 100 30 27 - c 0 24 21 18 ' 8 11 25 9 16 23 2 9 16 23 30 10 19 14 20 De Ja Ja Fe Fe Fe Mr Mr Mr Mi Mr Ap Ap My Jl dates Figure. 2. (a) Mean above-ground biomass of eelgrass bed (left scale) and plant height (right scale); error bars = standard deviation (n = 30), (b) scallop larvae density (left scale) and new spat abundance on col- lectors per week (right scale), and (c) seawater temperature during sampling period. All data shown are from December 1994 to July 1995. Coincidence of Eelgrass and Juvenile Scallops 417 2b). Spat in collectors were gradually detached when they reached n 20 to 30 mm. from April to May. Sizes of scallops observed in collectors were between 7 and 34 mm. Seawater temperature was between 18.2 °C in January and 22.5 °C in April, and reached 28 °C in July (Fig. 2c). DISCUSSION Our data show a high coincidence in timing of scallop spawn- ing, spat abundance, and the permanence of eelgrass beds. Spat abundance peaked in collectors installed one week after we ob- served peak in larval abundance on 9 March, therefore, it can be inferred that individuals in collectors came from the same cohort as the sampled larvae, because there were no more larvae in the water after that. Efficiency of collectors decreases rapidly after they have been installed, probably due to epibiosis. On the other hand, scallops take about 2 months to reach 20 to 30 mm, when they are able to become a free-living form, but scallop spat are able to detach themselves after reaching about 15 mm (Garci'a-Esquivel and Bricelj 1993). During the time when the scallop spat in Bahia Concepcion need to be fixed, the eelgrass bed was present. The height of the eelgrass canopy remained at high levels until plant death, but above-ground biomass began to decline in late March. Previously, it has been observed that most juvenile scallops detach themselves from the substrate to begin a free-living form before all the plants die in May (Felix-Pico et al. 1997). Similarly, populations of Ar- gopecten irradians on the east coast of the United States reproduce mainly in the season of major development of eelgrass meadows, although those beds are not annuals but perennials, and it occurs in summer, from July to September (Churchill and Riner 1978, Ro- man and Able 1988, Garcia-Esquivel and Bricelj 1993). The absence of juvenile scallops attached to eelgrass shoots in Bahia Concepcion during the sampling period was considered to be due to low adult scallop abundance, and spat abundance ob- served in collectors was also low. Normal levels of spat collections are consistent annual yields of 6,000 spat per collector. In 1994, spat collections were 41 to 852 spat per bag (Felix-Pico et al. 1997). but 2 years later, in spring 1997, a great quantity (about 500 spats/nr) of scallop spat were found together with plants sampled in April. Unfortunately it was not possible to quantify the density because distribution inside the bed is very contagious and system- atic sampling was not done. Other types of submerged macrophytes in Bahia Concepcion must not be ignored, such as Sargassum spp. stands, which are the most abundant macroalgae in the Bay and may also serve as a substratum to scallop spat. However, the main development of Sargassum is in June to July (Casas-Valdez et al. 1993) when scallop spat presence is not significant. In 1989, an unusually high eelgrass abundance coincided with the lowest temperature recorded in the Bay (Castro-Ortiz pers. comm.), and a boom in catches of scallops was also observed that year ( Felix-Pico et al. 1997). As annual eelgrass in the Gulf of California develops in the coolest season of the year because it is a temperate species living in a subtropical locality (Phillips and Backman 1983. Phillips and Menez 1988, Santamana 1996), and the main scallop spawning period in Bahi'a Concepcion also coin- cides with the coolest months of the year (Villalejo-Fuerte and Ochoa-Baez 1993). temperature probably plays a significant role in Pacific calico scallop recruitment and survival. However, other environmental factors that could also affect the abundance of eel- grass beds in their year-to-year variation must be considered in future studies. Findings obtained in this study are limited and more research is necessary to confirm and evaluate the relation of Pacific calico scallop spat to eelgrass beds, both in the laboratory and field. However, with these results it is possible to propose some sugges- tions for a better exploitation of Pacific calico scallop. If our hypothesis is confirmed, the protection of eelgrass beds would be necessary to preserve high abundance of scallop populations. Be- sides, with absolute certainty, eelgrass beds accomplish several other ecological tasks in the Bay (Phillips and Meiiez 1988). Avoidance of trawl fishing over the meadows is an urgent matter that should be addressed to protect the meadows. This activity, as we could see during field work in Bahi'a Concepcion, causes severe damage to the eelgrass coverage. Seagrass meadows in other lo- calities have been already protected successfully from trawl fish- ing (Guillen-Nieto et al. 1994) with benefits for artisanal fisheries (Martinez 1997). AKNOWLEDGMENTS This research was supported by the "Biologi'a y cultivo de la almeja catarina, y algunos invertebrados de importancia eco- nomica" project, under number 923460 (DEPI) for CICIMAR IPN. The first author was supported by the scholarships CONACyT and COFFA-IPN. Second and third authors had scholarships from COFFA-IPN and Desempeno Academico. Special thanks to Ar- mando Naranjo Mariscal and MASAVI S.A. for its collaboration on collector installation and all CICIMAR personnel who worked on this project. LITERATURE CITED Aviles-Quevedo, M. A. 1990. Crecimiento de la almeja catarina (Ar- gopecten circularis) en funcion del alimento, con anotaciones sobre su biologi'a y desarrollo. Tesis de Maestrfa. CICIMAR-IPN. La Paz, B.C.S. Mexico. 64 pp. Casas-Valdez, M. M., I. Sanchez-Rodriguez & G. Hemandez-Carmona. 1993. Evaluaci6n del Sargassum spp en la costa oeste de Bahi'a Con- cepcion, B.C.S.. Mexico. Inv. Mar. CICIMAR 8(2):61-69. Churchill. A. C. & M. L. Riner. 1978. Anthesis and seed production in Zostera marina L. from Great South Bay. New York, U.S.A. Aquat. Bot. 4:83-93. Ekman, J. E. 1987. The role of hydrodynamics in recruitment, growth and survival of Argopecten irradians (L.) and Anomia simplex (D Orhigny) within eelgrass meadows. J. Exp. Mar. Biol. Ecol. 106:165-191. Felix-Pico, E.. A. Tripp-Quezada, J. Castro-Ortiz. G. Serrano-Casillas. P. Gonzalez-Ramirez, M. Villalejo-Fuerte. R. Palomares-Garcfa, F. Gar- ci'a-Domfnguez. M. Mazon-Suastegui, G. Bojorquez-VenSstiga & G. Ldpez-Garci'a. 1997. Repopulation and culture of the pacific calico scallops in Bahi'a Concepcion. B.C.S.. Mexico. Aquacult. Int. 5:551— 563. Felix-Pico, E., A. Tripp-Quezada & J. Singh-Cabanillas. 1989. Anteced- entes en el cultivo de Argopecten circularis (Sowerby) en Baja Cali- fornia Sur, Mexico. Inv. Mar. CICIMAR 4:75-92. Garci'a-Esquivel, Z. & M. Bricelj. 1993. Ontogenic changes in microhabital distribution of juvenile bay scallops. Argopecten irradians irradians (L.). in eelgrass beds, and their potential significance to early recruit- ment. Biol. Bull. 185:42-55. Guillen-Nieto. J. E.. A. A. Ramos-Espla, L. Martinez & J. L. Sanchez- Lizaso. 1994. Antitrawling reefs and the protection of Posidonia oce- anica (L.) Delile meadows in the western Mediterranean sea: demand and aims. Bull. Mar. Sci. 55(2-3):645-650. Santamaria et al. Martinez, J. M. 1997. La pesca artesanal de El Campello (Alicante, S.E. Iberico): caracterizacion y elementos para una ordenacion. PhD Thesis, University of Alicante. 248 pp. Monsalvo-Spencer. P. 1998. Estudios sobre el cultivo de larvas y juveniles de almeja catarina Argopecten ventricosus ( = circularis) (Sowerby II. 1842) en laboratorio. Tesis de Maestria. CICIMAR-1PN. La Paz, B.C.S. Mexico. 70 pp. Peterson. C. H.. H. C. Summerson, S. R. Feghley & R. C. Prescott. 1989. Timing, intensity and sources of autumn mortality of adult bay scallops Argopecten irradians concentricus Say. /. Exp. Mar. Biol. Eeol. 127: 121-140. Phillips, R. C. & T. W. Backman. 1983. Phenology and reproductive biology of eelgrass {Zostera marina L.) at Bahia Kino, Sea of Cortez, Mexico. Aquat. Bot. 17:85-90. Phillips, R. C. & E.G. Menez. 1988. Seagrasses. Smithsonian Contribu- tions to the Marine Sciences. No. 34. Smithsonian Institution Press, Washington. 240 pp. Pohle, D. C. V. M. Bricelj & Z. Garcia-Esquivel. 1991. The eelgrass canopy: an above-bottom refuge from benthic predators for ju- venile bay scallops Argopecten irradians. Mar. Ecol. Prog. Ser. 74:47- 59. Roman, C. T. & K. W. Able. 1988. Production ecology of eelgrass {Zostera marina L.) in a Cape Cod salt marsh-estuarine system, Massachusetts. Aquat. Bot. 32:353-363. Santamaria. N.A. 1996. Ciclo de crecimiento y fenologi'a de la fanerogama Zostera marina L. en Punta Arena, Bahia Conception, B.C.S.. Mexico. Tesis de Maestria. CICIMAR-IPN. La Paz. B.C.S. Mexico. 97 pp. Thayer, G. & H. H. Stuart. 1974. The bay scallop makes its bed of seagrass. Mar. Fish. Rev. 36:27-30. Villalejo-Fuerte, M. & R. I. Ochoa-Baez. 1993. The reproductive cycle of the scallop Argopecten circularis (Sowerby, 1835), in relation to tem- perature and photoperiod, in Bahia Concepcion. B.C.S.. Mexico. Cien- cias Marinas 1 9( 2 ) : 1 8 1 -202. Journal of Shellfish Research, Vol. IX, No. 2, 419-423, 1999. OXYGEN CONSUMPTION AND AMMONIA EXCRETION OF LARVAE AND JUVENILES OF THE BAY SCALLOP, ARGOPECTEN IRRADIANS CONCENTRICVS (SAY) YANTIAN T. LU, NORMAN J. BLAKE, AND JOSEPH J. TORRES Department of Marine Science University of South Florida St. Petersburg, Florida 33701 ABSTRACT Rates of oxygen consumption and ammonia excretion were determined at 25 °C for larval and juvenile bay scallops Argopecten irradians concentricus. Oxygen consumption rate (Vo2, p.10, fr' ) varied with body size ash-free dry weight ( AFDW. mg) according to the relation Vo, = 7.199 AFDW"921 for larvae and Vo, = 2.142 AFDW"905 for juveniles. Mean weight-specific Vo, ranged from 14.66 to 15.84 p.10, mg AFDW ' h"' for larvae and from 1.80 to 5.28 p.10, mg AFDW ' h"' for juveniles. Weight- specific Vo, declined with increasing body size at > 2 mm shell height but became independent of body size at > 2 mm shell height. Swimming was estimated to cause 8 to 29% of the total larval oxygen demand. In juveniles of 3.5 to 5.0 mm shell height, Vo, increased as temperature increased from 15 to 30 °C, being 1.24 ± 0.35. 1.77 ± 0.70. 2.33 ± 0.85.. and 2.74 ± 0.57 p.lO, mg AFDW"1 h"' at 15, 20, 25 and 30 °C, respectively. The Q,„ was 2.99 at 15 to 20 °C, 1.74 at 20 to 25 °C and 1.37 at 25 to 30 °C. Ammonia excretion rate (U, u.gNH4-N IT1) increased with body size following the equation U = 0.161 AFDW""28. Energy loss through ammonia excretion was equal to 1.5 to 3.7% and 13.5 + 2.8% of the respiratory energy loss in larvae and juveniles, respectively. KEY WORDS: Bay scallop larvae, oxygen consumption, ammonia excretion INTRODUCTION Metabolism is an important component of physiological ener- getics, the study of energy gains and losses at the organismal level (Brett and Groves 1979). Metabolic rate is most often estimated as the rate of oxygen consumption, and it represents a major loss of energy in bivalves. Energy loss through ammonia excretion is also very significant (Barber and Blake 1985). Information on oxygen consumption exists for a variety of adult marine bivalves, but less is known about the respiratory physiology of early developmental stages, with the exception of a few commercially important species (Sprung 1984, MacDonald 1988. Beiras and Camacho 1994). Existing data suggest that, in both larval and adult bivalves, oxygen consumption increases with increasing body size according to the allometric equation Y = aX . where Y is oxygen consumption. X is body weight, and a and b are fitted parameters (Bayne et al. 1976a). In adult bivalves, b-values range from 0.65 to 0.84, with a mean of 0.7 (Bayne and Newell 1983); whereas, in larval bivalves, b-values are often close to 1 (Riisgard et al. 1981, Gerdes 1983, Beiras and Camacho 1994), demonstrating an isometric relationship between metabolic rate and body size in larvae. In energy balance studies, energy loss attributable to excretion of nitrogenous products is often overlooked (Bayne and Newell 1983). although it may substantially affect the general estimation of the energy budget. In most marine bivalves, ammonia is the dominant end product of protein catabolism, comprising 41 to 94% of the total nitrogen excretion (reviewed by Bayne et al. 1976b). Very limited data are available on ammonia excretion in pectinids; the only available data were collected by Barber and Blake ( 1985) on adult Argopecten irradians concentricus. The purpose of this study was to determine the energy loss of larvae and juveniles of the bay scallop A. irradians concentricus (Say) by measuring their rates of oxygen consumption and ammo- nia excretion and to determine the relationships between these physiological rates and body size. MATERIALS AND METHODS Bay scallops collected from Homosassa. Florida were spawned at the Department of Marine Science. University of South Florida. Culture of larvae and juveniles followed the methods described by Lu and Blake (1996). Before each determination of oxygen con- sumption, larvae and juveniles were filtered onto 35 u,m nylon screens, rinsed with, and then released to, 0.45 u.m filtered sea- water (salinity 26 ± Woo). Larger juveniles were also cleaned with a small, soft pen brush. Oxygen consumption rates of various size classes of larvae and juveniles were measured at 25 ± 0.5 CC. Larvae and small juveniles were placed in sealed respiration chambers filled with 0.45 u.m filtered seawater at a density of 20 to 80 larvae ml"1 or 1 to 10 juveniles ml-1, depending on shell size. Respiration chambers were made from plastic syringes (Torres et al. 1994). whose vol- ume could be adjusted (1-5 ml) according to the number and size of the experimental animals by adjusting the syringe plunger and/ or oxygen electrode. Experiments with juveniles larger than 5 mm in shell height were carried out in 25- to 65- ml Lucite chambers. Oxygen concentration in the syringes and chambers was measured every 2 minutes using Microcathode oxygen electrodes calibrated with air and nitrogen saturated seawater. Data were recorded using a computer-controlled digital data-logging system. Experiments lasted 4 to 8 hours, during which the oxygen concentration never dropped below 50%. The effect of temperature on oxygen consumption was deter- mined using juveniles of 3.5 to 5 mm shell height. Experimental procedures were the same as described above, except that experi- ments were run at four temperatures, 15, 20, 25. and 30 ± 0.2 °C, using circulating water bath to control temperature. The changing oxygen concentrations recorded for each respi- ration chamber during the course of a run were regressed against time to obtain a representative slope, a rate. The rate obtained was divided by the number of scallops in the chamber to give the amount of oxygen consumed per individual per hour (pJO, h"1). The oxygen consumption rates were further converted to u.10, mg AFDW"' h"1 using the weight data (total AFDW) of Lu and Blake (1996). All measured oxygen consumption rates were corrected using control runs without animals. Ammonia excretion was determined by placing larvae or juve- niles in 35- and 50-ml capped glass vials filled with 0.45 u,m filtered seawater at 20 to 80 larvae ml"' or 1 to 60 juveniles per vial. Filtered seawater without animals were used as blanks. Ex- 419 420 LU ET AL. periments lasted 4 to 7 hours at 25 ± 0.2 °C. At the end of each experiment, a 10-ml sample was drawn from each vial and placed in a 20-ml test tube. In experiments with larvae, the samples were passed through a 35-u.m mesh screen to remove the scallops. Am- monia concentrations in the samples and blanks were determined using the indophenol blue method (adapted from Solorzano 1969). To each sample and blank, 0.4 ml of phenol-alcohol reagent, 0.4 ml of nitroprusside reagent, and 1 .0 ml of oxidizing reagent were added. The blue color that developed in the dark for 1 hour was read on a Cary spectrophotometer at wavelength of 640 nm. Op- tical densities were converted to ammonia concentrations using a standard curve determined with a solution of ammonium sulfate. The following factors were used for converting oxygen con- sumption and ammonia excretion into energy unit: 1 ml O, = 19.9 J (Elliott and Davidson 1975) 1 mg NH4 - N = 24.8 J (Elliott and Davidson 1975) RESULTS Table 1 lists the mean oxygen consumption rates of various sizes of larvae and juveniles of the bay scallop. Mean oxygen consumption ranged from 1.35 to 4.10 x 10~3 u,102 IT1 for larvae of 120 to 180 u.m shell length, and 1.15 x 10_2 to 5.15 u.l02 h"1 for juveniles of 0.5 to 7.0 mm shell height. The oxygen consump- tion of larvae and juveniles (up to 10 mm height) is closely related to body size according to the following equations: Vo2 (u.!02 h" Vo2 (u,102 IV1 ) = 0.0444H- 410 r = 0.718 (for larvae) = 0.956 (for juveniles) where L is shell length of larvae in mm, and H is shell height of juveniles in mm. The two equations above can be further trans- formed to the following using the weight data determined by Lu and Blake (1996): Vo2 (ui02 IT1) = 7.199 AFDW1"521 (for larvae) Vo2(u.l02h~') : 2.142 AFDW0905 (for juveniles) where AFDW is the total ash free dry weight (including AFDW of the shells) in mg. The measured oxygen consumption rates and the fitted curves are shown in Figure 1 for both larvae and juveniles. Figure 1 was plotted in double logarithmic scales so that oxygen consumption of larvae and small juveniles can be compared on the 1 10 Shell length (larvae) or height (juveniles) (mm) 100 Figure. 1. Argopecten irradians concentricus. Oxygen consumption rate of larvae and juveniles versus shell size. same graph. It is clear from Figure 1 that oxygen consumption of larvae is above the extended line fitted to the datapoints of juve- niles, suggesting that larvae have higher relative metabolic rates than juveniles. Weight-specific oxygen consumption decreased with increas- ing body size at < 2 mm shell height (Fig. 2). Mean values ranged from 14.66 to 15.84 u.10, mg AFDW"1 h"1 for larvae, and from 5.28 to 2.28 u-102 mg AFDW"1 h"1 for juveniles 0.5 to 2 mm shell height. The relationship between weight-specific oxygen con- sumption rate and body size of juveniles < 2 mm shell height can be best described by the following equation: Vo2 (u.l02 mg AFDW ' h" ' ) = 3.443H" : 0.792 For juveniles > 2 mm in shell height, weight-specific oxygen con- sumption rates remained relatively constant, with a mean value of 2.16 ± 0.20 u-lO, mg AFDW"1 h_I. In the temperature experiments, oxygen consumption increased with increasing temperature between 15 to 30 °C (Fig. 3). At 15 °C. the mean weight-specific oxygen consumption was 1 .24 ± 0.35 U.10-, mg AFDW-1 h_1, representing 37.5% of the rate of oxygen consumption at 30 ° (2.74 ± 0.57 u-102 mg AFDW"1 h"1). At 20 and 25 °C. mean rates were 1 .77 ± 0.70 and 2.33 ± 0.85 u-102 mg AFDW ' h"\ being 64.8 and 85.3% of the rate at 30 °C, respec- tively. The Q10 was 2.99 at 15 to 20 °C, 1.74 at 20 to 25 °C and 1.37 at 25 to30°C. TABLE 1. Argopecten irradians concentricus. Rates (mean ± SD) of oxygen consumption and ammonium excretion of larvae and juveniles, n: number of estimates. Height (mm) AFDW (mg) O, Consumption Rate Ammonia Excretion Rate Ml ind"' h pi mg AFDW"1 h" pg ind ' h pg mg AFDW1 h" 0.12 0.0000915 5 O.O0135 + O.O0052 0.15 0.0001692 2 0.00268 ± 0.00074 0.18 0.0002796 6 0.00410 ±0.00077 0.50 0.00217 4 0.01 15 ±0.0034 1.00 0.0138 4 0.0518 ±0.0087 1.50 0.0406 2 0.0972 ± 0.0270 2.00 0.0875 6 0.1 994 ±0.0733 3.00 0.2576 2 0.61 10 ±0.0324 4.00 0.5543 3 1.2528 ±0.2060 5.00 1 .0045 3 1.8100 ±0.6429 7.00 2.4616 2 5.1473 ± 1.4575 14.784: 15.839: 14.664: 5.278 : 3.754 : 2.394 : 2.279 : 2.372 : 2.260 : 1.802: 2.088 : 6.334 4.034 2.801 1.099 0.630 0.665 0.838 0.126 0.372 0.640 0.592 (3.51 ±0.80)x 10" 6 (4.23 ± 1.04) x 10"' 5 (6.35 ± 1.94) x 10^ 4 0.00304 ± 0.00063 5 0.00778 ±0.00197 4 0.01540 ±0.00763 3 0.07604 ± 0.00505 4 0.12755 ±0.03924 5 0.18289 ±0.02421 1 0.42635 ± 0.00000 0.384 ± 0.098 0. 1 25 ± 0.035 0.292 ± 0.089 0.220 ± 0.045 0.192 ±0.048 0.1 76 ±0.087 0.295 ±0.019 0.230 ±0.071 0.1 82 ±0.024 0.1 73 ±0.000 Oxygen Consumption /Ammonia Excretion in Larval and Juvenile Bay Scallops 421 1 10 Shell length (larvae) or height (juveniles) (mm) 100 Figure. 2. Argopecten irradians concentricus. Weight-specific oxygen consumption rate of larvae and juveniles versus shell size. Mean ammonia excretion rates are also summarized in Table 1. In larvae, ammonia excretion was similar for young and old larvae; whereas, in juveniles, it was an increasing function of body size. The relationship between ammonia excretion rate (U) and body size of juveniles can be described by the following equations: U (u.gNH4 -Nh"') = 0.00302H2 472 (r = 0.875) U ( |J.gNH4 -N h"1) = 0.161 AFDW0 92S The measured ammonia excretion rates and the fitted U to H curve are shown in Figure 4. Mean weight-specific ammonia excretion decreased from 0.384 |xgN mg AFDW"1 h"' in 120 |xm larvae to 0.125 u.gN mg AFDW"1 IT1 in 180 u.m larvae. Energy loss through ammonia excretion was equivalent to 1.5 to 3.7% of the larval respiratory energy loss. In juveniles, mean weight-specific ammonia excretion was relatively constant, ranging from 0.173 to 0.295 u.gN mg AFDW-' IT1 with a mean of0.220± 0.046 jigNrng AFDW"1 h"1. Energy loss through juvenile ammonia excretion equaled 13.5 ± 2.8% of the respiratory energy loss. DISCUSSION Allometric exponents determined for the relationship between oxygen consumption and body size of larval and juvenile bay a % 2 E 6 o > 2.99 a-ttC V02 Hqio V^ 1.37 15 20 25 Temperature (°C) 30 10 1 0 1 001 0 001 1E-4 1E-5 Larvae Figure. 3. Argopecten irradians concentricus. Weight-specific oxygen consumption rate and Q]0 values of juveniles versus temperature. 01 1 10 Shell length (larvae) or heighl (juveniles) (mm) Figure. 4. Argopecten irradians concentricus. Ammonia excretion rate of larvae and juveniles versus shell size. scallops are very similar, being 0.921 and 0.905, respectively. These values are higher than those determined for adult bivalves, which range from 0.65 to 0.84, with a mean of 0.72 (Bayne and Newell 1983). A similar trend was found for the Japanese scallop Patinopecten yessoensis, in which a b- value of 0.81 was found for adults (Fuji and Hashizume 1974); whereas, a b-value of 1.39 was found for larvae (MacDonald 1988). High values of b were also reported for early stages of other bivalves; for example, 0.90 for Mytilus edulis larvae (Riisgard et al. 1981) and 1.09 for Ostrea edulis larvae (Beiras and Camacho 1994). Weight-specific oxygen consumption of bay scallop larvae de- termined in the present study ranged from 3.0to to.6 u-102 mg DW"1 IT1 (6.2 to 20.5 u.lO,mg AFDW"' h"1). These values fell at the lower end of the range 4.6 to 15.2 u.10, mg DW"1 h"1 deter- mined for larval Argopecten irradians (Siddall 1987) but were similar to the values of 1.6 to 10.0 jjlIO, mg DW"1 h"1 determined for larvae of other bivalves (reviewed by Holland 1978; Sprung 1984). Similar oxygen consumption was found for larvae of the Japanese scallop P. yessoensis, with a range of 5.2 to 11.6 u.10-, mg DW"1 h"1 (calculated from MacDonald 1988). Despite the similar values of b for larvae and juveniles of the bay scallop, the weight-specific oxygen consumption of larvae is 3 to 9 times higher than that determined for juveniles. High meta- bolic rates observed for larvae are likely a reflection of the energy expended in swimming during this planktonic stage. Swimming activity of bivalve larvae generally represents 8 to 50% of respi- ration loss (Zeuthen 1947). Sprung ( 1984) assumed that the energy expenditure of swimming larvae of Mytilus edulis was twice the amount needed to overcome the force of sinking (assuming the horizontal component equals the vertical component) and calcu- lated that the energy used in locomotion was less than 2% of the respiration loss. He pointed out that his values were low. because metabolic effort is transferred to the action of motion with certain efficiencies (Klyashtorin and Yarzhombek 1973), and the cost of swimming could be much higher. Projected oxygen consumption rates of bay scallop larvae were calculated by extrapolation using the allometric equation for small juvenile bay scallops of 0.5 to 2 mm shell height. If the calculated rates are assumed to be equivalent to the metabolic rates of non- swimming larvae, by comparing them with the metabolic rates measured for swimming larvae, we can estimate that the energy expended in swimming is 8.0% and 27.8% of the total metabolism 42: LU ET AL. of larvae 120 p.m and 180 |xm in shell length, respectively. Larger larvae spend a greater proportion of energy on swimming than smaller ones, a trend that was also found in Mytilus edulis larvae (Sprung 1984). The estimation is in accordance with the observa- tion that in Crassostrea virginica and C. gigas, oxygen consump- tion of larvae dropped approximately 40% when exposed to epi- nephrine, a metamorphosis inducer, probably because of the ces- sation of swimming activity (Haws et al. 1993). Although weight-specific oxygen consumption decreases with increasing body size in small juveniles (< 2 mm shell height), it is independent of body size in larger juveniles (2-16 mm shell height), with a mean of 2.16 (jl102 mg AFDW"1 h-1. This value is higher than the mean rate of 1.11 p-102 mg AFDW-1 IT1 deter- mined for adult Argopecten irradians concentricus (calculated from Barber and Blake 1985, assuming 85% of tissue dry weight is AFDW), consistent with the general trend that larger animals have lower weight-specific metabolic rates. Juvenile bay scallops are well adapted to temperatures between 20-30 °C. as indicated by the low Q1() values over this temperature range (1.73, 20-25 °C: 1.37, 25 -30 °C). This is in contrast with the case at low temperatures, where Q10 is much higher (2.99 at 15-20 °C). In adult bay scallops, a Ql0 of 1.38 was obtained from Vo2 values published by Barber and Blake (1985) over a tempera- ture range of 21-31 °C (calculated by Bricelj et al. 1987). This value is very close to the Ql0 found for juveniles over 20-30 °C in this study. Thus, respiration of juvenile bay scallops responds to temperature in a manner similar to adults. As compared to oxygen consumption, ammonia excretion has been a neglected area in the study of the physiological energetics of marine bivalves. In the present study, energy loss through am- monia excretion represented about 12% of the total energy loss (respiration + ammonia excretion) in juvenile bay scallops, making a significant contribution to the energy budget. Adult bay scallops A. i. concentricus were found to lose a similar percentage (7- 15.5%) of energy through ammonia excretion (calculated from Barber and Blake 19851. In contrast, the energy loss in bay scallop larvae attributable to ammonia excretion is much less, comprising only 1 .5-3.6% of the total energy loss. This may be a result of the high oxygen consumption rate, because weight-specific ammonia excretion of larvae is comparable with that of juveniles; whereas, weight-specific oxygen consumption is much higher. The highest weight-specific rate of ammonia excretion was found for 120 p.m larvae. Bay scallop life history begins with a brief lecithotrophic stage, during which energy metabolism is sup- ported by the energy reserve of eggs. As determined by Lu and Blake ( 1997). larvae of 120 p.m shell length have started to feed on phytoplankton. The observed high ammonia excretion rate may indicate that these larvae cannot assimilate sufficient energy for their metabolic demand and still must partially depend on the energy reserves from the eggs, which consist mainly of protein (Lu 1996). In eyed larvae, protein metabolism is dramatically reduced, as shown by the low weight-specific ammonia excretion of larvae at this stage. Heavy utilization of protein occurs during metamor- phosis, and, thus, a high rate of ammonia excretion would be expected for metamorphosing larvae. Data from the present study show that the smallest juveniles tested (500 p.m shell height) did not display significantly higher rates of ammonia excretion than larger ones as expected, probably because they had finished meta- morphosis and already lost the characteristics of metamorphosing larvae. Direct measurements on metamorphosing larvae may pro- vide some evidence on this matter. Weight-specific ammonia excretion found for juvenile bay scallops in the present study (146-250 ngN mg DW~' h"1) is higher than that found for adults (72 to 140 ngN mg DW"1 h-1). This is consistent with the findings that weight-specific physi- ological rates (feeding and respiration) are higher in juveniles than in adults. Oxygen consumption measured for larvae and juveniles of the bay scallop is comparable to that found for other bivalve larvae and juveniles (reviewed by Sprung 1984; Beiras and Camacho 1994), despite the fact that our measurements were made at a higher temperature (25 °C). Because metabolic rates are often determined at the optimum temperature range for each species, this may simply indicate that the bay scallop is adapted to the tem- perature encountered in its natural environment. Our data show that weight-specific ammonia excretion of larvae and juveniles is slightly higher but comparable to those determined for adult bay scallops, following the general trend that weight-specific physi- ological rates decrease as animals develop. ACKNOWLEDGMENT The authors thank Joe Donnelly, Steven Gustafson. and How- ard Lutherford for their lab assistance. Bayne. B. L. & R. C. Newell. 1983. Physiological energetics of marine mollusks. pp. 40-415. In: A. S. M. Saleuddin and K. M. Wilbur (eds.). The Mollusca. vol. 4. Academic Press. San Diego. Bayne. B. L.. R. J. Thompson & J. Widdows. 1976a. Physiology: I. pp. 121-206. In: B. L. Bayne (ed.). Marine Mussels. Cambridge University Press. Cambridge. UK. Bayne, B. L.. J. Widdows & R. J. Thompson. 1976b. Physiology: II. pp. 207-260. In: B. L. Bayne (ed.). Marine Mussels. Cambridge University Press, Cambridge, UK. Beiras. R. & A. P. Camacho. 1994. Influence of food concentration on the physiological energetics and growth of Ostrea edulis larvae. Mar. Biol. 120:427-435. Barber, B.J. & N.J. Blake. 1985. Substrate catabolism related to repro- duction in the bay scallop Argopecten irradians concentricus, as de- termined by O/N and RQ physiological indexes. Mar. Biol. 87:13-18. Brett. JR. & T. D. D. Groves. 1979. Physiological energetics, pp. 280- 352. In: W. S. Hoar, D. J. Randall & J. R. Brett (eds.). Fish Physiology, LITERATURE CITED vol. 8. San Diego, Academic Press. Bricelj, V. M., J. Epp & R. E. Malouf. 1987. Comparative physiology of young and old cohorts of bay scallop Argopecten irradians irradians (Lamarck): mortality, growth, and oxygen consumption. J. Exp. Mar. Biol. Ecol. 1 12 (2):73- 91. Elliott. J. M. & W. Davidson. 1975. Energy equivalents of oxygen con- sumption in animal energetics. Oecologia (Bert.) 19:195-201. Fuji, A. & M. Hashizume. 1974. Energy budget for a Japanese common scallop, Patinopecten yessoensis (Jay), in Matsu Bay. Bull. Fac. Fish. Hokkaido Univ. 25:7-19. Gerdes, D. 1983. The Pacific oyster Crassostrea gigas. part I. feeding behavior of larvae and adults. Aquaculture 31:221-231. Haws, M. C, L. DiMichele & S. C. Hand. 1993. Biochemical changes and mortality during metamorphosis of the eastern oyster, Crassostrea vir- ginica, and the pacific oyster, Crassostrea gigas. Mol. Mar. Biol. Bio- technol. 2:207-217. Oxygen Consumption /Ammonia Excretion in Larval and Juvenile Bay Scallops 423 Hickman. R. W. & L. D. Gruffydd. 1971. The histology of the larva of Ostrea edulis during metamorphosis, pp. 282-294. In: D. J. Crisp (ed.). Proceedings of the Fourth European Marine Biology Symposium. Cambridge University Press. New York. Holland, D. L. 1978. Lipid reserves and energy metabolism in the larvae of benthic marine invertebrates, pp. 85-123. In: D. C. Malins & JR. Sargent (eds.). Biochemical and Biophysical Perspectives in Marine Biology, vol. 4. Academic Press. San Diego. Klyashtorin. L. B. & A. A. Yarzhombek. 1973. Energy consumption in active movements of planktonic organisms. Oceanology 13:575-580. Lu. Y. T. 1996. Physiological energetics of larvae and juveniles of the bay scallop Argopecten irradians concentricus (Say). Dissertation, Univer- sity of South Florida. 160 pp. Lu. Y. T. & N. J. Blake. 1996. Optimum concentrations of Isochrysis gal- bana for growth of larval and juvenile bay scallop, Argopecten irra- dians concentricus (Say). J Shellfish Res. 15:635-644. Lu. Y. T. & N. J. Blake. 1997. Clearance and ingestion rates of Isochrysis galbana by larval and juvenile bay scallops. Argopecten irradians con- centricus (Say). J. Shellfish Res. 16:47-54. MacDonald. B. A. 1988. Physiological energetics of Japanese scallop Pa- tinopecten yessoensis larvae. J Exp. Mar. Biol. Ecol. 120:155-170. Riisgard. H. U.. A. Randlov & K. Hamburger. 1981. Oxygen consumption and clearance as a function of size in Mytilus edulis L. veliger larvae. Ophelia 20:179-183. Siddall. S. E. 1987. Comparative study of oxygen uptake rates in individual larvae and postlarvae of the bay scallop, Argopecten irradians. J. Shell- fish Res. 7:175. Solorzano, L. 1969. Determination of ammonia in natural waters by the phenolhypochlorite method. Limnol. Oceanogr. 14:799-800. Sprung, M. 1984. Physiological energetics of mussel larvae (Mytilus edu- lis). III. respiration. Mar. Ecol. Prog. Ser. 18:171-178. Torres, J. J.. A. V. Aarset, J. Donnelly. T. L. Hopkins. T. M. Lancraft & D. G. Ainley. 1994. Metabolism of Antarctic micronektonic Crustacea as a function of depth of occurrence and season. Mar. Ecol. Prog. Ser. 113:207-219. Zeuthen. E. 1947. Body size and metabolic rate in the animal kingdom with special regard to the marine microfauna. C. r. Trav. Lab. Carlsberg 26:17-161. Journal of Shellfish Research, Vol. 18, No. 2. 425-429, 1999. BIOCHEMICAL UTILIZATION DURING EMBRYOGENESIS AND METAMORPHOSIS IN THE BAY SCALLOP, ARGOPECTEN IRRADIANS CONCENTRICUS (SAY) YANTIAN T. LU, NORMAN J. BLAKE AND JOSEPH J. TORRES Department of Marine Science University of South Florida St. Petersburg, Florida 33701 ABSTRACT Protein, total lipid, and carbohydrate were measured for spawned eggs, D-shaped larvae, premetamorphic, and meta- morphic larvae of the bay scallop Argopecten irradians concentricus (Say). Spawned eggs were composed of 64.0% protein, 26.5% lipid, and 9.5% carbohydrate. After 48 hours of embryogenesis, 13.6% of the protein, 46.3% of the lipid, and 20.8% of the carbohydrate mass had been lost, providing 25.5%, 69.8%, and 4.8% of the total energy expenditure of 0.176 mJ per embryo. During 48 hours of metamorphosis, lipid was utilized first, followed by a heavy consumption of protein; protein, lipid and carbohydrate lost 55.6%, 59.4%, and 67. 3% of their mass respectively. Protein and lipid supplied a comparable amount of energy for metamorphosis, 47.9% and 43.5%, respectively; whereas, carbohydrate contributed only 8.6%, to the 4.35 mJ per larva metamorphic energy expenditure. INTRODUCTION Early survival of marine bivalves is limited by two critical developmental stages: embryogenesis and metamorphosis. Be- cause of the difficulties in obtaining such information from the natural environment, our knowledge on these subjects comes pri- marily from laboratory studies and hatchery production of com- mercially important species. Low survival of early developmental stages in laboratory studies has often been found to be associated with embryogenesis and metamorphosis (Castagna and Duggan 1971. Heffernan et al. 1992), with fluctuations in environmental factors often increasing mortalities (Tettelbach and Rhodes 1981, Lu 1989. Lu and Blake 1996). Embryogenesis of most marine bivalves occurs in surrounding waters rather than in the female and, thus, represents a environ- mental stage of negative energy balance, because embryos do not have a digestive system and cannot feed on particulates. Endog- enous energy reserves of eggs are mobilized to supply the energy necessary for embryogenesis. and this process lasts to at least the formation of a digestive tract. In bivalves, this represents a devel- opment from eggs to straight hinge veligers or D-shaped larvae. Early veligers of the bay scallop Argopecten irradians concentri- cus still need to rely partially on endogenous reserves in addition to energy obtained from feeding (Lu 1996). Metamorphosis represents the second phase of development, wherein stored energy reserves are consumed for metabolism dur- ing early development (Whyte et al. 1990). Energy reserves are accumulated by the planktonic larvae through feeding on organic particles and are subsequently used for supporting metamorphosis (Rodriguez et al. 1990. Haws et al. 1993). During metamorphosis, the larval velum disappears, and larvae lose their ability to feed until the development of gill filaments (Sastry 1965, Bayne 1965, Hickman and Gruffydd 1971 >. Success in completing embryogenesis and metamorphosis is determined to a large extent by the amount of energy reserves inherited from the female and/or accumulated through larval feed- ing. High survival has been found in the large eggs of Mercenaria mercenaria and Argopecten irradians (Kraeuter et al. 1982). prob- ably because of their higher energy content. Survival may depend upon the ability of embryos or larvae to complete development of feeding structures before energy reserves are depleted (Haws et al. 1993). The southern bay scallop was reported to produce smaller oo- cytes (Barber and Blake 1981) than its northern counterpart. Low energy reserves of small eggs and higher metabolic demand asso- ciated with higher temperature in its natural habitat could be dis- advantageous to the early development of the southern bay scallop. However, information on the energy metabolism of early devel- opment in the bay scallop is lacking. The objective of this study was to investigate changes in biochemical composition of eggs, larvae, and juveniles, and the energy expenditures associated with embryogenesis and metamorphosis of the southern bay scallop Argopecten irradians concentricus (Say). MATERIALS AND METHODS Sample Collection Bay scallops were collected from Homosassa. Florida and were spawned in the lab of the Department of Marine Science. Univer- sity of South Florida. Culturing of larvae and juveniles followed the methods described by Lu and Blake ( 1996). Larvae were raised at a density of 4-8 mL1 and fed daily with 10,000-30.000 cells mL1 of Isochrysis galbana, depending upon larval size. Seawater was replaced every day in the amount of 1/3 of the total volume. Daily food ration for juveniles was increased gradually from 30,000 to 100,000 cells mL"1 of /. galbana. Fertilized eggs were collected onto a 28 (xm nylon screen. A portion of the eggs were used for biochemical analysis, and the rest were released in 1.2 p.m filtered seawater (25% S) and allowed to develop at 26 °C. After 48 hours, D-shaped larvae were collected on a 35-jjim nylon screen. The egg and larval samples were washed with 1.2-p.m filtered seawater and pipetted to a graduated cylinder. Seawater was added to bring the volume to 100 mL and total eggs or larvae were determined by counting five 0.5-mL samples. Mean egg diameter and larval shell length were determined by measuring 50 individuals each using a microscope fitted with a micrometer. Subsamples were drawn from the graduated cylinder, washed three times with 3% ammonium formate solution to remove salt, and frozen at -20 °C until analyzed for biochemical composition. Each sample for chemical analysis contains 100,000-150.000 eggs or 150,000-200.000 D-larvae. Premetamorphic larvae (mean shell height 185 ± 7.8 (juti) were kept in 2000-mL plastic beakers containing 1.2-jim filtered sea- water without food. A sample of larvae was taken 24 hours later. Development was followed microscopically, and juveniles were observed in the culture at 48 hours. Veligers that settled on the 425 426 LU ET AL. beakers were brushed into a petri dish. Juveniles were separated from the pediveligers under a microscope. Samples of larvae and juveniles were washed with 3% ammonium formate solution, quantified, and frozen at -20 °C until analyzed. Each sample for chemical analysis contains 3.000-5,000 premetamorphic larvae or 1.000-2,000 postmetamorphic larvae. Egg and larval samples were homogenized in 1.5-mL DI water using a 4710 series ultrasonic homogenizer (Cole-Parmer Instru- ment Co.). Subsamples were taken from the slurry for protein, lipid, and carbohydrate analysis. Juveniles of various shell height (1-10 mm) were collected from a 300 Liter stocking tank at various times. Sample size ranges from 1-50 individuals, depending upon size. For juveniles less than 2 mm, it is difficult to completely separate the soft body from shells; consequently, whole animals were homogenized in DI wa- ter with a tissue grinder and then an ultrasonic homogenizer. For larger juveniles up to 10 mm. only the soft body was homogenized. In all cases, samples were analyzed within 1 month of collection. Biochemical Analysis Under most circumstances, three samples were analyzed for eggs and larvae of each development stage. The exception is with juveniles (>2 mm shell height), which were analyzed individually. Protein analysis followed the Folin phenol method of Lowry et al. (1951 ). Three samples of 20-p.L homogenate each were trans- ferred to test tubes, to which 60 u.L of DI water was added to make final volumes of 80 ptL each. Samples were hydrolyzed by adding 0. 12 mL of 0. 1 N NaOH to each test tube and heating at 1 00 °C for 10 minutes. After cooling, 1.2 mL Reagent B (0.5 mL 1% CuS04/ 5H20 + 0.5 mL 2% Na Tartrate + 50 ml 2% Na2CO,) was added and allowed to sit for 10 min. Then 0.12 mL reagent E (phenol reagent diluted to 1 N with water) was added and mixed immedi- ately. After 30 min. the optical density of the blue solution was read at 750 nm on a Cary 2000 spectrophotometer with bovine serum albumin as the standard. Lipid was extracted according to the method of Bligh and Dyer ( 1959). Three samples of 0.2-mL homogenate each were placed in test tubes. To each sample, 0.73 mL of 2: 1 MeOH:CHCl„ 0.24- mL chloroform, and 0.24-mL H20 were added and mixed after each addition. The samples were allowed to sit for 1 h for sepa- ration. The two phase solutions were poured into 0.45-p.m PTFF centrifuge filter units, and the lipid fraction passing through was collected in test tubes and dried at 30 °C under a flow of nitrogen. Total lipid content was determined using the charring method of Marsh and Weinstein (1966). To each dried sample, 1 mL of concentrated H2S04 was added, followed by heating at 200 °C for 20 minutes. The chairing samples were cooled in tap water and were then diluted with 3-mL H20. After cooling, the optical den- sity was read at 375 nm on a spectrophotometer with stearic acid as the standard. Total carbohydrate content was determined using the phenol- sulfuric acid method of Dubois et al. (1956). Three samples of 0.2-mL homogenate were pipetted to centrifuge tubes, washed consecutively with acetone and ether to remove lipids, and dried at 60 °C. The dried samples were hydrolyzed in 2.0 mL 5% trichlo- roacetic acid at 100 °C for 20 minutes. After cooling, the tubes were centrifugated at 5,000 rpm for 10 minutes and 1-mL super- natant was removed from each sample to 16 x 150-mm test tubes. To each test tube, 0.5 mL of 5% phenol and 2.5 mL of concen- trated H2S04 were added. After cooling for 30 minutes, the optical density of the orange-yellow solution was read at 490 nm on a Cary 2000 spectrophotometer with oyster glycogen as the stan- dard. Energy expenditures for embryogenesis and metamorphosis were estimated by the loss of protein, lipid, and carbohydrate during those two events. Energy conversion factors used were 20.0, 39.5. and 17.5 mJ u.g~' for protein, lipid, and carbohydrate respectively (Brett and Groves 1979). RESULTS Table 1 gives the content of protein, lipid, and carbohydrate at various developmental stages and shell lengths. The spawned eggs of the southern bay scallop had a mean diameter of 60.7 ± 0.8 u.m, and were composed primarily of protein (64.0%). lipid (26.5%), and carbohydrate (9.5%). Mean energy content of an egg was 0.631 mJ. During the 48 hours of embryogenesis, all three com- ponents decreased. Protein dropped 13.6% by mass, lipid 46.3%, and carbohydrate 20.8%, respectively, leading to a loss of 22.9% of the total organic matter (defined here as the total of protein, lipid, and carbohydrate) (Table 2). Lipid was the major substrate utilized for embryogenesis, supplying 69.8% of the total energy expenditure of 0.176 mJ per embryo, while protein and carbohy- drate contributed 25.5% and 4.8% respectively. Lipid was accumulated as the larvae developed, increasing from 18.5% in D-larvae to 26.9% in premetamorphic larvae. Mean energy content of premetamorphic larvae (185-p.m shell length) was 7.49 mJ per larva. During the first 24 hours of metamorphosis, protein, lipid, and carbohydrate lost 24.4%, 38.8%, and 53.6% of their original content, respectively. Lipid supplied half of the total energy expenditure, and protein and carbohydrate made up the rest. Carbohydrate was the substrate that lost the highest portion. but it only contributed 12.2% to the total energy expenditure be- cause of its low absolute content. During the second 24 hours of metamorphosis, more protein was lost (41.3%) than lipid (33.5%) and carbohydrate (29.4%). Protein surpassed lipid as the major substrate in fulfilling the en- ergy demand. The role of carbohydrate as an energy reserve was further reduced. Overall, the process of metamorphosis consumed 57.9% of the total organic substrate, contributed by 55.6%, 59.4%. and 67.3% of protein, lipid, and carbohydrate reserves respectively. Total energy expenditure during metamorphosis was 4.35 mJ per larva, with TABLE 1. Argopecten irradians concentricus. Changes in biochemical content (ng ind."') during embryogenesis and metamorphosis. Developmental Length Stages (jim) Protein Carbo- Lipid hydrate Total Fertilized eggs D-larvae Premetamorphic starved 0 hrs starved 24 hrs 61 98 185 187 Postmetamorphic 230 16.2 ±0.5 6.7 + 0.4 2.4 ±0.1 25.3 (64.0%) (26.5%) (9.5%) 14.0 ±0.1 3.6 ±0.2 1.9 ±0.1 19.5 (71.8%) (18.5%) (9.7%) 187.4 ±6.6 80.5 ±7.3 3 1.8 ±0.4 299.7 (62.4%) (26.9%) (10.6%) 141.7 ±2.8 49.3 ±5.6 14.7 ±0.2 205.6 (68.9%) (24.0%) (7.1%) 83.2 ±1.9 32.8+1.6 10.4 ±0.2 126.4 (65.8%) (25.9%) (8.2%) Biochemical Utilization in the Bay Scallop 427 TABLE 2. Argopecten irradians concentricus. Losses of biochemical substrates, their caloric equivalents and contribution to energy expenditures of emhryogenesis and metamorphosis. Protein Lipid Carbohydrate Total Embryogenesis Wt loss (ng ind."1) 2.2 3.1 0.5 5.8 Wt loss (%) 13.6% 46.3% 20.8% 22.9% E equiv. (mJ ind.-') 0.0449 0.1229 0.0084 0.1762 E contribution (%) 25.5% 69.8% 4.8% 100% Metamorphosis Wt loss (ng ind.-1) 104.2 47.8 21.4 173.4 Wt loss (%) 55.6% 59.4% 67.3 57.9% E equiv. (mJ indr1) 2.0841 1.8909 0.374 4.3487 E contribution (%) 47.9% 43.5% 8.6% 100% protein and lipid providing 47.9% and 43.5%, respectively, and carbohydrate contributing only 8.6%. Protein, lipid, and carbohydrate content of juveniles are sum- marized in Table 3. In contrast to the larval biochemical compo- sition, juveniles had significantly higher levels of protein (P < .001 ) and lower levels of lipid [P < .001 ). There was no significant difference in carbohydrate levels between larvae and juveniles. Protein was the major component in the biochemical composition of juveniles, constituting 72.6 ± 3.0%' of the total organic matter, and mean lipid and carbohydrate components were 18.6 ± 1.1% and 8.8 ± 2.5%, respectively. As in the larval stage, carbohydrate was the least significant component in juveniles. DISCUSSION Embryogenesis of marine invertebrates is an energy-consuming process during which embryos rely solely on energy reserves within the eggs provided by the female. Existing information has shown that energy stored in other parts of the body is transferred to the gonads during gametogenesis (Gabbott 1976, Bayne 1976, Barber and Blake 1981 ) and energy stored in eggs as protein, lipid, and carbohydrate substrates are subsequently utilized for growth and development of the embryo (Gallager et al. 1986. Whyte et al. 1990). Protein forms the main constituent of eggs, followed by TABLE 3. Argopecten irradians concentricus. Protein, total lipid, and total carbohydrate content (fig ind.-1) in juveniles. Mean Height iiiiiii) Protein Lipid Carbohydrate Total 1.07 8.8 (63.3%) 3.3 (23.7%) 1.8(12.9%) 13.9 2.1 68.5(71.6%) 18.0(18.8%) 9.2 (9.6%) 95.7 2.2 81.3(68.6%) 25.0(21.1%) 12.2(10.3%) 118.5 2.8 112.1 (67.9%) 30.2(18.3%) 22.7(13.8%) 165.0 3.0 162.4 (73.7%) 40.7(18.5%) 17.2(7.8%) 220.3 3.1 176.8(74.2%) 43.8(18.4%) 17.8(7.5%) 238.4 3.4 248.3 (69.4%) 64.9(18.1%) 44.7(12.5%) 357.9 4.8 566.1 (74.7%) 140.5(18.5%) 51.3(6.8%) 757.9 5.1 682.8 (74.5%) 164.6(18.0%) 69.4(7.6%) 916.8 5.2 685.3 (75.6%) 160.8(17.7%) 60.7 (6.7%) 906.8 5.5 837.4(71.6%) 2054(17.6%) 74.2(6.3%) 1170.0 5.8 953.7 (75.1%) 223.7(17.6%) 92.9(7.3%) 1270.3 lipid and then carbohydrate in marine invertebrates (reviewed by Holland 1978). The present study on the eggs of the bay scallop Argopecten irradians concentricus shows the same trend: protein forms the main biochemical constituent (64.1%), followed by lipid (26.4%). Carbohydrate is the smallest component of the three (9.5%). Energy reserves in the form of protein, lipid, and carbohydrate were utilized as eggs developed, which was indicated by the re- duction of corresponding substrates at the end of embryogenesis. In the bay scallops, lipid was used as the principal energy source for egg development, supplying 69.8% of the total energy expen- diture, more than protein (25.5%) and carbohydrate (5.5%) com- bined. Carbohydrate played a minor role as an energy reserve in the bay scallop because of its low content in eggs. High conversion efficiency (Wr,nalAVln,lial. where W is the weight of a given bio- chemical constituent (see Holland 1978) of protein (86.4%) and low conversion efficiency of lipid (53.7%) provide evidence that protein is conserved for the formation of planktonic larvae, and lipid is the major energy reserve fueling this process. Information on the utilization of major biochemical constitu- ents for embryogenesis of marine invertebrates comes mainly from studies on crustaceans. Although some species use lipid as the major energy reserve (Pandian 1967, Pandian and Schumann 1967. Shakuntala 1977), other species rely primarily on protein (Barnes 1965, Lucas and Crisp 1987). The utilization of lipid as the dom- inant energy reserve in developing eggs was also reported for the Pacific halibut Hippoglossus stenolepis (Schmidt) (Whyte et al. 1993) and the red drum Sciaenops ocellata (Vetter et al. 1983); whereas utilization of protein dominated in the rainbow trout Salnw gairdneri (Oliva-Teles and Kaushik 1987). In bivalves, 69 and 71% of total lipid was lost during embryogenesis of the clam Mercenaria mercenaria and the oyster Crassostrea virginica, re- spectively (Gallager et al. 1986), indicating heavy use of lipid: however, lack of information on changes in protein and carbohy- drate contents prevents estimation on relative importance between lipid and protein. Whyte et al. (1990) reported that lipid and pro- tein substrates contributed equally to the energy expenditure of embryogenesis of the rock scallop Crassadoma gigantea, account- ing for 46.7 and 43.5%, respectively, and carbohydrate supplied only 9.8%. During the planktonic stages of the bay scallop, larvae feed on organic particles, and their organic mass increases as larvae grow (Lu and Blake 1996). As a result, larvae build up energy reserves of protein, lipid, and carbohydrate through feeding (Table 1 ). Rela- tive lipid content increased from 18.5% in D-shaped larvae to 26.9% in premetamorphic larvae. Bay scallop larvae use lipid reserves at the beginning of meta- morphosis followed by a heavier consumption of protein reserves. During the first 24 hours, metamorphosing larvae derived 50.5% of their energy from lipid and 37.3% from protein. During the next 24 hours, lipid supplied 34.4% of the total energy expenditure of metamorphosing larvae, and protein supplied the bulk of it, 61.6%. On average, protein and lipid provided similar amounts of energy. 47.9 and 43.5%, respectively, during the 48-hour metamorphosis. As in embryogenesis. carbohydrate is the least important constitu- ent in metamorphosing larvae and contributed only 8.6% of the energy expenditure. Disagreement remains on which substrate serves as the major source of energy for metamorphosis in bivalves. In the oyster Ostrea edulis, Holland and Spencer ( 1973) reported that over half of the neutral lipid reserves were used for metabolism during rneta- 428 LU ET AL. morphosis, but the authors had no data for the use of protein, carbohydrate, or phospholipid. However, a later study on the same species by Rodriguez et al. (1990) found protein supplied most of the energy for metamorphosis (62%), more than twice that sup- plied by lipid (28%). In two other species of oysters. Crassostrea virginica and C. gigas, 50.2% and 51.1% of their total energy expenditure during metamorphosis came from lipid and 39.9% and 38.5% from protein, respectively (Haws et al. 1993). In the rock scallop Crassadoma gigantea (Gray I, Whyte et al. ( 1992) reported that protein formed 59.9% of the total energy expenditure during metamorphosis and lipid 38.5%. However, their data, like those of Whyte et al. (1990), were derived from relative values and were found hard to interpret. In the bay scallop, metamorphosis consumed 57.9% of the total organic reserves, equivalent to a total energy expenditure of 4.35 ml per larva. This may represent the minimum level of consum- able energy reserve of bay scallop larvae, below which larvae cannot complete metamorphosis without obtaining energy from their surrounding environment. Assuming that metamorphosing larvae have the same respiration rate as eyed larvae ( 1 4.664 u.L02 mgAFDW~' h-1; Lu 1996). energy expenditure during 48 hours of metamorphosis can be calculated to be 4.55 mj. This value is close to the value of 4.35 mJ determined in this study, to further support our estimation on the energy requirement for metamorphosis of the bay scallop. Those values are comparable to those determined for the oyster Crassostrea virginica and C. gigas (2.13 and 4.65 mj per larva, respectively, over a 36-hour period) (Haws et al. 1993). The oyster Ostrea edulis seems to lose more energy (5.62-14.65 mJ/larva over a 36^t8-hour period) during metamorphosis (Rod- riguez et al. 1990). One of the main difficulties in estimating the biochemical changes and the associated energy metabolism is the lack of knowledge on when premetamorphic larvae stop feeding and for how long. Our understanding of biochemical energetics in meta- morphosing larvae has been based on the assumption that meta- morphosing larvae lack the capability to feed and that metamor- phic energy demand comes solely from energy reserves accumu- lated during the planktonic stage. In a recent feeding study on the oyster Crassostrea virginica. Baker and Mann (1994) found that all prodissoconchal and dissoconchal metamorphs ingested the ex- perimental microspheres, and, except for only a few hours during the settler phase, feeding was possible throughout the oyster meta- morphosis. If this finding holds true for other bivalve larvae, en- ergy obtained through feeding by metamorphosing larvae has to be taken into account in the estimation of energy budget. However, it is not clear what the quantitative contribution of feeding to the total energy expenditure of metamorphosing larvae is. Results of this and other studies (e.g.. Rodriguez et al. 1990. Haws et al. 1993) demonstrate that bivalve larvae are able to complete meta- morphosis based solely on the accumulated energy reserves of their biochemical substrates. In the natural environment, larvae may cease feeding at a very late stage of metamorphosis, or larvae may stop feeding for only a few hours during metamorphosis, as observed by Baker and Mann ( 1994). As a result, the minimum energy level can be substantially lower, and larvae may not need as much energy reserve for meta- morphosis as previously thought. Therefore, the estimation made here may represent the upper range of the metamorphic metabolic- demand of the bay scallop. ACKNOWLEDGMENT The authors thank Joe Donnelly for his assistance in the bio- chemical analysis. REFERENCES Baker, S. M. & R. Mann. 1994. Feeding ability during settlement and metamorphosis in the oyster Crassostrea virginica (Gmelin. 1791) and the effects of hypoxia on postsettlement ingestion rates. J. Exp. Mar. Biol. EcoL 181:239-253. Barber, B. J. & N. J. Blake. 1981. Energy storage and utilization in relation to gametogenesis in Argopecten irradians concentricus (Say). J. Exp. Mar. EcoL 52:121-134. Barnes, H. 1965. Studies in the biochemistry of drripede eggs. 1. changes in the general biochemical composition during development of Balanus balanoides and B. balanus. ./. Mar. Biol. Ass. U.K. 45:321-339. Bayne. B. L. 1965. Growth and delay of metamorphosis of the larvae of Mytilus edulis (L.). Ophelia 2:419-443. Bayne, B. L. 1976. Aspects of reproduction in bivalve molluscs, pp. 432- 448. In: M. L. Wiley (ed.). Estuarine Processes. Academic Press, New York. Bligh. E. G. & W. J. Dyer. 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37:911-917. Brett. J. R. & T. D. D. Groves. 1979. Physiological energetics, pp. 280- 352. In: W. S. Hoar. D. J. Randall and J. R. Brett (eds.h Fish Physi- ology, vol. 8. Academic Press, New York. Castagna, M. & W. Duggan. 1971. Rearing the bay scallop. Aequipecten irradians. Proc. Natl. Shellfish. Assoc. 61:80-85. Dubois. M., K. A. Gilles. J. K. Hamilton. P. A. Rebers & F. Smith. 1956. Colorimetric method for determination of sugars and related sub- stances. Aiialxt. Chem. 28:350-356. Gabbott. P. A. 1976. Energy metabolism, pp. 293-355. In: B. L. Bayne (ed.). Marine Mussels: Their Ecology and Physiology. Cambridge Uni- versity Press, Cambridge. Gallager, S. M.. R. Mann & G C. Sassaki. 1986. Lipid as an index of growth and viability in three species of bivalve larvae. Aquacuhure 56:81-103. Haws. M. C. L. DiMichele & S. C. Hand. 1993. Biochemical changes and mortality during metamorphosis of the eastern oyster. Crassostrea vir- ginica. and the Pacific oyster, Crassostrea gigas. Mol. Mar. Biol. Bio- technol. 2:207-217. Heffernan, P. B.. R. L. Walker & J. W. Crenshaw. Jr. 1992. Embryonic and larval responses to selection for increased rate of growth in adult bay scallop, Argopecten irradians concentricus (Say. 1822). J. Shellfish Res. 11:21-25. Hickman. R. W. & L. D. Gruffydd. 1971. The histology of the larva of Ostrea edulis during metamorphosis, pp. 282-294. In: D. J. Crisp (ed.). Fourth European Marine Biology Symposium. Cambridge University Press, New York. Holland. D. L. 1978. Lipid reserves and energy metabolism in the larvae of benthic marine invertebrates, pp. 85-123. In: DC. Malins and J. R. Sargent (eds.h Biochemical and Biophysical Perspectives in Marine Biology, vol. 4. Academic Press, New York. 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. Ass. U.K. 53:287-298. Kraeuter. J. N.. M. Castagna & R. Van Dessel. 1982. Egg size and larval survival of Mercenaria mercenaria (L.) and Argopecten irradians (Lamarck). J. Exp. Mar. Biol. Ecol. 56:3-8. Lowry. O. H.. N. J. Rosebrough. A. L. Farr & R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:65- 275. Lu. Y. T. 1989. Effect of zinc on the growth and development of larvae of Biochemical Utilization in the Bay Scallop 429 the bay scallop Argopecten irradians. Chin. J. Oceanol. Limnol. 7:3 1 8- 326. Lu, Y. T. 1 996. Physiological energetics of larvae and juveniles of the hay scallop Argopecten irradians concentricus (Say). Dissertation, Univer- sity of South Florida. 160 pp. Lu. Y. T. & N. J. Blake. 1996. Optimum concentrations of Isochrysis gal- bana for growth of larval and juvenile bay scallop. Argopecten irradi- ans concentricus (Say). J. Shellfish Res. 15:635-644. Lucas, M. I. & D. J. Crisp. 1987. Energy metabolism of eggs during embryogenesis in Balanus balanoides. J. Mar. Biol. Assoc. U.K. 67: 27-54. Marsh, J. B. subject D. B. Weinstein. 1966. Simple charring method for determination of lipids. ./. Lipid Res. 7:574-576. Oliva-Teles. A. & S. J. Kaushik. 1987. Nitrogen and energy metabolism during the early ontogeny of diploid and triploid rainbow trout (Salmo gairdneri R.). Comp. Biochem. Physiol. A. 87A:157-160. Pandian, T. L. 1967. Changes in chemical composition and caloric content of developing eggs of the shrimp Crangon crangon. Helgoldnder wiss. Meeresunters 16:216—224. Pandian, T. L. & K. H. Schumann. 1967. Chemical composition and ca- loric content of egg and zoea of the hermit crab Eupagurus bernhardus. Helgoldnder wiss. Meeresunters 16:225-230. Rodriguez, J. L.. F. J. Sedano, L. O. Garcia-Martin, A. Perez-Camacho & J. L. Sanchez. 1990. Energy metabolism of newly settled Ostrea edulis spat during metamorphosis. Mar. Biol. 106:109-1 1 1. Sastry, A. N. 1965. The development and external morphology of pelagic larval and postlarval stages of the bay scallop, Aequipeclen irradians concentricus Say, reared in the laboratory. Bull. Mar. Sci. 1 5:4 17—435. Shakuntala, K. 1977. Yolk utilization in the freshwater prawn Macro- hrachium lamarrei. J. Anim. Morphol. Physiol. 24:13-20. Tettelbach. S. T. & E. W. Rhodes. 1981. Combined effects of temperature and salinity on embryos and larvae of the northern bay scallop, Ar- gopecten irradians irradians. Mar. Biol. 63:249-256. Vetter, R. D., R. E. Hodson & C. Arnold. 1983. Energy metabolism in a rapidly developing marine fish egg. the red drum (Sciaenops ocellata). Can. J. Fish. Aquat. Sci. 40:627-634. Whyte, J. N. C. N. Bourne & N. G. Ginther. 1990. Biochemical and en- ergy changes during embryogenesis in the rock scallop Crassadoma gigantea. Mar. Biol. 106:239-344. Whyte, J. N. C. N. Bourne, N. G. Ginther & C. A. Hodgson. 1992. Com- positional changes in the larva to juvenile development of the scallop Crassadoma gigantea (Gray). J. Exp. Mar. Biol. Ecol. 163:13-29. Whyte, J.N. C, W. C. Clarke. N. G. Ginther & J. O. T. Jensen. 1993. Biochemical changes during embryogenesis of the pacific halibut, Hip- poglossus stenolepis (Schmidt). Broodstock Manag. Egg Lanxil Qual- ity 24:193-201. Journal oj Shellfish Research, Vol. IS. No. 2. 431-135. 1999. CHROMOSOMAL LOCATION BY FLUORESCENCE IN SITU HYBRIDIZATION OF THE 28S RIBOSOMAL RNA GENE OF THE EASTERN OYSTER QUIYANG ZHANG,1 2 GANG YU,1 RICHARD K. COOPER,2 AND TERRENCE R. TIERSCH1 ' Aquaculture Research Station, and ^'Department of Veterinary Science, Louisiana Agricultural Experiment Station, Louisiana State University Agricultural Center, Baton Rouge, Louisiana 70803 ABSTRACT The physical location of the 28S ribosomal RNA gene (28S rDNA) was localized to the short arm of chromosome number 2 of the eastern oyster (Crassostrea virginica) by fluorescence in situ hybridization. The existence of a single locus of the 28S rDNA in the genome of the eastern oyster was concluded based on the findings from metaphase chromosomes prepared from gill, mantle, and embryos and from meiotic chromosome bivalents prepared from ripe gonad tissue. The region of the chromosome occupied by the gene was found to be GC-rich. and the location of probe DNA used to identify the 28S gene corresponded exactly with the location of the nucleolus organizer region. The transcriptional activity of the gene did not vary among different cell types but changed with different phases of mitosis. This study is the first report of physical mapping of a specific gene in mollusks and provides techniques for detection of integration of foreign DNA in the oyster genome. KEY WORDS: fluorescence in situ hybridization. 28S rDNA. chromosome. Crassostrea virginica INTRODUCTION The ribosomal RNA genes are a group of DNA sequences that produce structural rRNA to support protein synthesis. For ex- ample, the 28S ribosomal RNA gene (28S rDNA) encodes a major component of the large subunit of the eukaryotic ribosome. The nucleotide sequences of these genes are polymorphic, which has enabled taxonomic investigation of organisms with ambiguous phylogenetic relationships (Littlewood 1994). In addition. 28S rDNA is useful for the initiation of physical genome mapping, because the gene exists in multiple copies, which improves the ease of detection. There are various methods for verifying the chromosomal location of 28S rDNA. Fluores- cence in situ hybridization (FISH) (Pendas et al. 1993) or in situ polymerase chain reaction (Zhang et al. 1997) can be used for direct assignment of gene location. The 28S rDNA loci can be revealed indirectly by chromomycin A3 staining to target GC-rich regions (Amemiya and Gold 1987). Moreover, the 28S RNA genes are associated with the nucleolus organizer regions (NOR) (Long and Dawid 1980). By use of silver staining, which targets protein components associated with RNA synthesis (Howell and Black 1980), the active 28S rDNA loci can be detected. The eastern oyster has been harvested commercially in the United States for hundreds of years. An organized system of oyster leases has been maintained in Louisiana for over 100 years, and harvest in Louisiana was valued at more than $50 million in 1997 (Louisiana Summary of Agricultural and Natural Resources 1997). Overall, genetic improvement in this species has been hampered by the absence of basic genetic information. Study of the physical location of DNA sequences in oysters is still preliminary, and reports are restricted to nonspecific DNA elements derived from the mammalian genome (Guo and Allen 1997) or oyster genome (Clabby et al. 1996). The procedures for physical mapping of specific genes must be developed and validated in this species. The chromosomal location of the 28S rDNA in the eastern oyster has been preliminarily investigated using FISH techniques (Zhang et al. 1999a; Xu et al. 1999). Our goal was to use the 28S rDNA as a first step to verify techniques for physical mapping of genes in oysters. Specific objectives of the present study were to: 1 ) localize the 28S rDNA on eastern oyster chromosomes by fluo- rescence in situ hybridization; 2) compare these sites of the 28S rDNA to those detected by silver staining and chromomycin A3 staining; and 3) evaluate the transcriptional activity of the 28S rDNA in different cell types. MATERIALS AND METHODS Probe Construction Nuclear DNA was isolated from hemolymph collected from adult oysters (it = 5) using a QIAamp blood kit (Qiagen Inc., Chatsworth, CA). Probe DNA was synthesized by polymerase chain reaction (PCR). A pair of primers was designed to target the oyster 28S rDNA gene based on a published sequence (Littlewood 1994) and was synthesized by the Gene Probes and Expression Systems Laboratory, Louisiana State University, Baton Rouge. The primer sequences were (.V to 3'); GCTAAATACTTCCCG- AGTCCGATAGC and GCACCTTCCTCCAGCTCTTCTGAC. Conditions for PCR were initial denaturation at 94 °C for 2 min; 35 cycles at 94 °C for 30 s, 52 °C for 30 s, and 72 °C for 1 min. with a final elongation step at 72 °C for 7 min. The PCR products were labeled by inclusion of biotin-16-dUTP in the reaction mix- ture. Labeling was visualized by agarose electrophoresis, in which the shift of band size was detected because of incorporation of label into the PCR products. Chromosome Preparation Ten ripe oysters (five males and five females) were used in this study. Chromosomes were prepared from gonad tissues, embryos at 5 h after fertilization (Paniagua-Chavez et al. 1998) and gill tissues with methods reported elsewhere (Zhang et al. 1999b). For obtaining bivalent chromosomes, the gonad tissue was not treated with colchicine, but a prolonged hypotonic treatment in a 1% sodium citrate solution (-5-7 h) was used to promote separation of 431 432 Zhang et al. chromosomes. Chromosomes were prepared using standard air- drying methods (Zhang et al. 1999b). Fluorescence In Situ Hybridization Chromosomes were prepared on two-well, 7>/7o;i-coated slides (Cel-Line Associates, Inc., Newfield, NJ) and digested with RNase A ( 100 jig/ml) in 2-x SSC buffer at 37 °C for 1 h. After dehydra- tion with a series of ethanol concentrations (75%, 85%, 95%, and 100%), chromosomal DNA was denatured in 70% formamide at 70 °C for 5 min. The slides were chilled immediately in 70% ethanol at -20 °C, and were dipped in 100% ethanol and dried in a laminar-flow hood. The hybridization mixture was composed of biotin-labeled probe (1.2 ng/ml). 50% deionized formamide. 2-x SSC buffer. 10-x Denhardt's solution, 0.1% sonicated salmon sperm DNA, and 0.1% sodium dodecyl sulfate (Chen, 1994). The mixture was heated at 95 °C for 10 min and chilled immediately on ice. Twenty u.1 of the mixture was applied to each well of slides prewarmed to 37 °C. and coverslips were sealed using clear nail polish. The slides were incubated in a wet box at 37 °C for 24 to 36 h. Hy- bridization was detected with avidin-fluorescein isothiocyanate (FITC) (Zhang et al. 1997). Slides were counterstained with pro- pidium iodide (PI) (0.5 p,g/ml) prepared in an antifading medium (100 mg />-phenylenediamine in 100 ml glycerol. pH 7.0). Identification of NOR-Bearing Chromosomes by Silver or CMA Staining Slides were treated with 50% acetic acid for 20 min to remove background materials on embryonic and gonadal chromosomes. To determine the relationship between the 28S rDNA and NOR. chromosomes were stained with chromomycin A3 (CMA) made in a modified Mcllvaine's buffer (Amemiya and Gold 1987). After the images of chromosomes were recorded, slides were rinsed gently with 2-x SSC and left in the buffer for 20 min. The slides were dehydrated through a series of ethanol concentrations (10%, 80%, 95%. and 100%) and stained with a one-step silver staining procedure (Howell and Black 1980). Slides were covered with a solution of 33% silver nitrate and 0.7% gelatin and incubated for 8 to 10 min at 50 °C. Image Analysis and Map Construction Fluorescent images of chromosomes stained with CMA or FITC and PI were examined under a fluorescence microscope (Microphot-SA. Nikon Inc.. Garden City. NY) equipped with fil- ters for FITC and CMA (excitation wavelengths of 480 nm) and PI (excitation wavelength of 535 nm). Fluorescent images of nuclei and chromosomes were photographed using Kodak Ektachrome (400 ASA) color slide film. The negatives were scanned into a computer with a slide scanner (SprintScan 35. Polaroid scanner model CS-2770, Needham Heights. MA) for analysis. For silver staining, chromosomal images were captured and analyzed by a computer-based image analysis technique (Zhang and Tiersch 1998). Individual chromosomes were identified based on a previ- ously developed karyotype (Zhang et al. 1999b). Relative length (RL) and centromeric index (CI) of each chromosome were cal- culated using the following formulae: RL (%) = (length of the chromosome pair/total complement length) x 100 CI (%) = (length of short arm/total length of the chromosome) x 100 The location of the 28S rDNA gene was analyzed by microdensi- tometry, and a map of the chromosome bearing the 28S rDNA was constructed using Microsoft PowerPoint (Office 97 version). RESULTS The location and activity of the 28S rDNA in the genome of the eastern oyster was investigated using four cell sources (Table 1). Two chromosomes from embryo cells were found to hybridize with the 28S rDNA probe (Fig. la). The RL of one chromosome was 6.14 ± 0.12. and the CI was 42.3 ± 1.8: whereas, the other chromosome had a RL of 5.90 ± 0. 10 and a CI of 36.5 ± 2. 1 (n = 10 spreads). Two chromosomes from gill cells (Fig. lb) were found to hybridize with the 28S rDNA (Fig. lb & lc). The same region identified by the 28S rDNA probe was found to stain in- tensely with chromomycin A3, indicating the presence of GC-rich regions (Fig. 2a). and the location of the 28S rDNA was found to be the same as that of the NOR. as indicated bv silver stainina (Fia. 2b). One of 10 bivalent chromosomes prepared from gonad tissue was found to hybridize with the 28S rDNA probe (Fig. 3a). Two hybridization signals sometimes appeared on the same bivalent corresponding to the presence of homologous chromosomes that were in the process of separating (Fig. 3b). The chromosome biva- lent was consistently stained with CMA at diakinesis (Fig. 3c) and pachytene stages (Fig. 3d) and was stained positively with silver nitrate for the NOR at pachytene stage (Fig. 3e). However, in most cases (30 out of 50 spreads), this NOR site was not detectable by silver staining of diakinesis and pachytene chromosomes. Upon more detailed examination of the chromosome in ques- tion (number 2. from embryo), the location of the 28S rDNA, as measured by microdensitometry. was identified at the telomeric regions corresponding to the site of maximal hybridization inten- sity of the 28S rDNA probe (Fig. 4). DISCUSSION In this study, the 28S rDNA of the eastern oyster was found to be localized to the telomeric region of the short arm of chromo- some number 2 by fluorescence in situ hybridization. Although the two 28S rDNA-bearing chromosomes from embryo cells were different in size and centromeric index, a single chromosomal location of 28S rDNA was found using meiotic chromosome bivalents. The area of the gene was found to be GC-rich and to TABLE 1. Characterization of the 28S ribosomal RNA gene on chromosomes of the eastern oyster by different techniques." Tissue Phase of AgNOR CMA Type Cell Division Staining Staining FISH Embryo Prophase Two. NP Not studied Not studied Prometaphase Two. NP Two. NP Two, NP Early metaphase Two. NP Not studied Not studied Metaphase One Two. P. AS Two, P Gill Metaphase One Not studied Two. P Mantle Metaphase One Not studied Two. P Gonad Pachytene One bivalent One bivalent One bivalent Diakinesis One bivalent One bivalent One bivalent J Abbreviations: AgNOR, nucleolus organizer regions stained by silver nitrate: CMA. chromomycin A3; FISH, fluorescence in situ hybridization: NP. not pairable: P. pairable. and AS. asymmetric staining or different staining intensity between the two chromosomes. FISH in Oysters 433 Figure 1. Localization of the 28S ribosomal RNA gene (28S rDNA) of the eastern oyster by fluorescence in situ hybridization. Metaphase chromosomes were prepared from cells of la) embryos and (bl and (c) gill. Arrowheads point to the location of 28S rDNA; bars = 10 uni. match exactly with the location of nucleolus organizer regions (NOR). These results indicate that there is a single NOR-bearing chro- mosome pair in this species. In a previous study, two NOR-bearing chromosomes with measurable differences in size and centromeric index were found in preparations of embryo cells of the eastern oyster by use of silver staining (our unpublished data). Also, the number of the NOR sites was found to be variable among different phases of the cell cycle. Asymmetric features have been observed between homologous chromosomes prepared from embryo cells (Zhang et al. 1999b). Therefore, it was worthwhile to investigate chromosomes prepared from other tissue types for clarification of the NOR sites in embryo cells and to develop methods for physical genome mapping in this valuable species. In the present study, analysis of the location of the 28S rDNA on gonadal chromosomes and the transcriptional activity of the 28S rDNA among different tissues indicates that the two NOR-bearing chromosomes observed in embryo cells are homologous and pair during meiosis. Variation in chromosome morphology among closely related oyster species has been used to investigate evolutionary relation- ships (Thiriot-Quievreux and Insua 1992). However, our studies demonstrate that variation in morphology of homologous chromo- somes, such as in the pair of NOR-bearing chromosomes, can exist in the eastern oyster, especially in chromosomes obtained from cells of embryos. Centromeric position can change when pericen- tric inversions involve different lengths of the chromosomal seg- ments on each side of the centromere. This mechanism has been proposed to explain most of the chromosomal variation within oysters of the genus Crassostrea (Landron De Guevara et al. 1996) and could be used to interpret the difference of centromeric index between the two homologous 28S rDNA-bearing chromosomes found in this study, although it is likely not appropriate. The small difference we observed in the homologs of chromo- somes would be at the limit of detection when using basic proce- dures (Longwell and Stiles 1996). In catfish chromosomes, we have reported that a computer-assisted image system can routinely identify a size difference of -0.2% in relative length, which is equivalent to -0.1 |xm under the microscope (Zhang and Tierseh 1998). The size difference we observed between the two 28S rDNA-bearing chromosomes of the eastern oyster was -0.4% in prophase, prometaphase, and early metaphase of embryo chromo- somes, but the difference was reduced in late metaphase and was not distinguishable in highly contracted somatic chromosome spreads. Explanations for the size differences between chromo- somes include differential activity of the homologs. which may be useful for studying early development and gene expression in oys- ters. The results of the present study provide valuable information for physical genome mapping in mollusks. Probe DNA was syn- thesized by polymerase chain reaction and required only DNA sequence information from two primer regions, eliminating time and labor-intensive cloning and screening procedures. Primers pre- pared for physical mapping of oyster genes could be derived in some cases from mammalian species because of evolutionary con- servatism, which is especially beneficial in such species as the eastern oyster that lack DNA sequence information. Figure 2. Relationship between the 28S rDNA locus and the nucleolus organizer region of the eastern oyster. The same chromosome spread prepared from embryonic cells was subjected to chromomycin A3 staining (a), followed by staining with silver nitrate (b). Arrowheads indicate location of 28S rDNA and NOR; bars = 10 urn. 434 Zhang et al. Figure 3. Location and activity of the 28S rDNA on meiotic chromosomes of the eastern oyster. Diakinesis (a, b, and c) and pachytene chromosome bivalents (d and et were prepared from ripe gonad tissue. Fluorescence in situ hybridization (a and b), chromomycin A3 staining (c and d) and silver staining techniques (e) were used to analyze the chromosomal location of this gene (arrowheads); bars = 10 um. The 28S rDNA. because of its association with the NOR, can be informative for verifying such specific hybridization techniques as FISH and in situ PCR and provides an internal positive control for mapping studies. This is useful for such economically impor- tant species as the eastern oyster, which are poorly characterized at the chromosome level. Indeed, given the intrinsic difficulties of cytogenetic analysis in oysters, the association between NOR and Histogram * 28S rDNA secondary FISH AgNOR CMA Idiogram Figure 4. Analysis of the chromosomal location of the 28S rDNA of the eastern oyster by microdensitometry. Arrowheads indicate location of 28S rDNA, revealed by fluorescence in situ hybridization (FISH), sil- ver staining (AgNOR) and chromomycin A3 staining (CMA). 28S rDNA may be the only gene-level marker available at present to test physical mapping techniques, although the associations of such other genetic markers as microsatellite loci may prove ben- eficial in this regard (e.g., McGoldrick 1997). The techniques de- veloped in this study could be applied for identification of chro- mosomal integration of genetic material foreign to the genome of the eastern oyster, such as that in transgenic studies (Zhang et al. 1998). Localization of the 28S rDNA represents the first report of physical mapping of a gene in mollusks. More importantly, results of this study provide methodology for examining the validity of hybridization techniques. This is especially important in physical mapping of species without pre-existing genetic information, such as mollusks. Large-scale mapping studies in oysters await im- provements of techniques such as probe labeling to increase de- tection efficiency. ACKNOWLEDGMENTS This study was supported in part by the Louisiana Sea Grant Program, the USDA special grant program and the Louisiana Cat- fish Promotion and Research Board. We thank J. Buchannan, A. Pani. C. Paniagua, and B. Smith for technical assistance and J. Supan for providing oysters. This manuscript was approved by the Director of the Louisiana Agricultural Experiment Station as manuscript number 99-66-0256. FISH in Oysters 435 LITERATURE CITED Amemiya. C. T. & J. R. Gold. 1987. Chromomycin A3 staining of verte- brate chromosomes: enhancement of banding patterns by NaOH. Cy- tobios 49:147-152. Chen, T. R. 1994. Fluorescence in situ hybridization (FISH): detection of biotin- and digoxigenin-labeled signals on chromosomes. J. Tiss. Cut. Method 16:39-47. Clabby. C. U. Goswami, F. Flavin, N. P. Wilkins, J. A. Houghton & R. Powell. 1996. Cloning, characterization and chromosomal location of a satellite DNA from the Pacific oyster, Crassostrea gigas. Gene 168: 205-209. Guo, X. & S. K. Allen, Jr. 1997. Fluorescence in situ hybridization of vertebrate telomere sequence to chromosome ends of the Pacific oyster. Crassostrea gigas Thunberg. J. Shellfish Res. 16:87-89. Howell, W. M. & D. A. Black. 1980. Controlled silver staining of nucleo- lus organizer regions with a protective colloidal developer: a one step method. Experientia 36:1014-1015. Landron De Guevara, B.. F. Winkler, F. Rodriguez-Romero & C. Palma- Rojas. 1996. Comparative karyology of four American oyster species. Veliger 39:260-266. Littlewood, D. T. J. 1994. Molecular phylogenetics of cupped oysters based on partial 28S rRNA gene sequences. Mol. Phylogen. Evol. 3: 221-229. Long, E. O. & I. B. Dawid. 1980. Repeated Genes in Eukaryotes. Arm. Rev. Biochem. 49:727-764. Longwell. A. C. & S. S. Stiles. 1996. Chromosomes, biology, and breed- ing, in the Eastern Oyster: Crassostrea virginica. Chapter 12. In: V. S. Kennedy. R. I. E. Newell, and A. F. Eble (eds.). Maryland Sea Grant College. College Park, MD. Louisiana Summary of Agricultural and Natural Resources. 1997. Louisi- ana State University Agricultural Center. Louisiana Cooperative Ex- tension Service. Publ. 2382. McGoldrick. D. J. 1997. An experimental investigation of the genetic basis of hybrid vigor in the Pacific oyster, Crassostrea gigas (Thunberg). Ph.D. thesis. University of California, Davis. Paniagua-Chavez, C. G.. J. T. Buchanan. J. E. Supan & T. R. Tiersch. 1998. Settlement and growth of eastern oysters produced from cryo- preserved larvae. Cryo-Letters 19:283-292. Pendas, A. M., P. Moran & E. Garcia-Vazquez. 1993. Ribosomal RNA genes are interspersed throughout a heterochromatic chromosome arm in Atlantic salmon. Cytogenet. Cell Genet. 63:128-130. Thiriot-Quievreux, C. & A. Insua. 1992. Nucleolus organizer region varia- tion in the chromosomes of three oyster species. J. Exper. Mar. Biol. Ecol. 157:33-40. Xu. Z.. J. C. Pierce & X. Guo. 1999. Chromosome assignment of rDNA in the American and Pacific oyster by fluorescence in situ hybridization. Plant and Animal Genome Meeting VII. San Diego. CA. Book of Abstracts, p. 110. Zhang, Q., G. Yu. R. K. Cooper & T. R. Tiersch, 1999a. Localization of the 28S ribosomal RNA gene on chromosomes prepared from different tissues of the eastern oyster Crassostrea virginica. Annual Meeting ot the United States Chapter of the World Aquaculture Society, Tampa. FL. Book of Abstracts, p. 19. Zhang. Q.. G. Yu. R. K. Cooper & T. R. Tiersch. 1999b. High-resolution analysis of karyotypes prepared from different tissues of the eastern oyster {Crassostrea virginica). J. Shellfish Res. 18:115-120. Zhang, Q.. T. R. Tiersch & R. K. Cooper. 1998. Inducible expression of green fluorescent protein within channel catfish cells by a cecropm gene promoter. Gene 216:207-213. Zhang, Q., R. K. Cooper & T. R. Tiersch. 1997. Detection of a single-locus gene on channel catfish chromosomes by in situ polymerase chain reaction. Comp. Biochem. Physiol. 1 18B:793-796. Zhang, Q. & T. R. Tiersch. 1998. Identification and analysis of weak linear banding patterns of fish chromosomes with a densitometric method. BioTechniques 24:996-997. Journal of Shellfish Research, Vol. 18. No. 2. 437-444. 1999. IMPACTS OF SUSPENDED PEAT PARTICEES ON FEEDING AND ABSORPTION RATES IN CULTURED EASTERN OYSTERS (CRASSOSTREA VIRGINICA, GMELIN) K. B. STRYCHAR* AND B. A. MACDONALD Department of Biology and Centre for Coastal Studies and Aquaculture University of New Brunswick Box 5050, Saint John New Brunswick, E2L 4L5, Canada ABSTRACT A large portion of the oyster industry in eastern Canada is located in the northeastern part of New Brunswick. Peat mining industries exist in the same region and concern exists over the release of particulate matter and its potential impact on oyster feeding, growth, and survival. We examined the feeding response and absorption rates of several groups of Crassostrea virginica exposed to natural seston, cultured microalgae, and various concentrations of suspended peat (2 to 20 mg L"1 ). Clearance rates declined as the concentration of peat increased above 5 mg L_1, but relatively little of the peat cleared was rejected as pseudofeces (<5%); thereby, ingestion rates were maintained as concentrations increased. Absorption efficiency was unaffected by low concentrations of peat (<5 mg L_1), although, higher concentrations (>5 mg L~') produced negative absorption and metabolic fecal loss. We conclude that the presence of suspended peat will dilute the natural seston and negatively impact energy gain and growth in the oysters, because they indiscriminately ingest the peat; thereby, filling their guts with material that cannot readily be digested. The concentrations of peat where this dilution effect starts to affect the oysters nutrition adversely will vary with the concentration of the natural seston. KEY WORDS: oyster. Crassostrea, feeding, peat impacts, absorption efficiency, metabolic fecal loss INTRODUCTION The northernmost distribution of natural populations and com- mercial harvesting of the eastern oyster Crassostrea virginica (Gmelin 1791) occurs along Canada's eastern coastline. Large populations of the eastern oyster exist in Caraquet Bay (47°50'N. 65°W). New Brunswick and Bedeque Bay (46°22'N, 63°50"W). Prince Edward Island, with smaller populations found in Nova Scotia (Medcof 1961. Sephton and Bryan 1989. Lavoie 1995). Natural eastern populations of oysters have been greatly reduced in size over the past several decades by Malpeque disease, industrial and agricultural encroachment, mining, construction, forestry, and fishing activities (Medcof 1961, Chiasson 1991 ). Today, the oyster industry is based on both private natural leases and aquaculture operations with an estimated value of $7.5 million in eastern Canada in 1994 (Boghen 1995, Sephton 1997). Concerns have been expressed by the oyster leaseholders in New Brunswick that increasing harvesting of peatlands may jeopardize the growth and survival of oysters because of increased turbidity in local bays and estuaries (Sephton and Bryan 1989). In addition to the natural erosion of peatlands, peat particles enter the adjacent shallow ma- rine ecosystems from activities associated with peat harvesting, including primarily airborne transport of fine particles and runoff from the mounds of peat drying in the fields (Lavoie 1995). Many of the oyster-producing leases held in New Brunswick are in close proximity to large peat deposits (Fig. 1 ). Peat harvests continue to escalate although environmental im- pacts of harvesting peat are not well known (Glooschenko 1990). It has been suggested that drainage waters from developed peat- lands decrease light penetration, change water color through leach- ing, and increase suspended particulate matter that could be det- rimental to downstream aquatic organisms (Dunson and Martin 1973. Mitchell and McDonald 1992, Pattinson et al. 1994). Sub- *Current address: Department of Arts, Health, and Sciences, School of Biology and Environmental Sciences, Central Queensland University. North Rockhampton. Queensland, 4702 Australia. stantial increases in turbidity and sedimentation may have obvious impacts on benthic organisms by smothering and killing them. However, it is difficult to assess more subtle sublethal effects of smaller increases in particle concentration, including any reduction in light for phytoplankton production or interference with feeding activity and possible reduction in growth of suspension-feeders (Peterson 1985, Cloern 1987, Grant et al. 1990, Grant and Thorpe 1991 ). Environmental awareness has led to the installation of sedi- mentation ponds and stringent guidelines restricting the discharge of peat particles into the environment to <20 mg L"1 (Gemtech Ltd. 1991, Gemtech Ltd. 1993). Although concern regarding com- mercial harvesting of peatlands and the impact on receiving waters is increasing, there is little quantitative evidence of the possible effects on local benthic populations. Increases in particulate matter in the water column could po- tentially be a problem to suspension-feeding invertebrates espe- cially if the additional particles are poor in nutritional quality, such as inorganic sediments associated with natural resuspension events and anthropogenic activities including, agricultural runoff, dredg- ing, and dumping. In many cases, these additional poor quality particles are thought to dilute the higher quality background seston, consisting of higher proportions of phytoplankton and other nutritious organic particles, possibly interfering with feeding ac- tivity and reducing potential energy gain (Widdows et al. 1979, Berg and Newell 1986, Smaal et al. 1986). Many species of bi- valves, including C. virginica, reduce the potential impact of seston dilution by rejecting poor quality particles before ingestion through the production of pseudofeces (Kiorboe and Mehlenberg 1981. Newell and Jordan 1983. MacDonald and Ward 1994. Hawkins et al. 1996). C. virginica is known to thrive in estuaries where natural concentrations of seston range between 6.0 and 30.0 mg L~ , is very tolerant of high concentrations of suspended sedi- ments, and has been shown to maintain high clearance rates at elevated seston concentrations up to 75 mg L_1 (Berg and Newell 1986. Newell and Langdon 1996, Shumway 1996). The purpose of this study was to assess the potential impact of suspended peat particles on the ability of C. virginica to capture 437 438 Strychar and MacDonald Figure 1. Distribution of peatlands in the province of New Brunswick. Principle areas for producing oysters {Crassostrea rirginica) in the province are indicated as black circles. food particles and gain energy to support growth. Several groups of oysters were exposed to various concentrations of suspended peat, natural seston, and cultured microalgae while we measured clearance and ingestion rates, pseudofeces production, and absorp- tion efficiency. We tested the hypothesis that exposure to concen- trations of peat up to 20 mg L-' would have no impact on feeding and energy gain in the oyster. MATERIALS AND METHODS General Procedures Crassostrea virginica (-70 mm in shell length) were collected approximately every 2 to 4 days from an intertidal site in Oyster Point, Caraquet Bay, New Brunswick over a 3-wk period in June 1996 (Fig. 1 ). This area was selected, because eastern oysters are harvested commercially from this site, and it receives some runoff from channels draining local peatlands. Oysters were maintained in flowing filtered seawater ( 14 ± 1 °C SE; 25%) at the Marine Science Centre in Shippagan. NB for 2 to 4 days until experiments began. All epibiota were removed from the oyster shell immedi- ately, and the length of the oyster was recorded to the nearest 0. 1 mm using vernier calipers. Experimental Diets and Apparatus Ten groups of randomly selected oysters (;i = 9) were exposed to different experimental diets consisting of natural seston. filtered seawater. filtered seawater supplemented with microalgae (Iso- chrysis galbana, T-iso), and filtered seawater supplemented with various concentrations of suspended peat particles (2, 5. 10, and 20 mg LT1 ). The natural seston diet was supplied by a direct unfiltered seawater line submerged below the water surface in Shippagan Harbour. Filtered seawater was passed through a sand filter to remove particles greater than >10 p.m and stored in large under- ground tanks (454.000 L) to minimize variation in the experimen- tal diets caused by fluctuations in the background water over the 3-wk experimental period. In this way, any differences observed in the feeding activity may be attributed to the suspended particles rather than to any possible changes in the background water. /. galbana was grown in autoclaved ill medium, at 20 °C, under constant illumination and harvested for the diet when they were approximately at their exponential growth phase. Samples of local peat were dried at 60 °C for 24 h. ground into a fine powder with an ultrafine mill (Retschvibratory Mill, type MM-2), and sieved with stainless steel sieves to a particle size of less than <63 pm. The peat powder was then suspended in seawater. and a peristaltic pump added the desired concentration during the process of accli- mating the oysters ( 10 h) and throughout the experimental period. To determine whether oysters were influenced either by the peat fiber or the peat fiber and any associated microbial assemblages, two groups were also exposed to experimental diets (2 and 5 mg L_I ) containing autoclaved peat. Peat was autoclaved at 20 psi and 125.7 ± 0.6 °C SE for 15 min and passed through a 63-u.m sieve. The experimental apparatus used to measure feeding rates con- sisted of a 20-L header tank/mixing chamber supplied with natural seston or filtered seawater from the large storage tanks. Seawater from the header tank was gravity fed via Tygon tubing to a series of 1 1 2-litre plastic holding chambers. Flow rates were held con- stant (220-280 mL/min"') by maintaining head pressure and by inserting plastic flow-restricting plugs in each of the 1 1 delivery lines. Water exited the holding chambers through upright plastic standpipes fitted through the bottom of the chamber. A plastic baffle was inserted at the front of each chamber ensuring adequate mixing of the particles, and the chamber's oblong shape reduced recirculation of the experimental diet. A slurry of suspended peat particles was pumped to the header tank/mixing chamber using a peristaltic pump at a rate calculated to supply the desired experi- mental concentration. Submersible stirrers and air stones were used to mix the experimental diet in the header tank and prevent settling. Oysters were placed in the holding chambers and exposed to the experimental diet for 10 h before measurements began to acclimate them to the diet and purge the digestive system. All chambers were cleaned of biodeposits and any settled particles before the beginning of the experimental period. Nine of the eleven chambers contained individual oysters. The concentration and size distribution of the particles in the experi- mental diet was determined by sampling the outflowing water from the two remaining "empty reference" chambers. One of the two reference chambers contained only a pair of oyster shells to correct for any possible settlement of particles at the highest ex- perimental concentrations. Water samples were simultaneously collected from the standpipe drain of each chamber every 2 to 3 h over the 10- to 12-h experimental period. Water collection was done over a recorded time interval to allow calculation of flow rates for each chamber. The concentration of particles in the out- flowing water was determined using an electronic particle counter (Coulter Multisizer) equipped with a lOO-p-m aperture. Differences in particle concentration between the reference chamber contain- ing a shell and the other holding chambers was used to calculate clearance rates (see below). Analysis of Seston and Biodeposits Four replicate water samples were collected at the beginning, middle, and end of each experiment to verify the particle concen- tration (mg L"' ) and percentage organic content of the experimen- tal diets. Samples of each diet mixture were collected by filtering a known volume of water through a pre-ashed, pre-weighed 2.4- cm GF/C filter under low vacuum. To remove any salts, all filters were washed with 10 mL isotonic ammonium formate before dry- ing at a constant weight at 80 °C and then frozen at -25 °C and stored for later analysis. The percentage organics in each diet was determined by combusting them at 475 °C for 24 h in a muffle Impact of Peat Particles on C. virginica 439 furnace and reweighing them (±0.01 mg) after cooling in a desic- cator. Samples of feces were collected with a micropipette from each holding chamber at the end of the experimental period. The per- centage organic content of feces was determined using the same collection, treatment, and weighing methods as those used for the water samples. All biodeposits were removed from each holding chamber and discarded once during the experimental period to facilitate the quantitative collection of all pseudofeces produced during a subsequent recorded time interval. Subsamples of the pseudofeces were collected with a micropipette and placed in vials with 20 mL of filtered seawater, treated with several drops of Coulter Electronics type 1C dispersant and shaken vigorously for 5 to 10 min to disperse the particles so they could be counted using the Multisizer (Ward et al. 1992). If coincidence was high (>10%), samples were diluted with filtered seawater and recounted. The total volume of the sample was determined before counting so we could calculate the total number of particles above the background concentration expelled as pseudofeces. The weight of the pseu- dofeces were determined using the same collection, treatment, and weighing methods as those used for the water and feces samples. Measurements of Standardized Clearance Rates Clearance rate (CR), one indicator of feeding activity, is de- fined as the volume of water cleared of suspended particles >2 u.m in diameter per unit time (Bayne et al. 1977). Differences in the concentration of particles between the reference chamber and the holding chamber containing an oyster were used to calculate CR as follows: CR = FR x (C, -C2)/C,, where FR is the flow rate of water through the experimental cham- bers (L fr' ). C, is the particle concentration (number mL~' ) in the inflowing seawater (determined from the reference chamber), and C, is the particle concentration (number mL"1) in the outflowing water from a holding chamber after it has been processed by the oyster. The CR (L h"1 ) for each oyster was the minimum of three consistent rates measured over the experimental period. Although all the oysters were similar length (-70 mm), it was necessary to correct for any weight differences among the oysters to compare CR and other indicators of feeding activity between the groups of oysters exposed to different experimental diets. This was done by standardizing all rates (CRS) to an oyster with 1.0-g body dry weight using the following equation: CR. where CRS Ws = the standard weight of the oyster, W0 IRS = (CRS x TPM) - PS. Absorption efficiency (AE) was calculated using the Conover ( 1966) ratio, which compares the organic content of the food and feces as follows: (Ws/W0)b x CR0 the clearance rate for an oyster of standard weight. the observed weight of the oyster, CR0 = the observed or measured clearance rate, b = -0.70, the weight exponent for oyster clearance rate (Newell and Langdon 1996). At the end of each experiment, the soft body tissues were removed from the shell and dried to a constant weight (±0.01 g) at 90 °C. Ingestion and Absorption Rates Ingestion rates (mg h_1) standardized for a 1.0-g oyster (IRS) were calculated as the product of CRS (L h"1) and particle con- centration or total particulate material (TPM; mg L_1) after sub- tracting particles rejected as pseudofeces (PS: mg h"1) using the following equation: AE (F- E) x 100/(1 -E)xF where F = ash-free dry weight: dry weight ratio in the food or experimental diet, and E = ash-free dry weight:dry weight ratio in the egesta or feces. Absorption rates (mg organics h~') standard- ized for a 1.0-g oyster (ARS) were calculated as the product of IRS (mg fr') and absorption efficiency (%). Statistical Analysis All statistical procedures were carried out using the General Linear Model (GLM) procedure of the Statistical Analysis System (SAS Institute Inc.). One-way analyses of variance (ANOVA) were used to determine if oyster clearance rates, ingestion rates, absorption efficiencies, and absorption rates were significantly dif- ferent for the various diet treatments. The null hypothesis for these experiments was that the addition of peat would have no effect on oyster feeding activities. For all comparisons, a significance level of alpha = 0.05 was used. If significant F values were observed in any ANOVA tests, a Student-Newman-Keuls test was conducted to determine which means were significantly different from one another. Before statistical analyses, data were tested for normality and homogeneity of variance using a Shapiro-Wilk (W) test (Zar 1984) and an F-max test (Sokal and Rohlf 1981), respectively. If one of the above ANOVA requirements was not met, the data were log,,, transformed, and the ANOVA was repeated. If the assump- tions of the ANOVA were still not met after transformation, a nonparametric Kruskal-Wallis test was used. A paired t-test using a repeated measures design (Zar 1984) was performed to determine whether feeding activities would dif- fer when a group of oysters was exposed to the same concentra- tions of natural and autoclaved peat. The null hypothesis was that the presence of the living microbial community would have no impact on feeding activity in oysters. RESULTS Particle concentrations measured for each diet during the ex- perimental period with the targeted concentrations are presented in Table 1. There were significant differences in clearance rates (CR) between the 10 experimental groups of oysters (F = 15.62, df = 9. P < .001; Table 2). Clearance rates on the nonpeat diets, filtered water, natural seston. and algae (T-Iso). ranged from 2.5 to 3.0 L IT1 and were not significantly different (Fig. 2). Relatively low concentrations of peat (2 mg L"1) did not significantly reduce clearance in C. virginica. making it comparable to values observed for the nonpeat diets. Concentrations of suspended peat particles >2 mg L"1 reduce CR significantly. Generally, CR decreased as concentrations of peat increased to between 2 and 20 mg L"1, with values dropping from about 2.25 to about 0.50 L h"1 g"' (Fig. 2). Ingestion rate is a better indicator of feeding activity than CR, because it factors out the particles filtered from suspension but rejected as pseudofeces [e.g.. IR = (CR x TPM) - PS]. There were no significant differences in percentage of particles rejected as pseudofeces among the groups of oysters (F = 1.88, df = 9, P < .066; Table 2). The percentage of particles rejected as pseudofeces was relatively low, typically ranging from 1 to 2% by number for the nonpeat algal and natural seston diets (Fig. 3). Although a 440 Strychar and MacDonald TABLE 1. Target diet concentrations (mg L~') and concentrations actuall) obtained during the experiments. 3.5-1 Treatment Target Concentration (mg L"'l Actual Concentration (mg L 'l ±SD Algae (T-iso) 1.5 Natural seston 1.7 Filtered seawater 0.66 Peat 2 Peat 2* Peat 5 Peat 5+ Peat 5*+ Peat 10 Peat 20 1.51 ± 0.09 1.65 ±0.22 0.66 ± 0.46 2.15 ±0.19 2.10 ±0.05 5.18 ±0.20 4.17 ±0.16 4.92 ±0.14 10.23 + 0.15 21.28+ 1.21 * Autoclaved peat. + Same oysters used in a repeated measures design. higher percentage of the particles may be rejected for the peat diets, percentage rejection values of less than 5% indicate that the majority (95%) of the peat cleared from suspension is ingested. There were significant differences in ingestion rates for oysters exposed to the 10 experimental diets (F = 10.39. df = 9. P<.O0l; Table 2). With the exception of the 20 mg L~' peat diet and the filtered seawater diets (-1.5 mg h"'), ingestion rates ranged be- tween 2 and 5 mg h~' (Fig. 4). Overall, ingestion rates did not increase with increasing concentration, because CR continued to decrease. The exception to this trend was observed for oysters exposed to 20 mg L~' of peat that maintained higher CR values that were comparable to oysters exposed to lower concentrations (Fig. 2). We do not have sufficient evidence to reject the hypoth- esis that exposure to various concentrations of peat will have no impact on feeding activity in the oyster. There were significant differences in the oysters' ability to extract organic material (absorption efficiency) from the various experimental diets (F = 5.81, df = 9, P < .001: Table 2). Ab- sorption efficiencies (AE) were highest on the nonpeat diets but decreased as peat concentrations increased up to about 5 mg L"1 until negative values were observed at 10 and 20 mg L_l (Fig. 5). The only significant differences in AE were found between the algae and the 5, 10. and 20 mg L_1 groups. The organic content of carbon and nitrogen of peat in this study was 53.65% + 1.86 SD and 1 .76 ± 0.68 SD (respectively) and is comparable to previously published values (Table 3). The values recorded for the nonpeat diets are also comparable to published reports for bivalves in gen- TABLE 2. Summary of nonparametrie one-way ANOVA procedure for testing clearance and ingestion rate, absorption efficiency, absorption rate. and rejection rates (pseudofeces) of Crassostrea virginica exposed to various concentrations of suspended peat particles (alpha = 0.05). Response Variable F-Value df Error Pr>F Clearance rate 15.62 9 80 0.001 Ingestion rate 10.39 9 80 0.001 Absorption efficiency 5.81 9 39 0.001 Absorption rate 59.87 9 80 0.001 Pseudofeces [% rejection) 1.88 9 80 0.066 _otj 06 . 50 O c 25 5 mg L_1 ) did reduce absorption efficiency until negative values were observed. Negative values for AE have been observed before and have been described as metabolic fecal loss (Hawkins et al. 1983). Possible explanations for negative absorp- tion efficiencies include, ingestion of diets with very low organic content (Hawkins and Bayne 1985) or feeding on diets with high organic content but short retention times in the gut, reducing en- zymatic degradation of ingested food (Bricelj and Malouf 1984, Bayne et al. 1987). It is unlikely that gut retention times were reduced in these oysters, because clearance was reduced, and in- gestion remained relatively constant. Peat particles, because of their composition, may simply be more resistant to degradation in the digestive system than typical particles in the seston. Willows (1992) suggests that bivalves exposed to poor diets after short acclimation times could experience negative AEs; however, after longer acclimation periods. AE could become positive as gut resi- dence times increase. In this experiment, we exposed the oysters to the experimental diets for 10 h before initiating the experiment, but this may not be long enough to adjust gut residence times and absorb the peat, or perhaps the peat was simply too refractory. Negative AE seen in this study at high peat concentrations could be related to the absorptive quality of the peat. Peat consists of large porous hyaline cells that help this fibrous material absorb large quantities of water and nutrients (Warner 1992). It is possible that high concentrations of ingested peat will absorb organic ma- terial within the gut and. thereby, increase the organic content of the feces expelled. Bayne et al. ( 1987) suggested that the reasons for negative AEs could be the presence of exocytosed mucus ma- terial from the digestive cells or the adsorption of nitrogenous compounds to the fecal strip. It has long been thought that nonalgal particles, such as bacteria or plant detritus, suspended in the water column could represent an important supplementary food source for suspension-feeding bi- valves when concentrations of phytoplankton are low. This will depend upon the capabilities of the different species to accumulate bacteria and digest detritus. C. virginica has been shown to have a limited ability to absorb carbon from refractory cellulosic material (-3%); however, in the presence of cellulolytic bacteria, the effi- ciency of absorption was increased to 10% (Langdon and Newell 1990). These authors concluded that oysters may supplement their nitrogen requirements during the summer months through the in- gestion of bacteria, but the use of detritus and bacteria does not meet the carbon and nitrogen requirements of this species. Our results are somewhat similar when seston concentrations are very low (e.g., filtered water <0.7 mg L_I ). The addition of 2 mg L_l of peat actually increased the ingestion and absorption rate signifi- cantly above the rate for oysters fed on filtered water alone (Figs. 4 and 6). This observation suggests that peat may introduce a compliment of microbes favored by the oysters, the sand filter may not have efficiently removed all natural background seston (fil- tered seawater versus natural seston is approximately 60%. Table 1 ), and/or the peat fiber may have contributed additional organic Impact of Peat Particles on C. virginica 443 nutrients (e.g., carbon and nitrogen. Table 3). We also found that C. virginica does not have the ability to reduce the organic content of peat ingested at higher concentrations (>5 mg L"1). but the presence of a living microbial community increased the absorption rate in oysters above the rates observed for oysters fed autoclaved peat (Fig. 6). Microbes isolated within dry peat included 28 dif- ferent genera of bacteria and fungi, with Bacillus sp., Nitrosomo- nus sp.. Nitrobacter sp.. and Clostridium sp. the most common bacteria and Penicillium sp.. Fusidium sp., Aspergillus sp.. and Paecilomyces sp. the most abundant fungi (Strychar and Johnson, submitted). In summary. C. virginica clears and ingests peat particles at similar rates to those observed for microalgae and other suspended particles but is incapable of efficiently absorbing peat, resulting in metabolic fecal loss. C. virginica has previously been shown to reduce the impact of seston dilution by poorer particles, before ingestion, through the rejection of proportionately more of these particles in pseudofeces (Newell and Jordan 1983). However, un- less the pseudofeces are produced in quantities sufficient enough to significantly alter the quantity and quality of the ingested ration, it is not likely to be ecologically meaningful (Newell and Jordan 1983, Iglesais et al. 1992. MacDonald and Ward 1994). In this study, oysters produced very small amounts of pseudofeces, re- sulting in virtually indiscriminate ingestion and, consequently, a dilution of the ingested ration by inert peat particles. The ratio between suspended peat particles and the concentration of back- ground seston may be a more useful indicator of the potential impact of peat on oysters than simply using an absolute concen- tration of peat. For example, the addition of 2 mg L"1 of peat to very low background seston concentrations (<1.0 mg L~'l may have a neutral or positive impact; whereas, the addition of 5.0 mg L_I at the same time of year may have a very negative effect. However, if background seston concentrations were higher (5-10 mg L"1), the addition of 5.0 mg L_i or even higher concentrations of peat may not have adverse effects. ACKNOWLEDGMENTS We thank Drs. A. Boghen. J.-Y. Daigle, and T. Sephton for helpful comments on the manuscript and W. Morris for technical assistance. We gratefully acknowledge the assistance of: J. Thi- bault, N.B. Department of Natural Resources and Energy; R. Roiux, P. Cormier. S. Dioron, J. Mallet. N. Robichaud, and other staff members for the excellent facilities provided by the N.B. Department of Fisheries and Aquaculture. and Marine Science Centre. Shippagan; and the support of J.-Y. Daigle from the Peat Research and Development Centre, Shippagan. N.B. This work was supported by funds from a University of New Brunswick Environmental Research Grant and a Natural Sciences and Engi- neering Research Council of Canada research grant. We appreciate their support. This is contribution number 49 of the Centre for Coastal Studies and Aquaculture, University of New Brunswick in Saint John, NB. LITERATURE CITED Allison. F. E. 1973. Soil organic matter and its role in crop production. Elsevier. New York. 631 pp. Bacon. G. S.. B. A. MacDonald & J. E. Ward. 1998. Physiological re- sponses of infaunal {Mya arenaria) and epifaunal (Placopecton ma- gellanicus) bivalves to variations in the concentration and quality of suspended particles. I. feeding activity and selection. J. Exp. Mar. Biol. Ecol. 219:105-125. Bayne. B. L. & R. C. Newell. 1983. Physiological energetics of marine mollusks. pp. 407-515. In A. S. M. Saleuddin and K. M. Wilbur (eds.). The Mollusca. Physiology. Part I. Academic Press. San Diego. Bayne. B. L., J. Widdows & R. I. E. Newell. 1977. Physiological measure- ments on estuarine bivalve mollusks in the field, pp. 57-68. In B. K. Keegan, P. O'Ceidigh and P.J. S. Boaden (eds.). Biology of Benthic Organisms. Pergamon Press, Oxford. Bayne. B. L., A. J. S. Hawkins & E. Navarro. 1987. Feeding and digestion by the mussel Mytilus edulis L. (Bivalvia: Mollusca) in mixtures of silt and algal cells at low concentrations. J. Exp. Mar. Biol. Ecol. 1 1 1 : 1-22. Bayne, B. L., J. I. P. Iglesias, A. J. S. Navarro. E. M. Heral & J.-M. Deslous-Paoli. 1993. Feeding behaviour of the mussel. Mytilus edulis: responses to variations in quantity and organic content of the seston. J. Mar. Biol. Assoc. UK 73:813-829. Berg. J. A. & R. I. E. Newell. 1986. Temporal and spatial variations in the composition of seston available to the suspension feeder Crassostrea virginica. Est. Coastal Shelf Sci. 23:375-386. Boghen. A. D. 1995. Introduction: the state of aquaculture in Atlantic Canada, pp. 1-31. In A. D. Boghen (eds.). Cold Water Aquaculture in Atlantic Canada. Canadian Institute for Research on Regional Devel- opment. Moncton, NB, Canada. 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. Chiasson, R. 1991. Malpeque oysters: trying to best a 100- year-old record. World Aquacult. 22:36-41. Cloern. J. E. 1987. Turbidity as a control of phytoplankton biomass and productivity in estuaries. Com. Shelf Res. 7:1367-1381. Conover. R. J. 1966. Assimilation of organic matter by zooplankton. Lim- nol. Oceanog. 11:338-345. Dunson. W. A. & R. R. Martin. 1973. Survival of brook trout in bog- derived acidity gradient. Ecology 54:1370-1376. Galtsoff, P. S. 1964. The American oyster Crassostrea virginica (Gmelin). Fish. Bull. 64:1-480. Gemtech Ltd. 1991. Peatland hydrology. Bog 569, Pigeon Hill. New Brunswick. New Brunswick Department of Natural Resources and En- ergy, Open File Rept. 91-10. 91 pp. Gemtech Ltd. 1993. Design, installation, and monitoring of siltation ponds, peat bog 567, Lemeque Island. NB. New Brunswick Department of Natural Resources and Energy, Mineral Resources. Open File Rept. 93-94. Glooschenko, W. A. 1990. Effect of peatland on water quality, fish and wildlife habitat in Canada, a review. Tasks Vegetal. Sci. 25:85-93. Grant. J. & B. Thorpe. 1991. Effects of suspended sediment on growth, respiration, and excretion of the soft-shell clam (Mya arenaria). Can. J. Fish. Aquat. Sci. 48:1285-1292. Grant, J., C. T. Enright & A. Griswold. 1990. Resuspension and growth of Ostrea edulis: a field experiment. Mar. Biol. 104:51-59. Haven, D. S. & R. Morales-Almo. 1966. Aspects of biodeposition by oys- ters and other invertebrate filter feeders. Limnol. Oceanog. 1 1 :487- 498. Hawkins, A. J. S. & B. L. Bayne. 1985. Seasonal variation in the relative utilization of carbon and nitrogen by the mussel Mytilus edulis: Bud- gets, conversion efficiencies, and maintenance requirements. Mar. Ecol. Prog. Ser. 25:181-188. Hawkins, A. J. S., B. L. Bayne & K. R. Clarke. 1983. Coordinated rhythms of digestion, absorption, and excretion in Mytilus edulis (Bivalvia: Mollusca). Mar. Biol. 74:41-48. Hawkins. A. J. S., R. F. M. Smith, B. L. Bayne & M. Heral. 1996. Novel observations underlying the fast growth of suspension-feeding shellfish in turbid environments: Mytilus edulis. Mar. Ecol. Prog. Ser. 131:179- 190. Iglesias, J. I. P.. E. Navarro. P. Alvarez Jorna & I. Amentia. 1992. Feed- ing, particle selection, and absorption in cockles Cerastoderma edula 444 Strychar and MacDonald (L.) exposed to variable conditions of food concentration and quality. J. Exp. Mar. Biol. Ecol. 162:177-198. Jordan. S.J. 1987. Sedimentation and remineralization associated with biodeposition by the American oyster Crassostrea virginica (Gmelinl. Doctoral dissertation. University of Maryland, College Park. MD. pp. 200. Kiorboe. T. & F. M0hlenberg. 1981. Particle selection in suspension- feeding bivalves. Mar. Ecol. Prog. Ser. 5:291-296. Langdon. C. J. & R. I. E. Newell. 1990. Utilization of detritus and bacteria as food sources by two bivalve suspension-feeders, the oyster Cras- sostrea virginica and the mussel Geukensia demissa. Mar. Ecol. Prog. Ser. 58:299-310. Langdon, C. J. & R. I. E. Newell. 1996. Digestion and nutrition in larvae and adults, pp. 231-269. In V. S. Kennedy, R. I. E. Newell and A. F. Eble (eds.). The Eastern Oyster Crassostrea virginica. Maryland Sea Grant College, College Park, MD. Lavoie, R. E. 1995. Culture of the American oyster Crassostrea virginica. pp. 189-224. In A. D. Boghen (ed.). Cold Water Aquaculture in At- lantic Canada. Canadian Institute for Research on Regional Develop- ment, Moncton, NB. MacDonald. B. A. & J. E. Ward. 1994. Variation in food quality and par- ticle selectivity in the sea scallop Placopecten magellanicus (Mollusca: Bivalvia). Mar. Ecol. Prog. Ser. 108:251-264. Medcof, J. C. 1961. Oyster farming in the maritimes. Fisheries Research Board, Canada Bulletin. 131. 158 pp. Mitchell. G. & M. T. McDonald. 1992. Discoloration of water by peat following induced drought and rainfall simulation. Water Res. 26:321- 326. Navarro, E.. J. I. P. Iglesias & M. M. Ortega. 1992. Natural sediment as a food source for the cockle Cerastodertna edule (L.): effect of variable particle concentration on feeding, digestion, and the scope for growth. ./. Exp. Mar. Biol. Ecol. 156:69-87. Navarro, E.. J. I. P. Iglesias. M. M. Ortega & X. Larretxea. 1994. The basis for a functional response to variable food quantity and quality in cock- les Cerastoderma edule (Bivalvia, Cardiidae). Physiolog. Zool. 67: 468-196. Nelson, T. C. 1938. The feeding mechanism of the oyster. I. on the pallium and branchial chambers of Ostrea virginica, O. edulis, and O. angulata. J. Morphol. 63:1-61. Newell. R. I. E. & S. J. Jordan. 1983. Preferential ingestion of organic material by the American oyster Crassostrea virginica. Mar. Ecol. Prog. Ser. 13:47-53. Newell, R. I.E. & C.J. Langdon. 1996. Mechanisms and physiology of larval and adult feeding, pp. 185-229. In V. S. Kennedy. R. I. E. Newell and A. F. Eble (eds.). The Eastern Oyster Crassostrea vir- ginica. Maryland Sea Grant College, College Park, MD. Paquette. J. 1997. Revised statistics of the mineral production of Canada, by province. Natural Resources Canada, Minerals and Metals Sector: Minerals and Mining Statistics Division, Economic and Financial Analysis Branch. Natural Resources Canada. 2 pp. Pattison. V. A., D. P. Butcher & J. C. Labadz. 1994. The management of water color in peatland catchments. J. Inst. Water Manage. 8:298-307. Peterson, C. H. 1985. Patterns of lagoon bivalve mortality after heavy sedimentation and their palaeoecological significance. Paleobiology 11:139-153. Riley, J. L. 1989. Laboratory methods for testing peat. Ontario Peatland Inventory Project: Ontario Geological Survey. Miscellaneous Paper 145. 51 pp. Schnitzer. M. & S. U. Khan. 1978. Soil organic matter. Elsevier, New York. 311 pp. Sephton. T. W. 1997. Gulf of St. Lawrence shellfish aquaculture: future development, coldwater aquaculture to the year 2000. Aquaculture As- sociation of Canada. Special Publ. 2. pp. 77-84. Sephton, T. W. & C. F. Bryan. 1989. Changes in the abundance and dis- tribution of the principal American oyster public fishing grounds in the southern Gulf of St. Lawrence. J. Shellfish Res. 8:375-385. Shumway. S. E. 1996. Natural environmental factors, pp. 467-513. In V. S. Kennedy, R. I. E. Newell and A. F. Eble (eds.). The Eastern Oyster Crassostrea virginica. Maryland Sea Grant College, College Park. MD. Smaal, A. C, J. G. H. Verhagen, J. Coosen & H. A. Haas. 1986. Interaction between seston quantity and benthic suspension feeders in the Oster- schelde. The Netherlands. Ophelia 26:385-399. Sokal. R. R. & F. J. Rohlf. 1981. Biometry. 2nd ed. W. H. Freeman and Co., New York. 859 pp. Strychar. K. B. & J. A. Johnson. Microbial assemblages isolated from peat- lands in Shippagan, New Brunswick, Canada. Can. J. Microbiol, (sub- mitted). Thiessen. R. 1925. The microscopical constitution of coal. pp. 244-305. In C. A. Ross and J. R. P. Ross (eds.). Geology of Coal. Hutchinson Ross. Visser, S. A. 1964. The presence of micro-organisms in various strata of deep tropical peat deposits. Life Sci. 3:1061-1065. Ward, J. E.. J. S. Levinton. S. E. Shumway & T. Cucci. 1998. Particle sorting in bivalves: in vivo determination of the pallial organs of se- lection. Mar. Biol. 131:283-292. Ward, J. E., H. K. Cassell & B. A. MacDonald. 1992. Chemoreception in the sea scallop Placopecten magellanicus (Gmelin): I. stimulatory ef- fects of phytoplankton metabolites on clearance and ingestion rates. J. Exp. Mar. Biol. Ecol. 163:235-250. Warner, B. G. 1992. Peat: nature's compost. Earth 1:44-^9. Widdows. J.. P. Fieth & CM. Worrall. 1979. Relationships between seston. available food, and feeding activity in the common mussel Mytilus edulis. Mar. Biol. 50:195-207. Willows, R. I. 1992. Optimal digestive investment: a model for filter feed- ers experiencing variable diets. Limnol. Oceanog. 37:829-847. Zar, J. H. 1984. Biostatistical analysis. Prentice Hall, Upper Saddle River, NJ. 718 pp. Journal of Shellfish Research, Vol. IS. No. 2. 445-+50. 1999. ASSESSMENT OF REPRODUCTIVE HEALTH IN THE WILD SEED OYSTERS, CRASSOSTREA GIGAS, FROM TWO LOCATIONS IN KOREA MI SEON PARK,1 HYUN JEONG LIM,2 QTAE JO,' JANG SANG YOO,1 AND MINJEE JEON1 1 Department of Aquae ulture National Fisheries R&D Institute 40S-I Shining, Kijang Pusan 619-900, Korea 'Aquaculture Division West Sea Fisheries Institute 98-36 Puksung / lncheon 400-201, Korea ABSTRACT Spawning Pacific oysters. Crassostrea gigas, were collected monthly over 1 year from two distinct local seed grounds in southern coastal waters of the Korean Peninsula. One site. Tongyoung. is considered favorable for the collection of oyster seeds, the other, Koje, unfavorable. The reproductive health of the oysters was evaluated by means of reproductive stage and three condition indices. The health of the spawners was further evaluated by comparing the abundance and lipid content of D-shaped larvae they produced. Tongyoung oysters were distinct in spawning pattern, as compared to Koje oysters, which were characterized by a prolonged spawning stage, lower spawning prevalence, and earlier appearance of reproductive arrest. Spawns were dispersed rather than two cycles at both grounds, but D-shaped larval occurrence was considerably higher in Tongyoung oysters than in Koje oysters. The lipid content of D-shaped larvae was also higher in Tongyoung than in Koje. The condition indices of the two groups of oysters were similar except for the higher overall levels and earlier postspawning recovery time in Tongyoung oysters as compared to Koje oysters. The condition indices decreased after the start of spawning, but fluctuations were also noticed during the spawning period. These fluctuations were particularly significant [P < .05) in Tongyoung oysters, suggesting stronger spawning activities in the oysters. These results imply that reproductive health of spawning oysters might be determined by reproductive cycles and condition indices for a new seed ground of C. gigas seed. The abundance and lipid content of D-shaped larvae also seem to be indicators of physiological health of the spawners from which they were produced. KEY WORDS: Pacific oyster. Crassostrea gigas, nursery ground, reproductive health, condition indices, biochemical composition, reproductive cycle INTRODUCTION Koje and Tongyoung. located on the south coast of the Korean peninsula, have long been used as farming grounds for Pacific- oyster. Crassostrea gigas. one of the most commercially important bivalves cultured in the country. Oyster culture in these areas has shown a marked increase, including increasing application of ad- vanced technology, after FDA designation of these areas as blue belts (clean areas) for oyster culture in 1972. However, since pro- duction reached 288,000 M/T in 1987. it has shown a decrease, largely attributable to local shortages of healthy seed (NFRDI Report 1997). The local failure to collect healthy seed has been attributed to the appearance of apparently reproductively un- healthy broodstock. This is believed to be related to increased water pollution, intensified farming, and frequent outbreaks of pathogenic infection that endanger the ecological integrity of the local spawning areas (Choi et al. 1997, NFRDI Report 1997. Park et al. 1998). Thus, successful farming of the oyster has become strongly dependent on selection of a good spawning ground where there are apparently healthy wild broodstock. Oyster health condition indices have been routinely measured to grade the farming grounds. Typically, the biometry and bio- chemical composition of oysters undergo marked seasonal changes associated with both environmental factors and the annual repro- ductive cycles (Giese 1969, Sastry 1979, Gallager and Mann 1986, Ruiz et al. 1992). Lucas and Beninger (1985) reviewed several condition indices for assessing health conditions related to adult bivalve reproduction. The condition indices reviewed by them have been widely used (and, in some cases, modified) for bivalve condition (Brown and Hartwick 1988, Fisher et al. 1996). Bio- chemical composition alone has also been a good indicator for evaluating bivalve health (Gallager et al. 1986, Ruiz et al. 1992. Robinson 1992). Other indicators have been studied for assessment of bivalve health, as related to the growth and reproduction. These are lipid and glycogen content (Holland 1978, Pieters et al. 1979. Holm and Shapiro 1984), digestive tubule condition (Winstead 1995), vesicular connective tissue condition, and RNA/DNA ratio (Kenchington 1994. Paon and Kenchington 1995). In general, the biochemical composition of bivalves changes conspicuously with season in association with the reproductive cycle (Giese 1969, Sastry 1979). The total content of lipid and/or glycogen has. there- fore, been continuously monitored to get valuable information con- cerning general condition, metabolic state, and gonadal develop- ment of bivalves (De Zwaan and Zandee 1972, Helm et al. 1973, Barber and Blake 1981. Gabbott 1983). These seasonal changes have been well documented for C. gigas (Walne 1970. Whyte et al. 1990, Ruiz et al. 1992. NFRDI Report 1997). The present study is aimed at providing measurements of the physiological condition of C. gigas broodstocks from two different grades of spawning ground to see if any act as indicators of re- productive health. If so, such measurements may help select better sites for wild seed collection and enhancing culture of this valued oyster. 445 446 Park et al. MATERIALS AND METHODS Oysters and Experimental Sites Thirty ropes with 500 Pacific oysters. Crassostrea gigas per rope were suspended from a long line situated at different localities in Koje and Tongyoung in the southern part of the Korean Penin- sula (Fig. 1). The two locations were chosen to evaluate how condition indices and other indicators related to reproduction of the broodstocks from an apparent poor oyster ground for seed collection (Koje I compared with apparently normal seed collection (Tongyoung). Sixty oysters were randomly sampled from the 30 ropes on a monthly basis during 1997. Care was taken to minimize sample variation in size class differences by selecting oysters of similar sizes. Reproductive Cycle Shucked oysters were fixed in Bouin's fixative for 24 h. The visceral mass (gonad included) were washed in overflowing tap water for 24 h. dehydrated in an ethyl alcohol series dilution, and embedded in paraffin. Sections 3- to 5-u.m thick were made and stained with hematoxylin-eosin on slides. The reproductive pro- cess of the adult oysters was divided into stages, based on miero- histological observation of the slides (Barber 1996) in which the spawning stage of the oysters was represented by the appearance of mature ova in the follicles, and the arrest stage was identified by the appearance of shrunken follicles and refractory ova. Occurrence of D-Shaped iMrvae The occurrence of D-shaped larvae produced from the adult oysters was measured on a daily basis for 4 months from June 1997. Plankton samples to a depth of 3 m were collected three times from each ground using a plankton net (75-u.m mesh size), and fixed in 2% of neutral buffered formaldehyde solution. The total number of D-shaped larvae collected were counted under the microscope and divided by 3 as an average for each sample. Lipid Content of the D-Shaped iMrvae The lipid content of the D-shaped larvae from the two oyster grounds was measured every 10 days for 4 months, starting in the middle of June 1997. The larval lipid content was measured three times using the technique of Mann and Gallager (1985) and then 128" to' 65° oo' 34° 45' ar" 30' 35' 40' Figure 1. Map showing the two locations of Tongyoung and Koje on southern coast of Korean Peninsula. averaged for each sample. In brief, larval lipid was extracted twice in 1:2 v/v chloroform : methanol and 2:1 v/v chloroform:methanol. respectively, after being freeze-dried, and then purified in 0.7% w/v NaCl solution before being quantified gravimetrically at 500 g for 10 min. Ratio of Dry Flesh Weight to Wet Flesh Weight The flesh and adductor muscles from the shells were weighed after the mantle fluid was removed with a paper towel. The flesh and adductor muscles were freeze dried for 36 h to achieve dry weight. The index was calculated as the ratio of dry flesh weight to wet flesh weight multiplied by 100. Ratio of Dry Flesh Weight to Dry Shell Weight The shells of oysters transported to the laboratory were scrubbed to remove any attached epifauna and adhering sediment. The flesh and adductor muscles were removed from the shells and freeze dried in the drier (EYELA FD- 1 ) for 36 h after mantle fluid was removed with paper towels. The shells were rinsed with tap water and dried at 1 10°C until the weights were no longer reduced. The index was calculated as the ratio of dry flesh weight to dry shell weight multiplied by 100. Glycogen Content Wet flesh for glycogen analysis was processed immediately after the total weight was determined. The wet flesh was added to a mixture of 50 ml of chloroform: methanol: water (2:4:1, v/v) and homogenized for 1 min with a Sorvall omni-mixer. After centrifu- gation at 4000 rpm for 10 min. the supernatant was decanted, and the residue was extracted with a further 50 ml of solvent mixture. This step was repeated twice. The residue was then dried at 60C for 48 h and weighed as the polymeric fraction containing glyco- gen (Whyte and Englar, 1982). RESULTS Spawning Cycle Spawning periods of the oysters from the two sites were de- termined (Table 1). Koje oysters were characterized by a longer spawning period and an earlier appearance of reproductive arrest, as compared to Tongyoung oysters. In Koje, a partial spawning first occurred in May (8%), and the spawning continued to October ( 12%). However, the prevalence of the spawning stage never ex- ceeded 90%. The first spawning in Tongyoung oysters occurred a month later than in Koje oysters, but it happened synchronously. The spawning in Tongyoung was first observed in June with a prevalence of 92%. In September. 95% of oysters were still avail- able as spawners in Tongyoung; whereas, only 46% remained TABLE 1. The occurrence of spawning and resting stages of C. gigas in two spawning sites from APR to NOV. % spawning (degenerative/resting) Oysters APR MAY JUN JUL AUG SEP OCT NOV KO TO 0(0) 0(0) 8(0) 86(0) 90(8) 78(16) 46(54) 12(88) 0(0) 92(0) 95(0) 94(0) 95(5) 3(97) 0(100) 0(100) KO and TO stand for Koje and Tongyoung oysters, respectively. Reproductive Health in Korean Wild Seed Oysters 447 TABLE 2. Lipid content (ng/larva) of D-shaped larvae of C. gigas from two seed grounds for 3 months of spawning period. JUN JUL AUG Site M L E M L E M L KO TO 3.67 (0.31) 4.63 (0.41) 2.97 (0.15) 3.77 (0.35) 2.83 (0.26) 4.23 (0.28) 3.10 (0.20) 5.03 (0.41) 3.27 (0.19) 3.47 (0.25) 2.73 (0.25) 4.23 (0.21) 2.87 (0.24) 3.74 (0.40) 2.67 (0.32) 3.06 (0.15) Abbreviations: KO, Koje oysters; TO. Tongyoung oysters; E. early in the month; M. middle of the month; L, late in the month. The number in parentheses is standard error mean (SEM). reproductively viable in Koje. Some Koje oysters entered into a period of reproductive arrest (degenerative and resting stages) 2 months earlier than Tongyoung oysters. Fifty-four percent of Koje oysters showed reproductive arrest by September; whereas, only 5% of Tongyoung oysters were at the same stage. Occurrence of D-Shaped Larvae Occurrence of D-shaped larvae from the two spawning grounds was measured on a daily basis for the 4-month spawning period. Four peaks in larval production were observed during this time at both locations; however, the abundance of larvae differed signifi- cantly (Fig. 2). At peak abundance, there were about 30,000 lar- vae/net for Tongyoung and about 6,000 for Koje. At Tongyoung, the first three spawning cycles provided sufficient larvae for seed collection; none of the spawning cycles in Koje provided sufficient larvae. Lipid Content in the D-Shaped Larvae Lipid content of the two D-shaped larvae was measured for 3 months starting in the middle of June (Table 2). The lipid content of Tongyoung larvae ranged from 3.1 to 5.0 ng/larva while those from Koje ground ranged from 2.7 to 3.3 ng/larva. Dry Flesh Weight to Dry Shell Weight Ratio (Condition Index I) Condition index I , calculated as a ratio of dry flesh weight to dry shell weight multiplied by 100, exhibited similar patterns at the two sites (Fig. 3), with an initial increase followed by a decrease over the year cycle. Despite their similar patterns, the over-all levels of the condition index 1 in Tongyoung oysters were higher emlemleml em jln jul aug sb3 Figure 2. D-shaped larval occurrence during the spawning season of C. gigas from two seed grounds, Tongyoung (upper), a favorable, and Koje (lower), an unfavorable seed ground. D-shaped larvae to a depth of 3m were collected three times from each ground using plankton net (75pm mesh size) and averaged by dividing by three. Abbreviations: E, early; M, middle; L, late of the month. 0) ■o c o O Figure 3. Condition index 1 of C. gigas, expressed as dry flesh weight to dry shell weight multiplied by 100 (mean ± SEM) (Solid bars = Tongyoung oysters, open bars = Koje oysters). A total of 60 oysters. 30 from each seed ground (Koje and Tongyoung), were measured monthly over a year. The shells were removed for measurement of dry shell weight. The flesh weight includes all tissues removed from the shells. Asterisk indicates significant difference ip < 0.05) from previous value during spawning season. 448 Park et al. 30 20 S ■o c T3 C o o 10 I | Koje oyster | Tongyoung oyster i ~ ~ ~ r J F M A M 15 S O N D Month Figure 4. Condition index 2 of C. g/'gas, expressed as dry flesh weight to wet flesh weight multiplied by 100 (mean ± SEM) (Solid bars = Tongyoung, open bars = Koje). A total of 120 oysters, 60 from each ground (Koje and Tongyoung), were measured monthly over a year. The flesh weight includes all tissues removed from the shells. Asterisk indicates significant difference (p < 0.05) from previous value during spawning season. than in Koje oysters. This was particularly true when the off- spawning season was considered. In the off-spawning season (January to May). Tongyoung oysters were significantly higher than Koje oysters in the value of condition index 1 over 4 months (P < .05). With the commencement of spawning season (June to October), condition index 1 at both oyster grounds were marked by a decreasing trend. In the decreasing trends, only one significant decrease of condition index 1 between successive months (P < .05) was noticed in Koje oysters whereas, three significant decreases were observed in Tongyoung oyster, suggesting mass spawnings in Tongyoung oysters during the spawning season. The maximum values of condition index 1 were 10.4 for Tongyoung and 8.55 for Koje oysters, and minimum values were 3.7 for Tongyoung and 3.5 for Koje oysters. Dry Flesh Weight to Wet Flesh Weight Ratio (Condition Index 2) Condition index 2, expressed as a ratio of dry flesh weight to wet weight multiplied by 100, was similar to condition I at both sites, although greater fluctuations were noted in condition index 2 (Fig. 4). Condition index 2 showed an increase from January to May, and then a decrease from June, the time when most of the oysters started to spawn. Significant decreases in condition index 2 between successive months (P < .05) were noticed once (July) in Koje oysters and twice (July and October) in Tongyoung oysters. Interestingly, significant elevations of condition index 2 (P < .05) from the previous month were also noted at both locations during the over-all decreasing trend. This significant fluctuation of con- dition index 2 in Tongyoung oysters was reminiscent of spawning events, although spawning abundance was different at the two localities. The maximum values of condition index 2 were 23.8 for Tongyoung and 20.3 for Koje oysters, and minimum values were 9.0 for Tongyoung and 10.4 for Koje oysters. 10 - x ID o O 5 - [ | Koje oyster | Tongyoung oyster M A Month Figure 5. Condition index 3 of C. gigas, expressed as total glycogen content to dry flesh weight multiplied by 100 (mean ± SEM) (Solid bars = Tongyoung, open bars = Koje). A total of 120 oysters, 60 from each ground (Koje and Tongyoung), were measured monthly over a year. The flesh weight includes all tissues removed from the shells. Asterisk indicates significant difference (p < 0.05) from previous value during spawning season. Glycogen Content (Condition Index 3) Condition index 3. represented by total glycogen content to dry flesh weight multiplied by 100, was similar to condition indices 1 and 2 at both sites (Fig. 5). Condition index 3 also decreased significantly {P < .05) with the start of spawning season of the oysters. Unlike condition index 2, however, no significant in- creases of condition index 3 were observed during the spawning seasons of the two oyster localities. However, significant decreases of condition index 3 (P < .05) were marked once for Koje oysters (July) and twice for Tongyoung oysters (July and October) during their spawning seasons. The maximum values of condition index 3 were 1 1.2 for Tongyoung and 8.8 for Koje oysters, and minimum values were 4.6 for Tongyoung and 3.8 for Koje oysters. DISCUSSION Pacific oysters. Crassostrea gigas from two distinct spawning sites were studied over 1 year to assess their reproductive health. The timing and duration of gametogenesis of C. gigas are different in the different localities of the world (Ventilla 1984. Sphigel 1989). Generally, the reproductive strategy of C. gigas can be considered as an adaptation to ambient environmental factors, with temperature and nutrition as principal factors (Lubet 1976, Ruiz et al. 1992). Koje oysters were characterized by a prolonged prespawning stage, a lower prevalence of spawning oysters, and an earlier appearance of reproductive arrest, as compared to Tong- young oysters (Table 1 ). In addition. Koje oysters exhibited ex- tremely low production of D-shaped larvae. However, although the occurrence of D-shaped larvae at Tongyoung was high enough for seed collection, they were dispersed across four spawning cycles (Fig. 2). In general, spawning peaks of C. gigas are marked by two and. in some cases, three cycles in a season (Koganezawa 1972, Reproductive Health in Korean Wild Seed Oysters 449 Ruiz et al. 1992). Yoo and Ryu ( 1985) also observed two distinct cycles of D-shaped larvae. Lipid has been considered one of the principal energy sources for gametogenesis of adult bivalves ( Gabbott 1 983 ) and for normal development of eggs (Gallager and Mann 1986). Good survival of bivalve larvae also depends upon the optimum lipid level (Holland 1978,Whyteetal. 1987, Helm et al.. 1991. Lim et al. 1999). Bayne (1972) and Bayne et al. (1975) reported reduced growth in larvae of Mytilus ccliilis that developed from gametes of nutritionally stressed adults. Gallager and Mann (1986) also reported that a minimum threshold lipid level in eggs was necessary for optimal survival through nonfeeding embryonic and early larval stages. Lipid content of the D-shaped larvae from the two locations was studied for 3 months (Table 2). Tongyoung larvae ranged from 3.1 to 5.0 ng/larva on average; whereas, Koje larvae ranged from 2.7 to 3.3 ng/larva. The lower larval lipid content, together with lower larval occurrence suggested that the eggs produced from Koje oysters were not as healthy as Tongyoung oysters. Lucas and Beninger (1985) reviewed the physiological condi- tion indices most commonly used in bivalve aquaculture. Three of these were applied in the present study: condition index 1 for dry flesh weight to dry shell weight; condition index 2 for dry flesh weight to wet flesh weight; and condition index 3 for total glyco- gen content to dry flesh weight. Condition index 1, as indicated by Rheault and Rice (1996), was performed using the method de- scribed by Lucas and Beninger (1985) rather than that of Lawrence and Scott (1982) to avoid problems related to varied shell thick- ness and morphology. Condition indices 1 and 2 express the pro- portions of dry matter to whole tissues. High proportion of water in the tissues reflects an energy-depleted state of an organism attributed to losses of insoluble ash. lipid, glycogen, sugars, pro- tein, and nitrogenous compounds from body tissue (Whyte et al. 1990). Condition 3 also signifies an energy status of the oysters. Generally, glycogen has been regarded as the main source of en- ergy in bivalves (De Zwaan and Zandee 1972, Barber and Blake 1981). The natural gametogenic cycle in bivalve mollusks is closely linked to cycles of glycogen storage and subsequent de novo synthesis of lipid during vitellogenesis at the expense of stored glycogen (Gabbott 1975). Interruption of this cycle by ar- tificial conditioning may force the development of eggs before sufficient glycogen has accumulated for the synthesis of lipid. Thus, the consequence would be the production of either fewer eggs or eggs of suboptimal quality (Gallager and Mann 1986). All condition indices declined from May or June (P < .05) with the commencement of spawning. The condition indices remained low until the beginning of winter. The patterns for condition indices at the two oyster localities were similar except for the higher mag- nitude and earlier rebound in Tongyoung oysters. The condition indices declined throughout the spawning period but showed some month-to-month fluctuation. These fluctuations were particularly significant (P < .05) in condition indices 2 and 3 of Tongyoung oysters, perhaps indicating stronger spawning activities in Tong- young oysters. These results imply that the reproductive health of the spawners might be determined from measures of reproductive cycles and condition indices. This could be very useful for the new ground selection of C. gigas seed. The abundance and lipid content of D-shaped larvae were greater from the site that exhibited a shorter spawning season with higher prevalence of spawning organisms and a quicker physiological recovery. ACKNOWLEDGMENT We thank Dr. Sharon E. McGladdery at Fisheries and Oceans, Canada for her critical comments and suggestions on the manu- script. LITERATURE CITED Barber. B.J. 1996. Gametogenesis of eastern oysters. Crassostrea vir- ginica (Gmelin. 1791). and Pacific oysters, Crassostrea gigas (Thun- berg. 1793) in disease-endemic lower Chesapeake Bay. /. Shellfish. Res. 15:285-290. Barber. B. J. & N. B. Blake. 1981. Energy storage and utilization in rela- tion to gametogenesis in Argopecten irridians concentricus (Say). J. Exp. Mar. Biol. Ecol. 52:121-134. Bayne, B. L. 1972. Some effects of stress in the adult on larval develop- ment of Mytilus edulis. Nature. Lond. 237:459. Bayne, B. L., P. A. Gabbott & J. Widdows. 1975. Some effects of stress in the adult on the eggs and larvae of Mytilus edulis L. J. Mar. Biol. Assoc: U. K. 55:675-689. Brown, J. R. & E. B. Hartwick. 1988. Influences of temperature, salinity and available food upon supended culture of the pacific oyster, Cras- sostrea gigas. II. Condition index and survival. Aquaculture 70:253- 267. Choi. W.- J.. Y - Y. Chun. J- H. Park & Y.- C. Park. 1997. The influence of environmental characteristics on the fatness of Pacific oyster, Cras- sostrea gigas. in Hansan-Koje Bay. J. Korean Fish. Soc. 30:794-803 De Zwaan. A. & D. I. Zandee. 1972. Body distribution and seasonal changes in the glycogen content of the common sea mussel Mytilus edulis. Comp. Biochem. Physiol. A 43:53-58. Fisher. W. S., J. T. Winstead, L. M. Oliver. H. L. Edmiston & G. O. Bailey. 1996. Physiologic variability of eastern oysters from Apalachi- cola Bay, Florida. J. Shellfish Res. 15:543-553. Gabbott. P. A. 1975. Storage cycles in marine bivalve molluscs: a hypoth- esis concerning the relationship between glycogen metabolism and ga- metogenesis. pp. 191-211. In: H. Barnes (ed.). Proceedings of the Ninth European Marine Biology Symposium, Aberdeen University Press, Aberdeen, Scotland. Gabbott. P. A. 1983. Developmental and seasonal metabolic activities in marine mollusks. pp. 165-217. In: P. W. Hochachka (ed.). The Mol- lusca 2. Environmental Biochemistry and Physiology. Academic Press, New York. Gallager, S. M. & R. Mann. 1986. Growth and survival of larvae of Mer- cenaria mercenaria (L.) and Crassostrea virginica (Gmelin) relative to broodstock conditioning and lipid content of eggs. Aquaculture 56: 105-121. Gallager. S. M., R. Mann & G. C. Sakaki. 1986. Lipid as an index of growth and viability in three species of bivalve larvae. Aquaculture 56:81-103. Giese, A. C. 1969. A new approach to the biochemical composition of the mollusk body. Oceanogr. Mar. Biol. A. Rev. 7:175-229. 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 vigor. L. Mar. Biol. Assoc. U. K. 53:673-684. Helm, MM.. D. L. Holland. S. D. Utting & J. East. 1991. Fatty acid composition of early nonfeeding larvae of the European flat oyster, Ostrea edulis. J. Mar. Biol. Assoc. U.K. 71:691-705. Holland. D. L. 1978. Lipid reserves and energy metabolism in the larvae of benthic marine invertebrates, pp. 85-123. In: P. L. Malins and J. R. Sargent (eds.). Biochemical and Biophysical Perspectives in Marine Biology. Academic Press, San Diego. Holm. N. P. & Shapiro, J. 1984. An examination of lipid reserves and the nutritional status of Dapluua pulex fed Aphanizomenon flos-a Limnol. Oceanogra. 29:1137-1140. 450 Park et al. Kenchington. E. L. R. 1994. Spacial and temporal variation in adductor muscle RNA/DNA ratio in sea scallops (Placopecten magelkmicus) in the Bay of Fundy. Canada. J. Shellfish Res. 13:19-24. Koganezawa. A. 1972. Ecological studies at a seed oyster production area — II. studies on the spawning condition of the Pacific oyster in Matsushima Bay. Bull. Japan. Soc. Sci. Fish. 38:1315-1324. Lawrence, D. R. & G. I. Scott. 1982. The determination and use of con- dition index of oyster. Estuaries 5:23-27. Lim. H. J.. M. S. Park. J. Y. Cho & Y.-K. Hong. 1999. Dietary evaluation of the freeze-dned alga Isochrysis galbana for the larval survival of the Pacific oyster Crassostrea gigas. Bull. Korean. Fish. Soc. in press. Lubet, P. 1976. Ecophysiologiede la reproduction chez les mollusques lamellibranches. Haliotis 7:49-55. Lucas, A. & P. G. Beninger. 1985. The use of physiological condition indices in marine bivalve aquaculture. Aquaculture 44:187-200. Mann, R. and S. M. Gallager. 1985. Physiological and biochemical ener- getics of larvae of Teredo navalis and Bankia goul (Bartsch) (Bivalve: Teredinidae). J. Exp. Mar. Biol. Ecol. 85:211-228. NFRDI Report, 1997. Studies on the bad seed collection in oyster growing area and investigation of the new oyster seed collection area. National Fisheries Research and Development Institute, Pusan. Korea, pp. 1-226. Paon, L. A. & E. L. R. Kenchington. 1995. Changes in somatic and repro- ductive tissues during artificial conditioning of the sea scallop. Pla- copecten magellanicus (Gmellin, 1791). J. Shellfish. Res. 14:53-58. Park, M. S.. H. J. Lim & P. J. Kim. 1998. Effect of environmental factors on the growth, glycogen and hemoglobin content of cultured arkshell, Scapharca hroughtonii. J. Korean Fish. Soc. 31:176-185. Pieters, H, J. H. Kluytmans. W. Zurberg & D. I. Zandee. 1979. The influ- ence of seasonal changes on energy metabolism in Mytilus edulis (L)I. Growth rate and biochemical composition in relation to environmental parameters and spawning, pp. 285-292. In: E. Naylor and R. H. Hart- noil (eds.). Cyclic Phenomena in Marine Plants and Animals. Perga- mon Press, N.Y. Rheault, R. B. & M. A. Rice. 1996. Food-limited growth and condition index in the eastern oyster. Crassostrea virginica (GMEL1N 1791 ). and the bay scallop. Argopecten irradians (LAMARCK 1891). J. Shellfish Res. 15:271-283. Robinson. A. 1992. Dietary supplements for reproductive conditioning of Crassostrea gigas Kumamoto (Thunberg). II. Effects on glycogen, lipid, and fatty acid content of the broodstock oyster and eggs. J. Shellfish Res. 1L443-W7. Ruiz. C. M. Abad. F. Sedano. L. O. Garcia-Martin & J. L. Sanchez Lopes. 1992. Influence of seasonal environmental changes on the gamete pro- duction and biochemical composition of Crassostrea gigas (Thunberg: in suspended culture in El Grove, Galicia. Spain. J. Exp. Mar. Biol. Ecol. 155:249-262. Sastry, A.N. 1979. Pelecypoda (excluding Ostreidae). pp. 113-292. In: A. C. Giese and J. S. Pearse (eds.). Reproduction of Marine Inverte- brate, vol. V. Mollusks: Pelecypods and Lesser Classes. Academic Press. San Diego. Sphigel. M. 1989. Gametogenesis of the European flat oyster (Ostrea edulis ) and the Pacific oyster (Crassostrea gigas) in warm water in Israel. Aquaculture 80:343-349. Ventilla. R. F. 1984. Recent development in the Japanese oyster culture industry. Adv. Mar. Biol. 21:1-57. Walne. P. R. 1970. The seasonal variation of meat and glycogen content of seven populations of oysters. Ostrea edulis, and a review of the litera- ture. Fisheries Investigation. Ministry of Agriculture. Fisheries and Food, Series 2, 26:1-35. Whyte, J. N. C. & J. R. Englar. 1982. Seasonal variation in the chemical composition and condition indices of Pacific oyster, Crassostrea gigas, grown in trays and on the sea bed. Can. J. Fish. Aquat. Sci. 39:1984- 1094. Whyte, J. N. C. J. R. Englar & B. L. Carswell. 1990. Biochemical com- position and energy reserves in Crassostrea gigas exposed to different levels of nutrition. Aquaculture 90:157-172. Winstead. J. T. 1995. Digestive tubule atrophy in eastern oysters. Cras- sostrea virginica (Gmelin. 1791). exposed to salinity and starvation stress. / Shellfish Res. 14: 105-1 1 1 . Yoo, S. K. & H. Y. Ryu. 1985. Occurrence and survival rate of the larvae of Pacific oyster Crassostrea gigas in Hansan Bay. Bull. Korean Fish. Soc. Jmirnal of Shellfish Research. Vol. 18, No. 2. 451—458, 1999. GROWOUT OF BLACKLIP PEARL OYSTERS, PINCTADA MARGARITIFERA, ON CHAPLETS IN SUSPENDED CULTURE IN SOLOMON ISLANDS KIM J. FRIEDMAN1 2* AND PAUL C. SOUTHGATE1 James Cook University Townsville Queensland 481 I, Australia 2ICLARM Coastal Aquaculture Centre, P.O. Box 438 Honiara Solomon Islands ABSTRACT This study, conducted in the "open" reef systems of Solomon Islands, assessed growth and survival of blacklip pearl oysters (Pinctada margaritifera, L.) on chaplets in suspended culture. Oysters were robust and mortalities after handling and drilling were low (<0.6 %). Survival of an initial batch of P. margaritifera was 87% after 1 year. Groups of oysters with mean dorsoventral measurements (DVM) of 63 and 71 mm showed mean (± SE) annual growth rates of 64 ± 2 and 60 ± 1 mm. respectively. Growth rates compared favorably with those reported for P. margaritifera in Polynesia and indicate that oysters caught as spat (approx. 1 1 mm. DVM) would reach acceptable size for "seeding" (110 mm DVM) in around 16 months. Oysters on chaplets were subject to significant fouling by algae, however, survival of oysters cleaned every 2. 3, 4. and 6 weeks was 96-97% over a 7-month period. Growth of oysters cleaned every 3 weeks was significantly greater than those cleaned every 2 or 6 weeks. Oysters became detached from chaplets (through drill-hole breakage) in significant numbers; this problem was greater for smaller oysters. When oysters were attached singly on chaplets, approximately 54% made byssal attachments to the rope; however, 90% of oysters held in pairs on chaplets made byssal attachments to each other. Although paired oysters could be cleaned more rapidly than oysters hung singly, shell growth (DVM) of paired oysters was significantly reduced. KEY WORDS: pearl oyster. Pinctada. growth, survival, chaplets. suspended culture, open reefs INTRODUCTION Production of cultured black pearls from Pinctada marga- ritifera is seen as an appropriate, sustainable industry for remote regions of the Pacific (Lucas et al. 1995), and has expanded rapidly in eastern Polynesia over the last decade (Fassler 1995). French Polynesia has been at the forefront of this development and cur- rently earns approximately US $145 million annually from the sale of black pearls (Remoissenet 1996. Doubilet 1997). This success has not gone unnoticed by other small island nations in the region (Lucas et al. 1995). Historically, these nations have relied on more modest incomes from the sale of P. margaritifera shell for its nacre or "mother-of-pearl" (MOP) (Gervis and Sims 1992, Richards et al. 1994). In Cook Islands, pearl culture started with one family in 1982 (Sims 1993a). and. by 1994, pearl sales generated an annual income of US$ 4.5 million (Fassler 1995). At present, black pearl culture is also underway, being attempted or assessed in Japan (Lintilhac 1987). Marshall Islands (Sims pers comm. 1998). China (Meng and Xing 1991). Vanuatu (Anon 1996). Fiji (Ward 1995, Mercier and Hamel 1998), and Solomon Islands (Friedman et al. 1996, Mercier and Hamel 1998, Friedman et al, 1998). In the atolls of Polynesia. P. margaritifera are generally har- vested from collectors as spat and cultured on dropper ropes or "chaplets" (see Fig. 1 ) when their dorsoventral measurement (DVM. Nicholls 1931) reaches 65-90 mm (AQUACOP 1982. Preston 1990). This method is also widely used for scallops in Japan (Ventilla 1982). Oysters are drilled through the base of the shell (dorsal posterior region) and attached to chaplets with wire or monofilament fishing line. Oysters are grown on chaplets for the rest of their time in culture, only being removed temporarily when *Corresponding address: P.O. Box 20, North Beach. Western Australia 6020. E-mail: kfriedman@fish.wa.gov.au they reach 110 mm DVM. to be "seeded" for pearl production (Lintilhac 1987). Oysters large enough to be held on chaplets have "size refuge" from all but the largest fish and invertebrate predators (Coeroli et al. 1984). and, because longlines are set in relatively deep water, they are isolated from predators associated with reefs (Swift 1985, Sims and Sarver 1995). However, survival and growth of pearl oysters in suspended culture is also influenced by fouling (Ala- garswami and Chellam 1976. Mohammad 1976. Doroudi 1996, Taylor et al. 1997). In Cook Islands, chaplets are removed from the water once a year for washing with pressure hoses. At this time, algae ("soft" fouling) and such organisms as cementing bivalves and tubular polychaetes ("hard" fouling) are removed. Before oys- ters are returned to the water, the fastening of the oyster to the chaplet is checked for wear, and the oyster is re-drilled if necessary (J. Lyons, pers comm. 1997). Oysters are sometimes lost from chaplets when drill holes break, and oysters fall to the bottom. In Cook Islands, farmers lose -5% of stock in this way (R. Newnham. pers comm. 1995): how- ever, because atoll lagoons generally have a hard substrate (Sims 1992, Coeroli et al. 1984). a large percentage of these oysters can be recovered (J. Lyons, pers comm. 1997). Recently, experiments were conducted in Solomon Islands to determine whether collections of wild spat and culture of juvenile oysters would be successful in the "open" reef systems that are characteristic of that region (Friedman and Bell 1996, Friedman et al. 1996, Friedman et al. 1998. Friedman 1998). However, there is a paucity of published information on growth of P. margaritifera held on chaplets for open reef systems of the Pacific. This paper describes growth and survival of P. margaritifera held on chaplets in Solomon Islands. In addition, experiments were conducted to: ( 1 ) determine a cleaning regime that optimizes growth of oysters held on chaplets: and (2) assess the rate of oyster losses from chaplets because of failure of drill holes. 451 452 Friedman and Southgate /?*■&/*? SC. * ' ♦^aB' Figure 1. Chaplet system for culturing P. margaritifera. MATERIALS AND METHODS Growout of P. margaritifera on chaplets was carried out within Gizo lagoon in the Western Province of Solomon Islands (Fig. 2). The section of the lagoon chosen for oyster culture was approxi- mately 1 km across with a mean depth of -40 m. and numerous passages and sections of submerged reef linking the lagoon to the open ocean. Currents in this area measured over a representative tidal cycle with a Schiltknecht Mini Air 20 current meter ranged between 0.0-0.15 m s"1. P. margaritifera with a DVM of -65 mm were drilled (2-mm drill bit) and tied to 4-mm polypropylene rope chaplets using monofilament fishing line (37 kg breaking strain) of a similar width to the drill hole (Fig. 1). All oysters were threaded onto single pieces of monofilament and tied so that the flatter of the two valves lay against the chaplet rope (Fig. 1). Oysters tied singly, were attached at -20 cm intervals, when attached singly along the chaplet and 40 cm apart when in pairs (Fig. 1 ). Chaplets held 10 single oysters (or five pairs) and were attached to submerged longlines at 1.5-m intervals. The 100-m longlines (Fig. 3) were held at a depth of approximately 9 m and sited at least 50 m clear of fringing reef, over a sandy substrate. During the growth trials, oysters were brushed in situ by divers using SCUBA, to remove fouling on a monthly basis. Oysters that became detached from chaplets during the trials were re-drilled and replaced in their original position on the chaplet. Growth and Survival of Oysters on Chaplets A trial group of 90 oysters were drilled and hung singly on chaplets at the end of August 1996. These oysters were divided into two size classes: 58-65 mm and 66-78 mm DVM, with a mean (± SE) of 63 ± 0.3 mm (n= 51) and 71 ± 0.7 mm (n= 39), respectively. These oysters were removed from the water and mea- sured every 3 months. A second group of 1,440 oysters with a mean (SE) DVM of 66 ± 0.3 mm, were drilled and hung singly on chaplets between 1 1 November 1996 and 5 December 1996. These oysters were removed from the water and measured in February and again in September 1997. At the end of these trials (September 1997), oysters from ex- periments were measured for wet weight and shell thickness (Gervis and Sims 1992). as well as DVM. Additional data from pearl oysters grown on chaplets under similar conditions, but not included in experiments, where collated with data from experi- mental animals and used for analysis of morphometrie relation- ships. Morphometric analyses of pearl oysters were conducted on both experimental animals and other oysters grown under similar conditions. Modifying Cleaning Regimes of Oysters on Chaplets To identify a cleaning regime that provided satisfactory growth and survival, with acceptable labor input, 100 oysters that were hung singly on 10 replicate chaplets were cleaned every 2, 3, 4, or 6 weeks, for 7 months (20 February to 22 September 1997). At the start and end of this experiment, all oysters were measured (DVM). The mean DVM at the start of the experiment was 85 ± 0.4 mm (/; = 400). Fouling algae were saved from 25% of the oysters from each cleaning treatment in June 1997. at the time when oysters were scheduled for cleaning. Algae from individual oysters were placed into individual fine-meshed bags and rinsed to removed fine particulate matter (silt) and contaminants (e.g., shell, crabs) before being sun dried for up to a week. The samples were then oven dried for 24 h at 65 °C and weighed. Retention of Oysters on Chaplets Chaplets made up in November to December 1996 (n = 1440) and immersed for 7 months were assessed to determine any rela- tionship between the size of oysters at drilling and their retention on chaplets. All data used in this assessment were from chaplets where oysters had been hung singly. Strong byssal attachment reduces the chance of oysters being lost from chaplets. Observations that oysters readily made attach- ments to other oysters prompted an experiment to monitor byssal production and attachment of oysters hung in pairs and oysters hung singly. In April 1997. eight replicate chaplets ( 10 individuals per chaplet) for each treatment were deployed on longlines. The mean (±SE ) DVM at the start of the experiment was 67 ± 0.5 mm (n = 160). At the end of 4 months immersion, retention of oys- ters, byssal attachment, and growth (DVM and wet weight) of oysters were recorded. Chaplets holding "paired" and "single" oysters were cleaned monthly, and the time required for cleaning was determined. Following this experiment, byssal threads of all oysters were severed, and re-attachment was examined by SCUBA divers after 1. 2. 7. 14. and 21 days. Blacklip Pearl Oysters Suspended on Chaplets in the Solomons 453 7°S _ 9°S. WESTERN PROVINCE 157°E 159°E of '-CS ^- \ \Choiseul SOLOMON ISLANDS £> ^New Georgia 3." '?-?*, v»v V- *•- /^^ **.*-*p «■ GIZO ICLARM's Nusa Tupe Fieldstation 5km ^.-.'^ Figure 2. Gizo Lagoon in the Western Province of Solomon Islands. P. margarilifera were held in suspended culture east of Nusa Tupe Island (circled area). Data Analysis To examine differences in oyster growth (mm DVM) among cleaning regimes, a nested analysis of variance (ANOVA). (clean- ing schedule x chaplet [cs]) was used with data from six oysters per chaplet. To examine differences in dry weights of algae among cleaning regimes, a similar analysis was conducted using algal weights from five oysters from four chaplets within each treat- ment. To compare byssal attachment between oysters hung singly and in pairs, the number of oysters on each chaplet forming attach- ments with each other, or the rope, were compared for each treat- ment using a t-test. Growth (mm DVM and g wet weight) of oysters in each treatment were analyzed using t-tests. To analyze variation in the time taken to brush all oysters on a chaplet for each treatment, a two-way ANOVA, (method of attachment x diverl was used for two divers each cleaning four chaplets within each treatment. Before t-tests or ANOVA. data were checked for homogeneity of variance using Levene's or Cochran's test, respectively, and transformed to log,0 (x + l ) to meet this assumption where nec- essary. Significant differences among means were identified using Tukey's HSD test. RESULTS Growth and Survival of Oysters on Chaplets From the initial batch of 90 oysters, 78 (87%) were alive 1 year later. For the 1 2 oysters that were lost, only two dead shells were found attached to chaplets. The two subgroups drilled at sizes between 58-65 mm and 66-78 mm. had a mean (± SE) annual growth rate of 64 ± 1.9 mm and 60 ± 1.4 mm DVM. respectively (Fig. 4). Two of the oysters from each of these subgroups showed very low growth rates (<39 mm y"1 DVM). Of the 1,440 oysters drilled and hung in November to Decem- ber 1996. 1.342 (93.2%) were live in September 1997. Of the 98 that were lost, only eight dead shells were found on chaplets. Growth of these oysters is represented in Figure 4. by four sub- groups of oysters, delineated by their mean (± SE) sizes at drilling 454 Friedman and Southgate Substrate Floats Headline Anchor line for floats Angle line Figure 3. Diagram of longline system for suspended culture of P. margaritifera on chaplets. (DVM): 52 ± 0.2 mm (n = 641; 60 ± 0.2 mm (n = 225); 69 ± 0.2 mm (;? = 206); and 78 ± 0.3 mm (n = 63). Growth of these oysters was not as fast as the 90 oysters drilled in August 1996 (Fig. 4). Those hung on chaplets at a size of approximately 65 mm DVM attained 1 10 mm DVM. the size required for seeding, in 8 months. When oysters were attached to chaplets at a size of 77.8 mm DVM. they were large enough to seed in 6.5 months (Fig. 4). Morphometric relationships between wet weight and DVM and shell thickness and DVM are shown in Fig. 5 a & b. Graph a) is useful to farmers of P. margaritifera, because it allows for com- parisons of size/weight ratios between stocks of oysters grown in Solomon Islands and those grown elsewhere. The graph can also help farmers of P. margaritifera in Solomon Islands who want to calculate flotation needs for longline culture. Graph b) is added for general reference and will be of value should a relationship be established between shell thickness and the capacity for oysters of greater thickness to accept and retain larger nuclei. > D 0 30 60 90 120 150 180 210 240 270 300 330 360 390 Time on chaplets (days) Figure 4. Growth (DVM SE) trajectories for subgroups of P. marga- ritifera of different sizes placed on chaplets for growout. The two long-dash lines depict growth of the 90 oysters drilled and hung on chaplets in August 1996, and the four dotted lines are from the bulk of oysters drilled and hung on chaplets in November to December 1996. Modifying Cleaning Regimes of Oysters on Chaplets Survival of oysters cleaned every 2. 3, 4, and 6 weeks was 96. 96, 96, and 97%, respectively, despite the fact that algae grew heavily on oysters, sometimes covering them completely. Analysis of dry weight of algae on oysters cleaned every 2, 3, 4, or 6 weeks showed a progressive increase in weight of algae from 2 to 4 weeks, but no significant increase in weight between 4 and 6 700 600 500 400 300 200 100 0 b) 3 E E CO U) CD c -c o !c r2= 0 8625 n = 2272 40 60 80 100 120 140 160 180 200 45 40 35 30 25 20 15 10 r2= 0 5739 n = 2272 40 60 80 100 120 140 160 180 200 Oyster size (mm DVM) Figure 5. Size of P. margaritifera (DVM) relative to: a) wet weight; and b) thickness. Blacklip Pearl Oysters Suspended on Chaplets in the Solomons 455 TABLE 1. Results of the nested ANOVA for effects of cleaning schedule (fixed factor) and chaplet (random factor) on a) dry weight of algae on collectors, and b) growth of P. margaritifera (DVM). weeks (Table 1, Fig. 6a). The average mass of algae on oysters not cleaned for 6 weeks was 7.8 ± 0.8 SE g per oyster. The growth increment of oysters among the four cleaning treat- ments differed significantly (Table 1), with oysters cleaned every 3 weeks growing significantly faster (p < .05 ) than those cleaned on a 2- or 6-week schedule (Fig. 6b). Retention of Oysters on Chaplets Mortalities of oysters after handling and drilling were low (< 0.6%). However, greater proportions of smaller oysters were lost a) s b) E 35 Weeks between cleaning Figure 6. Variation in: a) mean (±SE) dry weight of algae removed from P. margaritifera; and b) growth of P. margaritifera (mean DVM ±SE) cleaned at 2, 3, 4. and 6 week frequencies. Means with different superscripts were significantly different {p < .05). n=271 Source of Variation df MS F P o a ) Dry weight of algae 0! Cleaning schedule 3 I.270 S2.452 0.0000 (D Chaplet (clean schedule] 12 0.015 7.630 0.0000 4 L r 1 14 Days after byssus was severed Figure 8. Mean (±SE) number of single (blank) and paired (solid) P. margaritifera per chaplet that made re-attachments after byssal threads had been severed. cultivating P. maxima in Indonesia and Australia. Whereas atoll lagoons in Polynesia are surrounded by a low-lying carbonate island (atoll) and are relatively nutrient poor (Littler et al. 1991 ). open reef systems are generally bordered by high islands that are the source of nutrient inputs from fresh water runoff (Chellam et al. 1987). The higher nutrient load in the lagoons of Solomon Islands may have been a factor in the good growth rates recorded in this study. Yukihira ( 1998). showed that increases in food avail- ability produced increased growth rates in P. margaritifera up to an optimum of 1-2 mg L~' (ca. 10.000-20.000 cells mL"' ). In the Cook Islands, Ponia (1997) found that water movement on a farm of 50,000 oysters needed to be > 0.01 m s~' to avoid 98% removal of microalgae by oysters hung on chaplets. His recordings of water movement in Manihiki lagoon in the Cook Islands ranged between 0-0.06 m s"1. In French Polynesia, surface water movement in Takapoto atoll (closed atoll) is 0.03 m s_1 (Salvat 1981). In Sol- omon Islands, tidal water flow is greater (0-0.15 m s~'), ensuring replenishment of food to culture areas. Although measurements of food abundance were not taken in this study, we suggest that the greater nutrient loading and relatively high water movement in Solomon Islands was likely to have stimulated greater pearl oyster growth (Chellam et al. 1987). Additional anecdotal evidence for the greater levels of nutrients in Solomon Islands than Polynesia is the fact that growth of algae on oysters and chaplet ropes was more of a problem in Solomon Islands than in the atoll lagoons of Polynesia. The noticeable dif- ference in the level of algal fouling may have been influenced by differences in the numbers of grazers between these two regions, although this is unlikely, because there was no observable evi- dence that fish or invertebrate grazers were less common in Sol- omon Islands than Polynesia. Despite oysters in Solomon Islands becoming completely covered with algae, observations in situ re- vealed that even heavily fouled oysters were able to open their valves normally. Although oysters required regular cleaning, there was relatively little fouling by cementing organisms such as bi- valves and polychaetes ("hard" fouling), which are more difficult to remove than algae. In addition, hard fouling and other byssally attached bivalves have been shown to cause shell deformity of other pearl oyster species during culture (Dharmaraj et al. 1987, Doumenge et al. 1991, Taylor et al. 1997). Such fouling may render oysters vulnerable to attack from predators such as small fish and crabs (J. Taylor, pers comm. 1998) and, in the worst cases, result in mortality (Dharmaraj et al. 1987). Hard fouling organisms present a greater problem than algal fouling, because they also compete directly with oysters for food and space (Ponia 1997). In Cook Islands, settlement of the pest species P. maculata, can cause longlines to sink to the substrate. This species is the dominant bivalve at pearl farms in Cook Islands, often comprising > 90% of the total tissue biomass on culture equipment. Management of this fouling organism places a considerable burden on farm husbandry (Ponia 1997). In Solomon Islands, there was no evidence that algal fouling presented any risk of mortality to oysters. In fact, the presence of an algal covering on shell valves may have prevented successful settlement and growth of hard fouling. Although survival was not threatened by algal fouling, growth of oysters was significantly greater when algae were brushed from chaplets on a 3^1 week cycle. This time interval is similar to that adopted for P. maxima culture in Indonesia and Australia (Gervis and Sims 1992. McGuinness 1994, Taylor et al. 1997). but shorter than that practiced by farmers in the atoll lagoons of Polynesia for P. margaritifera (J. Lyons, pers comm. 1999). Surprisingly, oys- ters cleaned on a 2-week schedule had some of the lowest growth Blacklip Pearl Oysters Suspended on Chaplets in the Solomons 457 rates. In addition, oysters from this treatment required re-drilling more often than oysters cleaned less frequently (K. Friedman. unpublished data 1997). In contrast. Taylor et al. (1997) showed that survival and growth of P. maxima in Indonesia did not differ significantly when cleaned at 2 or 4 week intervals; whereas, re- peated handling results in increased mortality of scallops (Ventilla 1982. Parsons and Dadswell 1992). Anecdotal accounts suggest that the lower growth rates at more frequent cleanings in Solomon islands are attributable to "stressing" of the oysters, although the precise causes are unclear. The number of oysters that became detached from chaplets was a concern, because the deep water (40 m) did not allow easy retrieval of lost oysters, and the sandy substrates in Gizo lagoon made the oysters difficult to find. Another notable difference be- tween atoll lagoons and the open reefs in Solomon Islands was the presence of relatively fast water movement in the growout area (up to 0.15 m s~'). Pearl oyster culture in this relatively high energy environment and regular brushing of chaplets to remove algae may have exacerbated losses from chaplets. Also, smaller sized (< 51 mm DVM) oysters with thinner shells detached from chaplets in greater numbers than larger, thicker shelled oysters. Oysters tied to chaplets in pairs made strong byssal attachments to one another, as is their habit in the wild (Herdman 1903. Gervis and Sims 1992). This behavior secures oysters to the chaplet even if one oyster becomes detached from the monofilament line. Pinctada margaritifera differ from their close relative P. maxima in this regard, because byssal attachment persists in adults (Dou- menge el al. 1991). Examination of re-attachment after byssal threads were severed showed that oysters attempted to make at- tachments even when hung singly, but that successful connections to polypropylene chaplet rope were less common than to other oysters. Because there was evidence that attaching oysters in pairs affected growth negatively, further experiments to find an alterna- tive material for chaplet rope, which is both hard wearing and suitable for byssal attachment, would allow oysters hung singly to make attachments to the rope. In conclusion: 1. Few mortalities resulted from drilling and attaching P. mar- garitifera to chaplets. 2. Growth of P. margaritifera in Solomon Islands compared well to that reported from lagoons in Polynesia; oysters caught as spat (±1 1 mm DVM) reached acceptable size for seeding in 16 months. 3. Oysters on chaplets were fouled quickly by algae. However, the algae did not cause mortalities and may have been ad- vantageous in preventing fouling by cementing bivalves and tubular polychaetes. A 3-4 week cleaning cycle resulted in significantly greater growth of oysters than more or less frequent cleaning regimes. 4. Oysters detached from chaplets in significant numbers; this problem was greatest for the smallest oysters drilled. ACKNOWLEDGMENTS We thank Gideon Tiroba and Ruth and Barley White Dunne for their assistance with experiments. Johann Bell. Sandra Shumway. and an unknown reviewer provided helpful comments on the draft manuscript. This research was conducted as part of the project entitled "Development of Small-Scale Village Farms for Blacklip Pearl Oysters in Solomon Islands Using Wild Spat" funded by the Australian Centre for International Agricultural Research (ACIAR). This is ICLARM Contribution Number: 1476. LITERATURE CITED Alagarswami. K. & A. Chellam. 1976. On fouling and boring organisms and mortality of pearl oysters in the farm at Veppaloidai. Gulf of Mannar. Indian J. Fish. 23 (1-2): 1-22. Anon. 1996. Production of black pearls in Vanuatu commences. South Pacific Comm. Pearl Oyster Bull. 9:40. AQUACOP. 1982. French Polynesia— Country Report, pp. 31-33. In: F. B. Davy and M. Graham (eds. ). Bivalve Culture in Asia and the Pacific. Proceedings of workshop held in Singapore. 16-19 February 1982. International Development Research Centre. Ottowa. Ontario, Canada. Chellam, A., T. S. Velayudhan & A. C. C. Victor. 1987. pp. 72-77. In: K. Alagarswami (ed.l. Pearl Culture Bulletin. Cent. Mar. Fish. Res. Inst. No. 39. Central Marine Fisheries Research Institute, Cochin, India. Coeroli. M.. D. De Gaillande, J. P. Landret & AQUACOP (D. Coatanea). 1984. Recent innovations in cultivation of mollusks in French Polyne- sia. Aquaculture 39:45-67. Dharmaraj, S. K... A. Chellam & T. S. Velayudhan. 1987. Biofouling. boring, and predation of pearl oyster, pp. 92-97. In: K. Alagarswami (ed.). Pearl Culture Bulletin. Cent. Mar. Fish. Res. Inst. No. 39. Central Marine Fisheries Research Institute. Cochin, India. Doroudi. M. S. 1996. Infestation of pearl oysters by boring and fouling organisms in the northern Persian Gull. Indian J. Mar. Set 25: 168-169. Douhilet, D. 1997. Black pearls of French Polynesia. Natl Geograph 191(6):30-37. Doumenge, F., A. Toulemont & J. Branellec. 1991. The South Sea Pearl. The Philippine Golden Pearl. Musee Oceanographique, Monaco. 54 pp. Fassler, C. R. 1995. New developments in pearl farming. World Aquacult. 26 (3):5-10. Freidman. K. J. & P. C. Southgate. 1999. Growout of blacklip pearl oys- ters, Pinctada margaritifera, collected as wild spat in Solomon Islands. J. Shellfish Res. 18:159-167. Friedman. K. J. 1998. Black pearl farms in the Pacific. The "open" reef systems of the Solomon Islands. Austasia Aquacult. 12(3):3-9. Friedman, K. J. & J. D. Bell. 1996. 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). J. Shellfish Res. 15(31:535-541. Friedman. K. J.. J. D. Bell, M. Gervis & G Tiroba. 1996. Progress on research on the potential of farming blacklip pearl oysters in Solomon Islands. South Pacific Comm. Pearl Oyster Inf. Bull. 9:11-13. Friedman, K. J., J. D. Bell & G Tiroba. 1998. Availability of wild spat of the blacklip pearl oyster. Pinctada margaritifera. from "open" reef systems in Solomon Islands. Aquaculture 167:283-299. Gervis, M. H. & N. A. Sims. 1992. The biology and culture of pearl oysters (Bivalva: Pteriidae). ICLARM Stud. Rev. 21. 49 pp. Herdman, W. A. 1903. Report to the Government of Ceylon on the Pearl Oyster Fisheries of the Gulf of Manaar. Part 1. The Royal Society of London. 146 pp. Lintilhac, J. 1987. Black Pearls of Tahiti. Royal Tahitian Black Pearl Book. Papeete. Tahite. 53 pp. Littler, M. M„ D. S. Littler & E. A. Titlyanov. 1991. Comparisons of N- and P- limited productivity between high islands versus low carbonate atolls in Seychelles Archipelago: a test of the relative-dominance para- digm. Coral Reefs 10:199-209. Lucas. J. S., M. Ito. M. Kamatie, J. R. Norton. P. C. Southgate. B. Tioti & J. R. Whitford. 1995. Pacific Island Pearl Oyster Resource Develop- ment Project, pp. 112-115. In: Y. Zhoa, F. Hu, H. Zhou, H. Cui, C. Yao, F. Din. and Y. Lu. (eds.). Proc. 4th Asian Fisheries Forum. Asian Fisheries Society. Manila. McGuinness. R. 1994. Australia's giant pearls. Australian Geographic Ju- ly-September 1994, 35:40-61. 45X Friedman and Southgate Meng, Z. & K. Xing. 1991. The effects of various factors on the nucleus- insertion of the black lipped pearl oysters Pinctada margaritifera Lin- naeus. Oceanologia et Limnologia Sinica, 22:8-13. Mercier, A. & J. F. Hamel. 1998. Pearls in the making. Island Business April: 16-17. Mohammad, M. B. M. 1976. Relationship between biofouling and growth of the pearl oyster, Pinctada fucata (Gould) in Kuwait. Arabian Gulf. Hydrobiologica 51 (21:129-138. Nicholls, A.G. 1931. On breeding and growth rate of the black-lip pearl oyster (Pinctada margaritifera). Report for Great Barrier Reef Com- mission 3:26-31. Parsons, G. J. & M. J. Dadswell. 1992. Effect of stocking density on growth, production, and survival of the giant scallop. Placopecten ma- gellanicus. held in intermediate suspension culture in Passamaquoddy Bay, New Brunswick. Aquaculture 103:291-309. Ponia, B. 1997. Feeding models and chlorophyll levels at black pearl farms in the Cook Islands: implications for management. M.Sc. thesis. James Cook University of North Queensland. Preston, G. 1990. Pearl culture in three French Polynesian atolls 1986- 1987. South Pacific Commission Pearl Oyster Information Bulletin 1:10-12. Remoissenet, G. 1996. From the emergence of mortalities and diseases on Pinctada margaritifera to the effects on the pearling industry, pp. 371- 386. In: Present and Future of Aquaculture Research and Development in the Pacific Island Countries. Proceedings of the International Work- shop held from 20-24 November 1995 at Ministry of Fisheries. Tonga. JICA. Richards. A. H . L. J. Bell & J. D. Bell. 1994. Inshore fisheries resources of Solomon Islands. Mar. Poll. Bull. 29(l-3):90-98. Salvat, B. 1981. Geomorphology and marine ecology of the Takapoto atoll (Tuamotu archipeligo) Proceedings of the 4th International Coral Reef Symposium. Manila. Philippines. Vol. 1. pp. 505-509. Sims, N. A. 1992. Abundance and distribution of the black-lip pearl oyster, Pinctada margaritifera (L.). in Cook Islands, South Pacific. Aust. J. Mar. Freshwater Res. 43:1409-1421. Sims, N. A. 1993a. Pearl oysters, pp. 409-430. In: A Wright (ed.!. Near- shore Resources of the South Pacific. IPS, Suva. 710 pp. Sims, N.A. 1993b. Size, age, and growth of the black-lip pearl oyster, Pinctada margaritifera (L.). (Bivalvia; Pteriidae). J. Shellfish Res. 12(21:223-228. Sims. N.A. & D. J. Sarver. 1995. Feasibility of commercial black pearl culture in Namrick Atoll. Marshall Islands. Final report for Saltonstall- Kennedy Grant Program. 24 pp. Swift. D. R. 1985. Training Manual Aquaculture. Fishing News Books. 135 pp. Taylor, J. J.. P. C. Southgate & R. A. Rose. 1997. Fouling animals and their effect on the growth of silver-lip pearl oyster, Pinctada maxima (Jame- sonl. Aquaculture 153:31-4(1. Ventilla. R. F. 1982. The scallop industry in Japan. Adv. Mar. Biol. 20: 310-381. Ward, F. 1995. Pearls. Gem Book Series, 64 pp. Yukihira, H. 1998. Feeding, energy budgets, and nutritional ecology of the pearl oysters, Pinctada margaritifera (Linnaeus) and P. maxima Jame- son. Ph.D. thesis. James Cook University, Australia. 177 pp. Journal of Shellfish Research, Vol. IS. No. 2. 459-464, 1999. EFFECT OF SPERM DENSITY ON HYBRIDIZATION BETWEEN CRASSOSTREA VIRGINICA, GMELIN AND C. GIGAS (THUNBERG) SUIFEN LYU1 AND STANDISH K. ALLEN, JR.2 Haskin Shellfish Research Lab Institute of Marine and Coastal Sciences Rutgers, The State University of New Jersey Port Norris, New Jersey 08349 ABSTRACT In (his study, effects of sperm density on hybridization between Crassostrea gigas and C. virginica were studied. Two-by-two crosses, C. gigas x C. gigas (GG). C. gigas x C. virginica (GV), C virginica x C. gigas (VG), C. virginica x C. virginica (VV), were conducted. Five levels of spenn density, which were measured by the relative number of attached sperm per egg ( AS/egg). were produced and tested in 3 replicates. At each of the 5 levels (2-3. 5. 10. 20, and 30 AS/egg). the affinity between sperm and eggs, fertilization success, and survival and yield at 48 h were examined and determined. Reciprocal differences in fertilization in GV and VG crosses were found, with GV less successful than VG. In GV crosses, 20 or 30 AS/egg could not be attained, while in VG. GG. and VV crosses all 5 AS/egg levels were accomplished. More sperm were needed to attain a given AS/egg level in GV crosses than in VG, VV, and GG. and likewise, in VG crosses than in VV and GG at more than 10 AS/egg. The affinity between C. virginica sperm and C. gigas eggs was significantly weaker than between C. gigas sperm and C. virginica eggs. The ratio of sperm to eggs at 10 AS/egg was significantly higher than at 2-3 or 5 AS/egg. Survival and yield at 48 h in GV crosses were significantly lower than in GG, VV, and VG crosses. There was no significant difference among the first 3 AS/egg levels in fertilization, survival and yield at 48 h, but embryo deformation was found from higher than 10 AS/egg (> 10%). The results showed that the affinity between heterologous gametes was weaker compared to homologous gametes. We suggest that 2-5 AS/egg was a safe and effective range for both pure crosses and hybrid trials between these 2 species, although more sperm is needed for hybrids. KEY WORDS: Crassostrea virginica, C. gigas, sperm density, fertilization, hybridization INTRODUCTION Hybridization has been used widely in plant, animal, and fish breeding (Briggs 1967. Tave 1986). In oysters, the first hybridiza- tion attempt was made between the Portuguese, Crassostrea an- gulata and the European flat oyster, Ostrea edulis, by Bouchon- Brandely in 1882 (cited in Davis 1950). Since then, many hybrid- izations have been investigated between different species of oysters (Menzel 1987; Gaffney and Allen 1993). While it seems obvious that sperm density can affect fertilization, there is scant information on its effect on hybridization success. For example, to what degree is sperm density responsible for asymmetrical fertil- ization rates in reciprocal crosses of C. gigas and C virginica'} Fertilization success is the first concern in hybridization. Some experiments found normal fertilization in crosses between C gigas (Thunberg, 1793) and C virginica (Gmelin, 1791) (Galtsoff and Smith 1932, Numachi 1977, Allen et al. 1993), versus reciprocal differences in fertilization between C. gigas with C. sikamea (Nu- machi 1977. Banks et al. 1991, Banks et al. 1994) and C gigas and C. rivularis (Allen and Gaffney 1993). Apparently sperm density can also affect early survival. In homologous crosses of C. gigas. higher ratios of sperm to oocytes could increase the fertilization rate from 75% to 100%. but decreased the larval survival (Staeger and Horton 1976). Studies reporting the effect of varying densities of heterologous sperm on fertilization and embryo survival among interspecific hybrids are lacking. The affinity between heterologous gametes — by our definition, the propensity for adhesion of the sperm to the egg — is probably determined by a number of factors. For mollusks, sperm appear to 'Corresponding author, current address: 297-D. Crowells Road. Highland Park, NJ, 08904. 2Current address: Aquaculture Genetics and Breeding Technology Center. Virginia Institute of Marine Science, Gloucester Point, VA 23062-1346. move in a random path to the egg without apparent attraction. As the sperm encounters the egg. it becomes oriented perpendicular to the ovum surface, then quickly binds with the vitelline layer in a process referred to as gamete binding (Hylander and Summers 1977), mediated by a substance called bindin. Bindin has been isolated from oyster sperm as a glycoprotein that attaches to gly- coprotein of the ovum surface (Brandriff et al. 1978). Since gamete binding in bivalves appears to be specific (Hylander and Summers 1977), it is reasonable to expect that gamete binding between heterologous gametes vary more than in homologous crosses. Successful fertilization during hybridization is likely deter- mined by other factors after binding, when the acrosome is ejected from the sperm head. The acrosomal reaction consists of 2 pro- cesses: ( 1 ) the opening of the acrosomal vesicle and the release of lytic enzymes; and (2) the exposure of the acrosomal vesicle pro- cess, a rodlike structure instrumental in gamete fusion. The lytic enzymes are responsible for lysing extracellular moieties sur- rounding the egg and assisting the sperm in its movements to the surface of the egg, and to facilitate gamete fusion. In associate with continuing trials on interspecific hybridization between C. gigas and C. virginica, experiments were designed to answer the following questions: (1) Should more sperm be added to fertilize eggs in hybrid crosses than in the pure crosses? (2) What is the optimal sperm density and how should this be assessed during fertilization trails'? MATERIALS AND METHODS Oysters and Gametes Crassostrea gigas brood stock were obtained from Washington in 1988, spawned, and destroyed. Progeny were held under quar- antine conditions at the Cape Shore hatchery. Oysters used in this work were an F3 generation made in 1992. Sexually mature adult 459 460 Lyu and Allen C. virginica were chosen from Delaware Bay native stocks. Ga- metes were obtained by dissection as follows. All surfaces and instruments contacting the oysters were cleaned and rinsed with hot freshwater between handling and opening of different indi- viduals. Each oyster was opened and a gonad biopsy was taken for determining the sex of the individual by light microscopy. Ga- metes from each oyster were dissected directly into individual beakers. Oocytes were separated from gonadal tissue and other debris by passing the suspension through a 60-u.m nylon screen. The oocytes were caught on a 25-p.m nylon screen (20-|xm for C. virginica ) and resuspended in seawater. The density of the eggs was checked under the microscope, and then estimated by counting aliquots of an appropriate dilution. The eggs of the same female were divided into separate groups of 150 x 103 eggs. Sperm were separated from debris by passing the suspension through a 15-p.m nylon screen. The density of sperm suspensions was first counted with a hemocytometer under a microscope and varied with each dissection. Controlling sperm density in fertiliza- tion was not practical; instead, the number of sperm used in fer- tilization (density of sperm suspension x volume added into 1 beaker of eggs) was recorded. This number was then divided by the amount of eggs in that beaker (150 x 103), and was called the ratio of sperm to eggs. The movement of sperm around eggs was also observed and recorded. Fertilizations were made between 30 and 90 min after dissec- tion so as to confirm that eggs were not self-fertilized and to reduce the susceptibility of oocytes to polyspermy. Stephano and Gould ( 1988) reported that when oocytes were inseminated within 15 min of their mechanical removal from the gonad, they were highly susceptible to polyspermy, even at low sperm density. When sperm were added to eggs, beakers were gently and evenly stirred to promote even distribution of sperm. The number of sperm observed around the perimeter of at least 10 eggs was counted within 2 min of fertilization. This number is called the relative number of attached sperm per egg (AS/egg). The reason we used relative sperm attachment as the experimental variable was 2-fold. First, during fertilization, it was more convenient (and practical) to check the amount of sperm around an egg's perimeter than to count the density of the sperm suspensions. Second, from our previous experience, we have observed considerable variabil- ity in the binding of sperm among crosses in which the same amount of sperm was added, suggesting variation in gamete affin- ity among matings. In this experiment. 5 levels of AS/egg were designated: 2-3, 5 (range 4-6), 10 (range 9-11), 20 (range 18-22) and 30 (range 28-32). To achieve these AS/egg levels in each cross, usually more than 5 beakers of unfertilized eggs were prepared and stored at 4 °C. First, a small amount of sperm was added to 1 beaker of eggs, and the number of sperm that associated with the egg was counted under the microscope within 2 min. If the observed AS/ egg was at the desired level, the beaker of eggs was kept for the experiment; then the next AS/egg level would be attained based on the amount of sperm used in previous trial. Sperm was added to the eggs only once. If the observed AS/egg was not at the desired level, the beaker of eggs would be withdrawn and discarded. The amount of sperm used in each AS/egg level was recorded. The final density of eggs in each group was maintained at 15 x lOVmL for fertilization and 150/mL for 48 h thereafter, without aeration. The temperature for all crosses and cultures was from 23-25 °C with salinity between 25-28 ppt. Experimental Design The purpose of this experiment was to quantify the differences in affinity between sperm and eggs and to measure the effect of sperm density on fertilization. 48-h embryo survival, and yield in reciprocal crosses. Abbreviations for oyster species are as follows: G = C. gigas and V = C. virginica. In our nomenclature, females are listed first. Crosses between C. gigas and C. virginica were produced by a 2 x 2 factorial mating of 1 female and 1 male from each species, i.e., 4 crosses were produced, VV, VG, GV, and GG. and replicated 3 times with different parents in each replicate. Fertilization rates at each AS/egg level in all crosses were assessed by light microscopy on at least 100 oocytes at 60 min post-fertilization (PF). Fertilization was considered successful if oocytes were at or beyond polar body I formation. To observe development, oocytes from each cross were checked before in- semination and 2 min PF, up to 90 min PF. Embryo survival to straight-hinge stage (48 h) was determined by sieving cultures onto a 25-p.m mesh screen and counting the number of larvae of normal appearance. Embryo survival was estimated as the proportion of larvae produced from fertilized eggs at 48 h: 48-h survival eggs) x 100 (No. of straight-hinge at 48 h -r No. fertilized Yield at 48 h was estimated as the proportion of larvae at 48 h produced from the total eggs, therefore not taking into account fertilization rate. This was calculated as: 48-h yield 100 ait. i ilia wa.t LaiLuiaiLU as. (No. of straight-hinge at 48 h h- No. eggs used) x Statistical Analyses In GV crosses, treatment levels of 20 and 30 AS/egg could not be obtained, and data were not available. Therefore, only data from the first 3 levels were analyzed in a 4-way mixed factorial ANOVA model. Dam (fixed) x Sire (fixed) x AS/egg level (fixed) x Replicate (random), using the computer program SPSS version 7.5. Fisher's Least Significant Difference (LSD) was used for mul- tiple comparisons among the first 3 AS/egg levels, when signifi- cant differences (P < .05) were indicated by ANOVA. Data from 20 and 30 AS/egg are only presented in figures. RESULTS Affinity Between Sperm and Eggs When sperm were added to egg suspensions, they swam ran- domly to the eggs. When sperm encountered an egg. they attached to the outer membrane rapidly and remained bound, if the gametes were homologous. In contrast. C. virginica sperm attached less readily to the surface of C. gigas eggs, with most of the sperm only bouncing off the surface of the eggs. C. gigas sperm could easily attach to the eggs of C. virginica. In VG crosses, when additional C. gigas sperm were added to C. virginica eggs, more sperm attached to the surfaces, so that all 5 AS/egg levels were accomplished, including higher AS/egg lev- els of 20 and 30. In contrast, in GV crosses, no matter how many C. virginica sperm were added. C. gigas eggs could not attain 20 or 30 AS/egg. The highest level attained was 18 AS/egg. Sperm Density in Hybridizations of Crassostrea 461 Results from the 4-way ANOVA showed that Sire (P = .022). AS/egg level (P = .01) and their interaction (P = .021) had significant effects on the ratio of sperm to eggs (Table 1, Figs. 1 and 2). There was no significant difference among Dams, and no significant interaction between or among any other 2 or 3 factors {P = .05). Fisher's LSD test showed that the ratio of sperm to eggs in 10 AS/egg was significantly higher than in 2-3 AS/egg (P = .002) or 5 AS/egg (P = .007). while there was no significant difference between 2-3 and 5 AS/egg (P > .05) (Fig. 1 ). Fertilization For fertilization rates, no significant effects were found from any main factors (Dam, Sire, Level and Replicate) or interactions (P = .05) (Table 1 and Fig. 3). Survival to 48 Hours (Embryo Survival) Embryo deformation (< 10%) was found from 10 to 30 AS/egg levels in all 4 crosses and consisted of irregularly shaped embryos and trochophores. Results from ANOVA showed that there was a significant interaction between Dam and Sire (P = 0.010. and the average survival of each cross in the first 3 AS levels. VV = 48%, VG = 41%, GG = 42%, and GV = 12%). but no significant effects from Dam. Sire. Level, Replicate, or interaction (P = .05) (Table 1 and Fig. 4). Yield at 48 Hours As with the 48-h survival, the 48-h yield was significantly affected by Dam x Sire interaction (P = 0.026, ANOVA. and the average yield of each cross in the first 3 AS levels. VV = 44%, VG = 33%. GG = 44%, and GV = 12%), but not other factors (P = .05) (Table 1 and Fig. 5). Figure 1. Mean ratio of sperm to eggs at different As/egg levels in reciprocal crosses of Crassostrea virginica (V) X C. gigas (G). Positive error bars represent standard deviation. DISCUSSION The results in this experiment seem to support species-specific gamete binding in oysters. Gamete binding is specific on pelecy- pods (Hylander and Summers 1977). A glycoprotein, bindin, has been isolated as the substance responsible for the attachment of the sperm acrosome to the glycoproteins of the egg surface ( Brandriff TABLE 1. P-values of 4-way ANOVA, Dam (fixed) x Sire (fixed) x AS/egg level (fixed) x Replicate (random), in the hybridization of Crassostrea virginica x C. gigas. /"-value of 4-Wav ANOVA Treatment Sperm Ratio Fertilization Rate Survival Vield Sources Dam-i Sire-i Dam x Sire-ij Level-k Dam x Level-ik Sire x Level-jk Dam x Sire x Level-ijk Replicate-m Dam x Replicate-im Sire x Replicate-jm Dam x Sire x Replicate-ijm Level x Replicate-km Dam x Level x Replicate- ikm Sire x Level x Replicate- jkm Dam x Sire x Level x Replicate-m(ijk) .141 .022* .184 .010* .3130 .021* .120 .423 .173 .346 .098 .510 .160 .062 .074 .231 .108 .058 .140 .336 .306 .584 .994 .348 .057 .657 .422 .473 .286 .410 .010* .054 .879 .849 .386 .371 .732 .563 .891 .926 .412 .376 .342 .362 .026* .321 .813 .715 .346 .345 .586 .314 .732 .882 .450 .338 Bold P-values with an asterisk indicate significant difference (P < .05). 462 Lyu and Allen Cf 1) V Q. 2000 1500 1000 500 ▼ Sire-G O Sire-V 100 90 80 j= 70 - CO Z 60 " g 50 - | 40 > w 30 20 10 ▼ GV cross v GG cross • VG cross O W cross AS/egg Level Figure 2. Interaction between Sire and AS/egg level in the ratio of sperm to eggs in reciprocal crosses of Crassostrea virginica (V) x C. gigas (G). Sire V represents the average of VV and GV crosses, and Sire G, average of GG and VG crosses. et al. 1978). Our observations and results show that sperm could more easily attach to homologous eggs than to heterologous ones, suggesting the bindin of the acrosome or the glycoprotein of the eggs could be different in C. virginica and C. gigas. We also noticed a significant difference in gamete binding between VG and GV crosses. First, the movement of C. virginica and C. gigas sperm was quite different on the surface of foreign AS/egg Level Figure 4. Mean survival rate ( % I at 48-h at different levels of AS/egg in reciprocal crosses of Crassostrea virginica ( V) x C. gigas (G). Positive error bars represent standard deviation. eggs: C. virginica sperm seemed to be unable to attach tightly onto the surface of C. gigas eggs, but rather bounced around them. In contrast, sperm of C. gigas could bind on the surface of C. vir- ginica eggs relatively quickly. Second, in VG crosses all AS/egg levels, including 20 and 30, were attained, while in GV crosses, only low AS/egg levels obtained with even large amounts of sperm addition. Therefore, hybridization was easier in the direction of VG crosses. Third, a significant interaction was found between se 100 90 80 70 60 50 40 30 20 10 0 GV cross GG cross VG cross W cross 12 3 4 5 AS/egg Level Figure 3. Mean fertilization rate (% I at different levels of AS/egg in reciprocal crosses of Crassostrea virginica (V) x C. gigas (G). Negative error bars represent standard deviation. CO 100 90 80 70 60 50 40 30 20 10 0 T GV cross V GG cross • VG cross O VV cross 12 3 4 5 AS/egg Level Figure 5. Mean yield (9c) at 48-h at different levels of AS/egg in reciprocal crosses of Crassostrea virginica (V) x C. gigas (G). Positive error bars represent standard deviation. Sperm Density in Hybridizations of Crassostrea 463 Sire and AS/egg level on the ratio of sperm to eggs. The interaction between Sire and AS/egg level (Fig. 2) suggests that the effect of V and G sperm on the ratio of sperm to egg varies as AS/egg levels go from low to high. From Figure Litis clear that ratios of sperm to eggs in VV and GG were the same, while VG and GV were quite different. The real cause for significant Sire x AS/egg level interaction is probably that different amounts of sperm were needed in GV and VG to attain a given AS/egg. Reciprocal differences in fertilization of hybrid oyster crosses have been reported previously. In the study of gametic incompat- ibility and genetic divergence in Pacific and Kumamoto (C. sika- mea) oysters. Banks et al. (1994) reported one-way genetic incom- patibilty. C. sikamea eggs x C.gigas sperm formed viable hybrid offspring, but C. sikamea sperm do not fertilize C. gigas eggs. Epifiuorescent staining revealed internal sperm pronuclei in the cross of C. sikamea eggs x C. gigas sperm, but not C. gigas eggs x C. sikamea sperm, and the sperm of C. sikamea were easily removed by washing C. gigas eggs. In contrast, sperm remained bound to eggs after washing in all other crosses. Banks et al. concluded that the complete fertilization failure for C. gigas fe- male x C. sikamea male indicated a discrete block on the molecu- lar mechanism of sperm-egg interaction and fertilization — the ap- parent failure of C. sikamea sperm to undergo the acrosome reac- tion at the C. gigas egg surface. In our study, the reciprocal difference in sperm and egg binding between C. gigas and C. virginica might be explained also by the reciprocal differences in acrosome reaction. Even though the fertilization rates in GV crosses were lower than in VG crosses at the lowest 3 levels of AS/egg. there was no significant effect from the main factors. For the same species. Downing (1989) reported reciprocal differences in fertilization in GV and VG crosses, with GV less successful than VG. Allen et al. ( 1993) reported that fertilization rates in GV and VG crosses were similar, 56% and 62%. Numachi (1977) and Galtsoff and Smith (1932) also attained similar fertilization in their GV and VG crosses. These inconsistent results could be caused by different quality of gametes or non-optimum sperm densities. No sperm density or sperm ratio to eggs were reported. Reciprocal differences in 48-h survival and yield were also apparent between GV and VG crosses, as revealed by the signifi- cant Dam x Sire interaction. Survival and yield in GV crosses were lower than in VG, while GG and VV crosses were similar. In the hybridization experiment of Allen et al. (1993) of C. virginica with C. gigas, survival of 48-h embryos was 42% in GG crosses and 38% in VV crosses. Their results were close to our optimum results: 48% (survival at 48-h) in GG. 51% in VV. But the result of their hybridizations were different from our experiment. Their survival in reciprocal GV and VG crosses were about the same. 45% for GV cross and 35% for VG. In this experiment, reciprocals behaved dissimilarly, with optimum 48-h survival (at 5 AS/egg ) in GV of 21% and VG. 51%. Optimum yields in the 2 pure crosses were obtained when sperm density was at 2-3 and 5 AS/egg. Yields of our pure crosses were higher than the results of Allen et al. (1993). Yields of our hybrid crosses at 48 h were also different from theirs. We had reciprocal differences in yield between GV and VG crosses, 10% and 41%. respectively. Their reciprocal yield rates were similar. 25% in GV and 22% in VG. The difference between these 2 experiments could be caused by the quality of gametes or the density of sperm. In this experiment, deformed embryos (< 10%) were found at higher sperm densities (10-30 AS/egg) in both pure and hybrid crosses. Polyspermy or poor quality seawater resulting from high sperm density could be reasons. In a study of polyspermy in the oyster C. gigas (Stephano and Gould 1988). the incidence of poly- spermy in naturally spawned eggs and in eggs artificially removed from the gonad but incubated in seawater for 1.5 h was much lower than the eggs removed from the gonads and immediately fertilized. They proposed that a polyspermy block is weak or absent in ovar- ian oocytes when they are removed from the gonad but develops during residence in seawater. They found that polyspermy of eggs incubated for 1-1.5 h was 5% using a ratio of 35 sperm/egg; 7% in 250 sperm/egg; 23% in 5.000 sperm/egg, and 69% in 50.000 sperm/egg. Eggs of our experiment were incubated for 0.5 to 1 .5 h, and the ratio of sperm to oocyte was from 50 to 1 2.500. Ac- cording to their results, polyspermy could have occurred in this experiment. What is the optimum sperm density? We suggest 2-5 AS/egg as a safe and effective range for fertilization within and between C. gigas and C. virginica, since no deformation was found at these levels. The amount of sperm should be judged by checking the number of sperm around the eggs under the microscope, not the number of sperm added. To attain lower levels of AS/egg, more sperm will be needed in GV crosses than in VG. VV. and GG; to attain higher levels of AS/egg (> 10). more sperm will be needed in VG crosses than in VV and GG. The affinity between heterolo- gous gametes seemed weaker than with conspecific gametes. AKNOWLEDGMENTS We thank Dr. Patrick M. Gaffney, Dr. Nicholi Vorsa, and Dr. Ximing Guo for advice and editing in this paper. This work was supported by NOAA Oyster Disease Research Program Grant NA47FL-0161. NJAES publication D-32100-xx-97. Marine and Coastal Sciences contribution #97-xx. and VIMS-ABC 002. Allen. S. K. Jr. and P. M. Gaffney. 1993. Genetic confirmation of hybrid- ization between Crassostrea gigas (Thunberg) and Crassostrea rivu- laris (Gould). Aquaculture 1 13:291-300. Allen. S. K... Jr. P. M. Gaffney. J. Scarpa & D. Bushek. 1993. Inviable hybridization of Crassostrea virginica (Gmelin) with Crassostrea rivu- laris (Gould) and Crassostrea gigas (Thunberg). Aquaculture 113: 269-289. Banks. M. A.. D. J. McGoldrick & D. Hedgecock. 1991. Discriminating Kumamoto and Pacific oysters using molecular markers. J. Shellfish Res. 10:511-512 (abstract). Banks, M. A.. D. J. McGoldick. W. Borgeson & D. Hedgecock. 1994. Gametic incompatibility and genetic divergence in Pacific and Kuma- LITERATURE CITED moto oysters, Crassostrea gigas and C. sikamea. Mar. Biol. 121:127- 135. Brandriff. B.. G. W. Moy & V. D. Vacquier. 1978. Isolation of sperm bindin from the oyster (Crassostrea virginica). Gam. Res. 1:89-99. Briggs. F. N. 1967. Introduction to plant breeding. Reinhold. New York. 223 pp. Davis, H. C. 1950. On interspecific hybridization on Ostrea. Science 1 1: 522. Downing. S. L. 1989. Hybridization, triploidy and salinity effects on crosses with Crassostrea gigas and Crassostrea virginica. J. Shellfish Res. 8:447 (abstract). 464 Lyu and Allen Gaffney. P. M. and S. K. Allen, Jr. 1993. Hybridization among Crassoslrea species: A review. Aquaculture 1 16:1-13. Galtsoff. P. S. & R. O. Smith. 1932. Stimulation of spawning and cross fertilization between American and Japanese oysters. Science 76:371- 372. Hylander. B. L. & R. G. Summers. 1977. An ultra-structural analysis of the gametes and early fertilization of 2 bivalve molluscs. Chama macero- phylla and Spisula solidissiua with special reference to gamete binding. Cell Tissue Res. 182:469^489. Menzel, W. 1987. Hybridization of oysters and clams, pp. 47-59. In: K. Tiews (ed.). Selection, Hybridization and Genetic engineering in Aqua- culture. Vol 2. Heeneman, Berlin. Numachi. K. 1977. Japanese species, breed and distribution, pp. 123-126. In: T. Imai (ed.). Aquaculture in Shallow seas: Progress in Shallow Sea Culture. Amerind. New Delhi. India. Staeger, W. H. & H. F. Horton. 1976. Fertilization method quantifying gamete concentrations and maximizing larvae production in Crassos- trea gigas. U.S. National Marine Fisheries Service. Fish. Bull. 74:698- 701. Stephano, J. L. & M. Gould. 1988. Avoiding polyspermy in the oyster (Crassostrea gigas). Aquaculture 73:295-307. Tave. D. 1986. Genetics for fish hatchery managers. AVI, Westport, Con- necticut. 299 pp. Journal of Shellfish Research, Vol. 18. No. 2. 465-473. 1999. COMPARATIVE FIELD STUDY OF CRASSOSTREA GIGAS (THUNBERG, 1793) AND CRASSOSTREA VIRGIN1CA (GMELIN, 1791) IN RELATION TO SALINITY IN VIRGINIA GUSTAVO W. CALVO, MARK W. LUCKENBACH, STANDISH K. ALLEN, JR., AND EUGENE M. BURRESON School of Marine Science Virginia Institute of Marine Science College of William and Mary Gloucester Point, Virginia 23062 ABSTRACT To evaluate and compare the performance of triploid juvenile C. gigas (mean shell height = 19.2 mm) and triploid juvenile Crassostrea virginica (mean shell height = 3 1 .7 mm), 600 oysters of each species were deployed for 1 year in floating mesh cages at three replicate sites within low, medium, and high salinity regimes (respectively, <15%o, 15-25%o. > 25%o) in the Chesapeake Bay and the Atlantic Coast of Virginia. The comparative performance of the two oyster species varied with salinity. At low salinity sites, cumulative mortality of C. virginica (10%) was significantly (P < .05) lower than that of C. gigas (63%), and over-all mean growth rate of C. virginica (2.9 mm mo"') was significantly (P < .05) higher than that of C. gigas (1.6 mm mo"'). At medium salinity sites, survival and growth rate of C. virginica and C. gigas were not significantly (P > .05) different. Both species experienced moderately high cumulative mortality at the medium salinity sites — 35% for C. virginica and 53% for C. gigas — but considerable variation among sites was observed. At high salinity sites, mean cumulative mortality was similarly low (<11%) for both species; whereas, over-all mean growth rate of C. gigas (7.1 mm mo"') was significantly (P < .05) higher than that of C. virginica (3.6 mm mo"'). At all sites, C. gigas was less susceptible than C. virginica to Perkinsus marinus infections. Infections by Haplosporidium nelsoni were present in C. virginica and absent in C. gigas. Infestations by mud-worm Polydora spp. were more prevalent and severe for C. gigas than for C. virginica at low and medium salinity sites in October 1997. but similar for both species at other times and locations. Condition index was significantly (P < .05) higher for C. virginica than for C. gigas at low salinity in May 1998. but similar for both species for other times and locations. Crassostrea virginica outperformed C. gigas in low salinity sites in the Chesapeake Bay. C. gigas outperformed C. virginica at high salinity sites in the Atlantic Coast, and performance was similar for both species at medium salinity sites in the Chesapeake Bay. KEY WORDS: Crassostrea gigas, triploid, growth, survival, disease susceptibility. Virginia INTRODUCTION As native eastern oyster, Crassostrea virginica (Gmelin, 1791 ) stocks have declined throughout much of the mid-Atlantic sea- board of the United States through overharvesting, disease, and water quality deterioration, interest in the potential of non-native oyster species to restore the fishery and ecological functions has grown. This has been particularly apparent in the Chesapeake Bay region, where standing stocks of eastern oysters have been reduced in the last decade to lr/r of late nineteenth-century levels (Newell 1988). Given that much of this decline has been caused by dev- astating Dermo and MSX epizootics resulting from, respectively. the protozoan parasites Perkinsus marinus and Haplosporidium nelsoni (Burreson and Ragone Calvo 1996), strategies aimed at rehabilitation of stocks largely depend upon the use of disease- resistant oysters. Although development of eastern oyster lines with resistance to MSX has been achieved (Ford and Haskin 1987) and development of lines with resistance to both Dermo and MSX is in progress (Ragone Calvo et al. 1997), applicability of selective breeding programs is mostly limited to aquaculture. Use of dis- ease-resistant eastern oysters for fishery enhancement or ecologi- cal restoration is constrained by dilution of their gene pool with that of susceptible oysters in the wild. Furthermore, the gene flow from relatively uninfected and highly susceptible populations in low salinity areas may limit the evolution of resistance in eastern oysters (Gaffney and Bushek 1996). The Pacific oyster, Crassostrea gigas (Thunberg, 1793), has been the species of choice to substitute for depleted local oyster populations decimated by disease and other factors in many coun- tries (Mann et al. 1991. Shatkin et al. 1997). Crassostrea gigas is the primary oyster species supporting shellfish industries around the globe, accounting for an estimated 80% of the world oyster production (Chew 1990). Shatkin and collaborators (1997) re- viewed the worldwide experience with introductions of C. gigas and presented an analysis of economic, legal, and ecological fac- tors relevant for introductions into the Gulf of Maine. Experience with the transfer of C. gigas beyond its native range in the Indo- Pacific coast of Asia, particularly in Japan, has been considered both successful and problematic. For example, transfer of C. gigas to the Pacific Northwest region of the United States has restored the shellfish industry that used to rely on the native oyster Ostrea lurida (Chew 1990). Transfer of C. gigas to France has rehabili- tated the industry by substituting for Crassostrea angulata. which was decimated by a viral disease (Grizel and Heral 1991). Prob- lems with the transfer of exotic oysters include parallel transfer of pests and disease agents and undesired competition of exotic spe- cies with their native counterparts. For example, spread of the viral disease affecting C. angulata in France has been correlated with the introduction of C. gigas, which was conducted in bulk and without proper measures for disease prevention (Andrews 1980. Grizel and Heral 1991). Following transplantation into southeast- em Australia. C. gigas successfully reproduced and displaced the native oyster, Saccostrea commercialis. from some of its habitat (Chew 1990). During the last decade, the possible introduction of C. gigas into the Chesapeake Bay has received considerable attention. Mann and collaborators (1991) developed the rationale and ana- lyzed the risks associated with such an introduction. Gottlieb and Sch weighofer ( 1 996 ) further discussed the potential of C. gigas for restoring the Chesapeake Bay ecosystem. In Virginia, a program to 465 466 Calvo et al. examine the suitability of nonindigenous oyster species to local conditions was established, while efforts to restore native oysters continued (VIMS 1996). Based upon ecological requirements and disease tolerance, two candidate nonindigenous oyster species within the genus Crassostrea. C. gigas and the Suminoe oyster, C. ariakensis (= rivularis) (Fujita. 1913) were initially selected for testing in the Chesapeake Bay (Mann et al. 1991, VIMS 1996). In this paper, we address field studies with C. gigas. No growth or disease challenge studies are available for C. ariakensis in the region; however, for locations on the West Coast of the United States. Langdon and Robinson (1991) reported growth rates simi- lar to that of C. gigas. Studies with C. ariakensis. currently un- derway at Virginia Institute of Marine Science (VIMS), will be the object of a future report. Both Mann et al. (1991) and Gottlieb and Schweighoffer (1996) have suggested that C. gigas has considerable potential for resto- ration in part of the Chesapeake Bay, but both indicated the need for more research. The need for field studies was particularly emphasized to assess the performance of exotic oysters under local conditions, and because there was no alternative way for challenge against MSX. Prior studies at VIMS indicated that C. gigas was more resistant to protozoan pathogens than the native oyster, at least under some environmental conditions. In laboratory disease challenge experiments with P. marinus, C. gigas exhibited lower disease prevalence and intensity and had lower mortality than C. virginica (Meyers et al. 1991, Barber and Mann 1994). A field challenge experiment conducted in the York River using triploid oysters also indicated that C. gigas had reduced susceptibility to P. marinus and H. nelsoni as compared to the native oyster (Burreson et al. 1994). In this field study, which lasted only 5 months, C. gigas had comparable shell growth rates to the native oysters, but became heavily infested by the polychaete Polydora websteri, re- sulting in poor meat quality. However, these studies were limited in duration and spatial extent, and more extensive field experi- ments were necessary to evaluate the performance of C. gigas better within a broader range of salinity and other environmental conditions. The present study was designed to ( 1 ) test the hypoth- esis that comparative performance of C. gigas and C. virginica would vary with salinity, (2) compare disease susceptibility in the same two species across salinity regimes, and (3) compare infes- tations by shell-boring organisms (e.g., mud worms and boring sponges). METHODS Study Sites Nine sites were selected on the basis of several criteria, includ- ing salinity regime, geographic location, available information on oyster growing conditions and water quality, safety, logistics, and relevance for the oyster industry. Sites were established at tripli- cate locations within low salinity (<15%25%c) areas (Fig 1). Low and medium salinity sites were established near the margins of rivers (Corro- toman, Great Wicomico, Coan, and York): or in shallow creeks surrounded by marshes (Woodas Creek, a tributary of the East River, and Nandua Creek). High salinity sites were located in well-flushed narrow channels surrounded by marshes and mudflats in the coastal lagoon system of the Atlantic Coast of Virginia. Temperature and salinity were measured during monthly site visits with a stem thermometer and a refractometer. To character- ize environmental variables further, hourly temperature, salinity. 77° 76° Figure 1. Location of study sites in the Chesapeake Bay and the At- lantic Coast of Virginia. A Low salinity (<15 ppt) sites, • medium salilnit) (15-25 ppt) sites, ■ high salinity (>25 ppt) sites. and turbidity were measured with Hydrolab-Minisonde® datalog- gers deployed at various sites for weekly to monthly intervals. Oyster Groups To ensure that this study resulted in neither the unintended reproduction of C. gigas nor the introduction of potential exotic pathogens, we used triploid oysters produced from progeny of quarantined brood stocks, in accordance with protocols developed by the International Council for the Exploration of the Seas (ICES). Triploid C. gigas (3CG) and triploid C. virginica (3CV) were produced for this study by Haskin Shellfish Research Labo- ratory (HSRL) during June to July 1996 (Table 1). Brood stock for 3CG was Miyagi strain C. gigas originating from the Pacific Northwest Coast of the United States and maintained in quarantine at HSRL for several generations. Triploid C. gigas were produced by mating tetraploid and diploid parent stocks, an approach that results in complete triploidy of progeny (Guo et al. 1996). Brood stock for 3CV was a Delaware Bay strain naturally selected against P. marinus and H. nelsoni in Delaware Bay. Triploidy in C. vir- ginica was chemically induced by treatment of fertilized eggs with TABLE 1. Oyster groups used. Species Group Code Hatchery Date Spawned Size in May 1997u C. xigas C. virginica 3CG 3CV HSRL HSRL 16 July 96 1 1 June 96 19.2 mm 31.7 mm y to group codes: 3 = triploid. CG = C. gigas, CV = C. virginica. 1 Mean shell height at the time of deployment. Key to ; Field Study of C. Virginica and C. Gigas 467 TABLE 2. Percentage market size (>76.2 mm) oysters in May 1998, based on the legal size for wild harvested oysters in Virginia. Oyster Group Salinity Regime 3CV 3CG Low Medium High 14% (38/268) 0%(0/69) 41% (65/159) 11% (10/91) 52% (13 1/252) 100% (260/261 ) Oyster group codes described in Table 1 . In parenthesis, number of market size oysters/total number of live oysters. cytochalasin-B using the methods described by Downing and Allen (1987) and Allen et al. (1989). Experimental Design Until field deployment in May 1997, juvenile 3CG were main- tained first in flow-through tanks with ambient Delaware Bay wa- ter and quarantined effluents at HSRL Cape Shore, NJ, and then with York River ambient water and quarantined effluents at VIMS Gloucester Point, VA. Juvenile 3CV were also maintained first at HSRL Cape Shore, NJ. and then at Gloucester Point, VA in flow- through tanks without quarantined effluents. Between 28 April and 16 May 1997, oysters were dispensed into triplicate 3.2-mm mesh bags and placed within individual floating trays at selected sites as described below. There were 200 oysters per bag and 600 oysters per floating tray. Floating trays (2.3 m x 0.5 m x 0.3 m) were constructed by fitting wire mesh trays (25-mm square 16-gauge mesh) into floating frames built with 4-inch (10.16 cm) PVC pipe, following the design of Luckenbach and Taylor (1997). Floating trays were cleaned of fouling organisms at least once a month during regular site visits and more often if necessary. All sites were visited monthly (±10 days). As oysters grew, they were transferred from 3.2-mm mesh bags to 9.5-mm mesh bags in July 1997. In March 1998, when 3CG at high salinity sites approached space limitation within bags, all oyster groups at high salinity sites were split by placing half of the oysters into new bags. Oysters in the new bags were placed in a float adjacent to the original one. A full factorial design, with three replicate sites within each of the three salinity regimes, was employed to examine the effects of triploid C. virginica and C. gigas (species), salinity regime, and time on final cumulative mortality, final condition index, preva- lence and weighted prevalence of P. marinus, and weighted preva- lence of Polydora spp.. Differences in mean variables, between species within salinity regime, between salinity regimes within species, and between times where appropriate, were further exam- ined by Newman-Keuls test (Zar 1974). Data were examined for compliance with analysis of variance (ANOVA) assumptions us- ing Bartlett chi-square test for homogeneity of variance and plots of means versus standard deviations. Arcsine and logarithmic transformations were used where appropriate (Zar 1974). Mortality, Growth, and Condition All live and dead oysters within each float were counted monthly to determine survival. Monthly mortality for each oyster group was calculated as the number of oysters that died during each month interval divided by the number of live oysters at the beginning of the interval, corrected for oysters removed by sam- pling. Cumulative mortality of each oyster group was calculated as the sum of interval mortality (Barber and Mann 1994, Krebs 1972). To follow growth. 100 oysters within each float were individu- ally labeled, and shell height was repeatedly measured to the near- est 0.1 mm, using calipers, once monthly, except January and February 1998. Mean monthly growth rates for individual oysters were calculated as the over-all shell height increment divided by the deployment time in days standardized for 30 days. To provide a measure of production potential, the proportion of individually labeled oysters that attained Virginia legal market size for wild stocks (3 in = 76.2 mm), within each salinity regime, was calcu- lated at the end of the experiment. Whole weight, shell weight, and tissue wet and dry weights were measured on the same oysters (n = 25) collected for disease diagnoses in October 1997 and May 1998. Following Lawrence and Scott ( 1982). condition index (CI) was calculated, by the for- mula: CI = tissue dry weight/(total weight - shell weight). ( 1 ) Oysters were allowed to air dry for 15-20 min before weighing, and whole oyster weight was recorded to the nearest 0.0 lg. Oys- ters were then shucked, shells weighed to the nearest 0.0 lg. and wet tissues were gently rolled on a paper towel and weighed on pre-tared vessels to the nearest 0.00 lg. Wet tissues were dried at 80 °C overnight, and tissue dry weight was measured the next day to the nearest 0.001 g. Diseases and Polydora A baseline sample (n = 25) was taken to assess the disease status of oyster groups before deployment in the spring of 1997. Subsequent disease samples (n = 25) were collected, depending upon group and site, during the summer and fall of 1997 and the spring of 1998. Perkinsus marinus was diagnosed using Ray's fluid Thioglycollate medium (RFTM) assays (Ray 1952) on com- bined mantle, gill, and rectum tissue. Infection intensity was rated based on Ray (1954) and Mackin (1962), and for the calculation of weighted prevalence, the following numerical values were as- signed to intensity categories: (1) light: (3) moderate: and (5) heavy. Weighted prevalence was calculated by the formula. Weighted prevalence = ((n, * I) + (n2 * 3) + (n3 * 5))/N, (2) where n, = number of cases rated as (z), N = total number of oysters examined in the sample. Haplosporidium nelsoni was diagnosed using standard paraffin histology procedures with oysters preserved in Davidson's AFA and 6-u.m tissue sections stained with Harris' hematoxylin and eosin (Burreson et al. 1988). Infection intensity was rated as light, moderate, and heavy based on Burreson et al. (1988). Histological sections were also used to document the presence of other parasites and to examine development of oyster gonads. All disease and histology analyses were performed by VIMS Shellfish Pathology Laboratory. The spionid polychaetes Polydora websteri and P. ligni are commensal with bivalves, including oysters. These suspension- feeding worms do not feed on the oyster, but the mechanical irritation caused by their presence causes the oyster to lay down additional layers of conchiolin over the worm's tube in what are often termed mud-blisters. At sufficiently high levels of infesta- tion, this can severely limit the growth of oysters and reduce their 468 Calvo et al. Date/Salinitv TABLE 3. Percentage genetic mosaics among C. gigas by salinity regime and date. Low Medium High Row Total 2-10 June 97 30 June-9 July 97 28 July-5 August 97 6-15 April 98 4-7 May 98 Column total 0.0% (0/105) 0.0% (0/105) 4.7% (5/105) 5.0% (3/60) 6.1% (20/325) 4.0% (28/700) 0.0% (0/105) 2.8% (3/105) ()»', (1/105) 8.391 (8/96) 1.7r-r (4/233) 2.591 (16/644) 0.0% (0/105) 0.0% (0/105) 0.0% (0/105) 4.8% (5/105) 2.5% (9/358) 1.8% (14/778) 0.0% (0/315) 0.9% (3/315) 1.9% (6/315) 6.1%- (16/261) 3.6% (33/916) 2.7% (58/2122) In parenthesis number of mosaics/number of oysters examined. condition index. Examination for mud-blisters associated with Polydora spp. was conducted on the same oysters collected for disease diagnoses in October 1997 and May 1998. Worms were not identified to species, but Polydora websteri is the most com- mon species affecting oysters in the northeast coast of the United States (Blake and Evans 1972, Wargo and Ford 1993). The internal surface of right valve shells was visually inspected and rated ac- cording to the presence and extent of mud-blisters. Examination was restricted to right valves as in Wargo and Ford (1993), who reported that infestations by Polydora spp. were equally found in right and left valves. Following the methods of Handley and Bergquist (1997), infestation was rated as: (0) no visible mud- blisters or any evidence of boring by Polydora spp.: (1) mud- blisters affecting less than 25c/c of the valve; (2) 25-50% of the valve affected; (3) 50-75% of the valve affected; or (4) more than 75% of the valve affected. Weighted prevalence was calculated by the following formula. Weighted prevalence = ((n, * 1 ) + (n2 * 2) + (n, * 3) + (n4 * 4))/N, (3) where n, = number of cases rated as (/), N = total number of oysters examined in the sample. Reproductive Status and Ploidy Before deployment, baseline samples of 3CV (n = 125 larvae) and 3CG (n = 35 juveniles) were taken to confirm ploidy status. During deployment, samples of 3CG (n = 35) were collected, depending upon site, at the beginning of the month in June, July, and August 1997 and May 1998. Only C. gigas was examined for ploidy during deployment, but an equal number of C. virginica were concurrently collected from trays to standardize the number of oysters removed by sampling. Ploidy was determined by flow cytometry of gill biopsies from individually labeled oysters. When gill tissues were found to contain any diploid cell (a condition termed mosaic), a biopsy of the gonad was examined by flow cytometry, and the remaining gonad tissue was processed by his- tology. Ploidy assays were conducted at HSRL and the VIMS Aquaculture Genetics and Breeding Technology Center. RESULTS Environmental Parameters Salinity was within the range established for low. medium, and high salinity sites for most of the monthly measures (Fig. 2). Low salinity sites experienced relatively high mean salinity (>15 %c) during September. October, and November because of drought conditions during the summer and relatively low mean salinity (<10%<) during March, April, and May because of high rainfall during the winter. The Coan River site experienced extreme low salinity with mean daily values of 3%c during April and May. Medium salinity sites experienced relatively low salinity (<15%o) during March, April, and May (Fig. 2). Temperature followed similar seasonal trends at all sites with a maximum of 27-29 °C in July and a minimum of 3-6 °C in March. High salinity sites experienced over-all cooler temperature with monthly means 2-A °C lower than medium or low salinity sites (Fig. 2). 30 LOW SALINITY REGIME ^ ?S o Ol 20 D ro 1b I^v - 3 — -D g 15- ■ \ -J Y " 0) 5 T Temp Ol 0 13 r ; 0- | . III1 1 30 25 20| 15.~ 'E >! 35- HIGH SALINITY REGIME — 30- O "S"25- ,5- -D- •n^ = 20- ro 4> 15- Q. E10- * 5- I ■-5 ''S.. i ,5 3' oJ i III!! 35 30 , , 25 n a 20 1 — ' >. 15 c 10 ro JJASOND-MAM Figure 2. Mean monthly (± SI)) temperature and salinity of three sites within low, medium, and high salinity regimes, using stem thermom- eter. * Break in monthly sampling Field Study of C. Virginica and C. Gigas 469 Turbidity, measured in Nephelometric Turbidity Units (NTU). was highest at the medium salinity Nandua Creek site and Woodas Creek site. Maximum daily mean turbidity at Nandua Creek and Woodas Creek was. respectively. 436 NTU and 149 NTU. and maximum daily mean values at other sites was <38 NTU. Mortality Species, salinity regime, and their interaction had significant (P < .05) effects on cumulative mortality. At low salinity sites, mean monthly mortality of 3CV was very low (<3%) at all times, and that of 3CG peaked at 28% in April 1998 (Fig. 3). By May 1998, mean cumulative mortality of 3CV (10%-) was significantly (P < .05) lower than that of 3CG (63%). At medium salinity sites, mean monthly mortality reached 17% for 3CV and 22% for 3CG in October 1997 (Fig. 3). By May 1998. mean cumulative mortality of 3CV (35%) was not significantly (P > .05) different than that of 3CG (53%). High variability in mortality, for both species, among medium salinity sites was attributable to extremely high mortality at Nandua Creek. At high salinity sites, mean monthly mortality LOW SALINITY REGIME 80 >. 80 £60 >-40 .c 5 20 HIGH SALINITY REGIME □ 3CV 3CG 80 60 JJASOND-MAM Figure 3. Monthly and cumulative mortality of triploid C. virginica (3CV) and triploid C. gigas (3CG) from June 1997 through May 1998. Bars = mean (+ SD) monthly mortality of three sites within salinity regimes. Dashed lines = mean cumulative mortality of 3CY. Solid lines = mean cumulative mortality of 3CG. * Break in monthly sampling. was very low (<37r ) for both species at all times (Fig. 3). In May 1998, mean cumulative mortality of 3CV (1 1%) was not signifi- cantly (P > .05) different from that of 3CG (4%). Within 3CV. there were no significant (P > .05) differences in mean cumulative mortality among salinity regimes. Within C. gigas, oysters at low and medium salinity experienced significantly (P < .05) higher mortality than those at high salinity, and no significant (P > .05) difference was detected between oysters at low and medium sa- linity. Growth At the initiation of the experiment, mean size of 3CV and 3CG was, respectively. 31.7 mm and 19.2 mm; subsequent growth var- ied with salinity regime (Table 2). At low salinity, 3CV increased its initial size advantage over 3CG. resulting in a mean shell height of 67.8 mm for 3CV and 41.1 mm for 3CG at the end of the study (Fig. 4). At medium salinity, the size differential between species was maintained throughout the study yielding a final mean shell height of 74.1 mm for 3CV and 65.1 mm for 3CG (Fig. 4). At high salinity, the initially smaller 3CG reached the same size as 3CV 3 mo after deployment, in July 1997, and continued to grow during fall and winter attaining a final mean shell height of 108.1 mm in D 3CV Market size 3" ....■5- ...-D x_ LOW SALINITY REGIME — • — 3CG n - i —9. x 1 1 1 1 1 1 1 I M J J A S O ND-M A M MEDIUM SALINI" 3CG I IT Q t rY REGIME D 3CV •— Market size 3" _ ...5 5 • ...a -' * — 4 — \ 1 1 — i i FP rT — 1 1 1 1 — i MJJASOND-MAM 120 m fc inn 40 > F n 80 3 .c 20 E 3 o '5 X 60 40 ~5 o °^ 0) .c 20 HIGH SALINITY REGIME D 3CV - — • 3CG x_I^— \ 1 1 y^Z. — : £}■ Q.-^Q-.^Q^Q^—i^- ■Us' D J* 1 r i i i i i i i i N M M M J J A S Figure 4. Monthly shell height of triploid C. virginica (3CV) and trip- loid C. gigas (3CG) from May 1997 to May 1998. Mean (± SD) of three sites within salinity regimes. 470 Calvo et al. May 1998. By comparison. C. virginica stopped growing after October 1997 and reached 78.4 mm in May 1998 (Fig 4). Species. salinity regime, and their interactions had significant (P < .05) effects on mean growth rate. At low salinity sites, mean overall growth rate of 3CV (2.9 mm mo1) was significantly (P < .05) greater than that of 3CG ( 1.6 mm mo"' ). with most of the growth in C. virginica occurring between July and October (Fig. 4). At medium salinity sites, mean over-all growth rate for both species (3.0 mm mo"1) was not significantly (P > .05) different, and the monthly pattern of growth was similar. At high salinity sites, mean over-all growth rate of 3CV (3.6 mm mo-1) was significantly (P < .05) lower and nearly half that of 3CG (7.1 mm mo-1). Within 3CV. growth rate did not significantly (P < .05) differ between salinity regimes. Within C. gigas, growth rate at high salinity was significantly (P < .05) higher than that at medium and low salinity regimes, and growth rate did not significantly I/5 > .05) differ between medium and low salinity regimes. Condition Index Salinity regime, time, and the interactions of salinity and spe- cies and salinity and time had significant (P < .0005) effects on final oyster condition. In October 1997. there were no significant (P > .05) differences in condition index between species within any salinity, or between salinities within a species (Fig. 5). In May 1998. at low salinity, mean condition index of 3CV (16.2%) was significantly (P < .05) higher than that of 3CG (8.7%); at other salinities, no significant (P > .05) differences were detected be- tween species. Within species, condition index increased signifi- cantly (P < .05) with salinity, except for C. gigas between medium and high salinity in May 1998. For both species within any salin- ity, except for C. gigas within low salinity, condition index in- creased with time. Mean condition indices for oysters at Nandua Creek and Woodas Creek were lower than those of oysters at the third medium salinity site (York River). Relative to whole oyster weight, shells of C. virginica were heavier than shells of C. gigas. For all samples combined, the percentage of shell weight relative to whole weight was 66% in 3CV and 57% in 3CG. Proportional shell weight remained fairly constant for 3CV at low, medium, and high salinity, between Oc- tober 1997 and May 1998, while it decreased in 3CG at low and medium salinity and increased in 3CG at high salinity. Disease Species, salinity regime, time, and the interaction of species and time had significant (P < .05) effects on prevalence and weighted prevalence of P. minimis infections. Higher prevalence and intensity of infections were observed in C. virginica and oc- curred at medium salinity during fall as compared to C. gigas and to other salinity regimes and times (Fig. 6). Infections in C. vir- ginica were low in prevalence and intensity during the first spring and summer of deployment and subsequently increased in the fall (Fig. 6). Infections in C. gigas were generally of low magnitude at most sites and times; however, infections at the Nandua Creek site in fall reached 67% prevalence with two heavy intensity infections. Maximum mean weighted prevalence for C. gigas (0.4) was sig- nificantly (P < .05) lower than that for C. virginica (1.4). At medium salinity sites, infections remained high in C. virginica during spring 1998 (prevalence >62%, weighted prevalence = 0.9), whereas, at low and high salinity sites, infections subsided in 30 x ■§20 c "15 o 5 10 ■a 8 5 o 0 LOW SALINITY REGIME D 3CV 3CG October 97 May 98 MEDIUM SALINITY REGIME October 97 30 g^25 x £20 c "15 o = 10 ■D § 5 O 0 May 98 HIGH SALINITY REGIME October 97 May 98 Figure 5. Condition index in triploid C. rirginica (3CV) and triploid ('. gigas (3CG). Mean (+ SD) of three sites within salinity regimes. spring 1998 (mean prevalence < 23%, mean weighted prevalence = 0.1-0.3) (Fig 6). Haplosporidium nelsoni was absent in C. gigas but was present at low prevalence (< 16%) in 3CV at medium and high salinity sites. At low salinity, no infections were detected in any of the samples. Polydora Mean prevalence of infestations by Polydora spp. was high 095%) for 3CV and 3CG at low and medium salinity sites re- gardless of time. At high salinity sites, however, although mean prevalence for 3CV remained at 64%. it decreased for C. gigas from 52% in October 1997 to 12% in May 1998. Differences in weighted prevalence between oyster species were more pro- nounced than differences in prevalence. Species, salinity regime, time, and the interaction of salinity regime and species had significant (P < .0005) effects on mean weighted prevalence. Triploid C. virginica had significantly (P < .05) lower weighted prevalence than C. gigas at medium and low salinity sites in October and similar levels of Polydora spp. infes- tation at all other times and locations (Fig. 7). For 3CV, within any salinity, mean weighted prevalence was not significantly (P > .05) different between October and May, whereas, for 3CG at low and Field Study of C. Virginica and C. Gigas 471 2.0^ 0) o c a 1.5 > 0) a 1.0 ■a ai LOW SALINITY REGIME LOW SALINITY REGIME □ 3CV 3CG 5 0.5 0.0 r^i Q M J-J S-0 A-M 2.0 ai u c ^ 1.5 > o a. 1.0 ■o S MEDIUM SALINITY REGIME id 5 0.5 0.0 2.0 a> u c ^ 1.5 n > £ 1.0 ■a □ 3CV ■ 3CG T " T - ■ — ■ Q , ■ 0 M J-J S-0 A-M HIGH SALINITY REGIME a; 5 0.5 0.0 □ 3CV ■ 3CG T 0 -* r- T -2— M J-J S-0 A-M Figure 6. Intensity of P. marinus in triploid C. virginica (3CV) and C. gigas (3CG) from April 1997 through May 1998. Mean (+ SD) of three sites salinity regimes. medium salinity, mean weighted prevalence significantly (P < .05) decreased from October to May. Within 3CG, at high salinity, mean weighted prevalence was not significantly (P > .05) different between October and May. Ploidy Baseline samples confirmed 1007r triploidy among naturally induced triploid C. gigas and revealed 85% triploidy among chemically induced triploid C. virginica. The proportion of C. gigas gill samples in which combinations of diploid and triploid cells (mosaics) were detected by flow-cytometry varied with time and salinity (Table 3). The proportion of mosaics, pooled for all salinity regimes, increased from 0.0% in June 1997 to 6.1% in April 1998. and then decreased to 3.6% in May 1998. The pro- portion of mosaics, pooled for all times within low, medium, and high salinity, was respectively, 4.0%. 2.5%, and 1.8%. For all samples collected during the study combined, regardless of salin- ity, the over-all proportion of mosaics was 2.7%. Examination of 23 oysters with mosaic gill cells revealed that 5 were females. 15 were males, and 3 were undifferentiated. Among oysters with mosaic gill cells, there was one individual in which haploid cells were detected in a gonad biopsy (a male col- lected in Bogues Bay on 14 April 1998). Concerns over the po- October 97 May 98 MEDIUM SALINITY REGIME October 97 5 0) u §4 50 May 98 HIGH SALINITY REGIME October 97 May 98 Figure 7. Intensity of Pulydora spp. infestations in triploid C. virginica (3CV) and triploid C. gigas (3CG). Mean (+ SD) of three sites within salinity regimes. tential reproduction of C. gigas following the finding of an indi- vidual oyster with potentially haploid gametes, resulted in termi- nation of the experiment. By 6 May 1998. all C. gigas were removed from the water and maintained in quarantine conditions at VIMS. DISCUSSION This study demonstrated that the comparative performance of C. virginica and C. gigas in the Chesapeake Bay and the Atlantic Coast of Virginia varied with salinity regime. At low salinity, survival, growth rate, final condition index, and resistance to in- festations by Polydora spp. were significantly greater for C. vir- ginica than for C. gigas. However. C. virginica was more suscep- tible than C. gigas to P. marinus infections. High mortality (63%) and poor growth (1.6 mm mo-1) observed for C. gigas at low salinity sites were not surprising considering the previously re- ported optimal salinity of 35%c for growth in this species (Mann et al. 1991). High mortality of C. gigas at the low salinity Coan River site in April (56%) can probably be attributed to a prolonged period of extreme low mean daily salinity (3%t for 1 month). Most of the growth for C. virginica and C. gigas occurred in the spring subsequent to deployment. 472 Calvo et al. At low and medium salinity, shells of C. gigas with severe Polydora spp. infestations were very fragile and often disinte- grated during monthly inspections of labeled individuals for growth. The decrease in the severity of Polydora spp. infestations between October 1997 and May 1998, primarily for medium and high salinity sites, can be attributed to oyster shell repair. In May 1998 nacre shell deposits were often observed to cover blisters. Comparing shell weight for oysters of similar size. Barber and Mann (1994) found that shell weight was significantly {P < .05) greater for similar sized C. virginica than C. gigas. Similarly, in the present study. C. virginica had heavier shells proportional to whole oyster weight relative to C. gigas. It is possible that the relatively thinner shells of C. gigas made it more susceptible to heavy Polydora spp. infestations. At medium salinity sites, mean cumulative mortality, growth rate, and final condition index of C. virginica were not signifi- cantly different than that of C. gigas. Crassostrea gigas was more susceptible to infestations by Polydora spp. and less susceptible to P. marinus than C. virginica in this salinity regime. Both C. vir- ginica and C. gigas experienced a high variability in mortality and growth rate because of extremely poor performance at Nandua Creek, relative to the other two medium salinity sites. High mor- tality and poor condition of C. virginica and C. gigas at Nandua Creek can be attributed to prevalent and severe P. marinus infec- tions. Oysters at Nandua Creek, and to a large extent at Woodas Creek, experienced the most prevalent and severe P. marinus in- fections recorded in this study. We speculate that high density of other oyster lots present in the immediate vicinity of the experi- mental oysters, coupled with relatively poor water exchange and high turbidity, resulted in high disease pressure and environmental stress at those sites. Barber and Mann ( 1994) reported greater growth rates for C. gigas than C. virginica at the York River site, although this study did not find significant differences in growth of the two species at the site. This incongruity may arise from different environmental conditions at the site between years or from differences in the timing of spawns and handling of oysters between the studies. Furthermore, the experiment of Barber and Mann ( 1994) involved exposing diploid oysters to unfiltered York River water in quar- antined tanks, while our study was conducted in situ with triploid oysters deployed within mesh cages. Growth rate of C. gigas at high salinity in the present study was higher than that reported in other studies for high salinity envi- ronments. In a study of C. gigas growth at Seto Inland Sea in southern Japan where temperature ranged from 8-30 °C (Koba- yashi et al. 1997). oyster shell height increased from 27.0 to 93.1 mm between May 1990 and January 1991. Studies with C. gigas in Canada and Korea reviewed by Kobayashi et al. (1997). reported similar growth rates. By comparison at high salinity sites in the present study, where temperature ranged from 4-27 °C, shell height of C. gigas increased from 19.2 to 101.6 mm between May and December 1997. Higher growth rates of C. gigas in the present study may be attributed to the use of triploid oysters; whereas, diploid oysters were used in the other studies cited above. In general, because gametogenesis is restricted in triploid oysters, more energy is available for somatic growth. Allen and Downing ( 1986) and Davis ( 1989) indicated that increased growth in triploid C. gigas mostly occurred during the normal reproductive season. Additional factors that would explain the difference in growth among C. gigas between studies may include different environ- mental conditions among study areas and times. In summary, during the course of the study C. gigas performed no better than C. virginica at low and medium salinity sites in the Chesapeake Bay. However, considering the large variability in performance between the two oyster species among medium sa- linity sites and given the wide temporal salinity fluctuations in the Chesapeake Bay, caution should be exercised in extrapolating per- formance of C. gigas at these sites over longer periods of time. In contrast, performance of C. gigas at high salinity sites in the At- lantic Coast of Virginia was clearly superior to that of C. virginica. The results of this study, however, are not sufficient to con- clude that C. gigas is or is not an appropriate species for intro- duction or use in these environments. Before reaching a decision concerning introduction of exotic species, ICES, as well as the European Inland Fisheries Advisory Commission (EIFAC) and the American Fisheries Society (AFS). have recommended that appro- priate authorities, including fishery managers, examine the candi- date species to: ( 1 ) assess the justification for the introduction: (2) assess its relationship with other members of the ecosystem and the possibility of introducing associated pathogens and parasites: and (3) examine the probable effects including a prediction of the range for the establishment of the species (Turner 1988). Use of reproductively capable diploid C. gigas would likely result in its introduction into some regions within the waters of Virginia and neighboring states. An important determinant of the extent to which this species might spread if introduced is the interactive effects of temperature and salinity on reproduction and larval de- velopment. Based on the review by Mann et. al (1991) and other reports indicating that optimal temperature and salinity ranges for C. gigas larvae are, respectively, 18-35 °C and 1 9— 35%c, Gottlieb and Schweighofer (1996) postulated that, if introduced. C. gigas would likely reproduce and establish resident populations in the lower portion of the Chesapeake Bay. Spreading would likely occur, via larval dispersal, into other areas of the Mid-Atlantic coast of North America. Interactions with other species — such as competitive interactions with C. virginica and predator-prey in- teractions may further influence the possible range extension. Ad- ditional investigations into environmental constraints on reproduc- tion, competitive interactions with native species and predator- prey dynamics would enhance our predictive capability to determine the potential range for establishment of C. gigas in habitats in the Mid-Atlantic region. ACKNOWLEDGMENTS We thank Rita Crockett. David Marshall, Jake Taylor, Shawn Stickler, Brian Trainum. Amanda Hayes, Tucker Terry. Francis O'Beirn. Tamara Hurlock. Caitlin Robertson. George Pongonis and the staff at VIMS vessels for assistance in the field. Juanita Walker and Rita Crockett conducted disease diagnoses. Stan Allen, Greg DeBrosse, and staff at Rutgers University produced the triploid oysters used in this study, and Valerie Harmon and the staff at the VIMS oyster hatchery produced the diploid oyster stocks. Ploidy analysis was conducted by Tom Gallivan and Aimee Howe under the direction of Stan Allen at both Rutgers University and VIMS. Wanda Cohen and staff at VIMS publications assisted with preparation of the report. The manuscript was improved with comments by Lisa Ragone-Calvo. VIMS contribution No. 2247. Field Study of C. Virginica and C. Gigas 473 Allen. S. K., Jr. & S. L. Downing. 1986. Performance of Iriploid Pacific oysters, Crassostrea gigas (Thunberg). I. survival, growth, glycogen content, and sexual maturation in yearlings. J. Experiment. Mar. Biol. Ecol. 102:197-208. Allen. S. K.. Jr.. S. L. Downing & K. K. Chew. 1989. Hatchery manual for producing triploid oysters. University of Washington Press. Seattle. 27 PP Andrews. J. D. 1980. A review of introductions of exotic oysters and biological planning for new importations. Mar. Fish. Re\: 42:1-1 1. Barber. B. J. & R. Mann. 1994. Growth and mortality of eastern oysters, Crassostrea virginica (Gmelin. 1791 ). and Pacific oysters. Crassostrea gigas (Thunherg. 1793) under challenge from the parasite, Perkinsus marinus. ./. Shellfish Res. 13:109-1 14. Blake, J. A. & J. W. Evans. 1972. Polydora and related genera as borers in mollusk shells and other calcareous substrates. Veliger 15:235-249. Burreson. E. M. & L. M. Ragone Calvo. 1996. Epizootiology of Perkinsus marinus disease of oysters in Chesapeake Bay. with emphasis on data since 1985. /. Shellfish Res. 15:17-34. Burreson. E. M., M. E. Robinson & A. Villalba. 1988. A comparison of paraffin histology and hemolymph analysis for the diagnosis of Hap- losporidium nelsoni (MSX) in Crassostrea virginica (Gmelin). J. Shell- fish Res. 7:19-23. Burreson, E. M, R. Mann & S. K. Allen. 1994. Field exposure of tnploid Crassostrea gigas to Haplosporidium nelsoni (MSX) and Perkinsus marinus (dermo) in the lower Chesapeake Bay. abstract. J. Shellfish Res. 13:293. Chew, K. K. 1990. Global bivalve shellfish introductions. World Aquacult. 21:9-22. Davis, J. P. 1989. Growth rate of sibling diploid and triploid oysters, Crassostrea gigas. J. Shellfish Res. 8:319. Downing, S. L. & S. K. Allen. 1987. Induced tnploidy in the Pacific oyster, Crassostrea gigas: optimal treatments with cytochalasin B depend on temperature. Aquaculture 61:1-15. Ford, S. E. & H. H. Haskin. 1987. Infection and mortality patterns in strains of oysters Crassostrea virginica selected for resistance to the parasite Haplosporidium nelsoni (MSX). J. Parasitol. 73:368-376. Gaffney, P. M. & D. Bushek. 1996. Genetic aspects of disease resistance in oysters. J. Shellfish Res. 15:135-140. Gottlieb. S. J. & M. E. Schweighofer. 1996. Oysters and the Chesapeake Bay ecosystem: a case for exotic species introduction to improve en- vironmental quality? Estuaries 19:639-650. Grizel, H. & M. Heral. 1991. Introduction into France of the Japanese oyster (Crassostrea gigas). J. Conseil Int. pour {'Exploration Mer 47: 399-403. Guo, X.. G. A. DeBrosse & S. K. Allen. 1996. All-triploid pacific oysters (Crassostrea gigas Thunberg) produced by mating tetraploids and dip- loids. Aquaculture 142:149-161. Handley, S. J. & P. R. Bergquist. 1997. Spionid polychaete infestations of intertidal Pacific oysters Crassostrea gigas (Thunberg). Mahurangi Harbour, northern New Zealand. Aquaculture 153:191-205. Kobayashi, M. K. E. E. Hoffmann. E. N. Powell. J. M. Klinck & K. Kusaka. 1997. A population dynamics model for the Japanese oyster, Crassostrea gigas. Aquaculture 149:285-321. Krebs, C. G. 1972. Ecology: The experimental analysis of distribution and abundance. Harper & Row. New York. 694 pp. LITERATURE CITED Langdon. C. J. & A. M. Robinson. 1991. Development of the commercial aquaculture of the Suminoe oyster (Crassostrea rivularis). abstract. J. Shellfish Res. 10:238. Lawrence. D. R. & G. I. Scott. 1982. The determination and use of con- dition index of oysters. Estuaries 5:23-27. Luckenbach. M. W. & J. Taylor. 1997. Oyster gardening in Virginia: an overview of techniques. Virginia Institute of Marine Science. School of Marine Science. College of William and Mary. Gloucester Point. Vir- ginia. 12 pp. Mackin. J.G. 1962. Oyster disease caused by Dermocystidium marinum and other microorganisms in Louisiana. Publication of the Institute of Marine Science. University of Texas. 7:132-229. Mann, R„ E. M. Burreson & P. K. Baker. 1991. The decline of the Virginia oyster fishery in Chesapeake Bay: considerations for introduction of a nonendemic species Crassostrea gigas (Thunberg. 1793). J. Shellfish Res. 10:379-388. Meyers, J. A.. E. M. Burreson, B. J. Barber & R. Mann. 1991. Suscepti- bility of diploid and Iriploid Pacfic oysters. Crassostrea gigas (Thun- berg, 1793) and eastern oysters. Crassostrea virginica (Gmelin. 1791 ). to Perkinsus marinus. J. Shellfish Res. 10:433—137. Newell, R. I. E. 1988. Ecological changes in Chesapeake Bay: are they the result of overharvesting the American oyster, Crassostrea virginica"!. pp. 536-546. In: M. P. Lynch and E. C. Chrome (eds.). Understanding the Estuary: Advances in Chesapeake Bay Research. Proceedings of a Conference. Chesapeake Bay Research Consortium Publication 129. Baltimore, Maryland. Ragone Calvo, L. M., V. Harmon & E. M. Burreson. 1997. Selection of oysters for resistance to two protozoan parasites, abstract. J. Shellfish Res. 16:327. 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. The Rice Institute Pamphlet. Special Issue. Houston. Ill pp. Shatkin. G.. S. E. Shumway & R. Hawes. 1997. Considerations regarding the possible introduction of the Pacific oyster (Crassostrea gigas) to the Gulf of Maine: a review of global experience. J. Shellfish Res. 16:463-478. Turner. G. E. 1988. (ed.). Codes of practice and manual of procedures for consideration of introductions and transfers of marine and freshwater organisms. International Council for the Exploration of the Sea. Coop- erative Research Rept. 159. Copenhagen. Denmark. 44 pp. Virginia Institute of Marine Science (VIMS). 1996. Strategic plan for molluscan shellfish research, including a rational plan for testing ap- plication of non-native oyster species. A report of the Virginia Institute of Marine Science to the Governor and the General Assembly of Vir- ginia. House Doc. 16. Richmond. 123 pp. Wargo. R. N. & S. E. Ford. 1993. The effect of shell infestation by Poly- dora sp. and infection by Haplosporidium nelsoni (MSX) on the tissue condition of oysters. Crassostrea virginica. Estuaries 16:229-234. Zar, J. H. 1974. Biostatistical analysis. Prentice-Hall, Upper Saddle River. New Jersey. 718 pp. Journal of Shellfish Research. Vol. IS. No. 2. 475-500, 1999. MODELING THE MSX PARASITE IN EASTERN OYSTER (CRASSOSTREA VIRGINICA) POPULATIONS. I. MODEL DEVELOPMENT, IMPLEMENTATION, AND VERIFICATION SUSAN FORD,1 ERIC POWELL,1 JOHN KLINCK,2 AND EILEEN HOFMANN2 [Haskin Shellfish Research Laboratory Rutgers University Port Norris, New Jersey 08349 2 Center for Coastal Physical Oceanography Old Dominion University Norfolk. Virginia 23529 ABSTRACT A mathematical model simulating the host-parasite-environmental interactions of eastern oysters (Crassostrea vir- ginica) and the pathogen, Haplosporidium nelsoni, which causes MSX disease, has been developed. The model has 2 components. One replicates the infection process within the oyster and the other simulates transmission. The infection-development component relies on basic physiological processes of both host and parasite, modified by the environment, to reproduce the observed annual prevalence cycle of H. nelsoni. Equations describing these rates were constructed using data from long-term field observations, and field and laboratory experiments. In the model, salinity and temperature have direct effects upon in vivo parasite survival and proliferation as well as on transmission rates. Cold winters depress transmission rates for 1 or 2 years after the event, even if temperatures return to normal. Warm winters have no effect on transmission in subsequent years. Hemocyte activity, parasite density, and the overall environmental quality provided to the parasite by the host also influence the modeled infection process. Hemocytes scavenge and eliminate parasites that die over the winter or that degenerate as a result of failed sporulation. Replication rates of H. nelsoni are slowed at high parasite densities. The environmental quality provided by the host, which is a function of oyster food availability and the oyster's potential growth efficiency, affects doubling times and also determines whether the parasite completes its life cycle by forming spores. Spore production is related to a threshold environmental quality, which occurs only in small oysters because of their high growth efficiency. Simulations that use environmental conditions characteristic of Delaware Bay reproduce the observed seasonal H. nelsoni cycle, consequent oyster mortality, and spore production in juvenile oysters. The oysier-H. nelsoni model provides a quanti- tative framework for guiding future laboratory and field studies as well as management efforts. KEY WORDS: Haplosporidium nelsoni. numerical modeling, MSX disease, marine pathogen, host-parasite environment INTRODUCTION Among the most important influences on population dynamics of eastern oysters. Crassostrea virginica, in the United States over the past half century has been disease. Two major diseases, both caused by water-borne protistan parasites, have severely dimin- ished the abundance of natural oyster populations, particularly in the middle Atlantic states (Ford and Tripp 1996). The first to be recognized was Dermo disease, caused by Perkinsus marinus. Al- though it was discovered in the late 1940s in the Gulf of Mexico, it had probably been present throughout the southeastern United States and Gulf of Mexico for many decades (Ray 1996). Between its discovery and 1990. Dermo disease was prevalent only in wa- ters south of Delaware Bay; since then however, epizootic out- breaks have been recorded as far north as Massachusetts (Ford 1996). The second, MSX disease, is caused by Haplosporidium nelsoni, a parasite believed to have been introduced to the east coast of the United States, where it began causing epizootic mor- talities in Delaware and Chesapeake Bays in the late 1950s and early 1960s. H. nelsoni is now present along the entire east coast, although its major impact has been from Virginia north to Maine (Ford and Tripp 1996. Barber et al. 1997). To synthesize available data and to investigate the factors influencing the interactions of host, parasite, and environment in Dermo disease, mathematical models for P. marinus and C. vir- ginica were developed (Hofmann et al. 1992, Hofmann et al. 1994, Hofmann et al. 1995). The individual models were then coupled to examine the effects of temperature, salinity, total seston, and food availability on the integrated host-parasite system (Powell et al. 1994. Powell et al. 1996). Simulations indicated that temperature controls on both host and parasite growth rates, and food avail- ability to the oyster, were the major elements influencing the in- teraction. High oyster reproduction and growth rates in southern latitudes allows populations to withstand Dermo disease pressure much better than in mid-latitudes, where both fecundity and growth rate are lower. Simulations also indicated that an important survival mechanism for the oyster is simply to increase body mass (i.e., growth) at a higher rate than the parasite can proliferate and thus to keep P. marinus densities from reaching lethal levels. In many of the locations where P. marinus is present, H. nel- soni is also. Particularly in the mid-Atlantic states and along the northeastern coast, both parasites cause major, recurring epizoot- ics. Therefore, to understand the effects of disease on oyster popu- lations in this region, it is necessary to consider the actions and interactions of both parasites on hosts at the individual and popu- lation level. Both parasites display distinct seasonal and inter- annual cycles of infection onset, development, and impact on the host. These cycles are largely a function of environmental factors, primarily temperature and salinity, to which both parasites and the oyster are sensitive. However, the ability to tolerate environmental extremes, or to profit from favorable conditions, is specific to each species. A numerical model offers an effective way to synthesize the many data available for the parasites and their host in a mecha- nism for understanding the complex interactions among these or- ganisms and their environment. The objective of this paper is to describe a model developed for Haplosporidium nelsoni in oysters. Like that for P. marinus. it is a physiological model structured around proliferation and death rates of H. nelsoni under different environmental conditions. Equa- tions describing these rates were constructed using data from long- term field observations, and field and laboratory experiments. The H. nelsoni model is described in the following section. The 475 476 Ford et al. MKTHODS AND MODEL DESCRIPTION Model Overview Conceptual Framework "I I I I I I I 1 I I r ASONDJFMAMJJ Figure 1. (A) Annual prevalence cycle for Haplosporidiitm nelsoni in- fections in eastern oysters in Delaware Bay, NJ, (first year of infection) showing relative contributions of epithelial (BFC = 1), subepithelial/ local (BFU = 2), and systemic (BFU = 3 & 4) infections to the overall prevalence (see text for definitions). The arrows and numbers indicate different phases of the infection cycle as described in the text. (B) Monthly and (C) cumulative nonpredation mortality for oysters un- dergoing first year exposure in the same location. Adapted from Ford and Haskin (1982). succeeding section presents a series of model outputs that illustrate its ability to simulate the seasonal cycle of H. nelsoni prevalence and intensity, and consequent oyster mortality, in a high-salinity enzootic area. The model described in the current paper is the basis for the studies presented in two subsequent papers: the effects of varying salinity on MSX disease development (Paraso et al. this volume) and a comparison of the disease in Delaware and Chesa- peake Bays with a discussion of the transmission issue (Powell et al. this volume). Haplosporidium nelsoni is classified in the phylum Haplospo- ridia (Corliss 1984. Perkins 1990). In the oyster, it occurs primarily as a multinucleated Plasmodium (Ford and Tripp 1996). A second life form, the spore, is found rarely in adult oysters, but can be common in juveniles (R. D. Barber et al. 1991, Burreson 1994). The method of transmission is unknown and may involve another host (Burreson 1988. Haskin and Andrews 1988). To replicate the oyster-W. nelsoni interactions, the mathematical model was di- vided into two principal components. One simulates the infection process within the oyster, including the formation of spores. The second simulates the transmission process, which occurs outside, and independent of. the oyster (Powell et al. this volume). Within the oyster, observed prevalence and intensity of H. nel- soni follows a defined seasonal pattern in all areas where it has been followed closely (Fig. la) (Andrews 1966. Farley 1975. Ford and Haskin 1982, Matthiessen et al. 1990). In the mid-Atlantic, infections are acquired from late May/early June through early fall. The earliest recognized stages are Plasmodia confined to the gill epithelium. Once established in the epithelium, parasites prolifer- ate, penetrate the basal lamina, and move into the circulatory sys- tem where they are carried to all tissues. Acquisition of new in- fections and in vivo parasite proliferation result in rising preva- lence and intensity levels throughout the summer and fall (Fig. 1A, point 1 ), and result in host mortality during late summer and fall. High infection prevalence and intensity occur in the autumn and into the winter, when low water temperatures have slowed the activity of both host and parasite (Fig. 1 A. point 2). In late winter and early spring, infection prevalence and intensity decrease, pre- sumably from the degeneration of H. nelsoni plasmodia as well as from the deaths of heavily infected oysters (Fig. 1A. point 3). In early spring, infection prevalence and intensity again increase, coincident with rising water temperature, reaching a peak in late May or early June (Fig. I A. point 4). This peak, which can be the most intense of the annual cycle, is often followed by a dramatic decrease in the number of infected oysters, again linked with the death of heavily infected oysters, but more so with the disappear- ance of parasites from live oysters (Fig. 1A. point 5). When sporu- lation occurs, it coincides with both the spring and the fall preva- lence/intensity peaks. The in vivo component of the H. nelsoni model is designed to replicate the above pattern. To simulate the in vivo relationship, the model relies on basic- physiological processes of host and parasite to reproduce the com- plex, bimodal, annual prevalence cycle observed in nature (An- drews 1966, Ford and Haskin 1982). Parasite proliferation, stage transition, and death rates, which are modified by environmental variables both external and internal to the host, form the basis of the model. Salinity and temperature have direct effects upon in vivo parasite survival and proliferation (Paraso et al. this volume). They also have both local and regional effects upon transmission. Hemocyte activity, parasite density, and the overall environmental quality provided by the host are additional factors that influence the parasite. The last affects not only parasite doubling times, but whether or not H. nelsoni completes its life cycle by producing spores. The environmental quality experienced by the parasite de- MSX Model Development and Verification 477 TABLE 1. Relationship of Haplosporidium nelsoni infection categories (see Ford and Haskin, 19821 to range and mean abundance of parasites, expressed in parasites (unit area)"1, in tissue selections of infected oysters. Counts were made from oysters with Little Ford Unit (LFll) ratings of I to 6 in either gill tissue, visceral mass tissue, or both. In each location, parasites were counted in a total area equalling 64,000 um\ The intensity ratings are indicated as Rare (R), Very Light (VL), Light (L), Moderate (M), and Heavy (H). The correspondence between LFUs and Big Ford Units (BFU) is also shown. INFECTION CATEGORIES GILL VISCERAL MASS Intensity BFU LFU Epithelial Subepithel ial LFU Epi thelial Suhepit Range helial Distribution Range Mian Range Mean Range Mean Mean Epithelial R. VL. L I 1 0.1-5.4 0.9 0 0 1 0 0 0 0 Subepith/Local VL. L II 2 0.5-3.9 1.4 0.1-1.1 0.5 2 II 0 O-fl.2 0.1 Systemic R. VL III 3 0-1.2 0.3 0.2-0.5 0.3 3 0 0 0.1-0.6 0.4 Systemic L III 4 0.2-0.4 0.2 0.2-2.1 LI 4 0-0.1 0.1 0.5-2.9 1.7 Systemic M IV 5 0.4-13.9 5.4 3.0-18.9 8.2 5 0-1.2 0.2 0.6-7.4 3.9 Systemic H IV 6 0.3-9.9 3.2 11.9-36.5 20.8 6 0-1.6 0.5 6.8-2S 5 17.5 pends, in turn, upon the quantity of food available to the oyster and its potential growth efficiency. Finally, the model simulates deaths of oysters as a consequence of parasitism. The transmission component of the model is discussed fully by Powell et al. (this volume). Unlike most disease transmission mod- els, including that for P. marinus, it does not rely on the density of nearby oysters as a measure of infective parasite concentration. In fact, there is no direct link between spore formation and transmis- sion in the model. Although spores are assumed to be an important element in the life cycle of H. nelsoni, it is not known if they are directly infective to other oysters. The infective stage is unknown, but histological observations of infected oysters suggest it is water- borne (Farley 1968. Ford and Haskin 1982). In the model, the relative abundance of these particles is influenced by salinity, on both local and estuary-wide scales, and long-term temperature fluctuations. The infection rate is a function of the abundance of infective particles and the filtration rate of oysters. This paper focuses on the in vivo model, which was constructed by applying rate functions developed from experimental and field data to an overall governing equation that controls the movement of oysters among infection classes according to the parasite load that they have at any time during a simulation. At each step in the construction of this model, output was compared with actual data and modifications implemented, if needed, to fit the model to field observations. To model the cycle in the absence of complete data on host-parasite interactions, and especially transmission, certain assumptions had to be made. The background and biological basis for these assumptions are stated briefly, along with the particular mathematical relationship, and are considered more fully in the Discussion. Model Units The first step in developing the H. nelsoni model was to define the units that provide the basic reference frame and that allow the model calculations and output to be consistent with measurements and to be compared with observations. The majority of the obser- vations on MSX disease are made by tissue-section histology and make use of scales that categorize H. nelsoni infection level ac- cording to parasite distribution (local or diffuse) and abundance in the oyster tissue. The scale reflects disease progression in the oyster as infections move from initial light lesions in the gill epi- thelium to heavy systemic (whole body) infections. The infection rating system used for the model is based on one developed for studies in Delaware Bay (Ford and Haskin 1982). This semi-quantitative scale involves 3 levels of distribution in the tissue (epithelial, subepithelial/local. and systemic) and 5 levels of abundance (see Ford and Haskin 1982 for details), resulting in a scale of 0 to 15 when the location and intensity for each oyster are multiplied. For reporting and statistical purposes, however, these 15 categories are reduced to 6 or 4, depending on need (Table 1 ). Similar ratings systems are used in Chesapeake Bay and elsewhere (Farley 1968, Y. Bobo, pers. coram., E. Burreson. pers. coram.. R. Smolowitz, pers. comm.l. In contrast to the rating systems in which most observations are reported, the oyster-W. nelsoni model is based on the number of parasites per oyster. It was therefore necessary to establish, at the outset, a relationship between parasites per oyster and the semi- quantitative scale. The 0-6 point scale (referred to as Little Ford Units [LFU]) was used as the basis for this relationship because it provided more precision than the 0—4 point scale. The 0—4 point scale was chosen as the final output from the model, however, because it is the simplest, because it can readily be compared with previous publications, and because it is most easily comparable to systems employed by other researchers. These units are referred to as Big Ford Units (BFU) (Table 1 ). Conversion between the scales simply involves combining the 4 highest LFUs into 2 BFUs for observational use (Table 1); however, the identical mathematical treatment results in a more complex conversion formula. A com- plete presentation of the conversion system is given below. Conversion of Infection Categories to Parasite Density The conversion of the LFU rating system into parasites per oyster was made by selecting archived slides with tissue sections in each of the categories (total n = 50. approximately equally distributed among the 6 categories of infected oyster). Each slide was then re-analyzed using a gridded ocular. All parasites were counted in 40 (40 u.m x 40 p.m) grids. 20 placed randomly over gill tissue and 20 over the remaining visceral mass. Resulting counts showed that the mean number of parasites per grid in each LFU category was similar for both the gill and the visceral mass (Table 1 ). The resulting empirical relationship between H. nelsoni 478 Ford et al. 2 w O 2 w O LFU LFU Figure 2. Haplosporidium nelsoni plasmodia per 64,000 unr tissue- section field versus infection intensity categories expressed as Little Ford Units (LFU) for a) epithelial and b) systemic tissues. infection category and the number of parasites in the oyster was exponential, with a rapid increase in numbers per grid as infections became systemic {LFU 4-6) (Fig. 2a, b). The location of the H. nelsoni cells in either epithelial or systemic tissue is included in the final relationship, which is based on a logarithmic scale of the form: temic is) tissue (the notation e/s will be used to denote constants that have different values for epithelial and systemic tissue), Ce/s is the number of H. nelsoni cells in the epithelial or systemic tissue, b is a scaling in grids per gram wet weight (gwwt-1) of oyster tissue, and ce/s is a constant. The coefficients represent the total dry weight in grams (gdwt-1) of the oyster tissue (W0) and the fraction of epithelial or systemic tissue ifrace/s) in the animal. The values of the coefficients in equation ( 1 ) are given in Table 2. The method described above for quantifying infection intensity introduced a bias when infections were confined to the epithelium, e.g., LFU 1 (Table 1). because gill epithelium comprised only about 20% of the tissue in each section. Thus, the values obtained in these instances were multiplied by a value of 5 so that the number of parasites per gram tissue was consistent with the values obtained for systemic infections. As a result, the constants a and c vary between the epithelial and systemic conversions and the num- ber of parasites per gram in the epithelial tissue is higher at an equivalent infection intensity than in the systemic tissue (Fig. 2). The constant /; in equation ( 1 ) is a conversion from the number of grids counted in a microscope field to the biomass of the tissue counted. In essence, this yields the weight of fixed tissue per grid. The conversion is based on the area of a grid (40 u.m x 40 p.m). the number of grids counted (40). and the thickness of a tissue cross-section (6 u.m). Included in the conversion is a factor of 0.5 to account for the expectation that, on average, an H. nelsoni Plas- modium would be observed in 2 consecutive cross-sections. The cal- culation of b also assumes a 10% shrinkage in tissue volume during fixation, thus correcting from fixed to wet tissue weight. The values for frace/s are obtained from weights of dissected oysters that show gill tissue to comprise about 20% of the total wet weight (Table 3). Half of this weight was estimated to be epithe- lium, based on point count stereology of tissue sections. The value for frace/s was therefore given a value of 0.1. Rare or very light epithelial infections (LFU = 1 ) may be iden- tified by as few as 1 or 2 parasites in the gill epithelium in a standard tissue cross-sectional analysis. With this method, however, it is likely that too few parasites are present in some oysters to be detected. Thus, some oysters diagnosed as having no infections (LFU = 0), are undoubtedly infected (Stokes et al. 1995). The model is constructed to reflect this circumstance. The distinction between an uninfected oys- ter (LFU = 0) and one in the very lightest infection category (LFU = 1 ) is based on a presumed detection limit and not on the absolute absence of infection. The detection limit, which differs for epithelial and systemic tissue, was obtained from the grid counts described above that were used to convert the infection scale to parasite densi- ties. The lowest level of detection for the conversion counts was 1 H. nelsoni cell per 20 grids, with an average value of 0.05 parasite per grid. However, tissue sections are routinely completely scanned for H. nelsoni to obtain observed prevalence. Twenty grids (64 x 103 ixirr) represented only an estimated 20% of the gill tissue and 10% of the visceral mass tissue present in a typical section. Therefore, the true detection limit of 1 parasite in either the gill or visceral mass after a complete search of the section would be 1 in 100 grids ( = 0.01 grid"' ) and 1 in 200 grids ( = 0.005 grid-1 ). respectively. This trans- lates into 1.3 x 104 and 6.5 x 103 parasites per gram wet weight for gill and visceral mass tissue, respectively. LFU = a., An C. (1) \.bce/sW0fract where «t.A is a constant that differs for the epithelial ie ) and sys- \fodel Equations The model is structured as a two-dimensional array (Figs. 3, 4) with 55 epithelial and 55 systemic infection categories. The infec- MSX Model Development and Verification 479 tion level in each category is defined by the average number of H. nelsoni in it. with the maximum difference between adjacent classes being 1 population doubling. The difference between in- fection classes at the higher parasite densities is less than 1 popu- lation doubling, because of the nonlinear distribution of LFUs with respect to parasite number (Fig. 2). The nonlinear arrangement was required to provide multiple infection classes within each LFU infection category and, consequently, necessitated scaling the transfers between infection categories by the ratio of the parasite cell number (C) between adjacent classes as: • for transfers up in epithelial tissue: CJ(Ce+1 - Ce) • for transfers down in epithelial tissue: CJ(Ce - Ce_x) • for transfers up in systemic tissue: CJ(Cs+l - Cs) • for transfers down in systemic tissue: CJ(CS - C,_,). For simplicity, these scalings are not explicitly stated in the equa- tions given below. In this array, only the [0.0] infection class is truly uninfected: however, a larger portion of the array contains infections not detectable by the tissue-section diagnostic method in which the model output is reported. To establish the boundaries of the patently uninfected class. LFU = 0. in the e x s array, the limits of detection described earlier were used to solve equation ( 1 ) and the array steps characterized by parasite densities below that value were defined as uninfected. For example, for a 1-g oyster, epithelial classes with LFUs < 0.8 and systemic classes with LFUs < -1.6 contained parasites below the detection limit. The lower LFU limit for systemic tissue originates from the much larger tissue cross-section area searched for the parasite, as discussed previously. The governing equation for determining the prevalence and intensity of//, nelsoni infections in the epithelial (e) and systemic (s) tissue of oysters (O) is given by: dO,. dt = ~ «,sO„ - B„0e.s + aes_xOe __, + P„_,,CUi > + Ote.s+lOe,s+l + KV+l.A+1.., - MesOes - 7e,i4 0,54 , N N + 8«i8S02^S'V"Oei (2) =54 6.5, without developing systemic infections, and this is modeled by an appropriately calibrated transfer function as dis- cussed later. Oysters in systemic classes >7.0 are automatically placed in the dead oyster category because parasite densities rep- resented by these classes are higher than those found in live oys- ters. Additional mortality processes will be discussed later. The diagonal line separating BFU 2 and BFU 3 (Fig. 4) is based on the observation that BFU category 2 is normally reached when advancing epithelial infections (BFU 1) give rise to local systemic infections (BFU 2). which then expand into BFU 3 and then BFU 4. BFU category 3 also includes infections decreasing in intensity. In the latter, parasite burdens diminish simultaneously in epithelial and systemic tissues, hence oysters move in a diagonal toward the undetectable infection category (BFU 0) rather than back through BFU 2 to BFU 1. Proliferation of H. nelsoni Transfers of oysters to different infection intensity classes are assumed to be due to proliferation and death of H. nelsoni cells, except in two cases. First, the acquisition of initial infections is determined by an external factor termed "transmission" (B00 in equation 3). Second, the development of an epithelial into a sys- temic infection is determined by an invasion rate that is not simply a function of cell division. These transfers are discussed in the following section and a schematic showing the many linkages in the oyster-//, nelsoni model is given in Fig. 5. To more easily describe the sequence of processes involved, the model is de- scribed as it simulates the yearly infection cycle (Fig. 1A) begin- ning with the onset of infection in June. Temperature-dependent proliferation of H. nelsoni. After the acquisition of//, nelsoni infections in early June (Fig. 1A). the proliferation of plasmodia in the epithelial and systemic tissue (ge/,(T)) is assumed to be exponential with a doubling rate that is modified by temperature as: ge/s(T) „MT-T„) (4) (3) where /;t) is the proliferation rate of the parasite in epithelial or systemic tissue based on a doubling time at a reference tempera- ture, T0. The reference temperature was taken as 15 °C instead of the standard 20 °C, because 20 °C did not produce adequate para- site division rates in the summer as compared to field observations. 480 Ford et al. TABLE 2. Definition, units, and values for the variables used in the oyster-//. nelsoni model equations. Delaware Bay and Chesapeake Bay are abbreviated DB and CB, respectively. TABLE 2. continued Variable a, C„ cs W0 t'rac,. frac, a P gJT) g,(T) crowd,,/s /Factor ccrowd,. ccrowd, cp do DD A° SM, K k, SDe l)l llllillllll Units Value constant none 1.244 constant none 0.919 epithelial or systemic number of calculated H. nelsoni cells cells scale factor grids (g wet wtr' 1.3 x 106 constant cells (grid)-' 0.135 constant cells(grid)-1 0.022 oyster dry weight g chosen fraction of W„ that is none 0.1 epithelial tissue fraction of Wn that is num systemic tissue growth rate d~' growth rate d_1 temp dependent d"' parasite growth rate in epithelium temp dependent d~' parasite growth rate in systemic tissue doubling time of d"1 parasite in epithelial tissue doubling time of d"' parasite in systemic tissue temperature effect on growth rate parasite growth rate reference temperature density-dependent control on growth oyster ingestion factor epithelial cell threshold for crowding systemic cell threshold for crowding rate of increase of crowding effect base cell diffusion rate degree days temperature differential maximum rate of cold susceptibility in epithelial tissue maximum rate of cold none susceptibility in systemic tissue DD value at which °C d reach one-half SM, DD value at which °C d reach one-half SMt. susceptibility decay (°C d)~ factor in epithelial tissue susceptibility decay (°C d)~ factor in systemic tissue 0.9 calculated calculated calculated calculated 0.23105 0.69315 oC-l 0.04 °c 15 none calculated none calculated number of cells 2.5 x 10" (g dry wtr1 number of cells 3.3 x 105 (g dry wt)"' none 1.5 d"1 0.138 °Cd calculated °C calculated none 2.0 8.0 20.0 10.0 0.2 0.1 Variable Definition Units Value HR(T) temperature dependent hemocyte rate a-1 calculated hr0 hemocyte activity base rate d-' 0.278 e hemocyte activity temperature rate (°C)-' 0.08 1 55 Th0 hemocyte activity base temperature °C 20.0 crowdw, epithelial cell threshold cells (g dry 3.0 x 106 for crowding for wtr1 hemocytes c/icrowd hemocyte crowding threshold in epithelial tissue none calculated NGER0 threshold value for modified net production (g dry wt)'1 0.25 NGERM minimum potential growth efficiency none 0.25 NGERqX minimum accumulated potential growth efficiency needed for sporulation (g dry wtr' 100 NGERql minimum accumulated potential growth efficiency needed for attempted sporulation (g dry wt)"' 10 SSR spore susceptibility decay rate d"' 0.1151 SporeS sporulation rate d-' calculated SporeSj sporulation rate none set at 1 .0 or modifier calculated SST0 spore temperature susceptibility °C 15 SSTsp susceptibility temperature switch °C 2.64 TempSS spore temperature susceptibility factor d none SpFrac fraction of H. nelsoni cells undergoing (oyster)"' 0.25 Total5 spores released cells calculated Spore/V number of spores number 25 formed per (cellr1 Plasmodium Sporek oyster death rate from sporulation d"' 0.1733 Smort salinity mortality factor none calculated SD, salinity mortality factor none 103.0 SD2 salinity mortality factor none 0.24065 5D, salinity mortality factor (pptr' 0.592456 SD4 salinity halving time d 4.0 Sdeath salinity mortality rate d"1 calculated Sfactor salinity effect of growth none calculated sg salinity effect on growth (pptr' 0.4605 So salinity growth effect reference salinity ppt 15.0 Sdeath,„,„ maximum salinity mortality rate d-1 0.01787 continued on next page MSX Model Development and Verification 4X1 TABLE 2. continued Variable Definition Units Value Sdiff salinity effect on diffusion rate none calculated SF\ salinity diffusion constant none 9.0 SF2 salinity diffusion constant none 2.65 SF3 salinity diffusion constant PPl 3.0 MortO mortality rate d"' calculated A/span mortality time span d 30 Ma mortality constant none 0.00747 Mb mortality constant (LFUr1 0.717 l\ infection constant none 0.0231 12 infection constant none i x ur4 B infection constant min particle -0.9 //'filter infective particles panicles calculated filtered min"' IPconc infective particle concentration particles 1"' calculated fill oyster filtration rate (Hofmann et al. 1992) 1 min"' calculated /Ptemp infection temperature effect none calculated /Psal infection salinity effect none calculated /Pseason infection seasonal effect none chosen SM, salinity mortality constant ppt 1.6 SM2 salinity mortality constant none 11.0 SMU salinity mortality reference salinity ppt 17.0 IP ' ' con Co base infective particle particles 1"' 450 iDB. 1960s) concentration 750 (DB. 1980s) 450 (CB) /Psalrate rate of change in spore concentration d"' calculated /Psalrateu change in spore concentration reference rate d"1 0.038376 IPsal0 change in spore concentration base salinity ppt 15.5 SIP salinity values from a specified time series for salinity oscillations ppt chosen IPsali change in spore concentration salinity constant ppt 5.0 /Pconcm (g) (g) (g) (%) Mantle 1.666 0.539 0.053 0.7 4.0 19 Digestive 1.715 0.55 0.054 0.7 4.0 19 Gland GUI 1.56 0.382 0.038 0.6 2.5 18 Adductor 1.677 0.549 0.054 0.6 3.2 19 Muscle Remainder 2.2S2 1.017 0.1 0.4 6.0 26 Diffusion between Epithelial and Systemic Tissue The crowding effect given by equation (5) modifies the tem- perature-dependent proliferation rates given by equation (4) to provide the final doubling times of H. nelsoni. These rates apply to H. nelsoni proliferation in all epithelial and systemic tissues, but not to the transfer of parasites from epithelial to systemic tissue. Although the mechanism by which H. nelsoni penetrates the basal lamina is not known, for purposes of the model, this transfer is assumed to be governed by a one-way diffusion process; i.e.. Plas- modia diffuse from the epithelial tissue to the systemic tissue. This process is described by an empirical equation of the form: diffusion = transfer by growth transfer by diffusion parasite death epigrow sysdie + * sysgrow parasite death ePidil -£> transfer by growth 0 Systemic infections (s) [55 classes] Figure 3. Conceptual model showing possible transfers of oysters through systemic (s) and epithelial ie) infection classes. The relative parasite density in each class is determined bv the proliferation (epi- grow and sysgrow I or death (epidie and sysdie I of Haplosporidium nelsoni. The [0,0] class represents uninfected oysters. Oysters cannot transfer from the uninfected class directly to a systemic infection class (open arrow with X), but must pass into an epithelial class first (i.e., initial infections are established in the epithelium), as indicated by the vertical arrow. Oysters may transfer to the uninfected class when the parasites die (horizontal and vertical arrows). Transfer of an oyster from an epithelial infection class into a systemic infection class is gov- erned by the diffusion (horizontal arrow), not proliferation (open ar- row with X), of parasites across the epithelial barrier. Oysters may transfer from systemic to epithelial infection levels by parasite death. in which the transfer rate depends on the density of parasites in both epithelial and systemic tissues. Equation (6) results in in- creased diffusion (or invasion) of H. nelsoni to systemic tissue as the number of parasites in the epithelial tissue increases (Fig. 8). Equation (6) applies to all transfer from epithelial to systemic tissue: however, there can be no transfer from the truly uninfected compartment (0,0 class) directly to the systemic compartment. Only an epithelial infection can give rise to a systemic infection (Fig. 3). From the above, the equations for proliferation of H. nelsoni in epithelial (Ge) and systemic (Gj tissue become: Ge = gj T) crowd,, G, = gs(T) crowds diffusion. (7) (8) Thus, the basic rate of H. nelsoni infection development includes only temperature- and density-dependent effects on doubling time, plus a diffusive contribution of parasites to the systemic tissue. H. nelsoni Mortality H. nelsoni prevalence and intensity decreases in early spring (Fig. 1A. point 3). Although many of the most heavily infected oysters die at this time, it is evident from histological observations that parasites are also in poor condition and probably dying (Ford and Haskin 1982). In the model, this loss of parasites cannot be accounted for by a simple reduction in H. nelsoni doubling rate at low winter temperatures or by a direct effect of cold temperatures LFU (systemic) 55 50 45 40 35 30 25 20 15 10 5 0 -5 0 2 4 6 777 1 1 1 1 ' 1 ' 1 / v/A /- -./ 6 / b v// 2 / Y// ** 4 / / : / Y / ; / 3 i i / A : + - i „Jc'" 7 - / *• " /- - < i <;■ 4 _ o 5 -5 -10 -15 10 15 20 25 30 35 40 45 50 55 Systemic Infection Class Figure 4. Schematic showing the relationship between the systemic and epithelial infection classes used in the oyster-//, nelsoni model and Eittle Ford I'nits (LFU). Model infection classes (0-55) for epithelial and systemic infections are shown on the left and bottom sides of the figure, respectively. Their respective LFU categories (0-6) are shown on the right and top sides. The model output is in LFUs, but these are converted to Big Ford Units (BFU) for plotting. BFUs are shown as areas within the plot, with numbers from 0-6. The cross-hatched re- gions indicate infection levels where the oysters are dead (BFU = 5) or epithelial infection levels that are not normally achieved in nature (BFU = 6). BFU category 0 represents infections that are below the level of detection. The dashed arrows indicate the H. nelsoni infection trajectory normally observed in oysters during infection proliferation and remission. MSX Model Development and Verification 483 on H. nelsoni survival. Simulations of both conditions failed to agree with field observations. In fact, the decrease occurs at a time when temperatures are rising. Obviously, an additional factor is operating in nature to cause the observed decrease in parasite density. Simulations agreed with observations when it was assumed that the spring decrease in parasite density is due to a combination of 2 factors: 1 ) in vivo conditions coinciding with low winter temperatures that debilitate H. nelsoni Plasmodia and increase their susceptibility to hemocyte attack; and 2) increasing activity of oyster hemocytes in the spring that removes the damaged parasites. In addition to the apparent degeneration of parasites in late winter, the biological rationale includes experimental data showing that oyster hemocytes do not attack and phagocytose live H. nelsoni, but readily ingest killed parasites or those with arrested metabolism (Ford et al. 1993. Ford and Ashton-Alcox 1998). The exact mechanism that results in reduced viability of H. nelsoni is unknown, but for convenience, the term "cold*' susceptibility is used for this factor. Cold susceptibility. The susceptibility of H. nelsoni to low temperature was assumed to depend on the number of winter days during which the parasites are exposed to temperatures below a threshold (i.e., degree days). The number of degree days (DD) was determined by summing the difference between the threshold tem- perature and the ambient temperature (A°) over time as: DD= ^ A°dt (9) where DD ranges between 0 and 200. The threshold temperature was taken to be 5 °C, based on simulations covering a range of temperatures between 0 °C and 10 °C. As long as A° was positive (ambient temperature declining), the number of degree days was related to H. nelsoni susceptibility to oyster-hemocyte attack in epithelial and systemic tissue (HSits^J as: HSus,,. = SMe/s DD L +DD (10) This hyperbolic saturation relationship results in increasing susceptibility of H. nelsoni to oyster hemocytes as cold exposure is prolonged. However, above a certain level of cold exposure, susceptibility no longer increases (Fig. 9). In the model, parasite burdens are made to diminish more rapidly in the systemic tissue than in the epithelium to match histological observations that, as infection intensity diminishes, the last parasites seen are in the epithelium (Ford and Haskin 1982, Ford 1985a). The data imply that parasites are eliminated faster from systemic locations and the biological rationale assumes that the number of hemocytes per parasite is higher in the circulation (systemic tissue) than in the epithelium and therefore the rate at which moribund parasites can be scavenged by the hemocytes is greater. As temperatures increase in the spring, the degree-day value decreases and eventually becomes negative. During this time, the assumption is that susceptibility of H. nelsoni plasmodia to the oyster hemocytes decreases as parasites recover from cold expo- sure, or because only undamaged parasites remain. This effect is incorporated into equation ( 10) through the addition of a term that attenuates H. nelsoni cold susceptibility as temperature increases: H Decav = - 1 + A° SD„ \° SD. Ill) Seasonal Salinity Cycle Particular Load Figure 5. Schematic of the linkages and processes included in the oyster-//, nelsoni model. 484 Ford et al. 2.0 1.5 1.0 0.0 - 1 — i — i- - 1 1 1 1 r- —I 1 1 1 1 1 1 1 1- Systemic Tissue Epithelial Tissue 0.15 10 20 Temperature ( C) 30 40 Figure 6. Haplosporidium nelsoni proliferation rates in epithelial and systemic tissues of oysters as a function of temperature at salinities > 15 ppt. where the values used to determine the rate of decay in suscepti- bility of H. nelsoni cells to hemocytes (S£>,.A) differ for the epi- thelium and systemic tissue. Parasites in the epithelium are as- sumed to recover more rapidly for the same reason that they are less susceptible to the degree-day factor. As long as A0 is negative (ambient temperature increasing) the parameter HSuseA is de- creased each time step by H Decay amount Hemocyte removal of damaged H. nelsoni Plasmodia. Once susceptible because of cold-associated damage. H. nelsoni Plas- modia can be removed by oyster hemocytes at a rate that is de- pendent on temperature. The hemocytes are assumed to become maximally active at 10 °C and their activity to decrease above 10 °C (Fisher and Tamplin 1988). This is given by an equation of the form: 1.00 0.80 H- 0.60 ■9 0.40 0.20 0.00 I I 1 I I 1 I 1 1 I 1 1 1 1 I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 - Systemic . - - 1 1 1 1 I 1 1 1 1 1 1 1 1 1 I - ■ ^J >r-^, Logio (H. nelsoni cells/gDW) Figure 7. The relationship between the reduction in Haplosporidium nelsoni proliferation rates in oyster epithelial and systemic tissue as a function of increasing parasite density (self crowding) at salinities > 15 ppt and a temperature of 20 °C. — 0.10 i 0.05 0.00 L°910 epithelial parasites oyster Figure 8. Relationship of the invasion (diffusion) rate of Haplospo- ridium nelsoni from oyster epithelial to systemic tissue and the number of parasites in the epithelial tissue. The rate shown is the one for the initial invasion of parasites into the systemic tissues. HR(T) = hr0e 0(max(T. 10)-77i(1) (12) where the base hemocyte activity rate. hr0, is related to the rate at 20 °C (Th0). The observation that the rate at which oyster hemocytes phagocytose foreign particles is reduced below 10 °C (Feng and Feng 1974. Alvarez et al. 1989) is incorporated into the model as a linear decline to zero in hemocyte activity from 10 °C down to 0 °C. Net Proliferation of H. nelsoni From the above relationships, the net doubling time of H. nelsoni in the epithelial (/VGt,) and systemic (NGj oyster tissue is given by: NG, = G„ - HR( T) HSiis,, ehcrowd NG„ = G, - HR{ T) HSus, crowd, (13) (14) where the final terms represent parasite density effects on overall hemocyte effectiveness. For systemic infections, this term is the same as that used for parasite crowding effects in equation (5) and accounts for the fact that the increase in circulating hemocyte concentrations stimulated by H. nelsoni infection is relatively less than the increase in parasite density (Ford and Kanaley 1988. Ford et al. 1993). Thus, the fraction of the H. nelsoni population re- moved by hemocytes becomes progressively lower as the number of parasites increases. The value for parasite concentration in the epithelium at which crowding occurs, crowdeh, as applied to hemocyte activity, is cal- culated from equation (5): however, the parasite concentration at which crowding occurs is 17% larger (Table 2) than the constant for epithelial tissue used in equation (5). Once again, this value was established through comparison of simulation results and field observations, because there are no direct observations of this ef- fect. The higher value for the coefficient. crowdchl indicates that hemocytes in the epithelial tissue remain active at proportionally higher H. nelsoni numbers than in the systemic tissue. In the ab- MSX Model Development and Verification 485 12.0 9.0 3.0 t — i — i — i — < — i — ' — r Systemic Tissue Epithelial Tissue _l_ to H. nelsoni after the oyster's respiratory and reproductive de- mands have been met. i.e., it is the oyster's potential growth effi- ciency. The term "potential," rather than "net." growth efficiency is used because some fraction of assimilated energy is utilized by H. nelsoni. rather than by the oyster, and this fraction should thus be subtracted from assimilated energy in the calculation of net growth efficiency (e.g.. Hofmann et al. 1995, Eq. 1). Potential growth efficiency would be energy available for oyster growth if H. nelsoni were not present. The value of NGER from Equation (15) is used to calculate / factor in the relationship that determines density-dependent crowding (equation 5) as: 80 100 120 140 Degree Days ( Cd) 160 180 200 Figure 9. Relationship between the number of degree days below 5 °C and Haplosporidium nelsoni susceptibility to destruction by oyster hemocytes in epithelial and systemic tissues. sence of processes discussed in subsequent sections, equations (13) and ( 14) determine the values of a and P in equation (2). H. nelsoni Proliferation in Spring The increase in H. nelsoni infection prevalence and intensity in early spring shortly after the late winter die off (Fig. 1A, point 4). which coincides with rising water temperature, cannot be repro- duced in the model through a simple temperature effect on dou- bling rate. The speed of the increase suggests that density- dependent control on parasite proliferation in the oyster has been released. The biological rationale for this argument is based on observations that, in spring, a rapid increase in oyster growth rate occurs associated with the spring bloom and rising water tempera- ture. For purposes of the model, the environment experienced by H. nelsoni inside its host is assumed to improve concurrently as a consequence of an inflow of nutrients, favoring rapid parasite pro- liferation. It is this fastidious dependency of H. nelsoni on nutri- ents supplied to its host that will dominate the remainder of the post-infection oyster-H. nelsoni model. The effect of changing nutrient supply in the spring was in- cluded in the model by relating the density-dependent control on H. nelsoni proliferation to food intake by the oyster through fil- tration and ingestion. In spring, when algal supply and oyster filtration rate are high, the density-dependent control on H. nelsoni proliferation is reduced, allowing the parasite to remain in the exponential phase of its growth with maximum cell division rates for a relatively long period. This effect is included in the model through a potential growth efficiency ratio (NGER) that is calcu- lated as: NGER- assimilation - respiration - reproduction assimilation (15) where oyster assimilation, respiration, and reproduction are cal- culated using the relationships given in Hofmann et al. (1992. 1994). Equation (15) gives the fraction of net production available I factor = max I NGER ' \NGER,, VV„ (16) where NGER0 is the threshold value above which the modified net production {NGER) is available to H. nelsoni. The threshold value was determined empirically through a series of simulations de- signed to reproduce the annual cycle of H. nelsoni infection and intensity observed in Delaware Bay (Ford and Haskin 1982). The release of the crowding effect occurs only when NGER 2 0. The effect of I factor is to increase the number of H. nelsoni parasites that must be present before density-dependent controls on parasite proliferation become a regulating factor. Sporulation of H. nelsoni The factors governing spore production in H. nelsoni -infected oysters and the role of spores in its life cycle are among the least understood aspects of this parasite (Haskin and Andrews 1988). The parasites rarely form spores in adult oysters, but may do so regularly in juveniles in both spring and autumn (R. D. Barber et al. 1991. Burreson 1994). Spores can be shed from live oysters, but it is likely that most oysters die during or after sporulation because their infections are so heavy (R. D. Barber et al. 1991). In histo- logical sections of adult oysters with advanced infections at the spring peak, parasites often appear degenerate, with large anoma- lous nuclei. These abnormal plasmodia may be evidence of a failed attempt at sporulation, after which the parasite dies without com- pleting its life cycle. For purposes of the model, sporulation or abortive sporulation is hypothesized to be responsible for the rapid disappearance ot H. nelsoni from oysters in late spring to early summer (Fig. 1A, point 5). In the model, parasites in heavily infected oysters, LFU 2 4, can attempt to sporulate, with two possible results. The first is that sporulation is successful, in which case spores are formed and released into the environment. Some oyster mortality is associated with this process. The second possibility is that sporulation is attempted, but is unsuccessful. Failed sporulation makes H. nelsoni more susceptible to oyster hemocytes. which remove the parasites and produce oysters with lighter infections. It may also happen that parasites in the heavily infected oysters do not attempt sporulation. The first part of modeling sporulation required determining whether or not H. nelsoni should attempt sporulation; that is, to model conditions within the oyster that would, or would not, favor spore development. The reason or reasons that small oysters sup- port sporulation whereas large oysters typically do not is unknown. The model, however, assumes that it is related to the higher growth efficiency of young oysters, which is reflected in higher NGER 486 Ford et al. values. The approach taken is based on the assumption that spol- iation requires a period of good environmental conditions, char- acterized by high oyster potential growth efficiency, which pro- vides a surplus of required nutrients or other factors to H. nelsoni and consequently permits sporulation. Thus, the model accumu- lates the value of NGER from equation 15 {NGERj over time to obtain a measure of the "internal environmental quality" of the oyster in terms of its ability to support H. nelsoni development (Fig. 10, step 1 ). This was done at each time step such that: NGERT = NGER? + max (NGER - NGERM. 0) Ar (17) where It is the time step of the model. As NGER exceeds the value of NGERd0. the quality of the parasite's environment improves and the parasite benefits from the improved conditions, e.g., NGERd is positive. Equation (17) provides the basis for the remainder of the approach used to simulate sporulation (Fig. 10). Thus, the equa- tions that control sporulation are structured around the seasonal cycle of oyster food availability (Fig. 9). When NGER - NGERM is negative, as during periods of low food, NGERd does not accumulate and sporulation cannot occur (Fig. 10, step 2). However, the time span of high nutrient avail- ability required for sporulation need not be continuous so NGERd does not decline during periods when nutrient availability is low. Times when NGERd is above zero have 4 possible outcomes. The first occurs if a positive NGERd occurs during times when Plasmodia are susceptible to cold (Fig. 10, step 3). It is assumed that cold-damaged plasmodia cannot take advantage of the im- proving quality of the internal host environment. When the sum of the cold-exposure death rates of H. nelsoni in the epithelial and systemic tissue (equations 10 and 12) exceeds 0.1 d"\ NGERd is not accumulated. The second and third possible outcomes occur when the H. nelsoni plasmodia are healthy and the internal quality of the host is improving (e.g.. NGERd is positive). At these times, sporulation HR(T) (Hsuse ♦ Hsusa) > 0.1 Figure 10. Schematic showing the potential pathways that may arise during sporulation and attempted sporulation by H. nelsoni. The num- bers correspond to specific sporulation processes that are described in the text. Figure 11. Relationship between NGERd, the accumulated potential growth efficiency ratio, and the sporulation triggers for oysters of 3 becomes a possibility. It is assumed that, as NGERd is accumulat- ing and the oyster quality is becoming more favorable, parasites are cued to begin the sporulation process. In the second possible outcome, sporulation is successful. For successful completion of this process, a certain level of internal host quality must be attained (Fig. 10. step 4). The quality trigger (NGERql) for sporulation was set at 100 gdwt"1. a value determined empirically through the comparison of a series of simulations and field observations. As noted above, sporulation success is related to the size of the oyster host, with successful sporulation predominating in small oysters. Thus, the quality trigger is scaled by the size of the host and, when NGERd exceeds the quality threshold (NGERd > NGER,/t W0), sporulation is triggered and NGERd is reset to zero (Fig. 1 1 ). The value of 100 gdwr1 permits sporulation in small (up to about 2 cm in length) oysters because of their higher potential growth effi- ciency, but does not permit sporulation in larger oysters. In the third possible outcome, sporulation is unsuccessful. In larger oysters, quality also improves as NGERd accumulates, but because of lower potential growth efficiency and, consequently, fewer resources available to the parasite, the sporulation trigger is rarely reached. In these oysters, the parasites prepare for sporula- tion. but the spring bloom ceases and nutrient levels decline before enough nutrients are obtained to sustain sporulation. When nutrient levels decline enough that NGER - NGERd(t becomes negative, abortive sporulation occurs in animals that have accumulated NGERd above a second weight-scaled quality trigger (NGER^ = 10 gdwt1): NGERd > NGERq2 W0 (Fig. 10. step 5). When this happens. NGERd is reset to zero. It is also possible that the accumulated value of NGERd will not exceed either quality trigger {NGERql, NGERi/2). In this fourth possible outcome, sporulation is not attempted and infection in- tensity continues to increase as determined by the parasite dou- bling time (Fig. 10, step 6). Sporulation and attempted sporulation do not occur instanta- neously in all oysters meeting the nutritive requirements for the process. The rate of sporulation or attempted sporulation (SporeS) is high immediately after the conditions of the nutritive triggers are met and decays over time. The base rate, SporeS, is defined as 0.1 LFU. which produces the desired result that sporulation and at- tempted sporulation events occur more frequently at higher infec- tion intensities. This rate decreases linearly over time by first setting SporeSd = 1. and then establishing a rate of decay. SporeST = SporeS? Q - ArS5/?) (18) MSX Model Development and Verification 487 where SSR sets the decay rate such that speculation or attempted sporulation ceases 60 days following the initial trigger. Sixty days provides simulations that best fit field observations of H. nelsoni infection intensity during the summer sporulation event. In any given time step. then, the number of oysters undergoing sporula- tion or attempted sporulation is: 0P3 = SporeS SporeS ,, Oeii (19) where O^. , are those oysters undergoing sporulation or attempted sporulation. Spores are formed at times of rapid parasite proliferation, in the spring and late summer/early fall (R. D. Barber et al. 1991, Bur- reson 1994), but the marked decline in prevalence and intensity that is hypothesized to occur, at least partly as a result of failed sporulation (i.e.. incomplete life cycle), occurs only in the spring as water temperatures exceed about 20 °C (Andrews 1966, Ford and Haskin 1982). This observation suggests an influence of tem- perature on sporulation and attempted sporulation such that neither process occurs at temperatures where H. nelsoni is cold susceptible and the process occurs at fastest rates above 20 °C despite ad- equate nutritive values (NGERd). Therefore, a temperature- dependent "spore susceptibility"' factor (TempSS) was used to modify equation (19). The temperature factor was defined as: 1 + tank TempSS = Ar - T-SSTu SST„ (20) which allows sporulation to be set in motion at about 9 to 10 °C and reach a maximum rate at 21 °C (Fig. 12). The coefficients. SST0 and SST determine the temperature at which TempSS is one-half its maximum rate and the temperature range over which the spore susceptibility switches from little to maximum effect (Fig. 12). The temperature effect modifies equation (19) as: 0%s = TempSS SporeS Spore d Oes (21) Failed sporulation results in death of H. nelsoni, their removal by hemocytes. and a lower intensity infection in the oyster. One-half 16 24 Temperature ( C) 32 40 Figure 12. Relationship between the rate at which sporulation can be attempted and temperature. of the oysters assumed to lose all parasites due to failed sporulation are placed in the uninfected oyster class ([0.0], equation 3). The remaining one-half are placed into the lowest epithelial, no- systemic infection ([1,0], equation 2) class. Successful sporulation occurs during periods when the quality of the host's internal environment increases to the point that the weight-scaled sporulation trigger (NCERc/1) is exceeded (Fig. 1 1 ). The factors that determine the number of oysters in which suc- cessful sporulation occurs are similar to those that affect the num- ber of oysters undergoing failed sporulation. with the exception that some oyster mortality also occurs in the process. Therefore, equation (21) is used to calculate the number of oysters surviving sporulation. Successful sporulation results in the death of some fraction of the affected oysters. The number of oysters with infec- tions in systemic LFU category 4 and all epithelial categories that die from the sporulation event is calculated as: 0;'s = Sporek SporeS Oes (22) where the initial rate at which oysters die as a result of sporulation {Sporek) is assumed to be equivalent to a four-day halving time. The dead oysters are removed from subsequent calculations. Successful sporulation releases H. nelsoni spores into the en- vironment. The total number of spores released {TotalS) can be calculated as: TotalS = Oes SpFrac SporeN (frace cellse + fracs cells J (23) where SpFrac is the fraction of the parasites that undergo success- ful sporulation and SporeN is the number of spores formed by each H. nelsoni Plasmodium (Table 2). Salinity Effects on H. nelsoni Laboratory (Sprague et al. 1969. Ford and Haskin 1988) and field observations (Farley 1975, Haskin and Ford 1982, Andrews 1983. Ford 1985b) have shown salinity to be a critical environ- mental factor regulating the spatial and temporal distribution of H. nelsoni in oyster populations. The following paper (Paraso et al. this volume) describes many of these interactions and provides detailed descriptions of how the coupled model simulates salinity- H. nelsoni interactions. However, a brief accounting of these pa- rameterizations is given here for completeness in the model de- scription. In the model, salinity affects H. nelsoni-oystcr interac- tions by controlling parasite proliferation rate, mortality rate, transfer rate from epithelial into systemic tissues, and infection rate. The basis for the effect of salinity on H. nelsoni proliferation in vivo is a relationship derived from measurements of acute in vitro salinity tolerance of the plasmodial stage of H. nelsoni (Ford and Haskin. 1988). This relationship shows that, at a salinity of less than 5 ppt. H. nelsoni survival is zero. Between 5 and 15 ppt, the parasites show an exponential increase in survival, and above 15 ppt little mortality occurs (see Paraso et al. this volume. Fig. 3). The salinity-caused mortality (Smart) was modeled as: / 0.01 SD, \ Smart = min / 1 . — 77: \ (24) 1 +■ SO, -50, SD, where 5 is the ambient salinity in ppt and SDJ, SD~. and SDJ are constants. The actual salinity-induced parasite death rate is calcu- lated as: Sdeath = ■ -Ini Smart) SDA (25) 488 Ford et al. where the death rate calculated from in vitro data is assumed to occur over four days (SDA) to account for the buffering effect of an in vivo situation (Galtsoff 1964, Shumway 1996). The salinity- caused mortality modifies the net proliferation rates in the epithe- lial and systemic tissues that are given by equations (7) and (8). Unlike other sources of mortality, salinity-caused mortality is as- sumed to be able to completely eliminate infections. This occurs at mortality rates above 0.01787 d"1. It was assumed that salinity effects on parasite proliferation rates would occur over the same salinity range as that producing parasite mortality; hence, the effect of salinities between 5 and 15 ppt on parasite doubling time was included through an exponential relationship: Sfactor sg(S-S0) (26) that varied between zero (S < 5 ppt) and 1 (S > S(1), where sg determines the rate of decrease of parasite proliferation rate with increasing salinity and S„ is 15 ppt. the salinity threshold above which no reduction in parasite proliferation rate occurs. Equation (26) modifies the temperature-dependent growth rate given in equation (4). In the initial simulations, the frequency of systemic infection decreased with decreasing salinity. Long-term observations in Delaware Bay. however, show that, after an initial decrease from the high salinity (20-23 ppt) planting grounds to the lower-most seed beds (18 ppt), the frequency of systemic infections remains unchanged along the remainder of the salinity gradient to the up- per-most seed bed (9 ppt). To simulate the observed pattern, the model increases the rate of parasite diffusion between epithelial and systemic tissue with decreasing salinity by including an addi- tional term of the form: Sdiff= \+SF\ lanh SF2 s-sf SFi (27) to equation (6). This relationship allows the rate of diffusion be- tween epithelial and systemic tissues to be maximum for salinities of 12 ppt and less, and to decrease to the base rate given by equation (6) between 12 and 18 ppt. It is presently unclear whether the biological basis of the field observations is actually tied to more rapid transfer of parasites, or whether some other mechanism is responsible. Thus, equation (14) can now be updated to its final form: NGS = GS- HR( T) HSus, crowd, - Sdeath + diffusion Sdiff. ( 28 ) Oyster Mortality The ultimate result of most H. nelsoni infections is the death of the oyster host. To model this effect, historical data on the intensity of infection (LFUs) in live and dead oysters was assembled. The percent of live and dead oysters in each infection category was calculated as a function of the total number of live or dead oysters, respectively, in the set of samples examined. The ratio of percent dead to percent live in each category was then computed. This ratio was considered a relative measure of the likelihood that an oyster will die with a given category of infection. Results showed that oysters in LFU categories 1 and 2 are no more likely to die than those in category 0: in categories 3-5. the likelihood rises to be- tween two and three; and oysters with category 6 infections are six times more likely to die than those without detectable infections. This relationship (Fig. 13) is of the form: MortO ■ -ln\ 1 - Ma eMhLFU) (29) Abundant field observations show that infected oysters can survive better at low temperatures than at high (Andrews 1968. Ford and Haskin 1982). For instance, as temperatures approach 7 °C in late November in Delaware Bay. the mortality rate drops to nearly zero. It is assumed that this happens because both host and parasite are quiescent at low temperature: the parasite no longer actively damaging the host and the host no longer actively feeling the effects of parasitism. It is a system "on hold" over the winter. Thus, a temperature effect was applied to the death rate given by equation (27) such that oyster mortality is reduced in a linear manner from the rate at 7 °C to zero at 0 °C. The total number of dead oysters in any infection class is then calculated as: O, (30) Spore k SporeS MortO Ot. s which is a modification of equation (22). In addition, any oyster in which the infection intensity exceeds that found in live oysters automatically is placed in the dead oyster category (Fig. 4). The dead oysters are removed from subsequent calculations of infec- tion dynamics, but they are accumulated over time to provide an estimate of mortality. H. nelsoni Transmission Transmission is dealt with fully in the third paper in this series (Powell et al. this volume). A condensed accounting of the param- eterizations used for this process is given here for completeness in the model description. The processes by which H. nelsoni is transmitted to uninfected oysters, and the form of the infective particle, are not known. However, observations that the earliest infections are in the nill \i.£. 1 ' ■ | i i i r T t --T- i ^ J Is ~ --- 5 C -. _ "O " - 0 to en > 1 0.1 i - c . 0 . 5 , - i 0) (0 >. o /J ^yy n n r" i ■ ' r " , ... 1 3.0 4.0 6.0 7.0 5.0 Systemic LFU Figure 13. Oyster mortality rate as a function of systemic LFU at 5 C and 25 C which span the range of temperature that is normally encountered in Delaware Bay. MSX Model Development and Verification 489 epithelium indicate that infective particles are acquired through filtration (Farley 1968. Ford and Haskin 1982). In addition, early studies with timed imports of oysters into enzootic regions of Delaware and Chesapeake Bays clearly showed that oysters be- came infected only during a period from late May through early October (Andrews 1968. Ford and Haskin 1982). suggesting that there is a seasonal dependence in the ambient concentration of H. nelsoni. The abundance of infective particles in the water is a critical element in modeling transmission, but no measurements are available to parameterize this process. Recently, however. Bar- ber and Ford (1992) reported finding haplosporidian spores, mor- phologically similar to those of H. nelsoni, in the digestive tract lumina of oysters in Delaware Bay and other regions enzootic for H. nelsoni. The spores, obviously ingested while feeding, predomi- nated from May through October, the known infective period for H. nelsoni. These may not be H. nelsoni spores, and if they are, they may not be the stage that infects oysters. Nevertheless, these data are the only ones available on which to base a rough estimate of likely seasonal fluctuations in ambient concentrations of H. nelsoni infective particles. Further, both simulations and observa- tions suggested that salinity and temperature, in addition to time of year, affect the abundance of infective particles (see below). The actual rate at which new H. nelsoni infections occur in uninfected oysters (Om) is dependent upon the number of infective particles filtered out of the water. This rate (B00) is given by: 3„o = - (31) where IP filter is the number of infective particles filtered by the oyster. The relationship assumes a threshold dose of 8,700 infec- tive particles filtered d~' needed to generate a new infection. The rationale for using this value is given in Powell et al. (this volume). The remainder of the transmission submodel is designed to esti- mate IP filter. The number of infective particles filtered by the oyster was modeled as: IP filter = IPConc filt(size) IPseason IPsal IPtemp (32) where IPconc is the ambient infective particle concentration in the water column, filt{size) is oyster filtration rate, IPtemp and IPsal are the temperature and salinity effects on infective particle abun- dance, respectively, and IPseason is the seasonal variation in in- fective particle availability. Oyster filtration rate is calculated us- ing the relationships given in Hofmann et al. (1992. 1994). The relationships used to specify the seasonal, salinity and temperature dependencies of the infective particles are described below. Seasonal effects. The base concentration of infective par- ticles, lPcono was chosen by comparing results of simulations using a range of values to field observations of prevalence (dis- cussed in Powell et al. this volume). The base concentration was then modified seasonally based on observations of ingested hap- losporidian spores, which revealed that spores were present pri- marily during the May-October period (Barber and Ford 1992). This time series (Fig. 14) was taken to reflect the relative abun- dance of infective particles and was included in equation (32) as IPseason. Local salinity effects. Initial simulations of H. nelsoni prevalence in low-salinity oysters showed that prevalences were I 1 X X n x r ^ _Q JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Figure 14. Time series of putative Haplosporodium nelsoni spores ob- served in sectioned Crassostrea virginica gut lumena as described in Barber and Ford (1992). higher than those observed and suggested that the rate of infection, as well as the rate of proliferation within oysters, decreases with decreasing salinity (Paraso et al. this volume). A function that decreased the concentration of infective particles in low salinity water resulted in simulated prevalence levels and patterns that better match those recorded on the low-salinity Delaware Bay seed beds (Paraso et al. this volume). The function was obtained by using the model to simulate infections over a broad range of sa- linities in Delaware Bay and comparing these to long-term time series (Haskin and Ford 1982, Fegley et al. 1994). Based on these comparisons, the effect of local salinity on transmission rate was modeled as: 1 + tank SM IPsal -- (5-5M„) SM, (33) The relationship makes biological sense because the salinity range affecting transmission is similar to the range affecting parasite mortality in the host and the somewhat wider range is anticipated for a potentially free-living infective particle. Whether the model simulates decreased survival of infective particles, their decreased ability to infect, or simply a dilution factor, is unknown. Bay-wide oscillations. Simulations with long-term time se- ries that were designed to test the adequacy of the transmission submodel, using the basic process of oyster filtration, infective dose, the seasonal cycle of infective particle availability, and a local effect of salinity on infectivity, showed adequate simulations for oyster populations over a wide salinity range in a specific bay, such as Delaware Bay (e.g., Paraso et al. this volume), during most years. However, the same parameterizations failed in Chesapeake Bay. Although the seasonal cycle of infective particle availability may be somewhat different, certainly the remaining processes should be equivalent in both bays. This suggested that an addi- tional process was needed to model transmission rate. Review of long-term time series taken simultaneously at mul- tiple sites across the salinity gradient in both bays revealed rela- tively simultaneous oscillations in disease prevalence with salinity change. Addition of bay-wide salinity-dependent multi-year oscil- lations in infective particle availability allowed both bays to be modeled with very minor differences in the values of only 2 vari- 490 Ford et al. ables, IPconc0 and IPconcmax. (Variations in IPconcmax are dis- cussed in the third paper in this series, Powell et al. this volume). These oscillations were parameterized as follows. The rate of sa- linity change was calculated as: IPsalrate = IPsalrate,, IPsaL (34) where lPsalrate0 specifies the response time of the infective par- ticles to changes in salinity, which was taken to be 180 days. The salinity value used to specify SIP can be considered representative of the salinity at which an hypothetical H. nelsoni secondary host lives or where some other reservoir of infective particles is found. For the simulations given in the following sections, the value of S/P was taken from the most down estuary (highest salinity) site showing strong salinity excursions across the 15 ppt isohaline in both Delaware and Chesapeake Bays. Lower salinity sites failed to provide adequate simulations in either bay. as discussed in Powell et al. (this volume) and higher salinity sites were not present in the suite of available Chesapeake Bay time series. The concentration of infective particles was updated at each time step based on this rate (IPsalrate) forced by the direction and migration of salinity change. So. for increasing salinities (IPsalrate a 0), dIPconc dt ■■ IPsalrate( IPconcmin - IPconc). ( 35 ) For decreasing salinities (/Psalrate < 0), dIPconc dt ■ = lPsalrate( IPconc - IPconcmin ) (36) and. at model initialization. IPconc IPconc was then inserted into equation (32). Temperature effects. Long-term observations from Dela- ware Bay show a cyclic pattern of H. nelsoni activity in which years of low infection prevalence follow, typically with a lag of 1 to 2 years, very cold winters (Ford and Haskin 1982). Examination of a 1989 to 1994 data set for Chesapeake Bay showed the same phenomenon. Thus, in some years, very few oysters become in- fected, even when appropriate salinity conditions are present (Haskin and Ford 1982. Paraso et al. this volume). This pattern suggests that, in some way. the abundance of infective particles is diminished after cold winters. In the model, direct temperature effects on infective particle abundance were included through a calculation of degree days that is based on 10 °C (DD10). This calculation differs from that for cold susceptibility (equation 9), which considers temperature ef- fects on H. nelsoni after it has infected the oyster. The number of days in which the temperature is below 10 °C from January to May is accumulated as: DD\0= 2 l°-T (37) where JD refers to Julian days. The value of DD10 is then used to determine an estimated degree to which cold temperature affects the survival of infective particles as: lPtempea ■■ 1.1, tank DD, DDW-DD, DD, (38) where DD„ is a threshold value at which the temperature effect becomes active. Equation (38) provides a value for the temperature effect that is based on the current degree-day calculation. To model the ob- served delay in the manifestation of winter temperature effects on H. nelsoni infective particles, the value of IPtemp determined from the current DD\0 value was modified based on the value calcu- lated for the previous year. A current value of DD10 less than one-half of the threshold value (DD„), indicates that the current year's winter is considerably warmer (an extreme difference) than that in the previous year, and the current value of [Ptempesl is used as IPtemp. If the current value of DD\0 is greater than one-half DD(I and less than the value for the previous year, such that the current year's winter is only slightly wanner than the previous year's winter, the current and previous year's values are averaged to obtain the value for IPtemp. This allows the conditions in the previous winter to affect the level of infectivity by H. nelsoni and thereby allows for persistence of the effects of harsh winters over a period of more than 1 year, as observed. If DD\0 is greater than one-half DD„ and greater than the value calculated for the previous year, then the current conditions are colder than previous year's conditions and also characteristic of a cold winter. In this case. IPtemp is specified using the current value of IPtempesr Data Sets Environmental Time Series The environmental inputs to the oyster population-//, nelsoni model are time series of temperature, salinity, food, and total seston (total suspended solids). The time series used for simula- tions presented in the next section are characteristic of the envi- ronmental conditions on the lower Delaware Bay planted grounds (Fig. 1 in Paraso et al. this volume). These reference simulations are intended to reproduce the annual H. nelsoni cycle in high salinity. Temperature measurements were made at a representative site, Miah Maull, by personnel from the Haskin Shellfish Research Laboratory at intervals of 1 to 3 measurements per month through- out the decade of the 1 960s. These data show that the winter of 1962 and those from 1968 to 1970 were particularly cold (Fig. 3a in Powell et al. this volume). Salinity time series for the 1964 to 1968 period were derived from monthly-averaged Delaware River flow measurements taken at Trenton. New Jersey, by the U.S. Geological Survey. Salinity time series were calculated using the relationship between Delaware River flow and salinity derived by Haskin (1972) as described in Paraso et al. (this volume). This relationship accurately represents salinity conditions during the 1960s in Delaware Bay. but may be less representative of salinities thereafter because of changing river flow to salinity relationships in the estuary (Haskin 1972). The 1960s were characterized by increasingly saline conditions in the first 6 years of the decade (Fig. 4 in Powell et al. this volume), followed by a freshening trend that began in 1967. The saline conditions in 1963 to 1967 coin- cided with a period of average-to-relatively mild winters. The salinity during this time was optimal for the proliferation and spread of H. nelsoni. The intent of the oyster-//, nelsoni model is to simulate the basic cycle observed for H. nelsoni prevalence and intensity. By using the time series for 1964 to 1968. the simula- tions were not influenced by anomalous environmental conditions that would limit H. nelsoni proliferation. Measurements of food and total seston at the Miah Maull site are not available for any time during the 1960s; however, total seston and chlorophyll measurements were made at other lower- estuary locations in Delaware Bay by Haskin Shellfish Research MSX Model Development and Verification 491 Laboratory (HSRL) scientists at about monthly intervals from 1981 to 1986, with the sampling frequency increased to bi-weekly between 1982 and 1984. The chlorophyll and total seston time series given by Powell et al. (1997) were used in the reference simulations. Measurements made at a site just south of Egg Island, New Jersey, were assumed to be representative of the Miah Maull planting grounds (Fig. 1 in Paraso et al. this volume). The 6-year time series from this site was averaged to obtain a single time series of 1-year duration that was used for each year of the simu- lations. Total suspended solids at the site showed variability throughout the year, with maximum values tending to occur in late spring to early autumn (Fig. 6 in Powell et al. this volume). The chlorophyll time series shows a distinctive spring bloom that occurs in March to May, with the maximum in March (Fig. 6 in Powell et al. this volume). A consistent fall bloom does not occur, although tran- sient increases in chlorophyll concentration do occur from time to time. Chlorophyll values drop to seasonally low levels in July and remain, for the most part, at or near these levels until the next spring. Chlorophyll a in u.g L"1 was converted to oyster food in nig DW L"' using the relationship derived by Powell et al. (1997) from Soniat et al. (1998): food = a x chlorophyll a + (3 (39) where a = 0.088 mgdw (p,g chlr1 and p = 0.26 mgdw L"1. H. nelsoni Prevalence and Intensity Time Series H. nelsoni prevalence and intensity were measured at numerous sites in Delaware Bay from 1959 to 1992 by personnel from the Haskin Shellfish Research Laboratory (Ford and Haskin 1982. Haskin and Ford 1982. Fegley et al. 1994). These measurements (Fig. 1 A) provide the calibration and verification for the reference simulation (described in the next section) obtained from the oyster- H. nelsoni model for lower Delaware Bay. Model Implementation The oyster-//, nelsoni model was solved numerically using a 2-step pseudo-steady state approximation scheme ( Verwer and van Loon 1994) with a time step of 1 hour. Each simulation begins on 1 June 1964 and extends through December 1968. The first simu- lation established a reference to which all other simulations were compared. The reference simulation was designed to reproduce the seasonal cycle of H. nelsoni prevalence and intensity as observed in a high-salinity location (Fig. 1A). Subsequent simulations were designed to show the modifications to this seasonal cycle that arise when some of the assumptions used in developing the oyster-//. nelsoni model were relaxed or removed (Table 4). In this regard, these simulations serve as a measure of the sensitivity of the model to the assumptions on which the model is based. Other simulations evaluate the response of the model to variations in environmental conditions. RESULTS Reference Simulation The simulated time-development of//, nelsoni infection in oys- ters from June 1964 to December 1968 (Fig. 15a), using the en- vironmental time series from the Miah Maull site in Delaware Bay. reproduces the observed annual cycle (Fig. 1A). The first (June 1964 to June 1965) and third (June 1966 to June 1967) years show TABLE 4. Simulations done with the oyster-//, nelsoni model to test the effect of certain model assumptions and environmental conditions on the simulated infection prevalence and intensity. For each simulation, the changes made in the environmental conditions, oyster size, or model dynamics relative to the conditions used to produce the reference simulation are given. The figure number showing the resultant simulation is indicated. Simulation Environmental data set Oyster size (g) Model change Figure number Reference Crowding effect Winter temperature Food effect on sporulation Spring food effect Oyster size-sporulation effect Oyster size-sporulation effect Winter temp-sporulation effect Cold winter Warm winter Miah Maull none 1964-1968 Miah Maull density-dependence 1964-1968 effect removed (equation 5) Miah Maull cold suscep. of 1964-1968 H. nelsoni removed (equation 10) decreased none food in 1965 no spring bloom none in each vear Miah Maull 0.3 none 1964-1968 Miah Maull 0.1 none 1964-1968 Miah Maull 0.1 winter temperature effect 1964-1968 on sporulation removed (equation 20) winter 1965-1966 1 none colder winter 1965-1966 1 none warmer 15a 15b 15c 16a 16b 17a 17b 18 19a 19b 492 Ford et al. iMiM n^^r^i it n fL k± Figure 15. Simulated time-development of Haplosporidium nelsoni in- fection for 1-g AFDW oysters in Delaware Bay using a) the environ- mental time series from June 1964 to June 1968. which represents the high-salinity, lower Bay grounds; h) with the density-dependent con- trol on Haplosporidium nelsoni growth, equation (5), removed; and c) with the cold susceptibility of Haplosporidium nelsoni, equation (10), removed. The term "cumulative fraction" means that the line for each BFU category represents the total prevalence of infections in that and all lower categories. the expected pattern in disease progression, with an increase in June to early fall (Fig. 1A. point 1 ), a plateauing in fall (Fig. 1A. point 2), the winter decrease (Fig. 1A. point 3), an increase the following spring (Fig. 1 A. point 4). and the decrease in late spring (Fig. 1A, point 5). Year 2 has a slightly modified version of this cycle, with the pattern during the late winter being less distinct. Year 4 of the simulation (Fig. 15a) shows the expected progression for the half year that is depicted. The simulated H. nelsoni infec- tions are initially primarily epithelial (BFU = 1) and progress rapidly to higher infection intensities. In the first and third years, about 30% to 40% of the oyster population has systemic infections of BFU > 2 by late summer. In the second year, over 50% of the oyster population is infected at this level. These year-to-year dif- ferences in prevalence result from the different environmental con- ditions in each year, as discussed in Powell et al. (this volume). The maximum total prevalences of about 60% to 80% that are attained in the early fall agree with the maximum prevalences reported for lower Delaware Bay at this time (Ford and Haskin 1982). Also, the partitioning of the disease between epithelial and systemic infections in the observed and simulated distributions is similar, with about 60% to 70% of the infections being systemic at peak prevalences (Fig. 1A). Thus, the simulated annual cycle of prevalence and intensity accurately reproduces both observed pat- terns and infection levels. Sensitivity of Density-Dependence and Cold Susceptibility Factors One of the assumptions made in the oyster-//, nelsoni model is that the plateau in disease intensity in late summer is due to self crowding by the parasites. However, since there is no direct ob- servation of this effect, it is instructive to determine how sensitive the model is to this assumption. To do this, equation (5) was set to zero. Without the density-dependent control, H. nelsoni prolifer- ates rapidly in the summer and triggers a large oyster mortality in December and January, sharply reducing prevalence by midwinter (Fig. 15b). Neither mortality nor a drop in prevalence is observed at this time in the field (Fig. IB). Even without the density depen- dent control, proliferation of H. nelsoni does slow in winter due to the cold temperatures, however, this reduction is not sufficient to limit oyster mortality. In particular, a second large oyster mortality event occurs in the late spring of the second, third, and fourth years of the simulation due to the very rapid increase in H. nelsoni cell number as temperatures increase in spring. The excessive mortali- ties of heavily infected oysters cause the simulated infection levels in surviving oysters to be lower than either the reference or ob- served values. Observed oyster mortality due to H. nelsoni does occur in late spring (Fig. IB), but it is only 10% to 15% of the oyster population rather than the nearly 50% that die in this simu- lation. Similarly, the removal of the cold susceptibility of H. nelsoni (equation 10) results in simulated disease prevalences and inten- sities (Fig. 15c) that do not reproduce the observed annual cycle. In the observed cycle (Fig. 1A) and the reference simulation (Fig. 15a), decrease in H. nelsoni prevalence and intensity does not occur in late winter. Removal of cold susceptibility predicts that the high parasite values that were present at the end of the previous summer and fall persist through the next spring. Increasing tem- peratures and subsequent rapid parasite proliferation result in in- fection prevalences (almost 80%) and intensities (nearly 80% sys- temic) that are higher than observed in late spring. These high disease levels are followed by a very large sporulation event and coincident oyster mortality in mid-summer, which is also not ob- served (Figs. IB. 15a). Sensitivity of Oyster Size and Environmental Conditions on Sporulation Of the many assumptions made in the development of the oys- ter-//, nelsoni model, those related to sporulation are mostly based on inferences made from observations of MSX disease progression in oyster populations and corresponding changes in the host, rather than from direct observation of the process itself. One of the basic assumptions made concerns the release of density-dependent con- trol on H. nelsoni growth in response to increased food levels in the spring. The sensitivity of the model to this assumption was tested by reducing the food supply in early 1965. which affects the calculation of I factor given by equation (16). The resulting simu- lation does not show an attempted sporulation event in the summer of 1965 (Fig. 16a). Rather, H. nelsoni prevalence remains high (BFU = 4) and about 70% of the oyster population is infected throughout the following year. In the spring of 1966, when the food levels return to the normal high values, a large attempted sporulation event occurs resulting in a sharp prevalence decline in 196S 1966 1967 1966 Figure 16. Simulated time-development of Haplosporidium nelsoni in- fection in BFUs (1 to 4) for a 1-g AFDW oyster in Delaware Bay after a) the oysters were exposed to low food values in 1965 and b) no spring hloom occurred in any year. MSX Model Development and Verification 493 early summer. Removal of the spring bloom in all years of the simulation ( Fig. 16b), disrupts the expected annual cycle com- pletely, indicating that food supply in the spring is crucial to at- tempted sporulation. Attempted sporulation events are either successful and spores are formed or unsuccessful in which case H. nelsoni mortality increases (Fig. 10). The difference in the two outcomes is assumed to be related to the size of the oyster. In the reference simulation (Fig. 15a), which uses a 1-g AFDW oyster, sporulation is at- tempted in early summer, but is unsuccessful. Parasite densities are reduced because failed sporulation leads to H. nelsoni death. How- ever, H. nelsoni cells in a 0.3-g AFDW oyster can undergo suc- cessful sporulation and release spores (Fig. 17a). In this simula- tion, one successful sporulation event occurred in each of the summers of 1966 and 1967. For smaller oysters, fall sporulation is also possible (Fig. 17b), as observed (Andrews 1979, Burreson 1994). There is no a priori reason to expect H. nelsoni to attempt sporulation at only 1 or 2 times per year. In fact, when the winter temperature effect on sporulation (equation 20) is removed, small oysters can sporulate into the winter and throughout the year (Fig. 18). However, observations indicate that this does not happen and therefore some factor, such as temperature, must be restricting this process to certain times of the year. Effect of Winter Temperature Many of the relationships in the oyster-//, nelsoni model are dependent on winter temperature. The sensitivity of the model to these assumptions can be tested by altering the winter temperature values in the temperature time series used as input to the model. Decreasing by 50% the 1965 to 1966 winter temperatures falling below 10 °C results in a prolonged period at temperatures of 0 °C to 5 °C, which increases the number of degree days during which H. nelsoni is exposed to cold. The resulting simulation (Fig. 19a) shows the expected annual cycle of disease progression, although prevalence is somewhat reduced relative to the reference simula- tion beginning in late 1965. Because cold winters affect transmis- sion in subsequent years (Powell et al. this volume), the major effect of the cold winter does not occur until the infection cycle beginning in the summer of 1966. Prevalences in that cycle and the following one are sharply reduced so that by the winter of 1968. only 10% of the oysters are infected. Thus, the effect of a single o n 6 U5 0) 4 a b 3 10 C ; u ° 6 4 - 1965 1967 1968 1966 Time Figure 17. Simulated time-development of successful sporulation events for a) a 0.3-g AFDW oyster and b) a 0.1-g AFDW oyster in Delaware Bay using the environmental time series from June 1964 and June 1968, which represents the high-salinity, lower Bay grounds. 1966 1967 1968 Time Figure 18. Simulated time-development of successful sporulation events for a 0.1-g AFDW oyster in Delaware Bay using the environ- mental time series from June 1964 to June 1968, representing the high-salinity lower Bay grounds. For this simulation, the winter tem- perature effect on H. nelsoni sporulation, equation (20), was removed. cold year can persist into subsequent years, even after winter tem- peratures have returned to normal. The effects of a warm winter were investigated by increasing by 50% the temperatures falling below 10 °C. In this simulation, the parasites spend little time at temperatures below 5 °C and do not experience the late- winter die off (Fig. 19b). As a consequence, parasite concentrations are already high at the start of the follow- ing spring. They increase further, resulting in heavy infections in the early summer of 1966 and consequent high oyster mortality. The return to normal winter temperatures in subsequent years re- sults in the same annual cycle as seen in the reference simulation. Thus, the effect of a single warm winter does not persist into subsequent years. DISCUSSION Model Characteristics A numerical model describing relationships between the pro- tistan parasite, Haplosporidium nelsoni, and its host, the Eastern oyster, Crassostrea virginica, has been developed. The model is unusually complex, particularly compared to that developed for the other major parasite of Eastern oysters, Perkinsus marinus ( Hof- mann et al. 1995. Powell et al. 1996). In the P. marinus-oystet model, in vivo parasite proliferation and death rates are a relatively simple function of temperature and salinity. Further, there is only a single life stage involved and transmission is dependent solely on the density of neighboring oysters and their infection level (Hof- mann et al. 1995, Powell et al. 1996). The complexity of the H. nelsoni model derives from the need to consider epithelial and systemic tissues as separate compartments, the failure of the para- BFU 1 ' l °'° BFU = 2 BFU = 3 - — BFU = 4 ; £ 0.6 ; 1 : i.U\ | 0.4 E u 0.2 = r if /V- -■..,:- -i i : :/, ty Figure 19. Simulated time-development of Haplosporidium nelsoni in- fection in BFUs (1 to 4) for a 1-g AFDW oyster in Delaware Bay with a) the winters of 1965-1966 made colder by decreasing by half the observed winter temperatures below 10 C and b) the winter of 1965 to 1966 made warmer by increasing by half the observed winter tem- peratures below 10 °C. 494 Ford et al. site to respond in a straightforward way to temperature and salinity change, the need to reproduce parasite sporulation only during certain times of the year and in certain size classes of oysters, and the decoupling of transmission from host infection levels or host density. Construction of the model involved making certain as- sumptions about the physiological or ecological processes under- lying the host-parasite relationship. Some of these assumptions are well grounded in experimental or observational data, or physi- ological principles; others are less so and may simply be surrogates for the true mechanism, but which happen to give the same answer. The following discussion considers these assumptions, as they occurred in the construction of the in vivo model. Assumptions made in the transmission component of the model are discussed in Powell et al. (this volume). Quantifying Infection Categories The model is quantitative: it uses parasites per oyster to track H. nelsoni infection development and decay. In contrast, the data used to construct and verify the model consist of semi-quantitative categories (LFU and BFU). which were converted into parasite densities by counting parasites in tissue section. Thus, a crucial assumption is that extrapolations from these counts adequately estimate total parasite burden, and that the conversion from LFUs to parasite numbers in the model is correct. In effect, the model converts from LFUs to parasite number for calculation and from parasite number back to LFUs (and then to BFUs) for data pre- sentation. As a result, most of the constants used in the model equations are dependent upon the conversion between LFU and parasite density given by equation 1. Should that relationship change with improved quantification methods, the absolute values of most model constants would also necessarily change. Diagnosis of P. marinus infections is also typically done using a semi-quantitative staging system (Mackin 1962). but a relatively accurate conversion between this system and parasite density ex- ists and was used in construction of the P. marinus model (Choi et al. 1989). The P. marinus conversion was achieved by a process that frees the parasites from oyster tissue for counting. Plasmodial stages of H. nelsoni are extremely fragile and would not survive this type of manipulation. Nevertheless, some comparisons be- tween extrapolated H. nelsoni densities and actual P. marinus counts are instructive. Estimates of H. nelsoni and P. marinus concentrations in the hemolymph of infected oysters have been made (Ford and Kanaley 1988, Ford et al. 1990, Gauthier and Fisher 1990. Bushek et al. 1994). For both parasites, maximum densities are in the range of 5 x 105 to 5 x 106 mL"1. Maximum densities of P. marinus in soft tissues are around 106 parasites gwwt"1 (Choi et al. 1989, Bushek et al. 1994). and our estimate of peak H. nelsoni concentrations from tissue sections was about the same. Further, the lethal level. 106 parasites gwwt"1, appears to be the same for both parasites, as higher densities are rarely found in live oysters. Interestingly, the estimated detection limit for H. nel- soni infections using tissue section histology (103-104 parasites gwwt-1 ) is similar to the detection limit found for P. marinus using the standard Ray/Mackin tissue subsample method (Choi et al. 1989, Bushek et al. 1994). These values suggest fundamental simi- larities in the per-parasite use of nutrients from, and the damage caused to. the oyster host. The Annual Infection Cycle within the Oyster The estimated in vivo doubling times for H. nelsoni used in the model were 1 to 1.4 days in the systemic tissues, and 3 to 4 days in the epithelium, over the 15-25 °C range. Over the same tem- perature range. P. marinus doubling times were estimated to range between 1.3 and 2.5 days (Hofmann et al. 1995). These rates fall well within the range for most free-living and symbiotic single- celled eukaryotes ( Laybourn-Parry 1987. Zaika 1973). The in vivo proliferation rate of//, nelsoni is based on a (?,„ of 3.2. This high value, set because lower values failed to provide adequate proliferation rates at elevated temperature, suggests that hi. nelsoni is very sensitive to temperature change. By comparison, the Ql0 used to model P. marinus cell division rates is 2.0 (Hof- mann et al. 1995). Under increasing temperature, then, H. nelsoni doubling rates should increase faster than those off. marinus and under decreasing temperatures, they should decrease faster. Over the temperature range where both parasites co-exist, approximately 0 °C to 35 °C. H. nelsoni has the higher proliferation rate. These comparisons of modeled proliferation rates are supported by field observations: when oysters are exposed to both parasites in the field. H. nelsoni begins killing before P. marinus does (Andrews 1967. Chintala et al. 1994). Declining autumn temperatures failed to slow the proliferation of H. nelsoni sufficiently to replicate the observed plateauing of infection levels at that time of year (Andrews 1966. Ford and Haskin 1982). Consequently, it was necessary to add a crowding factor such that, at high densities, proliferation is inhibited. There is no experimental evidence that this happens in H. nelsoni infec- tions, but it was also necessary to include a crowding effect in the P. marinus model (Hofmann et al. 1995) and there is experimental evidence that a density-dependent inhibition on proliferation does occur with this oyster parasite (Saunders et al. 1993, Ford et al. 1999). Further, a crowding effect is biologically defensible be- cause the host is a limited resource and at some point can no longer provide enough nutrients for all parasites. For both parasites, ample evidence exists that circulating and stored nutrients are di- minished by infection (Ford 1986. Barber et al. 1988. Chintala and Fisher 1991. Paynter 1996). The mechanism is analogous to cells in an in vitro culture, which reach a stationary phase of reduced division as culture-medium nutrients are exhausted and cellular byproducts accumulate. In the P. marinus and H. nelsoni models, crowding begins at similar parasite densities: 1 to 7 x lO'1 parasites gwwt-1. The P. marinus values were obtained from empirical data as described in Hofmann et al. (1995); those for H. nelsoni were determined by fitting model simulations to observed MSX disease prevalence and intensity. The similarities in the threshold values for the two parasites further supports evidence presented earlier, of fundamental similarities in the amount of nutrients and the damage produced by each parasite, be it a P. marinus or a H. nelsoni cell. The epithelium is one of the most important barriers to infec- tion encountered by an endoparasite. Although H. nelsoni readily enter the epithelium, it is truly a barrier because plasmodia pro- liferate along the base of epithelial cells, obviously prevented from immediate entry into the circulation and often accumulating con- siderable parasite loads in this layer before the first subepithelial parasites are observed (Farley 1968. Ford and Haskin 1982). In- fections confined to the epithelial layer are not lethal and often have few measurable effects on the oyster; further the ability to restrict parasites to the epithelium is one manifestation of resis- tance to MSX disease (Ford 1988, Ford and Tripp 1996). Conse- quently, the epithelium and the systemic tissues were considered as separate compartments in the model and the parasites behave somewhat differently in each. For instance, systemic parasites have faster division rates than do epithelial parasites, but become MSX Model Development and Verification 495 crowded at lower cell densities. It was necessary to assign different proliferation rates in order to fit the model to observed infection patterns, but there is good biological rationale based on histologi- cal observation and reasoning. Myhre (1973) pointed out that in the epithelium, plasmodia are located between oyster cells. Once they have become systemic, they are continuously bathed by hemolymph. Even though the shell cavity fluid of bivalves con- tains dissolved proteins, indicating the availability of nutrients to a parasite lodged in this compartment, levels are approximately half that in the hemolymph (Allam and Paillard 1998, Ford unpub- lished). Consequently, it seems reasonable to infer that the hemolvmph should provide more nutrients than the epithelium, and should allow faster multiplication. Why the crowding effects seems to run counter to this argument remains unclear, but without a higher crowding threshold in the epithelium, parasites rarely reached densities great enough to allow transfer into the systemic compartment. Although the crowding factor is based on the very plausible hypothesis of food limitation at high parasite densities, there may be another, less obvious, mechanism operating in the case of epithelial crowding. The mechanism by which plasmodia transverse the basal lamina and enter the circulatory system is not known, although structures known as haplosporosomes, which are common in the Haplosporidia, have been postulated to contain lytic enzymes that may aid in penetration of host tissues, including the basal lamina (Perkins 1968, Scro and Ford 1990). Nevertheless, it is clear that movement of plasmodia across the basal lamina is not a simple function of parasite replication; otherwise one would not expect to see an accumulation of parasites in this layer before they appear in the subepithelial space. The approach used to model the transfer was a simple diffusion equation that depends on the concentration of parasites in both compartments. This is admittedly an artificial mechanism for transporting an organism across a membrane; how- ever, the fact that it provided good results indicates that the true mechanism may have a similar basis. That is. the presence of large numbers of parasites is more likely to allow transfer, perhaps by weakening the basal lamina through the excretion of proteases, than is the presence of just a few plasmodia. In contrast, the P. marinus model does not consider the epithelium and systemic tissues as separate compartments and consequently the transfer of P. marinus across the epithelial barrier is a simple matter of para- site replication. The fact that this strategy works for P. marinus, but not for H. nelsoni, indicates an important difference in the way the two pathogens actually cross the barrier. In fact, it is likely that P. marinus is carried across within hemocytes, which routinely move between the epithelium and the circulatory system (Mackin and Boswell 1955, Alvarez et al. 1992). Thus, the chances of a phagocytosed P. marinus cell being carried across the basal lamina is likely to be the same for a single parasite as it is for one of many in an assemblage of parasites. In late winter, the observed infection cycle shows a marked prevalence and intensity decline, which is considered to be a com- bination of the deaths of heavily infected oysters and the mortality of H. nelsoni plasmodia in surviving oysters (Andrews 1966. Ford and Haskin 1982). The latter is concluded from the histological appearance of plasmodia at the time. They become dense, so that it is progressively more difficult to distinguish intracellular details, then begin to stain poorly, and finally are difficult to distinguish at all. Frequently they are inside hemocytes. It is not clear what the killing mechanism is. Low temperature is an obvious candidate, but enough parasites survive to initiate a new round of infection proliferation when temperatures begin to rise in the spring (Ford 1985a). Those parasites that do survive this period apparently are lodged in the epithelium, as that is the focus of renewed prolif- eration activity in spring. The initial attempt to model this observation was, in fact, to make H. nelsoni die as a direct result of exposure to low tempera- ture. This strategy failed to diminish the parasite burden fast enough, as did the use of an accumulator of low temperature, degree days. The addition of host hemocyte activity against para- sites made "susceptible" by prolonged (i.e.. degree day) cold, re- produced, in the model, the same infection decline recorded in nature. The use of degree days does not imply that low temperature alone is causing parasite deaths. Temperature could simply be a correlate for some other condition that the parasite experiences over the winter. Ford and Haskin (1982) hypothesized that a long period of anaerobiosis with a buildup of metabolic byproducts, rather than a direct cold effect, might be deleterious to H. nelsoni. In fact, the presence of abundant mitochondria in the plasmodia (Scro and Ford 1990) suggested a dependence on oxidative me- tabolism. Whereas the mechanism causing parasite degeneration over winter is unclear, the behavior of hemocytes toward them is explainable from experimental results. Hemocytes are becoming increasingly active with rising temperatures (Fisher and Tamplin 1988). Oyster hemocytes fail to attack and phagocytose live H. nelsoni, but they readily ingest and eliminate parasites in the post- winter period because the plasmodia are dead or damaged (Ford et al. 1993, Ford and Ashton-Alcox 1998). Thus, the need to add to the model, for the first time, an element of host activity is entirely in accord with both observed and experimental evidence. To fit the model to observations that declining infections persist longer in the epithelium than in the systemic tissues (Ford and Haskin 1982), systemic parasites made "susceptible" by cold are eliminated faster than those in the epithelium. Similarly, to reflect the observation that infections proliferate again from epithelial foci once temperatures begin to rise, the model sets faster recovery rates for the epithelial parasite population. This may reflect recov- ery of individual parasites or simply the component of undamaged parasites that remain. A possible biological explanation for the observed differences in epithelial and systemic locations is that there are probably more phagocytes per parasite in the hemolymph than in the epithelium so that the rate at which moribund parasites can be eliminated is consequently higher. Hemocyte numbers can become very high in epithelial lesions; however, they are fre- quently degenerate in appearance and being shed, along with para- sites, into the gill cavity (Farley 1968, Ford and Tripp 1996). Differences in hemocyte-to-parasite ratios appear plausible, but there is no evidence for this hypothesis and the actual reason may be quite different. The rate at which heavy infections decrease in late winter was observed to be slower than that for lighter infections. To model this event, it was necessary to have the overall effectiveness of the hemocyte population respond to parasite density, such that the response was relatively less effective at removing parasites at high H. nelsoni densities. It is reasonable that this could occur because of changing parasite-to-hemocyte ratios as infections intensify. The number of hemocytes in circulation and in tissues increases with increasing H. nelsoni infection intensity, but the change is relatively small (about 1.5-fold for circulating hemocytes. from a mean of 3.1 x 10'1 cells mLr1 in an uninfected oyster to a mean of 4.5 x 106 mLT1 in a heavily infected oyster) compared to the change in parasite concentration (from none to >105 mL-1) (Ford 496 Ford et al. and Kanaley 1988. Ford et al. 1993). The disproportionate increase in parasites means that the number of H. nelsoni cells removed by hemocytes becomes a progressively lower proportion of the total parasite population as the number of parasites increases. Once again, it was necessary to model different rates for the systemic and epithelial tissues to reproduce observed differences. Thus, in relation to their number, epithelial hemocytes remove more para- sites than do systemic hemocytes. There is no observational or experimental evidence for this model function other than the need for simulation to fit field observations of the H. nelsoni seasonal cycle. Up to this point in the annual infection cycle, late winter/early spring, the model relies on temperature, parasite-density, and hemocyte activity to replicate the observed seasonal changes in parasite loads. A new element was needed, however, to explain the rapid spring infection increase from pre-existing foci, and subse- quent sporulation. That element is oyster food, which remains of paramount importance throughout the remainder of the modeled annual cycle. Proliferation rates naturally increase with rising spring temperature, but the effect of temperature on parasite dou- bling time was inadequate to reproduce the observed, very rapid infection development in April and May. Particularly evident in field observations was the development of very heavy infections, indicating that high parasite division rates continued at densities where proliferation was otherwise restricted by self crowding. In addition to a rise in temperature in spring, the parasite experiences other changes inside the host. The oyster becomes active again after several months of quiescence over the winter. Oxygen avail- ability rises and the accumulation of end products from anaerobic metabolism ceases. A spring bloom typically occurs, and as oyster food consumption increases, the quantity of nutrients transported in the hemolymph rises (Fisher and Newell 1986). All of these changes should provide an increasingly favorable environment for H. nelsoni proliferation. Further, the fact that metabolic activity and nutritional status of the oyster is increasing in the spring should provide more or better resources for the parasite, and permit higher parasite densities before crowding interferes with replica- tion, than in late autumn when oyster metabolism is shutting down, even though nutrient reserves are generally high. Following this biological argument, the model eases the crowding effect so that higher parasite densities can be achieved rapidly in the spring. With this modification, simulations show the rapid infection in- tensification that occurs in the late spring and which culminates in what are often the highest parasite burdens of the year (Ford and Haskin 1982). Nutritional status, as modeled by oyster potential growth effi- ciency, is equally important in the next and last phase of the annual cycle, which is the production or attempted production of spores. It is also the most complex aspect of the annual cycle model. The observation that the model needed to fit was that the late May/early June prevalence peak is relatively brief, in contrast to the winter peak, and is followed by a rapid decline in prevalence (Andrews 1966, Ford and Haskin 1982). Like the loss of infections in late winter, part of this decline is due to the deaths of heavily infected oysters and part to the loss of parasites from live oysters. To simulate this event, a second life stage, the spore, was introduced into the model. In other members of the phylum Haplosporidia. Plasmodia regularly form spores (Perkins 1990), which presum- ably allow them to survive outside the host and are an important element in transmission. Haplosporidium nelsoni does form spores in adult oysters, but very rarely (Couch et al. 1966). Recent re- ports, however, suggest that spores are regularly formed in juve- nile oysters with advanced infections (R. D. Barber et al. 1991, Burreson 1994). Spore production coincides with the May/June infection peak and also occurs as infections intensify in the fall. Sporulation takes place in the epithelium of the digestive tubules and mature spores can be shed from live oysters; however, most oysters probably die during or after the sporulation process be- cause the overall infections are so heavy (R. D. Barber et al. 1991 ). Although spores are rare in adult oysters, histological observa- tions at the late May/early June infection peak suggest that some parasites may begin the sporulation process in adults. Oysters with advanced infections often have plasmodia in digestive tubule epi- thelia, sometimes with large, anomalous nuclei and a generally deteriorating appearance. We hypothesize that these plasmodia are evidence of failed sporulation. after which parasites die without completing their life cycle in the oyster. Their death consequently results in the post May/June drop in prevalence. Observational evidence, then, suggests a difference in the en- vironment experienced by H. nelsoni in young/small oysters, which allows the parasite to form spores, and that in larger/older hosts, which does not. This difference is not a question of differ- ential susceptibility or resistance because adult oysters of both types do not support spore formation. For purposes of the model, the internal environmental quality needed for sporulation was re- lated directly to the potential growth efficiency of the host and indirectly to food availability. Growth efficiency is an index to the amount of energy available after the host's basic metabolic re- quirements are met. This energy should be available to the parasite in the form of nutritional resources and relatively more of it should be available in younger oysters because of their higher growth efficiency. Spore formation, in the model, begins with the accumulation of nutritional reserves and the accompanying intensification of infec- tions. The parallel field observation is the movement of parasites into the digestive tubule epithelium, where they begin to undergo the many changes that accompany sporulation (Perkins 1969). The initial stages of simulated sporulation can happen regardless of oyster size, but to inhibit completion of the process in large oys- ters, the model establishes a threshold quantity of reserves that must be exceeded for spore production to occur. If that threshold is not reached, the process is not completed. Because of their higher growth efficiency, the threshold is exceeded only in small oysters, which consequently are the only oysters in which spores are formed. If the threshold is reached, sporulation is successful. Spores are shed from live oysters or after the host dies. The model considers that parasites that fail to sporulate are no longer viable. They become susceptible to hemocyte attack and are eliminated. In either case, resulting model simulations show a dramatic reduction in prevalence, as is seen in field observations. The growth-efficiency basis for sporulation used by the model is hypothetical, as is failed sporulation, to explain the early summer prevalence decline in adult oysters. Some other factor, perhaps a chemical or physical "cue" having nothing to do with growth efficiency or nutritional status, may well trigger sporulation. Or. there may be a suite of elements involved that occur in juveniles only. Nevertheless, the concept of a necessary threshold of some factor or factors remains a biologically defensible generalization for the fact that H. nelsoni can complete its life cycle in small oysters, but rarely in large ones. Modeling of the sporulation process needed to take into ac- count the observation that spores are formed in juveniles in the MSX Model Development and Verification 497 autumn, as well as in the spring (Andrews 1979. Burreson 1994). The process is probably set in motion in adults, too. but is rarely successful. In the fall, however, there is no abrupt prevalence decline. The model achieves this result in two ways. First, food supply is lower in the fall so only the smallest oysters have a potential growth efficiency adequate to trigger sporulation. Second, the model contains a temperature dependency on the loss of vi- ability of plasmodia that have failed to sporulate. Thus, if sporu- lation fails at relatively low temperature, plasmodia become less susceptible to hemocyte attack than those failing at relatively high temperature. Oysters remain infected and eventually die in late winter. In fact, if the "internal environment cue" hypothesis is correct and is related to the accumulation of nutrients, the slower reserve build up in adults compared to juveniles may simply retard the spore-formation process until the temperature is too low for parasite activity, so the plasmodia are never damaged. Salinity Effects Temperature is undoubtedly the most important environmental variable influencing the seasonal infection cycle, both directly and indirectly, and in the field and in the model. Salinity is also im- portant, but its effect is more obvious when considered on spatial or long-term temporal scales (Paraso et al. this volume. Powell et al. this volume). In the model, salinity affects H. nelsoni inside the oyster by affecting both survival and proliferation rates. Both are parameterized from in vitro experiments describing survival of plasmodia after acute salinity change (Ford and Haskin 1988). Results of these trials showed that survival was very low below about 9 ppt and very high above about 15 ppt. which roughly approximates its distribution in nature (Ford and Tripp 1996). Between those ranges, the parasite is highly sensitive to small salinity change. The model also considers that inside the oyster, parasites are buffered from rapid changes in salinity by the behav- ior of oysters themselves. When exposed to a large salinity change, bivalves typically close their valves and thereafter open them only briefly so as to allow entry of only small amounts of ambient water (Schoffeniels and Gilles 1972. Davenport 1979). The salt content of their body fluid thus changes more slowly than does the external water. Consequently, the model extends the in vitro death rate over a period of 4 days. In the absence of data on the effect of salinity on in vivo doubling times, it seems reasonable to assume that the salinity range over which it occurs is roughly the same as for survival, and that within this range, the response pattern is similar. In the model, salinity also affects the rate at which parasites move into the systemic tissue from the epithelium: at low salinity, the rate increases. This was a way to maintain the constant ratio of systemic to local infections observed along the salinity gradient (Haskin and Ford 1982. Fegley et al. 1994). Without it. the fre- quency of systemic infections decreased with decreasing salinity. Low salinity may. in fact, make it easier for parasites to make this transition, although the physiological mechanism is unclear. The actual reason may be quite different and this may be a case where the mathematical device provided a good approximation of ob- served patterns without a good biological rationale. Nevertheless. the need to include a factor that increased the proportion of sys- temic infections indicates that a simple salinity effect on parasite survival and growth is not sufficient to explain what is observed in field data. Oyster Mortality Oysters die. in the model, when H. nelsoni densities exceed that which is seen in live oysters. The same is true for the P. marinus model, but the H. nelsoni model also reflects the fact that the lethal parasite density for some oysters is lower than this maximally observed level. A few individuals die with relatively light infec- tions and an increasing proportion die as infections intensify. It is this variation in ability to tolerate infections that forms one of the bases for selective breeding: comparisons between oyster strains selected and unselected for resistance to MSX disease indicate that one measure of resistance is the ability to survive with relatively heavy infections (Ford and Haskin 1987, Ford 1988. B. J. Barber etal. 1991). Transmission Incomplete knowledge of the life cycle and mechanism of transmission of H. nelsoni is probably the single greatest impedi- ment to further understanding this important parasite and the dis- ease it causes. The sparsity of information about transmission made modeling this aspect of MSX disease particularly difficult because many assumptions had to be made. Yet the exercise was both intriguing and insightful. The transmission model is a sepa- rate component of the overall H. nelsoni-oyster model. It differs from most transmission models in that it simulates success or failure of transmission based on external environmental factors rather than on the density and infection levels of neighboring oys- ters. Modeling of the transmission process is detailed and dis- cussed by Powell et al. (this volume). SUMMARY The component of the H. nelsoni model that describes host- parasite interactions inside the oyster is constructed using func- tions describing physiological rates for both organisms: prolifera- tion, translocation, and death (or degradation) of the parasite; and hemocyte activity, filtration rate, and growth efficiency of the oyster. The rates, in turn, are controlled by four environmental variables: temperature, salinity, food, and total seston. Using only these few elements, the model is able to reproduce the bimodal annual infection cycle that includes infection intensification and remission, a life stage change of the parasite, response of the oyster's internal defense system, and. eventually, oyster death. With few exceptions, the physiological rate functions are based on experimental or observational evidence or general physiological principles. For instance, the effect of salinity on in vivo parasite survival, and the response of oyster hemocytes to dead or damaged parasites is well grounded with experimental, as well as observa- tional, data (Haskin and Ford 1982. Fisher and Tamplin 1988. Ford and Haskin 1988. Ford et al. 1993, Ford and Ashton-Alcox 1998). Parasite doubling times and the relationship between oyster mor- tality and infection intensity were computed directly from field data (Andrews 1966, Ford and Haskin 1982). Physiologically well- reasoned arguments were made for the self-crowding effect, the release of crowding in the spring, parasite degradation over the winter, differences in parasite growth and death rates between epithelial and systemic compartments, and the "threshold" trigger for sporulation. Whether failed sporulation in adult oysters is the cause for the rapid prevalence decline after the spring infection peak, whether lower salinity facilitates the movement of parasites from the epithelium into the systemic tissues, and the increased "efficiency" of the hemocyte component in the epithelium are 498 Ford et al. highly conjectural. Because virtually nothing is known about the transmission mechanism, this component of the model includes more hypothetical elements: specifically the infective dose thresh- old and the concentration of infective particles and their relation- ship to salinity and temperature (Powell et al. this volume). The fact that certain hypothetical mechanisms were used to fit the model to observation does not detract from its efficacy. Be- cause the simulations reproduce observed temporal and spatial patterns, and assuming that the major biological and physical sys- tems involved have, at some level, reasonably predictable re- sponses, the model suggests ways in which the host-parasite sys- tem must work. For instance, the modeling exercise clearly shows that temperature effects on parasite doubling times or salinity ef- fects on in vivo parasite survival, cannot by themselves, explain field observations. The model demonstrates that other factors must be involved and points to where efforts must be concentrated to gain a better understanding of the overall host-parasite relation- ship. Clearly, an improved knowledge of the complete system rests with a better understanding of the parasite's life cycle and mode of transmission, combined with an ability to infect oysters experi- mentally. Nevertheless, the fact that this very complex and detailed model works, with few modifications, in Chesapeake Bay as well as in Delaware Bay, is a measure of its power and potential use- fulness in other areas. ACKNOWLEDGMENTS We thank Bob Barber at HSRL for examination of oysters to provide the LFU-to-parasite abundance relationship. This research was supported by the Virginia Graduate Marine Science Consor- tium grant VGMSC 5-29222 and by the New Jersey Sea Grant under contract number 4-25238. Computer resources and facilities were provided by the Center for Coastal Physical Oceanography at Old Dominion University. The Delaware River and Bay Authority funded the 1981-1984 monitoring program that provided data for some of the environmental time series used in the model. Con- tinuation of the time series through 1986 was made possible by funds from the New Jersey Department of Environmental Protec- tion. Both programs were coordinated by Walt Canzonier. The States of New Jersey and Maryland provided funds for collection of the Delaware Bay and Chesapeake Bay Haplosporidium nelsoni time series. This is Contribution number 99-16 of the Institute of Marine Science at Rutgers University and NJAES Publication #D- 32405-1-99. LITERATURE CITED Allam, B. & C. Paillard. 1998. Defense factors in clam extrapallial fluids. Dis. Aquat. Org. 33:123-128. 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. Inver- tebr. Pathol. 59:290-294. Alvarez, M. R.. F. E. Friedl, J. S. Johnson & G. W. Hinsch. 1989. Factors affecting in vitro phagocytosis in oyster hemocytes. J. Invertebr. Pathol. 54:233-241. Andrews. J. D. 1966. Oyster mortality studies in Virginia. Epizootiology of MSX, a protistan parasite of oysters. Ecology 47:19-31. Andrews. J. D. 1967. Interaction of two diseases of oysters in natural waters. Proc. Natl. Shell. Assoc. 57:38-48. Andrews. J. D. 1968. Oyster mortality studies in Virginia. VII. Review of epizootiology and origin of Minchinia nelsoni. Proc. Natl. Shellfish. Assoc. 58:23-36. Andrews. J. D. 1979. Oyster diseases in Chesapeake Bay. Mar. Fish. Rev. 41:45-53. Andrews. J. D. 1983. Minchinia nelsoni (MSX I infections in the James River seed-oyster area and their expulsion in spring. Estuarine. Coastal and Shelf Sci. 16:255-269. Barber. B. J., S. E. Ford & H. H. Haskin. 1988. Effects of the parasite MSX [Haplosporidium nelsoni) on oyster (Crassostrea virginica) energy me- tabolism. II. Tissue biochemical composition. Comp. Biochem. Physiol. 91A:603-608. Barber. B. J., S. E. Ford & D.T.J. Littlewood. 1991. A physiological comparison of resistant and susceptible oysters Crassostrea virginica (Gmelin), exposed to the endoparasite. Haplosporidium nelsoni (Haskin. Stauber, and Mackin). J. Exp. Mar. Biol. Ecol. 146:101-1 12. Barber. B. J., R. R. Langan & T. L. Howell. 1997. Haplosporidium nelsoni (MSX) epizootic in the Piscataqua River Estuary (Maine/New Hamp- shire, U.S.A.). J. Parasilol. 83:148-150. Barber. R. D. & S. E. Ford. 1992. Occurrence and significance of ingested haplosporidan spores in the eastern oyster, Crassostrea virginica (Gmelin. 1791). J. Shellfish Res. 11:371-375. Barber. R. D., S. A. Kanaley & S. E. Ford. 1991. Evidence for regular sporulation by Haplosporidium nelsoni (MSX) (Ascetospora: Hap- losporidiidae) in spat of the American oyster. Crassostrea virginica. J. Protozoal. 38:305-306. Burreson. E. M. 1988. Use of immunoassays in haplosporidan life cycle studies, pp. 298-303. In: W. S. Fisher (ed.). Disease Processes in Ma- rine Bivalve Molluscs. Spec. Pub. 18, Am Fish Soc, Bethesda. MD. Burreson. E. M. 1994. Further evidence of regular sporulation by Hap- losporidium nelsoni in small oysters, Crassostrea virginica. J. Parasi- lol. 80:1036-1038. Bushek. D., S. E. Ford & S. K. Allen. 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 Diseases 4:201-217. Chintala, M. M. & W. S. Fisher. 1991. Disease incidence and potential mechanisms of defense for MSX-resistant and susceptible eastern oys- ters held in Chesapeake Bay. J. Shellfish Res. 10:439^*43. Chintala. M. M„ S. E. Ford, W. S. Fisher & K. A. Ashton-Alcox. 1994. Oyster serum agglutinins and resistance to protozoan parasites. J. Shell- fish Res. 13:115-121. 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. Quan- tification of the thioglycollate method. J. Shellfish Res. 8:1 17-125. Corliss, J. O. 1984. The kingdom Protista and its 45 phyla. BioSystems 17:87-126. Couch. J. A.. C. A. Farley & A. Rosenfield. 1966. Sporulation of Mini- chinia nelsoni (Haplosporida, Haplosporidiidae) in Crassostrea vir- ginica (Gmelin). Science 153:1529-1531. Davenport. J. 1979. Is Mytilus edulis a short term osmoregulator? Comp. Biochem. Physiol. 64A:91-95. Farley, C. A. 1968. Minchinia nelsoni (Haplosporida) disease syndrome in the American oyster Crassostrea virginica. J. Protozool. 15:585-599. Farley. C. A. 1975. Epizootic and enzootic aspects of Minchinia nelsoni (Haplosporida) disease in Maryland oysters. J. Protozool. 22:418-427. Fegley, S. R„ S. E. Ford. J. N. Kraeuter & D. R. Jones. 1994. Relative effects of harvest pressure and disease mortality on Eastern Oyster population dynamics in Delaware Bay. Final Report to NOAA. NMFS Oyster Disease Research Program. Rutgers. The State University of New Jersey. 208 pp. Feng, S. Y. & J. S. Feng. 1974. The effect of temperature on cellular reactions of Crassostrea virginica to the injection of avian erythro- cytes. J. Invertebr. Pathol. 23:22-37. Fisher, W. S. & R. I. E. Newell. 1986. Seasonal and environmental varia- MSX Model Development and Verification 499 Hon in the protein and earbohydrate levels in the hemolymph from American oysters (Crassostrea virginica Gmelin). Comp. Biochem. Physiol. 85A:365-372. Fisher. W. S. & M. Tamplin. 1988. Environmental influence on activities and foreign-panicle binding by hemocytes of American oysters. Cras- sostrea virginica. Can. J. Fish. Aqnat. Sci. 45:1309-1315. Ford. S. E. 1985a. Chronic infections of Haplosporidium nelsoni (MSX) in the oyster Crassostrea virginica. J. Invertebr. Pathol. 45:94-107. Ford, S. E. 1985b. Effects of salinity on survival of the MSX parasite Haplosporidium nelsoni (Haskin. Stauber, and Mackin) in oysters. J. Shellfish Res. 2:85-90. Ford, S. E. 1986. Comparison of hemolymph proteins between resistant and susceptible oysters, Crassostrea virginica. exposed to the parasite Haplosporidium nelsoni (MSX). J. Invertebr. Pathol. 47:283-294. Ford. S. E. 1988. Host parasite interactions in oysters, Crassostrea vir- ginica. selected for resistance to Haplosporidium nelsoni (MSX) dis- ease: survival mechanisms against a natural pathogen, pp. 206-224. In: W. S. Fisher, (ed.). Disease Processes in Marine Bivalve Molluscs. Spec. Pub. 18, Am. Fish. Soc. Bethesda, MD. Ford. S. E. 1996. Range extension by the oyster parasite Perkinsus marinus into the northeastern US: response to climate change? J. Shellfish Res. 15:45-56. Ford, S. E. & K. A. Ashton-Alcox. 1998. Altered response of oyster hemocytes to Haplosporidium nelsoni (MSX) plasmodia treated with enzymes or metabolic inhibitors. J. Invertebr. Pathol. 72:160-166. Ford, S. E. & H. H. Haskin. 1982. History and epizootiology of Haplospo- ridium nelsoni (MSX). an oyster pathogen, in Delaware Bay. 1957- 1980. J. Invertebr. Pathol. 40:118-141. Ford. S. E. & H. H. Haskin. 1987. Infection and mortality patterns in strains of oysters Crassostrea virginica selected for resistance to the parasite Haplosporidium nelsoni (MSX). J. Parasitol. 73:368-376. Ford, S. E. & H. H. Haskin. 1988. Comparison of in vitro salinity tolerance of the oyster parasite Haplosporidium nelsoni (MSX) and hemocytes from the host, Crassostrea virginica. Comp. Biochem. Physiol. 90A: 183-187. Ford, S. E. & S. A. Kanaley. 1988. An evaluation of hemolymph diagnosis for detection of the oyster parasite Haplosporidium nelsoni (MSX). J. Shellfish Res. 7:11-18. Ford. S. E.. S. A. Kanaley & K. A. Ashton-Alcox. 1993. In vitro interac- tions between bivalve hemocytes and the oyster pathogen Haplospo- ridium nelsoni (MSX). J. Parasitol. 79:255-265. Ford. S. E.. S. A. Kanaley. M. Ferris & K. A. Ashton-Alcox. 1990. "Pan- ning." a technique for enrichment of the parasite Haplosporidium nel- soni (MSX) from hemolymph of infected oysters. J. Invertebr. Pathol. 56:347-352. Ford, S. E., A. Schotthoefer & C. Spruck. 1999. In vivo dynamics of the microparasite Perkinsus marinus during progression and regression of infections in eastern oysters. J. Parasitol. 85:273-282. Ford, S. E. & M. R. Tripp. 1996. Diseases and defense mechanisms, pp. 383-150. In: R. I. E. Newell. V. S. Kennedy and A. F. Eble (eds.). The Eastern Oyster Crassostrea virginica. Maryland Sea Grant College. College Park, MD Galtsoff, P. S. 1964. The America Oyster. Crassostrea virginica Gmelin. United States Department of the Interior. Washington. D.C. 480 pp. Gauthier. J. D. & W. S. Fisher. 1990. Hemolymph assay for diagnosis of Perkinsus marinus in oysters Crassostrea virginica (Gmelin. 1791). J. Shellfish Res. 9:367-372. Haskin. H. H. 1972. Delaware River flow-bay salinity relationships. Report to the Delaware River Basin Commission: Phase III. 12 pp. Haskin. 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.). Disease Processes in Marine Bivalve Molluscs. Spec. Pub. 18, Am. Fish. Soc., Bethesda, MD. 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. Invertebr. Pathol. 40:388-405. Hofmann. E. E.. J. M. Klinck, E. N. Powell. S. Boyles & M. Ellis. 1994. Modeling oyster populations II. Adult size and reproductive effort. /. Shellfish Res. 13:165-182. Hofmann. E. E.. E. N. Powell. J. M. Klinck & G. Saunders. 1995. Model- ing diseased oyster populations I. Modeling Perkinsus marinus infec- tions in oysters. J. Shellfish Res. 14:121-151. Hofmann. E. E., E. N. Powell. J. M. Klinck & E. A. Wilson. 1992. Mod- eling oyster populations III. Critical feeding periods, growth and re- production. J. Shellfish Res. 1 1:399-+ 16. Laybourn-Parry, J. 1987. Protozoa, pp. 1-25. In: T.J. Pandian and F.J. Vernberg (eds.). Animal Energetics, vol. 1. Protozoa through Insecla. Academic Press, New York. Mackin, J. G. 1962. Oyster diseases caused by Dermocystidium marinum and other microorganisms in Louisiana, pp. 132-229. In: J. G. Mackin and S. H. Hopkins (eds.). Studies on Oysters in Relation to the Oil Industry. 7. Publication of the Institute of Marine Science, Texas A&M University. Mackin. J. G. & J. L. Boswell. 1955. The life cycle and relationships of Dermocystidium marinum. Proc. Natl. Shellfish. Assoc. 46:1 12-1 15. Matthiessen. G. C, S. Y. Feng & L. Leibovitz. 1990. Patterns of MSX (Haplosporidium nelsoni) infection and subsequent mortality in resis- tant and susceptible strains of the eastern oyster Crassostrea virginica (Gmelin. 1791) in New England. J. Shellfish Res. 9:359-366. Myhre. J. L. 1973. Levels of infection in spat of Crassostrea virginica and mechanisms of resistance to the haplosporidian parasite Minchinia nel- soni. M. S. Rutgers University, New Brunswick, NJ. 102 pp. Paraso, M. C, S. E. Ford. E. N. Powell, E. E. Hofmann & J. M. Klinck. 1999. Modeling the MSX parasite in eastern oyster (Crassostrea vir- ginica) populations. II. Salinity effects. J. Shellfish Res. Paynter, K. T. 1996. The effects of Perkinsus marinus infection on physi- ological processes in the eastern oyster, Crassostrea virginica. J. Shell- fish Res. 15:119-125. Perkins. F. O. 1968. Fine structure of the oyster pathogen Minchinia nel- soni (Haplosporida. Haplosporidiidae). J. Invertebr. Pathol. 10:287- 307. Perkins. F. O. 1969. Electron microscope studies of sporulation in the oyster pathogen. Minchinia costalis (Sporozoa; Haplosporida). J. Para- sitol. 55:897-920. Perkins. F. O. 1990. Phylum Haplosporidia. pp. 19-29. In: L. Margulis. J. O. Corliss, M. Melkonian and D.J. Chapman (eds.). Handbook of Protoctista. Jones and Bartlett, Boston. Powell, E. N., E. E. Hofmann & J. M. Klinck. 1996. Modeling diseased oyster populations II. Triggering mechanisms for Perkinsus marinus epizootics. J. Shellfish Res. 15:141-165. Powell, E. N.. J. M. Klinck. S. E. Ford, E. E. Hofmann & S. J. Jordan. 1999. Modeling the MSX parasite in eastern oyster (Crassostrea vir- ginica) populations. III. Regional application and the problem of trans- mission. J. Shellfish Res. 18:515-536. Powell, E. N.. J. M. Klinck. E. E. Hofmann & S. E. Ford. 1997. Varying the timing of oyster transplant: implications for management from simulation studies. Fish. Oceanog. 6:213-237. Powell, E. N.. J. M. Klinck. E. E. Hofmann & S. M. Ray. 1994. Modeling oyster populations IV. Rates of mortality, population crashes, and man- agement. Fishery Bull. 92:347-373. Ray, S. M. 1996. Historical perspsective on Perkinsus marinus disease of oysters in the Gulf of Mexico. J. Shellfish Res. 15:9-1 1. Saunders, G. L., E. N. Powell & D. H. Lewis. 1993. A determination of in vivo growth rates for Perkinsus marinus. a parasite of Crassostrea virginica. J. Shellfish Res. 12:229-240. Soniat, T. M.. E. N. Powell. E. E. Hofmann, & J. M. Klinck. 1998. Under- standing the success and failure of oyster populations: the importance of sampled variables and sample timing. J. Shellfish Res. 17:1149- 1165. Schoffeniels. E. & R. Gilles. 1972. Ion-regulation and Osmoregulation in Mollusca. pp. 393-420. In: M. Florkm and B. T. Scheer (eds.). Chemi- cal Zoology, vol. 7: Mollusca. Academic Press, New York. 500 Ford et al. Scro, R. A. & S. E. Ford. 1990. An electron microscope study of disease progression in the oyster Crassostrea virginica infected with the pro- tozoan parasite Haplosporidium nelsoni (MSX). pp. 229-254. In: F. O. Perkins and T. C. Cheng (eds.). Pathology in Marine Science. Aca- demic Press. Orlando. FL. Shumway. S. E. 1996. Natural Environmental Factors, pp. 467-513. In: R. 1. E. Newell. V. S. Kennedy and A. F. Eble (eds.). The Eastern Oys- ter Crassostrea virginica. Maryland Sea Grant College. College Park. MD. Sprague. V.. E. A. Dunnington & E. Drobeck. 1969. Decrease in incidence of Minchinia nelsoni in oysters accompanying reduction of salinity in the laboratory. Proc. Natl Shellfish. Assoc. 59:23-26. Stokes. N. A.. M. E. Siddall & E. M. Burreson. 1995. Detection of Hap- losporidium nelsoni (Haplospondia: Haplosporidiidae) in oysters by PCR amplification. Dis. Aquat. Org. 23:145-152. Verwer. J. G. & M. van Loon. 1994. An evaluation of explicit pseudo- steady-state approximation schemes for stiff ODE systems from chemi- cal kinetics. J. Comput. Pins. 1 13:347-352. Zaika. V. E. 1973. Specific Production of Aquatic Invertebrates. John Wiley & Sons. New York. 154 pp. Journal of Shellfish Research. Vol. 18. No. 2. 501-516, 1999. MODELING THE MSX PARASITE IN EASTERN OYSTER (CRASSOSTREA VIRGINICA) POPULATIONS. II. SALINITY EFFECTS MICHELLE C. PARASO,1 SUSAN E. FORD,2 ERIC N. POWELL.2 EILEEN E. HOFMANN,1 AND JOHN M. KLINCK3 Coastal Ocean Laboratory National Oceanographic Data Center. NOAA/NESDIS Silver Spring. Maryland 20910 "Haskin Shellfish Research Laboratory Rutgers University Port Norris, New Jersey 08349 ' Center for Coastal Physical Oceanography Old Dominion University Norfolk, Virginia 23529 ABSTRACT An oyster population model coupled with a model for Haplosporidium nelsoni, the causative agent of the oyster disease MSX. was used with salinity time-series constructed from Delaware River flow measurements to study environmentally-induced variations in the annual cycle of this disease in Delaware Bay oyster populations. Model simulations for the lower Bay (high salinity) site reproduced the annual cycle observed in lower Delaware Bay. Simulations at both upper Bay (low salinity) and lower Bay sites produced prevalences and intensities that were consistent with field observations. At all sites, low freshwater discharge resulted in increased disease levels, whereas high freshwater discharge produced decreased levels. At upper Bay sites, simulated changes in runoff produced high variability in disease prevalence; in the lower Bay. they produced a much lesser effect. Changes in salinity within the 10-20 ppt range produced the greatest changes in disease levels and patterns. Simulated shifts in timing of the spring runoff from March to either February or May affected the mid-Bay (13-19 ppt) only. A February runoff reduced the spring prevalence peak and caused a complete loss of systemic infections. In contrast, a May discharge occurred too late to affect parasite proliferation in the spring so that the spring peak was higher than average. Almost 100% of the infections were systemic by June, which resulted in high oyster mortality during July at this site. Model results indicate that parasite infection intensity under changing salinity is more complex than a simple function of salinity as it affects parasite proliferation and death rates within the oyster, and that the rate of infection is most likely reduced at low salinity. The simulated results demonstrate the ability of the model to reproduce field measurements and its usefulness in elucidating the association between the magnitude and timing of Delaware River discharge, its associated salinity variations, and the H. nelsoni annual cycle. KEY WORDS: MSX, model, eastern oyster. Crassostrea virginica, Haplosporidium nelsoni. Delaware Bay, salinity INTRODUCTION The Eastern oyster, Crassostrea virginica, has been an impor- tant economic resource in Delaware Bay (Fig. 1 ) for at least 2 centuries (Ford 1997). The practice of "farming" oysters, which began in the mid- 19th century, involves the transplantation of seed oysters during May and June from natural setting areas (seed beds) in the upper regions of the Bay to privately leased sections (planted grounds) in the lower Bay (Ford and Haskin 1982). Salinity of the water covering the seed beds ranges from 9 to 18 ppt under mid- tide, mean river-flow conditions, which protects the juvenile oys- ters from major predators (e.g., oyster drills) that are intolerant of low salinities and hence enhances oyster survival. The planted grounds are located in areas of higher salinity (20-23 ppt) that promotes growth and fattening of the oysters. Traditionally, oys- ters have been marketed the fall following the spring planting or have remained on the planted grounds an additional year or more before harvesting (Ford and Haskin 1982). In the late 1950s and early 1960s, the causative agent of the Eastern oyster disease MSX (Multinucleated Sphere Unknown), was first observed in Delaware and Chesapeake Bays (Andrews 1966. Haskin et al. 1966). It was described as a haplosporidian by Haskin et al. (1966) and named Haplosporidium nelsoni (Sprague 1978). By 1959, 90%-95% of the planted ground oysters and about 50% of seed bed oysters had died in Delaware Bay. Gradu- ally, in the late 1960s and the 1970s, Delaware Bay seed beds recovered as native oysters developed some resistance to MSX disease (Haskin and Ford 1979. Ford 1997). Following 15 years of modest success, the Delaware Bay oyster industry suffered another setback due to a resurgence of H. nelsoni, which made significant incursions onto the seed beds during a drought in the mid 1980s. The annual infection cycle (Fig. 2) of H. nelsoni in lower Delaware Bay oysters was described by Ford and Haskin ( 1982). New infections are acquired beginning in early June and for the remainder of the summer, infections increase in number and in- tensity, reaching peak infection levels in late fall. Prevalences remain relatively stable through late winter, and then drop as heav- ily infected oysters die and parasites are lost from the remaining oysters. Prevalence and intensity increase again in April and May, coincident with rising temperatures, and a resurgence of infections that were suppressed by low winter temperatures (Ford 1985b). The spring peak, often having the most intense infections of the year, is typically followed by a rapid loss of infections, the reason for which is not completely understood. By this time, infective particles are once again present in the water and the cycle begins again. Since poikilothermic and poikilosmotic animals are directly affected by the temperature and salinity of the environment in which they live, interactions between a parasite and a poikilother- mic or poikilosmotic host may be strongly influenced by external environmental conditions. For example, temperature and salinity- have well-documented effects on the relationship between H. nel- soni and C. virginica (e.g.. Andrews 1964, Andrews 1983, Ford 1985a, Ford and Haskin 1988). Early field observations suggested that the parasite was salinity limited (Andrews 1964, Andrews 501 502 Paraso et al. Figure 1. Location of Delaware Bay. Inset shows locations of the New Jersey oyster seed beds (sites 1—1) and the planted ground (sites 5-7) stations. The solid line indicates the area designated as planted grounds. The (*) indicates the locations where time series of tempera- ture, chlorophyll a, and total seston were acquired. The dashed line distinguishes between shoal and channel areas as designed by Pennock and Sharp (1986). 1983, Farley 1975, Haskin and Ford 1982). Although oysters are capable of activity at salinities as low as 5 ppt, infected oysters are extremely rare at salinities below 10 ppt and MSX disease does not become epizootic at salinities below 15 ppt (Andrews 1964, An- drews 1983, Haskin and Ford 1982). Ford (1985a) observed a rapid loss of parasites as mid-tide salinities decreased from 15 to 5 ppt. Full parasitic activity occurs at salinities above 20 ppt (Haskin and Ford 1982, Andrews 1983). Haskin and Ford (1982) proposed two explanations for the observed decrease in H. nelsoni prevalence along a salinity gradi- ent in Delaware Bay. The first hypothesis, which does not involve a direct salinity effect, is that the principal source of infective particles is in the lower Bay. Consequently, the concentration of H. nelsoni infective particles in the upper Bay would be reduced due to increased distance from their source (Haskin and Ford 1982). The second hypothesis is that reduced prevalence in low salin- ity areas was the result of a physiological response, either by the host or the parasite. Andrews (1983) suggested that low salinities might enhance the oysters' natural defenses, resulting in active expulsion of the parasite. Fisher and Newell ( 1986) found that in vitro rates of hemocyte locomotion were greater in low. compared to high, salinity and suggested that enhancement of hemocyte ac- tivity at low salinity might play a role in the expulsion of H. nelsoni from oysters in low salinity water. On the other hand, because the oyster does not osmoregulate, the parasite is exposed to salinities that mirror the ambient water. Thus, the parasite may be physiologically incapable of surviving in less than 10 ppt (An- drews 1964, Andrews 1983, Haskin and Ford 1982. Ford 1985a). Ford (1985a) proposed that a physiological limitation may restrict the distribution of H. nelsoni in low salinity water. Tolerance of plasmodial stages to acute in vitro salinity change roughly parallels BO 60 40 20 — i 1 i r Infection Period "i 1 1 1 r T * * Figure 2. Schematic of the annual Haplosporidium nelsoni infection cycle reproduced from Figure 4 in Ford and Haskin 1982. Filled ar- rows represent the percentage of oysters in each infection level (preva- lence), with epithelial infections representing the lightest oysters (BFU category 1 1 and general systemic infections representing the heaviest infections (BFU category 4). The infection period designates the time in which infections begin in uninfected oysters and are designated by the unfilled arrows. The length of the arrows represents the relative in- tensity of infection pressure. The area above the general systemic in- fection line represents the portion of oysters in the undetectable cat- egory. the parasite's distribution in nature and indicates that its reduced occurrence in low salinity water is most probable due to its physi- ological inability to tolerate reduced salinity, rather than to en- hanced effectiveness of host defense mechanisms (Ford and Haskin 1988). To investigate the effect of environmental factors, specifically salinity, on the time and space distribution of MSX disease in Delaware Bay oyster populations, simulations were done with a time-dependent numerical model for oyster populations that was coupled with a model for H. nelsoni. The oyster-//, nelsoni model includes the physiological processes that affect the parasite-host interaction and are responsible for the predictable cycle in infec- tion intensity and prevalence observed in Delaware Bay (Ford et al. this volume). Simulations were done to determine how seasonal variations (timing and strength) in Delaware River flow affect //. nelsoni prevalence and intensity at various sites within the Bay and the results from these were then compared with long-term field observations from the same sites. MATERIALS AND METHODS The Delaware Bay Study Site Delaware Bay was chosen as a study site because a long-term database on MSX-disease exists for its oyster population. The Bay (Fig. 1) is a funnel-shaped estuary, extending 215 km from the head at Trenton, New Jersey, to the mouth between Cape Hen- lopen. Delaware, and Cape May. New Jersey (Wong 1994. Wong Modeling the MSX in Eastern Oyster Populations II 503 1995, Ford 1997). The Delaware River, gauged at Trenton, New Jersey, contributes approximately 589<- of the total freshwater input to the estuary, with an average discharge of 330 nr s~' (Sharp et al. 1986. Wong 1995). The Schuylkill River entering through Philadelphia. Pennsylvania, near the head of the Bay. contributes another 15% of the total freshwater input. No other single fresh- water source provides more than 1% of the total discharge (Sharp et al. 1986). A direct, negative correlation exists between salinity in Delaware Bay and the Delaware River flow (Fegley et al. 1994. Wong 1995). Five sites in Delaware Bay were selected for this study (Fig. 1 ). Arnolds. Cohansey. Shell Rock, and Bennies Beds are on natural seed beds, whereas Miah Maull Grounds are located in the planted grounds. The depth of the seed beds ranges from 5 to 7 m. The location and area of each seed bed is described in Fegley et al. (1994). Arnolds Bed is located 64.1 to 67.0 km upbay from the mouth of the Bay and 0. 1 to 2.9 km off the New Jersey shore, and has an area of 2.32 km2. Cohansey Bed. located 54.4 to 58.5 km upbay and 0. 1 to 3.4 km offshore, is 5.45 km2. Shell Rock Bed with an area of 4.04 km2, is located 50.5 to 53.4 km upbay and 1 .4 to 3.6 km offshore. The seed bed site furthest downbay. Bennies Bed, is located 45.4 to 49.7 km upbay and 1.7 to 5.4 km offshore, and has an area of 6.36 km2. Miah Maull Grounds is the site closest to the mouth and is representative of the entire leased area. This site is located approximately 25.9-27.8 km from the mouth. 6.5 km from the New Jersey shore, and occupies an area of ap- proximately 32.3 km2. Oyster /Vputaio/i-Haplosporidium nelsoni Model General Characteristics A time-dependent oyster population-//, nelsoni model, de- scribed in Ford et al. (this volume), was used to stimulate infection prevalence and intensity along the salinity gradient in Delaware Bay. The host-parasite model contains components to simulate the dynamics of a single size-class oyster population (1 g dry wt. in this case), the proliferation and death of//, nelsoni in those oysters (Fig. 5 in Ford et al. this volume), and transmission of H. nelsoni (Powell et al. this volume). The model components are coupled by relationships that describe the infection of uninfected oysters, the removal of oyster energy by the parasite to support its metabolism, and the relationship of parasite density in host tissue to oyster mortality. Governing Equation The governing equation for the oyster-parasite model moves oysters among H. nelsoni infection classes as parasite densities increase or decrease in the epithelial ie) and systemic (s) tissue of the oyster (O) over time. It is formulated as: dO.. — = -«„sO„ - p.,,0,.., + a„_, (>,__, ( 1 ) + P,-i,Ol-1..> + a,,>+lO„+1 + 3..+ i,sO,+ i..,-M1,.lOt,v-8,,54Oe.,v4 . N N - s=s4e=0 where the first 6 terms represent the movement of oysters between infection intensity classes through gains or losses of H. nelsoni cells in the epithelial and systemic tissue (Fig. 3 in Ford et al. this volume). The coefficients, a and S, determine the rate at which H. nelsoni cells are gained or lost. The parameterizations used to determine these coefficients are given in Ford et al. (this volume). The seventh term in equation ( 1 ) represents the loss of parasites through oyster mortality from lethal infections as determined by the rate of mortality, M. The final 2 terms in equation ( 1 ) represent the transfer of oysters from heavy infection classes to lower in- fection classes due to the formation or attempted formation of spores by H. nelsoni in the oysters with advanced infections, which results in a loss of parasites from infection classes with BFU = 4 (s4 in equation 1 ). and a gain of infected oysters into infection class [1.0]. The 8 functions represent a step-function process in which the oysters are introduced into the [1,0] infection class only. The coefficient 7 determines the rate of this transfer process. The establishment of infection in uninfected oysters ([0.0] class) is determined by the equation: dOt ~jf = -Po.fAi.o + 2 8f.o8(U 2 2 1e,sO. (2) 4 p=0 where the first term represents the acquisition of H. nelsoni infec- tive particles at a rate determined by B00. The second term rep- resents addition of oysters to the uninfected class after hypoth- esized abortive H. nelsoni sporulation events, as described in Ford et al. (this volume), and is given by the eighth term in equation ( 1 ). These oysters are divided evenly between the [0.0] and [1,0] in- fection classes, as given by the last terms in equations ( 1 ) and (2). The Oyster Population Model The model component that describes the time-dependent de- velopment of the oyster population model is that described in Powell et al. (1992, 1994. 1995. and 1996) and Hofmann et al. ( 1992. 1994). For the purpose of this study, only a single size class of oyster ( 1 g dry wt) was used. The oyster model includes pa- rameterization for the processes that determine the production of somatic and reproductive tissue, which are based on assimilation efficiency, filtration rate, ingestion, and respiration as modified by the environmental conditions of food supply, total seston. tempera- ture and salinity. The H. nelsoni Model Parameterization of the processes that were used to develop the oyster population-//, nelsoni model, and the biological bases for the mathematical relationships, are described and discussed by Ford et al. (this volume). Processes that influence H. nelsoni in- teractions with its host include the rate of infection as a function of oyster filtration rate and ambient infective particle concentration; a density-dependent control on parasite proliferation based on oyster accumulated potential growth efficiency; parasite sporulation or attempted sporulation related to oyster accumulated potential growth efficiency; and H. nelsoni mortality due to cold exposure and increased susceptibility to oyster hemocytes, which them- selves are influenced by temperature. The processes by which oysters move among infection categories is described in Ford et al. (this volume). In addition to in vivo processes, the host-parasite model con- tains a subcomponent for disease transmission, which is dependent on infective dose, the filtration rate by which infective particles reach the oyster, and seasonal and salinity-dependent processes that control the concentration of infective particles in the water column (Powell et al. this volume). As this study is directed at salinity effects on the MSX disease process, the parameterizations involving salinity are discussed in more detail below. 504 Paraso et al. Salinity Parameterizations The effect of salinity on H. nelsoni mortality was parameterized using measurements of acute in vitro salinity tolerance of Plasmo- dia, the most common stage in oysters (Ford and Haskin 1988). This mortality effect begins at 1 5 ppt, increases in intensity from 15 to 5 ppt. and causes rapid H. nelsoni mortality below 5 ppt. A sigmoidal function was fit to observations of the fraction of live Plasmodia remaining after exposure to a range of salinities and modified to ensure that no parasite mortality occurred above sa- linities of 15 ppt (Fig. 3). It is formulated by Smort = min I 1, - 0.01 SD, 1 + - SD, -SD, (3) SD, where Smort is the fraction of parasites still alive after exposure. SD, 103. SD, .240650. SD, = 0.592456 ppt-'. and S is salinity in ppt. The salinity response was assumed to have a time scale of 4 days, which reflects the estimated period required for oyster hemolymph to equilibrate with external salinity after an ambient salinity change and for parasites to respond to the new conditions. The death rate of H. nelsoni, Sdeath, expressed as d~x is then given by: Sdeath ■■ -ln(Smort) SD4 (4) where SD4 = 4.0. Under normal circumstances it is assumed that H. nelsoni infections, once acquired, are never lost. Therefore, the oysters cannot decrease through H. nelsoni cell mortality to a level lower than an initial epithelial infection with no systemic infection (e,, s0). However, for salinity mortality, it is assumed that the infection can be lost at salinities <14 ppt. which is equivalent to Sdeath = 0.01787. It was assumed that salinity effects on parasite proliferation rates occur over the same salinity range producing cell mortality. The effect of salinity on cell growth was parameterized as an exponential relationship that produces reduced growth rates at sa- linities less than 14 ppt and essentially no growth at salinities less than 5 ppt. The form of the salinity-dependent H. nelsoni growth rate, Sfactor, is: Sfactor (5) where sg determines the rate of decrease of parasite proliferation rate with increasing salinity and S0 is 15 ppt, which represents the threshold above which no reduction in parasite proliferation rate occurs. The salinity effect on cell growth is such that it has the value of zero at S < 5 ppt and the value of one at S > S0. Equation (5) modified the temperature-dependent H. nelsoni growth rate that is given in equation (4) of Ford et al. (this volume). Initial simulations with the above model showed prevalences on the seed beds that were higher than those observed (Haskin and Ford 1982, Fegley et al. 1994). Subsequent simulations indicated that the source of this mismatch was related to the rate at which oysters became infected and suggested that reduced salinity, in some way. reduced infection efficiency. The model thus contains a parameter that varies transmission rate by reducing the number of infective particles at low salinity and is of the form: 100 1 1 1 1 1 ' 1 * - 80 */ - 60 - - 40 ■ m - 20 ■ X . 0 , , 1 , 1 , 0 10 20 30 Salinity (ppt) Figure 3. The relationship between salinity and acute in vitro H. nel- soni mortality based on observations (asterisks) by Ford and Haskin (1988) of the fraction of H. nelsoni plasmodia appearing viable (i.e., excluding Trypan Blue dye) after exposure to a range of salinities. + tunh\ SM IPsal ■■ (S-SM0)\ SM, ) (6) where IPsal is the fractional reduction in spore number that is applied to the calculation of spores filtered given by equation (29) in Ford et al. (this volume). The constants 5M,. SM,, and SM, have the values 1 .6, 11 .0, and 7.0, respectively, and S is the ambient salinity. Equation (6) is biologically valid because the salinity range affecting transmission is similar to the range causing cell mortality in vivo. However the effect described by equation (6) has a somewhat wider range, as might be expected by direct exposure to a free-living infective particle. Whether salinity causes cell mor- tality or just reduced infectivity in infective particles is unknown at present. The initial simulations for the seed bed sites also showed a decreasing infection intensity at upbay sites, which is inconsistent with observations that showed a single-step drop in infection in- tensity from the planted grounds to the seed beds, with little ad- ditional change related to decreasing salinity (Haskin and Ford 1982). Essentially, the simulations underestimated the systemic infections at low salinity. To include this effect the rate of H. nelsoni diffusion from the epithelial to systemic tissue, Sdiff, was increased at low salinity using the relationship: Sdiff-- + SF, 1 . - tank SF, S-S, HfJ (7) where the reference salinity, S0 is 15 ppt and the constants SF,, SF2, and SF, have the values 9.0, 2.65, and 3.0 ppt, respectively. Modeling the MSX in Eastern Oyster Populations II 505 The value of Sdiff was set to be 0.955 at 12.0 ppt. 0.5 at 15.0 ppt, and 0.005 at 18.0 ppt. Equation (7) gives the fractional reduction in H. nelsoni cell diffusion rate that modifies equation (6) in Ford et al. (this volume). This relationship is based on the rationale that osmotic stress increases the leakiness of the basement membrane thereby permitting H. nelsoni cells to more easily cross to the oyster circulatory system. Environmental Data Sets The environmental factors that force the oyster population-//. nelsoni model are salinity, temperature, food, and total seston. Data sets for temperature, chlorophyll <; (representing food), and total suspended solids (representing total seston), collected over a 5-year period (1981-1986), were obtained from the Haskin Shell- fish Research Laboratory of Rutgers University. Collections used for this study were made at Section E (representing the seed beds) and at the Ridge (representing the planted grounds) sites (Fig. 1). Data collected in each month were averaged over all years to construct a representative 1-year time series. Temperature Data Sets Temperatures for both Section E and Ridge time series began to increase during February and reached a maximum of about 26 °C in July and August (Fig. 4A). Temperature decreased with the onset of autumn and a minimum temperature of (2 °C) occurred in February. Food Data Sets Overall, chlorophyll a measurements, representing food avail- able to the oyster, were approximately 5-10 p.g L ' higher at Ridge than at Section E (Fig. 4B). Chlorophyll levels began to increase earlier (January) at the Ridge site, compared to the Sec- tion E site, where the increase began in February. Maximum chlo- rophyll concentrations occurred during March and were approxi- mately 55 (xg L~' at Ridge and 40 p,g L~' at Section E. Chloro- phyll then decreased steadily at Section E. with a minimum of approximately 3 p-g L"1 occurring in October. Levels at Ridge remained relatively high (32 p.g L"1) through May, then decreased to a minimum of 8 p.g L"1 in October. During verification studies of the oyster-/3, marinus model in Galveston Bay, Texas (e.g., Powell et al. 1995), agreement be- tween the simulated oyster populations and observed populations was improved if a non-chlorophyll component was added to the chlorophyll measurements to obtain total food supply. This modi- fication to the chlorophyll time series was done using an empirical relationship between chlorophyll and total available food (Soniat et al. 1998) of the form: Food ■■ (8) : a x chlorophyll + h where Food is given in mg DW L_I and chlorophyll is in p,g L~ . The constants a and b have the values of 0.088 mg DW u,g_1 and 0.52 mg DW L_1, respectively. This equation recognizes that total available food can include a significant non-chlorophyll compo- nent. Similarly, verification studies with the oyster-P. marinus model for Delaware Bay (Powell et al. 1997) showed that chloro- phyll alone did not provide an adequate food supply for the oys- ters. Therefore, the relationship between chlorophyll and total available food in Delaware Bay was assumed to be of the same 30 E Section E A : -^ z Ridge ^.<^ ~~~~*!^- - S-5 z sls^ ^~^s> - „, 20 — 6^ ^^. ~ 3 : ^s^ o / / N. \ - Z 20 o '" / ^""-0- ■ . 50 = Section E C ! p 40 | Ridge / "~-^ i % 30 o O 20 K .'""" "-. V- 10 0 I _ ---' \^_ -"1 I . Arnolds Bed . Cohansey Bed . Shell Rock Bed Bennies Bed Mioh Moull Grounds Jan Feb Mor Apr May Jun Jul Aug Sep Oct Nov Dec Month Figure 4. Time series of environmental data measured by the Haskin Shellfish Research Laboratory (monthly averages for 1981 to 1986), used as input to the model to simulate average conditions. Panels A, B, and C show time series of temperature, food, and total seston (TSP = total suspended particulates) data, respectively, for seed bed (Section E) and planted ground (Ridge) sites. Panel D shows mean salinity time series, estimated from freshwater inflow data, for the seed bed (Ar- nolds, Cohansey, Shell Rock, and Bennies Beds) and planted ground (Miah Maull Grounds) sites. form as that for Galveston Bay and the coefficients were deter- mined by an iterative procedure in which simulations and obser- vations were compared (see Powell et al. 1997 for details). This procedure resulted in a conversion to food supply given by equa- tion (7), but with b having the value 0.26 mg DW L"'. Total Seston Data Sets Total seston. as measured by total suspended solids, was rela- tively constant at Section E (10-12 x 103 p.g L_I) with a slight increase in May to 16 x 103 pg L~' (Fig. 4C). The Ridge total seston time series was more variable than in Section E, ranging from 20 x 103 p.g L"' in November to 41 x 103 p.g L"1 in August (Fig. 4C). Total seston values were greater at Ridge for all months. Salinity Data Sets Monthly-averaged Delaware River flow for the period 1913 to 1993 was measured at Trenton. New Jersey, by the United States Geological Survey. Salinity time series were then calculated for 506 Paraso et al. 1953 to 1993. using a relationship between Delaware River flow and salinity derived by Haskin (1972): v = c + dx'' (9) where y is salinity in ppt calculated from the preceding mean 30-day river flow (x) in ft3 s"1 (1 ft3 s"' = 2.83 x 10"2 m3 s"1). Derived values of c . d. and e for each of the 5 sites shown in Figure 1 are presented in Table 1 . The resultant salinity time series were used as input into the oyster population-W. nelsoni model. At each site, the 8 years (20%) between 1953 and 1993 with the highest calculated average monthly salinity were grouped together and a mean salinity calculated for each month at each site. The resultant time series were assumed to be representative of a year with low freshwater runoff (Fig. 5A). Similarly, the 8 years (20<7c ) with lowest monthly mean salinities were also grouped for each site and the salinity time series was taken to represent a year with high freshwater runoff (Fig. 5B). Salinities for the remaining 24 years (60%) were averaged to create a mean freshwater runoff time series (Fig. 4D). Although the salinity values changed (Table 2). the seasonal pattern for average and low runoff years was similar (Figs. 4D, 5A). Salinity minima occurred in March/April and the maximum values extended from July through October. The pattern for a high runoff year was only slightly different, with a distinct minimum occurring in April and a shortened period of maximum salinities extending from July through September (Fig. 5B). To simulate a shift in the timing of spring runoff, the maximum freshwater discharge, which occurred during March/April in the mean time series (Fig. 4D), was moved to either February or May (Fig. 5C. D). For these simulations, only the timing was changed: the quantity of freshwater discharge remained that of an average year. H. nelsoni Prevalence and Intensity Data Prevalence and intensity of H. nelsoni in Delaware Bay oysters was measured over the period 1 959 to 1 992 by personnel from the Haskin Shellfish Research Laboratory (Ford and Haskin 1982. Haskin and Ford 1982, Fegley et al. 1994). Oysters from the lower Bay planted grounds were obtained by dredge at regular intervals during the year. Seed bed oysters were typically collected during late autumn/early winter and the following late spring. These dates were chosen to coincide with peaks in the annual infection cycle (July 1 to June 30 of the following year) (Fig. 2). Stained tissue sections of each oyster were examined to determine the distribu- tion and abundance of H. nelsoni in the tissues, and these were TABLE 1. Derived constants for the Delaware River flow-Delaware Bay salinity relationship (y = a + bx') for the 5 sites displayed in Figure 1, where v is the salinity in ppt and x is river flow in ft3 s"1 (From Haskin 1972). a b c Site ppt ppt (s ft-3)' non-dimensional Arnolds Bed -37.85 99.32 -0.08036 Cohansey Bed -38.91 102.32 -0.07465 Shell Rock Bed -45.91 104.95 -0.06083 Bennies Bed -64.25 122.04 -0.04575 Miah Maull Grounds 19.86 152.16 -0.41722 Arnolds Bed . . Shell Rock Bed _ _ _ Cohansey Bed Bennies Bed Miah Maull Grounds 30 25 S. 20 F- — — -~~~ A : .J 15 o 10 5 izzzz}-- - - Ii>^^ ^^ ; 30 p 25 S 20 25 Q. S 20 | 15 o 10 5 30 25 Q. a. 20 I 15 o VI 10 ..an Apr May Jun Ji Month Aug Sep Figure 5. Idealized time series of salinity, constructed from Delaware River flow measurements, used as input to the model to simulate A) low freshwater runoff, B) high freshwater runoff. C) a year with an early spring (February) freshwater runoff period, and D) a year with a late spring (Mayl freshwater runoff period at seed bed (Arnolds. Cohansey, Shell Rock, and Bennies Beds} and planted ground (Miah Maull Grounds) sites. used to develop a scale. Big Ford Units (BFU, described in Ford et al. this volume), for comparing observed and simulated MSX dis- ease infection and intensity. The BFUs were to define intensity categories as: BFU = 0; no infections or infections too light for detection by standard histology (hereinafter referred to as "undetectable in- fections"). BFU = 1; Epithelial infections - H. nelsoni plasmodia in the gill epithelium only. BFU = 2: Subepithelial/local infections - H. nelsoni have broken through the basal lamina between the epithelium and the underlying tissues, but remain largely concentrated near that site. BFU = 3; Light systemic infections - H. nelsoni have spread through the tissues via the circulatory system, but parasite abundance remains relatively low. BFU = 4; Advanced systemic infections - H. nelsoni have spread throughout the tissues via the circulatory system and parasite abundance is high. These are typically lethal infec- tions. Modeling the MSX in Eastern Oyster Populations II 507 TABLE 2. Salinity ranges for high freshwater runoff, average freshwater runoff, and low freshwater runoff time series (Figs. 4D, 5A, and 5B| for the sites displayed in Figure 1. The high value in each range represents the late summer values and the low value represents the spring values. High Runoff Average Runoff Low Runoff Station (ppt) (pptt (ppt) Arnolds Bed 5.1-9.5 7.2-12.9 9.6-15.4 Cohansey Bed 7.8-12.2 9.9-15.6 12.3-18.2 Shell Rock Bed 9.5-13.7 11.5-16.9 13.8-19.3 Bennies Bed 11.2-15.5 13.3-18.7 15.6-21.1 Mian Maull Grounds 21.8-23.0 22.4-24.4 23.0-25.7 The distribution of BFUs in relation to epithelial and systemic infections is shown in Figure 4 of Ford et al. (this volume). Observed infection prevalences and intensities were used to verify model simulations. In the case of the planted ground simu- lations, comparisons were made with the complete annual cycle. For the seed beds, comparisons were made with the winter (No- vember/December) and spring (late May/early June) prevalence peaks. Observed category 1 and 2 infections were combined to represent the prevalence of localized infections and categories 3 and 4 were combined to represent systemic infections. These data were used to construct a time series of prevalence and intensity (localized vs. systemic infections) in oyster populations at each seed bed and planted ground site (Fig. 6). Simulations for Miah Maull Grounds represent the annual H. nelsoni cycle and infection levels of the entire leased area, since there is relatively little varia- tion within this area as compared to the variation among seed beds. Since measurements at Miah Maull Grounds were sparse, they were supplemented with data from Deepwater and Southwest Line planted grounds (Figs. 1,7). Model Implementation The oyster population-//, nelsoni model was solved numeri- cally using the 2-step pseudo-steady-state approximation scheme (Verwer and van Loon 1994) with a time step of 1 hour. Each simulation began on June 1 and extended for 3 years, with the first 2 years needed for the solution to reach a steady state. The same input data were used for each of the 3 years. The model assumes that the oysters have been continuously exposed to H. nelsoni, as is the natural stock, so that the population starts each cycle with the infection intensity and prevalence remaining at the end of the cycle. To investigate the effect of salinity variations on the prevalence and intensity of H. nelsoni in Delaware Bay oyster populations, a number of stimulations were considered under various environ- mental conditions (Table 3). A reference simulation was run for each site using mean environmental conditions. A series of simu- lations to determine the influence of high or low freshwater runoff on infection levels were subsequently run and compared to the reference simulation. Finally, the effect of the timing of the fresh- water runoff was investigated. Statistical Analysis To analyze variations in prevalence, intensity, and timing of//. nelsoni infections in Delaware Bay oyster populations as a result of the various environmental scenarios, several statistical analyses BO A -: hO 40 1 1 20 lln 0 -n nrflfi ■ ;. _ H _. H . N — 1 n " 1975 Year nln J i lNHWMNH^Mm.WH NN NN E - ^NNMHN NN UN NN NN NN NN NN hIM NN NN »J Figure 6. Prevalence and intensity of Haplosporidium nelsoni infec- tions for A) Arnolds Bed, B) Cohansey Bed, C) Shell Rock Bed, D) Bennies Bed, and E) Miah Maull Grounds constructed from long-term data of Haskin Shellfish Research Laboratory (Fegley et al. 1994). Solid portions of bars represent localized infections (BFU categories I and 2); clear portions represent systemic infections (BFU categories 3 and 4). Annual cycles are represented by 2 samples. Bars on the left of the tick mark represent the spring (May/June) measurements; bars on the right represent the early winter (November/December) measure- ments. Periods when no data were available are indicated by N. Note that after 1992, Perkins us marinus, cause of Dermo disease in oysters, became prevalent on the seed beds and confounded H. nelsoni patterns. were performed on the simulated output. The maximum differ- ence, average difference, root mean square (RMS), and correlation coefficient (r) were calculated for each site to compare each simu- lation to the reference simulation. Comparisons were made for each infection intensity (BFU) category between time series for each hypothetical extreme-condition simulated and the corre- sponding reference time series. The maximum difference is the greatest percentage point dif- ference in prevalence for a particular infection category between the simulated extreme and reference annual infection cycles. A high value means that an extreme condition caused a relatively large change in that category of infection compared to the refer- ence simulation; a low value means that the extreme condition caused relatively little change: maximum difference = max(\A, , - Bn\) (10) 508 Paraso et al. 1955 1960 1965 1970 1975 1980 1985 1990 1995 Yeor 100 20 B . wWN.^w»wmw 1960 1965 1970 1975 Year 1980 1985 1995 Figure 7. Prevalence and intensity of Haplosporidium nelsoni infec- tions in oysters at A) Southwest Line and B) Deepwater planted grounds from records of the Haskin Shellfish Research Laboratory (Fegley et al. 1994). Solid portions of bars represent localized infec- tions (BFU categories 1 and 2); clear portions represent systemic in- fections (BFU categories 3 and 4). Annual cycles are represented by 2 samples. Bars on the left of the tick mark represent the spring (May/ June) measurements; bars on the right represent the early winter (November/December) measurements. Periods when no data were available are indicated by N. Note that after 1990, Perkinsus marimis, cause of Dermo disease in oysters, became prevalent on the New Jersey planted and confounded H. nelsoni patterns. where A,, = A(t) or the MSX prevalence time series for a given location (/) and MSX intensity category (/) and B,, = B(t) for the reference simulation at the same site and intensity category. The average difference is the average percentage point differ- ence in prevalence of a particular infection category between simu- lated extreme and reference annual infection cycles. A high value means that an extreme condition caused a relatively large change in that category of infection; a low value means that the extreme condition caused relatively little change: average difference =— ?, I(A, , - B, , (ID where T is the length of the simulated time series produced by model simulations. TABLE 3. Environmental time series for model simulations. Salinity for Anrolds, Cohansey. Shell Rock, and Bennies Beds, and Miah Maull Grounds are indicated by A, C, S, B, and M, respectively. "Dry" and "wet" indicate low and high freshwater runoff, respectively. Timing of freshwater discharge is indicated by the month to which the spring discharge event was moved. R and E represent time series constructed from field measurements at Ridge and Section E sites, respectively. Simulation Site Temperature salinity Food Supply Turbidity (°C> (ppt) (ugchlaL1) (mgl1) Reference A E A E E C E C E E s E S E E B E B E E M R M R R Low fresh- A E A dry E E water C E Cdry E E s E Sdry E E B E B dry E E M R M dry R R High fresh- A E A wet E E water C E C wet E E s E S wet E E B E B wet E E M R M wet R R Early spring A E A Feb E E runoff C E CFeb E E s E SFeb E E B E B Feb E E M R M Feb R R Late spring A E A May E E runoff C E C May E E S E S May E E B E B May E E M R M May R R The RMS (root mean square) is a standard measure of the difference between 2 time series. Squaring the prevalence differ- ences between the extreme A(f) and B(t) reference time series before taking the average enhances the larger differences and di- minishes the smaller differences. Therefore, a relatively large dif- ference will result in a larger value than a smaller difference. Again, a high value means that an extreme condition caused a relatively large change in that category of infection; a low value indicates that the extreme condition caused relatively little change: RMS; f^(Au BUY (12) Spearman's Rank Correlation Coefficient (r) which is given by: Car (13) -AA V "-SB was used to evaluate correlations of ranked prevalences of each infection category between simulated extreme (A(t)) and reference (B(r)) annual infection cycles. The coefficient CAB is defined as the I,7! ,{A, , - A)(fi, , - S), where A and B are the mean for time series A,, and B, ,. respectively. Values of r near 1 indicate similar tem- poral patterns (although absolute values can still vary). Values of /- near zero indicate different temporal patterns. Modeling the MSX in Eastern Oyster Populations II 509 RESULTS Model Simulations Reference Time Series Prevalence The first set of simulations provided the basic time develop- ment of H. nelsoni infection prevalence and intensity at the seed bed and planted ground sites. These reference simulations were forced by average environmental conditions (Table 3) at each lo- cation and provide a basis for comparison with simulations pro- duced by extreme environmental conditions. At Arnolds Bed, the most upbay and least saline bed. mean environmental conditions produce no detectable infections (Fig. SA). Infections are first detected just downbay at Cohansey Bed during August, increase in prevalence and intensity during the fall, and reach a winter prevalence maximum of 9% by late October (Fig. 8B). At Shell Rock Bed. a similar pattern is seen with new infections appearing a little earlier, in July, and winter prevalence rising to 20% by November (Fig. 8C). At both sites, little change in total prevalence occurs over the winter. All detectable infections are lost in February and do not reappear in the spring. Bennies Bed is the first site where the simulation results in an annual cycle more typical of higher salinity regions (with 2 peaks) and retention of some infections throughout the year (Fig. 8D). Prevalence and intensity increase through summer and fall, reach- ing a maximum by November, when the prevalence is about 45%. Loss of infections in all categories, but primarily systemic, occurs in March: infections re-appear in the spring, coincident with rising water temperatures and salinities. Prevalence reaches 25% in the late May/early June peak. The reference simulation for Miah Maull, in the planted grounds, produces the highest infection levels (Fig. 8E). From a low of 5% to 107r in July, prevalence increases to 60% to 65%, with a little more than half the infections being systemic by No- vember. A loss of primarily systemic infections causes a total prevalence decline in March and April, followed by a second prevalence peak of 55% in late May, in which most infections are advanced. A rapid loss of infections follows by early summer (see also Ford et al. this volume). I0U c 80 60 40 20 0 4,1 . JJASONDJFMAMJ Month 100 B 80 - 60 40 .'0 - n 1,4 \-*-<*-, , ^A* 1,4 1 4,1 , JJASONDJFMAMJ Month Figure 8. Simulated, time evolution of Haplosporidium nelsoni preva- lence and intensity produced by average environmental conditions for A) Arnolds Bed. B) Cohansey Bed. C) Shell Rock Bed, Dl Bennies Bed, and E) Miah Maull Grounds. Numbers 1 through 4 refer to the cu- mulative percentage of total oysters (prevalence) with epithelial (1). subepithelial local (2). light systemic (3), and advanced systemic (4| infections. The area above the category 4 line represents the portion of oysters in the undetectable infection category. Reference Time Series Intensity Under average conditions, approximately half the simulated infections in seed bed oysters are systemic at the November/ December and late May/early June peaks (Fig. 8A-D). At the Miah Maull Grounds site, about half the winter infections are systemic. In contrast, the spring peak was predominantly systemic. with a large fraction in category 4. advanced systemic infections (Fig. 8E). Extremes in Freshwater Runoff Simulations Low Freshwater Runoff. During periods when freshwater run- off is lower than average, salinity in Delaware Bay increases (Fig. 5A) and elevated H. nelsoni infection prevalence results at all sites. Unlike average conditions, low freshwater runoff simulations pro- duce detectable infections at Arnolds Bed. beginning in August (Figs. 8A, 9A). Infections develop slowly through the fall, reach- ing a winter prevalence peak of 8%. At Cohansey Bed. detectable infections appear in June and reach a maximum prevalence of about 35% by November. At Arnolds Bed. all detectable infections are lost in January: the decline occurs in February at Cohansey. where nearly all infections are lost. At neither site do infections re-appear in the spring. Under the low runoff conditions, the simulated annual cycle for Shell Rock Bed is similar to that for Bennies Bed under average conditions: a 2-peak annual cycle with a substantial fraction of infections persisting throughout the year (Fig. 9C). The winter prevalence peak is approximately 45%. Loss of infections, primar- ily systemic, occurs in March. The spring peak prevalence is about 30%. Low-flow simulations produce almost identical cycle pat- terns and levels at Bennies Bed, representing the lower seed beds, as they do at Miah Maull. in the planting grounds (Fig. 9D. E). Both cycles show distinct winter and spring peaks with prevalence maxima of 60% to 65%. a moderate prevalence decrease in late winter, and a major loss of infections after the May peak. As in the reference simulations, about half the infections are systemic by early winter (Fig. 9). The spring peak, however, pro- duces relatively heavier infections, as well as higher prevalences at Shell Rock and Bennies Beds (Fig. 9C, D). The pattern at Bennies is nearly identical with that of the Miah Maull simulations, with a substantial number of category 4 infections (Fig. 9E). 510 Paraso et al. 100 A 30 - 60 - 40 ■ 20 0 l.V ASONDJ FMAMJ Month ' i.l: C BO ■ 50 ••• zc n * ^ ASONDJ FMAMJ Month JJASONDJ FMAMJ 100 uu D 80 BO ■!■: : : I n J Vi- JJASONDJFMAMJ Month Figure 9. Simulated time evolution of Haplosporidium nelsoni preva- lence and intensity produced by low freshwater conditions at A) Ar- nolds Bed, B) Cohansey Bed, C) Shell Rock Bed, Dl Bennies Bed, and E) Miah Maull Grounds. Numbers I through 4 refer to the cumulative percentage of total oysters (prevalence) with epithelial (I), subepithe- lial local (2), light systemic (3), and advanced systemic (4) infections. The area above the category 4 line represents the portion of oysters in the undetectable infection category. 100 Figure 10. Simulated, time evolution of Haplosporidium nelsoni preva- lence and intensity produced by high freshwater discharge conditions at A) Bennies Bed and B) Miah Maull (irounds. Numbers 1 through 4 refer to the cumulative percentage of total oysters (prevalence) with epithelial (1). subepithelial local (2), light systemic (3), and advanced systemic (4) infections. The area above the category 4 line represents the portion of oysters in the undetectable infection category. High Freshwater Runoff. Salinity conditions during periods of high freshwater runoff (Fig. 5B) result in no detectable H. nelsoni infections at Arnolds, Cohansey, or Shell Rock Beds. Infections appear at Bennies Bed in July, but reach a maximum prevalence of only 9% from November through January, considerably lower than the 40% to 45% prevalence found under average conditions (Figs. 8D, 10A). Infections disappear in February and do not reappear in spring. Compared to average conditions, there is little change in the Miah Maull Grounds cycle under high runoff. Changes in Timing of Freshwater Runoff Variations in the timing of maximum runoff from early (Fig. 5C) to late spring (Fig. 5D) affects H. nelsoni prevalence and intensity at Bennies Bed only. The major differences between the 2 extreme cycles (Fig. 1 1A. B) and that produced under average conditions (Fig. 8D) are in the degree of late winter infection loss and the extent of the late spring prevalence peak and subsequent decline. A February runoff produces a major loss of infections (from 45% to about 5%) in March followed by a relatively small resurgence, to 20%, in late May, and no subsequent decline (Fig. 1 1A). Understandably, a May runoff does not affect the late winter prevalence decline, which remains the same as under average con- ditions. It does, however, allow the late May prevalence peak to become higher than under average conditions (45% vs. 25%). More importantly, most of the infections become systemic (BFUs 3 and 4). resulting in a pronounced decline in prevalence in July (Fig. 1 1 B ) which was produced by a larger-than-normal number of heavily infected individuals in June that undergo attempted sporu- lation. After this decline, however, newly acquired infections quickly proliferate so that by late fall, and over the winter, there are no differences in prevalence or intensity among the early, average, and late discharge simulations. Statistical Comparisons Between Extreme and Average Model Simulations Because the first 2 years of the 3-year model runs were needed for the solution to reach a steady state, statistical analyses were done using third-year simulations only. The maximum difference. Modeling the MSX in Eastern Oyster Populations II 511 N D J Month uu ! 1 — 1 1 1 1 1 1 1 1 ! B BO - - FW Discharge 60 - - " /4^"4~^K T r 40 20 n X j-f i JJASONDJFMAMJ Month Figure 11. Simulated, time evolution of Haplosporidium nelsoni preva- lence and intensity at Bennies Bed for A) early (February) and B) late (May! high freshwater discharge events. Numbers 1 through 4 refer to the cumulative percentage of total oysters (prevalence) with epithelial ( 1 1, subepithelial local (2), light systemic (3), and advanced systemic (4) infections. The area above the category 4 line represents the portion of oysters with undetectable infections. Arrows show the timing of the freshwater discharge. average difference, root mean square (RMS), and correlation co- efficients (/•) between the reference simulation and each simulation with varied environmental conditions were calculated to quantify the effects of salinity variation of H. nelsoni on each infection category. Variations in the Magnitude of Runoff Low Freshwater Discharge. Conditions of low freshwater dis- charge have the largest effect on the undetectable infection cat- egory, as shown by high maximum difference, average difference, and RMS values (Table 4). Undetectable infections are. in effect, an inverse measure of percent infection. Thus, the decrease in this category in a high salinity regime simply means an increase in total prevalence. The largest differences associated with low discharge occurs at mid-bay sites, as suggested by the high maximum dif- ference, average difference, and RMS values. Low correlation co- efficients (<0.50) and relatively large maximum difference values were also calculated at Shell Rock and Bennies Beds for infection TABLE 4. Results of statistical analysis of the differences between low freshwater runoff and reference simulations at the seed bed and planted ground sites. Haplosporidium nelsoni infection intensity categories are UD = undetectable infections, 1 = epithelial, 2 = subepithelial local, 3 = light systemic, and 4 = systemic infections. "Maximum" and "Average" represent the maximum and average prevalence differences over an annual infection cycle. RMS represents the root mean square error, and r is the correlation coefficient. Bold is used to emphasize extreme differences. Site Intensity Maximum Average RMS category (%) (%) (%) r Arnolds Bed UD 8.03 2.42 4.02 0.44 1 3.62 1.07 1.70 0.93 2 2.03 0.44 0.78 0.87 3 4.86 0.90 1.76 0.92 4 0.10 0.01 0.03 — Cohansey Bed UD 34.93 15.26 18.62 0.84 1 22.29 7.86 9.60 0.71 2 11.20 3.04 4.47 0.89 3 11.46 2.24 3.41 0.87 4 5.95 2.10 3.11 0.91 Shell Rock Bed UD 44.68 29.32 30.12 0.72 1 27.85 13.92 15.08 0.44 2 15.99 5.39 6.22 0.60 3 19.67 4.85 6.61 0.48 4 12.15 5.29 6.63 0.71 Bennies Bed UD 40.03 19.00 21.19 0.81 I 16.82 9.21 9.83 0.92 2 30.84 4.30 7.37 0.81 3 20.89 3.52 5.50 0.73 4 37.87 5.73 9.89 0.40 Miah Maull Grounds UD 3.32 2.58 2.74 1.00 1 1.86 0.85 1.07 1.00 2 2.10 0.61 0.78 1.00 3 0.65 0.34 0.41 1.00 4 3.15 0.78 1.11 LOO categories 1. 3. and 4. reflecting increased intensity as infections move from category 1 to categories 3 and 4 under high salinity conditions. Correlation coefficients of 1.0 for every infection cat- egory at Miah Maull Grounds, as well as low values for the other 3 measures, indicate that there was almost no difference in preva- lence or intensity associated with low runoff at this site. High Freshwater Discharge. The greatest variation in simulated prevalences at all sites during high freshwater runoff occurs in the undetectable infection category (Table 5). This is consistent with the results of low freshwater discharge conditions and reflects the loss of parasites under low salinity conditions and a decrease in total prevalence. The greatest variability in maximum and total difference, and in RMS, is at Bennies Bed. Overall, correlation coefficients are positive and high O0.50 in all but one comparison, at Bennies Bed), and are 0.99 in all infection categories at Miah Maull Grounds. Values for the other three measures are lowest at the Miah Maull site, indicating little change in infection intensity and prevalence associated with high freshwater discharge at this location. Variation in Timing of Freshwater Discharge The simulated prevalence and intensity obtained for early and late freshwater runoff conditions at Arnolds. Cohansey. and Shell 512 Paraso et al. TABLE 5. Results of statistical analysis of the differences between high freshwater runoff and reference simulations at the seed bed and planted ground sites. Haplosporidium iielsoni infection intensity categories are UD = undetectable infections, 1 = epithelial, 2 = subepithelial local, 3 = light systemic, and 4 = systemic infections. "Maximum" and "Average" represent the maximum and average prevalence differences over an annual infection cycle. RMS represents the root mean square error, and r is the correlation coefficient. Bold is used to emphasize extreme differences TABLE 6. Results of statistical analysis of the differences between early spring freshwater runoff, late spring freshwater runoff and reference simulations at Bennies Bed. Haplosporidium nelsoni infection intensity categories are UD = undetectable infections, I = epithelial, 2 = subepithelial local, 3 = light systemic, and 4 = systemic infections. "Maximum" and "Average" represent the maximum and average prevalence differences over an annual infection cycle. RMS represents the root mean square error, and r is the correlation coefficient. Bold is used to emphasize extreme differences. Intensity Maximum Average RMS Station category (%) (%) (%) Simulation Intensity Maximum Average RMS category ( c, I (%) (%) Cohansey Bed Shell Rock Bed Bennies Bed Miah Maull Grounds US 1 2 3 4 UD 1 2 3 4 UD 1 2 3 4 UD I 2 3 4 8.82 4.31 2.58 5.15 0.22 19.72 12.67 6.43 11.77 0.70 44.43 26.06 1 3.50 12.93 10.04 5.36 2.91 3.25 1.27 4.89 2.73 1.27 0.53 0.92 0.02 6.57 3.45 1.36 1.65 0.11 28.04 13.43 5.60 4.56 4.46 4.14 1.47 0.93 0.55 1.20 4.49 2.01 0.93 1.81 0.06 10.18 5.24 2.47 3.42 0.23 29.35 14.28 6.70 5.55 5.92 4.40 1.75 1.19 0.64 1.71 0.86 0.74 0.84 0.74 0.88 0.90 0.96 0.81 0.40 0.64 0.55 0.74 1.00 0.99 1.00 1.00 1.00 Early freshwater runoff Late freshwater runoff UD 1 2 3 4 UD I 2 3 4 14.46 6.01 7.21 0.94 11.26 3.07 4.63 0.59 6.61 1.28 1.79 0.84 9.30 1.58 2.78 0.78 2.45 1.01 1.37 0.96 18.40 5.01 7.60 0.90 10.45 3.53 4.66 0.92 9.73 1.66 2.64 0.89 27.94 4.01 8.20 0.79 4.74 1.37 2.04 0.93 tions for Miah Maull in the planted grounds were compared with the annual infection cycle for that region, as described by Ford and Haskin ( 1982). On the other hand, data from the low salinity seed beds consist primarily of samples taken at winter and spring in- fection peaks (Haskin and Ford 1982. Fegley et al. 1994). Conse- quently, model output for this region of the bay was tested by comparing simulated prevalences and intensities at the peaks with observed values under similar ambient conditions. Rock Beds and at Miah Maull Grounds do not differ significantly from those obtained with the reference simulations (RMS < 0.37% and r > 0.98). The only significant modifications of the annual infection cycle associated with early or late discharge occur at Bennies Bed. Under early (February) runoff conditions, the great- est deviation from average was produced in undetectable and cat- egory 1 infections, as indicated by the maximum difference value at these categories (Table 6). A comparatively low correlation (0.59) in category 1 reflects the light infections present during the late spring and early summer, which never progress to higher levels (Fig. 1 1 A) as they do under average conditions (Fig. 8D). The undetectable and category 1 infections were also affected by late runoff, as indicated by the maximum difference values. The greatest maximum difference value was in category 3 infections (Table 6). reflecting the high percentage of light systemic infec- tions in the late spring peak (Fig. 1 IB) compared to the average cycle (Fig. 8D). DISCUSSION Comparison of Model Simulations with Field Observations A test of the accuracy of the oyster population-//, nelsoni model is to compare simulated with observed infection prevalence and intensity patterns and levels under different environmental conditions. This was done in two ways. For the high salinity region of lower Delaware Bay. where considerable data are available on H. nelsoni infections throughout the year (Fig. 2), model simula- The Annual H. nelsoni Infection Cycle in a High-Salinity Location The onset of the simulated infection cycle at the Miah Maull Grounds begins in June. Prevalences and intensities increase, reaching a maximum prevalence of 65% in December (Fig. 8E). This value is consistent with prevalences reported by Ford and Haskin ( 1982) for oysters in the planted ground area, which aver- aged 40% to 60% in most years, with minimum and maximum prevalences of 15% and 100%, respectively. As discussed by Ford et al. (this volume) the simulated proliferation of parasites during warm weather is primarily determined by temperature, but the reduction in proliferation rates apparent by late fall is caused by a combination of falling temperatures and high parasite density in oyster tissue. The latter inhibits proliferation because parasites are competing for a limited and declining nutrient resource within the host. In late winter, the simulated infection cycle shows a marked loss of infected oysters, as does the observed cycle (Figs. 2. 8E). Observed mortality of severely infected oysters occurs at this time, reducing both prevalence and intensity. In addition, intensity less- ens because parasites do not tolerate overwinter conditions well. Perhaps this is due to a direct low temperature intolerance or other unfavorable influences facing the parasite under the prolonged anaerobic conditions that occur inside the oyster, over the winter (Ford and Haskin 1982). For instance. H. nelsoni plasmodia con- tain numerous mitochondria (Muller 1967, Perkins 1968), which suggests that the parasite is well equipped for aerobic metabolism and may not be able to tolerate low oxygen well. The model Modeling the MSX in Eastern Oyster Populations II 513 simulates the degeneration and disappearance of parasites at this time by including a susceptibility factor, which is applied to para- sites and which increases as a function of degree days below 5 °C (Ford et al. this volume). As susceptibility increases, the parasites become more vulnerable to attack by host hemoeytes. Hemocytes themselves are becoming more active with increasing spring tem- peratures and. as a consequence, can readily eliminate the dam- aged parasites from the host. Both simulated and observed annual cycles contain a distinct second peak, reached in late May or early June. This peak contains many very heavily infected oysters due to increasing temperatures that cause a swift resurgence of those infections that were sup- pressed over the winter (Andrews 1966, Ford 1985b. Ford and Haskin 1982). The model, however, was unable to reproduce the very rapid infection development by temperature alone. A second factor was added, which eased the density-dependent limits on parasite growth as a consequence of improving conditions within the oyster (Ford et al. this volume). The increasing quality of the parasite's environment, hypothesized to occur at this time, is due to the spring bloom, increased feeding rates of the host, and the cleansing of waste products that have accumulated in the host over the winter. Finally, both simulated and observed cycles show a marked decline in prevalence following the spring peak (Figs. 2, 8E). Heavily infected oysters die at this time, accounting for some of the decline. Additionally, parasites are lost from living oysters for reasons that are not clear, but may have to do with attempted, but unsuccessful, sporulation and consequently a failure on the part ot the parasite to complete its life cycle (Ford and Haskin 1982. Ford et al. this volume). Peak Infection Levels at Low Salinity Seed Bed Locations To compare simulations for the seed bed sites (Fig. 1 ) with field measurements, the latter were grouped, by site, according to years of high, average, and low freshwater runoff. The grouping criteria were the same as those used to produce salinity time series for these conditions. That is, the average years were those in the middle 60% of the range; the high and low runoff years were the 20% at either extreme. For each flow regime and site, the mean, median, and range of observed prevalences at both late fall and spring peaks were obtained (Table 7). Because of the limited num- ber of observations (Fig. 6) Shell Rock Bed was not included in this comparison. Comparison of initial simulations with observations showed that the timing of infections was correct, but that prevalences at the seed bed sites were too high. The relationship used to model the host-parasite relationship under varying salinity was derived from a study in which parasites were subjected to acute in vitro salinity change (Ford and Haskin 1988). modified to simulate more gradual in vivo changes in salt content of host fluids (equation 3). The inability of the model to reproduce observed prevalences us- ing this relationship alone suggested that prevalence was influ- enced by an additional factor or factors. Although temperatures from only a single site. Section E (Fig. 4A). were used in reference simulations for all seed bed sites, monthly-averaged temperatures varied by a maximum of only 1.4 °C from the planted grounds to the seed bed areas (Fig. 4A). Further, temperature variations within the seed bed area are rela- tively small (Fegley et al. 1994). Thus, temperature differences are unlikely to be responsible for the disparity. Another possible factor could be salinity-associated differences in the rate of infection. The infective stage of H. nelsoni remains unknown. Spores, typically an important element in parasite trans- mission, are rare in adult oysters and their absence has led to the hypothesis that another host may be involved in the life cycle (Burreson 1988. Haskin and Andrews 1988). On the other hand, spores appear to be produced regularly in juvenile oysters with advanced infections (Barber and Ford 1992, Burreson 1994), so TABLE 7. Comparison of simulated total prevalence with observed mean, median, and range of prevalences on Delaware Bay sites (Fig. 1) under High. Average, and Low Delaware River runoff conditions. PG indicates various planted grounds (Fig. 7). High runoff Average runoff Low runoff %MSX Fall %MSX Spring %MSX Fall %MSX Spring %MSX Fall %MSX Spring Arnolds Model 0 0 0 0 8 0 observed mean 1 <1 13 2 11 4 observed median 0 0 0 0 8 0 observed range 0-5 0-5 0-60 0-10 0-30 0-15 Cohansey Model 0 0 9 0 37 3 observed mean 5 <1 16 5 19 8 observed median 5 0 10 0 20 5 observed range 0-10 0-5 0-52 0-25 0-40 0-20 Shell Rock Model 0 0 20 0 52 35 Bennies Model 9 0 47 25 67 60 observed mean 10 9 30 21 44 28 observed median 5 8 22 15 51 15 observed range 0-30 0-30 0-65 0-85 0-65 0-75 Miah Maull Model 60 55 65 55 67 62 PG observed mean 43-65 40-53 40-58 40-54 43-60 42-55 PG observed median 55-65 43-53 35-65 35-50 33-63 0 PG observed range 5-95 0-80 10-95 5-90 0-85 0-100 514 Paraso et al. direct transmission between oysters remains a possibility. Further. Barber and Ford (1992) found that haplosporidian spores, mor- phologically similar to those of H. nelsoni, are present in the water column in Delaware Bay in approximately equal abundance in upper and lower Bay sites. These authors speculated that if these spores are. in fact. H. nelsoni and are the infective element, the observed disparity between their concentration and infection prevalence in the upper and lower Bay could be explained by reduced infectivity under low salinity conditions. Alternatively, if another host is involved, the disparity could be explained by a salinity effect on another host or on its relationship with H. nel- soni. The discrepancy was resolved in the model by adding an in- fection efficiency factory (equation 6), which reduces infection rates with decreasing salinity. This relationship implies nothing about the source or form of the infective particle; however, it does suggest that some element in the life cycle other than the plasmo- dial stage in the oyster, is sensitive to low salinity. This includes, but is not limited to. an intermediate or reservoir (alternate) host, the abundance of infective stages in the water, or the ability of infective stages to invade and establish themselves in oysters. Addition of infection efficiency to the model resulted in simu- lated prevalences that largely matched those observed (Table 7). The simulated levels were most like the median prevalences, which they reproduced almost exactly in many cases. Notably, the model produced no detectable infections on Arnolds Bed except under low flow conditions, as observed, and reproduced the elimi- nation of the spring peak in cases where it was observed in the field. The model also accurately simulated the relative insensitivity to river flow changes of the lower bay infection cycle that is observed in nature. Only on Bennies Bed for average conditions and at Bennies and Cohansey Beds for low flow conditions, did the simulations not closely reproduce the mean or median observed prevalences (Table 7 ). The simulated prevalences were higher, especially under the low flow conditions, although they were clearly within the range of observations. Similarly, the modeled prevalence for Miah Maull Grounds tended to be at the high end of the observed mean or median range for the lower Bay planted grounds. It is clear that observed prevalence in the moderate to high salinity region of the bay varies widely and in some years no infected oysters are found, regardless of salinity (Figs. 6, 7; Table 7). Multiyear cycles in H. nelsoni prevalence, totally independent of river flow and salinity, were described by Ford and Haskin (1982) for lower Delaware Bay. The salinity in lower Delaware Bay, in fact, is always high enough to support full H. nelsoni activity so that year-to-year variations must be due to some other condition. The fact that disparities between observed and simulated prevalences were maximum at mid-Bay sites under low flow, high salinity condi- tions, which produce infection levels and patterns similar to the lower Bay. suggests that they are influenced by factors other than the local salinity. The possible factors are included in a more complex disease transmission model as described in Powell et al. (this volume). Another element of the basic model that was modified to simu- late observed conditions involved infection intensity. Long-term observed averages show that despite decreasing prevalence, the ratio of systemic to localized infections remains essentially 1:1 along the salinity gradient in the seed bed regions of Delaware Bay (Ford and Haskin 1982, Fegley et al. 1994). However, the simu- lated intensity decreased upbay. To improve the model, the trans- fer rate of parasites from the epithelium to the systemic tissue was increased with decreasing salinity. The addition of this formulation to the model caused about half of the infections to remain systemic at peak prevalence on seed bed sites, while still allowing more intense infections to develop at the spring peak under the normally high salinity conditions of the lower bay and. at low river flow, on the lower seed beds. When a strong spring peak occurs, which it does only under high salinity situations, oysters often have the most advanced infections of the year, so that simulations reproduce observed intensities under high salinity conditions also. The biological basis for the maintenance of high systemic in- fection levels under conditions of decreasing prevalence along the salinity gradient is unclear. Further, the mechanism used by para- sites to penetrate the basal lamina between the epithelium and the underlying tissue is not known (Scro and Ford 1990). Thus, there is no direct evidence that low salinity enhances the transfer rate between the two tissues. On the other hand, model simulations were run for a population in which all oysters were the same size ( 1 a dry weight). However, adult size generally decreases along the salinity gradient in Delaware Bay, and along with it occurs an increase in the surface (epithelium) to volume (systemic) ratio. A given concentration of parasites per unit area of epithelium in a small oyster is actually a larger total parasite load in the surface layer than it is in a larger oyster. In some way. this large overall burden may speed the movement of parasites into the underlying tissue and thus help explain the observed relatively high intensity levels on the seed beds. The addition of multiple oyster size classes to the model is needed to test this hypothesis. Simulations were also run with variations in food availability in the spring and with warm winter conditions (>5 °C) to determine the relative effects of salinity compared to food and temperature on oyster populations infected with H. nelsoni. Variations in salinity had a larger effect on the simulated infection levels than did varia- tions in food availability and winter temperatures in a population of 1-g oysters. However, the simulated time development of in- fections did demonstrate the importance of cold winters in pro- ducing a reduction in infections the following spring, and therefore modulating the annual cycle. When absolute differences between average and extreme con- ditions were compared statistically, the mid-Bay locations of Shell Rock and Bennies Beds showed the greatest variations. The ex- planation lies in the fact that mean mid-tide salinities at these locations range from 9.5-1 1.2 ppt to 19.3-21.1 ppt (Table 2). and it is in this range that a shift in salinity causes the greatest change in parasite survival (Fig. 3). In addition. Bennies Bed (salinity range 13.3 to 18.7 under average conditions) was the only site at which simulated variation in the timing of "spring" runoff caused a significant change from the reference annual cycle. This is be- cause Bennies Bed, under average conditions, was the only seed bed to exhibit a late spring prevalence peak (Fig. 8D) and thus the only bed that had any infections to be altered by differences in timing. Salinity was always low enough to permit a spring MSX prevalence peak at the upper bay sites irregardless of the timing of the spring runoff and salinity in the planted grounds was high enough (>20 ppt) nearly all the time to support infections so that changes in timing, like changes in volume of discharge, would have little effect on the spring peak at Miah Maull Grounds. Summary and Conclusions The simulations obtained with the oyster population-//, nelsoni model were able to reproduce observed infection patterns and lev- Modeling the MSX in Eastern Oyster Populations II 515 els along the salinity gradient in Delaware Bay. The base relation- ship used by the model was developed from in vitro parasite sur- vival data. The inability of initial simulations, using this direct salinity-parasite function, to produce infection levels that ad- equately represented field observations, indicate a more complex host-parasite relationship at low salinity. Two additional functions, both based on assumptions, were added to the model to improve simulations. One allowed infection intensity to remain stable along the decreasing salinity gradient by increasing the transfer rate of parasites from epithelial to systemic tissues. The biological ratio- nale may be linked to decreasing size of oysters and increasing surface-to-volume ratios along the salinity gradient. It is also pos- sible that the transfer rate change may simply be a substitute for some other mechanism that allows intense infections to occur at low prevalence. The second modification was to diminish infection efficiency w ith decreasing salinity. Because the mode of transmis- sion of H. nelsoni and the form of its infective stage remain un- known, this assumption cannot be verified. Reduced infection ef- ficiency may also be a surrogate for another mechanism, such as abundance of a potential intermediate or alternate host, the abun- dance of infective stages, or both. The importance of these modifications is not necessarily the assumptions that were made, but that they show that factors other than a simple, direct salinity effect on H. nelsoni proliferation and death rates in the oyster are required to explain field observations. The model results emphasize that in the H. nelsoni relationship, salinity is a permissive/restrictive element, modifying the degree and pattern of infections, during transmission as well as after oys- ters become infected. Another, overriding factor, is the presence and abundance of infective particles, which varies from year to year. Model simula- tions show that in low salinity environments, infection levels are clearly correlated with river flow and that salinity changes over the 10-20 ppt range have the greatest effect on infection patterns and level. This association, however, breaks down at high salinity. Infection levels in high salinity (>20 ppt) regions of Delaware Bay are uncorrected with river flow, as was noted earlier (Haskin and Ford 1982), and is attributed to fluctuations in infective particle abundance. The model shows that under drought conditions, the same factor or factors override local salinity to control infections in mid-Bay locations. The ability to investigate H. He/.wm'-oyster-salinity relation- ships with relative ease using the coupled model demonstrates its usefulness as a practical tool. In many regions other than Delaware Bay, low salinity areas are critical as setting and storage areas for seed oysters. The capacity of the model to simulate, with only minor modifications, observed H. nelsoni levels in mid- Chesapeake Bay demonstrates that it is robust and can be em- ployed in other areas (Powell et al. this volume). ACKNOWLEDGMENTS We thank Walter Conzonier and Bob Barber of the Haskin Shellfish Research Laboratory of Rutgers University for collecting and providing the environmental data sets necessary for this model study. This research was supported by the Virginia Marine Science Consortium Sea Grant program, under grant number VGMSC-5- 29222. Additional computer resources and facilities were supplied by the Center for Coastal Physical Oceanography. Old Dominion University. This is Contribution number 99-17 of the Institute of Marine Science at Rutgers University and NJAES Publication #D- 32405-2-99. LITERATURE CITED Andrews. J. D. 1964. Oyster mortality studies in Virginia. IV. MSX in James River public seed beds. Proc. Natl. Shellfish. Assoc. 53:65-84. Andrews. J. D. 1966. Oyster mortality studies in Virginia. V. Epizootiol- ogy of MSX. a protistan pathogen of oysters. Ecology. 47:19-31. Andrews. J. D. 1983. Minchinia nelsoni (MSX) infections in the James River seed-oyster area and their expulsion in the spring. Est. Coast. Shelf Sci. 16:225-269. Barber. B. J., S. A. Kanaley & S. E. Ford. 1991. Evidence for regular sporulation by Haplosporidium nelsoni (MSX) (Acetospora; Haplospo- ridiidae) in spat of the American Oyster. Crassostrea virginica. J. Protozool. 38:305-306. Barber. R. D. & S. E. Ford. 1992. Occurrence and significance of ingested haplosporidian spores in the eastern oyster, Crassostrea virginica (Gmelin. 1791). J. Shellfish Res. 11:371-375. Burreson, E. M. 1988. Use of immunoassays in haplosporidan life cycle studies. In: W. S. Fisher (ed.). Disease Processes in Marine Bivalve Molluscs. Amer. Fish. Soc. Spec. Puhl. 18:298-303. Burreson. E. M. 1994. Further evidence of regular sporulation by Hap- losporidium nelsoni in small oysters. Crassostrea virginica. J. Parasi- tol. 80:1036-1038. Farley. C. A. 1975. Epizootic and enzootic aspects of Minchinia nelsoni (Haplosporida) disease in Maryland oysters. J. Invertehr. Protozool. 22:418^127. Fegley, S. R.. S. E. Ford. J. N. Kraeuter & D. R. Jones. 1994. Relative effects of harvest pressure and disease mortality on the population dynamics of the Eastern oyster (Crassostrea virginica) in Delaware Bay. Final Report to NOAA. Grant No. NA26FL0388. 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. Ford, S. E. 1985a. Effects of salinity on survival of the MSX parasite Haplosporidium nelsoni (Haskin, Stauber. and Mackin) in oysters. J. Shellfish Res. 5:85-90. Ford, S. E. 1985b. Chronic infections of Haplosporidium nelsoni (MSX) in the oyster. Crassostrea virginica. J. Invertehr. Pathol. 45:94-107. Ford, S. E. 1997. History and present status of molluscan shellfisheries from Barnegat Bay to Delaware Bay. NOAA Tech. Rep. NMFS-127. Ford. S. E. & H. H. Haskin. 1982. History and epizootiology of Haplospo- ridium nelsoni (MSX). an oyster pathogen in Delaware Bay. 1957- 1980.7. Invertehr. Pathol. 40:118-141. Ford. S. E. & H. H. Haskin. 1988. Comparison on in vitro salinity tolerance of the oyster parasite, Haplosporidium nelsoni (MSX) and hemocytes from the host, Crassostrea virginica. Comp. Biochem. Physiol. 90A: 183-187. Ford, S. E., E. N. Powell. J. M. Klinck & E. E. Hofmann. 1999. Modeling the MSX in Eastern Oyster (Crassostrea virginica) populations. I. Model development, implementation and verification. /. Shellfish Res. 18:473^198 Haskin. H. H. 1972. Delaware River flow-bay salinity relationships. Report to the Delaware River Basin Commission: Phase III. 12 pp. Haskin. H. H. & J. D. Andrews. 1988. Uncertainties and speculations about the life cycle of the Eastern oyster pathogen Haplosporidium nelsoni (MSX). Amer. Fish. Soc. Spec. Puhl. 18:5-22. Haskin, H. H. & E. S. Ford. 1979. Development of resistance to Minchinia 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. 1982. Haplosporidium nelsoni (MSX) on 516 Paraso et al. Delaware Bay seed oyster beds: A host-parasite relationship along a salinity gradient. /. Invertebr. Pathol. 40:388-405. Haskin. H. H., L. A. Stauber & J. A. Mackin. 1966. Minchinia nelsoni n. sp. (Haplosporida. haplosporidiidae): Causative agent of the Delaware Bay oyster epizootic. Science 153:1414-1416. Hofmann. E. E., J. M. Klinck. E. N. Powell, S. Boyles & M. Ellis. 1994. Modeling oyster populations II. Adult size and reproductive effort. J. Shellfish Res. 13:159-176. Hofmann. E. E.. E. N. Powell, J. M. Klinck & E. A. Wilson. 1992. Mod- eling oyster populations III. Critical feeding periods, growth and re- production. / Shellfish Res. 1 1:399-416. Muller, E. 1967. Electron microscope studies of Minchinia nelsoni. Proc. Nat. Shellfish Assoc. 57:3. Pennock, J. R. 1985. Chlorophyll distributions in the Delaware Estuary: regulation by light-limitation. Est. Coast. Shelf Sci. 21:711-725. Pennock and Sharp. 1986. Phytoplankton production in the Delaware es- tuary: Temporal and spatial variability. Mar. Ecol. Prog. Ser. 34:143- 155. Perkins, F. O. 1968. Fine structure of the oyster pathogen. Minchinia nel- soni (Haplosporidia, Haplosporidiidae). J. Invertebr. Pathol. 10:287- 307. Powell, E. N., E. E. Hofmann. J. M. Klinck & S. M. Ray. 1992. Modeling oyster populations I. A commentary on filtration rate. Is faster always better?/ Shellfish Res. 11:387-398. Powell. E. N.. E. E. Hofmann, J. M. Klinck, E. A. Wilson-Ormond & M. S. Ellis. 1995. Modeling oyster populations V. Declining phytoplankton stocks and the population dynamics of American oyster {Crassostrea virginica) populations. Fish. Res. 24:199-222. Powell, E. N.. J. M. Klinck. S. E. Ford. E. E. Hofmann & S. J. Jordan. 1999. Modeling the MSX in Eastern Oyster (Crassostrea virginica) populations. III. Regional application and the problem of transmission. J. Shellfish Res. 18:515-536. Powell. E. N„ J. M. Klinck & E. E. Hofmann. 1996. Modeling diseased oyster populations. II. Triggering mechanisms for Perkinsus marimts epizootics. J. Shellfish Res. 15:141-165. Powell. E. N.. J. M. Klinck. E. E. Hofmann & S. Ford. 1997. Varying the timing of oyster transplant: Implications for management from simu- lation studies. Fish. Oceanogr. 6:213-237. Powell. E. N., J. M. Klinck, E. E. Hofmann & S. M. Ray. 1994. Modeling oyster populations. IV: Rates of mortality, population crashes, and management. Fish. Bull. 92:347-373. Scro. R. A. & S. E. Ford. 1990. An electron microscope study of disease progression in the oyster Crassostrea virginica infected with the pro- tozoan parasite Haplosporidium nelsoni (MSX). In: F. O. Perkins and T. C. Cheng (eds.). Pathology in Marine Science. Academic Press, Orlando. FL. pp. 229-254 Sharp, J. H.. L. A. Cifuentes, R. B. Coffin & J. R. Pennock. 1986. The influence of river variability on the circulation, chemistry, and micro- biology of the Delaware estuary. Estuaries 9:261-269. Soniat. T. M.. E. N. Powell, E. E. Hofmann & J. M. Klinck. 1998. Under- standing the success and failure of oyster populations: the importance of sampled variables and sample timing. /. Shellfish. Res.. 17:1149- 1165. Sprague. V. 1978. Comments on trends in research on parasitic diseases of shellfish and fish. Mar. Fish. Rev. 40:26-30. Verwer, J. G. & M. van Loon. 1994. An evaluation of explicit pseudo- steady-state approximation schemes for stiff ODE systems from chemi- cal kinetics. J. Comput. Phys. 113:347-352. Wong, K-C. 1994. On the nature of transverse variability in a coastal plain estuary. /. Geophys. Res. 99:14.209-14,222. Wong, K-C. 1995. On the relationship between long-term salinity varia- tions and river discharge in the middle reach of the Delaware estuary. J. Geophys. Res. 100:20,705-20.713. Journal o] Shellfish Research. Vol. IX. No. 2. 517-537. 1999. MODELING THE MSX PARASITE IN EASTERN OYSTER (CRASSOSTREA VIRGINICA) POPULATIONS. III. REGIONAL APPLICATION AND THE PROBLEM OF TRANSMISSION ERIC N. POWELL,1 JOHN M. KLINCK,2 SUSAN E. FORD,' EILEEN E. HOFMANN,2 AND STEPHEN J. JORDAN3 Haskin Shellfish Research Laboratory Rutgers University' Port Norris, New Jersey 08349-9736 'Center for Coastal Physical Oceanography Old Dominion University Norfolk, Vrginia 23529 Sarbanes Cooperative Oxford Laboratory Maryland Department of Natural Resources Oxford, Maryland 21654 ABSTRACT A model of transmission for Haplosporidium nelsoni, the disease agent for MSX disease, is developed and applied to sites in Delaware Bay and Chesapeake Bay. The environmental factors that force the oyster population-W. nelsoni model are salinity, temperature, food, and total suspended solids. The simulated development of MSX disease was verified using 3 time series of disease prevalence and intensity: 1960 to 1970 and 1980 to 1990 for Delaware Bay, and 1980 to 1994 for Chesapeake Bay, and for a series of sites covering the salinity gradient in each bay. Additional simulations consider the implications of assumptions made in devel- opment of the model for constraining the mode of transmission of H. nelsoni disease in oyster populations. Transmission of H. nelsoni includes non-local factors that exert a paramount influence on the transmission process. Key environ- mental forcing factors of season, salinity, and winter temperature exert a direct control on the transmission process, either by controlling the availability of infective particles in the water column or by controlling the population dynamics of an alternate host. Salinity's role is a dual one. Salinity acts on the local host population by varying the infectivity of infective particles as they impinge the oyster gill during the filtration process. In addition, salinity exerts a regional influence on the transmission process by controlling, in part and on a bay-wide scale, the concentration of infective particles in the water column (or perhaps the abundance of an alternate host). In addition to the effect of salinity, infective particle concentration also decreases for 1 to 2 years after a cold winter and returns to high levels faster after a warm winter. It is the presence in the H. nelsoni transmission model of these bay-wide influences of environmental change that make this model different from most other transmission models. Simulations suggest that epizootic cycles are principally the product of enhanced transmission rather than enhanced intensification. These influences of transmission on the course of infection, in many cases, have multiyear implications for prevalence and infection intensity, and the root of much of this mulliyear behavior is in the processes that control the concentration of infective particles in the water column. KEY WORDS: oyster, disease, parasite, population modeling, transmission INTRODUCTION observational studies (Mackin 1952, Andrews 1965. Andrews 1988, Andrews and Ray 1988, Ford 1992). MSX disease in oys- ters, produced by the haplosporidian protozoan Haplosporidium Perhaps the most significant task facing any parasite or disease nelsoni, is a good example of a disease that shows significant model is to adequately simulate the geographic distribution of temporal and geographic variation in prevalence and infection in- prevalence and infection intensity. Models attempting to do so tensity that cannot be modeled assuming a local process-driven always rely heavily on the dynamic modeling of transmission. mode of transmission. Regional-scale processes are inferred to be (Dietz 1982, Ackerman et al. 1984, Kermack and McKendrick important from observational studies (Andrews 1964. Haskin and 1991, Dwyer and Elkinton 1993), because the geographic variation Andrews 1988) and will be shown herein to be important in mod- in infection intensity cannot be adequately simulated if the geo- eling the disease. graphic and temporal dynamics of prevalence are inadequately The full life cycle of H. nelsoni is unknown (Ford and Tripp defined. Transmission may have local and regional dynamics. 1996). Transmission between oysters has never been confirmed Models of disease principally influenced by local processes usu- experimentally. Disease epizootics occur spontaneously over wide ally rely on estimates of contact rate, host density, and the fre- regional areas, more or less simultaneously across broad expanses quency of infected individuals (Godfrey and Briggs 1995. Heth- of the salinity gradient, and independent of the proximity of in- cote and van den Driessche 1995, Frank 1996). The spread of fected oysters (Andrews 1968. Ford and Haskin 1982. Haskin and disease beyond local populations, however, often involves the in- Ford 1982). These observations suggest bay-wide response pat- tegration of variations in landscape and environmental condition terns in transmission that are considerably decoupled from the that significantly influence the infection process (Gill 1928. Mol- local processes occurring on any one oyster bed. One well- lison 1987, Myers 1993, Ostfeld 1997). Dermo disease in oysters, considered hypothesis is that an intermediate host is involved in produced by the protistan Perkinsus marinus, is an example of a the life cycle, however no direct confirmation of this hypothesis disease modeled adequately using local processes of population exists (Haskin and Andrews 1988. Ford and Tripp 1996). The prevalence, infection intensity, and host density, a surrogate for purpose of this paper is first to develop a model of transmission in contact rate in an immobile species (Hofmann et al. 1995. Powell MSX disease. We will then apply this model to 3 case histories et al. 1996). and this seems to agree with the neighborhood model using long-term time series, 2 from Delaware Bay ( 1960 to 1970, of disease transmission that has evolved from experimental and 1980 to 1990) and 1 from Chesapeake Bay (1989 to 1994). that 517 518 Powell et al. come from a series of sites covering much of the salinity gradient in each bay. Finally, we will consider the implications of the assumptions made in formulating the transmission model for con- straining the mode of transmission of H. nelsoni in oyster popu- lations. THE CRASSOSTREA VIRGINICAHAPLOSPORIDIUM NELSONI MODEL The Crassostrea virginica-Haplosporidiwn nelsoni model can be separated into a transmission component and an intensification component. The Intensification Component The intensification component is described in detail by Ford et al. (this volume). Only a brief review is given here. Disease Quantification The majority of the observations on MSX disease use a semi- quantitative scale that categorizes H. nelsoni infection level ac- cording to parasite distribution (local or diffuse I and visual rank- ings of cell density. The infection rating system used for the model is based on a semi-quantitative scale developed for use in Dela- ware Bay (Ford and Haskin 1982). In contrast, the oyster-//, nel- soni model is based on the number of parasites per oyster. It was therefore necessary to establish, at the outset, a relationship be- tween H. nelsoni cells per oyster and the 0-to-6-point semi- quantitative scale (referred to as Little Ford Units [LFU]). This conversion, developed by Ford et al. (this volume), is based on a logarithmic relationship of the form: LFU = a. < In C. (1) Kbce/sWQfrm:e/sj where ot.A is a constant that differs for the epithelial (e) and sys- temic (s) tissue (the notation e/s will be used to denote constants that have different values for epithelial and systemic tissue), C^ is the number of H. nelsoni cells in the epithelial or systemic tissue. b is a scaling factor, and ce/s is a constant. W„ is the ash-free dry weight of the oyster and fraceA is the fraction of epithelial or systemic tissue in the animal. The values of the coefficients in equation (1) and all other equations are given in Ford et al. (this volume). Rare or very light epithelial infections (LFU = 1) may be identified by as few as 1 or 2 H. nelsoni cells in the gill epithelium in a standard tissue cross-section analysis. Thus, some oysters diagnosed as having no infections (LFU = 0) are probably in- fected (Stokes et al. 1995). The model is constructed to reflect this circumstance. The distinction between an uninfected oyster (LFU = 0) and one in the very lightest infection category (LFU = 1 ) is based on a detection limit of 1.3 x 104 and 6.5 x 103 parasites per gram wet weight for epithelial and systemic tissue, respectively and not on the absolute absence of infection. For simplicity, a 0-to-4-point scale was chosen to depict the final output from the model. These units are referred to as Big Ford Units (BFU). Conversion between the scales simply involves com- bining the 4 highest LFUs into 2 BFUs (Table I and Fig. 4 in Ford et al. this volume). Most of the information and parameterizations used in the model originated from data provided in terms of LFU. Thus, in effect, the model converts from LFU to parasite number for calculation and from cell number back to LFU (and then to BFU) for data presentation. As a result, the values for most of the constants used in the model equations depend upon the conversion given by equation ( 1 ). but. unlike the case for Perkinsus marinus where a relatively accurate conversion exists between cell density and Mackin units derived from whole body counts (Choi et al. 1989. Bushek et al. 1994), the relationship between LFUs and //. nelsoni number is more indirectly estimated and therefore subject to greater error. Model Equations The model is structured as a two-dimensional array (Fig. 3 in Ford et al. this volume) with 55 epithelial and 55 systemic infec- tion categories. The infection level in each category is defined by the average number of H. nelsoni in it, with the maximum differ- ence between adjacent classes being 1 population doubling. The difference between infection classes at the higher parasite densities is less than 1 population doubling because of the nonlinear distri- bution of LFUs with respect to parasite number. The nonlinear arrangement was required to provide multiple infection classes within each LFU infection category and, consequently, necessi- tated scaling the transfers between infection categories by the ratio of the parasite cell number (C) between adjacent classes as: • for transfers up in epithelial tissue: CJ(Cc+{ - Ce) • for transfers down in epithelial tissue: CJ(Ct. - Ct._,) • for transfers up in systemic tissue: CJ(CS+I - Cs) • for transfers down in systemic tissue: CJ{CS - Cv_,) The governing equation for determining the prevalence and intensity of//, nelsoni infections in the epithelial (e) and systemic (s) tissue of post-settlement infected oyster (0) populations is given by: = ~ «. A ., - Kfics + 6.5. without devel- oping systemic infections. Oysters in systemic classes > 7.0 are automatically placed in the dead oyster category because parasite densities represented by these classes are higher than those found in live oysters. Additional mortality processes are discussed in Ford et al. (this volume). The Transmission Model The transmission model deals with the initiation of true new infections. MSX. like Dermo. is assayed using semi-quantitative methods that have a limit of resolution defined as the H. nelsoni cell density below which false negatives commonly occur. Epide- miological data cannot distinguish false negatives from truly un- infected animals and. as such, particularly during certain times of the year, many infected oysters will appear to be uninfected (e.g., Bushek et al. 1994, Hofmann et al. 1995). Accordingly, a trans- mission model cannot be fit to field data alone; coupling to the intensification model is required to generate the false negatives that can be prevalent in field observational data. The development of this component of the Crassostrea vir- ginica-Haplosporidium nelsoni model faced a particularly strin- gent hurdle due to the limited information beyond epizootiological observations of the infection process. Accordingly, the strategy taken was to develop relationships that produced the broad re- gional and temporal trends in prevalence observed in observational data. An attempt was made to ground these relationships in ob- served or anticipated physiological and ecological processes. Nev- ertheless, the few available data do not permit verification of all of these processes and so the approach taken in the model may well be incorrect even though the answer would appear to be correct, and the relationships are admittedly crude in many cases. Never- theless, our approach has been to utilize a large number of sites covering a wide region with a significant salinity gradient and a significant temporal time series based on the assumption that any relationship fitting a multitude of sites is likely, to some significant extent, to express a real process controlling H. nelsoni transmis- sion. Finally, for convenience, we will refer to the H. nelsoni life stage responsible for infection as the infective particle. The H. nelsoni life cycle includes a spore phase (Couch et al. 1966, Ford and Tripp 1996) that is free in the water column, judging from the observation of spores in oyster guts (Barber and Ford 1992). How- ever, the evidence supporting spores as the primary infective stage for oysters is meager. An, as yet unknown, stage may be respon- sible for infection. Governing Equations The transmission model covers a series of basic processes. ( 1 ) The model is initialized with a base infective particle concentration (infective particles L~l). (2) The infective particle concentration is modified by season such that more new infections are acquired in the warmer months. (3) The infective particle concentration is affected by local changes in salinity such that lower salinity is assumed to increase infective particle mortality or decrease par- ticle infectivity. producing a salinity-dependent gradient in preva- lence. (4) The infective particle concentration is affected by 2 regional processes. Winters with unusually cold temperatures re- duce infective particle concentrations for up to 2 years. Oscilla- tions in average bay salinity above and below 15.59£c introduce a source of variatioin in infective particle concentration that occurs uniformly throughout the bay. (5) Changes in oyster filtration rate affect the rate of infective particle uptake and. hence, the rate of infection. Only the last process involves the host population. The processes by which H. nelsoni is transmitted to uninfected oysters, and the form of the infective particle, are not known. However, observations that the earliest infections are in the gill epithelium indicate that infective particles are acquired through filtration (Farley 1968. Ford and Haskin 1982). Hence the rate at which new H. nelsoni infections occur in uninfected oysters (Oqq) is dependent upon the number of infective particles filtered out of solution. This rate. P(H). is given by: Poo = " /, (4) h IPfilter where IPfilter is the number of infective particles filtered by the oyster. The relationship assumes a threshold level of infective particles needed to generate a new infection, the infective dose, of 8,700 particles filtered d_1. This infective dose was estimated from information provided by Barber and Ford (1992) on the number of haplosporidian spores in oyster guts, an estimate of the filtration rate of these oysters, and a series of simulations of varying infec- tive dose compared to field observations of the infection process. The remainder of the transmission submodel is designed to esti- mate IPfilter. Oyster Filtration Rate The number of infective particles filtered by the oyster was modeled as: IPfilter = IPeonc filt(size) IP season IP sal IP temp (5) where IPeonc is the ambient infective particle concentration in the water column, filt(size) is oyster filtration rate, IPseason is the seasonal variation in infective particle availability, and IPsal and lPtemp are the salinity and temperature effects on infective par- ticles, respectively. Oyster filtration rate is calculated using the relationships given in Hofmann et al. (1992. 1994) and is param- eterized as a function of temperature, salinity, and total seston as described in these references. The relationships used to specify the seasonal, salinity, and temperature dependencies of the infective particles are described in the next sections. Base Infective Particle Number The base concentration of infective particles, IPeonc a, was cho- sen by comparing results of simulations using a range of values to field observations of prevalence. This value was used to initialize the model and to establish, as twice its value, a maximum con- centration of infective particles (IPconcmaJ that could not be ex- ceeded during a simulation. 520 Powell et al. Seasonal Cycle in Infective Particle Number Early studies with timed imports of oysters into enzootic re- gions of Delaware and Chesapeake Bays showed that oysters be- came infected only during a period from late May through early October (Andrews 1968, Ford and Haskin 1982), suggesting a seasonal dependence in the ambient concentration of infective par- ticles. The concentration of infective particles in the water is a critical element in modeling transmission; however no measure- ments are available to parameterize this process. Recently, how- ever. Barber and Ford (1992) reported finding haplosporidian spores in the digestive tract lumena of oysters in Delaware Bay and other regions enzootic for H. nelsoni. These spores, morphologi- cally similar to spores produced by H. nelsoni, had obviously been ingested during feeding. Further, numbers were higher from May through October, the known infective period fori/, nelsoni. Barber and Ford (1992) extrapolated their findings to estimate that mean spore abundance in the water the oysters were filtering was several hundred spores per liter. These may not be H. nelsoni spores, and if they are. they may not be the stage that infects oysters. Never- theless, they represent the only available data on which to base a rough estimate of the seasonal variation in concentration of infec- tive particles. The base concentration of infective particles was modified sea- sonally based on the data obtained from examination of oyster gut contents. This time series (Fig. 1 ) was taken to reflect the relative abundance of infective particles. The seasonal change in relative abundance was included in equation (5) as IPseason. Salinity Effect on Infectivity Initial simulations of H. nelsoni prevalence in oysters from low salinity environments suggested that the rate of infection decreases with decreasing salinity. An effect of salinity on H. nelsoni mor- tality within the oyster (Paraso et al. this volume) did not ad- equately simulate changes in prevalence, even when false nega- tives were taken into account. A function that decreased infection efficiency at low salinity resulted in simulated prevalence levels that better matched those recorded on the low-salinity Delaware Bay seed beds (Paraso et al. this volume). The function was ob- tained by using the model to simulate infections over a broad range of salinities in Delaware Bay and comparing these to long-term 1.0 0.6 o 0.4 CO 0.2 0.0 I 1 X 1 R x j., SL, jQ time series (Ford and Haskin 1982. Haskin and Ford 1982). Based on these comparisons, the effect of local salinity on transmission rate was modeled as: JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Figure 1. Time series of putative Haplosporidium nelsoni spores ob- served in sectioned Crassostrea virginica gut lumena (from information in Barber and Ford 1992). (S-SM„) /Psal = - + ,a""[SM<^M7 (6) The relationship is biologically realistic because the salinity range affecting transmission is similar to the range affecting mortality in the host animal (Paraso et al. this volume). Whether equation (6) simulates the mortality of infective particles or their decreased ability to infect oysters at low salinity is unknown. Bay-wide Oscillations Induced by Salinity Simulations with long-term time series that were designed to test the adequacy of the transmission model, using the basic pro- cess of oyster filtration, infective dose, the seasonal cycle of in- fective particle availability, and a local effect of salinity on infec- tivity, showed adequate simulations for oyster populations over some portion of the salinity range in Delaware Bay (e.g.. Paraso et al. this volume) during most years. However, the same parameter- izations failed in Chesapeake Bay and over the entire salinity range in Delaware Bay. Although the seasonal cycle of infective particle availability may be somewhat different, certainly most factors in- fluencing transmission should be equivalent in both bays. This suggested an additional process was needed to model transmission rate. Review of long-term time series of H. nelsoni prevalence taken simultaneously at multiple sites across the salinity gradient in both bays revealed relatively simultaneous oscillations in disease preva- lence over a wide portion of the salinity gradient, even at locations where salinity remained high enough to have only a limited effect on H. nelsoni growth and mortality. Addition of bay-wide salinity- dependent multiyear oscillations in infective particle availability allowed both bays to be modeled with very minor differences in the values of only 3 variables. IPconc0, Ipconcmux. and DD„. (Variations in DD„ are discussed in the following section.) These oscillations are parameterized as follows. The rate of salinity change was calculated as: IPsalrate = IPsalrate^ lPsal0 IP sal. (7) where IPsalrate0 specifies the response time of the infective par- ticles to changes in salinity, which was taken to be 180 days. The salinity value used to specify SlP is assumed to be represen- tative of the salinity at which an H. nelsoni secondary host lives or where some other reservoir of infective particles is found. For the simulations given in the following sections, the value ot 5/P was taken from the most downestuary (highest salinity) site showing strong salinity excursions across the \5%c isohaline in both Delaware and Chesapeake bays. Lower salinity sites failed to provide adequate simulations in either bay and higher salinity sites were not present in the suite of available Chesapeake Bay time series. The concentration of infective particles was updated each time step, based on this rate (IPsalrate) forced by the direction and magnitude of salinity change. So, for increasing salinities (IPsal- rate > 0), dIPconc Jt IPsalrate (IPconc,, ■ IPconc). (8) Modeling MSX Transmission in Oysters 521 For decreasing salinities (IPsalrate < 0). dIPconc dt IPsalrate [IPconc - IPconc,,,,,,) (9) and. a model initialization. IPconc = lPconc0. The new value of I Panic was then inserted into equation (5). Not having any evi- dence to the contrary, the value of lPconc,„UK was routinely set at twice the initial base infective particle concentration [IPconca) and IPconc,,,, „ was set near zero. Values for all constants are given in Table 3 of Ford et al. (this volume). Effect of Cold Winters Long-term observations from Delaware Bay show a cyclic pat- tern of H. nelsoni activity in which years of low infection preva- lence follow, typically with a lag of 1 to 2 years, very cold winters (Ford and Haskin 1982). Examination of the long-term data sets for Chesapeake Bay showed the same phenomenon. Thus, in some years, very few oysters become infected, even when appropriate salinity conditions are present (Haskin and Ford 1982, Paraso et al. this volume). This pattern suggests that, in some way, the abun- dance of infective particles is diminished after cold winters. In the model, direct temperature effects on infective particle abundance were included through a calculation of degree days that is based on 10 °C. The number of degree-days in which the tem- perature (T) is below 10 °C (DD10) from January (Julian Day 1) to May (Julian Day 150) is accumulated as: DD10: io-r (10) and the value of DD 10 is then used to estimate the extent to which cold temperature affects the survival of infective particles (IPtcmp,sl) as: ' DD\0-DDo\~ IPtempest - 1 - tanli DD, DD, (11) where DD„ is a threshold value at which the temperature effect becomes active. Values for the constants are given in Table 3 of Ford et al. (this volume). Equation (11) provides a value for the temperature effect that is based upon the current degree-day calculation. To simulate the observed delay in the manifestation of winter temperature effects on //. nelsoni infective particles, the value of IPtemp in equation (5), determined from the current DD10 value, was modified based on the value calculated for the previous year. If the current value of £>£>10 is less than one-half of the threshold value (DD0), then this indicates that the current year's winter is considerably warmer (an extreme difference) than that in the previous year, and the current value of IPtemp e„ is used as IPtemp. If the current value of DD10 is greater than one-half DD„ and less than the value for the previous year, such that the current year's winter is only somewhat warmer than the previous year's winter, the current and previous year's values are averaged to obtain the value for IPtemp. This allows the conditions in the previous winter to affect the level of infectivity by H. nelsoni and thereby allows for persistence of the effects of harsh winters over a period of more than 1 year, as observed. If DD 10 is greater than one-half DDlt and greater than the value calculated for the previous year, then the current condi- tions are colder than the previous year's conditions and also char- acteristic of an average to cold winter. In this case, IPtemp is specified using the current value of lPtemp,.u. FIELD OBSERVATIONS The environmental factors that force the oyster population-//. nelsoni model are salinity, temperature, food, and total seston (to- tal suspended solids). We compared simulations of 3 time series to field data of H. nelsoni prevalence for verifying the model. In each case, the simulations and field observations covered a substantial fraction of the salinity gradient in which oysters live. These time series were: 1960 to 1970 for Delaware Bay, 1980 to 1990 for Delaware Bay, and 1989 to 1994 for Chesapeake Bay. Delaware Bay Prevalence and infection intensity of//, nelsoni were measured over the period 1959 to 1992 by personnel from the Haskin Shell- fish Research Laboratory at Rutgers University (Ford and Haskin 1982, Haskin and Ford 1982. Fegley et al. 1994). Oysters from the downbay planted grounds were obtained by dredge at regular in- tervals during the year. Oysters from the upbay seed beds were typically collected during late autumn/early winter and the follow- ing late spring. Five locations, identified in Fig. 1 of Paraso et al. (this volume), distributed along the salinity gradient, were chosen for simulation. From lowest to highest salinity, these were Arnolds Bed. Cohansey Bed, Shell Rock Bed, Bennies Bed, and Miah Maull Grounds. These sites are described in more detail in Paraso et al. (this volume). The time series of H. nelsoni prevalence for these sites show that the decades of the 1960s and 1980s contain epizootics of 3 to 5 years duration, preceded and followed by periods of relatively low H. nelsoni prevalence (Fig. 2). These decades, therefore, test the full range of the model's capabilities. Temperature time series for these 2 decades were obtained from long-term temperature recordings at the Haskin Shellfish Research Laboratory pier in the lower Maurice River, supple- mented by once-to-twice monthly recordings at each of the 5 sites for the 1960s and 1980s, except for 1989 to 1990 when very few measurements were available (Fig. 3). Simulations of the 1980s decade are credible only through 1988. The same temperature time series was used for all 5 locations along the salinity gradient. The DD 10 value, given for each winter (Fig. 3). is a measure of the severity of the winter, with higher values indicating colder winters. By this measure, the decade of the 1960s was characterized by particularly cold winters in 1962 and the period 1968 to 1970 (Fig. 3). A relatively warm period from 1963 to 1967 coincided with a prolonged drought and relatively high-salinity conditions. The 1980s were a period of consistently warmer temperatures. Only 1982 would be considered cold by 1960s standards and 7 winters were warmer than any winter recorded in the 1960s decade. Salinity time series were derived from monthly-averaged Dela- ware River flow measurements taken at Trenton. New Jersey, by the U.S. Geological Survey using the relationship between Dela- ware River flow and salinity derived by Haskin (1972) as de- scribed in Paraso et al. (this volume). The Haskin relationship was clearly appropriate for the 1960s time series. However, the 1980s time series followed the establishment of the relationship by about a decade and, so, salinities may not be predicted as accurately for this decade. In particular, the century-long gradual increase in Bay salinity observed by Haskin (1972) presents the possibility that the salinities used for the 1980s time series were somewhat underes- timated. The Delaware Bay salinity time series (Figs. 4, 5) is charac- terized by spring lows and summer/fall highs coincident with the typical variation in Delaware River flows (Fig. 2). The decade of 522 Powell et al. 25000 20000 15000 10000 5000 0 100 □ Spring River Flow Delaware l,„llllllllllllll,l|llllllll 195860 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 Figure 2. Delaware River flow and prevalence of Haplosporidium itel- soni for (a) Arnolds Bed, (b) Cohansey Bed, (e) Shell Rock Bed, (d) Bennies Bed, and (e) Miah Maull Grounds constructed from data in Fegley et al. (1994). Annual cycles are represented by two measure- ments. Open bars on the left are May/June conditions; shaded bars on the right are November/December conditions. Location map is Figure 1 in Paraso et al. (this volume). the 1960s was characterized by increasingly saline conditions, on the average, in the first 6 years of the decade, followed by a strong freshening trend that began in 1967 (Fig. 4). The 1963 to 1967 drought coincided with a period of relatively mild winters com- pared to other years in the decade. The decade of the 1980s (Fig. 5) had fewer anomalously wet or dry years, although 1985 was characterized by an unusually dry spring and salinity, on average, decreased in the latter years of the decade at all sites. Neither food nor total seston data were available for most of the 1980s or for any of the 1960s. Accordingly the food and total seston data sets developed by Powell et al. ( 1997) were employed. These time series are based on total seston and chlorophyll mea- surements made at several locations in Delaware Bay by Haskin Shellfish Research Laboratory (HSRL) scientists at about monthly intervals from 1981 to 1986. with the sampling frequency in- creased to bi-weekly between 1982 and 1984. For this study, the time series from a site just east of the Delaware Bay Ship Channel opposite Kelly Island. Delaware (39°14.44'N 75°16.47"W). was taken to be representative of conditions on the 4 seed beds. Mea- surements at a site just south of Egg Island, New Jersey) 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 Time Figure 3. Temperature time series used for the decadal runs of the (a) 1960s and (h) 1980s in Delaware Bay. Values for 1989 and 1990 are based on only a few measurements and should be discounted. Numbers above each winter are values of the model variable DDIO, a measure of the coldness of the winter. Higher values indicate colder conditions. Location map is Figure 1 in Paraso et al. (this volume). 39°10.46'N 75°5.05'W). were assumed to be representative of the planted grounds. The 6 years of data were averaged to obtain a 1-year average data set at both locations that was used each year during the decade-long simulations. The same data sets were used for the 1960s and 1980s simulations. Total suspended solids at both sites show variability throughout the year, with maximum values tending to occur in late spring to early autumn (Fig. 6). Total suspended solids average about twice as high on the planted grounds than over the 4 seed bed sites throughout the year. The chlorophyll time series shows a distinctive spring bloom that occurs in March to May. with the maximum in March downbay and about one-half month later over the seed beds (Fig. 6). A consistent fall bloom does not occur at either site, although transient increases in chlorophyll content do occur from time to time. Rather, at both sites, chlorophyll values drop to seasonally low levels in July and remain, for the most part, at or near these levels until the next spring. Chlorophyll values average higher over the planted grounds than any of the seed bed sites throughout the year, including the spring bloom and the much lower summer and fall mean. As a consequence, food supply drops to very low levels in the summer and fall over the seed beds. Chlorophyll a in u.g L~' was converted to oyster food in mg DW L~' using the relationship derived by Powell et al. (1997): 30 25 20 - 15 Modeling MSX Transmission in Oysters 20 523 _L _L _L J_ _L _L _L 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 Time Figure 4. Salinity time series used for the deeadal runs of the 1960s in Delaware Bay for (a) Arnolds Bed, (b) Cohansey Bed, (c) Shell Rock Bed, (dl Bennies Bed, and (e) Mi.ili Maull Grounds. Location map is Figure 1 in Paraso et al. (this volume). where a Food = a x chlorophyll a + (3 0.088 mg p.g~' and (3 = 0.26 mg LT (12) Chesapeake Bay H. nelsoni prevalence was derived from hemolymph assays measured by personnel at the Oxford Cooperative Laboratory. Maryland Department of Natural Resources (DNR), for the periods 30 25 20 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 Time F'igure 5. Salinity time series used for the decadal runs of the 1980s in Delaware Bay for (al Arnolds Bed, (b) Cohansey Bed, (c) Shell Rock Bed, Id) Bennies Bed, and (e) Miah Maull Grounds, Locatioin map in Figure 1 in Paraso et al. (this volume). 1990 to 1995 for a series of sites in the Maryland portion of Chesapeake Bay covering the full range of the salinity gradient in Maryland waters (averaging near 0%c to around \5%c). Sites TSOW (Tangier Sound-Old Woman's Leg). TSSS (Tangier Sound-Sharkfin Shoal), FBGC (Fishing Bay-Goose Creed). LCRP (Little Choptank River-Ragged Point). MESR (Middle Eastern Shore-Stone Rock), and PRCH (Potamic River-Cornfield Harbor) were selected for simulations (Fig. 7). Hemolymph prevalence data for each of these stations were used for verification (Fig. 8). Oyster shells were drilled. After purging the drill site in seawater for 12-24 h, 0.5 mL hemolymph was withdrawn from the adductor muscle with a sterile syringe. Four drops of hemolymph were placed in a well slide with 0.5 mL sterile seawater and allowed to settle for 30 min. After settling, seawater was decanted from the well and replaced with Davidson's fixative. The preparation was 524 Powell et al. a. >• ■C a o o 80 70 60 50 40 30 20 10 Kl b_L 1 M N M N a 50 M N N N^j ,0§J _L J L J L J L J_ 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 Time 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 Time Figure 6. Chlorophyll a and total seston time series used for the decadal runs of the 1960s and 1980s in Delaware Bay. Time series are composed of 10 repeats of an average year derived by averaging data obtained during the time period 1981 to 1986. (a,cl Data for Mian Maull Grounds: (b,d) data for the oyster seed beds, including Arnolds Bed, Cohansey Bed, Shell Rock Bed. and Bennies Bed. The same time series were used for the 1960s and the 1980s simulations. then stained with Mayer's hematoxylin and eosin. and cover- slipped. The resulting slides were read, with counts of MSX Plas- modia in standard microscope fields recorded on a semi- quantitative scale. Occasional values of tissue prevalence, done histologically, were always higher, so that the hemolymph values plotted in Figure 8 should be viewed as conservative estimates of prevalence. Time series of salinity, temperature, food, and total seston were obtained from stations near the oyster bed sampling sites that are part of the Environmental Protection Agency (EPA) Chesapeake Bay Monitoring Network (Fig. 7). Data were obtained from the EPA public access web site: www.chesapeakebay.net/ bayprogram/data/wqual/wqgate.html. Maryland DNR (EPA) sta- tion pairs are given in Figure 7. Data for the deepest depth sampled were used except in cases where the EPA station was in a channel. In this case, a mid-water depth was used to be more representative of the conditions on the nearby oyster bed. Salinity was highest at Tangier Sound-Old Woman's Leg (EE3.2) and lowest at Middle Eastern Shore-Stone Rock (CB4.2) (Fig. 9). Each of the 6 data sets was characterized by a relatively wet period in 1990/1991 and a relatively dry period in 1992/1993 followed by a period of extreme oscillations in 1994/1995. The temperature time series were nearly identical among the 6 sites (Fig. 10). Years 1990 and 1994 were markedly colder than the remaining years. Year 1992 was markedly warmer. Chlorophyll values typically showed a strong spring bloom and a much smaller, but distinct, fall bloom (Fig. 11). The spring bloom was more intense and predictable at Tangier Sound and Little Choptank River stations. With the exception of EE2.2 in the Little Choptank River, 1992 was an anomalous year with unusu- ally low chlorophyll values and the near absence of a spring bloom. Chlorophyll a data were converted to oyster food using equation (12). Total seston was chaotic in its time series, as it was in Delaware Bay (Fig. 12). Stations in the main bay stem. Middle Eastern Shore-Stone Rock (CB4.2), and Potomac River-Cornfield Harbor (LE2.3), had lower suspended loads in contrast to stations on the eastern side in Tangier Sound (Old Woman's Leg and Sharkfin Shoal) and Little Choptank River-Ragged Point (EE2.2), where total seston typically varied between 20 and 50 mg L~'. Modeling MSX Transmission in Oysters 525 PAIRED SAMPLING SITES ■ water quality sites O oyster disease sites 76° 75° Figure 7. Distribution of Oxford Cooperative Laboratory sampling sites for Haplosporidium nekoni prevalence and EPA sampling sites for environmental variables. The sites used for simulations are (Oxford/EPA) TSOW/EE3.2, TSSS/EE3.1, FBGC/EE3.0, LCRP/EE2.2, MESR/ CB4.2, PRCH/LE2.3. TSSS, Tangier Sound-Sharkfin Shoal; TSOW, Tangier Sound-Old Woman's Leg; LCRP, Little Choptank River-Ragged Point; FBGC, Fishing Bay-Goose Creek; MESR, Middle Eastern Shore-Stone Rock; PRCH, Potomac River-Cornfield Harbor. Additional information is available in Smith (1997). MODEL IMPLEMENTATION Details of model implementation are presented in Ford et al. (this volume). The limited understanding of the true process of transmission required that the model be verified using observations of the time series of prevalence in oyster populations. That is. the model was verified based on knowledge of the result of the process rather than the process itself. In approaching this task, our modus operandi included two principal constraints: (1) that processes induced in the model had some ecological and biological rel- evance, that is. that the processes existed and could have been responsible for the observed results; and (2) that the two bays, Chesapeake and Delaware, and the two time periods in Delaware Bay should be simulated adequately with very few differences in model construction and variable values, that is. that the transmis- sion process was common to all places and times. RESULTS Base Cases for Verification Delaware Bay, 1960s A drought during the middle 1960s generated an H. nelsoni epizootic in Delaware Bay. Prevalences were observed to increase down the salinity gradient from about 30% at Arnolds Bed to near 80% at Bennies Bed in 1965. Simulated H. nelsoni prevalence during the 1960s using an initial infective particle concentration of 900 infective particles L"1 shows a similar pattern. Prevalence increases downbay from about 20% at Arnolds Bed to 70% at Bennies Bed in 1965 (Fig. 13). About one-half of the simulated H. nelsoni infections are strong systemic (BFU = 3 or 4) infections, which corresponds to observation. The epizootic in the simulated population diminishes over time and H. nelsoni nearly disappears from the bay by 1970. as was observed. Observations at Bennies Bed showed that the epizootic was strongest in 1965 through 1967. This is also true in simulation. Simulated cumulative mortalities reach about 80% by 1968 at Bennies Bed and decline upbay to less than 10% during the same period as Arnolds Bed. Accordingly, the simulations of the time-dependence of H. nelsoni infection agree. to the level represented in the observed disease prevalence and intensity time series, with what occurred in Delaware Bay during the decade of the 1960s. Delaware Bay, 1980s The 1980s period was characterized by an H. nelsoni epizootic during 1982 and 1985. H. nelsoni prevalence declined late in the decade and during the 1982 to 1983 period (Fig. 2). In the simu- 526 Powell et al. 60 50- 40 30 20- 10- EE31 N = No data 0 ■ Zero blood prevelance _i 60 50 40 30 20 10- o ' Q. SO- TS O 40 _o m 30 20 10 0 60 50 40 30 20 10 0 60 50 40 30 20 10 EE32 LE23 CB42 EE22 EE30 Y777A V77A 8-Oct-90 12-NOV-91 16-NOV-92 26-Oct-93 26-Oct-94 Time Figure 8. Haplosporidium nelsoni prevalence from blood assays taken by the Oxford Cooperative Laboratory at the 6 sites identified in Figure 7. TSSS/EE3.1, Tangier Sound-Sharkin Shoal; TSOW-EE3.2, Tangier Sound-Old Woman's Leg; LCRP/EE2.2, Little Choptank River-Ragged Point; FBGC/EE3.0, Fishing Bay-Goose Creek; MESR/ CB4.2, Middle Eastern Shore-Stone Rock; PRCH/LE2.3, Potomac River-Cornfield Harbor. lations, the 1985 epizootic and the decline in prevalence of H. nelsoni in 1983/1984 and in the last few years of the decade are faithfully reproduced at most sites (Fig. 14). Prevalences in the 1982 epizootic are somewhat exaggerated because of adjustments that occur early in the siimulation as H. nelsoni prevalence and infection intensity come into equilibrium with environmental con- ditions. These adjustments normally take about 24 months of simu- lation, but the time required is exaggerated during epizootic con- ditions. Overall, the simulated prevalences for the 1980s underestimate observations throughout the salinity gradient, and particularly so at upbay sites, despite the use of a significantly higher maximum infective particle concentration (1,500 infective particles L_l ver- sus 900 particles L-1) than used in the 1960s simulations. The necessity of using varying concentrations of infective particles as initial conditions will be discussed in a later section. 1990 1991 1993 1993 199S Figure 9. Salinity time series for Chesapeake Bay obtained from the EPA stations identified in Figure 7. (a) CB4.2: (b) EE2.2; (c) LE2.3; (d) EE3.0; (e) EE3.1; and (f) EE3.2. Chesapeake Bay, Early 1990s Observations show that an H. nelsoni epizootic occurred in 1992 at all sites (Fig. 8) and H. nelsoni prevalence increased again in 1995. The 1992 to 1993 epizootic was restricted temporally at most low-salinity sites in 1992, whereas higher salinity sites Figure 10. Temperature time series for Chesapeake Bay obtained from the EPA stations identified in Figure 7. (a) CB4.2; (b) EE2.2; (c) LE2.3; (d) EE3.0; (e) EE3.1; and (f) EE3.2 Modeling MSX Transmission in Oysters 527 1 y 1 aJ b ; c ! *r 40 ' E - 30 - a 20 0 '■ . A i : Al l\ h 1A ,A ■ O 10 \r\iV\ AhA^ '■ ■. , ! v>f w y . ■ 1990 1991 1992 1993 1994 1995 1990 1991 1992 1993 1994 1995 Time Time Figure 11. Chlorophyll a time series for Chesapeake Bay obtained from the EPA stations identified in Figure 7. (a) CB4.2; (b) EE2.2; (c) LE2.3; (d) EE3.0: (e) EE3.1; and (f) EE3.2. " a BFU = 1 - BFU = 2 - : BFU = 3 - — BFU = 4 " i r ^dm 1962 1963 1964 1965 1966 1967 1966 1969 1970 1971 Time 961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 Figure 13. Time-development of Haplosporidium nelsoni infection in Delaware Bay during the 1960s using a maximum infective particle concentration of 900 infective particles L_I and a standard 1 g AFDW oyster, (a) Arnolds Bed. (b) Cohansey Bed, (c) Shell Rock Bed. (d) Bennies Bed, and (el Miah Maull Grounds, (f) Cumulative mortality during the simulation for each of the 5 simulated locaitons. - 0040 - '. c ' ■J\ 1 \ 1 ll V d v : 1990 1991 1992 1993 Time 1992 1993 Time Figure 12. Total seston time series for Chesapeake Bay obtained from the EPA stations identified in Figure 7. (a) CB4.2: (b) EE2.2; (c) LE2.3; (d) EE3.0; (e) EE3.1; and If] EE3.2. in Tangier Sound (TSOW. TSSS) were characterized by higher prevalences in 1993 as well and prevalences above 10% in most years. In contrast to the longer time span of the epizootic downbay. the epizootic reached highest prevalences upbay at the lower sa- linity sites, where prevalences in 1992. measured by hemolymph assay, exceeded 40%. The observations of H. nelsoni prevalence (Fig. 8) were taken in late October/early November of each year. It is during this time that the model simulations typically show rapid changes in preva- lence, making comparison to field data difficult. Nevertheless, the 1992 epizootic is present in each simulation, as are the low preva- lences before and after (Fig. 15). The epizootic, as simulated, is temporally more extensive at Tangier Bay sites, as observed, and tends to be restricted to 1992 at upbay sites, also as observed. Simulated upbay prevalences are not necessarily higher than downbay prevalences, however. Imposition of a spatially variable concentration of infective particles at the two most-upbay sites (Little Choptank River and Middle Eastern Shore) would correct this error. Nevertheless, except for Little Choptank River (under- estimated by about half) and Tangier Sound-Old Woman's Leg (overestimated by about a factor of 3), simulated prevalences are very close to observation. In both of the latter cases, similar preva- lences exist in the simulations within 1 month of observation, so that small changes in timing are critical. In particular, the fall rise in prevalence is keyed to a fall increase in salinity at most sites and the exact timing of this salinity excursion is crucial because H. nelsoni cannot increase in intensity as long as salinities stay much below 15%c. Finally, the model simulates tissue, not hemolymph prevalence, and these can be significantly different (Ford and 528 Powell et al. _i i i i_ *-■ * ' — J L - c \ M \ y \ ■ 11 « | | - :.l Ik. JL ~] ^lA_k_j" 1983 1984 1985 1986 1987 1988 1989 1990 1991 Time 1982 1983 19W 1985 19B6 1987 1 Time Figure 14. Time-development of Haplosporidium nelsoni infection in Delaware Bay during the 1980s using a maximum infective particle concentration of 1,5011 infective particles L_1 and a standard 1-g AFDW oyster, (a) Arnolds Bed, (b) Cohansey Bed. (cl Shell Rock Bed, (dl Bennies Bed, and (e) Miah Maull Grounds. Kanaley 1988). Accordingly, within the limitations of the model and observations, the model accurately simulates the time- dependent growth and decay of the 1992 to 1993 Chesapeake Bay epizootic. | 0.4 E " a BFU = 1 _ BFU = 2 . - BFU = 3 - " - - BFU = 4 : n M ( 1 ': - M - Pfri . -1. - i A. I\ -, „-^t-~ / V «& m Components of the Transmission Model 1990 1991 1992 1993 1994 1995 1990 1991 1992 1993 1994 1995 Time Time Figure 15. Time-development of Haplosporidium nelsoni infection in Chesapeake Bay during the early 1990s using a maximum infective particle concentration of 900 infective particles L~' and a standard 1-g AFDW oyster, (a) Middle Eastern Shore-Stone Rock; (b) Little Chop- tank River-Ragged Point; (cl Potomac River-Cornfield Harbor; (dl Fishing Bay-Goose Creek; (e) Tangier Sound-Sharkfin Shoal; (f) Tangier Sound-Old Woman's Leg. infective particle concentration except for estimates by Barber and Ford (1992). obtained from spore concentrations in gut lumena, that infective particle concentrations should be in the hundreds per liter. Oyster Filtration Rate Initiation of H. nelsoni infection requires filtration of an infec- tive dose. Total seston, among other environmental factors, can affect oyster filtration rate (Loosanoff 1962, Powell et al. 1992), as shown by a simulation in which the effect of a decrease in filtration rate was produced by an increase in total seston concentration (Fig. 16). Prevalence decreases as oysters filter infective particles less rapidly. Small changes in temperature, salinity, and total seston, as they affect filtration rate, will generate significant changes in prevalence. In this study, the rarity of environmental data and population data obtained from precisely the same location intro- duces a significant potential error in accurately simulating ob- served prevalence, because cross-bay and upbay-downbay varia- tions in environmental variables are significant (e.g., Garvine et al. 1992, Wong 1994). Initial Concentration of Infective Particles The model requires initialization with a beginning infective particle concentration {IPconcJ and establishment of a maximum particle concentration that cannot be exceeded during the simula- tion {IPconcmlx) set at twice lPconca. No information exists on 1.0 0.8 - u- 0.6 - 3 E E 3 o 0.4 - _ - BFU = 1 BFU =2 BFU = 3 BFU = 4 - - A I " I ^l I ^K J\ j \ A I I _L_ -j" 0.2 - 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 Time Figure 16. Time-development of Haplosporidium nelsoni infection on Cohansey Bed in Delaware Bay during the decade of the 1960s using a maximum infective particle concentration of 900 infective particles L"1 and a standard 1-g AFDW oyster, but using the higher total seston time series of Miah Maull Grounds (Fig. 6). The comparable reference case is given in Figure 13. Modeling MSX Transmission in Oysters 529 . a BFU= 1 BFU = 2 BFU = 3 - — BFU = 4 \ : J2v.~J^± 1990 1991 1992 1993 1994 1995 Time Figure 17. Time-development of Haplosporidium nelsoni infection in Chesapeake Bay during the early 1990s using a range of maximum infective particle concentrations and a standard 1-g AFDVV oyster. The simulations are run using the environmental conditions for Tangier Sound-Sharkfin Shoal. Maximum allowable infective particle concen- trations (lPconcmaj:) for these simulations are: (a) 70(1 particles L"1; (b) 900 particles L"1 (reference case in Fig. 15); (c) 1,100 particles L-1; (d) 1,500 particles L '. Initial particle concentration (ll'conc,,) was set at half this value. Variation in these two variables determining infective particle concentration significantly alters the simulated pattern of H. nel- soni prevalence. In Figure 17, simulations are presented in which the maximum allowed concentration of infective particles was var- ied from 700 to 1,500 particles L"'. Simulated prevalence in- creases in proportion with the concentration of infective particles. In Chesapeake Bay in the early 1990s and in Delaware Bay in the 1960s, the simulations agree with observations for a maximum infective particle concentration of 900 particles L_1; the value of which also agrees with estimates made by Barber and Ford ( 19921. In contrast, fitting simulated disease prevalences with observa- tions for the decade of the 1980s in Delaware Bay required using a maximum infective particle concentration of 1.500 particles LT . Simulations using lower concentrations failed to provide the ob- served prevalences (Fig. 18). This case represents one of two (the other to be discussed later) cases where changing model parameter values proved necessary to provide adequate agreement between simulation and observation between sites or times. Several pos- sible explanations exist for the higher infective particle concentra- tion required in the 1980s in Delaware Bay. ( 1 ) The infective dose has decreased over time in Delaware Bay. That is. oysters have become more susceptible to infection. A decrease in infective dose would allow infection to occur at a lower particle concentration. However, no evidence supports an increase in disease susceptibil- ity (e.g.. Andrews 1968, Farley 1975. Haskin and Ford 1979. Ford and Haskin 1982, Chintala and Fisher 1991). (2) H. nelsoni has become more resistant to low salinity in Delaware Bay. However, an increase in infective particle concentration was required even at the highest salinity sites (Bennies Bed and Mian Maull), where the influence of low salinity is minor or non-existent. Thus, a change in salinity tolerance fails to meet the requirements of all simula- tions. (3) Salinities have been underestimated in the 1980s using Haskin's (1972) regression equations between salinity and Dela- ware River flow. If salinities were higher, particularly at upbay ; a BFU = 1 BFU = 2 BFU = 3 BFU = 4 \ IjL ■Jill i ^x^^Aikt^^j^ \ 1981 Ida! 1983 1984 1985 1986 Figure 18. Time-development of Haplosporidium nelsoni infection in Delaware Bay during the 1980s using a range of maximum infective particle concentrations and a standard 1-g AFDW oyster. The simu- lations are run using the environmental conditions for Bennies Bed. Maximum infective particle concentrations [IPeoncmtuc) for these simu- lations are: (a) 700 particles L"1; (b) 1,000 particles L"1; (c) 1,500 particles L ' (reference case in Fig. 14). Initial particle concentration {IPconcJ was set at half this value. stations, than predicted from these equations, prevalence would increase for a given infective element concentration because the influence of low salinity on transmission is to decrease the rate of infection. Because the model simulations fit field observations increasingly poorly in the 1980s with decreasing salinity upbay, the evidence supports this option. However, an increase in infec- tive particle concentration was required even at high-salinity sites (Bennies Bed and Miah Maull). where the influence of low salinity is minor or non-existent. (4) The concentration of infective par- ticles has increased in Delaware Bay since the 1960s. Because modi- fications in the relationship of H. nelsoni with salinity do not ad- equately explain the discrepancy between simulation and observation and because H. nelsoni prevalence seems to have increased over time at high-salinity sites, this option remains the most viable of the four. Seasonal Cycle in Infective Particle Number Observations of H. nelsoni spores in oyster gut lumena suggest that infective particles are not present in the water column all year. Replacing the seasonal cycle with a constant value for infective particle availability gives an annual cycle of H. nelsoni prevalence (Fig. 19) that is relatively similar to the reference case (Fig. 13) because the temperature dependency of filtration rate reduces the filtration rate of infective particles in the winter. Nevertheless, infection intensities and prevalences average higher after removal of the seasonal cycle and this is due to higher infection rates during the spring warming than observed. Salinity Effect on Infectivity The effect of changes in salinity at the local population level has been investigated by Paraso et at. (this volume). Observations show that H. nelsoni prevalence declines with decreasing salinity. Although some part of this decline is induced by a reduction in infection in- tensity due to H. nelsoni cell mortality, which should increase the number of false negatives and thus decrease estimated prevalence, a direct effect of low salinity on cell survival in the host oyster could not fully explain the decline in prevalence with declining salinity. 530 Powell et al. P9* ^ii ^*^S£\£^^ 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 Time 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 Time Figure 19. Time-development of Haplosporidium nelsoni infection in Delaware Bay during the 1960s without the inclusion of a seasonal cycle in infective particle availability (Fig. 1) using a standard 1-g AFDW oyster. The simulation is run using the environmental condi- tions for Miah Maull Grounds. Maximum infective particle concen- tration {IPconcmaj:) was set at 90(1 particles L_1. A reference case com- parison plot is presented in Figure 13. Paraso et a), (this volume), accordingly, included a relationship that altered infective particle concentration according to local conditions of salinity, thereby reducing infectivity at low salinity. Bay-wide Oscillations Induced by Salinity H. nelsoni epizootics tend to occur simultaneously over large areas, including over significant fractions of the salinity gradient (Andrews 1968. Haskin and Ford 1982) (e.g.. Figs. 2 and 8). This generation of epizootics over large spatial scales requires that the infective particle populations be responsive to environmental change in two ways: the concentration of infective particles must increase dramatically prior to and during an epizootic and the concentration of infective particles must respond synchronously over large distances, even when environmental variables that might sometimes induce synchrony, such as simultaneous in- creases in salinity above 15%o, are not effective. Both of these responses must be non-local functions of salinity. Thus, the model includes a series of relationships that varies the concentration of infective particles according to the general rise and fall of salinity throughout the bay. These relationships do not use the specific values of the local salinity regime for each par- ticular site, because the absolute value and timing of salinity change vary considerably between sites. Rather, these relation- ships impose on all sites a simultaneous rise and fall in infective particle concentration based on the general, bay-wide variation in salinity caused by changes in freshwater inflow. The result is a simultaneous modification in infective particle concentration at all locations across the salinity gradient. When bay salinity increases. the concentration of infective particles increases. When bay salin- ity decreases, the concentration of infective particles decreases. The response time for this effect, determined by a series of simu- lations varying response time from 90 to 360 days, was determined to be 180 days. If this time-dependent and salinity -dependent variation in infective particle concentration is removed from the model, the simulated patterns of H. nelsoni prevalence no longer agree with field observations (Fig. 20). In particular, the magnitude Figure 20. Time-development of Haplosporidium nelsoni infections in Delaware and Chesapeake Bays without the imposition of bay-wide synchrony in the time-dependent variation of infective particle con- centration in a population of 1-g AFDW oysters. In this case, the relationship varying the concentration of infective particles according to the rise and fall in average bay salinity was deleted from the model. The simulation is run using the environmental conditions for (a) Ben- nies Bed in Delaware Bay during the 1960s and (b) Potomac River- Cornfield Harbor in Chesapeake Bay during the early 1990s. Refer- ence case comparison plots are presented in Figures 13 and 15. Maxi- mum infective particle concentration \lPconcmax) was set at 900 particles L . of change in prevalence between non-epizootic and epizootic con- ditions is significantly reduced. Clearly, generation of an epizootic requires an increase in the ambient concentration of infective par- ticles as well as environmental conditions conducive to the prolif- eration of H. nelsoni in the oyster and this increase in particle concentration must occur independent of the local salinity excur- sions experienced by any individual population. Possibly, a measure of residence time (e.g.. Marshall and Alden 1997). which should decrease with increasing freshwater inflow, or a measure describing the change in average bay salinity would more effectively determine temporal variation in infective particle concentration. Unfortunately, data were not adequate in either bay to calculate either variable at all sites. So. in developing the model, the need arose to choose a salinity time series as a surrogate to impose synchrony among all sites. A series of trials using the 6 Chesapeake Bay time series showed that the highest salinity time series available, from Tangier Sound-Old Woman's Leg, produced simulations closest to observations at all Chesapeake sites. This time series was characterized by significant salinity excursions across the 15%o line (Fig. 9f). A time series of nearly equivalent salinity range in Delaware Bay comes from Shell Rock Bed (Figs. 4c. 5c). This time series provided equally adequate simulations in Delaware Bay. Use of lower-salinity time series consistently re- duced prevalences below observations at all sites in the salinity gradient (e.g.. Fig. 21) and often modified the temporal pattern of an epizootic. For example, the 1960s epizootic on Bennies Bed is not only muted, but also shortened temporally by the utilization of a lower-salinity time series from Cohansey Bed (Fig. 21a). The results of these simulations clearly target higher-salinity conditions as the conditions that provide the best time series for varying infective element concentration in the water column. Be- cause the highest salinity time series available in Chesapeake Bay. which had concomitant information on H. nelsoni prevalence. came from Tangier Sound, an area characterized by average sa- linities around 15%o, this salinity range was used. The effective- ness of even higher salinity conditions remains to be tested. Effect of Cold Winters The year 1962 was particularly cold in Delaware Bay and 1990 was particularly cold in Chesapeake Bay. Simulations without an effect of cold winters on transmission rate showed epizootic con- ditions beginning early in both the decade of the 1960s in Dela- ware Bay and the early 1990s in Chesapeake Bay (Fig. 22). The Modeling MSX Transmission in Oysters 531 - b - f\s~ - : Jr^ W : -~-.y . i. . ._ i i c 08 -- - RFU - 1 BFU = 2 BFU = 3 BFU = 4 \ £ 0.6 i 04 £ 0.2 --^j — k^W.v i * i^g^' 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 Figure 21. Time-development of Haplosporidium nelsoni infections in Delaware and Chesapeake Bays with the imposition of hay-wide syn- chrony in the time-dependent variation of infective particle concen- tration in a population of 1-g AFDW oysters, however with a lower salinity time series used to generate synchrony. The simulations were run using the environmental conditions for (a) Bennies Bed in Dela- ware Bay during the 1960s but using the lower-salinity time series from Cohansey Bed rather than Shell Rock Bed as the time series controlling synchrony, and (bt Tangier Sound-Old Woman's Leg in Chesapeake Bay during the early 1990s, but using the lower-salinity time series for the Potomac River-Cornfield Harbor to control syn- chrony rather than the higher salinity times series from Tangier Sound. Reference case comparison plots are presented in Figures 13 and 15. Maximum infective particle concentration {lPconcmax) was set at 900 particles L"1. suggestion of Ford and Haskin (1982) that the effect of cold win- ters extends beyond the year in which the winter occurred is sup- ported by these simulations. Cold winters generate a multiyear response pattern in H. nelsoni transmission and, hence, in H. nel- soni prevalence. Fitting simulation to observation required that the degree of cold exposure necessary to trigger the effect of a cold winter be varied between Chesapeake Bay and Delaware Bay (see Ford et al. this volume. Table 3. variable DDj. In effect, a cold winter was defined to be colder in Delaware Bay than in Chesapeake Bay. 967 1963 1964 196S 1966 1967 1968 1969 1970 1971 " c ! - n : ~_ : '- ,Jts bm ; 1992 1993 Time 1994 1995 1992 1993 1994 1995 Time Figure 22. Time-de\elopment of Haplosporidium nelsoni infections in Delaware and Chesapeake Bays without the influence of cold winters on transmission rate in a population of 1-g AFDW oysters. The simu- lations were run using the environmental conditions for (a) Bennies Bed, and (b) Cohansey Bed in Delaware Bay during the 1960s, (c) Tangier Sound-Sharkfin Shoal, and (d) Fishing Bay-Goose Creek for Chesapeake Bay during the early 1990s. Reference case comparison plots are presented in Figures 13 and 15. Maximum infective particle concentration iIPconcmax) was set at 900 particles L . Temperature data came principally from moorings in Chesapeake Bay and from the Maurice River pier in Delaware Bay. Possibly, pier temperatures average lower than Delaware Bay temperatures during the winter. A small discrepancy would be sufficient to account for the variation required between the two bays. No alter- native explanation is available and the absence of field information on the location and type of reservoir of infective particles prohibits the formulation of additional hypotheses. The difference in value for this variable (DD„) that controls the definition of a cold winter represents the sole significant difference in model setup between the Delaware Bay 1960s case and the case for Chesapeake Bay in the early 1990s. All simulations reproduced observed trends in MSX disease except for the terminal years of the 1960s for Miah Maull Grounds in Delaware Bay where prevalences were considerably underesti- mated. The simulations suggest that high salinity, typical of Miah Maull grounds, might mitigate the effects of cold winters. Insuf- ficient information from similar salinity sites was available in Chesapeake Bay to further test this possibility. Disease Resistance H, nelsoni resistance can clearly be bred into oysters and the development of disease resistance has occured in wild populations (Andrews 1968, Haskin and Ford 1979, Burreson 1991). An in- crease in disease resistance might introduce another mechanism controlling prevalence, rather than transmission, by limiting pro- liferation and thereby increasing the number of false negatives. One mechanism of disease resistance is the restriction of H. nel- soni plasmodia to the epithelial tissue for longer periods of time. This effect was introduced into the model by increasing the dou- bling time (b0 in Table 3. Ford et al. this volume), thus decreasing the growth rate of H. nelsoni in the epithelial tissues. (Varying the diffusion rate, d0 in Table 3. Ford et al. [this volume], between epithelial and systemic tissue did not provide the desired result.) A simulated increase in H. nelsoni growth rate results in little increase in prevalence, however, the duration of epizootics is lengthened because environmental effects limiting H. nelsoni pro- liferation, such as salinity, are lessened in importance (Fig. 23). An increase in resistance significantly reduces prevalence (Fig. 23), Figure 23. Time-development of Haplosporidium nelsoni infections in Delaware and Chesapeake Bays with varying disease resistance in a population of 1-g AFDW oysters. The simulations were run using the environmental conditions for (a) Shell Rock Bed in Delaware Bay during the 1960s, and (b) Potomac River-Cornfield Harbor in Chesa- peake Bay during the early 1990s. Reference case comparison plots are presented in Figures 13 and 15. (a) is a case of decreased resistance obtained by decreasing the doubling time of H. nelsoni plasmodia in the epithelium from 3 days to 1.5 days, (b) is a case of increased disease resistance obtained by increasing the doubling time of H. nelsoni plas- modia in the epithelium from 3 days to 12 days. Maximum infective particle concentration iIPconcmax) was set at 900 particles L~'. 532 Powell et al. but not because infectivity changes. As anticipated, prevalence is reduced because false negatives are more frequently encountered. Simulations for Delaware Bay in the 1960s and 1980s and Chesapeake Bay in the early 1990s agree with field observations using the same H. nelsoni doubling time. These simulations sup- port the belief that resistance builds up slowly or not at all in natural populations following the first few years of exposure, pos- sibly because reservoir populations of oysters at low salinity are protected from H. nelsoni exposure and continue to supply non- resistant larvae to downbay populations. On the other hand, disease resistance might entail a decrease in susceptibility or an increase in infective dose. The need for a higher infective particle concentration in Delaware Bay during the 1980s might be explained by an increase in susceptibility or de- crease in infective dose (e.g., Fig. 23a), however, no evidence exists in observations of natural populations for a decrease in resistance and better explanations for the difference in simulation conditions between the 1960s and 1980s decades are available (see earlier discussion). The successful simulation of the 1960s decade in Delaware Bay and the early 1990s in Chesapeake Bay using the same model conditions argues against a long-term decrease in susceptibility. Accordingly, variations in disease resistance and susceptibility were not included among the factors affecting H. nelsoni transmission and disease intensification. DISCUSSION Perspective In the vast majority of disease models, transmission is modeled as a simple function of the number of infected individuals and the contact rate between infected and uninfected individuals (e.g., Ker- mack and McKendrick 1991, Hethcote and van den Driessche 1995, Frank 1996). Intermediate or alternate hosts and parasitoids can make the process more complex (e.g., Hassell 1982, Chesson and Murdoch 1986, Woolhouse and Chandiwana 1990). An ex- ample of the adequacy of the simple approach using prevalence and contact rate to model transmission in bivalve populations is the model of Perkinsus marinus in Crassostrea virginica, in which transmission was controlled by 3 variables, disease prevalence and infection intensity, both measures of the population of infected individuals, and host density, a surrogate of contact rate in an immobile host (Hofmann et al. 1995. Powell et al. 1996). Contrast this simplicity with the present model of transmission in H. nelsoni. In H. nelsoni, the standard approach involving the fraction of the population infected and contact rate with uninfected individuals failed utterly to adequately simulate prevalence. It was clear from the earliest studies that transmission in the field did not depend on the presence of infected oysters and that transmission could not be achieved experimentally (Andrews and Wood. 1967, Ford and Haskin 1982, Haskin and Andrews 1988). These obser- vations, along with the early failure to find spores, led to specu- lation that another host was involved in the life cycle, although a candidate species has yet to be identified. In the model, transmis- sion in H. nelsoni includes non-local factors that exert a paramount influence on the transmission process. In fact, population- dependent processes are not sufficiently important to even be in- cluded in the model, save for filtration rate that, with infective particle concentration, determines whether an individual animal receives an infective dose. The resulting model is unusually com- plex in comparison to other disease models. It is the source of this complexity and the implication of another reservoir of infective particles besides the oyster that is intriguing, because the full life cycle of H. nelsoni remains unknown and. in particular, present knowledge effectively provides no details of the transmission pro- cess beyond those that can be gleaned by observing temporal and spatial changes in prevalence. Although the simulations provide good approximations to field observations of H. nelsoni prevalence, a few trends are not as faithfully rendered as one might wish. Examples include the base case for Tangier Sound-Old Woman's Leg (Fig. 15f) which over- estimates prevalence and the upper bay cases for the decade of the 1980s in Delaware Bay where prevalence is underestimated (Fig. 14). Results presented in this contribution and in two companion papers (Ford et al. this volume, Paraso et al. this volume) show that the time-development of H. nelsoni proliferation is very sensitive to salinity, temperature, and food supply. In no case were we able to utilize a full suite of environmental data and observations of disease prevalence obtained from the same location at the same time. Soniat et al. ( 1998) provide a good example of the sensitivity of oyster population dynamics models to the frequency of collec- tion of environmental data. It is likely that one important source of the discrepancies that remain between field observations and model simulations is the inability of the presently-available envi- ronmental time series to fully describe the temporal dynamics of the disease process. Forcing Factors in the Transmission Model The transmission model has been constructed with little infor- mation on the transmission process itself beyond that gleaned from observations of the temporal progression of infection in host popu- lations and the correlation of this progression with environmental variables such as temperature and salinity. Nevertheless, the mod- eling process permits identification of key environmental forcing factors that can be expected to exert a direct control on the trans- mission process, either by controlling the availability of infective particles in the water column or by controlling the population dynamics of an alternate host. These forcing factors include sea- son, salinity, and winter temperature. Filtration Rate and Season The form of the infective stage is unknown, but is assumed to be waterborne, at least for a short period before it infects, because the earliest infections are observed in the gill epithelium where they presumably lodged as a consequence of filter feeding. Fur- ther, timed transplants of uninfected oysters into enzootic areas showed that they became infected only during a period from late May through October (Ford and Tripp 1996). Lack of additional information required that assumptions be made about the quantity and availability of infective stages. The model places infective stages in the water; they are not produced by infected oysters. In contrast, the density and infection level of oysters is a key element in the modeling of P. marinus transmission. The decoupling reflects the observed evidence that new infections are acquired independently of the infected oyster population. Based on a series of simulations compared to field observations of prevalence, an infective dose of 8.700 particles filtered d~' was set as the threshold needed to initiate an infection. Compared to P. marinus. which can initiate an infection with as few as 10-100 parasites (Valiulis 1973. Chu and Volety 1997), this dose appears to be very high. The P. marinus experiments, how- ever, were performed by injecting parasites directly into the gill Modeling MSX Transmission in Oysters 533 cavity, a method that is considerably more effective than allowing the oysters to filter P. marinus from the water as they do naturally (Chu 1996, Bushek et al. 1997). Recently. Li et al. (1993) reported levels of free Perkinsus spp. cells in estuarine water samples as high as 19.000 cells L"1. A 75-100-mm oyster filtering 133 L of water in a day (Powell et al. 1992), would encounter as many as 2,500,000 cells. Based on the results of Chu and Volety (1997) and Valiulis ( 1973) described above, it would seem impossible for any oysters in enzootic waters to be alive if all cells were infective. Ford et al. ( 1996) fed oysters a dose of 106 P. marinus g wet wt"1 and did succeed in infecting all animals, but mortalities and infec- tion intensities were very low through the 12-week experiment compared to groups injected with parasites, suggesting that a large fraction of parasites filtered from the water never cause infection. Thus. 8,700 infective stages filtered d"' for H. nelsoni is within the bounds set by observations. Using the above infective dose, simulations were run with a variety of infective particle concentrations and a value of 900 particles L_1 was found to provide the best results for the decade of the 1960s. Because no measurements of the abundance of in- fective stages in the water have been made, this figure cannot be verified, but it compares well with the 500-800 haplosporidian spores LT1 estimated to be present in Delaware Bay based on evidence from spores that oysters had ingested through feeding (Barber and Ford 1992). Although these spores are identical, at the light microscoope level, to those of H. nelsoni, their identity has not been confirmed. Nevertheless, because both these spores and H. nelsoni are abundant in Delaware Bay and because the temporal presence of the spores matched that of the known infective period for H. nelsoni. that they are, in fact. H. nelsoni must be seriously considered. Because of this strong circumstantial evidence linking the two, the seasonal abundance of ingested spores was used to vary the seasonal concentrations of H. nelsoni infective particles in the model. The fact that simulations subsequently reproduced the timing and degree of infection in oysters would seem to lend credence to the argument that these are H. nelsoni and that they are infective to the oyster. Nevertheless, some caution must be used in interpreting these results as correlation analysis failed to link in- terannual fluctuations in ingested spore abundance with subse- quent H. nelsoni infection levels, a finding that lends support to the secondary host hypothesis (Barber and Ford 1992). Salinity Salinity exerts the overriding influence on the transmission process and its role is a dual one. Initial simulations of prevalence and intensity at low salinity showed the appropriate timing, but produced prevalences that were higher than observed (Paraso et al. this volume). Decreasing the concentration of infective particles in low salinity water improved the model fit. Thus, salinity, in the model, acts on the local host population by varying the infectivity of infective particles as they impinge the oyster gill during the filtration process. The model makes no biological distinction between whether this salinity effect produces fewer infective stages or lower infec- tion efficiency of an undiminished number. In their report of in- gested H. nelsoni-Wke spores in Delaware Bay. Barber and Ford (1992) noted that the spores were abundant in the low salinity regions of Delaware Bay where infection prevalence is typically low. They postulated that salinity might affect excystation of spores, and consequently infection success, if these are indeed a life stage of W. nelsoni and responsible for infections in oysters. It is likely that a spore stage is involved in the transmission of H. nelsoni to oysters, even if it is not the one found by Barber and Ford (1992). This life stage should be more resistant to salinity change than the Plasmodium and it could be quite long-lived and widely dispersed in the environment. However, the sporoplasm. no longer protected within the spore wall, would be immediately ex- posed to ambient salinity during excystation and, assuming that it has the same salinity tolerance as plasmodial stages, would quickly be killed at low salinity. In addition, salinity exerts a regional influence on the trans- mission process by controlling, in part and on a bay-wide scale, the concentration of infective particles in the water column (or perhaps the abundance of an alternate host). This control on concentration must be distinguished from the local control on infectivity that may. in fact, be a local reduction in infective particle concentration due to cell mortality. The bay-wide control mechanism permits the concentration of infective particles to increase during high salinity times and decrease during low-salinity times, at all sites, regardless of the local salinity regime. Although the process has been mod- eled in a somewhat ad hoe fashion because of insufficient data, the process probably originates from the variation in residence time of bay waters. Very likely, as flushing rate declines with increased salinity, the concentration of infective particles increases. Temperature An intriguing observation from long-term monitoring of H. nelsoni in lower Delaware Bay was that infection prevalence showed distinct interannual cycles (Ford and Haskin 1982, Haskin and Andrews 1988). Years of low prevalence followed, by 1 or 2 years, a very cold winter. Between the onset of the MSX epizootic in Delaware Bay in 1957 and 1990, this pattern was repeated half a dozen times. The model was fit to these observations by decreas- ing the abundance of infective stages according to the "coldness" (measured in degree days) of the preceding two winters. Infective particle concentration decreases for 1 to 2 years after a cold winter. It returns to high levels faster after a warm winter. Simulations that included the cold-winter effect clearly repli- cated the cyclic prevalence pattern in Delaware Bay. The model also reproduced a post-cold-winter prevalence dip that appeared in a 1989 to 1994 data series from the Maryland portion of Chesa- peake Bay. indicating that neither the model nor the cold-winter effect is restricted to a single location. How widespread the effect is remains uncertain, however, long-term (30-year) comparisons of prevalence cycles in lower Delaware Bay and the lower York River near its juncture with the Virginia portion of Chesapeake Bay, revealed dissimilar patterns (Haskin and Andrews 1988). The cold-winter effect that was obvious in the Delaware Bay location was not obvious at the Virginia location. Nevertheless. Haskin and Andrews (1988) pointed out hints of the same pattern in the York River, where it may have been masked by the more pronounced salinity effect. Again, the modeling device says nothing about the biological mechanism behind the observation, which has been speculated to be an effect on a secondary host (Ford and Haskin 1982, Haskin and Andrews 1988). Local versus Bay-wide Processes Temperature, salinity (residence time effect), and season exert their influence over all sites regardless of the local salinity regime. It is the presence in the H. nelsoni transmission model of these 534 Powell et al. bay-wide influences of salinity, temperature and season that make this model different from most other transmission models and remarkably different from the other model most pertinent to oyster population dynamics, the model for P. marinus. In the latter case, temperature and salinity affect transmission indirectly by varying the infection intensity in the infected portion of the host popula- tion. Despite the influence of bay-wide processes, which might be expected to carry with them some unique attributes of individual bays, particularly considering bays as different as Chesapeake Bay and Delaware Bay. bay-to-bay differences in the transmission model were minor. The only variable that varied between bays was DD0, the variable defining a cold winter. Cold, as defined in the model, required more degree-days below 10 °C in Delaware Bay than in Chesapeake Bay. The possibility that this discrepancy originates in the acquisition of environmental time series and field observations of H. nelsoni prevalence from different locations can- not be discounted. Thus, this modeling exercise strongly supports the belief that an underlying basic process common throughout the disease range is responsible for transmission of H. nelsoni in oyster populations and that this process is dominantly one influenced by non-local, non-oyster-population-dependent processes. Multiyear-dependent Changes in the Infection Process Temporal changes in disease virulence and host resistance to disease are well-described and modeled (Lenski and May 1994. Frank 1996) and some evidence exists for geographic variation in virulence and resistance in another oyster disease-causing organ- ism, P. marinus (Bushek and Allen 1996). Early in the develop- ment of the first epizootics of H. nelsoni, after its introduction in the late 1950s, it was observed that oyster populations seemed to build up resistance to the disease over the first few years of ex- posure (Haskin and Ford 1979). Controlled breeding has clearly produced oysters distinctly more resistant to H. nelsoni disease than native populations (Haskin and Ford 1979, Ford and Haskin 1987). Accordingly, some evidence of long-term changes in the infectivity of H. nelsoni might be sought through this modeling exercise. In fact, such a search developed little supporting evidence. Simulations of disease-resistant oysters did not fit field observa- tions in either bay. The same relative susceptibility to disease proved adequate to simulate the 1960s in Delaware Bay and the early 1990s in Chesapeake Bay. However, an increase in infective particle concentration was required in Delaware Bay to adequately simulate the decade of the 1980s. The suggestion that this neces- sity might be caused by a reduction in infective dose (increased virulence or decreased susceptibility) is not supported by any other information. The possibility that the necessity originates from in- adequate field control on environmental time series cannot be dis- counted, but does not fit all nuances of the differences between field observation and simulation. A possibility that the concentra- tion of infective particles has built up in the Bay over the decades remains a viable explanation, and this might be the result of in- creased abundance of an alternate host or increased capacity of some other reservoir. The modeling exercise clearly supports the requirement for the existence of a reservoir, alternate host or oth- erwise, independent of the oyster population (Ford and Tripp 1996). The Influence of Transmission in System Memory In the P. marinus model, the influence of the infection inten- sification cycle overwhelms the influence of transmission. Most epizootics are triggered by processes influencing the degree to which infected individuals retain significant parasite burdens over the winter (Powell et al. 1996) or processes permitting an increase in P. marinus growth relative to oyster growth. The P. marinus model is somewhat unusual in this regard. Transmission is nor- mally an overwhelming component of the disease process in most animal diseases (e.g., Ackerman et al. 1984, Anderson 1991 ). In this limited perspective, the H. nelsoni model is more typi- cal, in that transmission has an overwhelming influence on the course of infection in oyster populations. Epizootic cycles seem to be principally the product of enhanced transmission rather than enhanced intensification. These influences of transmission on the course of infection, in many cases, have multiyear implications for prevalence and infection intensity, and the root of much of this multiyear behavior is in the processes that control the concentra- tion of infective particles in the water column (or reservoir). It is typical of population dynamics models for the system modeled to demonstrate a multiyear system memory in which the present state is a product not just of the present or immediately past environ- mental conditions, but also of the environmental conditions inte- grated over a much longer period of time. Oyster population dy- namics models are particularly good examples (e.g.. Soniat et al. 1998, Powell et al. 1996). System memory is less pervasive in the H. nelsoni model than in the P. marinus model. The course of epizootics and the periodic near-disappearance of the disease origi- nate through processes that have commenced within the year or, for cold winters. 1 or 2 years previously. In this extent. H. nelsoni is not typical of an epizootic-producing organism (e.g.. Gill 1928) and it is likely that this atypical nature is due to the limited time- dependent description required of the alternate reservoir for infec- tive particles. Either the reservoir is a large, persistent environ- mentally buffered one that does not involve an alternate host or involves a large long-lived animal or the alternate host has a 1-to- 2-yr life cycle that allows it to rapidly respond to environmental change. The volatility of these latter species, however, would seem to preclude their involvement. In either case, it seems unlikely that any significant feedback between the oyster and alternate reservoir can exist. Two interesting peculiarities can be considered with respect to the influence of large scale processes in H. nelsoni transmission. The first of these is the typical condition of H. nelsoni disease south of Cape Hatteras in which low H. nelsoni prevalence, but high infection intensity in infected animals, is characteristic of the disease process (e.g.. Lewis et al. 1992). This event is easily re- created in the model by establishing a low concentration of infec- tive particles, a condition very similar to the low-salinity cases from Delaware Bay where both simulation and observation show low prevalence, but high infection intensity in 50% or more of the infected population. Second, most observational time series show periods when H. nelsoni more or less completely disappears from large areas of the bay. Examples include the terminal years of the 1960s and 1980s on the Delaware Bay seed beds (e.g.. Fig. 2a-d). These events are faithfully reproduced in the model. They originate in processes set in motion 1 to 2 years earlier that gradually lead to the reduction in infective particle concentration and, thus, reduced transmission. They do not originate in processes resulting in the loss of patent Modeling MSX Transmission in Oysters 535 infections in the population produced by simultaneously occurring environmental conditions, although, of course, this must occur. It is particularly interesting, in Delaware Bay, that the decline of H. nelsoni in the latter part of the 1980s occurs even without the introduction of P. marinus. The coincidental change in the domi- nance of these two disease-causing organisms around 1990 in this bay has received some notoriety. It is significant, then, that the model indicates that processes set in motion in the late- 1980s were responsible for the disappearance of H. nelsoni at the end of the decade. The failure of H. nelsoni to return in the 1990s, however, may still indicate a significant interaction between the two dis- eases. This possibility cannot be investigated with the present model. The model suggests several processes that conspire to initiate and stifle epizootics, all related to changes in infective particle concentration. These include the presence of unusually warm or cold winters that would initiate and stifle, respectively, epizootics several years hence. Similarly, wet and dry years initiate the de- crease and increase, respectively, in concentration of infective par- ticles that then reduce and increase, respectively, the prevalence of disease. This process has a response time of about one-half year. The model strongly points to the need to evaluate the past several years of history of environmental change in understanding the process of epizootic initiation and cessation. The Intermediate Host Hypothesis The transmission model requires a reservoir of infective par- ticles independent of the host oyster population that is modified according to multiyear variations in temperature and salinity. Sev- eral components of the model lend credence to the hypothesis that H. nelsoni has an alternate host in the life cycle, although no such host has yet been identified (Burreson 1988, Haskin and Andrews 1988). These components include ( 1 ) the importance of bay-wide processes in transmission that impose a control on prevalence far beyond that exerted by the local population, including spore pro- duction in young oysters. This modeling exercise clearly indicates the importance of a time-varying concentration of infective par- ticles that varies according to change in bay-wide salinity (wet versus dry conditions) and temperature (warm versus cold win- ters). (2) With the exception of local effects on infectivity as the expected by-product of the influence of low salinity on what is essentially a polyhaline disease organism, local effects do not oc- cur in the transmission model. Oyster density, for example, plays no role in the transmission model. (3) The seasonal cycle of in- fection is an important component of the transmission model. However, the increase in infective particle concentration precedes the formation of any spores within the simulated oyster population (e.g.. Ford et al. this volume). (4) A contact rate model fails be- cause high prevalence and infection intensity can occur in host populations when infectivity is low and such a mismatch between the number of infected individuals and the rate of transmission cannot occur in a local-contact transmission model. (5) Cold win- ters significantly affect transmission rate, probably by reducing infective particle concentration, yet cold winters have a much less immediate effect on population infection intensity (Ford and Haskin 1982, Ford et al. this volume). All of these aspects of the transmission model support the alternate host hypothesis. What are the characteristics of the alternate reservoir? ( 1 ) It must be capable of rapid, continuous release of large numbers of infective particles into the water column, at least during the warmer months of the year. (2) It must be environmentally buff- ered to a significant extent, so that normal variations in tempera- ture and salinity do not affect its capacity for infective element generation. (3) It must be independent of infection intensity of the oyster population, at least over significant time periods, which suggests that the reservoir either is capable of supporting the full life cycle of H. nelsoni or that the reservoir has a multiyear ca- pacity under most conditions. (4) It must be sensitive to cold temperatures, but be capable of recovery over a 2-year period independent of the infection level in the oyster population. This sensitivity may be salinity-dependent. (5) It must exist at relatively high salinity to provide the observed response to salinity-driven changes in flushing rate. The Epidemiological Approach The transmission model points the way to epidemiological studies that might further elucidate the transmission cycle of H. nelsoni without requiring identification of the infective particle reservoir. That stumbling block has significantly impaired inves- tigation of the transmission process. Investigation of the multiyear influence of temperature and salinity, for example, would be ame- nable to standard epidemiological approaches. This modeling ex- ercise, however, emphasizes the need for such studies to extend over wide areas and significant time periods. Single-site studies and studies of 1 or 2 years duration are unlikely to significantly advance the understanding of transmission in this oyster disease. Formulation of the model, in fact, required the use of decadal time series. Clearly, large areas of bays are responding similarly and relatively synchronously and over time scales of several years or more. Understanding the generation of H. nelsoni epizootics will require epidemiological studies of similar scale. The transmission model has been run in hindcasting mode for this contribution. Hindcasting was necessary to verify the model. Good agreement of model simulations with field observations over broad expanses of the salinity gradient and between bays suggests that the model should now be tested in nowcasting mode. Devel- oping nowcasting. and then forecasting, capability will require a thorough understanding of environmental change, particularly near-real-time information on salinity, temperature, particulate load, and food supply. These data must be obtained in or very near the target population. Both salinity and temperature parameteriza- tion include highly non-linear equations that describe threshold responses in the oyster population. Accordingly, under certain cir- cumstances, small changes in temperature and salinity, typical of cross-estuary and depth gradients, will be significant and this puts a premium on the collection of data very close to the target popu- lation. ACKNOWLEDGMENTS This research was supported by the Virginia Graduate Marine Science Consortium grant VGMSC 5-29222 and by New Jer- sey Sea Grant under contract number 4-25238. Computer re- sources and facilities were provided by the Center for Coastal Physical Oceanography at Old Dominion University. The Dela- ware River and Bay Authority funded the 1981-1984 moni- toring program that provided data for some of the environmental time series. Continuation of the time series through 1986 was made possible by funds from the New Jersey Department of Environ- mental Protection. Both programs were coordinated by Walt Can- zonier. The States of New Jersey and Maryland provided funds for 536 Powell et al. collection of the Delaware Bay and Chesapeake Bay Haplospo- ridium nelsoni time series. Thanks to the staffs of the Haskin Shellfish Research Laboratory, the Maryland Department of Natu- ral Resources Shellfish Program, and the Sarbanes Cooperative Oxford Laboratory. Chesapeake Bay MSX diagnoses were done by Sara Otto. This is Contribution #99- 1 8 of the Institute of Marine Science at Rutgers University and New Jersey Agricultural Ex- periment Station Publication #D-32405-3-99. LITERATURE CITED Ackerman. E., L. R. Elvehack & J. P. Fox. 1984. Simulation of Infectious Disease Epidemics. Charles C. Thomas, Springfield. 202 pp. Anderson, R. M. 1991. Discussion: the Kermack-McKendrick epidemic threshold theorem. Bull. Math. Biol. 53:3-32. Andrews. J. D. 1964. Oyster mortality studies in Virginia IV. MSX in James River public seed beds. Proc. Null. Shell full. Assoc. 53:65-84. Andrews, J. D. 1965. Infection experiments in nature with Dermocystidium marinum in Chesapeake Bay. Chesapeake Set. 6:60-67. Andrews, J. D. 1968. Oyster mortality studies in Virginia. VII. Review of epizootiology and origin of Minchinia nelsoni. Proc. Natl. Shellfish. Assoc. 58:23-36. Andrews, J. D. 1988. Epizootiology of the disease caused by the oyster pathogen P. marinus and its effects on the oysler industry. Am, Fish. Soc. Spec. Publ. 18:47-63. Andrews, J. D. & S. M. Ray. 1988. Management strategies to control the disease caused by P. marinus. Am. Fish. Soc. Spec. Publ. 18:257-264. 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. Chesapeake Sci. 8:1-13. Barber, R. D. & S. E. Ford. 1992. Occurrence and significance of ingested baplosporidian spores in the Easter oyster. Crassoslrea virginica (Gme- lin, 1791). J. Shellfish Res. 11:371-375. Burreson, E. M. 1988. Use of immunoassays in haplosporidan life cycle studies, pp. 298-303. In: W. S. Fisher, (ed.). Disease Processes in Marine Bivalve Molluscs. 18. American Fisheries Society. Bethesda. MD. Burreson. E. M. 1991. Effects of Perkinsus marinus infection in the eastern oyster, Crassoslrea virginica: I. Susceptibility of native and MSX- resistant stocks. /. Shellfish Res. 10:417-423. Bushek, D. & S. K. Allen Jr. 1996. Races of Perkinsus marinus. J. Shellfish Res. 15:103-107. Bushek, D.. S. K. Allen, K. A. Alcox. R. Gustafson & S. E. Ford. 1997. Response of Crassoslrea virginica to in vitro cultured Perkinsus mari- nus: preliminary comparison of three inoculation methods. J. Shellfish Res. 16:479-485. 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. Crassoslrea virginica. Aram. Rev. Fish Dis. 4:201-217. Chesson, P. L. & W. W. Murdoch. 1986. Aggregation of risk: relationships among host-parasitoid models. Am. Nat. 127:696-715. Chintala, M. M. & W. S. Fisher. 1991. Disease incidence and potential mechanisms of defense for MSX-resistant and -susceptible Eastern oysters held in Chesapeake Bay. J. Shellfish Res. 10:439^143. 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. Quan- tification of the thioglycollate method. J. Shellfish Res. 8:1 17-125. Chu, F.-L. E. 1996. Laboratory investigations of susceptibility, infectivity, and transmission of Perkinsus marinus in oysters. J. Shellfish Res. 15:57-66. Chu, F.-L. E. & A. K. Volety. 1997. Disease processes of the parasite Perkinsus marinus in eastern oyster Crassoslrea virginica: minimum dose for infection, initiation, and interaction of temperature, salinity, and infective cell dose. Dis. Aqual. Org. 28:61-68. Couch. J. A.. C. A. Farley & A. Rosenfield. 1966. Sporulation of Min- chinia nelsoni (Haplosporida. Haplosporidiidae) in Crassostrea vir- ginica (Gmelin). Science 153:1529-1531. Dietz, K. 1982. Overall population patterns in Ihe transmission cycle of infectious disease agents, pp. 87-102. In: R. M. Anderson and R. M. May (eds.). Population biology of infectious diseases. Dahlem Konfer- enzen 1982. Springer- Verlag. Berlin. Dwyer, G. & J. S. Elkinton. 1993. Using simple models to predict virus epizootics in gypsy moth populations. / Anim. Ecol. 62:1-1 1. Farley, C. A. 1968. Minchinia nelsoni (Haplosporida) disease syndrome in the American oyster Crassostrea virginica. J. Protozoal. 15:585-599. Farley, C. A. 1975. Epizootic and enzootic aspects of Minchinia nelsoni (Haplosporida) disease in Maryland oysters. J. Protozoal. 22:418^127. Fegley, S. R., S. E. Ford. J. N. Kraeuter & D. R. Jones. 1994. Relative effects of harvest pressure and disease mortality on the population dynamics of the Eastern oyster (Crassostrea virginica) in Delaware Bay. Final Report to NOAA. Grant No. NA26FL0388. 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 1:539- 546. Ford. S. E. & H. H. Haskin. 1982. History and epizootiology of Haplospo- ridium nelsoni (MSX), an oyster pathogen, in Delaware Bay. 1957- 1980. J. Invertebr. Pathol. 40:1 18-141. Ford. S. E. & H. H. Haskin. 1987. Infection and mortality patterns in strains of oysters Crassoslrea virginica selected for resistance to the parasite Haplosporidium nelsoni (MSX). J. Parasitol. 73:368-376. Ford. S. E. & S. A. Kanaley. 1988. An evaluation of hemolymph diagnosis for detection of the oyster parasite Haplosporidium nelsoni (MSX). J. Shellfish Res 7:11-18. Ford, S. E„ E. N. Powell. J. M. Klinck and E. E. Hofmann. 1999. Modeling the MSX parasite in Eastern oyster ( Crassostrea virginica) populations. I. Model development, implementation and verification. J. Shellfish Res. 18:473-498 Ford, S. E. & M. R. Tripp. 1996. Diseases and defense mechanisms, pp. 383^*50. In: R. I. E. Newell. V. S. Kennedy & A. F. Eble (eds). The Eastern oyster Crassostrea virginica. Maryland Sea Grant College. College Park. Maryland. Frank. S. A. 1996. Models of parasite virulence. Quart. Rev. Biol. 71:37- 78. Garvine. R. W.. R. K. McCarthy & K-C. Wong. 1992. The axial salinity distribution in the Delaware Estuary and its weak response to river discharge. Esluarine Coastal Shelf Sci. 35:157-165. Gill. C. A. 1928. The genesis of epidemics and the natural history of disease. William Wood, New York. 550 pp. Godfrey, H. C. J. & C. J. Briggs. 1995. The population dynamics of patho- gens that control insect outbreaks. J. Theor. Biol. 176:125-136. Haskin, H. H. 1972. Delaware River flow-bay salinity relationships. Report to the Delaware River Basin Commission: Phase III. 12 pp. Haskin. 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.). Disease Processes in Marine Bivalve Molluscs. 18, American Fisheries Society, Bethesda, MD. Haskin, H. H. & S. E. Ford. 1979. Development of resistance to Minchinia 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. 1982. Haplosporidium nelsoni (MSX) on Delaware Bay seed oyster beds: a host-parasite relationship along a salinity gradient. J. Invertebr. Pathol. 40:388^105. Hassell. M. P. 1982. Patterns of parasitism by insect parasitoids in patchy environments. Ecol. Entomol. 7:365-377. Hethcote. H. W. & P. van den Driessche. 1995. An SIS epidemic model with variable population size and a delay. J. Math. Biol. 34:177-194. Hofmann. E. E„ J. M. Klinck. E. N. Powell. S. Boyles & M. Ellis. 1994. Modeling MSX Transmission in Oysters 537 Modeling oyster populations II. Adult si/e and reproductive effort. J. Shellfish Res. 13:165-182. Hofmann. E. E., E. N. Powell. J. M. Klinek & G. Saunders. 1995. Model- ing diseased oyster populations I. Modelling Perkinsus minimis infec- tions in oysters. /. Shellfish Res. 14:121-151. Hofmann. E. E., E. N. Powell. J. M. Klinek & E. A. Wilson. 1992. Mod- eling oyster populations. III. Critical feeding periods, growth and re- production. J. Shellfish Res. 11:399-416. Kermack, W. O. & A. G. McKendrick. 1991. Contributions to the math- ematical theory of epidemics-I. Bull. Math. Biol. 53:33-55. Lenski, R. & R. M. May. 1994. The evolution of virulence in parasites and pathogens: reconciliation between two competing hypotheses. J. Theor. Biol. 169:253-265. Lewis, E. J., F. G. Kern, A. Rosenfield, S. A. Stevens. R. L. Walker & P. B. Heffernan. 1992. Lethal parasites in oysters from coastal Georgia with discussion of disease and management implications. Mar. Fish. Rev. 54(21:1-6. Li, T., B. S. Roberson & C. F. Dungan. 1994. Interannual abundance varia- tion of environmental Perkinsus marinus cells in Chesapeake Bay wa- ters. J. Shellfish Res. 13:296-297. Loosanoff. V. L. 1962. Effects of turbidity on some larval and adult bi- valves. Proc. GulfCaribb. Fish. Inst. 80-95. Mackin, J. G. 1952. Incidence of infection of oysters by Dermocystidium in the Barataria Bay area of Louisiana. Natl. Shellfish Assoc. Com: Add. for 1951. pp. 22-35. Marshall, H. G. & R. W. Alden III. 1997. Dynamics of an estuanne eco- system: the influence of flow patterns on phytoplankton trends in the Chesapeake Bay. Oceanologica Acta 20:109-1 17. Mollison D. 1987. Population dynamics of mammalian diseases. Symp. Zool. Soc. bond. 58:329-342. Myers, J. H. 1993. Population outbreaks in forest Lepidoptera. Am. Sci. 81:240-251. Ostfeld. R. S. 1997. The ecology of lyme-disease risk. Am. Sci. 85:338- 346. Paraso, M. C. S. E. Ford. E. N. Powell, E. E. Hofmann & J. M. Klinek. 1999. Modeling the MSX parasite in eastern oyster {Crassostrea vir- ginica) populations. II. Salinity effects. J. Shellfish Res. 18:499-514 Powell, E. N., E. E. Hofmann. J. M. Klinek. & S. M. Ray. 1992. Modeling oyster populations I. A commentary on filtration rate. Is faster always better? J. Shellfish Res. 1 1:387-398. Powell, E. N., J. M. Klinek & E. E. Hofmann. 1996. Modeling diseased oyster populations. II. Triggering mechanisms for Perkinsus marinus epizootics. J. Shellfish Res. 15:141-165. Powell, E. N„ J. M. Klinek, E. E. Hofmann & S. Ford. 1997. Varying the timing of oyster transplant: implications for management from simu- lation studies. Fish. Oceanogr. 6:213-237. Smith, G. F. 1997. Maryland's historic oyster bottom. A geographic rep- resentation of the traditional named oyster bars. Maryland Dep Nat Res, Oxford. MD. 16 pp. 5 app. Soniat, T. M.. E. N. Powell. E. E. Hofmann & J. M. Klinek. 1998. Under- standing the success and failure of oyster populations: the importance of sampled variables and sample timing. / Shellfish Res. 17:1149- 1165. Stokes, N. A.. M. E. Siddall & E. M. Burreson. 1995. Detection of Hap- losporidium nelsoni (Haplosporidia: Haplosporidiidae) in oysters by PCR amplification. Dis. Aquat. Org. 23:145-152. Valiulis. G. A. 1973. Comparison of the resistance to Labyrinthomyxa marina with resistance to Minchinia nelsoni in Crassostrea virginica. Ph.D. dissertation. Rutgers University, New Brunswick, NI 180 pp. Wong. K-C. 1994. On the nature of transverse variability in a coastal plain estuary. J. Geophys. Res. 99:14209-14222. Woolhouse, M. E. J. & S. K. Chandiwana. 1990. Temporal patterns in the epidemiology of schistosome infections of snails: a model for field data. Parasitology 100:247-253. Journal of Shellfish Research, Vol. 18. No. 2. 534-546. 1999. REPRODUCTIVE CYCLES OF THE SURF BEACH CLAM PAPHIES DONACINA (SPENGLER, 1793) FROM NEW ZEALAND ISLAY D. MARSDEN Zoology Department University of Canterbury Christchurch New Zealand ABSTRACT The gametogenic cycle of the tuatua Paphles donacina is described from histological sections of clams collected monthly from November 1993 to April 1997. Trends in the dry weight condition index (CI) and gonad indices suggest an annual breeding cycle with continuous spawning over the summer and an inactive winter phase. There was considerable interannual variation with less than 50% of the population being reproductive during the 1994/5 season. There were significant correlations between seawater temperature, the population gonad index and gonosomatic indices of male and female tuatua. Food resources (chlorophyll a levels) fluctuated within and between seasons with values from 4 to 12 (j.g IT1. These values showed a positive correlation with seawater temperature, but not with any of the bivalve indices. Throughout the study, the population sex ratio was biased toward females, and often there were individuals of indeterminate sex within the population. Dry weight values for standard 40-mm and 90-mm length tuatua varied seasonally, with the smaller length group showing a higher coefficient of variation, consistent with seasonal patterns of somatic and gonad development. Both male and female tuatua show greater reproductive activity early rather than late in the breeding period and the proportion of reproductive tissue was similar for males and females of various lengths. The CI was not a good indicator of reproductive potential in male or female tuatua. The reproductive strategy of P. donacina is regarded as exploitive, closely determined by environmental seawater temperature. Somatic growth and gonad development occur simultaneously over the summer, but can extend into the winter if sufficient food resources are available. KEY WORDS: bivalve. New Zealand. Paphies donacina, reproduction, gametogenic cycle, gonosomatic index, condition index INTRODUCTION Surf clams form an important world-wide recreational and commercial shellfishery resource (McLachlan et al. 1996). and in New Zealand, several species have been identified as providing potential for aquaculture. These include two species of tuatua of the family Mesodesmatidae: Paphies subtriangulata, distributed mainly around the North Island, and P. donacina, the dominant tuatua around the South Island. It is optimally distributed from the low-tide down to about 3-m depth (Cranfield et al. 1994. Haddon et al. 1996). The habitat of P. donacina within Pegasus Bay includes high wave action and well-sorted fine sand, resulting in a low beach profile and cyclic changes in beach morphology (Kirk 1980). The exposed coastline, with its highly dynamic meso-tidal surf beach, is classified within the range intermediate to dissipative on a world-wide scale (Short 1990). Bivalves within the surf zone typi- cally are mobile and exposed to a shifting habitat with variable conditions of temperature, wave exposure, moisture levels, food supply, and salinity extremes (McLachlan and Erasmus 1983). Zonation patterns on exposed sand beaches are often indistinct, and environmental conditions are more extreme than those found in shallow estuaries or on rocky shores (Widdows 1978. Hummel 1985, McLachlan and Jaramillo 1995. Urrutia et al. 1997) Along with the two species of tuatua in New Zealand, the superfamily Mactridae also includes the toheroa. Paphies ventri- cosa, the pipi, P. australis, Mactra murchisoni, and Spisula aequi- latera (Redfearn 1974. Hooker and Creese 1995a, Hooker and Creese 1995b, Grant and Creese 1995, King 1997). The reproduc- tive cycle has been described for all of the above species, but there are no comparable data for P. donacina. McLachlan et al. ( 1996) have reviewed the fisheries biology of surf clams, and report highly variable reproductive patterns both within and between spe- cies. Although there have been numerous studies on the reproductive biology of a wide range of bivalves, including mussels, clams, and scallops, many of these have followed reproductive events over one or two annual cycles (Kanti et al. 1993, Garcia-Dominguez et al. 1994). Although some studies, including that on the sea scallop (Dibacco et al. 1995), include results from up to 7 years, it has not been possible to relate reproductive events to specific environmen- tal conditions. There is also continued debate about the importance of temperature and food resources in influencing the reproductive cycle of bivalves (Manzi et al. 1985, Garcia-Dominguez et al. 1994, Urrutia et al. 1997). The aims of the present study were to characterize the repro- ductive cycle of low-tide P. donacina and investigate the relation- ship between the gametogenic cycle and bivalve condition. This aspect has not previously been addressed in surf clams and is of interest in fisheries management. Another goal was to investigate the importance of temperature and potential food resources in the reproductive activity of the tuatua within the surf-zone habitat. For this reason the study was undertaken at approximately monthly intervals over a period of 3.5 years. MATERIALS AND METHODS Sampling Tuatua were opportunistically sampled along a 2-km section of South Brighton Beach within Pegasus Bay, New Zealand (43°33'S. 172°55'E) during extreme low-water spring tides be- tween November 1993 and April 1997. They were hand collected by detecting a hydroid on the sand surface. This hydroid attaches to the posterior part of the bivalve shell. Each month, during ex- treme low-water spring tides, at least 40 tuatua within the size range of the population were collected. They were returned to the laboratory within an hour and processed after cooling to 5 °C. This allowed the shell to be opened easily and the tissues extracted with 539 540 Marsden minimal damage. There were two occasions (July 1994 and May 1996) when tidal and weather conditions prevented sampling due to onshore winds or exceptionally rough sea conditions. At each sampling time, the low-tide water temperature and air temperature were recorded and a sample of seawater was returned to the labo- ratory, where estimates were made for chlorophyll a using a Turner Fluorimeter calibrated using chlorophyll standards. The tidal range in the habitat was 2 to 2.5 m, and the average seasonal air temperature range 6 to 17.5 °C, measured at the nearby Avon- Heathcote Estuary (Estcourt 1962). Weight Relationships The shell length and width of 20 tuatua representing the popu- lation length range were measured using vernier callipers and the total wet body weight recorded after cutting the adductor muscle and draining the shell fluid. The tissue was removed from the shell and the wet weight recorded before drying at 60 °C for 3 days. This allowed calculation of a dry weight/wet weight ratio to estimate the dry weight of tissue from samples used in the histological study. The shell of each bivalve was dried at 60 °C and the physi- ological condition index (CI) (Lucas and Beninger 1985) was cal- culated for each individual used in the population survey and the histological study. Histology For each sample, the length and width of 20 tuatua greater than 45-mm shell length were processed for the histological study. Bi- valves were measured and the whole-body wet weight recorded before fixing in Davidson's fluid (Morales-Alamo and Mann 1989). The sex of the tuatua could not be determined by eye, except for a few instances late in the breeding season when mature eggs were observed. The gonad of the tuatua is located along the posterior margin of the foot, and extends into the digestive diver- ticula. A standard cross section of approximately 5 mm was cut above the base of the foot, as in Hooker and Creese ( 1995a, 1995b) and Grant and Creese (1995). These cross sections were dehy- drated in an ethanol series, blocked in paraffin wax, then sectioned at 7-9 u.m. Two slides, each with 2 sections, were made from most blocks, and the slides were stained using Ehrlich's haematoxylin counterstained with eosin. Gonad Indices The slides were first examined under low power (x 40) to determine the presence of reproductive tissue. On one section ot each slide, the maximum percentage of the cross section occupied by the gonad was measured (gonosomatic index). The sections were then examined using a high power binocular microscope (x 200) to determine the stages of development, as in Redfearn (1974) and Grant and Creese (1995). For female clams, these stages were: early active follicles, mixed follicles, mature follicles, and spent reproductive or reabsorbing gonad tissue. In males only early and mature follicles could be distinguished. The population gonad index was calculated as the percentage of reproductive in- dividuals in each sample. To investigate the effects of body size on reproductive tissue mass of male and female tuatua. the gonosomatic index was plot- ted against shell length, dry weight, and condition index for the samples combined into 3-month intervals representing early and later parts of the breeding season. RESULTS Description of the Reproductive Tissues During the winter (June to September) most P. donacina were reproductively inactive, characterized by the relatively few indi- viduals with active gametogenic tissue. At this time it was not possible to distinguish between male and female tuatua and the sex ratio could not be accurately determined. Female gonad was found in bivalves of 52-101 mm shell length (Table 1). Early active developing female reproductive tissue was characterized by thick- TABLE 1. Sex ratio and shell length ranges of P. donacina from histological study. N = number of individuals used in sex ratio calculations; F length, female shell length range (mm): M length, male shell length (mm); I length, indeterminate shell length range (mm I. N = 20 for each sample. NA indicates that no calculation was possible. * indicates months with unequal numbers of females and males {\2 value, P< .05) Female:Male F Length M Length I Length Date Ratio N (mm) (mm) (mm) Nov 93 1.83 17 62-95 62-90 74-87 Dec 93 0.86 13 52-95 81-98 88-98 Jan 94 1.43 17 68-92 78-100 78-95 Feb 94 3.00 12 65-92 85-92 85-102 Mar 94 0.22* 11 85-90 82-92 84-102 Apr 94 2.00 9 65-95 58-90 74-95 Apr 94 2.16 19 65-90 54-62 84 May 94 1 .00 7 85-98 87-94 75-92 Jun 94 NA 0 - - 81-98 Aug 94 5.0 6 77-93 90 81-98 Oct 94 1.50 10 85-98 84-93 74-100 Nov 94 0.22* 11 83-95 67-88 66-93 Dec 94 0.66 10 83-92 66-95 68-92 Jan 95 3.00 12 85-91 72-98 84-93 Feb 95 2.00 3 85 100 79-95 Mar 95 3.00 4 81-87 82 56-94 Mar 95 2.00 6 87-92 87-90 81-98 Apr 95 NA i 78-85 - 67-94 May 95 2.00 3 83-87 90 66-92 Jun 95 1 .00 -) 83 88 63-94 Jul 95 2.00 3 77-80 82 55-92 Aug 95 NA 1 - 83 70-108 Sep 95 NA 1 - 85 47-98 Oct 95 5.50* 13 47-95 68-80 32-96 Oct 95 1.25 18 58-92 70-85 75-78 Nov 95 1.25 9 75-80 65-76 70-85 Dec 95 2.60 18 62-90 67-87 67-82 Jan 96 0.83 11 58-88 62-87 68-103 Feb 96 1.83 17 53-84 82-87 84-88 Mar 96 0.55 17 58-95 73-100 78-88 Apr 96 1.43 17 88-101 90-98 86-102 Jun 96 1.37 19 77-94 47-87 88 Jul 96 1.50 5 78-90 77-88 77-42 Aug 96 4.00 5 87-98 91 82-96 Aug 96 0.67 5 91-92 90-94 84-94 Sep 96 ] 1 .00* 12 67-92 78 80-90 Oct 96 7.00 8 71-100 80 73-95 Nov 96 3.30 13 68-86 72-80 50-93 Dec 96 1.00 16 70-84 72-90 84-88 Feb 97 1.70 19 69-90 64-82 77 Apr 97 1.00 16 68-88 65-85 73-95 Reproduction of P. donacina 541 ened follicle walls with many small oocytes. Later the follicle walls were thinner with both developing ova and mature oocytes (mixed stage). In the mature stage there was little ovogenic activ- ity, the follicle walls could not be distinguished, and the whole of the lumen was filled with mature ova. At the end of the breeding season early active follicles were observed in some sections con- taining partially spawned ova and there was histological evidence of gonad tissue undergoing reabsorption. Male gonad was found in individuals of shell length 47 to 108 mm throughout the year. This was separated into an early active reproductive stage, which was characterized by uniformly thickened follicular walls, and a mature phase, which included dense bands of spermatozoa appearing to rotate towards the centre of the lumina. Individuals of P. donacina are gonochoristic and no hermaphrodites were found. Se.x Ratio In most samples, the sex ratio of P. donacina deviated from a 1:1 female to male ratio (Table 1). In samples where more than 50% of individuals were able to be sexed, this ratio was 2.23 (SD = 2.08, n = 37). Unequal numbers of males and females occurred in 4 of the 23 samples, the combined ratio was biased in favor of females (\~ = 9.95, P < .05) and there was significant heteroge- neity between months (heterogeneity value = 37.88, P < .05). When the data were separated into two length groups, there was an obvious effect of shell length on the sex ratio. There were fewer reproductive males less than 80 mm shell length where the sex ratio was 3.33: 1 in favor of females. More than 80 mm shell length, the ratio was 1:1. Gametogenic Cycle The gametogenic cycle of P. donacina was annual with sig- nificantly reduced reproductive activity over winter (July to Sep- tember). In the early part of the cycle, the majority of females contained developing ova and there were many males with fully mature spermatogenic tissue (Fig. 1 ). The peak of breeding activity occurred in the summer (November to March) when the lumen of the female follicles contained large mature ova. The late part of the cycle, including both mature and partially spawned individuals, occurred in autumn (April), but extended some years into winter (June or July). Because few females were found with reabsorbing ova or gonads in an early development stage part way through the reproductive period, there is little evidence of multiple mass spawning events. The presence of mature ova throughout the breeding period suggests a prolonged breeding period with con- tinuously released gametes reaching a maximum during summer. However, male tuatua with early developing gonad were present throughout the breeding season. This may indicate more than one gametogenic cycle during the breeding period, or considerable variability in the onset of gamete formation in some male tuatua. At the end of the breeding season, the remaining gonad tissue appeared to be reabsorbed before the onset of winter. There was considerable variation in the timing of the reproduc- tive cycle from year to year. This variation was particularly evident during the 1994/1995 breeding season when less than 507r of the individuals were reproductively active. This contrasted with the other 3 seasons when more than 90% of the population showed reproductive activity during the summer. Environmental \ariables Seawater temperature varied seasonally, with minimum values close to 8 °C during winter and maximum values above 17 °C in Nov 93 lo Ocl 94 7 I lliJ D mature D mixed ■ F early □ late □ M devel NO J F MAjVMJJv Month Nov 94 lo Oct 95 □ mature □ mixed ■ F early □ late □ M devel N D J F Ml M2 A M J Jy Month li OND J FMAMJJyA1A2S Month Oct 96 to April 97 20 ll. 15 □ mature □ mixed 1 _ ■ F early h ■ □ late □ M devel 5 1 ■ 1 0 1 1 1 1 , ONDJ F M A M J Jy A1 A2 Monlfl Figure 1. Developmental stages of tuatua at approximately 4-week intervals from November 1993 to April 1997. Symbols show female stages, F early development, mixed oocytes, and mature oocytes. Male development is shown as M devel, early development and late, mature, ripe, and spawning clams. I shows elams with no gametogenic activity. slimmer, with an interannual variation of up to 3 °C (Fig. 2). Tuatua condition index followed the temperature cycle with simi- lar maximum and minimum values between years, and the popu- lation gonad index varied between years, with consistently low levels in February and March 1995 (Fig. 2). The chlorophyll a levels fluctuated widely, both within and between seasons. Maxi- mal values (> 12 p.g chl a L"1) were recorded in May 1995 and June 1996. and minimum values, close to 4 p.gchl a L_1, in August 1994, 1995, 1996. and December 1996 (Fig. 2). The relationship between environmental variables and tuatua reproductive condition was assessed using Spearman's rank cor- relation coefficient (Table 2). Of the factors examined, seawater temperature appeared to be the main variable affecting the num- 542 Marsden 2 ■? ' Monlhs 1993 to 1997 Figure 2. Variation in the seawater temperatures CO, chlorophyll a (ug L"'l, condition index, and °>c gonad, percent reproductive indi- viduals in tuatua populations from November 1993 to April 1997. bers of reproductive individuals present in the population. The tuatua condition index and the average gonosomatic index of male and female tuatua were also positively correlated with seawater temperature. However, none of the tuatua population or tissue indices correlated with seawater chlorophyll a levels, which in- creased at higher temperatures. Weight Relationships Regression analysis of 46 monthly samples supports a direct logarithmic relationship between dry tissue weight and shell length for tuatua (Table 3). The 3 samples where the tissue weight was independent of shell length were August 1994. February 1995. and March 1996. samples where there was a restricted length range. The average slope b of the regression lines was 2.55 (SD = 0.39, range 1.26-3.05), but there were high slope values, both at the peak of the breeding season and during winter. Cyclic patterns were observed in the estimated dry weight of small individuals, (shell length 40 mm) and of large 90 mm length) tuatua (Fig. 3). The smaller clams show high variability in dry weight, with a mean value of 0.61 g and a coefficient of variation of 32.0. For larger tuatua. the mean dry weight was 4.91 g and the coefficient of variation was 18.1. These are consistent with cycles of somatic and gametogenic activity. Gonosomatic Index The tissue gonad index for tuatua shows seasonal patterns (Fig. 4). Greater reproductive tissue mass occurred in males than fe- males in the first 2 years of the study, but this trend was reversed in the following seasons. Although these patterns between male and female tuatua differed between years (ANOVA males, F = 4.37. P = .02; females, F = 4.89, P = .01), they were similar within a particular season (t tests). When the gonad tissue data were combined for 3 monthly periods, representing early and later parts of the breeding season (Fig. 5), reproductive activity was consistently higher early rather than late in the season. Linear regression analysis investigating the effects of body size and condition on the gonosomatic index of male and female tuatua early and late in the breeding season are included in Table 4. Generally, the gonosomatic index was independent of shell length, dry tissue weight and condition index. However, there were five exceptions. In the 1994/5 season, when there were relatively few reproductive individuals in the population, large male tuatua early in the season produced less gonad than smaller individuals. This same relationship occurred in female tuatua late in the 1995/6 season. However, earlier in the same season, female gonosomatic index increased significantly with the condition index. For the 39 monthly samples, the average gonosomatic index for TABLE 2. Spearman's rank linear correlation coefficients relating reproductive variables and environmental variables: I, number of indeterminate individuals; Me, number of males with early development stages; MI, number of males with late developmental stages; Fe, female early stages; Fl, female late stages; % , percent individuals with gonad: T, sea-water temperature; chl a, chlorophyll a; CI, condition index; Mg, male gonosomatic index; Fg. female gonosomatic index. * indicates value significant at P < .05; -, negative correlation. Variable Me MI Fe Fl Chl a CI Mg Fg I Me -.35* MI -.66* Fe -.15 Fl -.63* % -.98* T -.54* Chl a .03 CI -.57* Mg -.53* Fg -.56* 35* -.66* -.15 -.63* -.98* -.54 .03 -.57 - -.12 -.00 .11 .35* .11 -.06 .02 12 - .29 .63* .69* .43* .08 00 -.29 - .39* .07 -.12 11 .63* -.39* - .65* 35* .69* .07 .65* - 11 .43 -.12 .48 06 .08 -.04 02 .58* 22 -.53* ,56* Reproduction of P. donacina 543 TABLE 3. Seasonal variation in regression coefficients for the equation v = and .v is the shell length (mini in P. donacina, from November 1993 to April 1997. a is the intercept: b, slope of the regression line; r, correlation coefficient; P, probability level; n.s.. not significant, n = 2(1 in all cases. Date a b r P 1 2741.93 2.48 x lO-5 2.77 0.95 0.01 2 15.2.93 4.43 x 10"5 2.63 0.92 0.01 3 8.1.94 6.34 x 10"6 3.04 0.96 0.01 4 1 1 .2.94 2.04 x lO"4 2.26 0.85 0.01 5 12.3.94 1.16 x 10~4 2.37 0.84 0.01 6 1.4.94 4.89 x 10-5 2.56 0.95 0.01 7 28.4.94 1.20 x lfr5 2.91 0.97 0.01 X 24.5.94 6.08 x lO"3 3.01 0.95 0.01 9 24.6.94 2.08 x 10"1 2.19 0.79 0.01 11 12.8.94 1.32 x 10"2 1.26 0.30 n.s. 13 10.10.94 1.75 x lO^4 2.23 0.89 0.01 14 6.11.94 2.23 x 10"4 2.19 0.82 0.01 15 5.12.94 5.08 x 10-4 2.05 0.83 0.01 16 1.1.95 9.96 x 10-5 2.43 0.96 0.01 17 2.2.95 132.2 0.75 0.17 n.s. 18 1.3.95 7.23 x 10"5 2.47 0.92 0.01 19 29.3.95 3.39 x 10"' 1.65 0.78 0.01 20 19.4.95 1.53 x 10"3 2.81 0.94 0.01 21 15.5.95 6.45 x lO"6 2.99 0.92 0.01 22 16.6.95 5.53 x 10"6 3.00 0.93 0.01 23 14.7.95 1.92 x 10"5 2.72 0.96 0.01 24 11.8.95 1.22 x 10"4 2.30 0.89 0.01 25 10.9.95 1.72 x 10'4 2.23 0.97 0.01 26 6.10.95 1.41 x 10"5 2.78 0.94 0.01 27 27.10.95 1.41 x 10"5 2.78 0.94 0.01 28 23.11.95 3.39 x 10-"' 2.60 0.98 0.01 29 27.12.95 8.17 x 10~6 3.02 0.99 0.01 30 25.1.96 7.53 x 10"6 3.05 0.99 0.01 31 21.2.96 1.57 x 10"5 2.86 0.97 0.01 32 20.3.96 5.75 0.25 0.37 n.s. 33 16.4.96 1.45 x 10"4 2.29 0.94 0.01 35 4.6.96 5.74 x 10'5 2.48 0.94 0.01 36 4.7.96 2.87 x irr5 2.63 0.97 0.01 37 2.8.96 5.28 x 10"6 2.99 0.93 0.01 38 31.8.96 1.21 x 1CT5 2.84 0.99 0.01 39 26.9.96 6.72 x 10-5 244 0.96 0.01 40 28.10.96 1.37 x 10"4 2.32 0.95 0.01 41 21.11.96 2.93 x 10~* 2.16 0.97 0.01 42 13.12.96 2.55 x 10"5 2.73 0.95 0.01 44 9.2.97 3.67 x lO"5 2.65 0.96 0.01 46 7.4.97 1.53 x irr5 2.85 0.92 0.01 male and female tuatua increased significantly with the population gonad index (linear regression analysis, Fig. 6). Therefore, as the number of reproductively active tuatua increased in the population, each individual was likely to have increased reproductive poten- tial. DISCUSSION Gametogenesis in Paphies donacina is similar to that described for many other surf clams, including the New Zealand mactrids (Redfearn 1974, Grant and Creese 1995, Hooker and Creese 1995a. McLachlan et al. 1996. King 1997). In P. donacina, as in some other surf clams, there was a resting or inactive stage during winter (Kanti et al. 1993. King 1997). This contrasts with P. rat- Figure 3. Seasonal variation in tissue dry weight (g) for P. donacina shell length 40 mm and 90 mm. Values calculated from the regression lines in Table 2. tricosca, P. australis, and P. subtriangulata from warmer regions of northern New Zealand, where gametogenesis was continuous throughout the year, and there were one or two distinct spawning periods (Redfearn 1974, Grant and Creese 1995, Hooker and Creese 1995a). Several studies of bivalves have found differences in the reproductive cycle with geographic location (Keck et al. 1975. McLachlan et al. 1996). For example. Donax trunculus, which is distributed from Brittany to southern Morocco and the Mediterranean, shows increased duration of the breeding season in Southern Spain (Tirado and Salas 1998) compared with other ar- eas. Also, within a species, populations in different areas may have two spawning periods (Ansell and Bodoy 1979) or spawnings may, as in P. donacina, be more evenly distributed throughout the sum- mer (Ansell et al. 1980). This timing of gametogenic cycles sug- gests an exploitive reproductive strategy for surf clams, contrasting with more conservative gametogenesis patterns, where bivalves Month Figure 4. Seasonal variation in the gonosomatic index for male (x) and female (•) in P. donacina from November 1993 to April 1997. 544 Marsden 1993E 1994L 1994E 1995L 1995E 1996L 1996E Year Figure 5. Comparison of gonosomatic index for male and female P. donacina early (E) (November to January) and (L) later (March to May) in the breeding season. Error bars show the standard error from the combined samples. N is shown in Table 4. spawn over the winter using up resources that have built up over the summer (Bayne 1976). In contrast with surf-clams such as D. trunculus, where males and females can be readily distinguished by their color during the breeding cycle, the sex of P. donacina could not be distinguished other than by histological sections. The sex ratio of 1.7 females to 1 male in this species differs from the more usual 1 : 1 ratio which is found in most other surf clams including D. trunculus (Tirado and Salas 1998), D. serra, D. cuneatus. Tivela mactroides (McLa- clan et al. 1996), and other New Zealand surf clams (Redfearn 1974, Grant and Creese 1995. Hooker and Creese 1995a, King 1997). Several possible explanations have been forwarded to ex- plain inequalities in the sex ratios of bivalves. These explanations include: a small sample size, differential rates of development, and differences in the maximal lengths for male and female bivalves. For P. donacina. the maximal shell length was similar for males and females, and the unequal sex ratio within the length range 45-108 mm was caused by females dominating the smaller length groups. The possibility of delayed sexual maturity in smaller males is supported by the presence of individuals of an indeterminate sex in the population throughout the year. In studying the reproductive biology of bivalves, some authors have suggested that the condition index of bivalves can act as an indicator of reproductive activity (Dix and Ferguson 1984, Cheung 1991). For P. donacina. populations with a high proportion of reproductive individuals would be expected to have a high CI 60 y = 0.2793x + 12.918 ♦ ♦ 1 50 r=0 53, N=39. P=0.01 * * ♦ • S 40 to E § 30- I 20 ♦ t TO ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ 0 20 40 60 80 100 Population gonad index 60 f 50 o § 40 E o £ 30 o Ol « 20 1 L- 10 0 y = 0 2663x ♦ 1 1 723 " r=060. N=39. P=001 ■ ■ ■ ■ - — ' ■ _— — —*~~m\ 1 ■ ■ ■ ■ ■ ■ ~~m ■ ■ " . ) 20 40 60 Population gonad index 80 100 Figure 6. Relationship between gonosomatic index of male and female P. donacina plotted against the population gonad index. Both of the regression lines are significant at the 1% level and the slopes and elevations are similar. index, but the index is not a good indicator of the reproductive potential. This is because both somatic tissue and gonad growth, as well as shell growth, may be occurring simultaneously in this species. This is supported by the seasonal weight relationships of smaller length tuatua that may not be reproductively active. The effects of temperature on the gametogenic and spawning cycles of bivalves is complex, and depends on the overall repro- ductive strategy and the environmental temperature range. Some clams have an inactive period, which is the result of high envi- ronmental temperatures (Hesselman et al. 1989). and others, like Spisula solidissima ( Kami et al. 1 993 ) and P. donacina, where low temperatures can delay gametogenesis and the periodicity of TABLE 4. Linear regression correlation coefficients relating shell length (L, mm), dry wt (W, g) and condition index (CI) to the gonosomatic index for male (M) and female (F) P. donacina early and late in the breeding season. N, number of individuals; * values significant at P < .05; ** values significant at P < .01; -. negative correlations. Season ML MD MCI MN FL FD FCI FN 1993/4 Early 0.13 0.02 0.14 22 0.11 0.14 0.35 28 Late 0.33 0.31 0.25 18 -0.43* -0.45* 0.04 22 1994/5 Earlv -0.77** -0.74** 0.10 17 0.36 0.28 0.02 15 Late 0.38 0.89 0.59 4 0.03 0.25 0.48 9 1995/6 Early 0.01 0.16 0.23 15 0.36 0.29 0.58** 23 Late 0.32 0.18 0.13 17 -0.55* -0.60* 0.16 16 1996/7 Early 0.20 0.56 0.22 11 0.29 0.05 0.31 14 Reproduction of P. donacina 545 spawning. In P. donacina and M. mercenaria (Manzi et al. 1985), the gametogenic cycle depends on temperature, but in others, like the European cockle. Cerastoderma edule (Navarro et al. 1989) and the semi-tropical subtidal clam. Megaitaria aurantica (Garcia- Dominguez et al. 1994). the gametogenic cycle was temperature- independent. Although in P. donacina, both the number of repro- ductive individuals and the gonosomatic index correlated with sea- water temperature, this factor was not a good predictor for reproductive success between years. A threshold temperature for spawning occurs in some oyster species (Burrell 1985) and in the northern tuatua P. subtriangulata (Grant and Creese 1995). How- ever, no such threshold was detected in P. donacina or the cockle Cerastoderma edule (Navarro et al. 1989), where spawning can occur over a wide temperature range. For the cockle, annual varia- tions are thought to be due to either the quantity of ingested food or to levels of previously stored nutrient reserves. Relatively few studies have followed the reproductive cycles of bivalves over a number of years, and of these only a few have provided values for both the temperature and available food re- sources. Kautsky ( 1982), studying the mussel Myrilus edulis popu- lations in the Baltic, concluded that annual differences in the dry weight of individuals was due to the food abundance not the tem- perature of the habitat. Although temperature has been identified as the major factor affecting the reproduction of P. donacina, it is interesting that there was no similar effect due to food levels as indicated by levels of chlorophyll a. Some studies have shown such a relationship may have a lag phase, which most likely results from the build up of reserves before gamete formation (Hawkins et al. 1985). There is some support for this observation in the present study, because chlorophyll a levels correlated with seawater tem- perature. Although sand beaches have been described as low production habitats compared with estuaries or rocky shores, this generaliza- tion does not hold for tropical and subtropical exposed sand beaches where productivity levels are high. Species of Dona.x dominate these habitats (Ansell 1983). and opportunistic features of their life style include small size, rapid growth, early sexual maturity, and high fecundity. In southern temperate latitudes, there are fewer clam species including a few Dona.x species and mac- trids such as Paphies and Mesodesma spp. These much larger surf clams are most abundant within highly productive areas. For ex- ample, D serra from South Africa occurs in warm water areas supplied with organic enrichment from upwellings (Brown et al. 1989). High production may also be due to surf zone algal blooms or inputs from kelp strandings. land runoff, or river debris. These nutrients, including inorganic nitrogen from the beach itself, can be trapped inshore by semi-enclosing water patterns and on-shore winds. If these processes result in high productivity levels through- out the year, then they most likely explain the fast growth rates of species like P. donacina, their large maximal size (McLachlan et al. 1996, Cranfield et al. 1996). and high fecundity. Dawson (1954) estimated egg production by P. donacina to be 30-40 million per individual and values for the toheroa of 80 and 90 million eggs per individual. Similarly high values have been found in other surf- clams such as Tivela stultorum (11-20 million, Coe and Fitch 1950). This large reproductive output, together with an extended breeding period, allows larval release over an extended period without the risk of food limitation. A longer recruitment period may also assist dispersion and might increase larval survival within this physically dominated habitat where mortality is likely to be high. For P. donacina, gonad development, somatic growth, and ac- cumulation of energy resources overlap during the spring and sum- mer and may extend into autumn because of continued high food levels. It is likely that P. donacina can maintain positive scope for growth at this time, because it shows seasonal adaptation in its oxygen consumption and feeding rate, maintaining a similar fil- tration rate at 10 °C in winter to 20 °C in summer (Marsden in press). The ecology and distribution patterns of P. donacina closely resemble M. donaciitm from southern Africa. Both species can occur — almost exclusively — intertidally (Dawson 1954) and also extend into the surf zone (Brown et al. 1989). The breeding cycle is exploitive, flexible, and largely determined by the tem- perature regime. As food is rarely limiting, these surf clam species would be expected to reproduce well during El Nino years, which are associated with increased seawater temperatures. ACKNOWLEDGMENTS I would like to thank Bryn and Celyn Fenwick for their help collecting tuatua and for laboratory assistance with the histological work. Thanks also to Reigel Gardener in Plant and Microbial Sciences Department, University of Canterbury, for the use of histological equipment. This research was funded by University of Canterbury Research Grant U61 15. This paper is dedicated to my friend and mentor Alan Ansell who generously shared with me his extensive knowledge, experience, and enthusiasm for sand beach bivalves. LITERATURE CITED Ansell. A. D. 1983. The biology of the genus Donax. pp. 607-635. In: A. McLachlan & T. Erasmus (eds.). Sandy beaches as ecosystems. Junk Publishers. The Hague. Ansell. A. D. & A. Bodoy. 1979. Comparison of the events in the seasonal cycle for Donax vittatus and Donax trunculus. pp. 191-198. In: E. Naylor & R. G. Hartnull (eds.) Cyclic phenomena in marine plants and animals. Pergamon Press. Oxford. Ansell, A. D., L. Frenkiel & M. Moueza. 1980. Seasonal changes in tissue weight and biochemical composition for the bivalve Donax trunculus on the Algerian coast. J. Exp. Mar. Biol. Ecol. 45:105-116. Bayne, B. L. 1976. Aspects of reproduction in bivalve molluscs, pp. 432- 448. In: M. Wiley (ed.). Estuarine processes. Vol. 1. Uses, stresses and adaptation to the estuary. Academic Press. London. Brown, A. C, J. M. E. Stenton-Dozey & E. R. Trueman. 1989. Sandy- beach bivalves and gastropods: a comparison between Donax serra and Bullia digitalis. Adv. Mar. Biol. 25:179-247. Burrell. V. G. Jr. 1985. Oyster culture, pp. 235-273. In: Crustacean and Mollusk Aquaculture in the United States. J. V. Hemer. and E. E. Brown (eds.). Cheung, S. G. 1991. Energetics of transplanted populations of the green- lipped mussel Pema viridis (Linnaeus) (Bivalvia: Mvtilidael in Hong Kong. 1: growth, condition and reproduction. Asian Mar. Biol. 8:117- 131. Coe. W. R. & J. E. Fitch. 1950. Population studies, local growth rates and reproduction of the Pismo clam ( Tivela stultorum). J. Mar. Res. 9: 1 88- 210. Cranfield, H. J.. K. P. Michael & R. I. C. C. Francis. 1996. Growth rates of five species of subtidal clam on a beach in the South Island, New Zealand. Mar. Freshwater Res. 47:773-784. Cranfield, H. J.. K. P. Michael. D. Stotter & I. J. Doonan. 1994. Distribu- tion, biomass and yield estimates of surf clams off New Zealand beaches. N.Z. Fish. Ass. Res. Doc. 94/1. N.Z. Ministry of Fisheries. 546 Marsden Dawson, E. W. 1954. Studies on the biology of Amphidesma, an intertidal mollusc of the sandy shore. M.Sc. thesis. University of Canterbury. New Zealand. Dibacco. C, G. Robert & J. Grant. 1995. Reproductive cycle of the sea scallop. Placopecten magellanicus {Gmelin, 1791). on northeastern Georges Bank. J. Shellfish Res. 14:59-69. Dix. T. G. & A. Ferguson. 1984. Cycles of reproduction and condition in Tasmanian blue mussels. Mytilus edulis planulatus. Aust. J. Mar. Freshw. Res. 35:307-313. Estcourt, I. N. 1962. An ecological survey of burrowing polychaetes of the Heathcote-Avon Estuary. M.Sc thesis. University of Canterbury. New Zealand. Garcia-Dominguez. F., S. A. Garcia-Gasga & J. L. Castro-Oritz. 1994. Spawning cycle of the red clam Megaptitaria aurantiaca (Sowerby. 1831) (Veneridae) at Isla Espiritu Santo. Baja California Sur, Mexico. J. Shellfish Res. 13:417-423. Grant. C. M. & R. G. Creese. 1995. The reproductive cycle of the tuatua- Paphies subtriangulala (Wood, 1828), in New Zealand. J. Shellfish Res. 14:287-292. Haddon, M., T.J. Willis. R. G. & V. C. Anderlini. 1996. Biomass and distribution of five species of surf clam off an exposed west coast North Island beach. New Zealand. /. Shellfish Res. 15:331-339. Hawkins. A. J. S.. P. N. Salked. B. L. Gnaiger & D. M. Lowe. 1985. Feed- ing and resource allocation in the mussel Mytilus edulis: evidence for time-averaged optimisation. Mar. Ecol. Prog. Set: 20:273-287. Hesselman, D. M.. B. J. Barber & N. J. Blake. 1989. The reproductive cycle of adult hard clams Mercenaria spp. in the Indian River Lagoon, Florida. J. Shellfish Res. 8:43-49. Hooker. S. H. & R. G. Creese. 1995a. The reproductive biology of the pipi. Paphies australis (Gmelin, 1790) (Bivalvia: Mesodesmatidae). 1. Tem- poral patterns of the reproductive cycle. J. Shellfish Res. 14:7-15. Hooker, S. H. & R. G. Creese. 1995b. The reproductive biology of the pipi. Paphies australis (Gmelin, 1790) (Bivalvia: Mesodesmatidae). 11. Spatial patterns of the reproductive cycle. /. Shellfish Res. 14:17-24. Hummel. H. 1985. Food intake of Macoma balthica (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. Kanti, A.. Haeffernan, P. B. & R. L. Walker. 1993. Gametogenic cycles in the southern surf clam Spisula solidissima similis (Say. 1822). from St Catherines Sound. Georgia. J. Shellfish Res. 12:255-261. Kautsky. N. 1982. Quantitative studies on gonad cycle, fecundity, repro- ductive output and recruitment in a Baltic Mytilus edulis population. Mar. Biol. 68:143-160. Keck. R. T.. D. Maurer & H. Lind. A comparative study of the hard clam gonad development cycle. Biol. Bull. 148:243-258. King. B. E. 1997. Comparative aspects of gonad development in field and laboratory conditioned surf clams from Foxton Beach. New Zealand. M.Sc thesis. Victoria University. Wellington, New Zealand. Kirk. R. M. 1980. Mixed sand and gravel beaches: morphology, processes and sediments. Prog. Phys. Geog. 4:189-210. Lucas, A. & P. G. Beninger. 1985. The use of physiological condition indices in marine bivalve aquaculture. Aquaculture 44:187-200. Manzi. J.. M. Bobo & V. Burrell. Jr. 1985. Gametogenesis in a population of the hard clam, Mercenaria mercenaria (Linnaeus), in North Santee Bay, South Carolina. Veliger 28:186-194. Marsden. 1. D. (in press). Respiration and feeding of the surf clam Paphies donacina. from New Zealand. Hydrobiologica. McLachlan. A.. J. E. Duggan. O. Defeo, A. D. Ansell. D. M. Hubbard. E. Jaramillo & P. E. Penchaszadeh. 1996. Beach clam fisheries. Oceanog. Mar. Biol. Ann. Rev. 34:163-232. McLachlan. A. & T. Erasmus. 1983. Sandy beaches as ecosystems. Dr Junk Pubf. The Hague. McLachlan. A. & E. Jaramillo. 1995. Zonation on sandy beaches. Oceanog. Mar. Biol. Ann Rev. 33:305-335. Morales- Alamo. R. & R. Mann. 1989. Anatomical features in histological sections of Crassostrea virginica (Gmelin, 1791 ) as an aid in measure- ments of gonad area for reproductive assessment. J. Shellfish Res. 8:71-8 Navarro. E.. J. I. P. Iglesias & A. Larranaga. 1989. Interannual variation in the reproductive cycle and biochemical composition of the cockle Cerastoderma edule from Mundaca Estuary (Biscay. North Spain). Mar. Biol. 101:503-511. Redfearn. P. 1974. Biology and distribution of the Toheroa. Paphies (Me- sodesma) ventricosa (Gray). Fish. Res. Bull. 1 1. New Zealand Ministry of Agriculture and Fisheries: 51 pp Short. A. D. 1990. Macro-mesal tidal beach morphodynamics-an overview. J. Coast. Res. 7:417^436. Tirado. C. & C. Salas. 1998. Reproduction and fecundity of Donax trun- culus L.. 1758 (Bivalvia: Donacidae) in the littoral of Malaga (Southern Spain). J. Shellfish Res. 17:169-176. Urrutia. M. B., J. I. P. Iglesias & E. Navarro. 1997. Feeding behaviour of Cerastoderma edule in a turbid environment: physiological adaptation and derived benefit. Hydrobiologica 355:173-180. Widdows. J. 1978. Combined effects of body size, food concentration and season on the physiology of Mytilus edulis. J. Mar. Biol. Ass. U.K. 58:109-124. Journal of Shellfish Research, Vol. 18, No. 2. 547-553, 1999. UPTAKE OF DISSOLVED FREE AMINO ACIDS BY NORTHERN QUAHOGS, MERCENARIA MERCENARIA AND ITS RELATIVE IMPORTANCE TO ORGANIC NITROGEN DEPOSITION IN NARRAGANSETT BAY, RHODE ISLAND MICHAEL A. RICE Department of Fisheries Animal and Veterinary Science University of Rhode Island Kingston, RI 02881 ABSTRACT Studies were undertaken to determine the relationship between size of northern quahogs Mercenaria mercenaria and the rate at which they transport aspartic acid. Quahogs ranging from 25 to 103 mm valve length were collected in Narragansett Bay and placed in seawater aquaria (27 ppt. 20 °C) and allowed to pump water actively. Uptake experiments were carried out using 1 p.mol/L C14 radiolabeled aspartic acid. Aspartate transport rates in (xmol/h can be related to valve length by the allometnc equation with a = 24.32 and b = 0.905 when valve length is in mm. In May 1990, near-bottom samples of seawater were taken from five locations in Narragansett Bay for analysis of dissolved free amino acids (DFAA) by high-pressure liquid chromatography (HPLC). Results showed that the mean total DFAA concentration was 667.6 nmol/L ± 167.3 SD, with the top five being serine, alanine, aspartic acid, glutamic acid, and glycine. A simple spreadsheet model was used to assess the relative importance of the uptake of DFAA vis-a-vis the filtration of particulate organic matter by M. mercenaria. In the model, environmental DFAA concentrations and uptake rates by quahogs determined in this study are compared with literature values for particulate organic nitrogen concentrations and filtration rates by quahogs. On an annual basis, uptake of DFAA can account for about 14% of the total organic nitrogen uptake by quahogs. Uptake of DFAA by these benthic filter feeders may be a pathway of benthic-pelagic nutrient coupling that is often overlooked in coastal ecosystem analyses. KEY WORDS: DFAA. quahog. Mercenaria, Narragansett Bay, filter feeding, benthic-pelagic coupling INTRODUCTION Since Stephens and Schinske ( 1961 ) demonstrated that several phyla of soft-bodied marine invertebrates are able to uptake dis- solved free amino acids (DFAA) directly across epithelial surfaces from seawater, subsequent studies have shown that uptake of DFAA occurs at concentrations that are typical in estuarine and marine environments (Manahan et al. 1982). DFAA in estuarine and coastal waters are believed to be released from sediments via microbial decomposition of complex organic matter and as exu- dates from phytoplankton and macrophytes (Jorgensen et al. 1980. Jorgensen 1982, Bronk et al. 1994). These DFAA in estuarine and coastal waters are of potential nutritional significance to most soft- bodied marine invertebrates (reviewed by Stephens 1988, Wright and Manahan 1989). Most work regarding the uptake of DFAA by invertebrates has focused on its significance at the organismal level. Several studies (e.g., Pequignat 1973, Manahan and Crisp 1983, Chien and Rice 1985) used histologic and autoradiographic methods to demon- strate the incorporation of transported DFAA into epithelial and subepithelial tissues. Other work has shown that DFAA can be transported by a number of invertebrate species at rates that are of nutritional significance, often supporting in excess of 50% of their measured oxygen consumption rates (e.g., Davis et al. 1985, O'Dell and Stephens 1986). In addition to nutritional significance, the uptake of DFAA has been demonstrated to be an important mechanism for the recovery of diffusionally lost solutes from high- concentration DFAA pools in epithelial tissues (Gomme 1982, Wright and Secomb 1984). In addition to work characterizing the organismal significance of DFAA uptake by marine invertebrates, some work has aimed at characterizing the nature and specificities of the DFAA carriers in epithelial tissues. Preston and Stevens (1982) showed that DFAA uptake by Glycera dibranchiata, polychaete, is a type of sodium cotransport with a Na+:amino acid coupling ratio of 3:1 necessary to energetically overcome a million-fold concentration gradient. Using competitive inhibition studies in bivalves, other workers have shown that the amino acid carriers, in general have broad specificities for a-amino acids, but there is evidence of separate carriers for fj-amino acids, such as taurine (Wright and Secomb 1984, Wright 1985, Rice and Stephens 1987). Eventually, Pajor and Wright (1987) isolated apical membrane vesicles from the ctenidium of the bivalve Mytilus edulis, and demonstrated active transport of alanine. Despite the extensive work describing the physiological sig- nificance of the uptake of DFAA and the characterization of the amino acid carriers within epithelial membranes, there is a dearth of available information about the ecological significance of DFAA uptake by populations of invertebrates in estuarine or coastal marine waters. Stephens (1981, Stephens 1982) proposed that uptake of dissolved organic material (DOM), including DFAA, can provide supplemental nutrition to invertebrates resid- ing in primary and secondary consumer trophic levels. He argued that this extra nutritional energy input directly to the higher trophic levels can potentially confound the accounting of trophic effi- ciency usually calculated based on the consumption of prey from lower trophic levels. Interestingly, Stephens's conclusions about trophic significance of DOM uptake mirrored that of Putter's (1909) postulation eight decades earlier. Aside from the potential trophic significance of DFAA uptake by invertebrates, the amount of organic nitrogen being cycled within assemblages of invertebrates may be a significant contri- bution to nitrogen mass balance budgets in estuarine and coastal environments. A priori, we could expect that on a per hectare basis upward of 8 kg of nitrogen could be cycled daily via uptake of FAA alone (assuming 5-g mussels and 14-g N/mol DFAA). This estimation is based on published rates of DFAA uptake by some 547 548 Rice marine mussels. Mytilus edulis, of 1 u.mol/g • h from mixed amino acids in the environmentally realistic micromolar range (Manahan et al. 1983b) and numbers of mussels in some natural assemblages of 500/nr (Newell and Shumway 1993). The significance of the uptake of DFAA and other forms of dissolved organic nitrogen (DON) by benthic filter feeders deserves further attention in light of its potential significance to recent discussions of estuarine ni- trogen budgets. Indeed, in their study of mussels in the Wadden Sea of Germany, Siebers and Winkler (1984) concluded that, ". . . mussel beds in shallow coastal waters exposed to tidal move- ments of large seawater masses play a significant role in recycling of dissolved organic material." But they did not estimate the rela- tive importance of DFAA uptake and PON filtration at their site. The first aim of this study is to determine the size-specific transport of a representative amino acid, aspartic acid, by an in- faunal bivalve mollusk. Mercenaria mercenaria, one of the dom- inant benthic filter feeders in Narragansett Bay. The next aim is to determine the mean concentration of DFAA in Narragansett Bay waters and then to apply the size-specific transport rates by M. mercenaria to available information on M. mercenaria populations in Narragansett Bay, thereby estimating the bay-wide magnitude of DON cycling attributable to the DFAA uptake. MATERIALS AND METHODS Northern quahogs. Mercenaria mercenaria, ranging in size from 24-103 mm valve length (/; = 20) were collected from Narragansett Bay and placed in 40-L opaque-walled aquaria in the laboratory approximately 12 hours before commencement of as- partic acid uptake experiments. The experimental aquaria con- tained aerated Narragansett Bay seawater (27 ppt and 20 °C). Nar- ragansett Bay seawater used during these experiments was passed through a 0.45 p.m nominal pore size filtration system and stored a minimum of 2 weeks in the laboratory to reduce the concentra- tions of naturally occurring DFAA. Labile dissolved organic ma- terial including DFAA are known to be greatly reduced in stored seawater (Waksman and Carey 1935, Stephens and Manahan 1984). Aspartic acid was chosen as a representative amino acid for study, because it is regularly among the top four most highly concentrated amino acids in seawater (Manahan et al. 1982. Sie- bers and Winkler 1984). Amino acid uptake experiments were begun only if the quahog siphons were visible as a sign of active water pumping. Water pumping by quahogs is easily disturbed in laboratory settings (e.g.. Rice and Stephens 1988), but dim illu- mination and limiting outside disturbances can help improve the chances of inducing the quahogs to extend their siphons actively and to commence pumping water. Despite these precautions, some of the quahogs collected in Narragansett Bay did not extend their siphons during the 12-hour preliminary conditioning period, thus they were not used for the uptake experiments. Uptake experiments were begun by the addition of stock as- partic acid to the experimental aquaria to give a final concentration of 1.0 (j.m. In addition, uniformly labeled 14C-aspartic acid (New England Nuclear) was added to a final specific activity of approxi- mately 150 Bq/mL (4 nCi/mL). Samples of the medium were taken initially and at the conclusion of the experiment at the end of a 2-hour period for liquid scintillation counting. Following the ex- periment, quahog valve lengths were measured, and they were shucked, blotted, and weighed, and the soft tissues were solubi- lized using formic acid. Samples of solubilized tissues and seawa- ter (1 mL) were prepared for scintillation counting by adding a scintillation cocktail suitable for aquatic samples (Aquasol. Du- pont/New England Nuclear). The radioactivity in each of the samples was determined using a Packard Instruments Tri-Carb liquid scintillation counter corrected for background and quench- ing to count at approximately 96% efficiency as determined by an internal radiation standard and an external C14 standard. Aspartic acid uptake data were analyzed and plotted using Axum 4.0 (Se- attle, WA) Statistical and Graphics software. To determine the concentration of DFAA in M. mercenaria habitat, near-bottom seawater samples were taken in five locations in Narragansett Bay (Fig. 1) in mid-May 1990. Samples were taken from either a small boat or at the end of a pier with a 2-Liter alpha water sampler tripped with a lead "bullet" messenger. The five locations were Outer Wickford Harbor (41°34.5'N; 71°26.5'W); South Ferry Pier (41°29.5'N; 71°25.1'W), Jamestown Harbor Pier (41°29.7'N; 71°21.9'W), Greenwich Bay (41°40.rN, 71°21.9'W). and Conimicut Point (41 42.6'N, 71°20.7'W). Once brought to the surface, triplicate samples of the water were taken from the sampler and filtered through a 0.1 -p.m pore size syringe filter into 1 .5-mL polyethylene microcentrifuge tubes and placed on ice for transport to the laboratory. All water samples were stored in a -60 °C freezer awaiting DFAA analysis. Amino acids in the seawater samples were analyzed by HPLC — — 1- — - j r— Narragansett Bay S Ge ographic Segments kLk 11 B Pron^( ^ ^S \_Grtxn*Kb ftr &y-4/ S_ «V\ am lest Passage J P I \ Upper ( f \ \^. 6 \ \ V> if *■ 1*^C fat PUsaje j Evt Paaage ( ft J /Siianoel L / firer \ /" H Jjv Rhode island Sound i A/ \ ,. r. n> e its Figure 1. Geographic segments of Narragansett Bay, Rhode Island, USA. Narragansett Bay has a total area of 389.3 km2 and is divided into 1 1 geographic regions. Closed circles on the map refer to five seawater collection stations for determination of dissolved free amino acids. DFAA Uptake in M. mercenaria 549 using a dual pump (Waters model 501) gradient mixing system (Waters model 481 controller). Separations were made on an Ultrasphere-ODS reverse-phase column (150 x 4.6 mm. 5 mm packing) and a Waters model 720 fluorescence detector was used with output to a 10 raV integrating chart recorder. Fluorescent derivatives of the DFAA were prepared using o-phthaldialdehyde reagent (Lindroth and Mopper 1979), and eluent buffers were pre- pared and used as described by Jones et al. (1981). with modifi- cations to improve separations. Duplicate samples of water from the five sites in Narragansett Bay were analyzed for DFAA, with the archived third sample maintained in reserve to be analyzed if there was a large discrepancy between the results of the duplicates. DFAAs in the natural waters were quantified by integrating peak area in comparison to 100 nanomolar amino acid standards. Fur- ther details on the procedure of DFAA analysis by HPLC are provided by Davis and Stephens (1984). To make projections on the magnitude of nitrogen cycling in Narragansett Bay by uptake of DFAA by M. mercenaria, a simple model was developed using the Microsoft Excel spreadsheet mi- crocomputer program. Using Geographical Information System (GIS) data from the University of Rhode Island Geographic Infor- mation Center, Narragansett Bay was broken up into 1 1 geo- graphic segments (Fig. 1), and the area for each segment was determined. Standing crop densities and length-frequency profiles of M. mercenaria in each of the 1 1 segments was estimated by using best available data from fisheries surveys undertaken in Nar- ragansett Bay since 1949 (most recently reviewed by Rice 1992). DFAA uptake rates are then compared to the best available data on particulate filtration rates by M. mercenaria (Doering and Oviatt 1986) along with estimates of phytoplankton, particulate detritus and total particulate organic nitrogen in Narragansett Bay (e.g., Oviatt and Nixon 1975, Durbin and Durbin 1981). RESULTS The uptake of aspartic acid by actively pumping northern qua- hogs, Mercenaria mercenaria, was demonstrated by disappearance of radiolabel in the medium and appearance of the label in the molluscan soft tissues. The rates of aspartic acid transport by in- dividual quahogs were determined by calculating a first-order rate constant based on radioactivity levels in the soft tissues at the end of 2 hours. The uptake rates of aspartic acid ranged from 0.70 p.mol/h in a 25 mm quahog and 2.55 p,mol/h in a 69 mm quahog. The rates of uptake of aspartic acid (J) by actively pumping M. mercenaria at 20 °C is size dependent (Fig. 2) and can be de- scribed by the allometric equation: J = a*Lb (1) when. J = rate of aspartic acid uptake in p.mol/h L = valve length in mm a = 24.32 and b = 0.905 The concentrations of various amino acids in Narragansett Bay waters are presented in Table 1 . Data represent the mean of rep- licate DFAA determinations from the five sites. Water was well mixed in the water sampler, so there was less than a 5% discrep- ancy between replicates. The top five amino acids in concentration at all five locations in Narragansett Bay were serine, alanine, as- partic acid, glutamic acid, and glycine, typically making up over 0.5 04 03 02 01 00 -0 1 -0.2 -1 366 + 0 9047'x 1 3 20 2.1 4 15 16 17 18 19 Log10 Valve Length in mm Figure 2. A double logarithmic plot relating aspartic acid uptake rate to quahog valve length. The uptake of aspartic acids by northern quahogs, Mercenaria mercenaria, is size dependent and can be related to valve length by the allometric equation with a = 24.32 and b = 0.905 when valve length is in mm. The correlation coefficient for the data is r2 = 0.917 (n = 20). Quahog valve length can be related to whole-body mass by the allometric equation with a = 6.45 x 10~4 and b = 2.81 (r2 = 0.985, ;i = 20), when valve length is in mm and mass is in g. 80% of the total DFAA. Total DFAA ranged from 478 nmol/L at Jamestown Harbor Pier to 852 nmol/L Outer Wickford Harbor, or nearly a two-fold difference in range. The sites closest to the mouth of Narragansett Bay (South Ferry and Jamestown Harbor) had the lowest concentrations of DFAAs, but it is uncertain if this difference is statistically significant, because only single water samples were taken at each of the five sites. From the data in Table 1. the mean DFAA concentration in Narragansett Bay is 667 nmol/L. so assuming there are 14 g of nitrogen per mole, the DON represented by DFAA is 9.35 p-g/L. The spreadsheet model of DFAA uptake by the population of M. mercenaria in Narragansett Bay is presented in Table 2. Based upon literature estimates of the standing crop of M. mercenaria in Narragansett Bay, there are, on average, three animals/nr (approx. 300 g/m2 whole animal weight) with patchy distribution through- out 389 km2 of the bay. In differing locations, average sizes range from 35 to 92 mm in valve length, based on the level of shellfish- ing. Higher levels of fishing effort for M. mercenaria are known to decrease the average valve lengths of a given subpopulation (e.g.. Rice et al. 1989). Rates of DFAA uptake by quahogs in this analy- sis ranged from 1.02 to 2.55 p.mol/h, depending upon average size. The total uptake of DFAA by the Narragansett Bay population of quahogs is estimated to be 2.08 kmoles/h or about 208 kg/h. and assuming 14 g of nitrogen per mole of DFAA, this represents an uptake of 2.9 1 kg of DON per hour. To compare the relative level of dissolved nitrogen uptake by quahogs to the filtration of particulate organic nitrogen, it is useful to estimate the filtration rates of phytoplankton and detritus. Key to this calculation is the estimation of the population filtration rates by quahogs. On average, the quahogs in Narragansett Bay filter 2.5 x 106 m3 of water per hour. With an average phytoplankton con- centration of 8,000 cells/mL and average PON associated with phytoplankton of 67.2 p-g/L, the bay-wide filtration of PON by quahogs if 14.6 kg/hr. Likewise, with average detritis PON of 17.8 p-g/L. the bay-wide population filtration is 3.88 kg/h. Thus, by way of comparison, the bulk of the organic nitrogen removal by qua- 550 Rice TABLE 1. Dissolved Free Amino Acids in Narragansett Bay in nmoles/L. Greenwich South Jamestown Conimicut Mean Amino Acid VVickford Bay Ferry Harbor Point Cone. SD Aspartate 149 103 91 88 152 116 31.4 Glutamate 131 87 82 65 135 100 31.2 Asparagine 9 3 3 4 5 4.8 2.48 Serine 201 173 171 118 198 172 33.3 Glutamine 3 1 1 1 2 1.6 0.89 Histidine 1 1 0 (I ] 0.6 0.54 Glycine 107 95 74 69 113 91 19.5 Arginine 62 48 31 18 56 43 18.2 Alanine 175 102 101 91 165 131 36.3 Tyrosine 2 1 1 1 1 1.2 0.45 Methionine 1 0 0 1 1 0.6 0.54 Valine 3 1 1 1 1 1.4 0.89 Phenylalanine 3 1 0 1 1 1.2 1.1 Isoleucine 2 1 0 0 1 0.8 0.84 Leucine 2 1 1 1 0 1.0 0.70 Lysine 1 0 0 0 1 0.4 0.54 Totals 852 618 557 478 833 667.6 167.3 hogs is by means of particulates, mostly phytoplankton, but on average about 14% of the organic nitrogen taken in by quahogs may be in form of DFAA. DISCUSSION Most work on the heterotrophic uptake and assimilation of DFAA in estuarine and coastal waters has focused on marine bac- teria (e.g., Jorgensen 1982. Carlucci et al. 1984. Carlucci et al. 1992, Keil and Kirchman 1991. Jorgensen et al. 1993). It is clear that uptake of DFAA by microheterotrophs is a major pathway for the removal of organic nitrogen from coastal waters, but dense assemblages of invertebrates, especially filter-feeding mollusks, are an important alternative pathway. It has been argued that mi- croheterotrophs have a competitive advantage over invertebrates in competition for DFAA in the environment, because they possess transport mechanisms that operate very efficiently at low (submi- cromolar) concentrations of DFAA (e.g., Sepers 1977). However, three lines of evidence show that marine invertebrates can effec- tively compete with bacteria for available DFAA. First, kinetic- studies have shown that environmentally realistic concentrations of DFAA in the hundreds of nanomolar to micromolar levels, oyster larvae can compete well against bacteria by virtue of higher mass-specific rates of transport at the higher DFAA concentrations (Manahan and Crisp 1982). Second, it has been demonstrated that amino acids are rapidly and efficiently removed in a single pass through the ctenidia of filter-feeding mollusks (Wright and Stephens 1978, Siebers and Winkler 1984). Finally, axenic sus- pensions of invertebrate larvae remove DFAA from seawater at rates comparable to those found in nonaxenic suspensions (Mana- han et al. 1983a, Davis and Stephens 1984). The data (Fig. 2) showing aspartic acid uptake rates from 1 mmol/L in the range of 1 to 2 u.mol/h by intact M. mercenaria correspond to dry weight specific rates of approximately 400-800 nmol/g*h. These data are in line with amino acid transport rates from such other bivalves as mytilids and oysters that have been more extensively studied (see review by Wright 1982). In a pre- vious study (Rice and Stephens 1988). we showed alanine trans- port by M. mercenaria occurring at only 22.9 nmol/g*h. This discrepancy in results between the experiments is a result of dif- ferences of their methods. In the present study, the quahogs were allowed to begin actively pumping before the uptake experiments commenced, and nonpumping quahogs were eliminated from the experiment. However, in the former study, a procedure of perfus- ing the mantle cavity was used to provide a steady water flow through the animals, but water was not necessarily passing through the ctenidia. The use of radiolabel to monitor the influx rates of amino acids into invertebrates is not an indicator of net amino acid flux ( Mana- han et al. 1983b). There is a possibility of efflux of amino acids from the animal that can only be monitored by chemical determi- nation of the concentration of amino acids in the medium by HPLC or other means. In a number of early studies (e.g.. Lum and Ham- men 1964, Hammen et al. 1966, Bayne and Scullard 1977) it was reported that bivalves can excrete nitrogen in the form of amino acids. In these early studies, bivalves were confined in vessels for several hours to obtain enough excretory products for adequate analytical detection and quantification. Later studies using shorter incubation times or flow-through systems along with analytical techniques with greater sensitivity have not shown excretion of DFAA or any other DON (e.g.. Jordan and Valiela 1982. Manahan et al. 1982. Siebers and Winkler 1984). Other studies have shown that excretion of amino acids occurs only in conditions of hypos- motic adjustment (Heavers and Hammen 1985. Rice and Stephens 1988). In the current study. I did not monitor the net flux of aspartic acid by directly following its change in concentration in the medium with time. However, I am confident that the influx of aspartic acid as represented by disappearance of radioactivity from the medium and its appearance in M. mercenaria soft tissues ad- equately reflects the net influx of aspartic acid. This is because we previously showed that influx of alanine as measured by depletion of radiolabel from the medium is tightly coupled with net flux of alanine as measured by HPLC when the quahogs are held in os- motically stable conditions with salinities above 17 ppt (Rice and Stephens 1988). DFAA Uptake in M. mercenaria 551 TABLE 2. Spreadsheet Model of DFAA Uptake and PON Filtration by Northern Quahogs. Est. Standing Mean DFAA Mean Geographic Area" Quahogs" Crop Sizec Uptake"1 Pop. Uptake1 FRf Pop FRB Pop FRh Segments (sq. km) (ind/sq m) Totals (mm) i in l ml i ml In (kg/DON/hr) (L/ind/hr) (kg-phvto-N/hrl (kg detrit-N/hr) Seekonk River 2.6 0 0 0 0 0 0 0 0 Providence River 21.2 12.5 2.65E+08 72 1.9 0.73 2.1 3.68 0.98 Upper Narragansett Bay 56.7 5.2 2.95E+08 48 1.3 0.56 1.4 2.77 0.73 Taunton River 12.2 0.1 I.22E+06 96 2.5 0.01 2.7 0.02 0.01 Mount Hope Bay 39.0 5 1 .95E+08 96 2.5 0.69 2.7 3.59 0.95 Greenwich Bay 13.0 5.4 7.01E+07 35 1.0 0.1 1.0 0.48 0.13 Upper West Passage 54.8 2.7 1 .48E+08 55 1.5 0.31 1.6 1.59 0.42 Upper East Passage 36.1 2.5 9.03E+07 59 1.6 0.21 1.7 1.04 0.27 Lower West Passage 47.6 1.2 5.71E+07 62 1.7 0.14 1.7 0.69 0.18 Lower East Passage 53.4 0.1 5.35E+06 96 2.5 0.02 2.7 0.1 0.03 Sakonnet River 52.7 1.1 5.79E+07 62 1.7 0.14 1.8 0.7 0.18 Narragansett Bay Total 389.3 3.0 1.18E+09 2.91 14.6 3.88 a Data on areas of Narragansett Bay geographic segments courtesy of Dr. Peter August of the URI Environmental Data Center. h Source of information on quahog populations include Stringer (1959), Saila et al. (1967), Pratt et al. (1992). Lazar et al. ( 1994). 1 Mean quahog sizes based on length-frequency analyses at individual geographic areas or estimates based on level of fishing effort (Pratt et al. 1992). d DFAA uptake rates calculated from data of present study. ' Calculation of DON uptake assumes 14 grams nitrogen per mole of DFAA. ' Filtration rate in L/h calculated from the allometric equation FR = 0.307L"t"'7 assuming valve length L in cm (Doering and Oviatt 1986). s Calculation of PON filtration assumes average phytoplankton densities of 8.000 cells/mL, mean phytoplankton cell dry weight of 0.28 ng/cell and nitrogen making up 3% of the cell dry weight (Durbin and Durbin 1981). Thus, average phytoplankton PON in Narragansett Bay is 67.2 u.g/L. h The calculation of rate of filtration of particulate organic nitrogen associated with detritus assumes an average detrital suspended particulate matter (SPM) of 3.7 mg/L (Morton 1972). organic dry weight comprising 12. lck of SPM (Oviatt and Nixon 1975), and nitrogen content comprising 3.8% of organic dry weight (Oviatt and Nixon 1975. Newell and Shumway 1993). Thus, average detrital PON in Narragansett Bay is 17.8 u.g/L. The spring concentrations of DFAA measured in Narragansett Bay waters ranging from 478 to 852 nmol/L (Table 1 ) are well within range of the DFAA concentrations found in other estuaries. Jorgensen ( 1982) reported that there were seasonal fluctuations in DFAA concentrations in the Kysing Fiord of Denmark ranging from minima of 200 nmol/L in summer and winter and maxima in the 700-900 nmol/L range in the spring and autumn, correspond- ing to peak phytoplankton blooms. Likewise, Poulet et al. ( 1985) found a significant correlation between various measures of pri- mary productivity and DFAA concentrations in the Morlaix Bay of Brittany. France, but DFAA concentrations were found to be highly variable and in the 1 to 4 u.mol/L range. Mopper and Lindroth (1982) sampled DFAA at an open water Baltic Sea sta- tion during the month of May and found a diel variation of DFAA concentrations ranging from 200 to 400 nmol/L in late evening and minimal concentrations of 30-40 nmol/L in the early day. Macko and Green ( 1982) carried out the most extensive study of DFAA in a New England estuary following the concentrations of 23 differ- ent amino acids for 1 year in the Damariscotta River estuary of Maine. That showed that DFAA concentrations fluctuated greatly throughout the year, but with a yearly average was 208 nmol/L. the dominant three amino acids being alanine, glycine, and serine. Based on results from these other estuaries, the May sampling in of Narragansett Bay reported in this study may represent maxi- mum DFAA levels, but this speculation awaits resolution pending more thorough sampling. The spreadsheet model shows that the uptake of DFAA by M. mercenaria may be nearly as important as filtered PON associated with detritus, which has been recognized as an important nutrition source for some bivalves (Newell and Shumway 1993). A weak- ness of this spreadsheet model is that it relies heavily on annual average estimates of PON and DFAA measurements taken during a single month. It is likely that, during different times of the year, the relative proportion of organic nitrogen removal in dissolved and particulate forms will exhibit strong seasonality. In addition, the relative magnitude of total organic nitrogen uptake will show strong seasonality because of the temperature dependency of filter feeding (e.g., Doering and Oviatt 1986). The total amount of organic nitrogen uptake by the population of M. mercenaria in Narragansett Bay should be viewed with caution. Although M. mercenaria are dominant filter feeders in Narragansett Bay. they are not the only ones. Other filter feeders in Narragansett Bay include such other bivalves as Mytilus editlis and Crassostrea virginica, the filter-feeding gastropod Crepidula fornicata, spionid polychaetes. ascidians. and others. However, the model is illustrative that uptake of DFAA by benthic filter feeders may be a pathway of benthic-pelagic nutrient coupling that is often overlooked. ACKNOWLEDGMENTS This study was funded by the Rhode Island Agricultural Ex- periment Station under project number H-879. This is external publication number 3610 of the College of Resource Development, University of Rhode Island. Thanks are extended to Dr. Peter August of the URI Environmental Data Center for providing GPS- GIS area measurements of Narragansett Bay. I also thank Dr. 552 Rice Terry Bradley of the URI Fisheries. Animal and Veterinary Sci- ence Department for use of the HPLC in his laboratory. I also thank Mr. David Johnson. URI Radiation Safety Officer, for the use of his liquid scintillation counter, and Dr. Candace Oviatt of the URI Graduate School of Oceanography for carefully reading through the manuscript and suggesting valuable improvements. LITERATURE CITED Bayne B. L. & C. Scullard. 1977. Rates of nitrogen excretion by species of Mytilus (Bivalvia:MolIusca). J. Mar. Biol. Assoc. U. K. 55:355-369. Bronk. D. A.. P. M. Gilbert & B. B. Ward. 1994. Nitrogen uptake, dis- solved organic nitrogen release, and new production. Science 265: 1843-1846. Carlucci, A. F., D. B. Craven & S. M. Henrichs. 1984. Diel production and microheterotrophic utilization of dissolved free amino acids in waters off southern California. Appl. Emir. Microbiol. 48:165-170. Carlucci. A. F.. D. M. Wolgast & D. B. Craven. 1992. Microbial popula- tions in surface fdms: amino acid dynamics in nearshore and offshore waters off Southern California. J. Geophys. Res. 94:5271-5280. Chien. P. K. & M. A. Rice. 1985. Autoradiographic localization of exog- enously supplied amino acids after uptake by the polychaete Glycera dibranchiata. Wasnumn J. Biol. 43:60-71. Davis, J. P. & G. C. Stephens. 1984. Uptake of free amino acids by bac- teria-free larvae of the sand dollar. Dendraster excentricus. Am. J. Physiol. 247:R733-R739. Davis, J. P., G. C. Stephens & M. A. Rice. 1985. Net entry of amino acids into the brittlestar Ophionereis annulata. Comp. Biochem. Physiol. 81A:899-903. Doering. P. H. & C. A. Oviatt. 1986. Application of filtration rate models to field populations of bivalves: an assessment using experimental me- socosms. Mar. Ecol. Prog. Ser. 31:265-275. Durbin, A. G. & E. G. Durbin. 1981. Standing stock and estimated pro- duction rates of phytoplankton and zooplankton in Narragansett Bay. Rhode Island. Estuaries 4:24 — 1 1 . Gomme. J. 1982. Epidermal nutrient absorption in marine invertebrates: a comparative analysis. Am. Zool. 22:691-708. Hammen, C. S„ H. F. Miller & W. H. Geer. 1966. Nitrogen excretion of Crassostrea virginica. Comp. Biochem. Physiol. 17:1199-1200. Heavers, B. W. & C. S. Hammen. 1985. Fate of endogenous free amino acids in osmotic adjustment of Crassostrea virginica (Gmelin). Comp. Biochem. Physiol. 82A:57 1-576. Jones. B N., S. Paabo & S. Stein. 1981. Amino acid analysis and enzymatic sequence determination of peptides by an improved o-phthaldialdehyde precolumn labeling procedure. J. Liquid Chromatog. 4:565-586. Jordan, T. E. & I. Valiela. 1982. A nitrogen budget of the ribbed mussel. Geukensia demissa, and its significance in nitrogen flow in a New England salt marsh. Limnol. Oceanog. 27:75-90. Jorgensen. N. O. G. 1982. Heterotrophic assimilation and occurrence of dissolved free amino acids in a shallow estuary. Mar. Ecol. Prog. Ser. 8:145-159. Jorgensen, N. O. G.. K. Mopper & P. Lindroth. 1980. Occurrence, origin, and assimilation of free amino acids in an estuarine environment. Ophelia 179-192. Jorgensen. N. O. G.. N. Kroer. R. B. Coffin. X.-H. Yang & C. Lee. 1993. Dissolved free amino acids, combined amino acids, and DNA as sources of carbon and nitrogen to marine bacteria. Mar. Ecol. Prog. Ser. 98:135-148. Keil, R. G. & D. L. Kirchman. 1991. Contribution of dissolved free amino acids and ammonium to the nitrogen requirements of heterotrophic bacterioplankton. Mar. Ecol. Prog. Ser. 73:1-10. Lazar. N., A. Ganz & A. Valliere. 1994. Quahog stock assessment and implementation of an interim managemenl plan in Greenwich Bay, Rhode Island, pp. 5-29. In M. A. Rice and E. Gibbs (eds.). Proceedings of the Third Rhode Island Shellfish Industry Conference, Pub. RIU- W-94-001. Rhode Island Sea Grant, University of Rhode Island, Nar- ragansett, RI. Lindroth. P. & K. Mopper. 1979. High-performance liquid chromato- graphic determination of subpicomolar amounts of amino acids by pre-column derivatization with o-phthaldialdehyde. Analyt. Chem. 5 1 : 1667-1674. Lum, S. C. & C. S. Hammen. 1964. Ammonia excretion of Lingula. Comp. Biochem. Physiol. 12:185-190. Macko. S. A. & E. J. Green. 1982. An investigation of dissolved free amino acids and their relation to phytoplankton cell density in the Dama- riscotta River Estuary, Maine. Estuaries 5:68-73. Manahan. D. T. & D. J. Crisp. 1982. The role of dissolved organic material in the nutrition of pelagic larvae: amino acid uptake by bivalve veligers. Am. Zool. 22:635-646. Manahan. D. T. & D. J. Crisp. 1983. Autoradiographic studies on the up- take of dissolved amino acids from seawater by bivalve larvae. J. Mar. Biol Assoc. U. K. 63:673-682. Manahan. D. T. S. H. Wright. G. C. Stephens & M. A. Rice. 1982. Trans- port of dissolved amino acids by the mussel. Mytilus edulis: demon- stration of net uptake from seawater by HPLC analysis. Science 215: 1553-1255. Manahan. D. T.. J. P. Davis & G. C. Stephens. 1983a. Bacteria-free sea urchin larvae: selective uptake of neutral amino acids from seawater. Science 220:204-206. Manahan, D. T. S. H. Wright & G. C. Stephens. 1983b. Simultaneous determination of net uptake of 16 amino acids by a marine bivalve. Am. J. Physiol. 244:R832-R838. Mopper. K. & P. Lindroth. 1982. Diel and depth variations in dissolved free amino acids and ammonium in the Baltic Sea determined by ship- board HPLC analysis. Limnol. Oceanog. 27:336-347. Morton, R. W. 1 972. Spatial and temporal distribution of suspended sedi- ment in Narragansett Bay and Rhode Island Sound, pp. 131-141. In B. W. Nelson (ed.l. Environmental Framework of Coastal Plain Estu- aries. Memoirs 133. Geological Society of America. Boulder. CO. Newell. C. R. & S. E. Shumway. 1993. Grazing of natural particulates by bivalve molluscs: a temporal and spatial perspective, pp. 85-148. In R. F. Dame (ed.). Bivalve Filter Feeders in Estuarine and Coastal Eco- system Processes. NATO ASI Series, Vol. G33. Springer Verlag. Berlin. O'Dell. S. J. & G. C. Stephens. 1986. Uptake of amino acids by Pareury- thoe californica: substrate interaction modifies net influx from the environment. Biol. Bull. 171:682-693. Oviatt, C. A. & S. W. Nixon. 1975. Sediment resuspension and deposition in Narragansett Bay. Estuar. Coast. Mar. Sci. 3:201-217. Pajor, A. M. & S. H. Wright. 1987. Characteristics of L-alanine uptake in apical membrane vesicles isolated from marine bivalve gill epithelium. J. Memb. Biol. 96:209-223. Pequignat. E. 1973. A kinetic and autoradiographic study of the direct assimilation of amino acids and glucose by organs of the mussel Myti- lus edulis. Mar. Biol. 19:227-244. Poulet. S. A.. V. Martin-Jezequel & D. Delmas. 1985. Gradient of dis- solved free amino acids and phytoplankton in a shallow by. Hydrobio- logia 121:11-17. Pratt. S. D.. A. R. Ganz & M. A. Rice. 1992. A species profile of the quahog in Rhode Island. Rep. RIU-T-92-00 1 . Rhode Island Sea Grant, University of Rhode Island. Narragansett. RI. 1 16 pp. Preston. R. L. & B. R. Stevens. 1982. Kinetic and thermodynamic aspects of sodium-coupled amino acid transport by marine invertebrates. Amer. Zool. 22:709-721. Putter. A. 1909. Die Ernahrung der Wassertiere und der Stoffhaushalt der Gewasser. Gustav Fischer Verlag, Jena, Germany. Rice, M. A. 1992. Overview of quahog management studies in Narragan- sett Bay, 1946-1992. pp. 49-61. In M. A. Rice and D. Grossman- Garber (eds.). Proceedings of the Second Rhode Island Shellfish In- DFAA Uptake in M. mercenaria 553 dustry Conference. Rhode Island Sea Grant pub. RIU-W-92-002. Uni- versity of Rhode Island. Narragansett. Rl. Rice. M. A. & G. C. Stephens. 1987. Uptake and internal distribution of exogenously supplied amino acids in the Pacific oyster. Crassostrea gigas (Thunberg). Aquaculture 66:19-31. Rice, M. A. & G. C. Stephens. 1988. Influx, net flux, and transepithelial flux of amino acids in the hardshell clam Mercenaria mercenaria (Linne): influence of salinity. Camp. Biochem. Physiol. 89A:63 1-636. Rice. M. A.. C. Hickox & I. Zehra. 1989. Effects of intensive fishing effort on the population structure of quahogs. Mercenaria mercenaria (Lin- naeus 1758) in Narragansett Bay. J. Shellfish Res. 8:345-354. Saila. S. B.. J. M. Flowers & M. T. Canario. 1967. Factors affecting the relative abundance of Mercenaria mercenaria in the Providence River. Rhode Island. Proc. Nat. Shellfisher. Assoc. 57:83-89. Sepers. A. B. J. 1977. The utilization of dissolved organic compounds in aquatic environments. Hydrohiologia 52:39-54. Siebers. D. & A. Winkler. 1984. Amino acid uptake by mussels. Mylilus edulis, from natural sea water in a flow-through system. Helgolander Meeresunters 38:189-199. Stephens, G. C. 1981. The trophic role of dissolved organic material, pp. 271-291. In A. L. Longhurst (ed.). Analysis of Marine Ecosystems. Academic Press, San Diego. CA. Stephens, G. C. 1982. Recent progress in the study of "Die Ernahrung der Wassertiere und der Stoffhaushalt der Gewasser." Am. Zool. 22:611- 619. Stephens, G. C. 1988. Epidermal amino acid transport in marine inverte- brates. Biochim. Biophys. Ada 947:1 13-138. Stephens, G. C. & D. T. Manahan. 1984. Technical advances in the study of nutrition of marine mollusks. Aquaculture 39:155-164. Stephens. G. C. & R. A. Schinske. 1961 . Uptake of amino acids by marine invertebrates. Limnol. Oceanogr. 6:175-181. Stringer, L. D. 1959. The population abundance and affect of sediment on the hard clam. In Hurricane Damage Control. Narragansett Bay and Vicinity, Rhode Island and Massachusetts. Appendix E: A Detailed Report on Fishery Resources. U.S. Fish and Wildlife Service. Wash- ington. DC. 17 pp. Waksman. S. A. & C. L. Carey. 1935. Decomposition of organic matter in seawater by bacteria. I. bacterial multiplication in stored seawater. J. Bacteriol. 29:531-543. Wright. S. H. 1982. A nutritional role for amino acid transport in filter feeding marine invertebrates. Am. Zool. 22:621-634. Wright, S. H. 1985. Multiple pathways for amino acid transport in Mylilus gill. J. Comp. Physiol. 156:259-267. Wright. S. H. & D. T. Manahan. 1989. Integumental nutrient uptake by aquatic organisms. Ann. Rev. Physiol. 51:585-600. Wright, S. H. & T. W. Secomb. 1984. Epidermal taurine transport in ma- rine mussels. Am. J. Physiol. 247:R346-R355. Wright, S. H. & G. C. Stephens. 1978. Removal of amino acid during a single passage of water across the gill of marine mussels. J. Exper. Zool. 205:337-351. Journal oj Shellfish Research. Vol. 18. No. 2. 555-560. L999. PERFORMANCE OF A TIDAL-POWERED UPWELLING NURSERY SYSTEM FOR NORTHERN QUAHOGS (HARDCLAMS) (MERCENARIA MERCENAR1A) IN SOUTH CAROLINA NANCY H. HADLEY,1 ROBERT B. BALDWIN,2 M. R. DEVOE,3 AND R. RHODES1 1 South Carolina Department of Natural Resources. Marine Resources Division, Charleston. South Carolina Lowcountry Seafood, McClellanville, South Carolina 3South Carolina Sea Grant Consortium, Charleston. South Carolina ABSTRACT Entry into the hard clam aquaculture industry on a small scale has been limited by the cost of plantable seed. Growers must either purchase expensive seed at a suitable planting size (usually 8 mm SL or larger) or raise small seed to this size in a nursery. Land-based nurseries foster high survival and rapid growth, but require expensive waterfront property and are energy- and labor- intensive to operate. Field-based nurseries are inexpensive to operate, but seed survival is often very low and success is site-specific. Floating upwelling systems (FLUPSYs) combine many of the advantages of land-based systems (high survival, rapid growth) with those of field-based systems (inexpensive operation). One particular type of FLUPSY. a tidal-powered upwelling system (TPU). is described here. TPU performance was tested in South Carolina over a 5-year period. Tidal currents averaging 0.53 m/s produced flow rates of 52 Lpm through upwelling bins. The TPU requires a mean current velocity of 0.26 m/s to produce a flow rate of 26 Lpm through the upwelling units, which is comparable to flow that land-based upwellers provide. Daily growth rates as high as 15% were observed. Growth in the TPU was more rapid than published reports of growth rates in land-based systems in South Carolina. The TPU described here can produce 122.000 (12 mm) seed, starting from 1-mm seed, in a 32-week growing season (March through October). Capacity is considerably higher (up to 1 .074,000) if seed are stocked at a larger size and/or harvested at a smaller size. The described system, built for $4,500 and with annual operating costs of less than $5,000, is a cost-effective nursery system that small-scale growers could implement to produce field-plantable seed. KEY WORDS: Mercenaria mercenaria. nursery culture, seed production, quahog. hardclam. aquaculture, FLUPSY INTRODUCTION The growout of clams to market size is a field-based operation that relies on the natural productivity of coastal waters to meet the high feeding requirements necessary to foster rapid growth. How- ever, to achieve high survival in the growout phase, the seed must be relatively large (8-12 mm) when planted. Nursery systems are used to raise hatchery-produced seed to the larger sizes needed for field growout. Bivalve nurseries have traditionally been located on waterfront property allowing seawater to be pumped to raceways or upwelling systems that contain the seed clams. Such land-based nurseries are expensive to operate and require access to ever scarcer and more expensive waterfront property. Alternatives to land-based systems include pond or impoundment-based nurseries (Manzi et al. 1988, Battey and Manzi 1990. Hopkins et al. 1993. Bayes 1981) and field nurseries (Vaughan 1988, Castagna 1984. Kraeuter et al. 1998, Flimlin and Kraeuter 1997, Walker et al. 1995). Pond-based nurseries suffer from many of the same prob- lems as land-based nurseries, in that they must be located near a source of clean, high-salinity seawater. Additionally, considerable expertise is required to maintain dense phytoplankton blooms re- quired for rapid growth. Field nurseries place the seed in naturally productive waters under more protected conditions than are re- quired for growout. Such nurseries employ various types of pro- tective devices (nets, wire cages, sand trays, soft bags, etc.) to deter predators. These devices require frequent maintenance to remove fouling organisms and silt (Manzi and Castagna 1989, Kraeuter et al. 1998), and even under optimal conditions they must be stocked Contribution No. 429 from the South Carolina Marine Resources Center, Charleston. South Carolina. at relatively low densities to achieve acceptable growth and sur- vival. Floating or suspended culture has been used for many years, usually in the form of suspended trays or nets (Manzi and Castagna 1989). Bayes (1981) described an upwelling system nursery lo- cated in an impoundment that offered protected conditions and rapid growth. Recently, floating upwelling systems (FLUPSYs) have been developed to capitalize on the features of land-based upwellers while avoiding the expense of waterfront property. A number of powered FLUPSYs have been designed and tested in recent years (Rivara and Bavaro 1997, Bishop 1998; RaLonde 1994. K. Brown, pers. comm.). These systems use airlifts, pumps, water wheels, or other electrical devices to move water through the system. However, one system, in use at Mook Seafarms. Dama- riscotta. Maine, uses tidal currents to provide water flow through the upwelling units (Mook 1988. Mook and Johnson 1988). Our tidal-powered upwelling (TPU) system was modeled after the Mook Seafarms system (Baldwin et al. 1994). This paper describes the performance of the South Carolina TPU. including flow rates, seed stocking densities, growth rates, production capacity and comparisons with other nursery systems. METHODS System Description A tidal powered upwelling system (TPU) was constructed fol- lowing plans drawn by Mook (1988). modified to have a more efficient scoop and shallower draft. The system construction is described in detail elsewhere (Baldwin et al. 1994). The TPU consists of a raft or floating dock (6 m x 3.6 m) with a tank in the center (4.8 m x 1.2 m) that houses 16 square upwelling bins, each 555 556 Hadley et al. providing bottom area of 0.23 m2 (Fig. 1). Each bin is 0.46-m deep, with window screen or other appropriate mesh attached to the bottom. The front of the raft has an open scoop that traps tidally-driven water and directs it into the tank. The water rises through the screens of the upwelling bins (which retain the seed clams) and exits through 7.6-cm PVC pipe located below the water line. The bins are arranged in 2 rows of 8 discharging into a common trough. The TPU has a single point mooring on the scoop end of the raft so that the scoop always faces the direction of tidal flow. The system was deployed in a tidal creek in the Cape Ro- maine Wildlife Refuge near McClellanville, South Carolina. Tidal Current Velocity and Flow Rate Determination Stocking densities for upwelling systems are expressed in terms of flow ratio: the volumetric water flow rate past a given volume of seed clams, usually expressed as liters of water per minute per liter of clams (Lpm-L-1). The actual quantity of seed that can be supported in an upwelling system is a function of the available water supply, and also varies with seed size, smaller seed requiring a greater flow ratio than larger seed. In land-based systems, maxi- mum water supply is a fixed quantity, and can only be altered by increasing the size of the pumps used. In tidal-powered systems, water supply is determined by the tidal currents at the deployment site and is not under operator control, except by relocating the system. Thus, in both types of upwelling system, higher flow ratios can be attained only by reducing the stocking density of seed clams. Consequently, before determining appropriate stocking densities for the tidal upweller. it was necessary to determine tidal current velocity at the deployment site and the resulting water flow rate through the upwelling bins. Tidal current velocity measurements were made with a Mont- edoro-Whitney digital current meter reading in ft/sec. These mea- surements were supplemented with additional estimates attained by timing the passage of a weighted cork past the length of the raft. Current speed was measured at 15-minute intervals over several tidal cycles on different days and plotted against tidal stage, ex- pressed in hours from slack tide. Mean tidal current velocity was estimated from this plot. Water flow rates through the upwelling system were calculated from measurements of effluent current velocities. It was not pos- sible to measure currents exiting from individual bins because of space constrictions. Therefore effluent current velocity was mea- sured in the common exit trough. These measurements, made con- currently with tidal velocity measurements, were then regressed against tidal current velocity with the intercept forced through the Working deck (partially cut away) To mooring Flotation (partially cut away) Outflow trough (front section cut away) Tidal scoop Figure 1. Diagrammatic representation of the tidal-powered up- welling system (from Hadley et al. 1997). origin. This regression equation was then used to predict mean effluent velocity from the previously estimated mean tidal current velocity. Mean effluent velocity was multiplied by cross-sectional area of the effluent trough to yield mean total effluent flow volume (Lpm) that was divided by the number of upwelling bins contrib- uting to the trough (8) to estimate volumetric flow rate through each bin. Stocking Densities A series of experiments was conducted to determine appropri- ate stocking densities for seed clams from 1 to 8 mm shell length. Two experiments were conducted in fall 1992, 4 between April and September 1993, 2 in summer 1996. and 6 between June and October 1997. Some of these experiments were conducted in one- quarter-scale experimental upwelling bins with discharge pipes constricted to provide water flow per unit area comparable to the production bins. A comparison of the experimental and production bins was made to verify that growth determined in the scaled-down version could be extrapolated to that in the production version (Fig. 2). For each of the seed sizes tested, replicate bins were stocked with a predetermined volume of seed to yield a desired flow ratio (Lpm-L "' ). Target flow ratios were based on published requirements for a land-based upwelling nursery in South Carolina (Manzi et al. 1984). For each seed size, 3 flow ratios were tested. Biovolume was determined at 2-week intervals and, if volume had doubled in any of the treatments, the seed were sieved and the upwellers were restocked with uniformly sized seed at a new den- sity appropriate to that seed size. A subsample of seed from each density was measured before sieving and a subsample of the pooled sieved individuals was measured before restocking. If seed had not doubled in any of the flow treatments, the experiment was continued for an additional week. Biovolume increases were converted to daily growth rates by the formula: DGR = (Fin vol/Init vol)1 1 (1) Daily growth rates from all 16 experiments were regressed against log-transformed flow ratios (Lpm-L"'). Within size classes, growth rate at different flow ratios was compared by ANOVA. Growth Determinations and Growth Comparisons Growth of different size seed at appropriate predetermined stocking densities was determined in multiple growth trials span- ning different seasons and years (1992 to 1997). Growth of 4-mm or o o 10% 8% 6% 4% 2% 0% □ Experimental □ Production 130 87 43 Flow Ratio (Lpm-L1 clams) Figure 2. Daily growth rate (DGR) of 4-mm clams at 3 flow ratios in experimental (0.05 nr) and production (0.23 nr) bins. Tidal-Powered Upwelling Nursery 557 seed was tested in each year and used as a baseline for adjusting for interannual variation in growing conditions. Growth of different size seed in the TPU was compared to growth in bottom trays (0.6 m x 1.2 m). landbased upwellers (Manzi et al. 1984. Manzi et al. 1986). and land-based raceways (Hadley and Manzi 1984). Biomass increases were converted to daily growth rates (DGR) to allow comparisons between trials of different duration. Where necessary, shell lengths for the published data sets were estimated by the formula: x SL (mm))* Volume (mL) (0.07: (2) Bottom-tray experiments were deployed concurrently with TPU trials and were analyzed by the paired comparisons test, treating dates as blocks. Comparison with land-based nurseries was based on published data (Manzi et al. 1984, Hadley and Manzi 1984. Manzi et al. 1986). Because of the large number of assump- tions that had to be made to compare these data sets, statistical analysis was not attempted. Production Capacity A model was developed to project seasonal production of the tidal-powered upweller (as built) for two stocking sizes (1 mm. 4 mm) and 3 harvest sizes (8 mm, 10 mm. and 12 mm). The com- puter model assumes a biweekly sieving and redistribution and factors in expected daily growth rates and expected mortality drawn from empirical observations. At each 2-week interval, the proportion of individuals expected in each size class produced by sieving is calculated from the properties of the normal distribution, using empirically determined standard deviations of different size clams. Our data sets, spanning many cohorts, seasons, years, and sizes, indicate that individual shell lengths for sieved populations were normally distributed in all size classes tested (4 mm to 12 mm SL). The production model assumes a 32-week growing season, and that multiple cohorts are stocked as space becomes available. RESULTS Tidal Current Velocity and Flow Rate Determination When plotted against tidal stage, tidal current velocity closely resembled a sine wave. Mean tidal velocity was therefore calcu- lated using the formula for the average value of the sine function: 2/tt x V„ (3) mulative flow through the effluent canal of 416 liters per minute (Lpm), or approximately 52 Lpm through each of the 8 upwelling bins. Actual flow rates measured were generally higher than this, so this may be considered a conservative estimate. Expected volu- metric flow rates for this system at various tidal current velocities are listed in Table 1. Each TPU will have unique characteristics that will influence the relationship between tidal current velocity and flow rate through the upweller system. Stocking Densities For each size class, 3 flow ratios were tested, beginning with the recommended flow for a land-based upweller and increasing that by 50% and 100%. The results of these experiments are sum- marized in Figure 3. Regression of DGR on log-transformed flow ratios was significant: where Vmax is the maximum velocity. Maximum tidal current velocity measured in this tidal creek was 0.84 m/s. yielding a mean tidal current velocity of 0.53 m/s. Flow rates through the upwelling system were more problem- atic to determine. The weight of a person on the raft causes it to tilt enough to change the head pressure and thus the flow. Wave action also causes the flow rate to vary. Friction in the relatively small effluent canal also appeared to influence current meter readings. Nonetheless, effluent flow rate regressed against tidal current ve- locity (measured simultaneously) was significant (r = 0.78, P < .01): Effluent flow rate (m/s) = Tidal current velocity (m/s) x 0.3785 (4) Substituting the previously determined mean tidal current ve- locity (0.53 m/s) in equation 4 yields a mean effluent flow rate of 0.20 m/s. Multiplying by the cross-sectional area of the effluent canal (0.033 m2) and converting to liters per minute yields a cu- DGR = 0.0284Ln(Flow ratio) = 0.7334, P < .001) 0.0673 ( R~ (5) At flow ratios equivalent to those recommended for land-based upwellers. growth was similar to published reports for land-based nurseries in South Carolina. However, increasing the flow ratio by 50% (to 1.5 times the ratio used in land-based nurseries) resulted in a substantial growth increase. Increasing the flow ratio by 100% sometimes but not always resulted in a further growth increase. To achieve this highest flow ratio, stocking densities must be low (about 80% of the stocking density for a land-based upweller) and the faster growth does not compensate for the reduced carrying capacity. Thus, we selected the middle flow ratio (50% higher than that recommended for land-based nurseries) to use in our growth trials and our production model. Recommended flow ratios for different size seed in a tidal upweller and land-based upweller (Manzi et al. 1984) are presented in Table 2. Growth Rate Determination and Comparisons with Other Nursery Systems Growth rates for different size seed varied between seasons as well as between years. Mean growth rates from different experi- ments and the grand mean for each size class are given in Table 3. In 1 997, growth of 4-mm seed was 407r below the average for all other years. For the purposes of modeling production capacity, TABLE 1. Expected mean flow rate (Lpm I through each upwelling bin of the TPU at different tidal current velocities. Average Current Velocity Flow Rate (m/s) (f/s) (knots) (Lpm) 0.15 0.50 0.30 9.60 0.23 0.75 0.44 14.39 0.30 1.00 0.59 19.19 0.38 1.25 0.74 23.99 0.46 1.50 0.89 28.79 0.53 1.75 1.04 33.59 0.61 2.00 1.18 38.39 0.69 2.25 1.33 43.18 0.76 2.50 1.48 47.98 0.84 2.75 1.63 52.78 0.91 3.00 1.78 57.58 0.99 3.25 1.93 62.38 558 Hadley et al. 8 5 4 mm mm 10% 9% 8% 7% A 6% A IT n 5% a 4% • A, A* 3% < ► T ' A AA 2% • / i * A 1% f i* 0% 2 mm 1 mm DGR = 0 0284Ln(Flow ratio) - 0.0673 R2 = 0.7334 A ■ 1 A " '^^■— — — "■" X — 8 ■ ■ * -" A #8-9 mm A 5-6 mm A 4 mm ■ 2 mm X 1-2 mm 25 50 75 100 125 150 175 200 225 250 275 Flow ratio (Lpm-L ) Figure 3. Daily growth rate (biomass increase per day, DGR) of different size seed clams in a tidal-powered upweller as a function of flow ratio. Vertical lines indicate recommended flow ratios for indicated seed size. 1997 growth rates were adjusted upward by 40%. The original and adjusted growth rates are shown in Table 3. Tidal Upweller vs. Bottom Trays Growth rate comparisons between the TPU and bottom trays are summarized in Table 4. Seed grew significantly faster in the bottom trays (P = .048) but growth in both systems was remark- ably fast, with daily growth rates as high as 18.9% in the bottom trays and 14.9% in the TPU for 1-mm seed. Growth rates of seed this size in land-based upwellers are estimated from published data to be no higher than 6.9%. For all size classes tested volume at least doubled in 2 weeks (DGR £5.5% ) as long as temperatures were higher than 20 °C. Seed produced in the upwellers tended to be more uniform in size than those produced in trays. Tidal Upweller vs. Land-based Nurseries At flow ratios recommended for land-based upwellers, clams grew faster in the TPU than in the land-based upwelling system and as fast as in land-based raceways (Table 5). At higher flow ratios, 50% higher than those used in land-based systems, even more rapid growth was obtained, with the exception of the 1-mm seed. The difference in the 1-mm seed growth rates was probably TABLE 2. Recommended flow ratios (Lpm-L-1 clams) for different size clam seed in tidal and land-based upwellers. Size (mm) Tidal I Lpm-L" Land" (Lpm-L -1 180 120 10(1 60 45 35 27 18 12 120 80 70 40 30 23 18 12 9 seasonal. Growth rates in the TPU at the lower flow ratio ranged from 14.9% DGR for 1-mm seed to 2.5% DGR for 5-6 mm seed. Growth rates in a land-based upweller were calculated to be 6.9% for 1-mm seed and 1.4% for 3-mm seed, based on data in Manzi et al. ( 1984). At the higher flow rates, growth in the TPU ranged from 8.2% DGR for 1- and 2-mm seed to 3.0% DGR for 5-6 mm seed. Not enough information is available on the land-based tests to permit statistical comparisons. TABLE 3. Daily growth rates (DGR) (% biomass increase per day) of different size seed in the TPU at recommended flow ratios. For trials performed in 1997, adjusted growth rates appear in parentheses. Size Date DGR (%) SE 1 Data from Manzi et al. 1984 1 mm Grand mean 2 mm Grand mean 3 mm Grand mean 4 mm Grand mean 5-6 mm Grand mean 8-9 mm Grand mean July 93 March 94 Sept 93 July 93 April 94 Aug 93 Oct 93 June 96 Oct 93 May 93 June 97 Aug 97 June 96 June 97 Sept 97 July 97 Sept 97 14.86 7.68 8.63 10.39 7.74 8.70 8.22 7.74 5.65 6.69 5.93 1.67 8.91 4.49 2.78 4.75 3.84 2.12 3.18 3.04 1.07 2.65 1.86 (6.29) (3.90) (5.34) (2.97) (4.45) (3.75) (1.50) (3.72) (2.61) 0.47 0.34 0.15 2.25 0.00 0.22 0.48 0.50 0.27 1.05 0.00 0.04 0.28 0.20(0.28) 0.05 (0.07) 2.83(1.22) 0.21 0.07(0.10) 0.03 (0.05) 0.05(0.43) 0.13(0.19) 0.22(0.31) 0.79(1.11) Tidal-Powered Upwelling Nursery 559 Production Capacity Seasonal production capacity of the TPU for different size seed was estimated from a computer model using recommended flow ratios, empirically determined growth rates and standard devia- tions for different size seed, and estimates of mortality based on empirical observations. When stocked with 4-mm seed, the pilot TPU can theoretically produce 4 cohorts of 8-mm seed (totalling 1.074,000). 2 cohorts of 10-mm seed (459.000). or 2 cohorts of 12-mm seed (244.000) in a 32-week growing season (Table 6). When stocked with 1-mm seed, the system is projected to produce 2 cohorts of 8-mm seed (591.000), 2 cohorts of 10-mm seed (426.000). or a single cohort of 12-mm seed ( 122.000) in a grow- ing season. DISCUSSION AND CONCLUSIONS Mook (1988) reported maximum flow rates in his tidal powered upweller of 38 Lpm at a tidal current velocity of 0.38 m/s, and minimum flow rates of 14 Lpm at a current velocity of 0.1 m/s. Our system is predicted to have comparable flow rates at the higher current velocity, but slightly lower flow rates at the low current velocity (Table 1 ). Since currents as low as 0.1 m/s were not measured, and our regression forced a zero intercept, the pre- dicted flow rates at low current speeds are conservative. The simi- larity in the flow rates of the 2 systems at operational current speeds suggests that flow rate for a TPU may be reliably predicted from tidal current velocity. Flow rates comparable to a land-based nursery (30-35 Lpm) are predicted at mean tidal current velocities of 0.25-0.35 m/s (Table 1). Based on tidal current tables (International Marine 1998). it appears that velocities of this magnitude would be avail- able along a large portion of the eastern seaboard. Only a few stations in the mid-Atlantic portion of the coast have maximum current velocities less than 0.4 m/s (the maximum current velocity needed to produce a mean current velocity of 0.25 m/s). A survey of regulatory agencies indicates that a tidal powered upweller would be permissible in all states where it would be operable (DeVoe and Nelson in prep.). Out of 18 coastal states surveyed, 15 indicated that a TPU would be allowed, although the number and type of permits required varied widely. The 3 states that would not allow a TPU also indicated that it would not be feasible because of slow current velocities. Growth rates in the tidal upweller were not as good as in TABLE 4. Daily growth rates (DGR) (% biomass increase per day) and final sizes of different size seed clams grown for 2 weeks in the TPU and in bottom travs. bottom trays deployed in the same creek (Table 4). However, the bottom tray culture system is labor-intensive to operate, requires frequent cleaning, and has a much lower production capacity than the TPU. The more rapid growth in the bottom tray may be due to the very low stocking density. The bottom trays were stocked at their known optimal density, rather than at densities comparable to those used in the TPU. The tidal upweller produced better growth rates than published reports for land-based upwellers (Manzi et al. 1984) and raceways (Hadley and Manzi 1984) in South Carolina (Table 5), particularly at the higher flow ratios recommended. Since the comparisons are made with data from previous studies, rather than with concurrent data, we cannot rule out the influence of interannual variation in growing conditions or the possibility that the clams used in the current study were naturally faster growing. The clams used in this study are descendants of those used in the published studies, which were hatchery reared seed bred for rapid growth. However, a num- ber of generations separate the seed cohorts and additional selec- tion may have occurred during this time span. It is interesting to note that the flow rate through the bins of the tidal powered up- weller was approximately 509c higher than through the land-based nursery with which it was compared. Thus, using the 50% higher flow ratio actually results in stocking the same density of clams per bin as in the land-based system. Other types of FLUPSYs used for growing bivalve seed range from small airlift-powered systems comparable in capacity to the TPU described here (RaLonde 1994, Rivara and Bavaro 1997), to large commercial facilities powered with pumps and having total flow rates of 6,000 gallons per minute (Bishop 1998). However, a TPU is the only type of FLUPSY which requires no external power source, so it can be located in remote areas where a powered system might not be feasible. Production data from other FLUPSYs have not been reported, so we are unable to compare the performance of the TPU with other FLUPSYs. Any performance differences would probably be attributable to the higher water flow that may be achieved with an externally powered system. The system described here was built for a total cost of $4,500 (1994 dollars) and operating costs were less than $5,000 per year (Baldwin et al. 1994, Baldwin et al. 1995). This is considerably more economical, both in construction and operation, than tradi- tional land-based systems (Baldwin et al. 1995) and compares TABLE 5. Daily growth rates (), suggesting that temperature variations alone cannot account for the cycles in the estuarine shell. The second major source of 8180 variation is mixing of sea- water with freshwater. Great variability occurs in the composition of coastal marine waters, which are commonly around 0-2%p rela- tive to SMOW, although can reach +4%o in hypersaline areas (Lloyd 1969) because of evaporative loss of H2I60. Depletion (of lsO) occurs because of the mixing of coastal precipitation and continental runoff waters and 8lsO values can be as low as -6%c (Lloyd 1964). Seasonal mixing of seawater and freshwater will cause depletion in 8I80 in the estuarine shell during annual floods. Peaks and troughs in the 8lsO data match with the color banding that has been interpreted as caused by seasonal differences (Agu- irre et al. 1998). The majority of the 8180 cycles in Figure 2a are characterized by an abrupt decrease in 8lsO within a 1-mm inter- val between two adjacent samples. This must reflect an abrupt change in environmental conditions from summer (warm season) to winter (cold season), perhaps representing the autumn flood or differential seasonal growth rates. The subsequent gradual rise in 8'xO values probably reflects the general decrease in supply of freshwater through the winter. It is interesting to note that the average 813C and 8,80 of the two shells are significantly higher in the marine specimen. It seems likely that the most enriched 8I80 values from the estuarine shell (maximum 8lsO for the five peaks in zones 2, 4, 6, 8, and 9 = 0.6 ± 0.6%c) represents a return to more typical marine conditions, because the mean 8180 for the marine shell is within error at 0.8%o. The variation in l3C/l2C ratios in shell carbonate is dependent on the source and amount of marine bicarbonate. Bicarbonate is derived from photosynthesis and respiration by marine organisms, the addition of terrestrial organic material, and exchange with atmospheric COv 813C is relatively insensitive to temperature changes. Typical 8'3C values for coastal marine carbonate precipi- tated in isotopic equilibrium with ambient seawater generally fall between -2 and +4%o. In the estuarine shell, depleted 813C values are coincident with depleted 8I80 values and are directly related to freshwater input, which brings a greater amount of l2C from ter- restrial plant material, via the Rio de la Plata river during winter flood events. This is consistent with other studies (e.g.. Klein et al. 1996). which have shown that the l3C content of total dissolved inorganic carbonate (TDIC) of marine water in coastal settings is directly related to salinity. The mean 813C value from the marine shell ( = 2.1%o) is typical of coastal marine waters. Trace Elements In most calcareous shells, it is unclear where the trace elements reside or in what form they are present (Rosenberg 1980). They could occur in the mineral or organic portions, in shell inclusions associated with detrital clay particles (Eisma et al. 1976). or ab- sorbed onto the shell surface (Foster and Chacko 1995). The amount of elemental incorporation into shells is dependent on a variety of factors, including such environmental effects as salinity and temperature, although studies by Lorens and Bender (1980), Lorens ( 1981 ). and Mucci ( 1986), among others, have shown that kinetic and metabolic effects can exert a greater influence. For example, fast shell growth can result in Sr/Ca ratios that exceed Variations in the Shell of Mactra 573 TABLE 1. Mean isotope and trace element concentration data from the estuarine and marine Mactra shell. Mean Value Marine Mean Value (1 SD) Estuarine Mean Value (1 SD) Estuarine Max. Value Estuarine Min. Value 5I80 (%<■) &"c c;,) Sr (ppm) Mg (ppm) Ba (ppm) Mn (ppm) Zn (ppm) Ph (ppm) Cd (ppm) 0.8 (±0.6) 2.1 (±0.3) 2601 567 25 2.7 17 2.2 0 -0.9 (±1.4) 0.0 (±0.7) 1819 183 111 0.6 31 1.5 0.5 1.5 1.6 2506 339 21 1.7 -2.9 -1.4 41 equilibrium values, and higher rates of metabolic pumping of Ca can result in lower than average Sr/Ca (Amiel et al. 1973). In Mactra, there is no statistically significant correlation be- tween isotopes and concentration of trace elements in either the marine or estuarine shell along incremental growth intervals, sug- gesting that environmental influences are not the only cause of trace element variation. Interestingly, there is a good visual match between isotopes and Sr. Mg. and Ba in the estuarine shell. The mean trace element concentrations in the marine shell are all much higher than those of the estuarine shell. The lowest concentrations of trace elements in the estuarine shell coincide with the lowest 8' 3C and 5lsO values, which suggests that addition of river water is the predominant, but probably not the only, environmental con- trol. Therefore, troughs in the elemental data are purely a function of ion availability, because the concentration of these elements is much lower (e.g., two orders of magnitude for Sr) in freshwater. However, the mean trace element concentrations from the marine shell are all higher than the mean trace element concentration of the 8180 peaks (at 15, 23. 34, and 42 mm along the shell) in the estuarine shell (Table 1 ), despite the fact that the peaks in 8lsO in the estuarine shell are thought to represent a change back to normal marine salinities. It is less clear where the other elements (e.g., Mn. Zn, Pb, Cd) reside. Mn can partly replace Ca (White et al. 1977) in the arago- nite lattice, although it equally can be found bound within pig- ments. In the estuarine shell, there are minor peaks in Mn in zones 2 and 4 (Fig. 2a), which coincide with summer conditions. Mn in the marine shell is highly variable, although the greatest peaks in zone 2 occur during summer growth. Other trace elements are all in low concentrations, although Zn and Cd are higher in the es- tuarine shell; whereas, Pb is slightly higher in the marine sample. CONCLUSIONS The two shells of Mactra, taken from extremes in Argentinean near-shore environments, show a cyclicity in 5'3C and 8180. which reflects seasonal changes in the coastal waters. This con- clusion is supported by the amplitude and form of the cyclicity in the isotope data. There are trends in the trace element concentra- tion data that can be compared with variations in the isotope data. In the estuarine shell, there are cyclic variations in Sr, Mg, and Ba that visually match the variation in 8180 and S'^C. The highest concentrations of these elements are coincident with 5lsO peaks and approach mean values for both isotopes and trace elements from the marine shell. The lack of statistically significant correla- tions between the trace elements and isotope data suggests that the patterns of trace element variations cannot be directly related to single environmental parameters, and variations are more likely to represent a combination of environmental and physiological vari- ables. In the estuarine shell, the dominant effect seems to be di- lution of marine water by river water, and this far outweighs other causes of variation that are seen in the marine shell. Mactra isabelleana shells, thus, have great potential as palaeosalinity indicators for the southern hemisphere, off the At- lantic coast of South America, which is especially significant, because palaeoclimate records tend to be scarce or contradictory in this area (Aguirre et al. 1998). LITERATURE CITED Aguirre. M. L. 1990. Holocene macrobenthic molluscan associations from northeastern Buenos Aires Province. Argentina. Quatem. South Am. Antarctic Peninsula 7:161-195. Aguirre. M. L., M. J. Leng & B. Spiro. 1998. Variation in isotopic com- position (C. O, Sr) of Holocene Mactra isabelleana (Bivalvia) from the coast of Buenos Aires Province. Argentina. Holocene 8:613-621. Aguirre, M. L. & R. C. Whatley. 1995: Late Quaternary marginal marine deposits from northeastern Buenos Aires Province. Argentina: a re- view. Quatem. Sci. Revs. 14:223-254. Aguirre, M. L., D. Q. Bowen, G. A. Sykes & R. C. Whatley. 1995: An aminostratigraphical framework for late Quaternary marine deposits in Buenos Aires Province, Argentina. Mar. Geo!. 128:85-104. Amiel, A. J., G. Friedman & D. S. Miller. 1973. Distribution and nature of corporation of trace elements in modern aragonitic corals. Sedimenlol- ogy 20:47-64. Cohen. A. & P. Tyson. 1995. Sea surface temperature fluctuations during the Holocene off the south coast of Africa: implications for terrestrial climate and rainfall. The Holocene 5: 304-312. Craig. H. 1965. The measurement of oxygen isotope palaeotempcratures. pp. 161-182. In: E. Tongiorgi (ed.). Stable Isotopes in Oceanographic Studies and Palaeotemperatures. CNR Lab. Geol. Nucl.. Pisa. Eisma, D., W. G. Mook & H. A. Das. 1976. Shell characteristics, isotopic composition and trace element contents of some euryhaline mollusks as indicators of salinity. Palaeogeog. Paleoclimatol. Palaeoecol. 19:39- 62. Foster. P. & J. Chako. 1995. Minor and trace elements in the shell of Patella vulgata (L.). Mar. Environ. Res. 40:55-76. Hendry, J. & R. Kalin. 1997. Are oxygen and carbon isotopes of mollusk shells reliable palaeosalinity indicators in marginal marine environ- ments'? a case study from the Middle Jurassic of England. /. Geolog. Soc. Land. 154:321-333. Jones. D. & I. Quitmyer. 1996. Marking time with bivalve shells: oxygen 574 Leng isotopes and season of annual increment formation. Palaios 1 1 :340- 346. Jones, D. S.. D. F. Williams & M. A. Arthur. 1983. Growth history and ecology of the Atlantic surf clam. Spisula sotidissima (Dillwyn) as revealed by stable isotopes and annual shell increments. J. Exp. Mar. Biol. Ecol 73:225-242. Klein. R.T.. K. C. Lohmann & C. W. Thayer. 1996. Sr/Ca and "C/12C ratios in skeletal calcite of Mytilus trossulus: covariation with meta- bolic rate, salinity, and carbon isotopic composition of seawater. Geochim. Cosmoch. Acta 60:4207-1221. Lloyd. R. M. 1964. Variations in the oxygen and carbon isotope ratios of Florida Bay mollusks and their environmental significance. J. Geol. 72:84-111. Lloyd, R. M. 1969: Palaeoecological interpretation of the Caloosatchee Formation, using stable isotope methods. J. Geol. 77:1-25. Lorens. R.B. 1981. Strontium, cadmium, manganese, and cobalt distribu- tion coefficients in calcite as a function of calcite precipitation rate. Geochim. Cosmoch. Acta 45:553-561. Lorens, R.B. & M. L. Bender. 1980. The impact of solution chemistry on Mytilus edulis calcite and aragonite. Geochim. Cosmoch. Acta 44: 1265-1278. Lutz. R. & D. Rhoads. 1980: Growth patterns within the molluscari shell, an overview, chap. 6. In: D. Rhoads and R. Lutz (eds.). Skeletal Growth of Aquatic Organisms. Plenum Press. New York. Marshall. J.D.. D. Pime. A. Clarke. C. P. Nolan & J. Sharman. 1996. Stable-isotopic composition of skeletal carbonates from living Antarc- tic marine invertebrates. Lethaia 29:203-212. Mitchell. L.. A. Fallick & G. Curry. 1994. Stable carbon and oxygen compositions of mollusk shells from Britain and New Zealand. Palaeo- geog. Palaeoclimatol. Palaeoecol. 111:207-216. Mucci, A. 19S6. Growth kinetics and composition of magnesian calcite overgrowths precipitated from seawater: quantitative influence of or- thophosphate ions. Geochim. Cosmoch. Acta 50: 2255-2265. Pearce. N. J. G.. W. T Perkins & R. Fuge. 1992. Developments in the quantitative and "semiquantitative" trace element analysis of carbon- ates by laser ablation inductively coupled plasma mass spectrometry. J. Analyt. Atom. Spectromet. 7:595-598. Pearce, N. J. G.. W. T. Perkins, J. A. Westgate, M. P. Gorton, S. E. Jackson, C. R. Neal, & S. P. Chenery. 1997. A compilation of new and published major and trace element data for NIST SRM 610 and NIST SRM 612 glass reference materials. Geostand. Newslett. 21:1 15-144. Perkins, W. T. & N. J. G. Pearce. 1995. Mineral microanalysis by laser- probe inductively coupled plasma mass spectrometry, pp. 291-325. In: P. J. Potts. J. F. W. Bowles , S. J. B. Reed, and M. R. Cave (eds.). Microprobc Techniques in the Earth Sciences. The Mineralogical So- ciety. Perkins. W. T, N. J. G. Pearce & J. A. Westgate. 1997. Calibration strat- egies for laser ablation ICP-MS: examples from the analysis of trace elements in volcanic glass shards and sulphide minerals. Geostand. Newslett. 21:175-190. Perrier. C. C. Ho;:aore-Marcel & L. Ortlieb. 1995. Paleogeographie litto- rale et enregistrement isotopique (l3C, lsO) d'evenements de type El Nino par les moilusques Holocenes et recents du Nord Ouest peruvien. Geographic Physique et Quaternaire 48:23-38. Rios. E. 1994. Seashells of Brazil. (2nd ed.). Funacao Universidade do Rio Grande. Rio Grande. Brazil. Rosenberg. G. D. 1980. An ontogenetic approach to the environmental significance of bivalve shell chemistry, pp. 133-168. In: D. C. Rhoads and R. A. Lutz (eds.). Skeletal Growth or Aquatic Organisms: Biologi- cal Records of Environmental Change. Plenum Press. New York. Taylor. J. D.. W. J. Kennedy & A. Hall. 1969. The shell structure and mineralogy of Bivalvia. introduction. Nuculacea-Trigonacea. Bulletin of the British Museum of Natural History (Zoology!. 3:1-128. White. L. K.. A. Szabo. O. Carkner. & N. D. Chasteen. 1977. An electron paramagnetic study of Mn II in the aragonite lattice of a clam shell. Mva arenaria. J. Phvs. Chem. 81:1420-1424. Journal oj Shellfish Research, Vol. 18. No. 2. 575-578, 1999. THE ROLE OF BIOTIC INTERACTIONS IN JUVENILE MORTALITY OF THE COCKLE (CERASTODERMA EDULE L.): FIELD OBSERVATIONS AND EXPERIMENT H. MASSKI AND J. GUILLOU* UMR CNRS 6539, Institut Universitaire Europeen de la Mer Technopole Brest-lroise, place Nicolas Copernic 29280, Plouzane, France ABSTRACT Recurrent drastic mortalities of juveniles in a population of the edible cockle. Cerastoderma edule, from Western Brittany, prompted a study of the role of biotic interactions. A predation hypothesis was tested through a field experiment using exclosure cages, in parallel with field observations of the natural population. Over 1 month, the juvenile mortality rate outside of the cages was more than 40%. Selective predation on juveniles below 1 1 mm in length was responsible for more than 85% of this mortality. Juveniles of the green crab. Carcinus maenas, were recognized as the most important predators of the young cockles. Intraspecific competition did not influence survival or growth of the cockle spat, despite the relatively high density of the experimental population i.e. more than 800 individuals. m~2. KEY WORDS: Cerastoderma edule. cockle, spat, mortality, predation, competition INTRODUCTION In temperate ecosystems, populations of the edible cockle, Cerastoderma edule (L.), are frequently affected by strong sea- sonal or interannual density fluctuations. During the benthic life span, critical stages may occur in all size classes, but the factors influencing the mortality of the early stages are major determinants of these fluctuations. Among the multiple sources of mortality, biotic interactions must be taken into account. These biological processes are related to competition and adult-larval interactions (Andre and Rosenberg 1991; Jensen 1993; Bachelet et al. 1992), or to predation (Hancock 1970; Brown and O'Connor 1974; Hylle- berg et al. 1978; Sanchez-Salazar et al. 1987(a, b); Kamermans and Huitema 1994). Critical stages affecting the population structure of the cockle were described as part of a concerted study of the the main cockle beds from the French Channel and Atlantic coasts (Ducrotoy et al 1991 ). In Northern Brittany and Arcachon, research focused on the role of intraspecific competition (Bachelet et al. 1992). In situ experiments clearly showed that negative juvenile-adult interac- tions were unlikely to explain early benthic mortalities in Brittany because of a relatively low density of adult population. In this area, a long term survey revealed recurrent features in spat mortality (Guillou and Tartu 1994). Every year, the main peak of recruit- ment occurred in autumn, followed by a winter decrease in density. After a short pause, strong mortality resumed in spring and sud- denly stopped in June. This study aimed to investigate whether biotic interactions were responsible for the spring juvenile mortality through a field experiment involving predator exclosure cages and juvenile cockle density manipulations. Cage experiments and density manipula- tions have been widely used in experimental ecology (Peterson 1979: Reise 1985; Ejdung and Bonsdorff 1992; Kamermans 1992) and may reveal species interactions affecting survival and growth. MATERIALS AND METHODS The cockle bed studied lies in the western part of the Bay of Morlaix, in Northern Brittany, in the intertidal area of the Penze Estuary. The sampling station (48°40.2' N; 3°56.8' W) was lo- *Corresponding author. E-mail: jacques.guillou@univ_brest.fr cated just below the mid-tide level, where the cockle density was highest. The main physico-chemical and edaphical factors had already been described (Guillou et al. 1990) and revealed a pri- marily oceanic environment. This sheltered area supports weak hydrodynamical processes, both in swell and currents. The sub- strate is an heterogeneous muddy sand with 20 to 40% silt. In order to determine the influence of biotic interactions on cockle mor- talities, a field experiment was carried out in parallel with a de- mographic survey of the natural population. Field Experiment Three replicates of a cage and an adjacent control were ran- domly located inside the cockle bed. The pairing (cage and con- trol) reduced the bias of local heterogeneity. The design consisted of PVC cylinder (30 cm in diameter and 15 cm high). These tubes were embedded in the substrate until only 3 cm protruded. Control enclosures were wide open whereas the exclosure cages were pro- tected by a 5 mm mesh size net. Predators and bivalve competitors were removed from the cages and controls after sieving the upper 3 cm of sediment through a 5 mm mesh sieve and large individuals of the macro- fauna (mainly adult cockles) were removed from the underlying substrate (10 cm depth). The experiment was conducted from 3 May to 7 June 1993. The size of the recruits in the natural population during this period (mean size on 3 May: 6.4 ± 2.6 mm) was sufficient to minimize handling mortality. Sixty juvenile cockles were randomly put in each enclosure, corresponding to 845 individuals. m":. This experi- mental density was more than 6 times the mean value in the natural environment at that time, but was similar to the highest density observed in this season during the long term survey i.e.. 850 in- dividuals. m2 in May 1990 (Guillou and Tartu 1994). Natural Population Sampling Analysis of the natural population of the juvenile cockles was performed between January and July 1993. The sampling fre- quency was monthly at the start of the study and twice a month during the critical period. The sampling was divided into 3 strata corresponding to the location of the 3 replicates in the experimen- tal design. A core sampler (unit-area: 1/16 m2, 10 cm depth) was used on the minimal basis of six random unit samples per stratum. The samples were preserved in formalin. The sieving was done 575 576 Masski and Guillou through 1 mm mesh size. Juveniles were measured along the an- teroposterior axis to the nearest 0. 1 mm. Data Processing A modal analysis was carried out on the size frequency distri- butions of the juvenile population in the natural environment using NORMSEP program (Abramson 1970). Mortality is expressed by a raw mortality rate which is the percentage of juveniles that died between two successive samplings: Q = [(Nl-N2)/N1]x\00, TV, and N-, being the numbers of juveniles in the consecutive samplings. This ratio was standardized as a 10-day mortality rate: R = Q+[(t2-t1)/lO], f, and /2 being the date of consecutive sam- plings. In the same way, growth was estimated by a 10-day growth rate: fe=[(X2-X1)xlOO/X1]-r[(f2-«l)/10], X, and X, being the mean body lengths in the consecutive samplings. Statistical tests were carried out in the data analysis of popu- lation parameters and experiment. The experimental results on the mortality rate (heterogenous variances) were analysed using a X" test (2I[(observed-expected.)"7expected]). Density and growth were analysed using a t test after checking homoscedasticity. The statistical analysis was performed with a minimum of 5% signifi- cance level. RESULTS Natural Population Analysis of the size-frequency distributions (Fig. 1) was per- formed on the newly recruited juveniles (GO) and on its cohort structure. The maximal density of juveniles (152 individuals. m~2) was reached at the end of March (Table 1 ). Significant decrease in the density could be observed from May. The 10-day mortality rates were not homogeneous after this time, the period of high mortality rate (21* from 3 to 24 May, 32% from 7 to 21 June) being interrupted by a notable pause from 24 May to 7 June (mor- tality =0). After 21 June, when mean size was 15.6 mm, density stabilized at a level of 40 individuals. m"2. This threshold of strong mortality was observed every year, density remaining at a nearly constant level over the following months (Guillou and Tartu 1994). The growth of the newly recruited year class started slowly between January and April, the value of the winter 10-day growth rate being slightly biased by the underestimate of the mean size until March. In Spring, the growth rate increased and maximum Figure 1. Size distributions of the juvenile cockles (sampled on 1 mm mesh size) recruited during the previous autumn and followed from January to July 1993. C'a and Cb are distinct cohorts of the year class (GO). values were reached between May and June. This period of strong growth occurred during the high mortality stage. A discriminate analysis of the GO group size-frequency distributions (Fig. 1 ) showed from 3 May a segregation into two cohorts (Ca and Cb) due to selective growth. The mean values of survival and growth rates of the cohorts were calculated during the critical period from 3 to 24 May (Table 2). TABLE 1. Mean density. 10-day mortality rate, mean size and 10-day growth rate of the juveniles in the natural population of the edible cockle. 10-day 10-day Number Density mortality rate Mean size growth rate Sampling date of samples (N. itT2 ± s.d.) (%) (mm ± s.d.) (%) 25/01/93 6 122 + 45 4.2 ± 3.0 24/02/93 6 133 ±53 -3 4.5 ± 2.4 2 26/03/93 6 152 ±73 -5 5.1 ±2.2 4 03/05/93 6 134 ±26* 3 6.4 ±2.6 7 24/05/93 6 75 ± 36* 21 9.2 ±3.1 21 07/06/93 6 77 ± 30 -2 11.9 ±4.4 21 21/06/93 6 43 ± 28* 32 15.6 ±4.4 it 16/07/93 10 42 ±20 1 18.4 ±3.0 7 * Significantly different from the previous sampling. Role of Biotic Interations in Cockles 577 A selective mortality was observed inside the newly recruited year class. Between 3 and 24 May, the mortality rate in the Cb cohort was less than 25% of that in Ca which supported more than 80% of the global juvenile mortality. TABLE 3. 10-day mortality rate and mean size of the juvenile cockles, in the field experiment and in the natural population (GO) from 3 May to 7 June 1993 (experimental period). Field Experiment At the end of the experiment, a highly significant difference (X2 = 37.21. df = 2, P < 0.05) was revealed in the densities between the cages and the controls (Table 3). In all the cages, the mortality rates were very low, which demonstrated their protective effect from predation. A comparison of the cages and controls showed that 85% of the young cockle mortality was eliminated by the cages. A possible bias could be expected from the experimental device (net covering the cages, fouling) at the level of the growth or the survival of the juvenile cockles. The mean mortality rate calculated over the 3 control replicates (10-day rate: 12.9%) was very close to the mortality rate in the natural environment (12.1%). showing that no significant disturbance resulted from the enclosure itself. Moreover, the mean size of the juveniles in the cages and in the controls showed no significant differences (r test. df=266, P value = 0. 19a(0.05)). confirming that the net covering the cages did not inhibit growth. DISCUSSION These field observations and experiment in 1993 confirmed the recurrent spring mortality of the juvenile cockles in Northern Brit- tany, revealed by a long term survey between 1987 and 1992 (Guillou and Tartu 1994). From 26 March to 21 June 1993, the juvenile mortality in the natural population reached 72% of the cohort number despite a low initial density. In the cage experiment, the presence of the net eliminated on average 85% of mortalities observed in the control, which empha- sized the major role of predation. High predation mortality in juvenile cockles has often been described (Hancock and Urquhart 1965; Hancock 1970). The green crab, Carcinus maenas, is known as a voracious predator of cockles (Jensen and Jensen 1985; Sanchez-Salazar et al. 1987a). In the site from Northern Brittany, this species appeared to be the main predator of the juvenile cock- les. More than 80% of the mortality affected the smallest cohort (6mm length). vs. less than 20% in the biggest (1 1mm length). This selective mortality is in accordance with the observations of Sanchez-Salazar et al. ( 1987a) concerning higher predation of the TABLE 2. Density, 10-day mortality rate, and mean size of the cohorts, Ca and Cb (see Fig. 1) in May 1993. Cage Control Natural pop. (GO) Sampling date GO Cohorts 03/05/93 24/05/93 Ca Cb Ca Cb Density (N.m"~) 10-day mortality rate (%) Mean length (mm) 78 56 8.4 29 30 6.3 48 7 11.3 10-day mortality rate (%) Mean length (final) (mm ± s.d.) 13.5 ± 2.3 13.3 ± 2.2 1.9 ±1.2 12.9 ±6.6 12.1 11.9 ± 4.4 green crab on the cockle juveniles smaller than 10 mm. A refuge size was also defined by Jensen and Jensen ( 1985) beyond which the prey are safe from their predators. In Brittany, the long term survey showed that mortality strongly decreased every year once the cockle size was between 10 mm and 15 mm (Guillou and Tartu 1994). In the present study the size at which mortality decreased was about 1 1 mm which correlated with the size of the predator as only juveniles of Carcinus were observed on the site. This agrees with relationships established between the size of the crabs and the size of the cockles (Ropes 1969; Seed and Brown 1978; Sanchez- Salazar et al. 1987b). the escape from predation resulting from a strengthening of the shell and a deeper burrowing into the sedi- ment. This selective mortality emphasized the role of predation in regulating and structuring populations from benthic communities (Vimstein 1977: Peterson. 1979: Holland et al. 1987). Migrations have also been advanced to explain density fluctua- tions in bivalve populations (Hall et al. 1990; Commito et al. 1995). In this study, however, passive migration was quite unlikely as hydrodynamical processes are weak in this sheltered environ- ment (Guillou et al. 1990). Active migration can take place in the early benthic stage of juvenile cockles (Baggerman 1953). but the experimental design (3 cm protruded controls) made it very im- probable. Moreover, broken valves observed in the controls em- phasized the role of predation. The high initial densities of juveniles in this experiment could also be a source of mortality due to intraspecific competition. However, analysis of growth and survival did not support this. Although a negative influence of the high density of juveniles on their survival was expected, the difference in mortality between the controls and the natural environment was very low. Moreover, the size distributions of the juveniles throughout the experiment (Table 3) showed that the mean size of individuals in cages and controls was slightly greater than in the natural environment, and hence was contrary to the expected consequences of high density, indicating a lack of intraspecific competition. These conclusions were supported by the data from the long term survey (Guillou and Tartu 1994). Survival curves of the juveniles showed the same trend every year whatever the density of recruits that ranged from a scale 1 to 6 according to the year. The growth curves also showed the same trend throughout the multiyear survey, confirming that the carrying capacity in juveniles was certainly much higher than the experiment density (845 individuals. m~2). ACKNOWLEDGMENTS We should like to thank Christian Tartu for his help in the field and in the laboratory. This work was funded by the French Na- tional Proeram on the Determinism of Recruitment (P.N.D.R.). 578 Masski and Guillou LITERATURE CITED Abramson, N. J. 1970. Computer programs for fish stock assessment. F.A.O. Fisheries Technical Paper 101. Andre. C. & R. Rosenberg. 1991. Adult-larval interactions in the suspen- sion-feeding bivalves Cerastoderma edule (L.) and Mya areanaria L. Mar. EcoL-Prog. Ser. 71:227-234. Bachelet, G.. M. Desprez, J. P. Ducrotoy, J. Guillou. P. J. Labourg, H. Rybarczyk, P. G. Saunau. B. Elkaim, & M. Glemarec. 1992. Role de la competition intraspecifique dans la regulation du recrutement chez la coque. Cerastoderma edule (L.). Ann. I. Oceanogr. Paris. 68:75-87. Baggerman. B. 1953. Spatfall and transport of Cardium edule L. Arch. Neerl. Tool. 10:315-342. Brown. R. A. & R. J. O'Connor. 1974. Some observations on the relation- ships between oystercatchers Haematopus ostralegus L. and cockles Cardium edule L. Strandford Lough Ireland Nat. J. 1 8:73-80. Commito, J. A.. S. F. Thrush. R. D. Pridmore. J. E. Hewitt, & V. J. Cummings. 1995. Dispersal dynamics in a wind-driven benthic system. Limnol. Oceanogr. 40:1513-1518. Ducrotoy, J. P.. H. Rybarczyk. J. Souprayen. G. Bachelet, J. J. Beukema, M. Desprez. J. Dorjes, K. Essink. J. Guillou. H. Michaelis, B. Sylvand. J.G. Wilson. B. Elkaim. & F. Ibanez. 1991. A comparison of the popu- lation dynamics of the cockle (Cerastoderma edule) in North Western Europe. In: Estuaries and coasts: Spatial and Temporal intercompari- sons. M. Elliot & J. P. Ducrotoy (eds.). Olsen & Olsen. Fredensborg, pp. 173- 184. Ejdung. G. & E. Bonsdorff. 1992. Predation on the bivalve Macoma bal- thica by the isopod Saduria entomon: laboratory and field experiments. Mar. Ecol.-Prog. Ser. 88:207-214. Guillou. J.. G. Bachelet. M. Desprez. J. P. Ducrotoy. I. Madani. H. Ryba- rczyk. P.G. Sauriau. B. Sylvand. B. Elkaim. & M. Glemarec. 1990. Les modalites de la reproduction de la coque {Cerastoderma edule) sur le littoral francais de la Manche et de 1'Atlantique. Aquat. Liv. Resour. 3:29-41. Guillou, J. & C. Tartu. 1994. Post-larval and juvenile mortality in a popu- lation of the edible cockle Cerastoderma edule (L.) from Northern Brittany. Neth. J. Sea Res. 33:103-1 1 1. Hall, S. J.. D. Raffaelli. & W. R. Turell. 1990. Predator-caging experiments in marine system: A reexamination of their value. Am. Nat. 136:657- 672. Hancock, D. A. & A. E. Urquhart. 1965. The determination of natural mortality and its causes in an exploited population of cockles {Cardium edule L.). Fish. Invest. 24. 1-10. Hancock. D. A. 1970. The role of predators and parasites in a fishery for the mollusc Cardium edule L. Proceedings of the Advanced Study Institute on Dynamics of Numbers Populations (Oosterbeek 1970). pp.4 19-439. Holland. A. F.. A. T. Shaughnessy. & M. H. Hiegel. 1987. Long-term variation in mesohaline Chesapeake Bay macrobenthos: Spatial and temporal patterns. Estuaries (Special issue: Long-term biological records). 10:227-245 Hylleberg. J.. V. Brock. & F. Jorgensen. 1978. Production of sublittoral cockles, Cardium edule L. with emphasis on predation by flounders and sea stars. Nat. Jutland. 20:183-191. Jensen, K.T. 1993. Density-dependant growth in cockles (Cerastoderma edule): Evidence from interannuel comparisons. J. Mar. Biol As- so, U.K. 73:333-342. Jensen, K.T. & J.N. Jensen. 1985. The importance of some epibenthic predators on the density of juvenile benthic macrofauna in the danish Wadden sea. J. Exp. Mar.Biol. Ecol. 89:157-174. Kamermans, P. 1993. Competition in deposit- and suspension-feeding bi- valves: experiments in controlled outdoor environments. J. Exp. Mar. Biol. Ecol. 162:113-135. Kamermans, P. & H. J. Huitema. 1994. Shrimp browsing upon siphon tips inhibits feeding and growth in the bivalve Macoma balthica (L.). /. Exp. Mar. Biol. Ecol. 175:59-75. Peterson. C. H. 1979. Predation. competition, exclusion and diversity in the soft-sediment benthic communities of estuaries and lagoons. In: Liv- ingston. R. J. (ed.) Ecological processes in coastal and marine systems. Plenum Press, New-York, pp.233-264. Reise. K. 1985. Tidal flat ecology. An experimental approach to species interactions. Springer Verlag, Berlin, pp. 1 — 191. Ropes. J. W. 1969. The feeding habits of the green crab Carcinus maenas. B. Fish. Wildlife Ser\:US. 67:183-203. Sanchez-Salazar, M. E.. C. L. Griffiths. & R. Seed. 1987a. The effect of si/e and temperature on the predation of cockles Cerastoderma edule (L.) by the shore crab Carcinus maenas. J. Exp. Mar. Biol. Ecol. 111:181-193. Sanchez-Salazar, M. E.. C. L. Griffiths. & R. Seed. 1987b. The interactive roles of predation and tidal elevation in structuring populations of the edible cockle, Cerastoderma edule. Estuar. Coast. Shelf S. 25:245-260. Seed, R. & R. A. Brown. 1978. Growth as strategy for survival in two marine bivalves. Cerastoderma edule and Modiolus modiolus. J. Anim. Ecol. 47:283-292. Virnstein. R. W. 1977. The importance of predation by crabs and fishes on benthic infauna in Chesapeake Bay. Ecology. 58:1 199-1217. Journal of Shellfish Research, Vol. 18. No. 2. 579-59?. 1999. UNIQUENESS OF THE GASTROPOD ACCESSORY BORING ORGAN (ABO): COMPARATIVE BIOLOGY, AN UPDATE MELBOURNE R. CARRIKER AND GREGORY L. GRUBER College of Marine Studies University of Delaware Lewes, Delaware 19958 ABSTRACT Unique among organs of invertebrate animals, the gastropod accessory boring organ is a compact, histologically, physiologically, biochemically complex mechanism by which boring predatory gastropods penetrate the calcareous armor of live molluscan prey to feed. The present review considers the close similarity of the ABO in naticoidean and muricoidean boring gastropod species, the worldwide, distribution of the ABO, its physiological-biochemical functions, and its possible evolutionary origin. The ABO occurs only in the midventral part of the foot of muricoidean neogastropods and under the tip of the proboscis of naticoidean mesogastropods. Many reports in the literature, and a new study of additional species from temperate and tropical regions by us, further confirm the quite remarkable similarity of the morphology of the ABO in a wide range of species from many different regions of the world. In 29 species and subspecies of live muricoidean snails ranging in shell height from 12.1 to 1 15.0 mm. the mean diameter of the relaxed ABO ranged from 0.9 to 4.4 mm. and the mean width of the radula. spread over the tip of the odonotophore. from 0.12 to 1 .40 mm. The fine structure and physiology of the secretory disc of the ABO is strikingly similar in the one naticoidean and four muricoidean species that have been studied in detail to date. The disc possesses features characteristic of highly active secretory cells. Cytologically, actively boring and inactive ABOs differ conspicuously. Active glands possess a thicker secretory epithelium, longer microvilli, more mitochondria, membrane-bound granules, vesicles, endoplasmic reticulum, and lysosomes, than inactive glands, and a denser concentration of hemocyanin molecules in the intercellular sinuses. Secretory granules probably pass to the surface of the ABO through interstitial ducts in the center of each secretory cell group. Glycogen is abundant in inactive ABOs and sparse in active glands. Shell dissolution in the borehole is primarily a chemical process involving enzymes (not yet identified), an inorganic acid (HC1). and chelating agents (not yet identified). Borehole size and shape mirror the external morphology of the extended ABO. Of the several enzymes that have been identified within the secretory disc, a prominant one is carbonic anhydrase. It plays a pivotal role in shell dissolution, but it does not function as a direct demineralization agent; it catalyzes the hydration of metabolic carbon dioxide producing hydrogen ions, which are selectively transported into the borehole for the release of calcium ions. These are transported into the ABO sinuses and thence into the circulatory system of the snail. The ABO develops early in the embryology of naticids as a patch of enlarged epidermal cells behind the mouth on the ventral side of the proboscis tip; the process has not yet been studied in muricids. Both naticid and muricid juveniles drill small prey upon hatching from the egg capsule. A dond nudibranch, a tonnacean mesogastropod, and a marginallid neogastropod also drill round holes in the shell of molluscan prey, but lack of an ABO. Possible anatomical precursors suggesting the line of evolution of the ABO from its late Cretaceous origins were not found. Although the position of the muricoidean ABO in the foot and that of Naticoidea on the proboscis tip differs, the organs are essentially identical, and represent a striking case of convergent evolution. KEY WORDS: Gastropoda. Naticoidea, Muricoidea, accessory boring organ, morpology. histology, physiology, biochemistry, dis- tribution, evolution INTRODUCTION MORPHOLOGY OF THE ABO The accessory boring organ (ABO) is unique among all organs The ABO occurs only in the foot of muricoidean neogastropods of invertebrates examined to date. In the compact, mushroom- and on the ventral tip of the proboscis of naticoidean mesogastro- shaped organ there has evolved a physiological-biochemical pods (Carriker 1981 ). In all muricoidean males, the ABO is located mechanism that permits predatory gastropods to penetrate the ex- deep in the midanterior ventral part of the foot (Fig. 1) within a ternal shell of their molluscan prey and within the protection of the vestibule out of which it balloons under blood pressure into the shell, to feed in comparative safety on the flesh within. borehole during the boring process. In the retracted position, the Since Troschel's (1854) first reported discovery of the ABO ABO is tightly folded and lies snugly suspended within the vesti- (he thought it was a muscular sucker for holding prey during bule. This chamber is formed by the invagination into a space in perforation) on the ventral tip of a naticid proboscis, the ABO has the pedal musculature of a thin area of the ventral pedal epithe- continued to excite interest among marine malacologists. espe- Iium. Around the circumference of the secretory disc, the epithe- cially about how the gland secretion weakens the shell at the lium thins and turns dorsally to form the wall of the cylindrical surface of the borehole permitting the radula to rasp off and swal- peduncle. When retracted, the free surface of the disc is a much low the softened shell crystals. The similarity in form of the ABO convoluted, thickened pad of specialized secretory tissue whose in predatory gastropods as distantly related as naticoidean meso- creases and folds run radially. In vivo, the everted disc appears gastropods and muricoidean neogastropods has also raised the translucent white, the color emanating from a white layer of tissue question as to whether the gland is homologous in the two taxa. To buried beneath the surface of the disc. The length and extensibility date the ABO has been found in no other invertebrates. The present of the peduncle permit eversion of the ABO into deep boreholes in review considers the similarity of the structure and ultrastructure thick shell. From the inner dorsal surface of the ABO disc, a pair of the gastropod ABO in naticoidean and muricoidean boring spe- of large blood vessels, many fine nerves, and minute muscle and cies, the physiological function of the ABO, and its evolutionary collagen fibers pass dorsally through the ABO sinus into the pedal origin. musculature. When everted into the prey borehole, the ABO takes 579 580 Carriker and Gruber ___ 1 mm Figure 1. Drawing of sagittal section of anterior part of foot of a male muricoidean, Rapana thomasiana, through the accessor) boring organ, ABO. S, ABO sinus containing arteries (A), nerves (N), and muscles (M) passing to back of ABO. V, ABO vestibule through which ABO is extended by blood pressure into borehole. P. propodium. T, transverse furrow of foot (from Carriker 1981). the shape of a stout-stalked mushroom, its cushiony, translucent, glistening disc suspended at the end of the thin-wall peduncle (Fig. 2). The diameter of the ABO is approximately equal to that of the proboscis. In females of some muricoidean species, the ABO is also lo- cated in the midanterior ventral region of the foot, but lodged anterior to. and against the ventral pedal gland (=egg capsule gland). Otherwise, its structure is identical to that of males (Fig. 3). ^__ / mm Figure 3. Drawing of sagittal section of anterior part of foot of a female muricoidean, L'rosalpinx cinerea follyensis, through the acces- sory boring organ. ABO, and ventral pedal gland (VPG). S, ABO sinus. V, ABO vestibule. P. propodium. T, trans\erse furrow. N. nerve. A, artery. M. muscle (from Carriker 1981). Curiously, in females of most species of the Rapaninae. the ABO within its vestibule rests atop (above), and its vestibule is continu- ous with the lumen of the ventral pedal gland (Fig. 4), so that a common duct serves the ABO and the egg capsule gland. Thus, during eversion, the ABO passes ventrally into the borehole through the lumen of the egg capsule gland. The structure of the ABO is identical to that of males. What evolutionary advantage the close proximity of the two glands provides the species, is unclear (Carriker 1981). In both male and female naticoideans. by contrast, the ABO has Figure 2. Accessory boring organ of Urosalpinx cinerea follyensis ex- tended from foot into an incomplete borehole ( 1 mm diam.) in a glass- shell model. Fight micrograph (from Carriker 1969). Figure -t. Drawing of sagittal section of anterior part of foot of a female muricoidean. Rapana thomasiana, through the accessory boring organ, ABO, and ventral pedal gland, VPG. The ABO is located atop the ventral pedal gland and in eversion passes through the lumen of the gland. S. ABO sinus. N, nerve. A, M, muscle. V, ABO vestibule. P. propodium. T, transverse furrow (from Carriker 1981). Uniqueness of Gastropod ABO 581 evolved under the ventral lip of the cephalic proboscis close to the mouth and the radula (Fig. 5). In the retracted position, the naticid ABO resembles a somewhat wrinkled cushion of tissue suspended from the ventral end of the proboscis, and takes up a part of the lumen of the mouth. When everted into the borehole, the fully tumescent ABO resembles a fungiform papilla with a short pe- duncle and has a diameter equal to that of the proboscis. From the ventral wall of the proboscis extend small muscle and collagen fibers, nerves, connective tissue, sinuses, larger muscles, and a large branch of the buccal artery, to the dorsum of the ABO. The large artery branches into a rich network within the basal /.one of the secretory disc (Bernard and Bagshaw 1969). The naticoidean ABO differs anatomically from that of muri- coideans primarily in possessing a z.one of subdermal mucocytes around the periphery of the secretory disc. The mucins are thought to provide a lubricant and "seal" for the ABO when applied to the boring site on the shell (Fretter and Graham 1994). Also, the peduncle is long in muricoideans. to accommodate the position of the gland deep within the foot and short in naticoideans. because the gland is exposed on the lower tip of the proboscis. ANATOMICAL UNIFORMITY OF ABO Many reports in the literature confirm the quite remarkable similarity in the gross anatomy of the ABO in a wide range of species from many different regions of the world. Tables 1 and 2 summarize these reports. Table 1 records the anatomical location of the ABO in male and female muricoidean families. In all muricoidean males, the ABO occurs in the same relative position in the midventral part of the foot. Of the 48 species and subspecies listed, the ABO is present atop the egg capsule gland in 28 taxa; and in the remaining, in front of the egg capsule gland. In all the naticoideans examined to date, the ABO occurs only on the ventral tip of the proboscis. The fact that these snails burrow through frequently abrasive sandy substrata, could have influenced the evolution of the ABO in the protected, retractable proboscis (Kool 1993a). In no boring gastropod species was a simple- structured, possible precursor gland to the ABO observed. ABO-RADULA RELATIONSHIPS Interested in the relationships among the variables of ABO diameter, radular width, and shell height in male and female boring Figure 5. Drawing of left side of proboscis of a naticoidean, Polinices duplicates, opened laterally to show accessory boring organ, ABO; buccal mass, BM; and proboscidial hemocoel, PH. M. mouth: R. radu- lar sac; E, esophagus; ORM, odontophoral retractor muscle; PRM, proboscis retractor muscle (from Carriker 1981 1. snails from widely different regions of the earth, we examined the boring mechanism of a large sample of muricoidean and nati- coiden gastropods. Snails were obtained by one of us collecting along the east coast of the United States (Carriker 1961 ) and by air shipments from colleagues abroad. Snails were placed in a little seawater in plastic bags supported in loose packing (see Acknowl- edgements section). For anatomical purposes, snails were relaxed in 10 ppm sevin in seawater, quickly frozen on dry ice. and stored frozen for at least 1 week (Carriker and Blake 1959). Dimensions of the ABO were obtained by excision of the organ from live snails and soaking the organ in distilled water for 3 minutes, causing it to swell to maximum size. Width of the radula was measured nor- mally extended over the front tip of the exposed odontophore of live snails under a binocular microscope. Results of the measure- ments are recorded in Table 3 and Figures 6-8. The height of the shell of muricoideans ranged from 18.5 mm {Urosalpinx perrugata) to 115.0 mm {Murex fulvescens). The mean diameter of the fresh, relaxed ABOs ranged from 0.9 mm (Murex cellulosus, Urosalpinx tampaensis) to 4.4 mm (Murex po- mum): and radular width spread over the tip of the odotophore (in the rasping position), from 0.9 mm (Urosalpinx tampaensis) to 1 .4 mm (Rapana thomasiana). Although, generally, the larger snails possessed larger ABOs and radulae, correlations were not close. Generally, the dimensions of the ABO and radula were smaller in males than in females, although, for the most part, this simply reflected smaller males than females. Comparison of the sizes of the ABO and radula in normally feeding and starved (for 6 months) snails (Urosalpinx cinerea follyensis) showed negligible loss in size because of lack of food, suggesting that these organs are ready to function normally and at full capacity as soon as food is encountered. Although, in a populational sense, the mean diameter of the ABO tended to increase slightly with increase in shell height, there was a wide range of ABO diameters within a given shell height, as, for example, in shell heights 18 and 23 mm (Table 3. Fig. 6). Larger snails, such as Murex fulvescens, with a shell height of 1 15 mm. would be expected to have a large ABO, in this case 4.2 mm. but one smaller species, Murex pomum (44.7 mm shell height), possessed the largest ABO (4.4 mm mean diameter) of all the species examined (Figs. 6, 8). Mean radular width showed a general increase with increasing shell height (Fig. 7). However, the range of mean radular widths was only 0.12 to 1.40 mm, considerably smaller than the range of mean diameters of ABOs. 0.9 to 4.4 (Fig. 8). Functionally, this size differential is important, because the radula evenly and completely scrapes around within the small space of the deepening borehole, softened by the ABO. during penetration of prey shell. A proportionally wider radula would be difficult for the odontophore to manipulate in producing the usu- ally quite smooth, symmetrical boreholes characteristic of boring gastropods. This is demonstrated in Figure 8 by the proportionate increase in ABO and radular sizes as the shell increases in height. At the upper limit, were individuals of Rapana thomasiana, with a mean radular width of 1 .4 mm and an ABO diameter of 2.7 (Fig. 8). In all species, the radula appears small enough to function efficiently within the confines of the forming borehole. Anatomically, the ABOs of all species listed in Table 3 were remarkably similar, confirming the uniformity of ABOs recorded in Tables 1 and 2. No possible precursor stages of the ABO were identified. All were fully formed. 582 Carriker and Gruber TABLE 1. Anatomical location of the ABO in species of muricoidean gastropods from different regions of the world reported in the literature. Subfamily and Species Reference Location of ABO Region Anatomy (Form I Histology Male Female Muricinae Chicoreus brevifrons (Lamarck) Chicoreus florifer (A. Adams) Favartia cellulosus (Conrad) Phylonotus pomum (Gmelin) Muricopsinae Muricopsis ostrearum (Conrad) Ocenebrinae Acamhina monodon (Pallas) Acamhina spirata (de Blainville) Eupleura caudata (Say) Eupleura caudata etterae (Baker) Eupleura sulcidentata Dall Forreria belcheri (Hinds) Haustrum haustorium (Gmelin) Muncanthus fulvescens ( Sowerby ) Nucella emarginata (Deshayes) Nucella lamellosa (Gmelin) Nucella lapillus (Linnaeus) Ocenebra erinacea (Linnaeus) Ocenebra inomata (Recluz) Ocinebrina edwardsi (Payr.) Pterorytis foliata Gmelin Urosalpinx cinerea (Say) Boqueron. Puerto Rico Florida. USA Florida, USA Florida. USA Florida, USA Valparaiso. Chile Tomales Bay. California. USA North Carolina. USA Eastern Shore. Virginia, USA Florida. USA San Francisco. California, USA New Zealand Shackleford Banks. North Carolina. USA Washington, USA Washington. USA Massachusetts, USA England. France. New Brunswick. Canada England Japan; Washington, USA Livorno, Italy Washington. USA England. Florida to Massachusetts. USA Camker 1961 Carriker 1961 Carriker 1961, Carriker 1981 Carriker 1961 Carriker 1961, Carriker 1981 Camker 1961 Carriker 1961, Carriker 1981 Carriker 1961 Carriker 1961. Carriker 1981 Carriker 1961 Carriker 1961. Carriker 1981 Carriker 1961 Kool 1993a Wu 1985 Hemingway 1975 Camker 1955, Camker 1958. Camker 1959. Carriker 1961. Carriker 1981 Carriker 1959. Camker. 1961. Camker 1981: Provenza et al. 1966 Carriker 1961. Carriker 1981 Kool 1993a Kool 1993a Carriker 1961. Carriker 1981 Kool 1993a Carriker 1961. Camker 1981 Carriker 1961. Carriker 1981 Carriker 1961. Carriker 1981 Chetail and Binot 1967 Chetail et al. 1969 Chetail and Fournie 1970 Chetail et al. 1982 Derer 1975 Fretter 1941, Fretter 1946 Kool 1993a Webb and Saleuddin 1977 Carriker 1961, Carriker 1981 Fretter 1941. Fretter 1946 Carriker 1961. Carriker 1981 Bolognani-Fantin et al. 1981 Franchini et al. 1983 Carriker 1961. Carriker 1981 Carriker 1943. Camker 1955. Camker 1958, Camker 1959, Carriker 1961 Fretter 1941, Fretter 1946 Fretter and Graham 1994 T T T T A A A Carriker 1961 S A S A S T Camker 1961 s T Carriker 1961 s A Carriker 1961 s A Carriker 1961 s A Chetail and Binot 1967 Chetail et al. 1968 Chetail and Fournie 1970 Chetail et al. 1982 Derer 1975 Fretter 1941, Fretter 1946 Webb and Saleuddin 1977 Carriker 1961 s A Fretter 1941, Fretter 1946 Carriker 1961 s A Bolognani-Fantin et al. s A 1981 s A Fretter 1941, Fretter 1946 s A continued on next page Uniqueness of Gastropod ABO 583 TABLE 1. continued Subfamily and Species Reference Location of ABO Region Anatomy (Form) Histology Male Female Urosalpinx Baker) . follyensis (B. Urosalpinx perrugata (Conrad) Urosalpinx tampaensis (Conrad) Ergalataxinae Bedeva hanleyi Angas Trophoninae Trophon geversianus (Pallas) Rapaninae Acanthais brevidentata (Wood) Concholepas choncholepas (Bruguiere) Cronia amygdala Kiener Cymia lecta (Wood) Dicalhais orbita (Gmelin) Drupa morum Rdding Mancinella alouina (Roding) Mancinella deltoidea (Lamarck) Morula uva (Roding) Nasa serta (Bruguiere) Neorapana muricata (Broderip) Pinaxia versicolor (Gray) Plicopurpura patula (Linnaeus) Purpura clavigera Kuester Purpura persica (Linnaeus) Rapana rapiformis (Born) Rapana thomasiana (Grosse) Stramonita haemastoma (Linnaeus) Thais nodosa (Linnaeus) Tribulus planospira (Lamarck) Vasula melones (Duclos) Ve.xilla vexillum (Gmelin) Eastern Shore. Maryland and Virginia. USA Carriker 1943. Carriker 1959. Carriker 1961, Carriker 1963, Carriker 1966, Carriker 1969, Carriker 1978. Carriker 1981 Carriker and Van Zand! 1972a. Carriker and Van Zandt 1972b Carriker and Chauncey 1973 Carriker and Williams. 1978 Carriker et al. 1967 Carriker et al. 1978a. Carriker et al. 1978b Provenza et al. 1966 Person et al. 1967 Smarsh et al. 1969 Carriker 1961, Carriker 1981 Person el al. 1967 Smarsh et al. 1969 Florida. USA Carriker 1961. Carriker 1981 Carriker 1961 S A Florida. USA Carriker 1961. Carriker 1981 Carriker 1961 S A Port Jackson, Australia Carriker 1961. Carriker 1981 Carriker 1961 S T Argentina Harasewych 1984 S T Eastern Pacific Vermeij and Kool 1994 S T Chile Arias 1983 Gruber and Carriker 1990 Kool 1993a Arias 1983 Gruber and Carriker 1990 S A Queensland, Australia Kool 1993a s T Panama Kool 1993a s A New South Wales, Kool 1993a s T Australia Guam. USA Kool 1993a s T Queensland. Australia Kool 1993a s T Bimini. Bahamas Carriker 1961. Carriker 1981 Carriker 1961 s T Guam, USA Kool 1993a s T Guam. USA Kool 1993a s T Sonora. Mexico Kool 1993a s T Madagascar Kool 1993a s T Florida, USA Kool 1993a s T Japan Carriker 1961, Carriker 1981 Lo et al. 1980 Carriker 1961 Loet al. 1980 s T Krakatoa. Indonesia Kool 1993a s T Mahe. Seychelles Kool 1993a s T Japan Carriker 1961. Carriker 1981 s T Florida. USA Carriker 1961, Carriker 1981 Carriker 1961 s T Bimini. Bahamas Kool 1993a Ascension Island Kool 1993a s T Galapagos Islands, Kool 1993a s T Ecuador Panama Kool 1993a s T Oahu, Hawaii. USA Kool 1993a (parasitic on urchins) Absent Note: Nomenclature based on Kool (1993a) and Vaught (1989). updated by J. Harasewych. In all specimens, shape of the ABO was fungiform, and the histology (when reported) was typically muricid. S = in sole of foot, in all males and in some females; A = anterior to the ventral pedal gland in sole of foot, in some females; T = atop the ventral pedal gland, in some females. 584 Carriker and Gruber TABLE 2. Anatomical location of the ABO in species of naticoidean gastropods from different regions of the world reported in the literature. Reference Species Location of ABO Region Anatomy (Form) Histology Male Female Massachusetts, USA Carriker 1981 P P Washington. USA Carriker 1961. Carriker 1981 P P Germany, Sweden Ankel 1937. Ankel 1938 Ziegelmeier 1954 Ankel 1938 Ziegelmeier 1954 P P Massachusetts, USA Carriker 1981 P P France Fischer 1922 Fischer 1922 P P Naples, Italy Schiemenz 1891 Simroth 1896-1907 Schiemenz 1891 Simroth 1896-1907 P P Germany Haller 1892 P P Naples, Italy Troschel 1854 Schiemenz 1891 Hirsch 1915 Schiemenz 1891 P P Korea Carriker 1961. Carriker 1981 Carriker 1961 P P Korea Carriker 1961. Carriker 1981 Carriker 1961 P P Beaufort. North Carolina, Carriker 1961. Carriker 1981 Carriker 1961 P P USA Europe Troschel 1854 Bernard and Bagshaw P P Washington. USA Schiemenz 1891 1969 Lunatia heros (Say) Lunatia lewisii (Gould) Lunatia nitida Donovan Liuuitia triseriata (Say) Nalica catena da Costa Nalica josephina Risso Nalua lineata Chemnitz Nalica millepunctata Lamarck Natica severa (Gould) Neverita didyma (Roeding) Polinices duplicates (Say) Polinices lewisii (Gould) (see L. lewisii above) Simon perspectivum (Say) Beaufort, North Carolina, USA Fischer 1922 Ziegelmeier 1954 Bernard and Bagshaw 1969 Carriker 1981 Page and Pedersen 1998 Carriker 1961, Carriker 1981 Page and Pedersen 1998 Carriker 1961 Species names reported in the cited references. The shape of the ABO was a cylindrical pad on a short peduncle, and the histology (when reported) was similar to that of muricid ABOs. P = under ventral tip of proboscis. ABOs OF TROPICAL GASTROPODS In a further search for possible precursor stages of the ABO. we examined additional naticoideans and muricoideans from Guay- mas, Mexico; Santa Marta, Colombia; Oahu. Hawaii; and Marin- duque Island. The Philippines. Pedal cubes of tissue containing the ABO were processed by standard histological techniques, cut into sections 7- to 10-p.m thick, stained with a standard triehrome stain, and mounted on slides for microscopic examination. The number of tissues processed per species ranged from 1 to 5 (depending upon the snails available). Because tissues were fixed in the field for the most part, it was not possible to measure the diameter of the ABO. Suffice it to say that all muricoidean species and one species of Naticoidea examined possessed a typical accessory boring organ (Table 4)! The most interesting species was Drupa ricina. Wu (1965) noted that this species possesses the primitive features of the Stenoglossa. is not thought to be a typical predator of hard- shelled mollusks. and feeds on soft prey. However, we found that four individuals did possess a typical ABO. The tropical venture did add more species of gastropods that possess an ABO, but did not reveal any information on the possible early evolution of the group. FINE STRUCTURE OF THE ABO The fine structure of the secretory disc of the ABO is strikingly similar in the five species of boring gastropods studied to date: • Nucella lapillus (Derer 1975, Webb and Saleuddin 1977, Che- tail et al. 1982). • Ocenebra erinacea (Humphrey 1990), • Ocinebrina edwardsi (Franchini et al. 1983), • Polinices lewesii (Bernard and Bagshaw 1969), • Urosalpinx cinerea (Provenza et al. 1966. Nylen et al. 1969). The excellent, detailed studies of Nylen et al. (1969) on U. cinerea and Derer (1975) on N. lapillus, are especially informative. In brief, the secretory disc consists solely of a single layer of very tall columnar cells, arranged in compact cell groups with blood spaces around them (Fig. 9). The distal ends of the cells form a continuum over the surface of the ABO. Blood flows into the intergroup spaces in elaborately branching arterioles and cap- illaries that pass through the ABO sinus. At high magnifications, the blood pigments, hemocyanin particles, are identifiable in the blood vessels and intergroups sinus spaces. Nerves and muscles accompany the vessels into the base of the disc and pass among the cell groups to the undersurface of the disc continuum. From the intergroup spaces, blood flows back into the open ABO sinus, and thence returns into the principal venous channels of the foot. This complex architecture is admirably suited to supply abundant oxy- gen and nutrients to all parts of the metabolically highly active, extended ABO during shell boring (Peson et al. 1967). Longitudinally, the tall cell groups in the secretory epithelium are divisible into three distinct zones: the distal zone, bearing the brush border, and containing dense populations of long mitochon- dria; the middle zone, with relatively few organelles; and the basal Uniqueness of Gastropod ABO 585 TABLE 3. Mean diameter (mm) of relaxed ABO relative to mean width (mm) of radula spread over odontophore in species of boring gastropods from different temperate and tropical regions of the world. Species Region Snail Sex Snail Mean Height ABO Mean Diam. Radula Mean Width Bedeva hanleyi [5] Angas [3] Eupleura caudata [6] (Say) [3] Eupleura caudata [6] etterae (Baker) [6] Eupleura sulciden lata Dall [1] Murex brevifrons Lamarck [2] Murex cellulosus Conrad [2] Murex florifer A. Adams [6] Murex fulvescens Sowerby [4] Murex pomum Gmelin [6] Muricopsis ostrearum (Conrad) [6] Neverita didyma (Roeding) [2] Nucella lapillus [3] (Linnaeus) [3] " [4] [2] " [2] [2] Ocenebra erinacea [3] (Linnaeus) [3] Ocenebra inornata [4] (Recluz) [2] Polinices duplicates (Say) [4] Pterorytis foliata (Gmelin) [3] Purpura clavigera Kuester [6] Rapana thomaslana Crosse [3] Slnum perspectivum Say [1] Thais deltoida [41 (Lamark) [2] Thais emergenata (4] Deshayes [2] Thais haemastotna [4] floridana (Conrad) [6] Thais haemastotna [3] canaltculata (Gray) [3] Thais haemastotna [ 1 1 floridana (Conrad) [3] Thais lamellosa [4] (Gmelin) [2] Urosalpinx cinerea [3] (Say) [3] " [6] [7] " [V[ [5] " [6] [6] Urosalpinx cinerea [6] follyensis Baker [6] " [6] [6] Urosalpinx perrugata [6] (Conrad) [6] Urosalpinx tampiensis [2] (Conrad) [1] Port Jackson. Australia F 19.4 1.2 0.17 M 18.7 1.3 0.28 Virginia. USA F 22.5 1.2 0.20 M 18.7 11 0.17 Virginia, USA F 35.5 1.7 0.30 M 26.2 1.3 0.23 Honda, USA F 23.0 1.1 0.15 Puerto Rico, USA F 94.1 3.4 0.95 Honda, USA ? 21.5 0.9 0.20 Horida, USA 7 43.7 2.3 0.45 North Carolina. USA 7 115.0 4.2 0.85 Horida. USA ? 44.7 4.4 0.65 Florida. USA ? 22.7 1.0 0.19 Inchon. Korea ? 24.5 1.9 0.50 Plymouth. England F 25.7 1.5 0.29 M 27.3 1.4 0.27 Plymouth. Massachusetts. USA F 23.3 1.2 0.32 M 22.9 1.1 0.30 St. Andrews. New Brunswick, Canada F 33.5 2.4 0.48 M 29.0 2.2 0.38 Plymouth, England F 30.3 1.9 0.28 M 27.0 1.6 0.23 Washington. USA F 39.7 1.5 0.40 M 34.8 1.5 0.39 North Carolina, USA 7 34.3 2.6 0.73 Washington. USA 7 64.7 2.2 0.60 Japan ? 43.0 1.8 0.63 Japan ? 78.5 2.7 1.40 North Carolina. USA ? 25.0 1.6 0.63 Bimini. Bahamas F 37.0 2.4 0.51 M 34.5 2.4 0.48 Washington, USA F 29.3 1.6 0.35 M 26.8 1.5 0.32 East Coast, Florida F 57.3 2.4 0.65 M 58.7 2.5 0.78 Pensacola. Florida. USA F 50.0 2.2 0.54 M 41.0 1.9 0.47 Bimini. Bahamas F 19.0 1.6 0.35 M 20.0 1.5 0.40 Washington, USA F 40.3 1.5 0.50 M 36.5 1.4 0.42 Horida. USA F 26.7 1.6 0.28 M 21.7 1.3 0.23 North Carolina. USA F 26.2 1.5 0.26 M 22.7 1.7 0.27 Massachusetts. USA F 21.7 1.9 0.25 M 22.2 1.9 0.22 Bumham. England F 31.5 1.7 0.30 M 30.7 1.7 0.30 Eastern Shore, F 37.8 2.2 0.38 Virginia, USA M 30.0 1.8 0.31 ' (starved for 6 months) F 36.2 1.8 0.34 M 32.2 1.8 0.29 Florida, USA F 23.3 1.5 0.23 M 18.5 1.1 0.19 Florida, USA F 19.5 1.1 0.18 M 12.1 0.9 0.12 Note: Species names based in part on Keen (1971). Abbott (1974). and Diaz and Puyana (1994). [ ] = number of specimens examined. 586 Carriker and Gruber 4.0 - E J. i— E D Q c D c O CO ?» ? n — V U < 50 70 Shell Height (mm) Figure 6. Relationship between shell height and diameter of the ac- cessory boring organ in snails from widely different regions of the Earth. zone, containing the nuclei. Golgi complexes, endoplasmic reticu- lum, ribosomes, autophagic vacuoles, secretion granules, and other vacuoles. The basal zone is highly active physiologically during shell boring. The exterior surface of the ABO possesses a prominant brush border of long microvilli, with an occasional, longer tuft of cilia (possibly sensoryl among the microvilli. A viscid secretion coats the brush border. Distally, each cell group is nearly circular in cross section. An interstitial "star-shaped"" duct passes up the cen- ter of each cell group to the base of the microvilli. This is likely a passageway for secretion granules moving to the exterior of the ABO. The basal zone is extremely complex histologically, made so by the interdigitation of the secretory cells and their processes, and the maze of capillaries, muscle fibers, and nerve fibers that intertwine among them. Mitochondria are mostly located close to the cell membranes bordering the hemocyanin-filled intergroup sinuses. Blood pig- ment is conspicuously more abundant in active than in inactive ABOs. the increased blood flow into the gland augmenting the supply of required oxygen and nutrients. The dense network of nerves throughout the secretory disc indicates an organ whose extension and withdrawal are closely synchronized with the move- ments of the proboscis, odontophore. and radula during hole bor- ing, as well as, possibly, peristaltic pulsations of the muscular gland (such pulsations have not been observed) during discharge of secretion granules and other secretory products. E E o 5 D 0./ — D 30 50 70 0.9 Shell Height (mm) 90 Figure 7. Relationship between shell height and width of radula in snails from widely different regions of the earth. The ABO secretory epithelium possesses the cytological fea- tures characteristic of highly active cells: elaboration and extra- cellular discharge of secretory products, synthesis of glycogen in the inactive stage, and changes in the population of organelles relative to the physiological state of the organ. Golgi complexes in the basal zone produce the membrane-bound secretion granules and vesicles. How these are discharged onto the microvillar sur- face remains unclear. In electronmierographs, granules are visible apparently fusing with plasma membranes at the base of the mi- crovillar zone. This suggests that granules could be transported from the Golgi complexes at the base of the cells through the cell cytoplasm to the microvillar border, a distance of about 100 u.m. a discharge that would require considerable energy. A second, more probable course is the movement of granules into the inter- stitial ducts in the center of the cell groups, and extrusion by contractions of the musculature onto the microvillar border. How- ever, electronmierographs have not shown granules in the ducts (Nylen et al. 1969), possibly because the granules were expulsed quickly to the ABO surface by the fixative. Also. Derer (1975) observed no interstitial ducts in the cell groups, possibly because of a different method of tissue preparation. However. Carriker (1973) confirmed the existence of interstitial ducts using scanning electron microscopy. Freshly excised active ABOs fixed slowly with agitation in increasing concentrations of glutaraldehyde and freeze dried by a special technique clearly showed openings on the surface of the ABO among the microvilli and discharged granule- like spheres (Figs. 10. 1 1). Each extremely dilated duct opening bore a conspicuous flange and was traceable into the interior of the ABO in fracture sections. The secretion-like granules appeared swollen and larger than those (0.2-u.m diameter) in Nylen et al.'s (1969) micrographs, probably a result of the preservation proce- dures. Abundant immature secretion products in the Golgi complexes of active ABOs (in contrast to inactive ones) indicated that elabo- Uniqueness of Gastropod ABO 587 4.0 o E g Q c o c o ct> 6 CD O CO 3.0 - 2.0 1.0 TABLE 4. Condition of the ABO in naticoidean and muricoidean gastropods from tropical regions. 0.1 0.3 0.5 0.7 0.9 1.1 Radula, Mean Width (mm) 1.3 Figure 8. Relationship hetween width of radula and size of accessory boring organ in snails from widely different regions of the earth. ration and discharge of secretion granules are highest during pe- riods of active boring. The glycogen in the secretory disc of inac- tive ABOs provides a readily available source of energy for the change from the inactive to the active phase. Junctional complexes between the interstitial spaces and the free surface of the active ABO probably facilitate the passage of other secretion components (specific molecules, ions, water) from the cells onto the exterior brush border; the reverse movement is also likely. PHYSIOLOGY OF THE ABO Comprehensive studies on the physiology of the gastropod ABO have been reported for four species of Muricidae (Neogas- tropoda). These studies include: • Nucella lapillus (Linnaeus). Chetail and Binot (1967), Rosen- berg et al. (1968). Chetail et al. (1968). Chetail and Fournie (1969), Chetail and Fournie (1970), Webb and Saleuddin (1977), Chetail and Fournie (1980). Chetail et al. (1982); • Ocenebra erinacea (Linnaeus). Humphrey (1990); • Ocenebrina edwardsi (Payr), Bolognani-Fantin et al. (1981 ); • Urosalpinx cinerea (Say) and U. c.follyensis Baker Carriker et al. (1963). Provenza et al. (1966), Carriker et al. ( 1967). Person et al. (1967). Smarsh et al. (1969), Carriker ( 1969), Carriker and Chauncey (1973), Carriker (1978), Carriker and Williams (1978). Carriker et al. (1978a), Carriker et al. (1978b), Evans (1980). Carriker (1981). There have been no comparable studies on the physiology of the naticoidean ABO. Because of its prominence and accessibility on the ventral tip of the proboscis, the naticid ABO attracted the attention of early Species Region ABO Acanthina angelica | 3 1 Guaymas. Mexico Present I. Oldroyd Boreotrophon aculeatus [2] Bahamas Present (Watson) Chicoreus brevifrons [2] Santa Marta, Colombia Present (Lamarck) Coralliophila pan'a [2] Guaymas, Mexico Absent (E. A. Smith) Drupa ricina [4] Oahu. Hawaii Present (Linnaeus) Eupleura murieiformis [4] Guaymas. Mexico Present (Broderip) Haustellum haustellum 1 1 1 Marinduque, The Philippines Present (Linnaeus) Hexaple.x erythrostomus [2] Guaymas. Mexico Present (Swainson) Morula ferruginosa [3) Guaymas. Mexico Present (Reeve) Murex troscheli [2] Marinduque. The Philippines Present troscheli Lischke Muricanthus nigritus [2] Guaymas. Mexico Present (Philippi) Muricanthus princeps [3] Guaymas, Mexico Present (Broderip) Muricopsis annalus [5] Guaymas. Mexico Present (A. Adams) Natica chemnitzii [2] Guaymas. Mexico Present Pfeiffer Neorapana tuberculata |4] Guaymas. Mexico Present (Sowerby) Preropurpura erinaceoides [ 1 ] Guaymas. Mexico Present (Valenciennes) Purpura pansa [3] Guaymas. Mexico Present Gould Purpura patula [5] Santa Marta, Colombia Present Linnaeus Siralus plicifesoides | 1 ] Marinduque, The Philippines Present (Kuroda) Thais biserialis [4] Guaymas, Mexico Present (Blainville) Thais coronata trini [3] Santa Marta. Colombia Present tatensis (Guppy) Thais dehoidea [3] Santa Marta, Colombia Present (Lamarck) Thais haemastoma | 1 1 Santa Marta, Colombia Present floridana (Conrad) Thais rustica [3] ( Lamarck I Santa Marta. Colombia Present Thais speciosa [4] Guaymas. Mexico Present (Valenciennes) Thais triangularis [2] Guaymas. Mexico Present (Blainville) Note: Species names based in part on Keen (1971), Abbott (1974). and Diaz and Puyana ( 1994). and not updated. [ ] = number of specimens examined. naturalists long before the muricid ABO was known. Reaumur (1711). observing the smoothness of boreholes in molluscan prey, was the first to suggest that penetration was probably effected by chemical means. This report stimulated Schiemenz (1891 ) to study 588 Carriker and Gruber Figure 9. Histological sagittal section of accessory boring organ of Urosalpinx cinerea follyensis extended from foot. S, secretory epithelium. C, connective tissue in ABO sinus supporting retractor muscles, capillaries, and nerve fibers. ABO 1 mm in diameter. Light micrograph (from Carriker 1981). naticids (at the Zoological Station in Naples) and to conclude that the ABO. which he named the "boring gland," secreted sulfuric acid. Blue litmus paper touched to the boring gland in three ex- periments was changed to red. This, he reported, supported his theory of shell penetration by chemical means. Hirsch (1915) re- futed Schiemenz's results by using Congo red paper, which, he noted, is a true test of a free acid, and this acid did not change color when applied to the gland. He suggested that a specific enzyme, perhaps a calcase, could be involved in the shell boring. Fischer (1922) likewise tested the pH of the fluid on the ABO and the borehole with litmus paper and obtained neutral results; as did Ankel ( 1938) in later experiments. It would seem. then, that at least by these unreliable tests and conflicting results, the ABO secretion was neutral. Carriker et al. (1963), questioning these determinations of hy- drogen ion concentration, attempted more quantitative tests. Minute strips of short-range pH indicator papers, their color veri- fied against standardized buffer solutions checked with a pH meter, were applied to normally extended ABOs, excised ABOs. and ABO homogenates of muricid snails. All tests proved neutral to slightly alkaline, confirming the observations of most earlier workers. Not satisfied with these results, Carriker et al. (1967) devised a glass-shell bivalve model that permitted observation of the boring process and access to the extended ABO through a pore in the glass plate. Using specific glass microelectrodes developed by G. Charlton, they were able quantitatively to determine the pH of the viscid secretion on the surface of the ABO extended nor- mally into an incomplete borehole, and to demonstrate that the secretion of the functioning ABO in the borehole is distinctly acid, ranging in pH from 3.8 to 4.1. The failure of earlier workers to detect convincing acid reactions with pH papers was because of the insensitivity of their methods and the basic fact that forcefully extruded or excised ABOs produce no, or minimal quantities of acid. Subsequently. Carriker et al. (1978a). using a similar bivalve model, were able to monitor both hydrogen and chloride ions with microelectrodes touched to live, normally extended ABOs. They confirmed the acidity of the ABO secretion determined earlier ( 1967). and in addition observed that chloride ions were plentiful and increased stepwise in concentration from the time of extension of the ABO to its withdrawal. Qualitative analysis of dry ABO secretion by energy dispersive X-ray analysis confirmed the pres- ence of chloride and of sodium. It is important to note that there had been no contact of the ABO with seawater during the tests. Metabolically (relative to oxidase activity in mitochondrial rich particulate fractions), the boring ABO is much more active than the resting ABO (Person et al. 1967). Several investigators have identified a number of enzymes in the ABO: cytochrome oxidase, succinate dehydrogenase, and lactate dehydrogenase (Person et al. 1967); lipase and alkaline phosphatase (Chetail et al. 1968); car- Uniqueness of Gastropod ABO 589 Figure 10. A part of the exterior surface of the secretory disc of the accessory boring organ of V. c. follyensis showing a field of dilated pore openings in the brush border. Outer diameter of largest pores 10 urn. The light spheres could be secretion granules. Scanning electron micrograph. bonk anhydrase. acid phosphatase, alkaline phosphatase, ATPase (Webb and Saleuddin 1977): carbonic anhydrase (Chetail and Binot 1967, Smarsh et al. 1969. Chetail and Fournie 1969. Carriker and Chauncey 1973); alkaline and acid phosphatase, carbonic an- hydrase. Na-K- ATPase (Bolognani-Fantin et al. 1981); and Ca- ATPase, Na-K-ATPase. acid phosphatase, adenylate cyclase (Humphrey 1990). Whether the snails in these tests had been ac- tively boring before their ABO was excised, was not indicated. The most prominent enzyme in the secretory cells of the ABO is carbonic anhydrase (Carriker et al. 1968, Chetail et al. 1968. Webb and Saleuddin 1977). Furthermore, it is considerably more concentrated in the secretory epithelium than in the surrounding pedal tissues (Chetail and Binot 1967, Smarsh et al. 1969). Smarsh also found carbonic anhydrase in the ABO brush border; whereas, Webb and Saleuddin (1977) did not. Some researchers found the enzyme about equally abundant in both active ( = boring) and in- active (= resting) ABOs (Chetail and Fournie 1969. Webb and Saleuddin 1977): whereas, others found it more abundant in active ABOs (Chetail et al. 1968. Bolognani-Fantin et al. 1981). This inconsistency could have resulted from different ways of excising and treating the ABOs. Carbonic anhydrase is associated with secretory cells noted for active transport of hydrogen ions, bicarbonate ions, and carbon dioxide, and with some cells that transport sodium and chloride ions (Webb and Saleuddin 1977). Smarsh et al. (1969) and Car- riker and Chauncey (1973) concluded that carbonic anhydrase plays a vital role in shell dissolution during boring, but does not function as a direct demineralizing agent. Whether this is true in shell penetration by other invertebrates (Carriker and Smith 1969) remains to be determined. The required involvement of carbonic anhydrase in shell drilling was demonstrated by the significant inhibitory effect of Diamox. a specific inhibitor of the enzyme, on the incidence of boring by live snails in seawater to which the Diamox had been added (Chetail and Binot 1967. Rosenberg et al. 1968, Chetail and Fournie 1969, Carriker and Chauncey 1973). Chetail and Fournie (1969) and Webb and Saleuddin (1977) as- sumed that carbonic anhydrase is not released in the secretion during boring; yet Carriker and Chauncey (1973) identified the enzyme on blots of secretion taken from live, active ABOs. Again, handling of the ABO could explain the differing results. As released on the surface of the active ABO, the secretion is highly viscid, granular, generally insoluble in seawater, hyper- tonic, and about 65% volatile, when dried it becomes highly hy- groscopic. Its granular consistency is attributable to the secretory, membrane-bound granules and vesicles; its mucoid consistency prevents its dispersion during boring (Carriker 1981 ). Excised, live ABOs when placed on polished mollusk shell etch the surface: whereas, excised ABOs treated with heat and papain do not (Car- riker and Van Zandt 1964), suggesting inactivation of enzymes. Cytologically, there is a conspicuous difference between ac- tively boring and inactive ABOs. Active glands possess a thicker secretory epithelium, longer microvilli, more secretory granules. 590 Figure 11. Another field of pores in the secretory disc of the ABO. twice the magnification of that in Figure 10. Outer diameter of largest pores 10 um. Scanning electron micrograph. vesicles, endoplasmic reticulum, and lysosomes than inactive glands, and a higher concentration of hemocyanin molecules in the intergroup sinuses (Provenza et al. 19661. Active ABOs contain little glycogen; whereas, in inactive glands, it is abundant (Chetail et al. 1968). The marked structural and microstructural correspondence of the ABOs of different species of boring gastropods suggest that their physiology and biochemistry are probably quite similar. Car- riker and colleagues hypothesized that shell penetration is primar- ily a chemical process in which a combination of enzymes, an inorganic acid, and chelating agents solubilize the shell at the bottom of the borehole (Carriker et al. 1963. Carriker and Smith 1969, Carriker and Williams 1978, Carriker 1981; also reviewed by Bubel 1984). Borehole size and shape are. thus, a reflection of the external morphology of the extended ABO. Dissolution of the shell surface occurs first at the exterior, insoluble layers of the organic matrix of shell prisms and then proceeds into the interior organic-calcareous structure of individual prisms (Carriker 1996). weakening them for removal by the radula (Carriker 1981). Membrane-bound secretion granules and vesicles in the dis- charged hypertonic ABO secretion (Carriker 1973. Carriker et al. 1978a. Nylen et al. 1969, Humphrey 1990) and solubilization by the secretion of the organic matrix of shell prisms provide circum- stantial evidence for the presence of enzymes (possibly a conchio- linase). Inactivation of the shell-dissolving activity of the excised ABOs by heat and papain further suggest this view. That proteins are present in the ABO secretion of a muricid was demonstrated by Evans ( 1980). Hydrogen, chloride, and sodium ion concentrations in the secretion indicate its acidic (probably HC1) and hypertonic (NaCI) characteristics (Carriker et al. 1978a). A chelating agent, although not chemically identified, and an acid mucopolysaccha- ride present in the secretory epithelium of the ABO (Bernard and Bagshaw 1969. Smarsh et al. 1969. Carriker and Chauncey 1973) could function in chelation. The intricate distal zone of the secretory epithelium of the ABO is elaborately microstructured for active ionic transfer: long, dense microvilli, gap junctions, numerous microtubules, a rich capillary system, and a dense nerve net (Carriker et al. 1963. Nylen et al. 1969, Carriker 1973, Chetail et al. 1982, Webb and Saleuddin 1977. Humphrey 1990). The highly hydrophylic, negatively charged layer of acid mucopolysaccharides on the microvillar bor- der of active ABOs facilitates the flow of ions and the function of secretion and absorption (Chetail and Fournie 1980). Initial contact of the ABO epithelium with the borehole surface probably activates the carbonic anhydrase within the secretory cells. The enzyme catalyzes the hydration of metabolic carbon dioxide, producing hydrogen ions (from the dissociation of the carbonic acid), which are selectively transported into the borehole. These ions are involved in the release of calcium ions, water, and carbon dioxide from the shell calcium carbonate (and other shell carbonates. Carriker et al. 1978b). The carbon dioxide readily permeates the secretory cell membranes, and functions synergis- tically with cellular metabolic carbon dioxide and ATP to increase the rate of acid production (Webb and Salleuddin 1977). Simulta- Uniqueness of Gastropod ABO 591 neously. an increase in oxygen and pedal blood raises the meta- bolic rate and the activity of the mitochondrial population in the secretory cells (Person et al. 1967). The sodium ions in the secre- tion (Carriker et al. 1978a) are active in exchange, reabsorption. and transport of solutes across the secretory cell membranes (Humphrey 1990). As sodium ions pass into the secretory cells along the chemical gradient, hydrogen ions transfer out. acidifying the narrow space between the ABO and the borehole wall. The activity coefficients of calcium and carbonate ions, thus, probably decrease because of the low pH and because of the ion pairing of carbonate and bicarbonate ions with sodium ions. Removal of the free calcium ions from the borehole by chelation and ion transport across the cells of the ABO further decreases the concentration of calcium ions and increases the solubility product of the borehole shell calcium carbonate. Calcium ions may also serve as counter- ions to hydrogen ion transport. The role of chloride ions (Carriker et al. 1978a) in ionic transport is less clear. Calcium ions resulting from the solubilization of the shell pass through the secretory epithelium into the ABO intergroup sinuses and thence into the snail's pedal blood stream (Chetail and Fournie 1970). When the borehole has been completed, the blood supply in the ABO is reduced, and metabolic rate and production of acid diminish (Webb and Saleuddin 1977). Glycogen in inactive ABOs is prob- ably used to fuel ionic exchange pumps and for the synthesis and release of the secretory products in active ABOs. BORING GASTROPODS LACKING AN ABO The dorid nudibranch. Okadaia elegans Baba. bores holes in the calcareous tubes of spirorbid and serpulid polychaetes and feeds on them (Young 1969). Its holes are smooth, round, and beveled. However, the nudibranch does not possess a typical gas- tropod-type accessory boring organ: instead, a collar of conspicu- ous glandular cells surrounds the lumen of the proboscidial sto- modeum. During shell penetration, the stomodeum is everted fully, releasing a secretion, while simultaneously rasping with its radula. Cassis tuberosa (Linnaeus) and Cypraecassis testiculus (Lin- naeus) (tonnacean mesogastropods) drill round holes in the tests of echinoids to feed on them (Hughes and Hughes 1981 ). These snails possess two large salivary glands that deliver a secretion rich in sulfuric acid (pH 1) through long ducts passing through the nerve ring, along the proboscis, to the buccal cavity. Penetration of the prey test is achieved within about 10 minutes by the combined action of the acid and the radula. The marginellid neogastropods, Austroginella johnstoni (Pet- terd) and A. muscara (Lamarck), which bore round holes in the shell of their molluscan prey, also lack an ABO. The anterior end of the proboscis is supplied with an abundance of subepithelial gland cells, which may secrete the solvent for dissolving the shell. The surface of the hole is highly corroded, lacks evidence of radular scraping, or wear of the radular teeth, and its inner diam- eter is proportionally very small. These observations suggest that penetration is dominantly solutional (Ponder and Taylor 1992). The marginellid holes are similar to those bored by octopods. The composition of the shell-dissolving secretion in the dorids and marginellids has not been determined. EMBRYOLOGY OF THE ABO The ABO develops early in the embryology of boring gastro- pods. In Lunatia lewisii larvae, for example, the ABO is first recognizable as a patch of slightly enlarged epidermal cells located a short distance behind the ventral lip of the mouth on the pro- boscis tip. As development of the embryo progresses, the epithelial cells of the prospective ABO proliferate, grow taller and columnar in shape, and acquire darkly staining inclusions. In the laboratory within 3 to 5 days of metamorphosis, these larvae drill and ingest small bivalves and ostracods (Page and Pedersen 1998). In Nucella lapillus, the ABO opening is soon visible in the foot of veligers (Ball et al. 1997). The early development of the ABO in muricids has not been reported. It has been demonstrated by several investigators that naticid and muricid juveniles drill small prey and feed on them very soon after hatching: Concholepas concholepas, young barnacles and conspecifics (DiSalvo and Carriker 1994); Nassarius festivus, con- specifics (Morton and Chan 1997); Nalica gualtieriana, Bitiutiti sp. (Berg 1976); Ocenebra erinacea, small barnacles and conspe- cifics (Humphrey 1990); Polinices duplicates, young Gemma gemma (Wiltse 1980); Lunatia lewisii, small bivalves and ostra- cods (Page and Pederson 1998); Urosalpinx cinerea, small oysters, barnacles, clams (Carriker 1957). Fretter (1946) noted that in newly hatched Nucella lapillus, the ABO is very large, its diameter being equal to nearly one-third of the width of the foot. Humphrey (1990) observed the same pro- portionately large ABO in the foot of young Ocenebra erinacea. It seems that cannibalism among these young snails occurs primarily when no other prey are present. Morton and Chan (1997) were the first to report the evidence for shell boring by a species of the Nassaridae. The boring mecha- nism has not been described. EVOLUTION OF THE ABO The fossil record shows that shell boring by gastropods evolved late in the biological history of the Earth, probably in the Upper Cretaceous. After this, the frequency of bored shells increased dramatically, attesting the success of shell boring in obtaining food (Carriker and Yochelson 1968. Sohl 1969). The abundant, almost worldwide presence of gastropod borings in recent marine mol- luscan valves further demonstrates the ubiquity and dominance of boring gastropod species. Shell boring not only permits the preda- tor to feed in relative safety, but allows it to consume prey many times larger than itself. Although the position of the muricoidean ABO in the foot and that of the Naticoidea on the proboscis tip differs, the function of the ABO in the two taxa is similar, and thus analagous. However, the glands in the two groups are not homolo- gous; their development undoubtedly represents a case of conver- gent evolution (Wiley 1981. Kabat 1990). Although possible modern precursors suggesting the line of evolution of the ABO from its late Cretaceous beginnings have not been found, embryology does suggest a possible course. In the larvae of Polinices lewisii (Page and Pedersen 1998). for example, the ABO is first recognized as a patch of slightly enlarged epider- mal cells located a short distance behind the ventral lip of the mouth on the proboscis tip. As development progresses, the epi- dermal cells of the prospective ABO proliferates, become tall and columnar in shape and acquire darkly staining inclusions. From this sequence, we can hypothesize that in muricoideans mucus- secreting cells in the midventral pedal epithelium also evolved into unicellular glands that came to secrete a shell softener. For un- known reasons, when the specialized secretory pad invaginated. innervation and vascularization increased, and retracting muscles formed. Cilia from the original mucus cells, for the most part 592 Carriker and Gruber disappeared and were replaced by a dense brush border of mi- crovilli. The few remaining cilia assumed sensory functions. Be- cause of the position of the organ in the sole of the foot, only evolution in the direction of invagination was practical. Other possible evolutionary experiments, such as evagination, which would have impeded locomotion, or remaining at the level of the pedal epithelium, which might have limited full development of the organ, must have failed. Evolvement of the ABO peduncle made possible lengthy eversion and the capacity to bore through thick-shelled prey. In those species of female muricoideans in which the ABO lies atop the egg capsule gland utilizing a common duct, an origin of the ABO similar to that of males is difficult to envision. Kool ( 1993a) suggested that the condition of the ABO atop the capsule gland evolved first, and from this position, the two separate open- ings to the outside evolved. It is also likely that the two positions formed independently: that of the ABO in front of the capsule gland, in a way similar to that of the ABO in males; and that of the ABO atop the capsule gland, by specialization of a patch of epi- thelium in the roof of the capsule gland. Table 1 shows that the ABO and ventral pedal gland in all species of Muricinae, Muricopsinae, Ergalataxinae, Trophinae, and Rapaninae (less Concholepas concholepas and Cymia tecta) possess a common duct, and that all species of the Ocenebrinae (less Haustrum haustorium and Muricanthus fulvescens) are char- acterized by separate openings. These data support Kool's ( 1993a) findings that ABOs with separate ducts are restricted primarily to the Ocenebrinae. and those with a common duct, to the Rapaninae. His report that C. concholepas possesses an ABO atop the ventral pedal gland may be in error (Gruber and Carriker 1990). Evolution of the ABO under the end of the naticoidean pro- boscis, compared to the evolution of the muricoidean ABO, was probably a relatively simple one. It could have involved a trans- formation of mucus-secreting epithelium of the proboscis sheath into a pad of tall epithelium secreting a shell softener, and a con- sequent thickening of the pad in situ. Its formation as a slightly projecting, nonretractile gland was successful because of its posi- tion on the proboscis and the protection from mechanical injury afforded by a retractable proboscis. The literature discloses little evidence of possible precursors of the ABO in modern gastropods. Fischer (1922) observed that the naticid Sigaretus sp. possesses under the end of its proboscis a bi-lobed organ that he suggested could be a homologue to the naticid ABO. This interesting possibility has not been pursued. In another case, Fischer searched for the ABO in Natica catena da Costa, but found it only in individuals "of great size." He consid- ered this an example of tardy development in ontogeny. In any event, presence of the long cephalic proboscis was un- doubtedly necessary before initiation of the evolution of the ABO. Conversely, elongation of the cephalic snout into a proboscis did not necessarily serve as the catalyst for the evolution of the ABO. as demonstrated by the many species of gastropods that possess a lengthy proboscis, but lack an ABO. What stimulated the original evolution of these specialized shell-dissolving epithelial pads in the first place, is a fascinating question for which there seems to be no answer. However, we do know that external molluscan epithelia are highly plastic physi- ologically, versatilely capable of depositing shell or removing it as the state of development or other conditions require. During the ontogenetic development of the shell, for example, new shell is deposited along the valve margins and inside the valve(s) accom- modating the increase in size and reshaping the form of the soft parts. Several instances can be cited. A striking case is that of the muricid gastropods Chicoreus brevifrons (Lamarck) and Muricanthus fulvescens (Sowerby). whose shells are ornamented with conspicuous spines that run the breadth of the whorls. In the process of the spiral growth of the shell, the snail must remove the older spines that come to lie along the inner lip of the aperture in order to make room for the new shell of the enlarging body whorl. If the spines were not removed, they would block the movement of the snail body in and out of the shell, entombing the animal in its own shell. The response of the snail is partially to dissolve the base of the offending spines with secretion from the mantle edge until the spines fall away. What part of the complex mantle margin secretes the dissolving substance and what part deposits new shell, or whether the same secretory epithelium functions alternately in shell dissolution and shell formation, is not yet known (Carriker 1972). Another case is that of bivalves that employ a shell dissolving secretion to open burrows for themselves in molluscan shell. An unusual example is that of the pholad bivalve Perihelia conradi Valenciennes that dissolves burrows in the shell of the abalone Haliotis rufescens Swainson. The burrowing process proceeds mainly by chemical dissolution of the calcareous substrate by a shell-dissolving secretion released by the mantle. This remains in close contact with the burrow wall as the excavation is enlarged to permit increase in the growth of the pholad (Smith 1969). A final, unusual case is that in which minute, single-cell ex- tensions of the mantle epithelium of such bivalves as Corbicula flwninea (Miiller) dissolve microtubules that penetrate the shell to the outer periostracal covering. The function of the mantle exten- sions into the microtubules is unclear, but the tubules demonstrate the flexibility of the normally shell-depositing mantle epithelium in producing the microtubules (Aranjo et al. 1994). It is clear that shell formation and shell dissolution, two sides of the physiological coin, are natural processes in the shelled mem- bers of the Phylum Mollusca. Consequently, it is not difficult to conceive of the evolution of special shell-dissolving organs like the ABO. The titillating question, however, is what stimulus ini- tiated the development of such an organ in the first place, and then step-by-step carried the evolution to climax in the morphological- physiological complex that has successfully served many species of gastropods since at least the Cretaceous — seemingly with little. if any, change. A parallel question is why the ABO evolved in such different parts of the snail body in naticoideans and muricoideans. Bernard and Bagshaw (1969) liked to think of the variation as "one of the most interesting parallels in molluscan morphology." In- deed! ACKNOWLEDGMENTS The main part of the research on the distribution of the ABO was carried out at the University of North Carolina. Institute of Fisheries Research. Morehead City. 1954 to 1960. supported by a grant from the U.S. Fish and Wildlife Service. Aspects of the research were continued at the Marine Biological Laboratory. Woods Hole. Massachusetts. 1962 to 1973, and completed at the University of Delaware, College of Marine Studies in Lewes be- ginning in 1973. In 1959. Dennis Crisp. John Blake, and the senior author col- Uniqueness of Gastropod ABO 593 lected boring gastropods along the east and west shores of Florida. In 1966, during a brief expedition to Guaymas. Mexico, organized by the senior author's brother, Frederick R. Carriker, several of us collected intertidal and subtidal boring snails. And in 1981, the senior author, hosted by Director Jose A. Lozano, INVEMAR (Instituto de Investigaciones Marinas de Punto de Betin), searched for boring snails along shores in the vicinity of Santa Marta. Co- lombia. During these years, at different times. John Ballard, John Blake, Gregory Gruber, Alex Marsh, Mackie Willis. Langley Wood, and Dirk Van Zandt assisted in the ABO research. In addition to the collections by us, live boring gastropods were airshipped to the senior author in Morehead City by friends from other parts of the world: • Australia, Sydney Harbor: D. F. McMichael • Bahamas. Bimini: Langley Wood • Canada, St. Andrews: Neil Bourne • England, Burnham-on-Crouch: D. A. Hancock. Duncan Waugh • England. Plymouth: D. P. Wilson • Hawaii, Oahu: E. Allison Kay • Japan. Sendai: Akimitsu Koganezawa • Korea, Seoul: Yongbok Cho • Philippines, Marinduque (these were preserved snails from the DMNH, Smithsonian Institution!: Jerry Harasewych • Puerto Rico, Mayaguez: Juan Rivero, Jeff Rogers. Paul Shave • United States, Massachusetts, Woods Hole, Gloucester: Lang- ley Wood • United States, Virginia, Chincoteague: Thomas Carver, Michael Castagna • United States. Virginia, lower Chesapeake Bay: William Hargis • United States, Washington. Quilcene: Lee Fosdick. Cedric Lindsey. Identifications were based, in part, on Keen (1971). Abbott (1974), and Diaz and Puyana (1994). R. Tucker Abbot, William Clench, and Ruth Turner kindly confirmed some of the identifi- cations. It is a pleasure to express our thanks to all these persons and institutions that so generously and courteously facilitated this research. We are especially indebted to Jerry Harasewych for his detailed review of the manuscript. Present address of Gregory L. Gruber: Maryland Department of the Environment. Water Quality Monitoring Division. 416 Chin- quapin Round Road. Annapolis, Maryland, U.S.A. LITERATURE CITED Abbott, R. T. 1974. American seashells. The marine mollusca of the At- lantic and Pacific coasts of North America. 2nd. ed. Van Nostrand Reinhold Co.. New York. 663 pp. Ankel. W. E. 1937. Wie bohrt Natica? Biol. Zentralbl. 57:75-82. Ankel. W. E. 1938. Erwerb und Aufnahme de Nahrung bei den Gas- tropoden. Tool. Anz. Suppl. 11:223-295. Aranjo. R.. M. A. Ramos, & J. Bedoya. 1994. Microtubules in the shell of the invasive bivalve Corbicula fluminea (Miiller, 1774) (Bivalvia: He- terodonta). J. Moll. Stud. 60:405-413. Arias, E. 1983. Variaciones morfologicas en juveniles de Concholepas (Bruguiere, 1789). Prof, thesis. Department of Oceanology. University of Concepcion, Chile. Ball. A. D.. E. B. Andrews & J. D. Taylor. 1997. The ontogeny of the pleurembolic proboscis in Nucella lapillus (Gastropoda: Muncidae). J. Moll. Stud. 63:87-99. Berg, C. J., Jr. 1976. Ontogeny of predatory behavior in marine snails (Prosobranchia: Naticidae). Nautilus 90:1-4. Bernard, F. R. & J. W. Bagshaw. 1969. Histology and fine structure of the accessory boring organ of Polinices lewisi (Gastropoda. Prosobran- chiata). J. Fish. Res. Board Can. 26:1451-1457. Bolognani Fantin. A. M. B.. P. Nardi, E. Ottaviani, A. Franchini & P. Tongiorgi. 1981. Food deprivation in the muricid drill Ocenebrina edwardsi (Mollusca: Prosobranchia): histomorphological and behavior research. Mar. Ecol. 2:307-316. Bubel, A. 1984. Epidermal cells, pp. 400^147. In: J. Bereiter-Hahn, A. G. Matolsky and K. S. Richards (eds.). Biology of the Integument. VII. Mollusca. 1. Invertebrates. Springer- Verlag, Berlin. Carriker. M. R. 1943. On the structure and function of the proboscis in the common oyster drill. Urosalpinx cinerea Say. /. Morphol. 73:441—198. Carriker. M. R. 1955. Critical review of biology and control of oyster drills Urosalpinx and Eupleura. Special Science Rept.. Fisheries. U.S. Fish and Wildlife Service. 148:1-150. Carriker, M. R. 1957. Preliminary study of behavior of newly hatched oyster drills, Urosalpinx cinerea (Say). J. Elisha Mitchell Sci. Soc. 73:328-351. Carriker. M. R. 1958. Additional information on the mechanical-chemical nature of drilling by the gastropods Urosalpinx and Eupleura. Assoc. Southeast. Biol. Bull. 5:5. Carriker, M. R. 1959. Comparative functional morphology of the drilling mechanism in Urosalpinx and Eupleura (muricid gastropods). Proceed- ings of the XVth International Congress on Zoology, London. 373-376. Carriker, M. R. 1961. Comparative functional morphology of boring mechanisms in gastropods. Am. Zool. 1 :263-266. Carriker, M. R. 1966. Color cover photograph of normally extended ABO of Urosalpinx cinerea. BioScience 16. Carriker, M. R. 1969. Excavation of boreholes by the gastropod. Urosal- pinx: an analysis by light and scanning electron microscopy. Am. Zool. 9:917-933. Carriker. M. R. 1972. Observations on removal of spines by muricid gas- tropods during shell growth. Veliger 15:69-73. Carriker. M. R. 1973. Discovery of duct system in accessory boring organ of Urosalpinx cinerea follyensis. Proc. Nat. Shellfish Assoc. Com: 63:1. Carriker. M. R. 1978. Ultrastructural analysis of dissolution of shell of the bivalve Mytilus edulis by the accessory boring organ of the gastropod Urosalpinx cinerea. Mar. Biol. 48:105-134. Carriker. M. R. 1981. Shell penetration and feeding by naticacean and muricacean predatory gastropods: a synthesis. Malacologia 20:403- 422. Carriker. M. R. 1996. The shell and ligament, pp. 75-168. In: V. S. Kennedy, R. I. E. Newell and A. F. Eble (eds.). The Eastern Oyster Crassostrea virginica. Maryland Sea Grant College, University of Maryland, College Park. Carriker, M. R. & J. W. Blake. 1959. A method for full relaxation of muricids. Nautilus 73:16-21. Carriker, M. R. & H. H. Chauncey. 1973. Effect of carbonic anhydrase inhibition on shell penetration by the muricid gastropod Urosalpinx cinerea. Malacologia 12:247-263. Carriker. M. R. & E. H. Smith. 1969. Comparative calcibiocavitology: summary and conclusions. In: M. R. Carriker, E. H. Smith and R. T Wilce (eds.). Penetration of Calcium Carbonate Substrates by Lower Plants and Invertebrates. Am. Zoologist 9:1011-1020. Carriker. M. R. & D. Van Zandt. 1964. Use of polished shell for testing demineralization activity of accessory boring organ of muricid boring gastropods. Biol. Bull. 127:365. Carriker. M. R. & D. Van Zandt. 1972a. Predatory behavior of a shell- 594 Carriker and Gruber boring muricid gastropod. In: pp. 157-244. H. E. Winn and B. L. 011a (eds.). Behavior of Marine Animals, vol. 1. Plenum. New York. Carriker. M. R. & D. Van Zandt. 1972b. Regeneration of the accessory boring organ of muricid gastropods after excision. Trans. Amer. Mi- cros. Soc. 91:455^166. Carriker. M. R. & L. G. Williams. 1978. The chemical mechanism of shell dissolution by predatory boring gastropods: a review and an hypothesis. Malacologia 17:143-156. Carriker. M. R. & E. L. Yochelson. 1968. Recent gastropod bore holes and Ordovician cylindrical borings. Contr. Paleontology. Geological Sur- vey Prof, paper 593-B:Bl-B26. Carriker, M. R.. D. B. Scott & G. N. Martin. 1963. Demineralization mechanism of boring gastropods, pp. 55-89. In: R. F. Sognnaes (ed.). Mechanisms of Hard Tissue Destruction. Publ. 75. American Associa- tion for the Advancement of Science (AAAS). Washington. DC. Carriker. M. R.. D. Van Zandt & G Charlton. 1967. Gastropod Urosal- pinx: pH of accessory boring organ while boring. Science 158:920- 922. Carriker. M. R . P. Person. A. Smarsh. S. Lipson & H. H. Chauncey. 1968. Role of carbonic anhydrase in decalcification by Urosalpinx cinerea follyensis (oyster drill). Intemat. Assoc. Dental Res. Program Ab- stracts. March 1968. Abstract No. 604. p. 188. Carriker. M. R.. L. G Williams & D. Van Zandt. 1978a. Preliminary char- acterization of the secretion of the accessory boring organ of the shell- penetrating muricid gastropod Urosalpinx cinerea. Malacologia 17: 125-142. Carriker. M. R.. D. Van Zandt & T. J. Grant. 1978b. Penetration of mol- luscan and non-molluscan minerals by the boring gastropod Urosalpinx cinerea. Biol Bull. 155:511-526. Chetail. M. & D. Binot. 1967. Histoenzymologie. Mise en evidence et role de 1'anhydrase carbonique dans 1'organe accessoire de perforation de Puipura lapillus L.. Gasteropode Prosobranche. C.R. Acad. Sc. Paris 264:946-948. Chetail. M. & J. Fournie. 1969. Shell-boring mechanism of the gastropod Purpura {Thias) lapillus: a physiological demonstration of the role of carbonic anhydrase in the dissolution of CaCO,. Am. Zool. 9:893-990. Chetail. M. & J. Foumie. 1970. Mecanisme de perforation chez Thais lapillus L. (Gasteropode Prosobranche. Muricide): mise en evidence d'une entree d'ions calcium durant l'activite de l'organ de perforation. C.R. Acad. Sci. 271:118-121. Chetail. M. & J. Fournie. 1980. Mise en evidence et role des mucopolysac- charide de surface dans 1'organe de perforation de Thais lapillus L. Gasteropode Prosobranche. Haliotis 10. Chetail. M.. D. Binot & M. Bensalem. 1968. Organe de perforation de Purpura lapillus (L.) (Muricide): histochimie et histoenzymologie. Cahiers Biol. Mar. 9:13-22. Chetail. M.. M. Derer & J. Foumie. 1982. L'epithelium de 1'organe de perforation de Thais lapillus L. (Mollusca. Prosobranchia): un epithe- lium transporteur d'ions. Malacologia 22:305-311. Derer. M. 1975. L'organe de perforation de 77ia/s lapillus L. (Gasteropode. Prosobranche). Etude en microscopie optique et electronique. Arch, a' Anal. Micros. Morph. Exp. 64:1-26. Diaz. J. M. & M. Puyana. 1994. Molluscos del Caribe Colombiano. COLCIENCIAS. Fundacion Natura, INVEMAR. Santa Fe de Bogota. Colombia. 291 pp. Disalvo. L. H. & M. R. Carriker. 1994. Planktonic, metamorphic. and early benthic behavior of the Chilean loco Concholepas concholepa (Muri- cidae. Gastropoda. Mollusca). J. Shellfish Res. 13:57-66. Evans, T. B. 1980. Optical and ionic characterization of the secretion of the accessory boring organ of the predatory gastropod Urosalpinx cinerea (Say). Am. Zool. 20:769. Fischer. P -H. 1922. Sur les gasteropodes perceurs. J. Conchyliologie 67: 1-56. Franchini, A.. A. M. Bolognani-Fantin. P. Caselli. 1983. Fine structure of the accessory boring organ of starved and satiated specimens of Ocine- brina edwardsi (Payr. ). J. Exp. Mar. Biol. Ecol. 72:59-66. Fretter. V. 1941. The genital ducts of some British stenoglossan proso- branchs. J. Mar. Biol. Assoc. 25:173-211. Fretter. V. 1946. The pedal sucker and anal gland of some British Steno- glossa. Proc. Malacolog. Soc. Lond. 27:126-130. Fretter. V. & A. Graham. 1994. British prosobranch mollosks: their func- tional anatomy and ecology, vol. 161. Royal Society Publ. 820 pp. Gruber. G. L. & M. R. Carriker. 1990. A note on the accessory boring organ and shell boring by the "loco". Concholepas concholepas Mu- ncidae. Gastropods). Bull. Mar. Sci. 47:739-744. Haller. B. 1892. Die Morphologic de Prosobranchier. III. Naticiden und Calyptraeiden. Morph. Jahrb. 18:451-543. Harasewych, M. G. 1984. Comparative anatomy of four primitive muri- cacean gastropods: implications for trophonine phylogeny. Am. Mala- colog. Bull. 3:11-26. Heminguav. G T. 1975. Functional morphology of feeding in the preda- tory whelk Acanthina spirata (Gastropoda Prosobranchia). Am. Mala- colog. Bull. 41:64-65. Hirsch, G. C. 1915. Die Ernahrungsbiologie fleischfressender Gas- tropoden. Zool. Jahrb.. Abt. allgem. Zool. 35:357-504. Hughes. R. N. & H. P. 1. Hughes. 1981. Morphologial and behavioral as- pects of feeding in the Cassidae (Tonnacea. Mesogastropoda). Mala- cologia 20:385-102. Humphrey. E. C. 1990. Feeding studies of the muricid gastropod Ocenebra erinacea (Linnaeus. 1758). with ultrastructural and histochemical in- vestigations of the ABO (accessory boring organ). Dissertation. Uni- versity of Southampton. UK. 169 pp. Rabat. A. R. 1990. Predatory ecology of natricid gastropods with a review of shell boring predation. Malacologia 32:155-193. Keen. A. M. 1971. Sea shells of tropical west America. 2nd. ed. Marine mollusks from Baja California to Peru. Stanford University Press, Stan- ford. CA. 1064 pp. Kool. S. P. 1993a. Phylogenetic analysis of the Rapaninae ( Neogastropoda: Muricidae). Malacologia 35:155-259. Kool. S. P. 1993b. The systematic position of the Genus Nucella (Proso- branchia: Muricidae: Ocenebrinael. Nautilus 107:43-57. Lo. C. F.. C. H. Wang. F. Huber. C. C. Shen. 1980. The study of the boring organs in ovster drill Thais clavigera Kuester (Gastropoda: Muricidae i. Rept. Fish Disease Res. 3:25-38 (CAPD Fish. Ser. 3). Morton. B. & K. Chan. 1997. First report of shell-boring predation by a member of the Nassariidae (Gastropoda). J. Molluscan Studies 63:476- 478. Nylen. M. U„ D. V. Provenza & M. R. Carriker. 1969. Fine structure of the accessory boring organ of the gastropod. Urosalpinx. Am. Zool. 9:935- 965. Page, L. R. & R. V. K. Pedersen. 1998. Transformation of phytoplanktivo- rous larvae into predatory carnivores during the development of Po- linices lewisii (Mollusca. Caenogastropoda). Invert. Biol. 117:208- 220. Person. P.. A. Smarsh. S. J. Lipson & M. R. Carriker. 1967. Enzymes of the accessory boring organ of the muricid gastropod Urosalpinx cinerea follyensis. 1. aerobic and related oxidative systems. Biol. Bull. 133: 401-110. Ponder. W. F. & J . D. Taylor. 1992. Predatory shell drilling by two species of Austroginella (Gastropoda: Maginellidae). J. Zool. Lond. 228:317- 328. Provenza. D. C. M. U. Nylen & M. R. Carriker. 1966. Some cytologic observations of the secretory epithelium of the accessory boring organ of the gastropods Urosalpinx and Eupleura. Am. Zool. 6:322. Reaumur. R. 1 7 1 1 . De la formation et de l'accroissement des coquilles des animaux tant terrestre qu'aquatiques. soit de mer soit de riviere. Mem. Hist. Acad. Sci. Annee 1 709:364-400. Rosenberg. A. J.. M. Chetail & J. Fournie. 1968. Intervention de 1'anhydrase carbonique dans le mecanisme de perforation des valves de Lamellibranches par Purpura (Thias) lapillus L. (Gasteropode. Proso- branche. Muricide). C.R. Acad. Sci. Paris 266:944-947. Schiemenz. P. 1891. Wie bohrt Natica die Muscheln an? Mitt. Zool. St. Neapel 10:153-169. Uniqueness of Gastropod ABO 595 Simroth. H. 1896-1907. Gastropoda Prosbranchia. Klassen und Ordnungen des Tier-Reichs, Vol. 3. Mollusca. C. F. Winter' sche Verlagshandlung. Leipzig. Smarsh. A.. H. H. Chauncey. M. R. Carriker & P. Person. 1969. Carbonic anhydrase in the accessory boring organ of the gastropod. Urosalpinx. Am. Zool. 9:967-982. Smith. E. H. 1969. Functional morphology of Penitella conradi relative to shell penetration. Am. Zool. 9:869-880. Sohl, N. F. 1969. The fossil record of shell boring by snails. Am. Zool. 9:725-734. Troschel, F. H. 1854. Ueber die Spiechel von Dolium galea. J. prakt. Chemie 63:170-179. Vaught. K. C. 1989. A classification of the living mollusca. American Malacologists, Melbourne. FL. 189 pp. Vermeij. G. J. & S. P. Kool. 1994. Evolution of labral spines in Acanthais, new genus, and other Rapanine muricid gastropods. Veliger 37:414- 424. Webb, R. S. & A. S. M. Saleuddin. 1977. The role of enzymes in the mechanism of shell penetration by the muricid gastropod. Thais lapillus (L.). Can. J. Zool. 55:1846-1857. Wiley, E. O. 1981. Phylogenetics. John Wiley & Sons, New York. 439 pp. Wiltse, W.I. 1980. Predation by juvenile Polinices duplicatus (Say) on Gemma gemma (Totten). ./. Exp. Mar. Biol. Ecol. 42:187-199. Wu. S-K. 1965. Comparative functional studies of the digestive system of the muricid gastropods Drupa ricina and Morula granulata. Malaco- logia 3:211-233. Wu. S-K. 1985. The genus Acanthina (Gastropod: Muricacea) in west America. Spec. Publ. Mukaishima Mar. Biol. Sta. 1985:45-66. Young, D. K. 1969. Okadaia elegans, a tube4x>ring nudibranch mollusk from the central and west Pacific. Am. Zool. 9:903-907. Ziegelmeier, E. 1954. Beobachtungen iiber den Narungserwerb bei der Naticidae Lunatia nitida Donovan (Gasteropoda Prosobranchia). Hel- golander Wiss. Meeresuntersuch. 5:1-33. Journal oj Shellfish Research, Vol. 18, No. 2. 597-600, 1999. EFFECT OF FRESHWATER IMMERSION ON ATTACHMENT OF THE JAPANESE OYSTER DRILL, CERATOSTOMA INORNATVM (RECLUZ 1851) KARL W. MUELLER1 AND ANNETTE HOFFMANN2 Washington Department of Fish and Wildlife ]P. O. Box 1100 La Conner Washington 9X257 2600 Capitol Way North Olympia Washington 98501 ABSTRACT The Japanese oyster drill, Ceratostoma inomatum, has plagued the northeast Pacific oyster industry for most of the 20th century. To prevent its spread in Washington state, the Washington Department of Fish and Wildlife regulates the intrastate transfer of shellfish between growing areas. Immediately after each transfer into or from an area infested with C. inomatum, all working surfaces (e.g., decks of vessels or beds of vehicles) must be rinsed or washed down, preferably with freshwater. The motivation for this rule was based on a previous study that showed that oyster drills detach under hyposaline conditions (7.2-18.0 ppt). The objectives of our study were to test the hypothesis that freshwater (0 ppt) causes C. inomatum to detach and to estimate the freshwater bath time. T so that the probability of oyster drills remaining attached after Tp is less than P (range = 1 x 10"' to 1 x 10"" ). To determine the time to detach in freshwater T. individual C. inomatum (n = 373) were placed on oyster valves at the bottom of a seawater (28.5 ppt) holding tank and allowed to attach themselves. Once attached, the oyster valves and oyster drills were removed and immersed in a bath of freshwater (0 ppt). Detachment was indicated by C inomatum rolling off the oyster valves; T was measured using a digital stopwatch. T ranged from 0.7 to 190.5 s and varied considerably for all size classes. A linear regression of log-normal Ton shell length revealed a significant and positive size correlation. Large C. inomatum took longer to detach than smaller conspecifics. Assuming a log-normal probability distribution for T, Tr of oyster drills measuring s= 40 mm shell length (n = 53) ranged from 82 to 1,213 s (1.4 to 20.2 min) for P = 10"' to 10~LI (i.e., 1 in 10 oyster drills is expected to remain attached after being immersed in freshwater for 82 s. and so on). Our results can be used by resource managers and shellfish growers alike to reduce the risk of C. inomatum infestation to both wild and cultured shellfish stocks. KEY WORDS: Ceratostoma inomatum, oyster drill, shellfish pest control, shellfish transfer protocol INTRODUCTION Washington state (USA) is a leading producer of cultured bi- valves, including the Pacific oyster, Crassostrea gigas (Thunberg 1793) (Cheney and Mumford 1986). The Japanese oyster drill, Ceratostoma inomatum, is a predatory snail (Neogastropoda: Mu- ricidae) that has plagued Washington's oyster industry ever since its accidental introduction to the state during the early 1920s with shipments of C. gigas from Japan (Galtsoff 1932, Quayle 1969, Chew 1990). Predation by C. inomatum can cause major losses to shellfish farmers in Washington. According to one study (Westley 1965) by the Washington Department of Fish and Wildlife (WDFW). formerly the Washington Department of Fisheries ( WDF), in oyster drill-infested areas of the state, mortality in out- planted oyster seed increased by at least 25% because of predation during the first 6 months after planting, resulting in lower total yields. Furthermore, oyster drill control measures, such as trans- planting oyster crops to uninfested areas, increased production costs by 17%. These combined with seed losses. Westley (1965) predicted, decreased net profits by 55% because of predation by C. inornalam. By 1945. C. inomatum was detected in most of the commercial shellfish growing areas of Washington, the result of unchecked oyster transplanting activities during the previous decades. At the time, total eradication of the oyster drill was the favored solution to the problem. In response, the WDF explored a number of pos- sibilities, including chemical molluscicides, barriers, and attrac- tants. However, most methods failed or were considered imprac- tical in the field because of logistic or environmental concerns (Chambers et al. 1972). Therefore, to prevent the spread of oyster drills, the WDF began regulating the intrastate transplantation or transfer of oysters and oystering equipment between growing areas (Glud[Glude] 1947). Given the impact of C. inomatum on the shellfish industry, and the improbability of eradicating the pest, this practice is continued today by the WDFW. The WDFW classifies all marine waters, tidelands, and shell- fish handling facilities within the state as "restricted" with respect to the presence of aquatic diseases, including such harmful pests as C. inomatum. Those marine habitats and facilities lying outside restricted shellfish areas are designated "unrestricted." A WDFW pamphlet (Mueller et al. 1997) describes the guidelines and re- quirements for most types of shellfish transfers between these areas. According to the pamphlet, immediately after each shellfish transfer into or from a restricted area, all working surfaces (e.g., decks of vessels or beds of vehicles) must be rinsed or washed down, preferably with freshwater, at a location where the rinsed material or debris cannot reach tidelands. The purpose of the fresh- water rinse or wash-down is to remove oyster drills from products and equipment used during a shellfish transfer. The original motivation for this condition was based on the observations of Chapman and Banner (1949), who reported that oyster drills [C. inomatum as well as the native oyster drill. Nu- cella lamellosa (Gmelin 1791)] detach under hyposaline condi- tions («18.0 ppt). In their study. Chapman and Banner monitored the activity of oyster drills exposed to different treatment salinities (7.2. 11.6. 15.2. 18.0. 22.4, 26.3. and 28.9 ppt) for 11 d. Each treatment consisted of a single holding jar containing 10 oyster drills immersed in water of prescribed salinity. After 15.5 h expo- 597 598 Mueller and Hoffmann sure to salinities =£18.0 ppt. the percentage of C. inornatum at- tached and crawling on the sides of holding jars declined with decreasing salinity. For example, in 18.0 ppt. 809r of the oyster drills remained attached: whereas, in 7.2 ppt. no oyster drills re- mained attached. One objective of our study was to extend the earlier study of Chapman and Banner by rigorously testing the hypothesis that freshwater (0 ppt) causes C. inornatum to detach. A second objective was to estimate the time to detach in freshwa- ter, as a basis for improving methods of controlling a pest that affects both wild and cultured shellfish stocks alike. MATERIALS AND METHODS The study was conducted at the WDFW's Point Whitney Shell- fish Laboratory located on the Hood Canal, Puget Sound, Wash- ington during 18 June to 14 August 1996 and 12 to 14 May 1997. Several hundred juvenile and adult C. inornatum were collected from the Point Whitney Lagoon (47 °45'44"N. Lat: 122 °51 '07"W. Long.) and maintained in an aerated, seawater (28.5 ppt) holding tank (53 L) at ambient temperature (16-18 °C). To determine the time to detach in freshwater T, we placed individual C. inornatum (n = 373; size range = 5.7-50.5 mm shell length; Fig. I ) on single oyster valves at the bottom of the holding tank. Oyster drills were allowed to attach themselves, which was indicated by crawling movements across the oyster valves (elapsed time <2 min). The oyster valves and attached oyster drills were then removed from the holding tank by hand, tilted slightly (<30 ° from horizontal), and immersed in a plastic- tub filled with freshwater (0 ppt) at ambient temperature. Detach- ment was indicated by C. inornatum rolling off the tilted oyster valves; time to detach was measured for each oyster drill using a digital stopwatch. To serve as a control, and to confirm that fresh- water rather than handling caused C. inornatum to detach, we followed the same procedures described above, using an additional 50 oyster drills (size range = 8.5—44.6 mm shell length) with one difference: C. inornatum and the oyster valves were held in a bath of seawater instead of freshwater. During preliminary testing, we observed longer times to detach for larger oyster drills. Based on the skewed shape of the histogram of times to detach in freshwater (Fig. 2). we log-transformed the data, then conducted a linear regression to confirm that size and time to detach were significantly and positively related. The linear regression model was ln(7",) = a + B (Z.,) + e, 120 100 - •o 80 - 22 ° 60 - c | 40 2 20 - 140 120 100 - o nn 80 120 Time to detach (sec > in freshwater Figure 2. Frequency distribution of the time to detach in freshwater for all C. inornatum in treatment group (n = 373; size range = 5.7-50.5 mm shell length). where Ti was the time to detach of the ith oyster drill, L,was the length of the ith oyster drill, and e, - normaHO. o2). Because B was significantly positive, the oyster drills in the largest size class (>40 mm shell length: Fig. 1) were expected to have the longest times to detach. Consequently, we estimated freshwater bath time. Tp, so that the proportion of the largest oyster drills that remain attached alter Tp is less than P (range = I x 10~' to 1 x 10""). Because the regression was conducted on ln(7" ), we first esti- mated the required bath time. T on a log scale and then exponen- tiated the result to obtain the required real time. T„ To estimate T„ we used the condition that the probability of ln(7~) being greater than 71 must be p. that is 20 - 29 30 - 39 Shell length I mm l Figure 1. Length-frequency distribution of C. inornatum in treatment group (n = 373). P(ln(7~) > Tp) = p This led to the solution Tp = Zpa + p. where ln(7") was normally distributed with mean p, and standard deviation o\ and Zf, was the ( 1 - /))th quantile from a standard normal distribution (Zar 1984). We estimated p. and a from the ln(D of the oyster drills in the largest size class, so that 9p = Z/ya + p. which resulted in an estimate of Tp, as tp = exp( ip) RESULTS Time to detach in freshwater ranged from 0.7 to 190.5 s, or roughly 3 min (Fig. 2). Although there was considerable variation in time to detach within size classes, a linear regression of the natural log of the time to detach on shell length revealed a slight, but positive significant relationship (f < 1 x I0"'\ n = 373). On average, large C. inornatum took longer to detach in freshwater than smaller conspecifics (Fig. 3). In the control group, all but one of the oyster drills (n = 50) remained attached to the oyster valve after 210 s. This individual (size = 23.8 mm shell length) de- tached after 178 s while crawling along the thin lip of the oyster valve (i.e., lost its foothold and fell). The estimated mean and standard deviations for ln(D of the largest oyster drills (S40 mm shell length, n = 53) were (1 = 3.68 and & = 0.57. The estimated freshwater bath times, Tp, so that the proportion of the largest oyster drills remaining attached after T was less than p are given in Table 1. For example, after being Immersion of Oyster Drills in Freshwater 599 — 5 v = 2.47 + 0.03x r - = 0.16 10 20 30 40 Shell length (mm) 60 Figure 3. Relationship of the natural log of the time to detach (s) on shell length (mm) of C. inornatum immersed in freshwater (P < 1 x 1015; n = 373). soaked in freshwater for 82 s, one in ten of all C. inornatum >40 mm shell length are expected to remain attached. DISCUSSION The use of fresh- and brackish water has long been declared effective in controlling oyster drills. For example, Federighi (1930) described a procedure called "floating." wherein oysters infested with the Atlantic oyster drill. Urosalpinx cinerea (Say) 1822, were immersed in large containers of brackish water (12-14 ppt) for 10 days. Evidently, this treatment was sufficient to kill U. cinerea without harming the oysters. Butler (1953) concluded that, although the southern oyster drill, Stramonita haemastoma Lin- naeus, 1758, was capable of surviving prolonged exposure to freshwater, the only way to prevent the spread of S. haemastoma was through sustained hyposaline conditions (<15 ppt). Accord- ingly. Pollard ( 1973) proposed the diversion of freshwater from the Mississippi River to control populations of S. haemastoma by flooding Louisiana state oyster grounds with brackish water. A follow-up study by Breithaupt and Dugas (1979) indicated that Pollard's proposal was plausible and, since 1991. S. haemastoma has been eliminated from some oyster grounds because of a suc- cessful freshwater diversion program (Greg Laiche, Louisiana Wildlife and Fisheries Commission, pers. comm.). Our study showed unequivocally that freshwater causes C. in- ornatum to detach from its foothold. We also showed that time to detach was positively correlated to size and that the bath times required to remove the largest oyster drills could be over 1.200 s or 20 min. Because the regression was significantly positive, the largest oyster drills represented the "worst case scenario," requir- TABLE 1. Estimated freshwater bath times Tr, so that the proportion of the largest C. inornatum (240 mm shell length) remaining attached after Tp is less than P (range = 1 x 10"1 to 1 x 10"''). Values are based on the log-normal distribution of bath times. T (n = 53), with ii = 3.68. d = 0.57, and Z„ = (1 - pth) quantile from a standard normal distribution (Zar 1984). p z. K = exp{(0.57 x Zp) + 3.68) 1 x 10"" (1 in 1 billion) 6.00 1,213 s (20.2 mm) 1 x 10"6 (1 in 1 million) 4.75 596 s (9.9 min) 1 x 10"' (1 in 1 thousand) 3.09 231 s (3.9 mini 1 x 10"' (1 in ten) 1.28 82 s (1.4 min) ing the longest bath times. Because the motivation for this study was removal of most, if not all. oyster drills, estimating the fresh- water bath times for the largest individuals was the most conser- vative approach, precluding the need to estimate bath times for smaller oyster drills. We confirmed the observations made by Chapman and Banner ( 1949) that hyposaline conditions cause oyster drills to detach and support the recommendation to use freshwater, if available, when rinsing or washing down working surfaces, products, or equipment after each shellfish transfer into or from areas infested by C. in- ornatum. Our results strongly suggest that a short, haphazard freshwater rinse or wash-down will not be sufficient to remove all oyster drills, because C. inornatum is capable of maintaining its foothold for potentially 1,213 s or 20.2 min when completely immersed in freshwater (Table 1 ). Still, we have shown that longer exposures to freshwater substantially decrease the probability of C. inornatum maintaining its foothold. Our results can be used by resource managers and shellfish growers alike when developing shellfish transfer protocol to reduce the risk of C. inornatum in- festation to both wild and cultured shellfish stocks (sensu Elston 1992). ACKNOWLEDGMENTS We thank R. T. Burge. J. H. Beattie. W. A. Bradbury, and R. E. Sizemore of the Washington Department of Fish and Wildlife's (WDFW) Point Whitney Shellfish Laboratory for encouragement and support. This study was conducted during the first author's tenure managing the WDFW shellfish disease, pest, and predator control program. The first author dedicates this work to the memory of Elizabeth C. Mueller, who adored her "beautiful, but deadly," muricids. LITERATURE CITED Butler. P. A. 1953. The southern oyster drill. Proc. Natl. Shellfish. Assoc. 44:67-75. Breithaupt. R. L. & R. J. Dugas. 1979. A Study of the southern oyster drill [Thais ( = Stramonita) haemastoma]: distribution and density on the oyster seed grounds. Louisiana Wildlife and Fisheries Commission Tech. Bull. 30. 20 pp. Chambers, J., M. Fraidenburg. T. Mecklenburg & W. Hoffman. 1972. Oyster drill investigations. Washington Department of Fisheries Man- agement Research Division Comp. Rept. 31 pp. Chapman. W. M. & A. H. Banner. 1949. Contributions to the life history of the Japanese oyster drill, Tritonalia japonica ( = Ceratostoma inorna- tum). with notes on other enemies of the Olympia oyster, Ostrea lurida. Washington Department of Fisheries Biology Bull. 49:169-200. Cheney. D. P. & T. F. Mumford. 1986. Shellfish and seaweed harvests of Puget Sound. Washington Sea Grant Program. Seattle. WA. 164 pp. Chew, K. K. 1990. Global bivalve introductions. World Aquacult. 21:9-22. Elston, R. A. 1992. Effective application of aquaculture disease control regulations: recommendations from an industry perspective, pp. 353- 359. In: A. Rosenfield and R. Mann (eds.). Dispersal of Living Organ- isms into Aquatic Ecosystems. Maryland Sea Grant Program. College Park, MD. Federighi, H. 1930. Control of the common oyster drill. U. S. Bureau of Fisheries Econ. Circ. 70. 7 pp. Galtsoff. P. S. 1932. Introduction of Japanese oysters into the United States. U. S. Bureau of Fisheries Fish. Circ. 12. 16 pp. 600 Mueller and Hoffmann Glud [Glude], J. B. 1947. Oyster investigation, pp. 11-14. In: 1947 Reports grounds and to control the southern oyster drill. Louisiana Wildlife and on Dungeness Crabs, Pacific and Olympia Oysters, and the Japanese Fisheries Commission Tech. Bull. 6. 82 pp. Clam Industry. Washington Department of Fisheries, Olympia, WA. Quayle. D. B. 1969. Pacific oyster culture in British Columbia. Fisheries Mueller, K., B. Sizemore & L. Timme. 1997. Guidelines and Requirements Research Board of Canada Bull. 169. 192 pp. for the Import and Transfer of Shellfish, Including Oysters. Clams, and Westley, R. E. 1965. Impact of drills on oyster culture in Washington State. Other Aquatic Invertebrates in Washington State. Washington Depart- Washington Department of Fisheries. Olympia, WA. 3 pp. ment of Fish and Wildlife. Olympia, WA. 32 pp. Zar. J. H. 1984. Biostatistical Analysis, 2nd ed. Prentice-Hall. Upper Pollard, J. F. 1973. Experiments to re-establish historical oyster seed Saddle River, NJ. 718 pp. Journal of Shellfish Research, Vol. 18, No. 2. 601-604. 1999. FATTY ACIDS AND STEROLS OF RAPANA VENOSA (VALENCIENNES, 1846) KASIM CEMAL GUVEN,1 ZELIHA YAZICI,2 SERAP AKINCI,1 AND ERDOGAN OKU§' Institute of Marine Sciences and Management, University of Istanbul 34470, Vefa. Istanbul 'Department of Pharmacology, Cerrahpasa Faculty of Medicine, Istanbul, Turkey ABSTRACT The composition of fatty acids and sterols was investigated in different organs of Rapana venosa. The isolated lipids were saponified then esterified. and fatty acid esters were analyzed by gas chromatography (GO. The total fatty acid content in the hepatopancreas was > right massive gland salivary > gonad > flesh. In the flesh, the ratio of saturated to unsaturated fatty acids was 0.50. The highest amounts of fatty acids in the flesh were stearic acid in the saturated fatty acids, and n-3 and n-6 acids in the polyunsaturated fatty acids at 0.30, 0.32. and 0.38 (j.g/mg. respectively. The sterols isolated from R. venosa were analyzed by gas chromatography/mass spectrometry (GC/MS). Two sterols were identified in the flesh and nine sterols in the whole organs. KEY WORDS: Rapana venosa organs, fatty acids, sterols INTRODUCTION Lipids and sterols of marine invertebrates have been intensively studied, but relatively little is known for shellfish, Rapana venosa (Valenciennes 1846) (formerly R. thomasiana thomasiana), Mol- lusca, Gastropoda. In shellfish, lipids percentage (Rosoiu and Ser- ban 1981, Rosoiu and Panait 1992) and whole body fatty acid content of lipids (Christie et al. 1988) were investigated. Cholesterol is the only sterol reported in R. venosa (Tsujimoto and Koyanagi 1934). Other contents of R. venosa, such as heparin (Giiven et al. 1991 ), insulin, and some enzymes (Akinci et al. 1998 a, Akinci et al. 1998b. Akmci et al. 1998c) were also investigated. Fatty acids (FAs) composition in marine organisms includes highly polyunsaturated fatty acids (PUFAs). The medically impor- tant PUFAs are 20:5n-3, eicosapentaenoic acid (EPA) and 22:6n-3 docosahexaenoic acid (DHA). These PUFAs can reduce the risk of cardiovascular diseases and reduced platelet aggregation (Seidelin et al. 1992; Eritsland et al. 1995; Daviglus et al. 1997). Shellfish as a food source in such countries as Japan entails the need to inves- tigate fatty acids as nutritional components. In this paper, the com- position of fatty acids and sterols in different organs of R. venosa is reported. MATERIALS AND METHODS Rapana venosa was collected between July to September 1997 from the Black Sea near the entrance of the Bosphorus at a depth of 20-30 m and stored at -30 °C. Flesh, hepatopancreas, right massive gland salivary, and gonad were separated according to Lupu (1977) and stored at -70 °C until analysis. The standard fatty acid methyl esters were obtained from Sigma and authentic cho- lesterol from Merck. Total Lipid Extraction The frozen organs were thawed, carefully weighed (200-300 mg), and homogenized in cold 154 mM NaCl. The total lipids were extracted according to the method of Folch et al. (1957) modified by Yazici et al. ( 1994). 0.1 ml internal standard (200 u.g heptadecanoic acid in chloroform), 0.1 ml of methanol containing 20 g/1 butylated hydroxytoluene, as antioxidant. 2 ml methanol, and 4 ml chloroform added to the homogenated tissue. The sample was mixed vigorously using a vortex for 2 min, followed by cen- trifugation at 2.000 g for 10 min at 4 °C. The chloroform phase was separated and evaporated to dryness at 37 °C under a stream of nitrogen. Saponification of Lipids, Methylation and Analysis of Fatty Acid Esters The total lipid extract was hydrolyzed in 8 ml of KOH/ metha- nol (2% v/w) under reflux in a water bath for 30 min. After cooling, 1 ml of 14 % BFrmethanol was added and heated for 2 min at 100 °C than cooled, 5 ml distilled water and 2 ml hexane were added, vortex-mixed for 2 min. and centrifuged (2,000 g; 10 min at 4 °C). The hexane phase was separated, evaporated to dryness under a stream of nitrogen, and the residue was dissolved in hexane in a volumetric flask and applied to gas chromatography (GO. Isolation of Sterols Sterols were extracted from the flesh and in whole organs of R. venosa with dichloromethane (DCM) in Soxhlet apparatus for 4 h. The extract was distilled and hydrolyzed with 5 % KOH in metha- nol for 30 min under reflux in water bath. After hydrolysis, two volumes of water were added and extracted with DCM. and the organic phase was separated then distilled. The residue was col- lected with hexane and applied to gas chromatography/mass spec- trometry (GC/MS) analysis. GC Analysis The FA methyl esters were analyzed by capillary gas chroma- tography (Perkin-Elmer 8420 Capillary Gas Chromatography. Gouda, The Netherlands). Column: 25 x 0.25 mm ID, QC2/BP x 70, 0.25 p.m film; flame ionization detector temperature 300 °C; split injector temperature 300 °C; oven temperature program from 150 to 230 °C at 2 °C min; carrier gas N2. FAs were identified by their retention time and compared to those of the standards. Their amounts were estimated by calculat- ing the corresponding areas. GC/MS Analysis The analysis of sterols was run on an HP6890 capillary gas chromatograph connected to an HP MSD and controlled by an HP 601 602 GtJVEN ET AL. TABLE 1. Fatty acid profiles of lipids of Rapana venosa organs. Right Massive Gland Fatty Acid Flesh Hepatopancreas Salivary Gonad P 14:0 0.07 ± 0.00 B 2.21 ±0.27C 1.28 ±0.17 A 0.2 1 ± 0.03 B 0.0000 15:0 0.04 ±0.01 B 0.62 ± 0.03 A 0.42 ±0.11 A 0.42 ± 0.07 A 0.0016 15:0 iso 14-methyl Trace 0.20 ± 0.02 0.14 + 0.06 Trace 16:0 0.20 ±0.01 B 5.63 ± 0.52 C 3.35 ± 0.73 A 0.65 ± 0.09 B 0.0001 I6:ln-7 0.09 ± 0.00 B 1.73 ±0.17 C 0.64 ± 0.05 A 0.23 ± 0.04 B 0.0000 16:ln-9 trace 0.49 + 0.21 B 0.24 ± 0.09 A 0.06 + 0.01 A 0.0084 16:0 iso 14 ■methyl trace 0.45 + 0.06 0.27 ±0.06 trace 16:0 iso 15- methyl 0.02 ±0.01 B 0.33 ± 0.02 C 0.24 ± 0.03 A 0.09 ± 0.02 B 0.0006 17:0 iso 16- methyl 0.09 ± 0.02 D 0.84 ± 0.04 C 0.61 ±0.04 A 0.31 ±0.04B 0.0001 17:0 iso 15-methyl 0.07 ± 0.02 B 0.53 ± 0.09 A 0.31 ±0.10AB 0.23 ± 0.06 B 0.0137 18:0 0.28 ±0.01 D 2.02 ±0.01 C 1.44 ±0.28 A 0.89 ± 0.03 B 0.0001 18:ln-13 0.07 ±0.01 A 0.45 + 0.03 B 0.27 ± 0.08 A 0.19 + 0.04 A 0.0090 18:ln-9 trace 0.68 ± 0.090 0.34 ±0.18 0.06 ±0.01 KS: In- 1 1 trace 0.5 1± 0.05 B 0.22 ± 0.08 A 0.03 ±0.01 A 0.0090 18:ln-7 trace 0.34 + 0.02 B 0.16 + 0.03 A trace 0.0115 18:2n-6 0.09 ± 0.07 B 0.40 ± 0.03 A 0.29 ± 0.07 A 0.32 ± 0.05 A 0.0043 18:3n-3 0.01 ±O.00B 0.35 ± 0.04 A 0.21 ±0.06 A 0.09 ± 0.00 B 0.0140 20:0 trace 0.36 ± 0.05 0.18 ±0.09 trace 20:ln-9 0.12 ± 0.02 B 4.78 + 0.47 A 2.94+ 1.31 A 0.50 + 0.03 B 0.0047 20:1 0.02 ± 0.00 B 1.20 + 0.01 C 0.78 + 0.12 A 0.08 + 0.01 B 0.0001 20:1 0.02 ± 0.00 B 1.17 ±0.13 C 0.67 ±0.12 A 0.10 ± 0.02 B 0.0012 20:2n-9 0.03 ±0.01 B 1.01 ±0.29 A 0.55 ± 0.28 AB 0.22 ± 0.06 AB 0.0401 20:2n-7 0.05 ±0.01 0.36 ± 0.24 0.30 ± 0.05 0.35 ± 0.04 20:3n-6 0.03 ± 0.00 D 0.33 ± 0.05 C 0.23 ±0.01 A 0.12 ±0.01 B 0.0002 20:4n-6 0.19 ±0.00 1.49 ±0.06 1.24 ±0.19 1.59 + 0.59 0.0456 20:5n-3 0.08 + 0.01 B 3.24 ± 0.17 C 1.91 ±0.57 A 0.64 ± 0.09 B 0.0003 22:2n-9 0.15 + 0.I0C 4.57 ± 1.17 A 3.10 ± 0.80 B 1.09 ± 0.09 BC 0.0093 22:2n-7 0.13 ±0.01 B 2.14±0.14A 1.64 ±0.27 A 0.53 ± 0.07 B 0.0001 22:3n-6 0.03 ± 0.00 B 0.51 ±0.06 A 0.38 ±.01 2 A 0.15 ±0.01 B 0.0031 22:4n-6 0.05 ±0.01 C 0.26 ± 0.03 A 0.24 ± 0.05 A 0.39 ± 0.03 B 0.0007 22:5n-3 0.14 ±0.01 B 1.34 ±0.21 A 1.21 ±0.18 A 0.52 ± 0.06 B 0.0009 22:6n-3 0.10 ±0.02 B 3.85 + 0.41 A 2.49 ±0.93 A 0.52 ± 0.06 B 0.0026 n-3 0.32 ± 0.04 B 8.78 ±0.77 A 5.83 ± 1.72 A 1.77 ± 0.20 B 0.0009 n-6 0.38 ±0.01 B 3.00 ±0.23 A 2.38 ± 0.40 AB 2.24 ± 0.94 AB 0.0360 n-9 0.40 ±0.01 B 12.74 ±2.03 A 7.92 ± 2.24 A 1.91 ±0.50B 0.0016 Saturated 0.77 ± 0.04 B 13.19 ± 0.95 C 8.24 ± 1.69 A 2.99 ±0.31 B 0.0001 Unsaturated 1.4 ± 0.10 B 30.52 ± 3.40 C 20.05 ± 5.34 A 7.78 ± 1.10 B 0.0007 saturated/unsaturated 0.55 ± 0.50 0.43 ±0.16 0.41 ±0.03 0.384 ±0.19 Total 2.17 ± 0.14 B 43.71 ±3.33C 28.29 ± 7.02 A 10.77 ± 1.34 B 0.0003 One-way analysis of variance. The means are compared by the Student-New man-Keuls test: values with no common capital letter differ at P < .05. Mean values are from three replicate (u.g/mg wet weight). ChemStation. Capillary column; Column 50 m x 200 u.m ID. fused HP PON A (methyl siloxane). Column temperature program was from 1 10 °C to 290 °C at 6 °C min"1 and 290 °C at 10 min-1; split injector temperature 250 °C: carrier gas helium, 44.7 psi. The cholesterol was identified using a cholesterol standard, and the other sterols were identified by comparing the spectrum of each peak with its corresponding spectrum from HP memory. RESULTS AND DISCUSSION Fatty acid content of lipids isolated from different organs of R. venosa are shown in Table 1 . The main saturated FAs were pal- mitic and stearic acid in all organs studied. Palmitic (16:0) and stearic (18:0) acid contents were lower in flesh as compared to the other organs tested. The important PUFAs, were EPA and DHA with percentages of the total FAs in the flesh of 3.7 EPA, 4.6 DHA, in the hepatopancreas 7.4 EPA. 8.8 DHA. in the right mas- sive gland salivary 6.8 EPA. 8.8 DHA. and in the gonad 5.9 EPA and 4.8 DHA. EPA level was lower in the flesh than in the hepato- pancreas and the right massive gland salivary. The amount of EPA TABLE 2. Pentaenoic and hexaenoic acids level in R. venosa and fish flesh ( It ). Animals Acids R. venosa Salmon ia Herring" Mackerel" Pentaenoic Hexaenoic 10.1 4.6 8 16 7 5 7 5.5 ' Notevarp and Cyvin ( 1962). Fatty Acids of Rapana Venosa 603 TABLE 3. Sterols identified in R. venosa whole organs (1), and flesh (2). Sterols GC (Rt) MS Peaks 22 Dehydrocholesterol ( 1 ) 33.002 27-Norergosta-5, 22-dien-3. ol, (3. beta. 22 Z) ( 1 ) 33.512 Cholesterol (1). (2 1 34.450 7-Dehydrocholeslerol ( 1 ) 35.130 Crinosterol ( 1 ) 35.473 24-Methyl-24-dehydrocholesterol (1) 36.814 24. beta Methyl cholesterol ( 1 ) 37.058 Cholest-4 en-3 one ( 1 ) 37.501 Cholesterilene ( 1 ). (2) 31.584 55, 69. 95. 111. 145. 159. 213. 255. 271, 300, 351, 366. 384 (M+) 55. 69, 95. 111. 145. 161. 213. 255. 273. 300. 351. 366. 384 (M+l 55, 95, 145, 161, 213. 231, 255. 275, 301, 353. 368 lM+1 55, 95, 1 19. 145. 159. 211, 325. 351, 366. 384 (M+) 69, 95. 133. 159. 255. 271, 300, 337. 355. 380, 398 (M+) 55, 81. 105. 121. 145, 161. 185. 213, 229. 271. 299. 314, 365, 383. 398 (M+) 55, 81, 95. 119. 145. 161, 213. 273. 289, 315, 382,400 (M+) 55. 95. 124. 147. 229, 245. 261, 342. 384 (M+) 55, 81, 95, 107, 120, 133, 147. 159. 213. 247. 260, 326. 353 (M+) (0.08 u.g/mg) and DHA (0.10 p.g/mg) was similar to that of arachi- donic acid (0.19 p-g/mg). EPA and DHA competitively inhibits the utilization of arachidonic acid (Dyerberg et al. 1978). Notevarp and Cyvin (1962). studied pentaenoic and hexaenoic acid levels in fish flesh. Their levels were high in fish but low in animal lipids. Table 2 shows the comparison of the pentaenoic (EPA. C20:5n-3; Docosapentaenoic acid, C22:5n-3) and hexaenoic (DHA, C22:6n-3) acids level in flesh of R. venosa with fish. The pentaenoic acids level was higher in R. venosa than in salmon, herring, or mackerel (Table 2). The amount of saturated FAs in flesh was 0.77 p,g/mg in com- parison to 1.40 p.g/mg unsaturated FAs. So far. little is known about the exact FAs composition of different organs of R. venosa. Total organ lipid content (Rosoiu and Serban 1981; Rosoiu and Panait 1992) and total body fatty acids content (Christie et al. 1988) were investigated earlier. In the present investigation, quan- titative and qualitative studies were carried out on R. venosa FAs. Comparison of our findings to those of Christie et al. (1988) as calculated from their tables on saturated and unsaturated fatty acid levels shows similarity (the percentage of saturated 29.20, 29.66, and unsaturated 70.70. 70.34, respectively). The ratio of saturated to unsaturated fatty acids were 0.55 in flesh. 0.43 in hepatopancreas. 0.41 in right massive gland salivary. and 0.38 in gonad, thus unsaturated FAs were twice the level of the saturated FAs. The total fatty acid (saturated and unsaturated FAs) contents were ranked as hepatopancreas greater than gland salivary greater than gonad greater than flesh. When the lipid levels of Mytilus galloprovincialis (Christie et al. 1988) are compared with the flesh, which is the consumed part of/?, venosa, the lipid content of the latter was lower than the total mussel consumed. In R. venosa, cholesterol was the only sterol that was reported by Tsujimoto and Koyanagi, (1934). In this work, nine sterols were identified in whole organs and two in the flesh of R. venosa (Table 2). Sterols were identified by using standard cholesterol and the others from the HP memory. GC/MS spectral data are given in Table 3. The mass peaks were compared in the spectra of the lipid extract obtained from R. venosa organs with the spectra taken from the HP memory. The similarities of both spectra (quality 96-99) were noted. In contrast, more sterol compounds were isolated from M. galloprovincialis, and differed from those of R. venosa. Con- sidering the lipid and sterol contents of R. venosa, its flesh has an advantage over mussels. In conclusion, the flesh of R. venosa can be considered as suitable for human consumption based on its content of FAs and sterols. Akinci, S.. K. C. Guven, M. Kuciik, M. Hacibekiroglu. M., H. Koyuncuo- glu. & E. Oku§. 1998a. Insulin from Rapana venosa (Valenciennes 1846). Pharmazie 53:650-651. Akinci, S„ M. Hacibekiroglu. M. Kucuiik. E. Okus & K. C. Guven. 1998b. Enzymes of hepatopancreas and thromp of Rapana venosa (Valenci- ennes 1846). Turkish J. Mar. Sci. 4: 29-38. Akinci. S.. K. C. Guven. M. Hacibekiroglu. M. Kiiciik. & E. Okus. 1998c. Enzyme activities of right massive gland salivary of Rapana venosa (Valenciennes 1846). Acta Pharm. Turcica. 30:197-200. Christie. W. W.. E. Y. Brechany & K. Stefenov. 1988. Silver ion high- performance liquid chromatography and gas ehromatography-mass spectrometry in the analysis of complex fatty acid mixtures: application to marine invertebrates. Chem. Pins. Lipids 46:137-135. Daviglus. M. L.. J. Stamler, A. J. Orencia, A. R. Dyer, K. Liu. P. Greenland M. K. Walsh. D. Morris & R. B. Shekelle.1997. Fish consumption and the 30-year risk of fatal myocardial infarction. N. Engl. J. Med. 336: 1046-1053. Dyerberg. J.. H. O. Bang. G. Stoffersen. S. Moncada & J. R. Vane. 1978. LITERATURE CITED Eicosapentaenoic acid and prevention of thrombosis and atherosclero- sis. Lancer 2:117-119. Eritsland. J.. H. Arnesen. I. Seljeflot & T. A. Hostmark. 1995. Long-term metabolic effects of n-3 polyunsaturated fatty acids in patients with coronary artery disease, ,4m. J. Clin. Nutr. 61:831-836. Folch, J.. M. Lees & G. H. Stanley. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:497-509. Guven. K.C., Y. Ozsoy. B. Oztiirk. B. Topaloglu & O. N. Ulutin. 1991. Raparin. a new heparinoid from Rapana venosa (Valenciennes). Phar- mazie 40:547-548. Lupu. D. 1977. Contribution a l'etude de l'anotomie chez Rapana thoma- siana Crosse 1862 (Gastropoda. Muricidae. Rapaninae) de la mer Noire. Trav. Mus. Hist. Nat. "Gr. Antipa". 18:57-65. Notevarp, O. & B.N. Cyvin. 1962. Polyunsaturated fatty acids in fish fat. in the diet and in the blood, pp. 286-291. In: E. Heen and R. Kreuzer (eds.). Fish in Nutrition. Fishing News (Books) Ltd.. London. 604 GiJVEN ET AL. Seidelm. K. N.. B. Myrup & B. Fisher-Hansen. 1992. N-3 fatty acids in adipose tissue and coronary artery disease are inversely related. Am. J. Clin. Nun: 55:1117-1119. Tusijimoto. M. & H. Koyanagi. 1934. The sterols in marine invertebrates: composition, biosynthesis, and metabolites. U. Soc. Chem. Ind. Japan. 37: 436B: through; L.J. Goad), p. 98. In: P.J. Scheuerted.). Marine Natural Products. Academic Press, Inc.. San Diego. 1978. Rosoiu. N. & M. Serban. 1981. Quelques donnees biochimiques chez Rapana thomasiana Grosse. Rapp. Comm. Int. Mer. Medit. 27: 31-32. Rosoiu, N. & M. Panait. 1992. Donnees biochimiques sur les principales especes d'invertebres du littoral Roumain de la mer Noire. Rapp. Comm. Int. Mer. Medit. 33: 201. Yazici, Z.. 1. A. Zavares. I. F. Stamford. P. M. Bishai & A. Bennett. 1994. Changes in tissue fatty acid composition in murine malignancy and following anticancer therapy. Br. J. Cancer, 65:163-170. Journal of Shellfish Research, Vol. 18, No. 2, 605-609. 1999. ANALYSIS OF KARYOTYPE, CHROMOSOME BANDING, AND NUCLEOLUS ORGANIZER REGION OF PACIFIC ABALONE, HALIOTIS DISCUS HANNAI (ARCHAEOGASTROPODA: HALIOTIDAE) SEI-ICHI OKUMURA, SHOUJIRO KINUGAWA, AIKO FUJIMAKI, WATARU KAWAI, HIDETAKA MAEHATA, KAZUHIRO YOSHIOKA, RYOUKO YONEDA, AND KUNIO YAMAMORI School of Fisheries Sciences Kitasato University Sanriku Kesen Iwate 022-0101 Japan ABSTR.ACT Chromosome preparations of Haliotis discus hannai larvae were subjected to karyotype analysis using a scanning electron microscope (SEM). banding analysis using a salt solution treatment, and nucleolus organizer region (NOR) analysis by silver staining. Standard values of the relative length and arm ratio of each chromosome pair in this species were determined by SEM measurements of the chromosome arm length. The arm ratios indicated that this species possessed 1 1 pairs of metacentric and seven pairs of submetacentric chromosomes (2 n = 36). This is the first study of Haliotidae in which banding is attempted (i.e.. not partial banding, such as C-banding ) as well as NOR analyses. The characteristic banding pattern served as a useful marker for the identification of homologous chromosomes that were difficult to identify using only morphological analysis. The NORs were located terminally on the long arms of two chromosome pairs with variations in the NOR-bearing chromosome. The techniques of karyotype identification, banding, and NORs employed in this study may serve as useful indexes in future research on cytogenetics in Haliotidae. KEY WORDS: chromosome, karyotype, banding. NOR, abalone, Haliotis discus hannai, moliusca INTRODUCTION Abalone (Haliotidae) is an economically important marine product in Japan. Therefore, wide-ranging studies on Haliotidae from various perspectives, including seed production, taxonomy, genetics, and ecology have been intensively conducted. Attention has focused on cytogenetics, because this field furnishes useful information on chromosome manipulation and taxonomy in Hali- otidae. To conduct cytogenetic studies, it is necessary to establish chromosome identification techniques for reliable comparisons of intra- and interspecific karyotypes. Chromosomal studies on Haliotidae have been performed using Haliotis aquatilis (Nishikawa 1962, Nakamura 1986), H. crach- erodii (Minkler 1977). H. discus hannai (Arai et al. 1982). H, d. discus (Arai et al. 1982. Miyaki et al. 1997). H. lamellosa (Co- lombera and Tagliaferri 1983), H. tuberculata (Colombera and Tagliaferri 1983, Arai and Wilkins 1986). H. diversicolor aquatilis (Nakamura 1985). H. varia (Nakamura 1986. Arai et al. 1988, Jarayabhand et al. 1998), H. gigantea (Nakamura 1986, Miyaki et al. 1997), H. planata, H. diversicolor diversicolor (Arai et al. 1988), H. asinina, H. ovina (Jarayabhand et al. 1998). and H. madaka (Miyaki et al. 1999). These studies have revealed the chromosome number (2 n = 28 ~ 36) in 14 species of the above- mentioned Haliotidae. In most of these abalone species, morpho- logical classification of the chromosome (metacentric, submeta- centric, or others) has been established. However, none of these researchers has described the concrete numerical values of relative length, arm ratio, or centromeric index of the chromosomes, or reported the chromosome banding. The only chromosome banding study on Haliotidae was performed with C-banding in H. d. hannai (Okumura et al. 1995). Therefore, at this time the systematic ap- proaches to Haliotidae cytogenetics based on reliable measurement of chromosome size and morphology (relative length and arm ratio) and banding patterns are difficult. Accordingly, the estab- lishment of reliable indexes of karyotypes based on accurate mea- surements of chromosome arm length and the application of band- ing or nucleolus organizer region (NOR) analysis in Haliotidae are required. In moliusca, some species of Bivalvia and Gastropoda have been analyzed with NORs and chromosome banding. Chromosom- al NORs of Cerastoderma glaucu (Thiriot-Quievreux and Wolow- icz 1996), Mytilus galloprovincialis (Martinez-Exposito et al. 1997), Ostrea angasi (Li and Havenhand 1997), and Nucella lapil- lus (Pascoe and Dixon 1994, Pascoe et al. 1996) have been ana- lyzed by the silver staining method. In addition. C-banding has previously been employed in investigations of Ostrea denselamel- losa (Insua and Thiriot-Quievreux 1991 ) and M. galloprovincialis (Martinez-Exposito et al. 1997). Furthermore, a banding technique has been established for M. galloprovincialis using 2 x SSC/ Giemsa-stain treatment (Mendez et at. 1990). Based on these stud- ies, banding and NOR patterns have been shown to be useful markers not only for the identification of homologous chromo- somes, but also to serve as effective indexes for discriminating intra- and interspecific genetic differences. Because these ap- proaches have been successfully performed in the moliusca species noted above, banding and NOR analyses on Haliotidae may pro- vide an avenue for differentiating chromosomal construction in Haliotidae. In the present study, we demonstrated a reliable chromosome index from accurate measurements of the chromosome length us- ing a scanning electron microscope (SEM) and applied the banding and Ag-NOR-staining techniques to H. d. hannai chromosomes. MATERIALS AND METHODS H. d. hannai larvae were obtained at 15 - 20 h after fertilization by artificial hatching in a hatchery (Marine Development Com- pany, Iwate, Japan). The larvae were treated with 0.1% colchicine. 0.075 M KG and fixed in Camoy solution according to the method of Arai et al. (1984). The chromosome preparations used for con- ducting SEM determinations of chromosome arm lengths, band- ing, and NOR analyses were made by the chopping method 605 606 Okumura et al. TABLE 1. Chromosome measurements derived from 12 metaphase plates in Haliotis discus hannai. Chromosome Relative Length Arm Ratio Pair No. Mean SDa Mean SDa Classification1' O 1 6.88 0.20 1.21 0.08 m +J 2 6.69 0.33 2.09 0.19 sm 3 6.49 0.32 1.36 0.1 1 m E 4 6.48 0.25 1.59 0.14 m 5 6.10 0.32 2.59 0.22 sm < 6 5.78 0.32 1.28 0.15 m 7 5.75 0.35 2.00 0.12 sm 8 5.74 0.35 2.24 0.17 sm 9 5.71 0.41 1.63 0.08 m 10 5.47 0.24 1.84 0.10 sm 11 5.32 0.35 1.45 0.14 m 12 5.20 0.15 1.15 0.07 m 13 4.95 0.41 2.13 0.13 sm 14 4.74 0.37 1.31 0.14 m 15 4.73 0.24 1.48 0.07 m 16 4.70 0.27 1.91 0.13 sm 17 4.66 0.24 1.61 0.09 m 18 4.64 0.22 1.10 0.05 ni a Standard deviation. b m. metacentric; sm. submetacentric. (Yamazaki et al. 1981 ) using 30 to 50 fixed larvae per slide. Slides were either dried at room temperature for SEM and NOR analyses, or incubated at 50 °C before drying for banding analysis, as de- scribed below. For observations using the SEM. slides were stained with Gi- emsa solution according to the method of Okumura et al. ( 1995). and slides with well-spread metaphases were monitored by light microscopy (LM; Nikon Optiphot). The slides were cut with a glass cutter into 15 x 15 mm2 pieces bearing well-spread metaphases before attachment to a brass disk (specific for the SEM, 26-mm diameter, 7-mm thick) using adhesive carbon tape. The disk-mounted pieces were lightly coated with gold for 3-5 min using an ion coater (Jeol JFC-1 100), and the metaphase plates were photographically analyzed using SEM (Jeol JSM-6400) at 5,000 - 10,000 magnification. The short and long arm lengths of chromosomes in 12 well-spread metaphase plates were displayed in SEM photographs and measured with a digital map meter (Koi- zumi Sokki Mfg. Co.). Relative lengths were calculated from the total chromosome length (Thiriot-Quievreux 1984). and arm ratios were derived (Levan et al. 1964) accordingly. The arm ratios and relative lengths were used for the morphological classification (Le- van et al. 1964) and the numbering of chromosomes, respectively. For banding analysis, cell suspensions from the chopped larvae expanded with Carnoy solution on the slides were incubated at 50 °C until dry. The slides were treated with a salt solution (Ohnuki 1968) at room temperature for 150 min before rinsing with tap water. The slides were then stained with Giemsa solution by the above method. Nine well-banded metaphase plates were photo- graphed using LM or SEM, and chromosomes were measured and paired on the basis of their relative lengths, arm ratios, and banding patterns. Ag-NOR staining was performed according to the method of Howell and Black (1980)with slight modifications. After Giemsa staining for 40 min, the slides were briefly treated with a mixture 3.0 -I 2.5 - 2.0 - 1.5 - 1.0 r-18H »£* 4.0 5.0 6.0 7.0 Relative length (%) I I — *1^ 1 •4 * T " *2 *3 Figure. 1. A two-dimensional reference plot of the relative chromo- some length and arm ratio distributions in H. d. hannai derived from Table 1. *1: Chromosome number and position of mean value. *2: Mean ± standard deviation of arm ratio. *3: Mean ± standard devia- tion of relative length. of silver nitrate, gelatin, and formic acid (Howell and Black 1980) for 5 min at 70 °C. Four-well NOR metaphase plates were pho- tographed before measuring chromosome arm lengths. RESULTS The mean values of the relative lengths and arm ratios of 1 8 chromosome pairs were estimated from arm length measurements in 12 well-spread metaphase plates from the larvae using an SEM (Table 1 ). Chromosome pairs were numbered in the order of their relative lengths. The arm ratios indicated that this species pos- sessed 1 1 pairs of metacentric and seven pairs of submetacentric chromosomes. The distribution of relative chromosome lengths and arm ratios (Fig. 1) was derived from the values in Table 1. This two-dimensional graph served as a reference plot for the identification of karyotypes in subsequent individual H. d. hannai analyses, and can be used for comparisons with members of other Haliotidae species in future research. Chromosome pairs 7 and 8. 7 and 10. and 14 and 15 were difficult to identify using only such morphological characteristics as the relative length and arm ratio, because of their mutually overlapping standard deviations (Fig. 1 ). Banding was accomplished in H. d. hannai by heat and salt solution treatment using nine well-spread metaphase plates. Some intermetaphase variations in the banding pattern, such as shading or numbers of bands, were found. Nevertheless, the banding pat- terns indicated the same characteristics within each homologous chromosome pair in the same metaphase plate. A typical banded karyotype of H. d. hannai is shown in Figure 2. In Fig. 3, a Chromosomal Analysis of Pacific Abalone 607 1 5 9 2 3 4 678 10 11 12 13 14 15 16 17 18 Figure. 2. Banded karyotype of H. d. hannai after treatment with heat and salt solution (scale bar = 5 urn). diagram of the banding patterns of chromosome pairs no. 7, 8, 10, 14, and 15 from Figure 2 is shown. These chromosomes were among those in the combinations mentioned above for which it was difficult to identify homologous chromosomes using only morphological characteristics. Chromosomes 7 and 15 had unique bands deeply stained on the short arm (7) or the long arm (15) (Figs. 2,3). These staining characteristics simplified the differ- entiation between chromosome pairs 7 and 8, 7 and 10, and 14 and 15. Ag-NOR staining was performed on four well-metaphase plates of H. d. hannai. The NORs were located terminally on the long arms of two chromosome pairs in all metaphases. The chro- mosome numbers of each pair bearing the NOR in those four metaphases were identified by referring to the relative chromo- some lengths and arm ratios illustrated in Figure 1 . In the meta- phase shown in Figure 4. the two chromosome pairs bearing the NOR were identified as nos. 5 and 6. In a similar manner, NOR- bearing chromosome pairs in the other three metaphases were identified as nos. 8 and 11,5 and 1 1, and 5 and 6, respectively DISCUSSION Based on the accurate measurement of chromosome arm lengths using an SEM, we determined the mean values of the relative lengths and arm ratios of 18 chromosome pairs in this study. This is the first report of concrete numerical values de- scribed for the Haliotidae species. These values serve as reliable indexes of morphological characteristics in H. d. hannai chromo- somes, and our novel approach may furnish information on the H. d. hannai karyotype useful for future research. Although previous findings using LM (Arai et al. 1982; Okumura et al. 1995) indi- cated the presence of 10 pairs of metacentric and eight pairs of submetacentric chromosomes, the present study using SEM mea- surements indicated that 1 1 pairs of metacentric and seven pairs of submetacentric chromosomes exist in this species. This difference may be attributable to the existence of chromosomes 4. 9. or 17 showing borderline values of the arm ratio in the metacentric/ submetacentric classification (Fig. 1). Comparisons of chromo- some classifications based on observations using LM between H. d. hannai and H. d. discus (Arai et al. 1982), as well as between H. d. discus and H. gigantea (Miyaki et al. 1997). suggest that there is no difference in chromosome number (2 n = 36) or morphology among these species. However, if comparisons using the mean values of SEM-determined chromosome arm lengths and arm ra- tios are performed, some morphological differences might be found, despite their being within the same chromosome classifi- cation. 608 Okumura et al. \J t 8 10 14 15 Figure. 3. Diagram of banding patterns of chromosome numbers 7, 8, 10, 14, and 15 from Figure 2. Arrowheads indicate the deeply stained bands. This is the first study in which chromosome banding analysis (i.e.. not partial banding, such as C-banding) of Haliotidae by heat and salt solution treatment is attempted. This solution was origi- nally used for the observation of chromosome spiral structures in humans (Ohnuki 1968); however, in the present study, when the metaphase plates of H. d. hannai were treated with this solution after heat treatment, clear chromosome bands appeared. These findings may well be applicable to mollusca species that have not been successfully subjected to banding analyses using the general methods of G-banding. The banding patterns may be similar to patterns of G-banding; however, we cannot compare our patterns to G-banding patterns, because although some general methods of G-banding (Sumner et al. 1971. Seabright 1971 ) have been previ- ously attempted on this species (Okumura unpublished), the G- bands were not convincingly clear as compared to our present banding patterns. The characteristic banding patterns serve as re- liable and useful markers for discriminating homologous chromo- somes in the same metaphase plates, which had hitherto been difficult to identify using only morphological analysis. Neverthe- less, the banding patterns showed some intermetaphase variations. Mendez et al. ( 1990) described variations in the quality of banding attributable to differing extents of chromosomal condensation in mollusca. We speculate that the variation observed in the present study was caused by the same factors. NOR analysis was performed here for the first time in Hali- otidae. and variations in NOR-bearing chromosomes were ob- served. Some species of shellfish and finfish have been reported to manifest intraspecific variations in NOR patterns: Ostrea angasi shows variations in three chromosome pairs (Li and Havenhand 1997), and differences in NOR number and NOR-bearing chro- mosome types have been observed in Erythrinidae fish (Antonio and Bertollo 1996). In those studies, the NOR variations have been used in comparative studies of intra- or interspecific genetic fea- tures. Variations in NOR chromosomes may confuse the identifi- f m t 4 ' V 5 6 Figure. 4. Ag-NOR-stained metaphase plate from H. d. hannai. Ar- rows indicate the NOR-bearing chromosomes. Arrowheads indicate NOR spots on the two chromosome pairs cut out from the upper photograph of the metaphased plate (5 and 6 represent the chromo- some pair numbers. Scale bar = 5 um). cation of chromosome karyotypes when NOR markers are used. However, the variation may serve as a useful index for monitoring genetics and taxonomy in Haliotidae. In the present study, we developed a method of reliable iden- tification of karyotype, chromosome banding, and NOR analyses in H. discus hannai. In fact, we showed the morphological values and banding patterns in this species. Because the present study was successfully performed for this species, similar analyses in other species of Haliotidae are highly likely to succeed. In the future, comparative studies of Haliotidae chromosomes based on the re- sults for H. d. hannai in this study are warranted. Recently, studies on fluorescent in situ hybridization of mollusca chromosomes have been undertaken (Pascoe et al. 1996. Guo and Allen 1997, Mar- tinez-Exposito et al. 1997, Insua et al. 1998). This technique is useful for chromosome and gene locus identification, and it may well be applicable for the Haliotidae species. ACKNOWLEDGMENTS The authors are grateful to Mr. Katsuhiro Furukawa and Mr. Suehiro Furukawa of the Marine Development Company for the kind gift of research materials. The authors are indebted to Mr. Hideo Hajima, Mr. Hisashi Hatano, and Mr. Mamoru Kawashima for their expert technical assistance. Chromosomal Analysis of Pacific Abalone 609 UTKRATURE CITED Antonio, L. & C. Bertollo. 1996. The nucleolar organizer regions of Eryth- rinidae fish, an uncommon situation in the genus Haplias. Cytologia 61:75-81. Arai. K., H. Tsuhaki, Y. Ishitani & K. Fujino. 1982. Chromosomes of Haliotis discus hamuli Ino and H. discus Reeve. Bull. Japan. Soc. Sci. Fish. 48:1689-1691. Arai, K.. F. Naito. H. Sasaki & K. Fujino. 1984. Gynogenesis with ultra- violet ray irradiated sperm in the Pacific abalone. Bull. Japan. Soc. Sci. Fish. 50:2019-2023. Arai, K. & N. P. Wilkins. 1986. Chromosomes of Haliotis tuberculata L. Aquaculture 58:305-308. Arai, K., K. Fujino & M. Kudo. 1988. Karyotype and zymogram differ- ences among three species of the abalones Haliotis planata, H. varia, and H. diversicolor diversicolor. Nippon Suisan Gakkaishi. 54:2055- 2064. Colombera, D. & F. Tagliaferri. 1983. Chromosomes from male gonads of Haliotis tuberculata and Haliotis lamellosa (Haliotidae, Archaeogas- tropoda. Mollusca). Caryologia 36:231-234. Guo. X. & S. K. Allen, Jr. 1997. Fluorescence in situ hybridization of vertebrate telomere sequence to chromosome ends of the Pacific oyster, Crassoslrea gigas Thunberg. J. Shellfish Res. 16:87-89. Howell. W. M. & D. A. Black. 1980. Controlled silver-staining of nucleo- lus organizer regions with a protective colloidal developer: a 1-step method. Experientia 36:1014-1015. Insua, A. & C. Thiriot-Quievreux. 1991. The characterization of Ostrea denselamellosa (Mollusca, Bivalvia) chromosomes: karyotype, consti- tutive heterochromatin, and nucleolus organizer regions. Aquaculture, Amsterdam 97:317-325. Insua, A., M. J. Lopez-Pinon & J. Mendez. 1998. Characterization of Aequipecten opercularis (Bivalvia: Pectinidae) chromosomes by dif- ferent staining techniques and fluorescent in situ hybridization. Genes Genet. Syst. 73:193-200. Jarayabhand. P., R. Yom-la & A. Popongviwat. 1998. Karyotypes of ma- rine molluscs in the family Haliotidae found in Thailand. J. Shellfish Res. 17:761-764. Levan, A., K. Fredga & A. A. Sandberg. 1964. Nomenclature for centro- meric position on chromosomes. Hereditas 52:201-220. Li, X. X. & J. N. Havenhand 1997. Karyotype, nucleolus organizer regions, and constitutive heterochromatin in Ostrea angasi (Molluscae: Bi- valvia): evidence of taxonomic relationships within the Ostreidae. Mar. Biol. 127:443^448. Martmez-Exposito, M. J., J. Mendez & J. J. Pasantes 1997. Analysis of NORs and NOR-associated heterochromatin in the mussel Mytilus gal- loprovincialis Lmk. Chrom. Res. 5:268-273. Mendez. J.. J. J. Pasantes & M. J. Martinez-Exposito. 1990. Banding pattern of mussel {Mytilus galloprovincialis) chromosomes induced by 2 x SSC/giemsa-stain treatment. Mar. Biol. 106:375-377. Minkler. J. 1977. Chromosomes of the black abalone {Haliotis crachero- dii). Experimental 33:1143. Miyaki. K., O. Tabeta & H. Kayano. 1997. Karyotypes of the two species of abalones Nordotis discus and N. gigantea. Fisheries Sci. 63:179- 180. Miyaki, K., M. Matsuda & O. Tabeta. 1999. Karyotype of the giant aba- lone, Nordotis madaka. Fisheries Sci. 65:317-318. Nakamura, H. K. 1985. The chromosomes of Haliotis diversicolor aqua- tills (Archaeogastropoda: Haliotidae). Mai. Rev. 18:113-114. Nakamura, H. K. 1986. Chromosome of Archaeogastropoda (Mollusca: Ptosobranchia). with some remarks on their cytotaxonomy and phylog- eny. Publ. Seto. Mar. Biol. Lab. 31:191-267. Nishikawa, S. 1962. A comparative study of the chromosomes in marine gastropods, with some remarks on cytotaxonomy and phylogeny. J. Shimonoseki Coll. Fish. 11:149-186. Ohnuki. Y. 1968. Structure of chromosomes. I. morphological studies of the spiral structure of human somatic chromosomes. Chromosome 25: 402-428. Okumura. S.. S. Yamada. T. Sugie. D. Sekimiya, A. Toda. H. Hajima. H. Hatano & K. Yamamori. 1995. C-banding study of chromosomes in Pacific abalone. Haliotis discus hamuli (Archaeogastropoda: Hali- otidae). Chrom. Inf. Sen: 59:7-9. Pascoe, P. L. & D. R. Dixon. 1994. Structural chromosomal polymorphism in the dog-whelk Nucella lapillus (Mollusca: Neogastropoda). Mar. Biol. 118:247-253. Pascoe, P. L., S. J. Patton, R. Critcher & D. R. Dixon. 1996. Robertsonian polymorphism in the marine gastropod, Nucella lapillus: advances in karyology using rDNA loci and NORs. Chromosoma 104:455—460. Seabright. M. 1971. A rapid handing technique for human chromosomes. Lancet 2:971. Sumner. A. T.. H. J. Evans & R. A. Buckland. 1971. A new technique for distinguishing between human chromosomes. Nature New Biol. 232: 31-32. Thiriot-Quievreux, C. 1984. Chromosome analysis of three species of Mytilus (Bivalvia: Mytilidae). Mar Biol Lett. 5:265-273. Thiriot-Quievreux, C. & M. Wolowicz. 1996. Karyotypes of Cerasloderma glaucum (Bivalvia) from Baltic and Mediterranean populations. Hy- drobiologia 324:149-155. Yamazaki, F., H. Onozato & K. Arai. 1981. The chopping method for obtaining permanent chromosome preparation from embryos of teleost fishes. Bull. Japan. Soc. Sci. Fish. 47:963. Journal of Shellfish Research. Vol. 18, No. 2. 61 1-619. 1999. EFFECT OF pH ON GROWTH RATE, OXYGEN CONSUMPTION RATE, AND HISTOPATHOLOGY OF GILL AND KIDNEY TISSUE FOR JUVENILE GREENLIP ABALONE, HALIOTIS LAEVIGATA DONOVAN AND BLACKLIP ABALONE, HALIOTIS RUBRA LEACH JAMES O. HARRIS,1 GREG B. MAGUIRE,1 2 STEPHEN J. EDWARDS,1 AND STEPHEN M. HINDRUM1 School of Aquaculture University of Tasmania P.O. Box 1214 Launceston Tasmania. Australia, 7250 'Fisheries Research Division Western Australia P.O. Box 20 North Beach, Western Australia Australia, 6020 School of Applied Science University of Tasmania P.O. Box 1214 Launceston Tasmania, Australia, 7250 ABSTRACT Juvenile greenlip abalone, Haliotis laevigata, (mean whole mass 2.30 ± 0.73 g, mean ± SD. n = 561) and juvenile hlacklip abalone, Haliotis rubra, (mean whole mass 1.56 ± 0.64 g, mean ± SD. n = 559) were grown for 50-68 days in bioassay tanks at a range of pH levels adjusted using hydrochloric acid (HCI) or sodium hydroxide (NaOH). For greenlip abalone, specific growth rate (SGR) was significantly affected by pH. whether measured on a length or whole mass basis (P < .001 ). For blacklip abalone. SGR was significantly affected by pH whether SGR was measured on a length or whole mass basis. For growth, expressed on a whole mass basis, the EC5 values (5% growth reductions) were at pH 7.78 and 8.77 for greenlip abalone, and at 7.93 and 8.46 for blacklip abalone. The EC5„ values (50% growth reductions) were at pH 7.39 for greenlip abalone. and 7.37 and 9.02 for blacklip abalone. Survival of both species was significantly reduced at pH 6.79, and survival of blacklip abalone was also significantly reduced at pH 7.76. At the end of the bioassay, groups of abalone were transferred to respiratory chambers. A significant reduction in respiratory activity was observed at both high and low pH values for greenlip abalone (P < .001). Greenlip abalone exposed to pH 7.16 showed alterations in kidney definition, tubule and lumen size, and an increase in gill hyperplasia and abnormalities. Blacklip abalone exposed to pH 7.16 demonstrated alterations to kidney and gill definition, and lumen size was increased. KEY WORDS: abalone. Haliotis laevigata. Haliotis rubra, growth, mollusks, pH, histology INTRODUCTION The current emphasis toward water reuse for land-based aba- lone culture systems is likely to have an impact on pH levels. The With the demand for premium abalone products rising steadily process of nitrification, central to bioftlter operations in recircu- (Oakes and Ponte 1996), abalone culture is expanding in both land- lating water systems, causes pH levels to decline (Wickins 1983). and sea-based culture systems (Fleming and Hone 1996). With In abalone culture systems, diatom surfaces are often used for land-based culture, some recirculation of water is often employed juvenile rearing (Fleming and Hone 1996). Both respiration by the to reduce costs, and a variety of conditions can be experienced abalone and nitrification in a biofilter will depress the water pH; with sea-based culture, depending upon site (Hindrum et al. 1996). however, photosynthetic activity by diatoms will cause pH to in- In both cases, abalone may be exposed to levels of ammonia. crease. In a previous experiment with abalone. pH values ranged nitrite. pH. and dissolved oxygen that may. at least, vary from their from 7.96-8.06. pH levels outside a range of 5-9 are lethal to optima. Biochemical, physiological, and/or morphological many aquatic animals (Randall 1991 ); whereas. pH variation can changes can occur as a response to water quality levels that are in also have such secondary effects as altering ammonia toxicity excess of those tolerated in aquatic animals (Meyers and Hen- (Thurston and Russo 1981). However, limited information is cur- dricks 1985). Often the gills are among the organs most affected by rently available regarding the effects of pH on mollusks and is waterborne pollutants (Mallat 1985). because the respiratory sur- mostly concerned with bivalves (Calabrese and Davis 1966. Bam- face provides an extensive interface with the aquatic environment. ber 1987. Bamber 1990). In many fish, the kidney often forms a site of histological changes Oxygen uptake has been widely used to help indicate the health in response to toxicants (Russo 1985). In a previous study, both of animals and their over-all energy expenditure or activity levels gill and kidney tissue of greenlip abalone provided some indicators (Innes and Houlihan 1985) and is also a critical factor in assess- of environmental stress (Harris et al. 1998a). ments of stress in aquatic organisms (Beitinger and McAuley 611 612 Harris et al. 1990, Willows 1994), including abalone (Harris et al. 1997, Harris et al. 1999). Wells et al. (1998) demonstrated the New Zealand abalone, Haliotis iris and Haliotis australis, to have reversed Bohr and Root effects at low pH. leading to an increase in the binding affinity of the respiratory pigment, haemocyanin, and oxygen, and subsequent decrease in released oxygen. Previous bioassays on Australian greenlip abalone have deter- mined the effects of some aspects of water quality to abalone, such as the chronic toxicity of nitrite (Harris et al. 1997). ammonia (Harris et al. 1998b) and dissolved oxygen (Harris et al. 1999). The aim of this research is to determine the effects of chronic exposure to a variety of pH levels on growth, survival, food consumption, oxygen consumption, and histopathology of gill and kidney tissue of the greenlip abalone, Haliotis laevigata, and the blacklip aba- lone, Haliotis rubra, because both species are of increasing im- portance to abalone aquaculture in Australia. MATERIALS AND METHODS The juvenile greenlip abalone used in these experiments were approximately 2 years old and were obtained from a commercial hatchery at Bicheno, Tasmania, Australia, where the research was conducted (El 48' 1 8", S41'53"). The juvenile blacklip abalone were approximately 12 months old and were obtained from a com- mercial farm at Swansea. Tasmania. Australia. The initial mean length and mass of the greenlip and blacklip abalone were 26.49 ± 2.83 mm and 2.30 ± 0.73 g and 22.92 ± 2.92 mm and 1.56 ± 0.64 g, respectively (mean ± SD; n = 561 and 559). For 2-3 months before experimentation, the greenlip abalone were maintained on a mixture of formulated abalone feed and benthic diatoms, and the blacklip abalone had been maintained on a formulated abalone feed (Adam & Amos). Blacklip abalone were initially removed using a spatula before transport to the experimental site, and ac- climatized for 3 days in flowing, aerated seawater before further handling. All abalone were anesthetized (0.1% benzocaine) until they could be easily removed from the tank surfaces. Subse- quently, they were weighed to the nearest 0.01 g, measured with callipers to 0.1 mm, tagged (Hallprint, Adelaide, Australia), and randomly distributed into 18 bioassay units to give 30 of each species within each tank. Mortalities from pH 7.76 resulted in the stocking of this treatment 15 days after the trial commenced, using more from the initial group of blacklip abalone. Bioassay System Seawater from an exposed coastline, free from freshwater run- off, was filtered through a commercial sand filter and delivered to six 1.100 L reservoirs. pH was adjusted using AR grade NaOH or HC1. thoroughly dissolved in each reservoir. Each reservoir was connected to a constant head chamber ( 150-mm diameter, vertical PVC pipe, operating volume 30 L) that supplied constant flow to three bioassay chambers via standard lengths of black 4-mm poly- propylene tubing that entered the bioassay tanks. These tanks were cylindrical with a conical base to concentrate solid wastes. In each 70-L bioassay tank, there were two cages ( 100-mm x 35-cm PVC tube with 6-mm mesh floor and 8-mm mesh wall sections) sus- pended vertically, containing greenlip and blacklip abalone. Daily flow rates averaged 193 ± 1.4 mL/min"1 (n = 108; 18 tanks on six occasions) giving an effective replacement rate of 90% of bioassay tank volume in 10-12 h. This was within the recommended flow rates for aquatic toxicological studies by Sprague (1969) of 90% replacement in 8-12 h. Identical 5 W submersible pumps were placed in each tank to stimulate similar current flow (8.7 l.min"' output at zero head). The experiment was conducted using 200 and 300 W aquarium heaters in the bioassay tanks and constant head chambers, respectively, to maintain relatively uniform daily tem- perature at 19.0 ± 1.0 °C (mean ± SD) (n = 71 days) (range 16.5- 21.7 °C) (Table 1). Water Quality Analysis The pH. temperature, salinity, and DO in all tanks were mea- sured on all days (Table 2). A pH meter and combination glass electrode (TPS) were calibrated with phosphate (pH = 7.00) and borate (pH = 9.28) buffers daily before use (Bruno and Svoronos 1989). A TPS oxygen electrode, used for daily measurements, was calibrated before use in "air-saturated" seawater. The efficiency of this calibration was validated occasionally using Winkler's titra- tion. Water samples were collected in acid-washed glassware, and ammonia was measured using the indophenol blue spectrophoto- metric method (Solorzano 1969, as modified by Dal Pont et al. 1974). The concentration of ammonia was measured as total am- monia-nitrogen (TAN), and free ammonia-nitrogen (FAN) was calculated from appropriate temperature, pH. and salinity tables (Bower and Bidwell 1978) (Table 1 ). Nitrite was measured occa- sionally, using the diazotisation method (Grasshoff 1989). TABLE 1. Food consumption and survival of greenlip abalone, Haliotis laevigata and blacklip abalone. PH Food Consumption (g/g-'/day-1) % Survival mean ± SE Min Max Greenlip Blacklip Greenlip Blacklip 9.01 ±0.01 8.34 9.40 0.087 ± 0.01 Cb 0.139±0.024-' 81.9 ± 10.3a 55.0 ± 15.1ab 8.27 ± 0.00 7.86 8.77 0.070 ± 0.003°b 0.088 ± 0.007ab 95.8±2.1a 78.8±8.7J 7.76 ±0.01 6.71 8.27 0.102 ±0.010" 0.428 ± 0.294f 94.4 ± 1.1° 20.0 ± 18.4b 7.46 ± 0.02 6.97 8.17 0.061 ±0.()llab 0.074 ± 0.0 15ab 98.9 ± 1.1° 55.2+ 13.4°b 7.16 + 0.01 6.45 7.93 0.036 ± 0.006b 0.034 ± 0.005h 70.4 ± 14.8" 75.8 ± 8.6° 6.79 ± 0.03 6.04 7.62 0.052 ± 0.007'* 0.078 ±0.0152 3.1 ±3.1" 0b Haliotis rubra exposed to a range of pH conditions (mean ± SE) (means sharing a common superscript are not significantly different [P > .05]). * Water quality: ammonia concentrations ranged from 0-0.026 mg FAN/1"1, nitrite concentrations ranged from 0.003-0.005 mg NCK-N/T'. temperatures ranged from 18.6- 19.3 °C, flow rates ranged from 184.2-202.9 ml. min"'. salinity ranged from 33.8-34.6 ppt. and oxygen levels ranged from 6.96-7.19 mg DO.l"'. f Data for pH 7.76 and pH 6.79 were not included in statistical analyses because of lack of replicates. pH and Growth for Two Australian Abalone Species 613 TABLE 2. Water quality parameters and biomass for respiratory experiments on greenlip abalone. Haiiotis laevigata, and hlacklip abalone, Haliotis rubra. Greenlip Abalone Blacklip Abalone PH Biomass (g) Temperature Salinity PH Biomass (g) Temperature Salinity - - CO (ppt) - - (°C) (ppt) 9.25 ± 0.07 76.31 +7.16 18.4 ±0.2 34.4 ± 0.3 9.27 ±0.00 29.08 ± 17.05 1 8.5 ± 0.5 8.45 ±0.15 105.57 ± 12.56 18.5 ±0.2 34.7 ± 0. 1 8.56 ±0.01 50.07 ± 1.80 18.8 + 0.3 7.95 ±0.1 2 68.65 ±7.71 16.7 ±0.2 33.7 ±0.1 8.02 ±0.19 26.05a 19.5J 34.2a 7.30 ± 0.03 68.22 ±2.66 16.9 ±0.2 34. 1 ± 0. 1 7.11 ±0.31 44.48 ± 14.98 20.4 ±1.1 34.8a 6.72 ± 0.06 51.12 ± 0.75 19.0 ±0.3 34.2 ± 0.3 6.88 ± 0.02 37.76 ±5.47 18.9 ±0.2 35.0 ±0.1 6.08 ± 0.05 5.77J 18.8 ±0.3 34.4 ± 0.0 All measurements expressed as mean ± SE. 'n = 1. Experiment I: Chronic pH exposure Six experimental treatments were established (Table 1 ); aver- age pH ranged from 9.01-6.79. The abalone were acclimatized to the bioassay system for 4-6 days before pH adjustment com- menced. pH adjustment occurred over several days, with a gradual increase in chemical levels (HC1 or NaOH) each day, until the desired level was attained. All cages were checked daily for mor- tality. All tanks were fed a proprietary, formulated abalone diet (AB- CHOW) every 2 to 3 days. The feeding ration was adjusted in response to food consumption data as the trial progressed. Food consumption was estimated on four occasions from uneaten food removed from the base of the cages after 2 days and drying it for 24—48 hours at 55-60 °C. Residual food mass was not corrected for soluble and particulate nutrient losses over the 2 days. Appar- ent food consumption (amount of food supplied minus residual food as g dry mass) was divided by the initial tank biomass, less the mass of any mortalities to that point, and expressed as g dry mass food remaining per g whole wet body mass per day. A valve in the base of each bioassay tank was opened daily to remove organic wastes. Tanks were also cleaned more thoroughly, on average, every 9 days. Cleaning involved lowering the water level, siphoning enough water from the bioassay tank into a 20-L bucket to cover the cages, removing cages to the bucket, draining the tank, scrubbing the tanks and cages, refilling the tanks directly from the reservoirs, and returning the cages to the tanks. This took under 10 minutes for any tank. Abalone remained in the bioassay system for up to 68 days and were removed in staggered groups for respirometry over 14 days. This is unlikely to be sufficient time for significant differences in growth because of stocking density to arise. All abalone were weighed and measured for the final growth data. Specific growth rate data were calculated for mass and length of each abalone as SGR = [ln(final) - In(initial)] 100 days"1. Experiment 2: Oxygen Consumption Rates at End of the Chronic Bioassay The respirometer system included five elliptical perspex cham- bers (of 2.3 I) normally set up with two replicate chambers for each treatment and one chamber as a control (no animals), as described in Harris et al. (1997). Commencing on day 57, abalone from the bioassay system were transferred to respirometer chambers for a series of 3-day experiments. All abalone remaining in two of the three replicate bioassay tanks for each treatment level were transferred to the respirometer system so that data could be obtained for duplicate tanks of each species at each nominal treatment level. These ani- mals had been fed before removal. Abalone that did not attach to transferable plastic strips in the cages within the bioassay units were removed manually, either by sliding them directly from the substrate or by inserting a thin, plastic card underneath each aba- lone's foot. Temperature and pH levels were measured within the constant head chambers (Table 3). TABLE 3. Scoring schedule for histological sections of abalone. 1 Gill Definition Well-defined brush border Hypertrophy No cells Hyperplasia No evidence Abnormalities No evidence Right kidney Definition Well-defined cells Tubule size Small, plenty of tubule contents Lumen size Small, plenty of lumen contents Cytoplasm Small % of each cell vacuoles Small % of each cell Some filaments showing irregular brush borders Isolated cells Isolated cells Isolated incidences Some difficulty in defining tubules Some enlargement of tubules Some enlargement of lumen >half cell volume >half cell volume Poorly defined filaments widespread widespread several per gill Tubules very difficult to define large tubules large lumen space < half cell volume < half cell volume 614 Harris et al. Histological Sample Preparation Five abalone were sampled from two of the triplicate bioassay tanks for each treatment. These abalone were dissected to remove the posterior portion of the viscera containing the gills and kidney. This tissue was fixed in phosphate-buffered formalin at room tem- perature (15-18 °C) then dehydrated through a graded ethanol se- ries to xylene in a Tissue-Tek II tissue processor. Dehydrated tissue samples were embedded in paraffin resin on a Shandon Histocentre 2 and sectioned on a Microm HM 340 microtome at 4 (o.m. Routine Harris' Hematoxylin and Eosin (H & E) staining were carried out on all tissues processed using a Shandon Linistain GLX automatic tissue stainer. All sections were mounted in DPX and examined under a light microscope. Insufficient animals remained in pH 6.79 for histological analy- sis, so abalone from pH 7.16 were the most extreme treatment considered. The tissue sample from each abalone was examined and scored regarding several aspects of gill and kidney structure (Table 3). Statistical Analysis Data were subjected to one factor analysis of variance (ANOVA) after meeting assumptions of normality using the Sha- piro_Wilk test (Zar 1996) and homogeneity of variance using Co- chran's test (Underwood 1981). Replicates were considered to be independent, and pH concentration was analyzed as a fixed factor. Survival data (as percentage) and whole wet body mass (WWBM): shell length (SL) ratio were transformed (arcsin V%x0.01 and log, respectively) to satisfy assumptions of normality and homogeneity of variance before analysis. Results for each pH level were com- pared using Tukey's HSD (Sokal and Rohlf 1995). Preliminary analysis indicated that initial abalone size did not affect growth rate. All analyses included assessment of FAN. nitrite-N, DO and temperature as covariates (Sokal and Rohlf) and were conducted using JMP 3.0 software (SAS Institute. Cary. North Carolina). The effect of each toxicant on gill and kidney structure was examined using Chi-squared (x2) analysis to compare two proportions (one- tailed) (Sokal and Rohlf). RESULTS Experiment 1: Chronic pH Exposure For greenlip abalone. SGR was significantly affected by pH (P < .001 ). whether measured on a SL or WWBM basis. SL growth (Fig. 1) and WWBM gain (Fig. 2) were highest at pH 8.27-7.76. however, significant growth rate reductions occurred at pH 9.01 and 7.46-6.79 (P < 0.05). Second order regression of the SGR length data indicated that there was no shell growth below pH 6.90 (Fig. I ). The EC values (effective concentration where reductions of x% occur) from the modeled mass data were pH 8.77 and 7.78 (EC5), and pH 7.39 (ECSI)) (Fig. 2). Highest WWBM and SL growth rates were observed for greenlip abalone at pH 8.27. For blacklip abalone, SGR was significantly affected by pH whether measured on a length (P < .001) or whole mass (P < .01 ) basis. SL growth rates were highest at pH 8.27-7.76. with further depression of SL growth rates at pH 9.01 and at pH 7.46-6.79 (P < .05). Second-order regression of the SGR length data indicated that there was no shell growth below pH 6.99 (Fig. 3). For WWBM gain, significant growth rate reductions occurred at pH 7.16 (Fig. 4) (P < .05). The EC values from the mass data were pH 8.46 and 7.93 (EC5), and pH 9.02 and 7.37 (EC5„). Highest WWBM and SL growth rates were observed for blacklip abalone at pH 7.76 al- 0 4 0.3 H, 02 0 1 0.0 -0.1 y = 2.23x-0.14x-8.97 r2 = 086 f d -, 1 1 1 1 1 1 1 1 1— 9 00 8 75 8 50 8.25 8.00 7 75 7 50 7 25 7 00 6 75 6 50 pH Figure 1. Specific growth rate (length) of juvenile greenlip abalone, Haliotis laevigata, subjected to chronic pH conditions in Experiment I (mean ± SE, n=3(. Regressions based on data for each replicate rather than treatment means. though WWBM growth rate data at this pH were highly variable (Fig. 4). Survival of greenlip abalone was significantly affected by pH exposure (P < .0001 ). Survival was high in all but pH 6.79, where significant mortalities occurred (P < .05) (Table 1 ). Survival of blacklip abalone was also significantly affected by pH exposure (P < .01). Survival was high in all but pH 7.76 and 6.79, where 1 2 10 - 0.8 - 0.6 S3 E 04 0.2 0.0 - -0.2 -0.4 -0.6 VEC» y = 10.40x-0 63x -42 16 r2 = 0.92 a Ec,y V ec 9.00 8.75 8.50 8.25 8.00 7 75 7 50 7 25 7 00 6 75 pH Figure 2. Specific growth rate (mass) of juvenile greenlip abalone. Haliotis laevigata, subjected to chronic pH conditions in Experiment 1 (mean ± SE, n = 3). Regressions based on data for each replicate rather than treatment means (1; n=l). pH and Growth for Two Australian Abalone Species 615 020 0 15 0.10 0 05 0.00 & -0 05 -0 10 -, , , , 1 , 1 , 1- 9.00 8 75 8.50 8.25 8.00 7 75 7.50 7 25 7 00 6 75 6.50 PH Figure 3. Specific growth rate (length) of juvenile blacklip abalone, Haliotis rubra, subjected to chronic pH conditions in Kxperiment 1 (mean ± SE, n = 3). Regressions based on data for each replicate rather than treatment means. significant mortalities occurred (P < .05) (Table 1). During the experiment, if the pH fell below 6.2, significant mortalities for both species followed for up to 7 days. Abalone exposed to these conditions lost attachment and collected toward the bottom of the cages. At the end of the experiment, all blacklip abalone at pH 6.79. all greenlip abalone in two replicates at pH 6.79. and all blacklip abalone of one replicate at pH 7.76 had died, so only length measurements until their day of removal were calculated, not mass data. pH had a significant effect on WWBM:SL for both greenlip abalone (Fig. 5) (P < .001) and blacklip abalone (P < 0.05) (Fig. 6). Greenlip abalone from pH 6.79 had significantly lower ratios than abalone from pH 8.27 (P < .05). For blacklip abalone. pH 7.16 produced a significantly (P < .05) lower ratio than pH 7.76. There was a significant effect of pH on food consumption by greenlip abalone (P < .01 ), where abalone from pH 7.76 had sig- nificantly higher food consumption than abalone from pH 7.16 (P > .05) (Table 2). The variability of the data for abalone in treat- ments that also had significant mortality rates prevented the con- ditions of homogeniety of variance being satisfied for statistical analysis, so these treatments (pH 7.76 and pH 6.79) were omitted from analysis. A significant effect of pH on food consumption by blacklip abalone was observed (P < .05), with abalone held at pH 7.16 demonstrating significantly lower rates than abalone from pH 9.01 (Table 1). Significantly lower temperatures were recorded at pH 9.01 and 7.46 than other treatments (P < .001). DO levels were also sig- nificantly different between treatments (P < .001), although aver- age treatment oxygen saturation was 96.7 ± 0.2%. Salinity was also significantly affected by pH (P < .001). The salinity at pH 9.01 was significantly lower than the controls (Table 1 ) (P < .05). Statistical analysis of log-transformed ammonia levels determined all treatments to be significantly different to the control (P < .05). Nitrite levels were at or below 0.005 mg N02-N.l. Experiment 2: Oxygen Consumption Rates at End of Chronic Bioassay Oxygen consumption rate of juvenile greenlip abalone was sig- nificantly affected by pH (P < .001 ). with greenlip abalone of pH 9.25 and 6.72-6.08 recording significantly lower (P < .05) oxygen consumption rates than the controls (pH 8.45) (Fig. 7). Mortality 1.00 PH Figure 4. Specific growth rate (mass) of juvenile blacklip abalone, Haliotis rubra, subjected to chronic pH conditions in Experiment 1 (mean ± SE, n=3). Regressions based on data for each replicate rather than treatment means. i ' ' ' 1 ' ' ' r 9 00 8.75 8.50 8.25 8 00 7 75 7 50 7 25 7 00 6 75 PH Figure 5. Whole wet body mass: shell length of juvenile greenlip aba- lone, Haliotis laevigata, subjected to chronic pH conditions in Experi- ment 1 (mean + SE, n = 3). Regressions based on data for each repli- cate rather than treatment means. 616 Harris et al. o.ioo 0 095 0 070 0.065 — I ' ' — 9.00 8.75 8.50 8.25 8.00 PH 7.75 750 725 7.00 Figure 6. Whole wet body mass: shell length of juvenile hlacklip aba- lone. Haliotis rubra, subjected to chronic pH conditions in Experiment 1 (mean ± SE, n = 3). Regressions based on data for each replicate rather than treatment means. among blacklip abalone before and during respirometry prevented statistical analysis of this species. Histological Sample Examination Chi-squared analysis of histological observations from greenlip abalone exposed to slightly acidified seawater (pH 7.16) demon- strated significantly different kidney definition (\2 calc. = 8.57; v 100 - k_ ■' 80 - / * \ 60 - 40 - 4 1 be y=415.76x-26.32x2-155163 r!=0.80 *\ 20 - 0 - -20 - 1 1 1 ' 1 c 1 9 0 85 80 7.5 70 65 60 PH Figure 7. Oxygen consumption rate of juvenile greenlip abalone, Hali- otis laevigata, subjected to a range of pH conditions (mean ± SE, n = 2). Regressions based on data for each replicate rather than treatment means. = 1 . P < .005) and tubule enlargement (x2 calc. = 8.57; P < .005. v = 1 ) (Fig. 8). Kidney lumen size was also significantly larger in greenlip abalone from pH 6.79 (x2 calc. = 15; v = 1. P < .001). Analysis of gill structure in greenlip abalone revealed significant increases in hyperplasia (x2 calc. = 4.29; v = 1, P < .05), and abnormalities (x2 calc. = 6.2; v = 1. P < .005), with exposure to pH approaching acidity (pH 7.16) (Fig. 9) Blacklip abalone ex- posed to pH 7.16 demonstrated similar significant differences in kidney definition (x2 calc. = 4.8; v = I, P < .05), kidney lumen size (x2 calc. = 4.8; v = 1, P < .05) (Fig. 10). and gill definition (X2 calc. = 8.24; v = 1. P < .005) (Fig. 11). High pH did not induce any detectable alterations to the structure of the gill or kidney tissue of the abalone examined. DISCUSSION The fastest growth rates of greenlip abalone (SGR mass = 0.87 ± 0. 1 1 %/day"1; SGR length = 0.29 ± 0.03 %/day-' ) and blacklip abalone (SGR mass = 0.97 ± 0.22 %/day-1; SGR length = 0.18 ± 0.01 %/day-') in this experiment were much higher in compari- son to a previous bioassay conducted at a similar temperature in the same experimental system with greenlip abalone only (Harris et al. 1997) (SGR mass = 0.48 ± 0.04 %/day" '; SGR length = 0.12 ± 0.01 %/day-'). One difference in the system design in- volved the incorporation of small submersible pumps into the tanks to improve water flow, as increased water movement stimu- lates feeding for Australian abalone (Shepherd 1973. Higham et al. 1998). The abalone in this study seem to be less tolerant to alterations in pH than other species. From the EC5 estimations, greenlip aba- lone have a wider range of pH over which whole body growth is not inhibited, although outside this pH range, inhibition seems more severe than for blacklip abalone. In comparison, bivalves exhibited slightly higher levels of tolerance to pH than the abalone in this study. Ostrea edulis and Crassostrea gigas grown for 30-60 days lost shell at pH 6.0 and 7.0. respectively (Bamber 1990) and young Venerupis decussate grown for up to 30 days also lost shell at pH 7.0 (Bamber 1987). The flatfish Paralickthys orbignyanus demonstrated no adverse effects at pH 6.0 (Wasielesky et al. 1997), and the pH level where a 5% growth reduction occurred for the marine shrimp Penaeus monodon was pH 5.9 (Allan and Maguire 1992), both substantially lower than for abalone. The greenlip abalone also demonstrated a different pattern of 4 £ 7.76. The decline in WWBM:SL of blacklip abalone outside the ECS range indicates a decline in body growth, as opposed to shell growth, outside this range. The data for WWBM:SL suggest that pH can affect whole animal growth (mass) and shell growth (length) differently. In a previous bioassay, we argued that ammonia affected shell growth more than whole body growth (mass) at low ammonia concentra- tions but that this pattern was reversed at high concentrations (Harris et al. 1998b). The low ratio at more extreme pH may reflect a limitation on whole body growth imposed by depressed shell growth rates in gastropods (Palmer 1981, Preston et al. 1996), in addition to the effect of the toxicant on body growth. In another bioassay on chronic nitrite toxicity, a more complex pattern of WWBM:SL growth appeared (Harris et al. 1997). Further work is required to examine the relationship of body mass growth to shell growth. ^ * .«» - m i • «** * r 1 - . 4dL, v* f4 4 '" «k 4 %:,: A at a* *. •' •V • ■» '• »> e s V i f A f *>- m Pit i « • * 1 V ■ ■ ■J ■ 1 • Figure 10. Right kidney of Haliotis rubra exposed to pH 7.16. Magni- fication 400X. A: enlarged lumen of kidney tubule. Figure 11. (Jill lamellae of Haliotis rubra exposed to pH 7.16. Magni- fication 400X. A; poorly defined brush border. The low pH level at which reductions in survival occurred for greenlip abalone and blacklip abalone is comparable to bivalves, because significant mortalities were observed for abalone at pH 6.79. Small O. edulis, C. gigas, and Mytilus edulis demonstrated reduced survival at pH 6.6, 6.0, and 6.6. respectively (Bamber 1990) and for V. decussata at pH 6.1-6.4. However, other species have much higher tolerance to acid stress, including P. monodon (96 h LC 50 = pH 3.7) (Allan and Maguire 1992) and Paralich- thys orbignyanus. which had 100% survival after 96 h in pH 5.2 (Wasielesky et al. 1997). No growth reduction occurred for blacklip abalone at pH 7.76, a level where survival was affected, compared with the lower level, where both growth and survival were affected. The decrease in survival for blacklip abalone at pH 7.76 does not seem related to treatment levels and is more likely caused by stress on the abalone from handling. This treatment was restocked on day 15, when all blacklips from the most acidic treatment (pH 6.64. n = 3; 12 days) were replaced because of total mortality, hence the possibility of differences to handling for these abalone. In this study, greenlip abalone demonstrated a similar pattern of oxygen consumption in response to pH as had been observed pre- viously for the prosobranch gastropod Viviparus contectoides (Buckingham and Freed 1976). In their study, V. contectoides demonstrated two peaks in oxygen consumption at pH values 7.1 and 8.9, with an intervening trough. They suggested that, although energetically possible for V. contectoides to exist at pH 7.1 and 8.9, it involved substantial energy cost. In rainbow trout, this de- cline in oxygen consumption rate is a response to the reduced scope for activity at pH levels beyond those at which the organism can easily metabolise (Ye et al. 1991). In the case of greenlip abalone, the low pH experienced by the abalone depressed oxygen consumption rates. According to Wells et al. (1998). low internal pH should produce conditions where oxygen-hemocyanin affinity is highest. However, abalone rely heavily on anerobic metabolism during exercise or environmental hypoxia, and the subsequent metabolic acidosis will conserve this oxygen further (Wells et al. 1998). It is likely that the abalone from this experiment have altered oxygen-hemocyanin affinity caused by the experimental conditions, which is further exacerbated by the products of anero- bic metabolism, thus causing the decline in oxygen consumption The alterations to gill epithelium that were observed in this study for both greenlip and blacklip abalone are of note, because similar chanaes are known to occur in fish, as the cells become 618 Harris et al. damaged through accumulation of bicarbonate in the mucus layer (Randall 1991). The decrease in kidney definition and increased lumen size noted for both species of abalone indicate that pH can alter kidney structure. A decrease in nuclear size and staining intensity was observed in kidney cells of the brook trout, Salveli- nus fontinalis at pH 4.0 (Mudge et al. 1977). The variations in water quality experienced during this experi- ment are not believed to have influenced the results. Although a significant reduction in salinity was observed at pH 9.01, the re- duction was in the order of 0.5 ppt from the over-all mean of all other treatments ( 1 .5f/r reduction). Short-term survival of greenlip abalone is known to be affected at 23 ppt, and at 28 ppt, if inap- propriately fed (Boarder 1997). The daily temperature average of 19.0 °C is little different to the preferred temperature of greenlip abalone (18.3 °C) and blacklip abalone (17.0 °C) (Edwards 1996). FAN levels were higher at increased pH because of the influence of pH on ammonia ionisation (Bower and Bidwell 1978). though they were below 0.041 nig FAN I"1, the EC5 for greenlip abalone (Harris et al. 1998b). Within recirculation systems, there is greater likelihood of a combination of adverse water quality factors occurring, rather than just one factor deteriorating. Because nitrification is a complex of processes by which ammonia is converted first to nitrite then ni- trate, with concomitant acidification (Collins et al. 1975. Wickins 1983). then studying the effects of all these parameters in combi- nation would provide much greater understanding of the tolerances of abalone to recirculating systems. Presently, the effect of chronic- exposure to ammonia (Harris et al. 1998b). nitrite (Harris et al. 1997). dissolved oxygen (Harris et al. 1999) and pH (this study) are known, although this knowledge would be enhanced though subsequent combination studies. ACKNOWLEDGMENTS The authors thank the Fisheries Research and Development Corporation and the School of Aquaculture for research funding, the Tasmania Research Council for scholarship funding. Marine Shellfish Hatcheries for hosting this work, and Mr. Deon Johns for technical assistance. We also thank Mr. Rob Scharkie of Tas. Aqua Co. for the supply of the blacklip abalone. We also thank Dr. Natalie Moltschaniwskyj for critical assessment of the manuscript. LITERATURE CITED Bamber, R. N. 1987. The effects of acidic seawater on young carpet-shell clams Venerupis decussata (L.) (Mollusca: Veneracea). J. Exp. Mar. Biol. Ecol. 108:241-260. Bamber, R. N. 1990. The effects of acidic seawater on three species of lamellibranch mollusc. J. Exp. Mar. Biol. Ecol. 143:181-191. Beitinger, T. L. & R. W. McAuley. 1990. Whole animal physiological approaches for the assessment of stress in fishes. J. Great Lakes Res. 16:542-575. Boarder, S. J. 1997. Effects of dietary vitamin and mineral inclusion levels on the greenlip abalone Haliods laevigata Donovan. Unpublished hon- ors thesis. University of Tasmania. 125 pp. Bower, C. E. & J. P. Bidwell. 1978. Ionization of ammonia in seawater: effects of temperature. pH, and salinity. J. Fish. Res. Board Can. 35: 1012-1017. Bruno. T. J. & P. D. N. Svoronos (eds.). 1989. CRC handbook of basic tables for chemical analysis. CRC Press, Boca Raton. Buckingham. M.J. & D. E. Freed. 1976. Oxygen consumption in the prosobranch snail Viviparus contectoides (Mollusca: Gastropoda )-II. Effects of temperature and pH. Comp. Biochem. Physiol. 53A:249-252. Calabrese. A. & H. C. Davis. 1966. The pH tolerance of embryos and larvae of Mercenaria mercenaria and Crassostrea virginica. Biol. Bull. Woods Hole 131:427-436. Collins. M. T.. J. B. Gratzek. E. B. Shotts, Jr.. D. L. Dawe, L. M. Campbell & D. R. Senn. 1975. Nitrification in an aquatic recirculating system. J. Fish. Res. Board Can. 32:2025-2031. Dal Pont, G. M. Hogan & B. Newell. 1974. Laboratory techniques in marine chemistry II. determination of ammonia in sea water and the preservation of samples for nitrate analysis. CSIRO Division of Fish- eries and Oceanography 55. Cronulla. Edwards, S. J. 1996. Thermal preferences and tolerances of greenlip and blacklip abalone. pp. 65-67. In: P. W. Hone (ed.). Proceedings of the 3rd Annual Abalone Aquaculture Workshop, July 1996, Port Lincoln, SA. SARDI, Adelaide, Australia. Fleming. A. E. & P. W. Hone. 1996. Abalone aquaculture. Aquaculture 140:1-1. Grasshoff, K. 1989. Methods of seawater analysis. Velag Chemie. New York. Harris, J. O., G. B. Maguire, S. J. Edwards & S. M. Hindrum. 1997. Effect of nitrite on growth and oxygen consumption for juvenile greenlip abalone, Haliotis laevigata Donovan. /. Shellfish Res. 16:395-401. Harris, J. O. G. B. Maguire & J. H. Handlinger. 1998a. Effects of chronic exposure of greenlip abalone, Haliotis laevigata Donovan, to high am- monia, nitrite, and low dissolved oxygen concentrations on gill and kidney structure. J. Shellfish Res. 1 7:683-687. Harris. J. O.. G. B. Maguire. S. J. Edwards & S. M. Hindrum. 1998b. Ef- fect of ammonia on growth rate and oxygen consumption rate for juvenile greenlip abalone, Haliotis laevigata Donovan. Aquaculture 160. 259-272. Hams. J. O., G. B. Maguire, S.J. Edwards & D. R. Johns. 1999. Low dissolved oxygen reduces growth rate and oxygen consumption rate for juvenile greenlip abalone, Haliotis laevigata Donovan. Aquaculture 174:265-278. Higham, J.. P. Hone, S. Clarke, R. Baudinette & M. Geddes. 1998. The effect of flow on growth in juvenile greenlip abalone. Haliotis laevi- gata (Donovan), pp. 1 15-122. In: P. W. Hone (ed.). Proceedings of the 5th Abalone Aquaculture Workshop. July 3-6 1998, Hobart. FRDC. Henley Beach. Hindrum, S. M.. M. Cropp, D. O'Brien. N. Savva, G. B. Maguire & D. R. Johns. 1996. Performance of greenlip (Haliotis laevigata) and black- lip— greenlip hybrid abalone in land-based or sea-based production sys- tems, pp. 15-38. In: P. W. Hone (ed.). Proceedings of the 3rd Annual Abalone Aquaculture Workshop. August 1996, SARDI. Adelaide, South Australia. SARDI, Adelaide. Innes. A. J. & D. F. Houlihan. 1985. Aerobic capacity and cost of loco- motion of a cool temperate gastropod: a comparison with some Medi- terranean species. Comp. Biochem. Physiol. 80A:487— 493. Mallat, J. 1985. Fish gill structural changes induced by toxicants and other irritants: a statistical review. Can. J. Fish. Aquat. Set. 42:630-648. Meyers. T. R. & J. D. Hendricks. 1985. Histopathology. pp. 283-331. In: G. M. Rand & S. R. Petrocelli (eds.). Fundamentals of Aquatic Toxi- cology. Hemisphere Publishing. Washington. Mudge. J. E.. J. L. Dively. W. H. Neff & A. Anthony. 1977. Inter-renal histochemistry of acid-exposed brook trout. Salvelinus fontinalis (Mitchill). Gen. Comp. Endocrinol. 31:208-215. Oakes. F. R. & R. D. Ponte. 1996. The abalone market: opportunities for cultured abalone. Aquaculture 140:187-195. Palmer. A. R. 1981. Do carbonate skeletons limit the rate of body growth? Nature 292:150-152. Preston, S. J., D. Roberts & W. I. Montgomery. 1996. Crab predation as a selective agent on shelled gastropods: a case study of Calliostoma zizyphinum (Prosobranchia: Trochidae). pp. 313-325. In: J. Taylor pH and Growth for Two Australian Abalone Species 619 led.). Origin and Evolutionary Radiation ot the Mollusca. Oxford Uni- versity Press, Oxford. UK. Randall, D. 1991. The impact of variations in water pH on fish. pp. 90-104. In: D. E. Brune & J. R. Tomasso (eds.l. Aquaculture and Water Qual- ity. World Aquaculture Society, Baton Rouge. LA. Russo, R. C. 1985. Ammonia, nitrite, and nitrate, pp. 455—171. In: G. M. Rand & S. R. Petrocelli (eds.l. Fundamentals of Aquatic Toxicology. Hemisphere Publishing. Washington. Shepherd, S. A. 1973. Studies on southern Australian abalone (genus Hali- otis). I. ecology of five sympatric species. Australian J. Mar. Fresh- water Res. 24:2 1 7-257. Sokal, R. R. & J. F. Rohlf. 1995. Biometry, the principles and practice of statistics in biological research. W. H. Freeman. Bristol, PA. Solorzano. L. 1969. Determination of ammonia in natural waters by the phenolhypochlorite method. LimnoL Oceanogr. 14:799-801. Sprague. J. B. 1969. Measurement of pollutant toxicity to fish. I. bioassay methods for acute toxicity. Water Res. 3793-82 1 . Thurston. R. V. & R. C. Russo. 1981 . Ammonia toxicity to fishes, effect of pH on the toxicity of the un-ionized ammonia species. Environ. Sci. Techno/. 15:837-840. Underwood. A. J. 1981. Techniques of analysis oi variance in experimental marine biology and ecology. Oceanogr. Mar. Biol. Ann. Rev. 19:513- 605. Wasielesky. W.. Jr.. A. Bianchini. M. H. S. Santon & L. H. Poersch. 1997. Tolerance of juvenile flatfish Paralichthys orbignyanus to acid stress. ./. World Kquacu.lt. Soc. 28:202-204. Wickins, J. F. 1983. Studies on marine biological filters. Model filters. Water Res. 17:1769-1780. Willows, R. I. 1994. The ecological impact of different mechanisms of chronic sublethal toxicity on feeding and respiratory physiology, pp. 88-97. In: D. W. Sutcliffe (ed.l. Water Quality and Stress Indicators in Marine and Freshwater Ecosystems. Freshwater Biol. Assoc. 6-7 Sept. 1993. Edinburgh. Ye. X.. D. J. Randall & X. He. 1991 . The effect of acid water on oxygen consumption, circulating catecholamines and blood ionic and acid-base status in rainbow trout (Salmo gairdneri, Richardson). Fish Physiol. Biochem. 9:23-30. Zar, J. H. 1996. Biostatistical analyses. Prentice-Hall of Australia. Pty. Ltd.. Sydney. Journal of Shellfish Research, Vol. 18. No. 2, 621-625. 1999. GENETIC VARIATION FOR SURVIVAL AND SHELL LENGTH OF CULTURED RED ABALONE (HALIOTIS RUFESCENS) IN ICELAND J. JONASSON,' S. E. STEFANSSON,1 A. GUDNASON,2 AND A. STEINARSSON3 1 Stofnfiskitr Ltd. Laugavegi 103 PO Box 5166 125 Reykjavik, Iceland 2Saebyli Ltd. 190 Vogar Iceland Marine Research Institute Skulagata 4 101 Reykjavik. Iceland ABSTRACT A research program was started in 1996 to study genetic variation in survival and size of imported red abalone cultured in Iceland. A total of 100 families was produced from a hierarchical mating design using 29 males and 88 females. Larvae from each family were settled and grown in separate tanks until tagged at the age of 10 months, at which time individuals from all families were reared in one common environment, but at two farms, until the age of 24 months. After tagging, the abalone were fed with the macro algae species Palmaria palmata and Laminaria digitata. The mean survival rate of the population was estimated at 9.6% at the age of 4 months, with large variation among families where the highest survival was 31.5% but the lowest 0.02%. The average shell length at the age of 8, 10, 18. and 24 months old was 13.10, 15.88, 34.99, and 50.90 mm. respectively. Heritability for survival at 4 months of age was estimated to 0. 1 1 (0.33 on the liability scale) and for shell length at the age of 8, 10, 1 8. and 24 months to 0.08, 0.06. 0.27, and 0.34. respectively. A low, but negative, genetic correlation between the survival rate to 4 months and shell length was observed. Genotype by environment interaction was negligible after rearing individuals from the same families in two different farms for 14 months. Based on the heritability estimates observed, it was concluded, that, theoretically, it is possible to double the growth rate of cultured red abalone in just four generations by selection at the end of 2 years of rearing. INTRODUCTION Selection of quantitative traits has been one of the most im- portant tools in the improvement of production from farm animals and plants over the past few decades. The red abalone {Haliotis rufescens) is one of over 100 abalone species belonging to the genus Haliotis (Howorth 1978). The red abalone is the largest in the world often reaching shell lengths greater than 27.5 cm and weighing over 1.7 kg (McAlister 1976). A population of 2,300 individuals was imported to Iceland from California in 1988 to establish an aquaculture industry (Steinarsson 1993). Abalone cul- ture is a rapidly growing industry throughout the world, because abalone is one of the most valuable shellfish on the international market (Gudnason pers. comm.l. Despite the great interest in aqua- cultural production of abalone, no study for estimating genetic parameters in abalone has yet been reported. Results from breeding experiments with other shellfish species suggest that additive genetic variance leads to successful selection programs for growth rate. Hara and Kikuehi (1989) demonstrated an increase in daily growth rate in the Japanese abalone. Haliotis discus hannai of 21% in animals of shell size 20-30 mm and a 65% increase in those of 30-70 mm after three generations of selection, when compared to the maximum growth of control com- mercial animals. Toro and Newkirk ( 1991 ) demonstrated response to selection for shell height in the Chilean oyster (Ostrea chilensis) with a realized heritability estimated to be 0.34 ± 0.12. Jarayab- hand and Thavornyutikam ( 1995) also demonstrated a response to selection for growth rate of the oyster Saccostrea cucullata with a realized heritability to 0.28 ± 0.006. The importance of estimating phenotypic and genetic param- eters of economically important life history traits, the genetic cor- relation between them, and predicting responses to selection and estimating breeding values are crucial for planing a breeding pro- gram. The magnitude of additive and nonadditive genetic variance in relation to the total variance of each trait will determine which breeding method (purebreeding and/or crossbreeding) and selec- tion method (individual and/or family selection) should be applied. A research project was initiated in 1996 to study the economi- cally important life history traits of survival, growth rate, meat yield, and age at maturity in farmed red abalone in Iceland. The aim of the project was to produce a breeding plan that would reduce production costs. In this paper, we report on estimates of the genetic parameters (heritabilities and genetic correlation) for survival at 4 months posthatching and shell length at 8, 10, 18, and 24 months of age. MATERIALS AND METHODS The broodstock used in this study were 9 years old from the original population imported to Iceland in 1988. The average shell length of the 88 females used for spawning was 14.6 cm (SD2.01), body weight 517.5 g (SD 185.7); the average shell length of the 29 males was 14.6 cm (SD 2.40) and body weight 576.3 grams (SD 245.7). Spawning and Settlement of Larvae Spawning at the Saebyli Ltd.'s farm (Farm 1, Table 1) was induced by increasing the water temperature and by exposure to UV light-treated seawater. The temperature was increased from 1 3 to 17 °C overnight before introduction of UV-treated seawater for 621 622 JONASSON ET AL. TABLE 1. Shell length and number of animals in), and coefficient of variation during the experiment, reared in two farms after the age of 10 months. Trait Age Months Farm Mean C.V. Length 1 S Length 2 II) Length 2 10 Length 3 18 Length 3 18 Length 4 24 Lentrth 4 24 9.648 13.10 27.9 3.46X 15.91 24.3 2.391 15.83 23.6 2.438 37.45 16.0 2,098 32.14 19.9 1,701 51.30 16.8 1 .354 50.40 17.3 a few hours. Altogether. 29 males and 88 females were used to produce 100 full- and half-sib families over a period of 1 month (September 1996). A hierarchical mating design was used where each male (sire) fertilized eggs of 2-6 females (dams). After fer- tilization and hatching, the larvae were held in separate tanks for 6 days at which time, the number of larvae per three 5-mL samples were counted in each tank. Thereafter, a sample of 5.000 larvae from each family were put into a 20-L tank in a greenhouse with feeder plates covered with microalgae. The feeder plates had been pregrazed by older abalone to induce settlement, these were re- moved before input of the larvae (Seki and Kan-no 1981). Each family was put in one tank, and the families were randomly dis- tributed over two stands and three shelves (Fig. 1 ) to reduce pos- sible variation caused by different environmental conditions within the greenhouse. Growing of microalgae was induced by direct sunlight, but at later staees. additional electrical lighting was used. continued eulturing. For families with lower survival. 5% of the remaining individuals were counted and returned. The mean shell length at 4 months of age was 3.30 mm (SD 0.91 ). At the age of 4 months, some of the animals in each family were too small to be measured. Therefore, the first growth measurement was postponed until the age of 8 months, when approximately 100 individuals were measured per family. The widest diameter of the shell was measured to the nearest 0.01 mm. At the age of 10 months, a random sample of approximately 60 individuals from each family were tagged and measured; one family was not tagged because of high mortality. At tagging, each family was split up in two groups; one for each of two rearing sites. Two weeks after tagging, one group of all families was transported to the Marine Institutes re- search farm (Farm 2. Table 1 ) 20 km from the Saebyli's farm. All tagged individuals were placed together in one big common tank at each farm. Two weeks later, the animals were randomly placed into 40-L tanks with about 170 animals in each tank. At the age of 18 and 24 months, all tagged individuals were measured again. Tag-loss was observed; therefore, fewer than 30 individuals were measured per family at each site. Temperature The average water temperature was 14.2 °C during the first 4 months of growth. 15.8 °C for the next 4. and increased to 17.1 °C for the next 2 (to age 10 months). At Saebyli's farm the average temperature from the age of 10 to 18 months was 17.3 °C and 16.9 °C from the age 18 months to 24. At the Marine Institute's farm, the average temperature from age 10 to 18 was 16.0 °C, and from age 18 to 24 months it was 16.8 °C. Statistical Analysis Feeding with Macroalgae At the age of 6 months, fresh algae Palmaria palmata was added to the diet, and at the same time, the greenhouse lights were turned off for 5 hours during the night. At the age of 8 months, the greenhouse lights were turned off and fresh Laminaria digitata was added to the diet. Recording Traits Survival was estimated from the age of 6 days posthatching. at the time larvae were settled into 20-L tanks, until the age of 4 months, when all surviving individuals within each tank were counted. Into each tank, a random 250 survivors were returned for Figure 1. Location of the shelves and stands and the type of tank used from larval stages until tagging. Variance components for each of the four shell length mea- surements were estimated according to an animal model. The ani- mal model allows both the recorded individuals and the parents without records to be included in the analysis in order to account for all known additive genetic relationships between animals; also, the model includes both fixed and random effects. In matrix no- tation, the model can be written. y = Xb + Za + Zf + e where y is the observation vector; b is the vector of fixed (envi- ronmental) effects (for length, these were farm, number of animals in tanks, and age from hatching; and for survival, they were shelf and stand) a is the vector of random animal effects: f is the vector of random effects of full-sib group caused by factors other than additive genetics; e is the vector of individual random error effects. X and Z are the corresponding design matrices. The additive effects, the common full-sib effects attributable to other factors than additive gene effects, and the residual effects were assumed to have independent normal distributions with zero means and variances of 7.2 7.8 8.1 10.7 14.2 Genetic Correlation Between Survival and Shell Lenfftli For estimation of the genetic correlation between survival and shell lengths 1 (8 months) and 4 (24 months), a bivariate analysis was run in which survival and shell length were regarded as two different traits. A genetic correlation between survival and length 2 and length 3 was not computed. The error covariance between the traits was set to zero, because they are not measured on the same animal when running a multitrait analysis for the two traits. The same models as above for survival and shell lengths were used, but excluding the common full-sib effect. When estimating the genetic correlations between the four shell lengths 2 (10 months), 3 (18 months), and 4 (24 months), the error covariance was computed, because the measurements are made on the same individuals at different ages. The genetic correlation between length 1 and lengths 2. 3, and 4 was not computed, because lengths 1 and 2 were measured with only a 2-month interval. Genotype by Environmental Interaction For estimation of the genetic correlation between shell length at the age of 18 (length 3) and 24 (length 4) months in the two different farms, a univariate analysis was run as described above but excluding fixed effect of farm. Subsequently, a bivariate analy- sis was run with shell length at each farm for both age classes, regarded as two different traits. The error covariance between the traits was set to zero, because they are not measured on the same animal. The common full-sib effect was not included. RESULTS The mean rate of survival for the 100 families at the age of 4 months was 9.6%, with large variation among families where the highest survival was 31.5% but the lowest 0.02%. Significant ef- fect (99.5% level) was found for the position of the family tank on the two stands and shelves (Table 2). Higher survival was in ob- TABLE 3. Shell length (standard deviation) at the age of 8 months in stands and shelves. Stand/Shelf // Mean (SD) n Mean (SD) 3 1464 13.39(3.6) 2 1907 12.49(3.1) 2117 12.72(3.3) 1 1 770 13.36(3.8) 1889 13.42(4.2) Total 5641 13.08(3.5) 4006 13.05(3.8) TABLE 4. Proportion of partial sum of square for shell length, effect of each fixed (environment) effect, as R" in v/t on shell length during the experiment. Effect Length I Length 2 Length 3 Length 4 Shelf Stand Animals (») in tank Animals iir) in tank Age from hatching Age2 from hatching Farm Total 0.01 0.0 18.22 0.56 1.1 I 0.16 11.6 1.28 3.53NS 0.01NS 0.0NS 17.7 0.2 0.97 2.74 0.62 2.45 2.01 14.96 24.0 0.19' 1.13 0.01 0.92 3.40 2.46 0.33 8.4 ' Nonsignificant; all other effects significant at 99.5% level. served on stand 2 and particularly on shelf 1, at 14.2%. No sig- nificant effect of tank position was observed for shell length at 8 months of age (Table 3). The importance, expressed as percentage of partial sums of squares (R2). of each environmental effect included in the model for the shell length measurements are given in Table 4. The dif- ference in number of animals in a tank had the greatest effect on shell length at 8 and 10 months of age. with minimum effect evident by 24 months of age (length 4). The difference between farms was most evident at 18 months (length 3). with the main effect at 24 months being age-at-hatching. The effect of stand and shelves in the first 10 months of growth had minor effect on all the models although the effect was significant on the 99.5% level. The mean shell lengths observed were 13.10 mm at 8 months (length I), 15.88 mm at 10 months (length 2), 34.99 mm at 18 months (length 3), and 50.90 mm at 24 months (length 4) (Table 1 ). A substantial variation between families was observed in shell length at the age of 24 months (Fig. 2). The largest mean shell length for a family was 66.9 mm and the smallest was 42.2 mm. Heritabilities for survival (0. 1 1 ) and shell lengths 1 and 2 (0.08 and 0.06) were low (Table 5). but when survival was transformed to the underlying liability scale, the heritability was 0.33. A higher growth heritability estimate was observed after 18 months (0.27) and even higher after 24 months (0.34). Effect other than additive effect (f2) was lower than the heritability for all traits (Table 5). 60,9 59,7 45 4 4c p .. **3 ^ 45 1 fli i Hill 2 27 1 3 29 Family number 33 58 39 34 Figure 2. Mean shell length of family groups for the five largest and the five smallest families at the age of 24 months from hatching. Mean shell length for the 100 families tested was 50.9 mm. 624 JONASSON ET AL. TABLE 5. Heritability estimates (diagonal), with then full-sib group effect caused by factors other than additive genetics (in parenthesis) and genetic correlations for the 5 traits recorded. Survival Length 1 Length 2 Length 3 Length 4 Survival 0.11 (0.0) Length 1 -0.18 0.08 (0.06) Length 2 0.06 (0.04) Length 3 0.28 0.27 (0.0) Length 4 -0.12 0.09 0.09 0.34 (0.0) The estimated genetic correlation between survival and shell lengths 1 and 4 were low and negative at -0.18 and -0.12, respec- tively. As shown in Table 5, the genetic correlation between shell lengths 2 and 3 was low and even lower between shell lengths 2 and 4. Genetic correlation between shell lengths 3 and 4 was close to unity (0.9). The genetic correlations between shell length at the age of 18 months and 24 months in the two farms was estimated to 0.94 and 0.95. respectively. DISCUSSIONS A low survival rate of 10% during the first 4 months is quite common in red abalone culture (Fallu 1991 ). The mean growth rate during the experimental period of 24 months was 2.12 mm/month. Such a growth rate was similar to the 2.4—2.5 cm/year growth rate reported by Leighton ( 1974), Ebert and Houk (1984). An effect of density on growth during the first 8 months of rearing was ob- served in our families. A large variation was observed between full-sib families in survival and shell length. Limited literature is available on the subject of growth variation in abalone. Hara ( 1990) observed significantly lower growth in one of three families over a 180-day growing period. The heritability for survival during the first 4 months was low at 0. 1 1 but does suggest that it may be possible to improve the trait by selection. However, in general, survival is not of great eco- nomic importance, because of the high fecundity of females (Ault 1985); a single female can easily produce over 3 million eggs allowing sufficient spat production even at a survival rate of only 10%. Heritability for shell length was low for lengths 1 and 2 but higher for lengths 3 and 4. An increase in heritability with age has also been reported for other shellfish species; for example, oysters (0.24-0.5; Longwell 1976, Newkirk et al. 1977. Losee 1978) and blue mussel (0.12-0.43; Innes and Haley 1977, Newkirk 1980). It must be emphasized that all the estimates presented for oysters and blue mussels were estimated from data from few families. It is not uncommon that heritability estimates at early life stages are lower than when estimated nearer to market size. This is also the case for Atlantic salmon and rainbow trout. (Jonasson 1993, Jonasson et al. 1997, Refstie 1980, Gjerde and Schaeffer 1989). The statistical model used allowed estimates of effects other than additive genetics (f2). The observed low estimates for shell lengths 1 and 2 are most likely a transitory common environmental effect introduced when the groups were reared in separate tanks during the first 10 months of the experiment. This is demonstrated in Table 4, where it is shown that the effect of shelf, stand, and number of animals in tanks in the first 10 months is greatly re- duced as the animals get older and are reared together in a common environment. However, nonadditive genetic effects cannot be ruled out. Genetic correlation between survival and shell lengths 1 and length 4 was negative but low. Worth keeping in mind is the categorical nature of the survival observation. The estimated ge- netic correlations are, therefore, systematically underestimated (Kendall and Stuart 1961). The present results indicate an unfa- vorable genetic correlation between survival in the first 4 months and shell length until 2 years of age. This is not in agreement with studies in oyster (Jarayabhand and Thavornyutikarn 1995). where survival of groups selected for fast growth was relatively higher than groups selected from the medium and slow growing part of the population. In studies on Atlantic salmon ranching. Jonasson (1992) estimated genetic correlation between survival and body length of Atlantic salmon fingerlings to 0.39 ± 0.26. and Jonasson et al. ( 1997) estimated the genetic correlation between body weight and survival of grilse at sea in Atlantic salmon ranching to 0.16 ± 0.16. More data for abalone are needed; furthermore, survival during the first 4 months will most likely be correlated to other life history traits. The low genetic correlation between shell length at the age of 10 months and at the age of 2 years indicates that shell length at young age is a poor estimate of the individual shell length at older stages. Market size is between 50-100 mm (Gudnason pers. comm.), and the results show that, in a breeding program for abalone, selection for growth should be taken as close to market size as possible. Genetic correlation between the two farms for shell length is close to unity. This indicates that genotype by environmental in- teraction is negligible in this experiment for the first 2 years of rearing. It may, therefore, be concluded that selective breeding program for red abalone in Iceland can be based on one breeding population only and shell lengths of full- and half-sib groups can be recorded based on rearing in one farm. The heritability for shell length at age 24 months was estimated at 0.34. indicating that prospects for improving growth rate are good. By using the standard formula for genetic gain for individual selection (AG = /*crp*/r. Falconer 1989), assuming that 5% of the largest animals will be selected as broodstock and using the phe- notypic standard deviation (8.8 as in Table 1) obtained after 24 months of age, the genetic gain will be 6.0 mm per generation. Such a gain represents a shortening in production time of 2.83 months to the mean size of 56.90 mm. Should this predicted gain be real, then it would take just over four generations to double the growth rate in red abalone in Iceland by selection. By applying a combined individual and family selection for growth, the genetic gain during the first 2 years will be higher, because combined individual and family selection is more effective for traits with low heritabilities (Falconer 1989). Genetic Variation of Red Abalone 625 At this stage, it is too early to suggest a breeding plan for red abalone culture, because market size is usually reached in 2-3 years of age. Genetic parameters for such other life history traits as age at market size, meat yield, and age at maturity will be esti- mated on these same families, and economic evaluation of all traits will be performed. However, our results do suggest that significant production gains can be made in the culture of abalone through a selective breeding program. ACKNOWLEDGMENTS The National Research Counsel of Iceland as well as Saebyli Ltd.. Stofnfiskur Ltd. funded the project together with the Marine Research Institute in Iceland. Special thanks to Professor Trygve Gjedrem at AKVAFORSK, Norway and Dr. Nicholas G. Elliot at the CSIRO Marine Research in Tasmania. Australia for comments on the work and manuscript. LITERATURE CITED Ault J. 1985. Some quantitative aspects of reproduction and growth of the red abalone. Haliotis rufescens Swainson. J. World Maricult Soc. 16: 398. Boldman. K. G„ L. A. Knese. L. D. Van Vleck. C. P. Van Tassell & S. D. Kachman. 1995. A manual for use of MTDFREML. A set of programs to obtain estimates of variances and covariances (draft). U.S. Depart- ment of Agriculture. Agricultural Service, Washington. DC. Ebert. E. E. & J. L. Houk. 1984. Elements of innovation in the cultivation of red abalone Haliotis rufescens. Aquaculture 39:375. Falconer, D. S. 1952. The problem of environment and selection. Am. Nat. 86:75-86. Falconer. D. S. 1989. Introduction to quantitative genetics. Longman. Har- low. UK. 438 pp. Fallu. R. 1991. Abalone farming. Fishing news books. Blackwell Scien- tific, Oxford, UK. 192 pp. Giorgi, A. E. & J. D. DeMartini. 1977. A study of the reproductive biology of the red abalone Haliotis rufescens Swainson, near Mendocino, Cali- fornia. Calif. Fish Game 63:80. Gjerde. B. & L. R. Schaeffer. 1989. Body traits in rainbow trout. II. esti- mates of heritabilities and of phenotypic and genetic correlations. Aquacuhure 80:25-44. Hara, M. 1990. The effects of genetics on growth in three groups of abalone seeds. Bulletin Tohoku National Fisheries Research Institute 0(52). 73-78. In Japanese. Hara, M. & S. Kikuchi. 1989. Increasing the growth rate of abalone. Haliotis discus hannai, using selection techniques. National Oceanic and Atmospheric Administration NOAA Tech. Rept.. National Marine Fisheries Service Rept. NMFS 106. 21-26. Howorth. P. C. 1978. The abalone book. Naturegraph, Happy Camp. CA. 1 . Innes. D. J. & L. E. Haley. 1977. Genetic aspects of larval growth under reduced salinity in Mytilus edulis. Biol. Bull. Woods Hole 153: 312-321. Jarayabhand, P. & M. Thavomyutikarn. 1995. Realized heritability esti- mation on growth rate of oyster. Saccostrea cucullata born, 1778. Aquaculture 138:111-18. Jdnasson J. 1993. Selection experiments in salmon ranching. I. genetic and environmental sources of variation in survival and growth in freshwa- ter. Aquaculture 109:225-236. Jonasson J., B. Gjerde & Gjedrem. 1997. Genetic parameters for return rate and body weight of sea-ranched atlantic salmon. Aquaculture 154:219- 231. Kendall. M. G. & A. Stuart. 1961. The advanced theory of statistics, vol. 2. Griffin. London. Leighton, D. L. 1974. The influence of temperature on larval and juvenile growth in three species of Southern California abalone. Fish. Bull 72:1137. Longwell, A. C. 1976. Review of genetic and related studies on commer- cial oysters and other pelecypod mollusks. J. Fish. Res. Board Can. 33:1100-1107. Losee, E. 1978. Influence of heredity on larval and spot growth. Proceed- ings of the 9th Annual Meeting of the World Mariculture Society, pp. 101-107. McAllister. R. 1976. California marine fish landings for 1974. Calif. Fish Game Fish Bull. 166:1. Meyer, K. 1989. Restricted maximum likelihood to estimate variance com- ponents for animal models with several random effects using a deriva- tive-free algorithm. Genet. Set Evol. 21:317-340. Newkirk. G F. 1980. Review of the genetics and the potential for selective breeding of commercially important bivalves. Aquaculture 19:209- 228. Newkirk, G F., L. E. Haley. D. L. Waugh & R. W. Doyle. 1977. Genetics of larval and spat growth rate in oysters (Crassostrea virginica). Mar. Biol. 41:49-52. Refstie. T. 1980. Genetic and environmental sources of variation in body weight and length of rainbow trout fingerlings. Aquaculture 19:351- 357. Robertson. A. & I. M. Lerner. 1949. The heritability of all-or-none traits: liability of poultry. Genetics 34:394-41 1. Seki, T. & H. Kan-no. 1981. Induced settlement of the Japanese abalone. Haliotis discus hannai. veliger by mucous trails of the juvenile and adult abalones. Bull. Tohoku Reg. Fish. Res. Lab. 43:29. Steinarsson A. 1993. Breeding and grow-out of red abalone in Iceland. Report from the Marine Institute in Iceland. 27 pp. Toro, J. E. & G. F. Newkirk. 1991. Response to artificial selection and realized heritability estimate for shell height in the Chilean oyster Os- trea chilensis. Aquatic Living Resources 4: pp. 101-108. Journal of Shellfish Research, Vol. 18, No. 2. 627-635. 1999. GROWTH OF TAIWAN ABALONE HALIOTIS DIVERSICOLOR SUPERTEXTA FED ON GRACILARIA TENUISTIPITATA AND ARTIFICIAL DIET IN A MULTIPLE-TIER BASKET SYSTEM JIANN-CHU CHEN AND WON-CHUNG LEE Department of Aquaculture National Taiwan Ocean University Keelung, Taiwan, 20224 Republic of China ABSTRACT Taiwan abalone. Haliotis diversicolor supertexta, juveniles (26.64 ± 2.30 mm) were placed in 4 sets of 7-basket tiers. 35 abalone in each basket, and reared indoors with running seawater (31-35%o) for 395 days. Food was generally given every other day at a rate of 30% and 3% of total body weight for Gracilaria tenuistipitata and artificial diet, respectively. Survival was 64.3-7 1 .4% and 61 .4-78.6% in the Gracilaria and artificial diet, respectively. Growth of abalone placed on the top tier was inferior to those placed on the middle and bottom tiers. In the first 33 days, the growth rate of H. diversicolor supertexta fed Gracilaria and the artificial diet was 61-101 and 105-163 u,m/day. respectively. The abalone fed an artificial diet gained twice as much total body weight as those fed Gracilaria after 96 days. The overall growth rate of abalone fed an artificial diet was 24.4. 47.7, and 47.9 mg/day, or 1 .8, 2.2. and 2.2 times higher than abalone fed Gracilaria, on the top, middle, and bottom tiers, respectively. The FCR (feed conversion ratio) in dry weight of food was 3.10. 2.68. and 2.76 for the abalone fed an artificial diet, and 4.43. 3.03. and 3.04 for those fed Gracilaria. It is concluded that H. diversicolor supertexta placed in a multiple-tier basket system and fed an artificial diet could grow to market size (> 40 mm shell length) in less than half the time in comparison to those fed on Gracilaria. KEY WORDS: Haliotis diversicolor supertexta, Gracilaria tenuistipitata, artificial diet, growth INTRODUCTION Taiwan abalone (also known as small abalone) Haliotis diver- sicolor supertexta Lischke, which live in the littoral zone of rocky shores along the south coast of Japan and northeast coast of Tai- wan, grow to 100 mm shell length in the wild (Nie 1992). Culture of H. diversicolor supertexta has expanded greatly since 1986 due to development of successful propagation and larval rearing tech- niques (Chen and Yang 1979, Lin 1986. Yang and Ting 1986). The farmed production of H. diversicolor supertexta in Taiwan was 502 tonne in 1986, doubled in 1987. and reached 2.213 tonne in 1997 (Taiwan Fisheries Bureau 1998). Culture of H. diversicolor supertexta in Taiwan is commer- cially divided into 3 phases. The first stage is the culture of newly settled spat on corrugated plastic plates until a shell length of 2-3 mm is attained. The second stage is the culture of 2-3 mm larvae on cement plates to 20-30 mm juveniles. The third stage is the culture of 20-30 mm juveniles on cement plates or baskets to reach a market size shell length of more than 40 mm. Gracilaria has been cultured commercially since 1962 on land farms. Most Gracilaria produced is used as food for the Taiwan abalone. According to the Taiwan Fisheries Bureau (1998). farmed Gracilaria totaled 9,232 and 12,576 ton in 1986 and 1997, respec- tively. This amount is not sufficient to meet the needs of the abalone produced, and recently a number of artificial diets have been developed for abalone culture. Growth of abalone feeding on macroalgae and artificial diets has been studied for H. discus (Ogino and Ohta 1963). H. discus hannai (Uki et al. 1985. Nie et al. 1986, Uki et al. 1986a. 1986b). H. fulgens (Viana et al. 1993). H. asinina (Capinpin and Core 1996), H. iris (Stuart and Brown 1994). H. tuberculata (Koike et al. 1979, Mgaya and Mercer 1995). H. laevigata (Morrison and Whittington 1991) and//. midae(Bntx 1996a. 1996b, Knaueretal. 1996). However. little is known about the growth of H. diversi- color supertexta (Chen 1984). A multiple-tier basket system for the culture of abalone has recently been developed in Taiwan. In this system, abalone juve- niles are placed inside the baskets and the baskets are stacked in 4 to 14 tiers in indoor cement ponds. This system is the best way to culture H. diversicolor supertexta in Taiwan where land is limited. Most farmers install a mechanical device to lift the basket for feeding and cleaning. However, we do not know how much varia- tion there is in the harvest when using a multiple-tier system. The objectives of this research were to quantify the growth rate of H. diversicolor supertexta fed on Gracilaria and artificial diet, and compare the growth of abalone placed on the top. middle, and bottom tiers in a multiple-tier basket system. This is the first report of a comparatively long experimental trial to show the effect of diet on growth performance for H. diversicolor supertexta. MATERIALS AND METHODS Test abalone H. diversicolor supertexta, which were hatched and reared for about 3 months at a private farm in Kaohsiung. Taiwan, were shipped to our University on October 10. 1996. They were reared on cement plates in a concrete pond and fed red macroalga Gracilaria tenuistipitata Var. liui Zhang et Xia prior to shipping. They were placed inside the greenhouse and acclimated to running seawater at 35%< for 20 days before experimentation. During the acclimation period, the juveniles were divided into 2 groups. One group of abalone juveniles was fed Gracilaria, and the other group was fed artificial diet (see Experiment Diet for details). For the experiment, 420 abalone juveniles were used, mean (± SD) body weight 2.137 ± 0.52 g and shell length 26.64 ± 2.30 mm. No significant difference in weight and shell length was observed among the treatments. Experimental System Perforated plastic baskets (39-cm long. 31 -cm wide. 12-cm high; with perforations 9.2 x 9.6 mm) were employed in the study. 627 628 Chen and Lee Each basket was stocked with 35 juvenile abalone. Seven baskets were stacked into a tier and connected to a cement block to the bottom of the tier. In all, 4 sets of the 7-basket tiers were used: 2 for feeding with Gracilaria and 2 for feeding with artificial diet. The 4 sets of baskets were placed in an oval fiberglass tank. 2.0-m diameter and 1 .2-m deep. The bottom of the tank was covered with gravel and sand 12-cm deep, and aerated with airlift to create a biological filter (Fig. 1 ). Seawater pumped from the coast adjacent to the University passed through a sand filter into the tank con- tinuously at a flow rate of 479 ± 6 L/h. The abalone placed in the 1 st. 4th, and 7th basket of each tier served as the top. middle, and bottom, respectively. There were 6 treatments (2 diets x 3 tiers - top, middle, and bottom), and each treatment was conducted in 2 replicates. Experiment Diet Two diets were used in the study. The red macroalga, Gracilaria tenuistipitata var. Hue Zhang et Xia, was harvested from farms in Tainan. Taiwan, and shipped to the University. The percentage composition (± SE) of Gracilaria was moisture 89.03 ± 0.85, crude protein 26.08 ± 2.64. crude lipid 1.52 ± 0.70. crude fiber 6.58 ± 0. 15. ash 35.29 ± 0.88. and nitrogen-free extract 30.53 ± 3.03 in dry base. The artificial diet was manufactured by Tonlee Feed Company (Pingtung. Taiwan) based on seaweed powder, soy M bean powder, and wheat powder as main ingredients. The artificial diet was prepared in a row of 3 cylinder-shaped pellets (4-mm diameter and 10-30-mm long). The percentage composition of artificial diet was moisture 12.99, crude protein 30.02, crude lipid 4.41. crude fiber 3.14. ash 10.93. nitrogen-free extract 51.50 in dry base. Feeding experiment Food was generally given every other day at a rate of 30 and 3% of body weight of abalone for Gracilaria and artificial diet, respectively, based on preliminary observation and the experience of abalone growers. However, feeding amount was reduced to 20 and 27c of body weight for Gracilaria and artificial diet, respec- tively, from March to April 1997, the period considered by farmers to be dangerous for the animals due to the south wind. In general, feeding was 15 times a month, but was reduced to 7 times in August 1997 by power failure caused by 2 typhoons; feed was not given during this period to avoid accumulation of toxic ammonia. Fresh Gracilaria was collected every 4-5 days, blotted on filter paper, weighed, and fed to the abalone. The uneaten food was removed every other day. blotted on filter paper, and weighed. The artificial diet kept its shape for 48 h. but its stability in water over 48 h is 73.4%. Consumption of artificial diet corrected for leaching was calculated using the formula of Britz (1996b). The experiment started November 1. 1996. and lasted for 395 days. The baskets were checked every other day for abalone sur- vival, and dead abalone were removed. Growth was generally measured every month as body weight and shell length. An elec- tronic balance was used to register the weight in mg, and a vernier caliper was used to measure the shell length in mm. Percentage weight gain, percentage length increase, growth rate in weight (mg/day) and shell length (p.m/day), and FCR (feed conversion ratio) were calculated. FCR was calculated based on the wet weight feed consumed (g) per wet weight gain (g). based on the dry weight feed consumed (g) per wet weight gain (g), and ex- pressed as FCR (Wet) and FCR (Dry), respectively. During the experimental period, the temperature range was 14.5-31.0 °C (Fig. 2) (measured with thermometer model HOBO HTEA. Onset Computer Corp., Pocasset, MA). The illuminance in the top. middle, and bottom tier was measured with an illuminom- eter (model HOBO HLI OK-P) and averaged 0.3, -2.1, and -2.5 log lm/ft2. respectively (Fig. 3). Salinity, monitored with a refractom- eter, ranged from 31 to 359^. and the dissolved oxygen, measured with a YSI Model 58 DO meter (YSI, USA), ranged from 5.8 to 35 i 30 25 - 20 - 15 10 Figure. 1. A schematic diagram of the 7-tier basket system, a, water inlet; b, water outlet; c, airlift; d, gravel; e, basket; f, cement block; g, abalone. Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month Figure. 2. Water temperature in the culture system from November 1996 to November 1997. Growth of Taiwan Abalone 629 Weight 28 °C) were recorded, abalone growth was low compared to the other periods. The reduction in growth rate was ascribed to reduction in the feeding rate of the animals due to typhoon strikes. Low growth rate may also relate to sexual matu- ration (Ohba 1964). However, we did not observe the maturation status of abalone in this study. In the present study, despite the low temperatures (16-24 °C) recorded during November to February, there was consistently higher growth in abalone fed on artificial diet. This indicates that H. diversicolor supertesta fed on artificial diet grew significantly faster during in the first 96 days. The larvae of Taiwan abalone are generally available during October to March. We suggest that the larvae can be fed on artificial diets when they are just removed from the plastic plates. In fact, it has become a common practice to raise larvae to 20-30 mm in nursery ponds using artificial diets for the juvenile growers, when larvae are placed inside cement plates as shelters in concrete ponds. Our research indicated that H. diversicolor suptrtexta fed on artificial diet still grew well when the temperature dropped to below 20 °C. Uki and Watanabe ( 1992) reported that the growth rate of aba- lone increased with increasing protein level in diets. They also reported that the optimal level of protein for H. discu hamuli was 28%. Mai et al. ( 1995) reported that the optimum protein level was 22.3-32.3% and 23.3-35.6% for H. tuberculata and H. discus hannai, respectively. The protein content in dry weight used in the present study was 26% in Gracilaria tenuistipitata and 30% in the artificial diet. The fact that growth of the H. diversicolor fed on Gracilaria was inferior to those fed on artificial diet is considered to be due to a deficiency of essential nutrients, lower content of protein, lipid and nitrogen-free extract, and higher content of crude fiber and ash in Gracilaria as compared to artificial diet. In the present study, the abalone placed in the multiple-tier basket system were reared in the same tank and were exposed to the same environmental conditions, with the exception of light intensity. The abalone placed on the top tier that received more light grew significantly less than those placed on the middle and the bottom tiers. Hahn (1989) reported that abalone are easily stressed by light. Low growth rate of abalone in the top tier may also be attributed to the water flow passed from the bottom through airlift. In conclusion. Taiwan abalone fed on artificial diet performed better than those fed on Gracilaria tenuistipitata, displaying higher rates of weight gain and length increase, and lower FCR. The present study indicated that growth of the abalone placed on the top tier in a multiple-tier basket system was inferior to those placed on the other tiers. The present study demonstrated that Taiwan abalone H. diversicolor supertexta fed on artificial diet in a mul- tiple-tier basket system could grow to market size in less than half the time in comparison to those fed on Gracilaria. ACKNOWLEDGMENTS The authors would like to thank the Council of Agriculture, Republic of China, for support of this work (Grant No.: 88-Chung- Mei-l-4-Ho-01-3). We appreciate Mr. W. Z. Chen and C. C. Lin for their assistance in the culture. We also thank Mr. S. W. Liu, Mr. M. T. Huang, and Mr. T. S. Huang for providing the abalone juveniles and Gracilaria. LITERATURE CITED Bendschneider, K. & R. J. Robinson. 1952. A new spectrometric method for the determination of nitrite in the sea water. J. Mar. Res. 1 1:87-96. Britz. P. J. 1996a. Effect of dietary protein level on growth performance of South African abalone. Haliotis midae, fed fishmeal-based semi- purified diets. Aquaculture 140:55-61. Britz, P. J. 1996b. The suitability of selected protein sources for inclusion in formulated diets for the South African abalone. Haliotis midae. Aqua- culture 140:63-73. Capinpin, E. C. & K. G. Corre. 1996. Growth rate of the Philippines abalone, Haliotis asinina fed an artificial diet and macroalgae. Aqua- culture 144:81-89. Chen. H. C. 1984. Recent innovations in cultivation of edible mollusks in Taiwan, with special reference to the small abalone Haliotis diversi- color and the hard clam Meretrix lusoria. Aquaculture 39:1 1-27. Chen. H. C. & H. S. Yang. 1979. Artificial propagation of the abalone. Haliotis diversicolor supertexta. China Fish. Monthly 314:3-9. Fleming. A. E.. R. J. Van Barneveld & P. W. Hone. 1996. The develop- ment of artificial diets for abalone: A review and future directions. Aquaculture 140:5-53. Hahn, K. 1989. Nutrition and growth, pp. 135-180. In: K. Hahn (ed.). Handbook of Culture of Abalone and Other marine Gastropods. CRC Press, Boca Raton, FL, USA. Lin, T. S. 1986. Experiment on the utilization of fertilizer in adhesive stage larvae of abalone. Bull. Taiwan Fish. Res. Inst. 41:23-28. Knauer. J.. P. J. Britz & T. Hecht. 1996. Comparative growth performance and digestive enzymes activity of juvenile South African abalone. Hali- otis midae. fed on diatoms and a practical diet. Aquaculture 140:75-85. Koike. Y.. J. Flassch & J. Mazurier. 1979. Biological and ecological stud- ies on the propagation of the ormer Haliotis tuberculata Linnaeus. II. Influences of food and density on the growth of juveniles. La Mer 17:43—45. Mai. K.. J. P. Mercer & I. Donlon. 1995. Comparative studies on the nutrition of two species of abalone. Haliotis tuberculata L. and Haliotis discus hannai Ino. Optimum dietary protein level for growth. Aqua- culture 136:165-180. Mgaya. Y. D. & J. P. Mercer. 1995. The effects of size grading and slocking density on growth performance of juvenile abalone. Haliotis tuberculata Linnaeus. Aquaculture 136:297-312. Morrison. J. & J. Whittington. 1991. Development of an artificial feed for abalone. Australia Aquaculture 5:53-54. Nie. Z. Q.. 1992. A review of abalone culture in China, pp. 592-602. In: S. A. Shepherd. M. J. Tegner & S. A. Guzman Del Proo (eds.). Abalone of the World. Biology, Fisheries and Culture. Fishing News Books, Oxford, England. Nie, Z. Q.. P. Wang, Z. Q. Wang & J.-P. Yan. 1986. Experiments on preparing of formulated feed and feeding efficiency of young abalone. Haliotis discus hannai Ino. Mar. Fish. Res. 7:53-64. Ohba. T. 1964. Studies on the propagation of an abalone Haliotis diversi- color supertexta Lischke-I on the spawning habits. Bull. Jap. Soc. Sci. Fish. 30:742-747. Ogino. C. & E. Ohta. 1963. Studies on the nutrition of abalone -I. Feeding trials of abalone. Haliotis discus Reeve with artificial diets. Bull. Jap. Soc. Sci. Fish. 29:691-694. SAS. 1988. SAS/STAT user's guide 6.03 edition. SAS Institute Inc., Cary, North Carolina. Growth of Taiwan Abalone 635 Solorzano. L. 1969. Determination of ammonia in natural waters by the phenolhypochlorite method. Limnol. Oceanogr. 14:799-801. Stuart, M. D. & M. T. Brown. 1994. Growth and diet of cultivated black- footed abalone. Haliotis iris (Martyn). Aquaculture 127:329-337. Taiwan Fisheries Bureau. 1998. Fisheries yearbook Taiwan area, 1997. Taipei. Taiwan. Uki. N.. A. Kemuyama & T. Watanabe. 1985. Nutritional evaluation of several protein sources in diets for abalone Haliotis discus hannai. Bull. Jap. Soc. Sci. Fish. 51:1835-1837. Uki. N.. M. Sugiura & T. Watabane. 1986a. Dietary value of sea- weeds of occurring on the Pacific coast of Tohoku for growth of the abalone Haliotis discus hannai. Bull. Jap. Soc. Sci. Fish. 52:257- 266. Uki. N., A. Kemuyama & T. Watanabe. 1986b. Optimum protein level in diets for abalone. Bull. Jap. Soc. Fish. 52:1005-1012. Uki, N. & T. Watanabe. 1992. Review of the nutritional requirements of abalone [Haliotis spp.) and development of more efficient artificial diets, pp. 504-517. In: S. A. Shepherd. M. J. Tegner & S. A. Guzman Del Proo (eds.l. Abalone of the World. Biology. Fisheries and Culture. Fishing News Books, Oxford. England. Viana. M. T.. L. M. Lopez & A. Salas. 1993. Diet development for juvenile abalone Haliotis fulgens-Evdludlion of two artificial diets and macroal- gae. Aquaculture 1 17:149-156. Yang. H. S. & Y. Y. Ting. 1986. Artificial propagation and culture of abalone iHaliotis diversicolor supertexta Lischke). Bull. Taiwan. Fish. Res. Inst. 40:195-201. Journal of Shellfish Research, Vol. 18, No. 2, 637-644, 1999. MORPHOLOGICAL CHANGES IN THE RADULA OF ABALONE (HALIOTIS IRIS) DURING POST-LARVAL DEVELOPMENT RODNEY D. ROBERTS,1 2* TOMOHIKO KAWAMURA/ AND HIDEKI TAKAMI1 Cawthron Institute Private Bag 2 Nelson, New Zealand University of Otago P.O.Box 56 Dunedin, New Zealand 3Tohoku National Fisheries Research Institute Shinhama-cho, 3-27-5 Shiogama, Miyagi 985, Japan ABSTRACT H. iris Gmelin post-larvae fed with diatoms {Cylindrorheca closterium) were sampled weekly from 10-60 days post-settlement. Radula morphology was examined by SEM, and compared with that of competent larvae (260-280 p.m shell length (SU) and adults ( 125-130 mm SL). Larvae (130 "C.days old) had a well-formed radula with -10 rows of teeth. Each row comprised 1 rachidian tooth, 2 pairs of lateral teeth, and 1 pair of marginal teeth. The number of rows of teeth increased to 26-29 by -500 p.m SL (Day 10 post-settlement) and stayed at that level throughout the post-larval period (maximum size/age observed was 2.7 mm SL at Day 60). Marginal teeth were added steadily to give -30 pairs per row at -2.5 mm SL (compared to 60-80 in adults). Lateral teeth (L3-L5) were added between 1.0 and 1.7 mm SL to complete the adult complement. Progressive changes in the post-larval radula above 1-1.5 mm SL included: (1) reduction of serrations on the margins of rachidian and lateral teeth, particularly the more central (R. LI, L2) teeth; (2) increased spacing between adjacent rows of teeth; (3) lengthening of the outer lateral teeth (L3-L5) making them larger and taller than the central (R, LI. L2) teeth. The clearance angle of rachidian and lateral teeth was variable, but generally increased as post-larvae grew. Post-larvae <1 mm SL had highly curved teeth and apparent clearance angles around or below 0°. Larger post-larvae had positive clearance angles. These developments suggest that the teeth of post-larvae <1 mm SL probably function as "scoops" which slide across the surface collecting small diatoms and other fine, loose particles. Radulae of post-larvae > 1mm SL become more suitable for collecting larger particles, and gouging feeding substrata. KEY WORDS: benthic diatom; radula; development; feeding; growth; post-larval abalone INTRODUCTION Post-larval abalone have low survival in both hatcheries (e.g.. Searcey-Bernal et al. 1992. Roberts et al. 1998) and natural habi- tats (eg, Sasaki and Shepherd 1995. McShane and Naylor 1995. Preece et al. 1997). The nature and quantity of their food is likely to be a critical factor controlling survival. As post-larvae grow, their food consumption increases rapidly (Roberts et al. 1999) and there is evidence that post-larval survival rates can be density- dependent (McShane 1992. Preece et al. 1997). As post-larval abalone grow, their sources of nutrition change, both on artificial surfaces and on coralline algae (Kawamura et al. 1998a). Benthic diatoms are a principal food of post-larval abalone (Kawamura et al. 1998a). Post-larval growth and survival differ among diatom diets, due largely to differences in the post-larva's ability to ingest and digest the diatoms (Kawamura et al. 1998a,b, Roberts et al. 1999). Both ingestion and digestion appear to be affected by the radula. Physical rupturing of diatom cells probably relies solely on the radula (Kawamura et al. 1995) or other buccal apparatus, as the abalone gut lacks any grinding mechanism (Crofts 1929). The ability of post-larvae to ingest large (or stalked) diatoms increases as post larvae grow (Kawamura et al. 1998b, Roberts et al. 1999). This appears to be a function of food handling by the radula rather than mouth size (Roberts et al. 1999). Author to whom all correspondence should be sent. Tel: + 64 3 548 2319 E-mail: rodney@cawthron.org.nz Diatoms in the genus Cocconeis are particularly interesting in the context of abalone development. Cocconeis is often the dom- inant diatom on coralline algae (authors' unpubl. data) and on pre-grazed abalone settlement plates. However, the food value of Cocconeis spp. for abalone post-larvae is dependent on the stage of post-larval development (Kawamura et al. 1995, Matthews and Cook 1995. Daume et al. 1997, Roberts et al. 1999). Newly settled post-larvae are unable to effectively ingest Cocconeis cells, and growth curves may flatten out at 500-700 u.m shell length (SL) unless other sources of food are available (Kawamura and Takami 1995, Takami et al. 1997). However, once post-larvae reach -800 p,m SL they can efficiently ingest Cocconeis cells, and grow rap- idly. This transition in feeding efficiency on Cocconeis may be gradual (Roberts et al. 1999) or more sudden (Takami et al. 1997. Daume et al. 1997). perhaps depending on the specific character- istics of the culture. The characteristics of Cocconeis species that make them resis- tant to grazing are their very high attachment strength and their low profile. Grazing efficiency could be affected by either the detailed morphology of the feeding apparatus, or the force with which it is used. In post-larval abalone, the radula is the organ of key interest, and changes in feeding efficiency are likely to relate to radula morphology, rather than the strength of the radula action. If post-larval strength was inadequate, then radula strokes would falter. This was not seen in small post-larval Haliotis iris Gmelin. which grazed smoothly over Cocconeis scutellum Cleve (Roberts et al. 1999). 637 638 Roberts et al. The radula begins to develop during the larval phase (Tong and Moss 1992). It is used for feeding within a day or two of settle- ment, and is capable of collecting small diatoms and other particles at this time (Hahn 1989. Roberts et al. 1999). However. Kitting and Morse ( 1997) found that post-larval Haliotis rufescens Swain- son grazing on coralline algae did not ingest particles until 10 days after metamorphosis. This contrast suggests that feeding is strongly influenced by the interaction between available food, and the developing radula. There have been few observations of radulae in post-larvae of abalone or other molluscs. Several reports present a photograph and limited information on the abalone radula at a single point of larval/post-larval development (Tong 1984 - H. iris. Garland et al. 1985 - Haliotis ruber Leach. Dinamani and McRae 1986 - H. iris, Daume et al. 1997 - Haliotis rubra Leach, Kitting and Morse 1997 - H. rufescens). None of these studies describes the progres- sive development of the radula throughout the post-larval period. Studies of radula development in other larval/post-larval molluscs are limited to chitons, aplacophorans. opisthobranchs and pulmo- nates (Eernisse and Kerth 1988). In this paper we describe changes in radula morphology of H. iris throughout the post-larval stage, and compare the post-larval radulae with those of larvae and adults. Radula morphology is discussed in relation to changes in post-larval feeding. MATERIALS AND METHODS Abalone Rearing H iris larvae were obtained as described previously (Roberts et al. 1999). Competent larvae were transferred to tissue culture dishes (Falcon 3046) with 10 ml of 0.2 p.m filtered natural sea- water (FSW) containing 150 u.g/ml each of Penicillin G sodium (Bioehemie) and Streptomycin sulphate BP (Sigma). These larvae were induced to attach and metamorphose by the addition of 1 u.M GAB A (Roberts and Nicholson 1997). Two days after settlement induction, a diatom culture (Cylindrotheca closterium (Ehrenberg) Reimann and Lewin 1964) was added to the wells as a food source. These post-larvae were held at 17.5 ± 1.0 °C with supplementary diatom (C. closterium) added as necessary, and water replaced each 3—4 days with new FSW without antibiotics. We measured post-larval size as the longest shell dimension (shell length =SL) and post-larval age as days since settlement (settlement induction = Day 0). Observations of Radulae Samples of post-larvae were preserved in 49c formalin in sea- water approximately weekly from 10-60 days post-settlement. Shell length of each individual was recorded. A sample of larvae was taken from the rearing vessel after 8 days of development at 16 °C, and preserved as above. To obtain clean radulae, we dis- solved larval/post-larval tissues by bathing the abalone in sodium hypochlorite (0.6% final CI concentration, Wako Pure Chemical Industries Ltd.. Osaka) for several minutes. When free of sur- rounding tissues, the radula was removed with a pipette and trans- ferred through several distilled water baths before the radula length, width and number of rows of teeth were determined using a light microscope and video micrometer. Adult radulae were dis- sected from fresh animals and rinsed with distilled water. Radulae were then transferred to SEM stubs, laid flat with teeth upwards. and allowed to air dry before sputter coating with gold for SEM observation. The radula formulae, terminology and measurements used are illustrated in Figure 1. The length of the rachidian and L3 teeth was measured along the centre line of the cusp in plan view (Fig 1 A). Whole radula width was measured with marginal teeth in resting configuration (not splayed) and whole radula length was measured along the centre line of the radula (Fig. ID). Clearance angle of rachidian and lateral teeth, as defined by Padilla ( 1985), was mea- sured by viewing flat-lying radulae from the side and averaging the angle of 10-14 teeth. On printed photos, a line was drawn repre- senting the approximate slope along the midline of the back sur- face of the tooth. The angle between this line and the long axis of the radula was the clearance angle (Fig. IB. 1C). Imperfect view angles and highly curved teeth made precise measurements diffi- cult, so data should be treated as indicative rather than absolute. Rake angles could not be measured directly on most teeth due to their shape and orientation. However, rake angle can be calculated from clearance angle if the angle across the tooth tip is known (Rake angle = 90 - Tooth tip angle - Clearance angle) (Fig. IB). Accurate counts of marginal teeth were only possible in small '\r-^N, /-\ / \> I Serrations 7 1- inure 1. Illustration of terms used in the text to describe radula morphology. (A) Two transverse rows of radula teeth in plan view. Tooth types: R = rachidian. L1-L5 = lateral teeth 1-5. M = marginals (not all shown). Measurements: VVR = width of rachidian tooth. LR = length of rachidian tooth, LL3 = length of L3 tooth, G = gap between rachidian teeth of adjacent transverse rows. (B) A single radula tooth in profile illustrating: CA = clearance angle (positive). R4 = rake angle (both as defined by Padilla 1985), TTA = tooth tip angle. (C) A single tooth in slightly oblique profile showing tooth components (after Hick- man 198(1) and the negative clearance angle of a strongly curved tooth. (D) Whole radula laid flat showing 27 transverse rows of teeth pro- jecting upwards. \YR\V = whole radula width (marginals are in resting position). WRL = whole radula length. Radula Development in Post-Larval Abalone 639 post-larvae with few marginals, and in specimens with the mar- ginal teeth fortuitously splayed. RESULTS During the post-larval period the radula underwent various morphological developments toward the adult form. Most devel- opments correlated slightly more strongly with post-larval shell length than post-larval age (Table 1 ). so the following description relates primarily to shell length. The overall length of the radula increased linearly with post- larval shell length (Fig. 2A). This length increase was caused by different factors at different stages of post-larval development. The number of rows of radula teeth increased only during the first 10 days after settlement. The larval radula contained 9-10 rows of fully formed teeth, and an additional 2-3 rows of developing teeth. By Day 10 post-settlement, the number of rows of teeth had in- creased to 26-28. and remained between 26 and 29 throughout the post-larval period correlating very weakly with shell length (Fig. 2A). From 500-1700 p.m SL the increase in radula length reflected both an increased gap between rows of teeth, and increased length of the rachidian and L1-L2 teeth (Fig. 2B). Above 1700 u.m SL the gap between adjacent rows of teeth increased rapidly (Fig. 2B) as differentiation of the L3-L5 teeth developed (see below). Radula width also increased linearly with post-larval shell length (Fig. 3A) due to an increase in the number of teeth per row (Fig. 3B), and to a lesser extent, the width of individual teeth (Fig. 3A). The number of marginal teeth per row increased from just one pair in competent larvae to -30 pairs in post-larvae of 2000-2700 p.m SL (Fig. 3B). The adult radula contained -60-80 pairs of marginal teeth per row (data not presented). Lateral teeth were added relatively late in post-larval develop- ment. In the larval radula, lateral and marginal teeth were not clearly differentiated, but the radulae appeared to contain 2 pairs of lateral teeth and 1 pair of marginal teeth per row. All post-larvae less than 1 mm SL contained only the two pairs of lateral teeth (LI and L2) present in the larval radula. An additional 3 pairs of lateral teeth (L3-L5) were added progressively as post-larvae grew from -1.0 to -1.7 mm SL (Fig. 3B) completing the adult complement of laterals. TABLE 1. Correlation of radula variables with post-larval age and post-larval shell length. Post-larvae were chosen to include a range of sizes for each age. Data are Pearson's correlation coefficients. All correlations were significant {P < 0.01) except where marked (NS). Variable Post-larval age Post-larval shell length Overall length of radula Overall width of radula at rest No. of rows of teeth in radula Width of rachidian tooth Length of rachidian tooth Length of third lateral (L3) tooth Gap between adjacent rachidian teeth Clearance angle of rachidian/lateral teeth Number of lateral teeth per row Number of marginal teeth per row 0.878 0.967 55 0.866 0.942 48 0.260 (NS) 0.363 51 0.706 0.847 32 0.855 0.891 28 0.776 0.875 18 0.820 0.954 28 0.880 0.810 21 0.907 0.891 29 0.981 0.976 8 500- - A r2 = 0.1315. p<0.01 © © © © o © *xXc#d5 ©o ©y © © o » © * • 000 - ♦ Length © No. of rows • • *> ♦ 500 - * ♦ 1^ = 0.9342 p<0.001 0 - jn — 1 1 1 —I 1 30 20 « o -- 10 0 500 1000 1500 2000 2500 3000 30 - B 25 - • C Tooth length ♦ f- n 20 - 4 Gap ♦ ra ♦ O) o 15 - ♦ CT C ♦ ttJ 10 - » ^1 o © o© © © o o o H 5 - 0 - — i — i — > ♦ © oo o H h © -1 i 500 1000 1500 2000 2500 Post-larval shell length (urn) 3000 Figure 2. Relationship between post-larval size and factors relating to radula length: (A) whole radula length and number of rows of trans- verse teeth in radula: IB) length of the rachidian tooth and gap be- tween rachidian teeth in adjacent transverse rows. Morphology of Individual Radula Teeth The serrations on the working edges of the rachidian and lateral teeth became less pronounced as post-larvae grew. The rachidian and lateral teeth initially had long, pointed serrations along both lateral margins (Fig. 4A, B). Above 1 mm SL, these serrations became progressively shallower, first on the rachidian and LI teeth (Fig. 4C, D). and later on L2, and to some extent, L3 and L4 (Fig. 4E). By about 2 mm SL, post-larvae had lost nearly all serrations from R-L2 teeth, but retained them on L3. and particularly L4-L5. Marginal teeth remained finely serrated throughout the post-larval period. The adult radula lacked serrations on the rachidian and lateral teeth (Fig. 4F). but retained them on marginal teeth. Differentiation of the lateral teeth was evident in larger post- larvae. The rachidian and L1-L2 teeth were similar to one another throughout the post-larval period (but some differentiation was evident in the adult radula). The L3-L5 teeth were longer than R. LI, and L2 from the time they first appeared, but the size differ- ence became more marked in larger post-larvae, with rapid length- ening of the L3-L5 teeth (Fig. 5 for L3. data for L4 and L5 similar but not shown). The morphology of marginal teeth did not change markedly over the post-larval period. Marginal teeth were long, narrow and 640 Roberts et al. 250 200 -■ E — 150 I 100 50 -- T A ♦ Radula 8 » Rachidian O OO *' o ^ " O 0 -• • • r2 = 0.7181 -* *" p<0.001 0 o^ ^*^° ♦ -"^ o o o ♦ *■■* o • * ^ • ♦• : ~^T"* ■- «» «» *^* £^ * -£{** r2 = 0.8865 ■- y* pfcX ^ 4 »3 <*2 V J J fcbJ 5/im 500/nm Figure 4. Developmental stages of the post4ar\al radula. as seen by SEM. R = rachidian tooth, Ll-LS = lateral teeth. M = marginal teeth. Radula formulae represent numbers of M+L+R+L+M (Voltzow 1994). (A) Competent larva 270 urn SL after developing for 8 days at 16 C, radula formula = 1+2+R+2+1. (B) Post-larva of 530 urn SI. at Day 10 post-settlement, -4+2+R+2+4. (C) Post-larva of 1050 urn SL at Day 31, 12+3+R+3+-12. (D) Post-larva of 1.2 mm SL at Day 53. Note that the laterals have started to differentiate. -14+(2+2)+R+<2+2)+-14. (E) Post-larva of 2.7 mm SL at Day 60, with L3-L5 now much larger than R-L2, and the latter having lost their coarse serrations. (F) Adult of 130 mm shell length. Marginals are numerous and small (not all visible) and L3-L5 are specialised gouging teeth without serrations: -80+(3+2)+R+(2+3)+-80. strongly curved shape may make them effective "scoops" for small particles. Larger post-larvae had positive clearance angles (aver- ages of 15-35°) which would allow the tip of the tooth to dig in, meaning that rake angle becomes important. Tooth rake angles in post-larvae > 1 mm SL were around 40-70°, which is probably adequate to make them "cut" rather than "plough" (sensu Padilla 1985). Daume et al. (1997) reported a low clearance angle of -8° on the radula teeth of 53-day-old post-larval H. rubra. They made no mention of variation within/between radulae. Hardness and Flexibility of Radula Teeth In adult gastropods, newly formed radula teeth are initially soft, consisting of chitin and proteins (Runham 1961). The hardness of 642 JZ tz> - •> Rachidian ?n - ♦ Lateral 3 ♦ jr S ♦ 15- r2 = 0.7647 p<0.001 y / ♦ 10- 5 - s* * • — t> o o r2 = 0.7934 p<0.001 0- 1 1 1 — —I 1 1 Roberts et al A 0 500 3000 1000 1500 2000 2500 Post-larval shell length (pm) Figure 5. Relationship between post-larval size and the length of the rachidian and L3 teeth. Each data point represents the mean of 3-6 teeth on one radula. the teeth develops through precipitation of inorganic salts (Ca, Fe, P in the rhipidoglossa) via a complex biomineralisation process (Sollas 1907, Rinkevich 1993). There have been few observations of biomineralisation of post-larval radulae. Eernisse and Kerth (19881 observed dark-capped lateral teeth in even the most re- cently formed radulae of chitons. They analytically confirmed mineralisation of the teeth in juvenile chitons 16 days post- metamorphosis. No such dark caps are present in abalone, and there are no data on the mineralisation of post-larval abalone teeth. We guess that the teeth of young post-larval abalone would be relatively soft and flexible (especially those most recently formed) and that this would affect diatom ingestion, especially with respect to tightly attached and/or low-profile diatoms, such as Cocconeis spp. (see below). It would also limit the ability of the radula to penetrate firm substrata such as foliose macroalgae or crustose 50 40 30 i. 20-- D) c CO CD 10 o c CO o -10 -20 \* -+- ■+■ -+- 500 1000 1500 2000 Post-larval shell length (pm) 2500 3000 Figure 6. Relationship between clearance angle of teeth (see Fig. 1) and post-larval shell length. Each point represents the mean of 10-14 teeth on one radula. ■MflBHBMHb Figure 7. SEM photographs of the post-larval radula showing: (A) Strongly curved cusps on radula teeth, producing clearance angles around 0° or less. These were typical in small post-larvae (<1 mm SL). (B) Straight-cusped radula teeth with positive clearance angles - typi- cal of larger post-larvae (>1.5 mm SL). corallines. While the radula teeth are still soft and flexible, it would make sense for them to act as scoops rather than excavators. Grazing on Cocconeis spp. Ingestion of Cocconeis spp. by post-larval abalone is of par- ticular interest in the context of hatcheries and natural recruitment (see Introduction). In feeding and growth experiments conducted in parallel with this study, the post-larvae became progressively more efficient at ingesting Cocconeis scutellum between shell lengths of 0.5 and 1.2 mm (Roberts et al. 1999). Faeces of 0.5 mm post-larvae contained few C. scutellum cells, and post-larvae were observed grazing rapidly and smoothly over C. scutellum cells without removing or rupturing them. By 0.8 mm SL. post-larval grazing was efficiently removing C. scutellum. An increase in clearance angle would result in a gouging, rather than sliding, tooth (Padilla 1985) and could explain the ingestion pattern seen on C. scutellum. Our data suggested that clearance angle does increase (and become positive) during early post-larval develop- ment, but the change was relatively small in post-larvae less than 1 mm SL. and variability was high. Furthermore, the data are based on radulae removed from post-larvae and prepared for SEM, so any discussion of the feeding implications of increasing clearance angle is conjectural. The hardness and flexibility of the radula teeth (see above) may also explain the change in ingestion of Cocconeis cells. Radula Development in Post-Larval Abalone 643 Changing Radula Function and Diet Based on the observed changes in radula morphology, we can speculate about changes in food sources for post-larval H. iris. The radula teeth of post-larvae <1 mm SL probably function as scoops, which slide across the feeding substratum and collect small and loosely attached particles. On hatchery surfaces these will be pre- dominantly small to medium sized naviculoid diatoms and other biofilm components such as bacteria and extracellular secretions. On coralline algae in natural habitats the food particles may in- clude biofilm components (e.g., bacteria, loosely attached diatoms and extracellular secretions) or loose epithallial cells. Above 1 mm SL the apparent clearance and rake angles mean that the radula teeth may begin to penetrate at least soft foods, rather than just sliding across the substratum. The reduction of tooth serrations may indicate less emphasis on very small food items, and the increasing spacing of tooth rows may render the radula more capable of handling larger food particles. The L3-L5 teeth appear to become specialised "gouging"" teeth, and the in- creased number of marginals improves the particle collecting ac- tion of the radula. These changes are consistent with observations that larger abalone post-larvae consume even large or tightly at- tached diatoms (e.g.. Roberts et al. 1999). and can ingest delicate red seaweeds (Crofts 1937) and the surface cuticle/cells of coral- line algae (Garland et al. 1985. Kitting and Morse 1997). Steneck and Watling (1982) regarded rhipidoglossan radulae as "brooms'" with very limited excavating force. They stated that no rhipidoglossans feed on calcareous algae (which includes the cor- allines). However, Shepherd and Cannon ( 1988) found a predomi- nance of coralline algal fragments in the guts of 5-10 mm Haliotis laevigata Donovan and H. scalaris Leach. Garland et al. (1985) showed that even 6-13 week old H. rubra (average of 700-2400 p.m SL respectively) can ingest the cuticle and some cytoplasm from coralline algae. Thus, abalone are capable of ingesting at least the outer layers of coralline algae from an early age. but determination of the relative importance and food value of coral- line algae requires quantitative studies of feeding and food utili- sation. Many of the structural developments of the abalone radula occurred within 60 days post-settlement but the radula of 2.5 mm juveniles was still quite different from that of the adult. The adult H. iris radula has the same number of rachidian and lateral teeth, but at least twice as many marginal teeth per row. The adult's L3-L5 teeth lack serrations and are pointed blocks (apparently specialised for gouging macroalgae) (Fig 4F) and the L2 teeth were differentiated from the R and LI teeth. The sequence of radula development we observed was for post- larvae feeding on a film of loosely attached diatoms at 17 °C. Development correlated more strongly with post-larval size than post-larval age (Table 1). so we would expect slower radula de- velopment with lower temperatures and poor food sources. Graz- ing resistant food sources (such as coralline algae) may accelerate tooth wear and tooth loss, resulting in differences in radula length, number of rows of teeth, and tooth serrations. Ontogeny of Radula Development Previous work on radula development in larval and post-larval molluscs was reviewed by Eernisse and Kerth (1988). In all gas- tropods studied (two opisthobranch species and seven Pulmonate families) lateral teeth developed before a median (=rachidian) tooth was added. The same pattern was observed for four species of chiton (Eernisse and Kerth 1988). In H. iris, by contrast, the rachidian tooth was present from the time that the radula first became visible in the larva. Eernisse and Kerth ( 1988) found no evidence of radula forma- tion in the larvae of four chiton species, but they state that Min- ichev and Sirenko (1974) reported primordial radulae in unidenti- fied chiton trochophores. In contrast, the radula of abalone is well developed by the time larvae are competent to settle (Fig. 4A of this paper. Moss and Tong 1992. Roberts and Nicholson 1997, Salas-Garza et al. 1994). If metamorphosis is delayed. H. iris larvae continue to rapidly add rows of teeth to the radula reaching -15 rows after 15 days at 16 °C, and up to 30 or more rows after -30 days (R. Roberts, unpublished). The development of the radula in abalone larvae provides a functional food-collecting or- gan within a day or two of settlement (Hahn 1989. Kitting and Morse 1997, Roberts et al. 1999). In H. iris, additional lateral teeth were added "outside" ( = lat- eral to) the existing lateral teeth, but "inside" ( = central to) the marginal teeth. This pattern is consistent with other gastropods, but differs from chitons, which add new lateral teeth between existing laterals during post-larval development (Eernisse and Kerth 1988). When additional pairs of lateral teeth are forming in H. iris, they are similar to marginal teeth in their drooping, comb-like form (Fig. 2D). This, and their place of formation, suggest that new lateral teeth may be added by progressive differentiation of the inner-most marginal teeth. ACKNOWLEDGMENTS We thank Henry Kaspar. Yoh Yamashita. and Christine Han- dley for constructive review of drafts. Larval abalone were sup- plied in part by Island Hatcheries Ltd. This research was supported by contract CAW 801 with the New Zealand Foundation for Re- search Science and Technology, the Asia 2000 Foundation of New Zealand, the New Zealand Ministry for Research Science and Technology, an Alliance Group post-graduate fellowship through the University of Otago, and the Japan Fisheries Agency. II ERATURE CITED Crofts. D.R. 1929. Haliotis. L.M.B.C Memoir XXIX. 174 pp. Crofts, D. R. 1937. The development of Haliotis tuberculata, with special reference to organogenesis during torsion. Roy. Soc. Land. Philos. Trans.. Ser. B. 228:219-268. Daume. S.. S. Brand & W. J. Woelkerling 1997. Effects of post-larval abalone (Haliotis rubra) grazing on the epiphytic diatom assemblage of coralline red algal surfaces. Moll. Res. 18:119-130. Dinamani, M. & C. McRae 1986. Paua settlement: the prelude. Shellfish- eries Newsletter 30 in Catch 13(3):9. Eernisse. D. J. & K. Kerth 1988. The initial stages of radular development in chitons (Mollusca: Polyplacophora). Malacologia 28:95-103. Garland. C. D.. S. L. Cooke. J. F. Grant & T. A. McMeekin 1985. Ingestion of the bacteria on and the cuticle of crustose (non-articulated) coralline algae by post-larval and juvenile abalone [Haliotis ruber Leach) from Tasmanian waters. J. Exp. Mar. Biol. Ecol. 91: 137-149. Hahn. K. O. 1989. Induction of settlement in competent abalone larvae, pp. 101-112 //; K.O. Hahn Handbook of Culture of Abalone and Other Marine Gastropods. CRC Press. Florida. 348 pp. 644 Roberts et al. Hickman. C. S. 1980. Gastropod radulae and the assessment of form in evolutionary paleontology. Paleobiology 6:276-294. Kawamura. T. & H. Takami 1995. Analysis of feeding and growth rate of newly metamorphosed abalone Haliotis discus hamuli fed on four spe- cies of benthic diatom. Fisheries Sci. 61:357-358. Kawamura. T.. T. Saido. H. Takami & Y. Yamashita 1995. Dietary value of benthic diatoms for the growth of post-larval abalone Haliotis discus haimai. J. Exp. Mar. Biol. Ecol. 194:189-199. Kawamura. T.. R. D. Roberts & H. Takami 1998a. A review of the feeding and growth of post-larval abalone. J. Shellfish Res. 17:615-625. Kawamura, T.. R. D. Roberts & C. Nicholson 1998b. Factors affecting the food value of diatom strains for post-larval abalone Haliotis iris. Aqua- culture 160:81-88. Kitting, C. L. & D. E. Morse 1997. Feeding effects of post-larval red abalone. Haliotis rufescens (Mollusca: Gastropoda) on encrusting cor- alline algae. Moll. Res. 18:183-196. Matthews, I. & P. A. Cook 1995. Diatom diet of abalone post-larvae {Haliotis midae) and the effect of pre-grazing the diatom overstorey. Mar. Freshwat. Res. 46:545-548. Minichev, Y. S. & B. I. Sirenko 1974. Development and evolution of radula in Polyplacophora. Zoologicheskii Zhurnal 53:1133-1 139. [ In Russian.] Moss, G. A. & L. J. Tong 1992. Effect of stage of larval development on the settlement of the abalone Haliotis iris. N. Z. J. Mar. Freshwat. Res. 26:69-73. McShane, P. E. 1992. Density-dependent mortality of recruits of the aba- lone Haliotis rubra (Mollusca : Gastropoda). Mar. Biol. 1 10:381-389. McShane, P. E. & J. R. Naylor 1995. Depth can affect post-settlement survival of Haliotis iris (Mollusca: Gastropoda). J. Exp. Mar. Biol. Ecol. 187:1-12. Padilla. D. K. 1985. The structural resistance of algae to herbivores: a biomechanical approach. Mar. Biol. 90:103-109. Preece. P., S. A. Shepherd. S. M. Clarke & J. K. Keesing 1997. Abalone stock enhancement by larval reseeding: effect of larval density on settlement and survival. Moll. Res. 18:265-274. Rinkevich. B. 1993. Major primary stages of biomineralisation in radula teeth of the limpet Lottia gigantea. Mar. Biol. 1 17:269-277. Roberts, R. D. & C. M. Nicholson 1997. Variable response from abalone larvae (Haliotis iris, H. virginea) to a range of settlement cues. Moll. Res. 18:131-141. Roberts. R. D.. R. Searcy Bemal. & P. A. Cook 1998. A workshop on the culture of larval and post-larval abalone. Third International Abalone Symposium. Cawthron Rep. 477. 5 pp. Roberts. R. D.. T. Kawamura. & C. M. Nicholson 1999. Growth and survival of post-larval abalone (Haliotis iris) in relation to their devel- opment and diatom diet. /. Shellfish Res. 18:243-250. Runham, N. W. 1961. The histochemistry of the radula of Patella vulgata. Q. Jl. microsc. Sci. 102:371-380. Salaz-Garza, A„ E. Carpizo-Ituarte, L. S. McAnally-Salas & J. A. Zertuche 1994. Metamorphosis in Haliotis rufescens abalone larvae of different ages, related to the number of rows of teeth in the radula. Abstracts of the Second International Symposium on Abalone Biology Culture and Fisheries. Hobart. Sasaki, R. & S. A. Shepherd 1995. Larval dispersal and recruitment of Haliotis discus hannai and Tegula spp. on Miyagi coasts, Japan. Mar. Freshwat. Res. 46:519-529. Searcy Bernal. R., A. E. Salas-Garza. & R. A. Flores-Aguilar 1992. Re- search in Mexico on the critical stage of abalone (Haliotis spp.) seed production, pp. 547-560 In S.A. Shepherd. M.J. Tegner, and S.A. Guz- man del Proo (eds). Abalone of the World: Biology, Fisheries and Culture. Fishing News Books, Oxford. Shepherd, S. A. & J. Cannon 1988. Studies on southern Australian abalone (genus Haliotis) X. Food and feeding of juveniles. J. Malucol. Soc. of Aust. 9:21-26. Sollas, I. B. J. 1907. The molluscan radula: its chemical composition and some points in its development. Q. Jl. microsc. Sci. 51:115-136. Steneck. R. S. & L. Watling 1982. Feeding capabilities and limitation of herbivorous molluscs: a functional group approach. Mar. Biol. 68:299- 319. Takami, H., T. Kawamura & Y. Yamashita 1997. Survival and growth rates of post-larval abalone Haliotis discus hannai fed conspecific trail mucus and/or benthic diatom Cocconeis scutellum var. parva. Aqua- culture 152:129-138. Tong. L. 1984. Paua larvae put under the microscope. Shellfisheries News- letter 24 in Catch 11(11):7. Tong, L. J. & G. A. Moss 1992. The New Zealand culture system for abalone. pp. 583-591 In S. A. Shepherd. M. J. Tegner, and S.A. Guz- man del Proo (eds). Abalone of the World: Biology. Fisheries and Culture. Fishing News Books, Oxford. Voltzow, J. 1994. Gastropoda: Prosobranchia. pp. 111-252 In F.W. Har- rison and A.J. Kohn (eds) Microscopic Anatomy of Invertebrates Vol- ume 5. Mollusca I. Wiley-Liss Inc. New York. Journal of Shellfish Research, Vol. 18. No. 2. 645-650, 1999. THE EXTRACELLULAR MINERAL CONCRETIONS IN ANODONTA CYGNEA (L.): DIFFERENT TYPES AND MANGANESE EXPOSURE-CAUSED CHANGES GABRIELA MOURA,1 RICARDO GUEDES,2 AND JORGE MACHADO1 ' Laboratorio de Fisiologia Aplicada, Instituto de Ciencias Biomedicas Abel Salazar, Lg. Prof. Abel Salazar. 2. 4099-003 Porto. Portugal. 2Centro de Materials da Universidade do Porto (CEMUP), 4150 Porto. Portugal 3Centro de Investigaqao Maritima e Ambiental (CIMAR). 4050 Porto. Portugal ABSTRACT The mantle and gill extracellular mineral concretions of normal unpolluted and manganese-incubated specimens of the freshwater mussel. Anodonta cygnea (L.), were studied for their elementary composition and morphology. The mineral concretions of the gills, as well as those analyzed on the haemolymph side of the intrapallial epithelium, revealed a phosphate nature, being related to manganese accumulation. On the contrary, mineral formations located on the haemolymph side of the extrapallial epithelium are mainly built of carbonate salts, probably being more related to the shell biomineralization process than to the toxic metal detoxification. The concretions also differ markedly in their morphological aspects, with the phosphate formations being spherical, and the carbonate formations quite irregularly shaped. KEY WORDS: biomineralization, detoxification, mussel, gill, mantle, concretion, manganese INTRODUCTION Many invertebrate phyla are known to produce calcium con- cretions in one or more of their tissues (George 1982). The fresh- water bivalve. Anodonta cygnea, has been reported to use this insoluble store of calcium for shell deposition (Istin and Girard 1970a). The extracellular concretions of Anodonta and other related species have been described in the mantle (Pekkarinen and Valov- irta 1997) and gill (Silverman et al. 1983. Silverman et al. 1985, Silverman et al. 1987a). The composition and structure of these concretions has been studied in Ligumia subrostrata (Silverman et al. 1983), Margaritifera margaritifera (Pekkarinen and Valovirta 1997), and Anodonta sp. (Pynnonen et al. 1987). through histo- chemistry, electron microscopy, and chemical analysis. We now report the results of an annual monitoring of these extracellular concretions in normal specimens of the freshwater mussel. Anodonta cygnea, with respect to their morphology and mineral composition. Both concretions from the mantle and gill epithelia were examined, as well as the alteration of their normal characteristics after manganese exposure, to establish a connection between these microstructures and their possible physiological roles, namely those related to shell production and heavy metal detoxification. MATERIALS AND METHODS Lake mussels, belonging to the species Anodonta cygnea (L.). were collected through dragging from the clay bottom of the Mira lagoon (Aveiro, Portugal). The length of the animals treated varied from 12.0 to 18.0 cm, with a mean of 15.0 cm. Six specimens were collected each month, were transported to the Institute laboratory facilities in iceboxes containing natural pond water; kept in the laboratory in aerated pond water without feeding; and processed within 24 h of collection. The animals were considered healthy if they showed active ventilation, powerful valve closing or water ejection upon disturbance, and if their nacre presented a smooth and shiny look. For the manganese-exposure assay, 12 specimens of about 200-g body weight, collected in the same way during March, were randomly assigned to two treatment categories, control and heavy metal exposure, and kept in plastic tanks with naturally collected water at a temperature similar to natural conditions. The tanks were placed in relative darkness and the water was aerated and used without any pretreatment. For the heavy metal treatment, manganese, as MnCl2, was added to the tank water to reach a final concentration of around 10-6 M (between 1 and 2 ppm) from a stock solution previously prepared with distilled water. Manganese was the heavy metal chosen because of the well-known relationship with the mineral structures (Silverman 1989). The exposure period lasted for 16 h, during which the mussels showed normal ventilation. After expo- sure, both manganese-exposed and control animals were subjected to the same treatment as the specimens collected each month. Gill and mantle fragments were extracted by cutting the mussel adductor muscles, opening the shell, and cutting the middle region of the tissue free from the body. After dissection, small (3-4 mm) mantle pieces were separated into their two composing epithelia. i.e., the one facing the pallial cavity (intrapallial epithelium) and the other facing the shell side (extrapallial epithelium). These frag- ments were then placed in aluminum cylinders with their haemo- lymph-bathed side up (showing the basal face of both epithelia). and left to dry at room temperature. The composition of concretions from both mantle epithelia, on the haemolymph side, was analyzed on the assembled material by energy dispersive X-ray spectroscopy (EDS), with a spectrum col- lection time of 60 seconds. The elements chosen for analysis were Ca, P, C, CI, Na, Mg, and Mn. from pilot determinations. Gill tissue fragments, on their haemolymph-bathed side and equally assembled, were also analyzed with this technique, to establish 645 646 MOURA ET AL. eventual differences between the concretion populations from both organs. The morphology of the different mineral concretions, de- termined by their elemental composition, was then studied by scanning electron microscopy (SEM). also using the dried tissue samples without any other treatment. RESULTS Results of the mineral concretion composition reveal no major variations, either between individuals or throughout the year, the only significant differences being those between tissue-type ana- lyzed, and those exposed to manganese. Therefore, comments will focus on these elements, using the EDS records of each group of animals (i.e.. normal and metal-exposed) and tissues (i.e., gill, extrapallial and intrapallial epithelia). Using the March records, which also correspond to the control group for the contamination study, we find high amounts of phos- phate-like concretions on both the gill (see Fig. 1) and the intra- pallial epithelium, on its haemolymph side (Fig. 2). The major phosphate composition of these mineral formations is demon- strated by their higher content of phosphorus, compared to car- bon— which is probably more related to the surrounding tissue present — as shown by the record of the intrapallial epithelium itself (Fig. 4). In these phosphate concretions, and throughout the whole year, there seems to be a good correlation with the manga- nese level, suggesting a constant relationship between them. On the other hand, at the haemolymph side of the shell-facing extrapallial epithelium, the mineral concretions reveal a carbonate nature (Fig. 3), although with some phosphate, which is usually rather low (residual). This second group of concretions does not show the same tight relationship with the manganese composition, with a lower level being the general rule. The two kinds of mineral concretions, now detected also show morphological differences (Fig. 5). with the carbonate concretions (3) being usually larger and more irregular. The phosphate forma- tions (1,2) are present in higher amounts and show a regular spherulitic morphology, although with important size variation. They are also tightly covered by organic material presenting a net-like structure, in which they seem to be formed. Both concre- tion types are well attached to the respective epithelium, always giving the same distribution results when both tissues are separated from each other (see Fig. 6 for better a understanding of the overall arrangement). These general features were not significantly changed in the manganese-incubated specimens. These results show that the man- ganese is preferentially bound by the phosphate-bearing granules, being increased only in those formations in response to the con- tamination situation, compared to the normal control animals. The association between manganese and phosphate minerals is also supported by our observations on the carbonate-composed concre- tions, where higher amounts of this metal were associated with the samples with the highest level of residual phosphate. DISCUSSION Anodonla cygnea uses calcium concretions as calcium stores for shell growth (Istin and Girard 1970b), associated with anhy- drase activity as a regulatory feature (Istin and Girard 1970a). Concretions are reported both in the gills (Silverman et al. 1983, Silverman et al. 1985, Silverman et al., 1987a) and in the mantle (Pekkarinen and Valovirta 1997) connective tissues of this and other related (i.e.. freshwater) species. NORMAL ANIMALS 650 0 600- SSO 500- P 450 Ca 1 C 400- ■ (I « aso t * 300- C 2S0 ZOO- ISO 100- Ca ,.J ' y^ WJ VL-^Jl . "^ 1 2 3 - S 6 7 Energy (keV) MANGANESE-EXPOSED ANIMALS Ca too p 600 700 0 c 600 t 1 400 300 200 C* too "L^ s A Hn o-l- 1 2 3 « S 6 7 Energy (keV) Figure 1. EDS records from analyses of mineral concretions in the gills of Anodonla cygnea specimens collected in March, from the natu- ral environment (above) or after manganese-contamination (below). In the unionid, the calcium of gill concretions is bound to inorganic or organic phosphate (Silverman et al. 1983. Silverman et al. 1987b: Pynnonen et al. 1987. Lautie et al. 1988) and is associated with an organic matrix (Silverman et al. 1983; Silver- man et al. 1987b). According to Silverman et al. ( 1983), 25% of the concretion weight is organic, calcium makes up 25%, and phosphate represents 36%-39%. These extracellular concretions are composed of phosphorus, calcium, manganese, and iron, with smaller amounts of Mg. Al. S. CI, Zn, and Ba (Lautie et al. 1988). No carbonate can be detected in these concretions by Raman analysis (Lautie et al. 1988). These features resemble those reported here for the gill and intrapallial epithelium, except for the trace element composi- tion, which were not studied in the present work. The phosphate concretions we studied are probably identical to the ones report- ed by Lautie et al. (1988) and, therefore, the carbon amount de- Extracellular Mineral Concretions in Bivalves 647 NORMAL animals 1900 1*00 1700 isoo 1300 1400 1300 1200 1100 1000- 900 800 700 too 500 400 300 ZOO 100 Cm U^^wJ C* IA_ Kn S 6 Energy (keV) MANGANESE-EXPOSED ANIMALS Cm Cj VU Hn Energy (keV) Figure 2. EDS records from analyses of mineral concretions on the intrapallial epithelium of Anodonta cygnea specimens collected in March, from the natural environment (above) or after manganese- contamination (below). NORMAL ANIMALS £200 Cs 0 2000- c 1000 1 itoo MM n l20° 1 t X 1000- 000 600 400 Iaw^^^ c« 200 li . u 1 2 3 « 5 6 Energy (keV) 7 MANGANESE-EXPOSED ANIMALS 1400 C» 1300 1200 1100 C 1000- 900 c e u MO l 700 a 0 600 soo 400 300 200 p A? i» 100 vUmMA^J IL ». 12 3 4 5 6 Energy (keV) 7 Figure 3. EDS records from the analyses of mineral concretions on the extrapallial epithelium of Anodonta cygnea specimens collected in March, from the natural environment (above) or after manganese- contamination (below). tected could be related to the organic material present. The cal- cium/phosphate proportions are. in fact, parallel to the data now presented. The phosphate spherules react positively to PAS following amylase reaction, suggesting the presence of polysaccharides (Sil- verman et al. 1983). However, although being partially composed of polysaccharides, the organic core of the concretions described by these authors does not appear to have sulfate groups, in contrast to the results of Davis et al. (1982), who found sulfur with X-ray microanalysis. The negative reaction for the carboxyl and sulfhydryl radicals of the organic matter from concretions has been explained by the tightly bound metal amounts, which can keep the active groups from reacting (Lautie et al. 1988). Alternatively, these contradic- tory results can be explained by the presence of two kinds of mineral concretion differently located, and separated from each other, one having sulfated organic material and the other being poorer in that respect. This is, in fact, what we found in both epithelia of Anodonta cygnea mantle. The extracellular spherocrystals of the mantle are usually shown to be composed of calcium carbonate (Istin and Girard 1970b). a conclusion taken after studying the action of the pH. CO, partial pressure, and carbonic anhydrase activity in the cal- cium movements, and the location of Ca and this enzyme in the concretions. Alternatively. Pekkarinen and Valovirta (1997) and others reported the presence, in this same tissue, of calcium phos- phate concretions in both Anodonta and Margaritifera, similar to those found in the gills of the same animals. The present work demonstrates, for the first time to our knowledge, that both types of concretion composition are present simultaneously in the mantle 648 MOURA ET AL. NORMAL ANIMALS 1608 1500 1400 1300 1200 1100 C 1000- u 900 n 1- 000 I 700 600 500 400 d c. Mn Energy (keV) Figure 4. EDS records from the analyses of the intrapallial epithelium from normal Anodonta cygnea specimens collected in March. Figure 6. Schematic representation showing the localization of the different pallial concretion types of Anodonta cygnea. Ill external me- dium; (2) intrapallial epithelium; (3) haemolymph compartment; (4) extrapallial epithelium; (5) extrapallial compartment; (6) shell; CF = spherical calcium phosphate concretions; CC = irregularly shaped cal- cium carbonate concretions. of Anodonta cygnea, which explains these contradicting results, and confirms the hypothesis proposed by Pekkarinen and Valovirta (1997). On the other hand, lamellar concretions have been described singly or clustered (Silverman et al. 1983), both in muscle (Kapur and Gibson, 1968) and mantle (Davis et al. 1982) tissues; they are related to the biomineralization process of the shell, because they disappear during shell regeneration (Watabe et al. 1976). The morphological characteristics reported for mantle concre- tions can also be explained by the presence of both kinds of min- eral formations. We found spherulitic concretions together with more lamellar-shaped formations, both described in an isolated fashion by these authors. Perhaps each different kind show, in fact, some variation in their amounts inside the mantle (as those re- Figure 5. SEM photographs showing the morphology of Anodonta cygnea mineral concretions. ( 1 1 Spherical concretions, located on the surface of the intrapallial epithelium, on its haemolymph side. Bar 10 urn. (2) Higher magnification of the structures seen in (1). Bar 1 urn. (3) Irregular-shaped concretions, located on the surface of the extra- pallial epithelium, on its haemolymph side. Bar 10 urn. ported for the shell regenerating specimens, with respect to the lamellar formations), which can explain why both types were not found by each worker mentioned. Such a fluctuation is not de- tected in our results, where, although with some variations, both types coexist throughout the year. The hypothetical functions of these mineral concretions have been studied by several authors (e.g., Istin and Girard 1970a). Concretions may be involved with calcium storage for shell re- generation or normal growth (Vaidya and Nagabhushanam 1980). pH buffering during metabolic and/or respiratory acidosis (Istin and Girard 1970b). and detoxification of heavy metals (Mason and Simkiss 1982). During acidosis, bivalves dissolve their calcium reserves to buffer the pH (as reviewed by Burton 1983), causing a significant increase of the circulating levels of calcium (e.g., Pynnonen, 1990a, Byrne and McMahon 1991, Pekkarinen and Suoranta, 1995), carbonates and Pco, (Burton 1983, Pynnonen 1994). while the internal pH stays only slightly decreased (Collip 1921 ). The excessive calcium arising from this situation, on the other hand, can partially leak to the external mantle cavity (Pekkarinen and Suoranta 1995) and be lost to the environment, or be again absorbed after the acidic stress period. The calcium remaining in the biological fluids can be incorporated into the shell (Pekkarinen and Suoranta 1995). in the same way as the one seasonally liber- ated by the succinic acid formed in the summer (Machado et al. 1990). The reported high permeability of the mantle to calcium ions (Coimbra et al. 1988) will allow this excess calcium to be rapidly transported to the mineralizing front, causing the fast formation ot a calcified pellicle on the inner side of the shell (Machado et al. 1988). A different model is proposed by other authors, in which the pH is buffered with the dissolution of calcium carbonate from the shell itself (Sorokina and Zelenskaya 1967). However, their stud- ies dealt mainly with marine species that present much smaller amounts of calcium microspherules (Machado et al. 1988). In ad- dition, freshwater organisms do not have an abundant supply of Extracellular Mineral Concretions in Bivalves 649 calcium and. therefore, try to avoid any loss. During a metabolic hypoxia, for example, the animal needs a mechanism to retain the calcium liberated from dissolved minerals. Accordingly, Machado et al. (1988) propose significant differences between marine and freshwater mollusks with respect to the origin of calcium to be dissolved for pH buffering purposes. The calcium carbonate of the shell would be more susceptible to leaching by acidic metabolites than the calcium phosphate of the concretions (Howard et al. 1981), because it is more soluble. This agrees with the fact that these concretions do not seem to be implicated with pH buffering during anoxia in these animals (Sil- verman et al. 1983), thus different bivalves may have different calcium reserves to be used in this mechanism (Pynnonen et al. 1987). Keeping Anodonta cygnea in acidic water leads to further shell growth, demonstrated by the formation of a calcified new pellicle on the inner surface of the shell (Machado et al. 1988). Therefore, it was suggested that calcium may come from the phosphate con- cretions (Pekkarinen and Suoranta, 1995). at least in this species. In other works, however, calcium concretions were not mobi- lized under the most severe hypoxia or acidic conditions (Silver- man et al. 1983, Pynnonen, 1990a. 1990b). On the contrary, the increase of calcium in the blood that follows hypoxia was accom- panied by an increase of the concretions level, showing that the concretions in the gill of these mussels were gaining calcium from the circulating fluid during a time of increased blood calcium, which would serve to avoid calcium loss to the environment fol- lowing its liberation from the shell, as a result of hypoxia (Silver- man et al. 1983). Radiolabeled studies also showed that the source of calcium for the elevated haemolymph levels seems to be the shell (Crenshaw and Neff. 1969). In the freshwater bivalve. Ligumia subrostrata, the calcium released from the shell during hypoxia (McMahon. 1979; Silverman et al. 1983) is deposited into calcium phosphate concretions in the gill tissue (Silverman et al. 1983). Finally, these contradicting results lead to the hypothesis that a fraction of the calcium stored in the concretions is associated with carbonate, being more easily solubilized than calcium phosphate (Pekkarinen and Valovirta, 1997). All previous results, together with the present observations of gill and mantle mineral concretions in Anodonta cygnea, can be explained if we consider this dual nature of the mantle mineral formations. In fact, the phosphate concretions are most probably related to the other possible functions usually assigned to these structures, namely the metal sequestering and detoxification pro- cesses, as suggested by the higher incorporation of manganese reported here and in other works (Silverman et al. 1987b). This detoxification mechanism seems to be more important in freshwa- ter species, because this environment is subjected to higher heavy metal fluctuations, according to Mersch et al. (1996). As explained by this author, this type of concretion probably evolved as a form of retaining the calcium that would otherwise be lost during aci- dosis. Its lower solubility and higher calcium binding ability (Sil- verman et al. 1987b) allowed for its later use as a detoxifying agent, in view of the higher binding of metal, as calcium analogues (Silverman et al. 1987b). The phosphate granules are also closely related to the reproduction cycle, being rapidly mobilized for the glochidia shell formation (Silverman et al. 1985), in what seems to constitute its primer function (Silverman et al. 1985; Silverman et al. 1987b), an idea also confirmed by the fact that these are the only concretions found in the gills of unionid bivalves, their mar- supial organs. On the other hand, there are the carbonate lamellar granules, formed only by the extrapallial mantle epithelium, directly related to the inner shell growth and bearing some resemblance in struc- ture and composition. These formations can, in turn, be related to the deposition cycling of the nacre formations, a situation sug- gested by Mason and Simkiss (1982). As calcium carbonate enti- ties, these formations are more readily dissolved in response to metabolic or respiratory acidosis than the phosphate ones. Under the normal seasonal internal pH decrease postulated by Machado et al. (1988), these concretions would dissolve before the shell itself, in view of their particular spatial localization, partially spar- ing the nacre from this dissolution event. The increased calcium, following this dissolution, would then cross the extrapallial fluid, fulfilling the mechanism proposed by Machado et al. (1988). This, however, in view of the small amounts of carbonate con- cretions, would be a limited event, particularly if the acidosis situations, like those experimentally caused, were strong and pro- longed. Under these conditions, the shell would be the next can- didate for the acid action, being dissolved as commonly reported. ACKNOWLEDGMENTS The authors wish to thank to Graca Casal for her help with editing the photographs. This work was supported by a JNICT grant (Junta Nacional de Investigacao Cientffica e Technologica. Portugal). LITERATURE CITED Burton. R. F. 1983. Ionic regulation and water balance, pp. 291-352. In: A. S. M. Saleuddin & K. M. Wilbur (eds.). The Mollusca V. Academic Press. New York. Byrne. R. A. & B. R. McMahon. 1991. Acid-base and ionic regulation, during and following emersion, in the freshwater bivalve, Anodonta grandis simpsoniana (Bivalvia: Unionidae). Biol. Bull. 181:289-297. Coimbra, J.. J. Machado, P. L. Fernandes, H. G. Ferreira & K. G. Ferreira. 1988. Electrophysiology of the mantle of Anodonta cygnea. J. Exp. Biol 140:65-88. Collip. J. B. 1921. A further study of the respiratory processes in Mya arenaria and other marine mollusca. J. Biol. Chem. 49:297-310. Crenshaw, M. A. & J. M. Neff. 1969. Decalcification at the mantle-shell interface in molluscs. Amer. Zool. 9:881-885. Davis, W. L.. R. G Jones. J. P. Knight & H. K. Hagler. 1982. An electron microscopic histochemical and X-ray microprobe study of spherites in a mussel. Tissue Cell. 14:61-67. George. S. G 1982. Subcellular accumulation and detoxification of metals in aquatic animals, pp. 3-52. In: W. B. Vernberg, A. Calabrese, F. P. Thurberg & F. J. Vernberg (eds.). Physiological Mechanisms of Marine Pollutant Toxicity. Academic Press, New York. Howard, B.. P. C. H. Mitchell. A. Ritchie. K. Simkiss & M. Taylor. 1981. The composition of intracellular granules from the metal- accumulating cells of the common garden snail (Helix aspersa). J. Biochem. 194:507-511. Istin, M. & J. P. Girard. 1970a. Carbonic anhydrase and mobilisation of calcium reserves in the mantle of lamellibranchs. Calcif. Tissue Res. 5:247-260. Istin. M. & J. P. Girard. 1970b. Dynamic state of calcium reserves in freshwater clam mantle. Calcif. Tissue Res. 5:196-205. 650 MOURA ET AL. Kapur, S. P. & M. A. Gibson. 1968. A histochemical study of calcium storage in the foot of the freshwater gastropod. Heliosoma duryi eu- discus (Pilsbry). Can. J. Zool. 46:987-992. Lautie. N.. A.-M. Carru & M. Truchet. 1988. Bioaccumulation naturelle de manganese et de fer fans les tissus mous A' Anodonta cygnea (Mol- lusque, Lamellihranche. Metabranchie). Malacologia. 29:405—117. Machado, J., J. Coimbra, C. Sa & I. Cardoso. 1988. Shell thickening in Anodonta cygnea by induced acidosis. Comp. Biochem. Physiol. 91A: 645-651. Machado, J.. K. G. Ferreira. H. G. Ferreira & P. L. Femandes. 1990. The acid-base balance of the outer mantle epithelium of Anodonta cygnea. J. Exp. Biol. 150:159-169. Mason, A. Z. & K. Simkiss. 1982. Sites of mineral deposition in metal- accumulating cells. Exp. Cell Res. 139:383-391. McMahon, R. F. 1979. Tolerance of aerial exposure in the Asiatic fresh- water clam, Corbicula fluminea (Muller). pp. 227-241. In: J. C. Britton (ed.). Proceedings First International Corbicula Symposium. Texas Univ Research Foundation. Fort Worth. Mersch, J., P. Wagner & J. C. Pihan. 1996. Copper in indigenous and transplanted zebra mussels in relation to changing water concentrations and body weight. Environ. Toxicol. Chem. 15(61:886-893. Pekkarinen, M. & R. Suoranta. 1995. Effects of transportation stress and recovery and sample treatment on Ca2* and glucose concentrations in body fluids of Anodonta anatina (Linnaeus). J. Shellfish Res. 14:425- 433. Pekkarinen, M. & I. Valovirta. 1997. Histochemical and X-ray studies on tissue concretions and shells of Margaritifera margaritifera (Lin- naeus). /. Shellfish Res. 16:169-177. Pynnonen. K. 1990a. Physiological responses to severe acid stress in four species of freshwater clams (Unionidae). Arch. Envirom. Contain Toxicol. 19:471-478. Pynnonen. K. 1990b. Effect of acidic conditions in cadmium kinetics and electrolyte balance in the freshwater clam Unio pictorum. Ann. Zool. Fennici. 27:351-360. Pynnonen, K. 1994. Hemolymph gases, acid-base status, and electrolyte concentration in the freshwater clams Anodonta anatina and Unio tu- mulus during exposure to and recovery from acidic conditions. Physiol. Zool. 67:1544-1559. Pynnonen. K„ D. A. Holwerda & D. I. Zandee. 1987. Occurrence of cal- cium concretions in various tissues of freshwater mussels, and their capacity for cadmium sequestration. Aquat. Toxicol. 10:101-114. Silverman, H. 1989. Form and function of calcium concretions in unionids. pp. 367-384. In: R. E. Crick (ed.). Origin, Evolution and Modern As- pects of Biomineralization in Plants and Animals. Plenum Press. New York. Silverman. H.. W. T. Kays & T. H. Dietz. 1987a. Maternal calcium contribution to glochidial shells in freshwater mussels (Eulamellibran- chia: Unionidae). J. Exp. Zool. 242:137-146. Silverman, H.. J. W. McNeil & T. H. Dietz. 1987b. Interaction of trace metals Zn. Cd. and Mn, with Ca concretions in the gills of freshwater unionid mussels. Can. J. Zool. 65(4):828-832. Silverman, H., W. L. Steffens & T. H. Dietz. 1983. Calcium concretions in the gills of a freshwater mussel serve as calcium reservoir during pe- riods of hypoxia. J. Exp. Zool. 227:177-189. Silverman. H.. W. L. Steffens & T. H. Dietz. 1985. Calcium from extra- cellular concretions in the gills of freshwater unionid mussels is mo- bilized during reproduction. J. Exp. Zool. 236:137-147. Sorokina. K. A. & V. S. Zelenskaya. 1967. Peculiarities of eletrolyte com- position of molluscan haemolymph. J. Evol. Biochem. Physiol. 3:25- 30. Vaidya, D. P. & R. Nagabhushanam. 1980. Calcium regulation in the fresh- water snail Indoplanorbis exuslus during shell repair. Hydrobiologia. 69:29-32. Watabe. N.. V. R. Meenakshi. P. L. Blackwelder, E. M. Kurtz & D. G. Dunkelberger. 1976. Calcareous spherules in the gastropod. Pomacea paludosa. pp. 283-300. In: N. Watabe & K. M. Wilbur (eds). The Mechanisms of Mineralization in the Invertebrates and Plants. Univ. of South Carolina Press Columbia. Journal of Shellfish Research, Vol. 18. No. 2, 651-656. 19m). FEEDING ENRICHED ARTEMIA BIOMASS TO PENAEUS VANNAMEI BROODSTOCK: ITS EFFECT ON REPRODUCTIVE PERFORMANCE AND LARVAL QUALITY R. VVOUTERS,1 L. GOMEZ,1 P. LAVENS,2 AND J. CALDERON1 lFundacidn CENAIM-ESPOL Centra National de Acuicultura e Investigaciones Marinas Campus Politecnico Casilla 0901 4519, Guayaquil, Ecuador 'Laboratory of Aquaculture and Artemia Reference Center University of Gent Rozier 44. B-9000 Gent, Belgium ABSTRACT Two experiments were conducted co-feeding Pendens vannamei broodstock with frozen Artemia biomass. In the first experiment, animals were fed natural diets supplemented with squid (treatment SQ). Artemia (A), or enriched Artemia (EA). In the second experiment, animals received a supplement of Artemia enriched with different products; rich in polyunsaturated fatty acids (PUFA) and cholesterol (treatment L). rich in vitamin c, vitamin e. and astaxanthin (treatment V). or a complete enrichment (treatment LV). In experiment 1, treatment SQ gave poor results for most parameters. Supplementation with Artemia resulted in higher survival, higher maturation frequency, a higher incidence of repeated spawns, and an improved larval quality. The best results were obtained in the treatment that received enriched Artemia. In experiment 2. the highest reproductive performance was obtained through enrich- ment of Artemia with both lipids and vitamins (LV). By reducing the concentration of PUFA and cholesterol in the enrichment product. a decline in egg fertilization, a lower incidence of repeated spawns, and a lower egg production per female was observed. High vitamin levels played a positive role only when provided in combination with high levels of PUFA and cholesterol. If not, symptoms of oversaturation occurred. KEY WORDS: Artemia, reproduction, Penaeus, shrimp broodstock, nutrition INTRODUCTION In Ecuador — the world's second largest shrimp producer in 1997 — stocking of growout ponds depends largely on wild Pe- naeus vannamei postlarvae (PL) and to a lesser extend on PL grown in hatcheries. Until recently, wild PL were preferred over hatchery PL by all the farm managers. Today, thanks to the prog- ress made in reproduction and larviculture techniques, many man- agers consider both PL types of equal quality, and have begun to focus on closing the shrimp life cycle. In June 1998, 30 hatcheries had their proper maturation facilities and four more were under construction in Ecuador. This evolution spawned an urgent and increased need for applied research and technical assistance. Cost and availability of maturation diets are among the major problems in shrimp maturation. All maturation units base the nutrition of their reproducers on a mixture of fresh frozen natural diets ( squid, mussel, oyster) locally available, with supplements of bloodworm imported from Maine, USA or Panama and relatively small por- tions of commercial dry diets. Traditionally, bloodworm has been the key to success in P. vannamei maturation: however, it is the most expensive component, and quality product is not available yearlong. A search for alternatives pointed toward Anemia bio- mass. Naessens et al. (1997) demonstrated that the replacement of bloodworm with adult Artemia does not negatively affect the re- productive performance of/3, vannamei. In 1997. fresh-frozen en- riched biomass of Artemia from the USA and occasionally from Peru was available on the Ecuadorian market and substituted com- pletely or partially the bloodworm supplement in many of the hatcheries. The effect of bloodworm on shrimp maturation has been at- tributed to its polyunsaturated fatty acid (PUFA) profile (Mid- dleditch et al. 1980, Lytle et al. 1990). Although Browdy et al. ( 1989) and Naessens et al. ( 1997) demonstrated the importance of co-feeding Artemia to P. semisulcatus and P. vannamei brood- stock, respectively, it remains unclear what constituents are re- sponsible for triggering maturation. Therefore, more detailed re- search on Artemia enrichment components is needed. The present study consisted of two experiments. The first experiment was run from July until September 1996 and sought to identify which re- productive parameters are affected by Artemia biomass and en- riched Artemia biomass. respectively. The second experiment was run from March until May 1997 and sought to identify the relative importance of certain enrichment nutrients. MATERIALS AND METHODS Experiment 1 Wild P. vannamei reproducers, captured at night at Jama (Manabi, Ecuador), were transported to the CENAIM research center and kept in maturation tanks for 2 to 3 weeks to acclimate to experimental conditions. After acclimation, female shrimp were unilaterally eyestalk-ablated by cutting and pinching and marking with eye-tags. A unisex system was used as described in Browdy et al. (1996): three tanks were stocked with 40 females each and three tanks with 45 males each. At the time of stocking, the aver- age weights of the male and female shrimps were 47.5 and 63.3 g. respectively. The postablation phase of the experiment lasted 77 days, during which females with fully developed ovaries were transferred to one of the three male tanks. If females mated, they were placed in spawning tanks. If not. they were returned to their maturation tanks. Animals were fed a base diet that consisted of fresh frozen squid, mussel, oyster, and clam at a ratio of 2.5:1.3:1:1 and at a rate of 12% x d"1 of the tank live biomass wet weight basis (WWB). Administering two different fresh-frozen diet supplements at a rate of 6% x d_l WWB resulted in the following treatments: A and EA received a supplement of adult Artemia and enriched Artemia, respectively. Treatment SQ did not receive a separate supplement: therefore. 6% x d"1 WWB more frozen squid was added to the base diet. The Artemia were harvested from San Francisco Bay 651 652 WOUTERS ET AL. ponds by San Francisco Bay Brand Co. (CA. USA) and were enriched after harvesting according their standard procedure with an experimental emulsion provided by the Anemia Reference Cen- ter (Gent, Belgium). The booster consisted of an ICES 30/4/E reference emulsion containing 30% PUFA to which 2% choles- terol, 3,000 ppm ascorbic acid (AA) equivalent (ascorbyl palmi- tate; Roche, Belgium), 1,000 ppm a-tocopherol (a-TOH) equiva- lent (DL-a-tocopherol acetate, ATA, Roche), and 1,000 ppm as- taxanthin (AX) equivalent (Carophyl Pink, Roche) were added. Feeds were administered in five daily rations, two of which con- sisted of the dietary supplement only. The three treatments were applied in the same way for female as for male broodstock. The maturation tanks were oval-shaped (5 m x 3 m; 19.6 m2) black Fiberglas tanks in which sand-filtered and UV-treated sea- water (salinity 33 g x L , pH 7.8-8.2) was exchanged at a rate of 250% daily. Water temperature was 24.0 °C during acclimation and was heated to 28.5-29.0 °C from ablation onward. A timer- controlled, inverted photoperiod of 14 h light: 10 h dark was adopted, with gradual transition between light and dark hours. Mated females were transferred to individual 300-L black spawn- ing tanks, and from each spawn, the eggs were hatched out in 20-L buckets. Nauplii were collected after phototaxism selection and stocked in 1-L bottles at a density of 100.L"' and a temperature of 29 °C until metamorphosis to zoea 1 (Zl). Eight-day larviculture trials were run in 3-L glass bottles. Late nauplii 5 (N5) were stocked at a density of 100 ind x L~' in seawaterof 33 g x L"1 and 28.5 °C. Water was exchanged 90% daily, and an algae concen- tration of 100.000 cells x raL"' Chaetoceros sp. was maintained. The hatching percentage was estimated by concentrating the viable nauplii in a 10-L bucket and counting five subsamples. The percentage of egg fertilization was determined by the presence of a double membrane and/or embryonic development. Zoea 1 length was measured with a profile projector on samples of 30 zoea each. Spermatophore quality was based on sperm count and spermato- phore weight (Alfaro and Lozano 1993). Experiment 2 Wild reproducers were captured at San Pablo (Guayas, Ecua- dor) and transported to CENAIM. Acclimation and ablation tech- niques were similar to those used in experiment 1. Three tanks with mixed sex were monitored during 55 days postablation. Each tank was stocked with 45 males and 40 females with average weights of 55.6 and 64.5 g, respectively. A similar feeding strategy as in experiment 1 was adopted, but diet supplements were adult frozen Artemia enriched with three different boosters. For treat- ments L and LV. the oil component of the enrichment product consisted of the ICES 30/4/E reference emulsion with inclusion of 2%' cholesterol. In treatment V, this oil component was replaced by the ICES 0/0/E reference emulsion based on a PUFA-free coconut oil. Furthermore, boosters V and LV contained high vitamin lev- els: 3,000 ppm AA equivalent, 1,000 ppm a-TOH .and 1.000 ppm AX equivalent. All treatments were isocaloric. The same infrastructure and conditions as in experiment 1 were used, but this time water temperature ranged from 27.0 to 29.5 °C. Larval monitoring was only continued up to stage Zl. Data Processing Female reproducers were considered as experimental units. For statistical analysis, either animals or spawns were considered as treatment replicates. Infertile spawns were not considered. In the case of ovarian maturation frequency, daily observations were used as replicates, as in Nascimento et al. (1991). Pearson mo- ment-product correlation was used to determine correlations be- tween the independent variables, female weight and spawn order, and the dependent variables related to spawn size and spawn qual- ity. Data were analyzed with analysis of variance: a two-way ANOVA for experiment 1 with male tank as second variable and a one-way ANOVA for experiment 2. Female weight and spawn order were included as covariates in an analysis of covariance (ANCOVA) for the number of eggs per spawn and Zl length, respectively, as correlations were found between them. When nec- essary, data expressed in percentages or fractions were asinV trans- formed to obtain normal distribution, although unadjusted means are presented. Duncan's new multiple range test was used to iden- tify differences among treatments. All references to statistical sig- nificance were at the 5% level or lower. RESULTS Experiment I Mean survival rates of 35.0. 62.5, and 82.5% were registered for male reproducers and 35, 37.5, and 47.5% for female repro- ducers of treatments SQ, A, and EA, respectively. On average, female reproducers survived 41 days out of the total 77 days of the experiment, for which no differences were detected among treat- ments. In total 9, 25, and 55 fertile spawns were obtained in treatments SQ, A, and EA, respectively. Spawn size (eggs per spawn) was not affected by dietary treat- ment (Table 1 ). On the other hand, ovarian maturation and rematu- ration as well as repeat spawning differed among treatments. The maturation frequency was higher when Artemia (A and EA) was supplemented to the diet. The number of females with repeated spawns increased significantly in the order SQ-A-EA, spawning frequency and total egg production per female followed the same trend. Also, all parameters related to egg or larval quality were significant better in A and EA as compared to SQ (Table 2). The total nauplii production per tank after 77 days, expressed as per- centages of the production in treatment SQ, were 451% and 875% for the Artemia and enriched Artemia treatments, respectively. No significant effect of the diet on spermatophore quality could be detected (sperm count, spermatophore weight presented in Table 3; egg fertilization presented in Table 1 ), neither was there a male tank effect in the analysis. Experiment 2 During the 55 days postablation mean survival rates of 77% for male and 58% for female reproducers were registered. On average, female reproducers survived 45 days out of the total 55 days of the experiment, for which no differences were detected among treat- ments. Over the whole period. 91. 61, and 24 fertile spawns were recorded for treatments LV, L, and V. respectively. Table 4 illustrates how different treatments affected the repro- ductive performance. The maturation frequency was significantly higher for spawners of treatment L than for spawners of treatment LV. No statistical testing was possible on spawn frequencies (no normal distribution), but a decreasing trend is observed in the order LV-L-V. Spawn size did not differ between treatments. Spawners that received a supplement of Artemia enriched with high vitamin levels only (V) exhibited the lowest number of eggs produced per female, the lowest egg fertilization, and the lowest incidence of Feeding Enriched Artemia Biomass to Shrimp Broodstock 653 TABLE 1. Effect of different dietary supplements on P. vannamei spanners: maturation frequency, spawn frequency, number of females that spawned more than once, fecundity, and fertilization (experiment 1). Dietary Supplement SQ EA Maturation/female/day* Spawns/female/day # Females that spawned more then oncet Eggs/spawn (x 103) Eggs/female (x 103) Egg fertilization (%) 0.023 ± 0.037J 0.003 ± 0.009 7" 213.0 + 67.9" 58.6 ± 142.9a 47.0 ± 26.9a 0.067 ± 0.0651, 0.011 ±0.024 16b 214.4 ±60.8" 188.9 ±388.7al 65.9 ±21.6" 0.082 ± 0.068h 0.023 ± 0.032 36c 210.1 ±69.1" 337.3 ± 507.327"1 65.9 ± 23.8a * Observation of ovarian maturation stage 3 or 4 according to King (1948). t Statistical differences detected with \2 Iesl repeated spawns. In Table 6, the mean sperm count in both sper- matophores of male reproducers is given, a recording that was significantly higher in treatment L as compared with treatment V. No significant effect of the dietary treatments on egg quality or larval quality was observed (Table 5). However, the number of zoea I produced per spawn (a combination of spawn size and larval survival) was lower in treatment V as compared with treat- ment PV. The decrease of larval survival with successive spawns (spawner exhaustion) seemed to be more critical in treatment V as in the remaining treatments (Fig. 1 ). DISCUSSION A dietary regime consisting of squid, oyster, clam, and mussel (treatment SQ) gave poor results for most reproductive parameters. This is most probably attributed to the noninclusion of bloodworm (Middleditch et al. 1980, Lytle et al. 1990). Supplementation with Artemia biomass resulted in higher survival, improved maturation and reproduction, and better offspring quality. Lavens and Sorgeloos (1991), Cahu et al. (1991). and Palacios et al. (1998) demonstrated with their work on Macrobrachiwn rosenbergii, P. indicus, and P. vannamei, respectively, that offspring quality is associated with the level of metabolic fuel (mainly lipids) in eggs and nauplii. Because fuel levels in eggs or nauplii depend on the nutritional status of the female broodstock, which is affected by TABLE 2. Effect of different dietary supplements on mean percentage hatch, percent larval survival from nauplii 2 to zoea 1, number of zoea 1 per spawn, zoea 1 length and percentage larval survival from zoea 1 to mysis 2 (experiment 1). Dietary Supplements SQ EA Hatch (%) Larval survival N2-Z1 (%) Zoea 1/spawn (x 10')* Zoea 1 length (p.m) Larval survival Z1-M2 (%) 31.6±36.4a 41.1 +32.0a 42.8 ± 56.5" 804.7+ 116.3a 8.3 ± 4.7a 60.8 ± 27.5b 71.8±22.2h 93.4 ± 60.7h 900.5 ± 36.5b 43.9 ± 23.7" 61.6 ±30.2" 69.4±21.1h 94.0 ± 60.8ah their dietary regime, we can attribute the positive effect of Artemia biomass on egg and larval quality to its good nutritional value. The nutritional composition of adult Artemia is well documented in reviews by Leger et al. (1986) and by Lavens and Sorgeloos (1996). It seems to be quite similar to the body composition of penaeid shrimp, and therefore, it is likely to contain appropriate protein and lipid levels. Considering that biomass of adult Artemia was used in this study, its effect on ovarian maturation and repro- ductive activity might also be attributed to hormones or sexual steroids in addition to the nutritional input. Indeed, it is likely that the reproductive hormones within crustaceans are of the same nature and, therefore, could be effective in other species. The findings of Mendoza et al. ( 1997) on the effect of squid extracts on vitellogenesis in P. vannamei. and of Alava and Kanazawa ( 1991 ) on the effect of clam on ovarian maturation in P. japonicus, sug- gest a role of methanol-water-soluble extracts (hormones, ste- roids) on shrimp maturation. Further research on this topic could help with the identification of the Artemia component responsible for triggering maturation. Supplementing the feeding regime with enriched Artemia as compared to regular Artemia promoted mating and spawning. Also, a positive effect on maturation seems to exist, but because of high within-treatment variation, no significant differences were detected. The booster emulsion was particularly rich in lipids, which provide essential nutrients as well as energy. Our studies on the biochemical composition of wild P. vannamei reproducers (Wouters et al. in preparation) as well as studies on other shrimp species (Middleditch et al. 1979; Read and Caulton 1980; Jeckel et al. 1989, Castille and Lawrence 1989, Mourente and Rodriguez 1991 ) demonstrated a remarkable increase of lipids in the ovaries TABLE 3. Spermatophore quality of P. vannamei male reproducers at the end of experiment 1 estimated by mean sperm count (number of sperm cells in both spermatophores) and mean spermatophore weight. Dietary Supplements SQ EA * Calculated data (number of nauplii per spawn x larval survival/100) 884.4 ± 36.4 Sperm count (x 10'') 12.97±12.89a 13.44±11.59a 13.56±6.75a 4X.4±19.5h Spermatophore weight ( g ) 0.0397 + 0.0 161" 0.0827 ± 0.0468a 0.0682 ± 0.0208a 654 WOUTERS ET AL. TABLE 4. Effect of different enriched Anemia supplements on maturation frequency, spawn frequency, number of females that spawned more than once, fecundity, and fertilization of P. vannamei spawners (experiment 2). Artermia Supplements LV L V Maturation/ female/dav* 0.141 ±0.066" 0. 1 73 + 0.079b O.I40±0.080"b Spawns/female/day (1.064 ± 0.052 0.048 ± 0.052 0.025 ± 0.035 # Females that spawned more then oncet 28" 19" llh Eggs/spawn (x 103) 288.1 ±98.4" 296.6+ 100.1° 278.5 ± 69.0a Eggs/female (x 103) 865.8 ± 577.3a 7 17.6 ±467.4J 371.4+ I42.9b Egg fertilization (%) 71 ±27'' 64 ± 24"b 41 ± 16b * This is the maturation frequency: observation of ovarian maturation stage 3 or 4 according to King ( 1948). t Also referred to as rematuration: Statistical differences detected with \2 test. during maturation. In addition, maturation and reproduction cause increased metabolic energy demands (Harrison 1990). which can be met by augmenting maternal nutrition ( Clarke 1982, Bray et al. 1990). The results of the present study indicate that through the enrichment of Anemia more adequate lipid levels were obtained that could sustain optimal reproductive performance of P. van- namei. Experiment 2 further confirms that the role of the enrichment product is not only energetic but also nutritional, all treatments were iso-caloric. Most studies on shrimp broodstock nutrition have focused on the use of various natural diets, but few attempted to elucidate specific nutritional requirements (Harrison 1990) and/or possible effects on maturation, spawning, or offspring quality. Studies on lipid requirements indicate the importance of polyun- saturated fatty acids (PUFA; Middleditch et al. 1980, Chamberlain 1988, Jeckel et al. 1989; Lytle et al. 1990; Teshima and Kanazawa 1983. Xu et al. 1994; Cahu et al. 1995) and cholesterol (Mid- dleditch et al. 1980; Kanazawa et al. 1988). Others studied the role of vitamins in shrimp reproduction: ascorbic acid (Cahu et al. 1991. Alava et al. 1993a. Cahu et al. 1995). alpha-tocopherol (Chamberlain 1988, Alava et al. 1993b. Cahu et al. 1995). and TABLE 5. Mean percentage hatch, percent larval survival from nauplii 2 to zoea 1 IN2-Z1I, number of zoea 1 per spawn and zoea 1 length for the three Artemia treatments (experiment 2). Artemia Supplements LV L V Hatch (%) 50.9 ±33.6" 44.4 ± 29.4J 32.2 + 24.6" Larval survival N2-Z1 (%) 62 ±24" 54 ±30" 47 + 27" Zoea 1 /spawn (x 103)* 87.2±S4.4" 69.2±75.5ab 40.6±42.2h Zoea 1 length Ijjun) 895.7+116.3" 862.6 ± 193.5" 799.9 ± 264.9" * Calculated data (number of nauplii per spawn x larval survival/100). TABLE 6. Spermatophore quality of P. vannamei male reproducers at the end of experiment 2 estimated by mean sperm count (number of sperm cells in both spermatophores) and mean spermatophore weight. Artemia Supplements LV \ Sperm count (10b) 1I.05+4.24"'1 28.72 ±2.40" 12.37 ±4.50h Spermatophore weight (g) 0.0956 ±0.0172" 0.0879 ± 0.0422" 0.0836 ± 0.0175" vitamin A (Alava et al. 1993b, Dall 1995). Also, astaxanthin can be considered as a vitamin (it may serve as a vitamin A precursor), and several biochemical studies have detected considerable levels of this carotenoid in hepatopancreas and gonads of crustacean reproducers (Vincent et al. 1988. Dall 1995, Dall et al. 1995; Sagi et al. 1995. Mantiri et al. 1996). The second experiment does not allow determination of the optimum dietary levels of each enrich- ment component, but may give us an estimation of the importance and role of the vitamin and lipid fractions in penaeid reproduction. Best results were obtained with the treatment that received a supplement of Artemia biomass enriched with lipids and vitamins (LV). However, by reducing the concentration of PUFA and cho- lesterol in the enriched Artemia, a decline in egg fertilization, a lower incidence of repeated spawns, and a lower egg production per female was observed, which clearly demonstrates the impor- tance of PUFA and cholesterol. Obviously, good reproductive performance also depends on male testis maturation and spermatophore quality. Several authors evaluated spermatophore quality by monitoring such parameters as spermatophore regeneration, spermatophore weight and color, sperm count, and percentage of abnormal and dead sperm cells (Chamberlain 1988. Leung-Trujillo and Lawrence 1991, Alfaro and Lozano 1993). Chamberlain detected a dietary effect on sper- matophore quality. Others adopted a unisex system to detect male diet effects on mating and fertilization but failed to do so (Naes- sens et al. 1997). In experiment 1 of the present study, no male effect was observed (Table 3), but in experiment 2. significant differences in mean sperm count were detected at the end of the experiment (Table 6). The sperm count in treatment L was twice that of the remaining treatments, but only significantly different from treatment V. The latter result suggests a negative effect of high vitamin levels on sperm production. The nonsignificant dif- ferences with the LV treatment might be explained by the fact that part of the vitamins were used as antioxidative agents protecting the high PUFA levels. In contrast. Chamberlain ( 1988) reported a positive effect of vitamin E on spermatophore quality; he detected lower percentages of abnormal and lysed sperm cells in a treatment that received a diet high in vitamin E (approximately 500 mg/kg ATA) and suggested that this is related to the membrane stabiliz- ing properties of tocopherol. It is possible that the vitamin con- centrations in the Artemia biomass of treatments LV and V were excessively high, causing oversaturation or hypervitaminosis, par- ticularly of fat-soluble vitamins. This would also explain the ob- served treatment differences in frequencies of ovarian maturation of the female reproducers. However, vitamins and astaxanthin do seem to play a positive role on shrimp reproductive performance if they are provided together with high PUFA and cholesterol levels: Feeding Enriched Artemia Biomass to Shrimp Broodstock 655 ,^ 100 — 100 _ 100 5? 80 - _ LV 3« ^T 80 _ L - 80 V NJ M N N f (N 60 A 60 4 _ ~ ci 60 Z z Z ra 40 « 40 To 40 A T > > > | 20 — £ 20 "• A"^*-^^ 4 £ 20 \* 3 a A 3 0 I i ™ CO o "> 0 * ^S« 12 3 4 5 1 2 " 4 5 12 3 4 5 Spawn Order Spawn Order Spawn Order Figure. 1. Mean percentage larval survival from Nauplii 2 to Zoea I according spawn order tor the three dietary treatments of experiment Best fitted lines and standard deviation bars are shown. the best performing treatment in most aspects is the one that re- ceived the LV enriched Artemia. We assume that the positive role of high vitamin levels can be attributed to their antioxidant prop- erties. Cowey et al. (1985) demonstrated that vitamins C and E are very efficient lipid antioxidants in eggs of salmon. Also Cahu et al. ( 1995) suggest that the beneficial action of vitamins C and E can be found in their antioxidant properties, through the protection of biological membranes from oxidation and degradation by free radicals. Further research on the role of vitamins and astaxanthin and their optimal diet inclusion levels would help to formulate performing broodstock diets. It is interesting to note that between treatment EA of experi- ment 1 and treatment LV of experiment 2, which were fed similar diets, very distinct values were recorded for percentage hatch, larval survival, and zoea length. The wild reproducers used in this study were obtained at different times of the year (June to July 1996 and February to March 1997) and at different locations (Jama and San Pablo). Hansford and Marsden ( 1995) and Marsden et al. (1997) also reported high variability in reproductive performance between P. monodon prawns captured at different seasons and explained this by differences in age and environmental effects. In the present study, spawners of both experiments had similar weights and. therefore, presumably similar ages. As such, the ob- served differences are probably attributable to seasonal- environmental and/or geographical effects. For reproducers of ex- periment 1. a high mortality was recorded during the purchase as well as a high occurrence of necrosis. It seems that the adopted acclimation period (2-3 weeks) on natural diets did not help to overcome the initial poor condition and poor nutritional status of the spawners used in experiment 1 . Finally, it is observed that spawner exhaustion; that is, decreas- ing larval survival with successive spawns, was less pronounced in treatments LV and L as in treatment V. In the latter treatment, no larval survival was obtained from spawns of the third order. This is a clear indication that rematuration and repeat performance can be promoted by providing adequate diets, a finding that is in agree- ment with the reported data for P. monodon by Marsden et al. (1997). CONCLUSION Supplementing regular Artemia biomass to P. varmamei brood- stock diets promotes maturation and spawning, and improves lar- val quality. Better results are obtained if animals are co-fed with Artemia biomass that is enriched with high levels of polyunsatu- rated fatty acids (PUFA) and cholesterol, vitamin C, vitamin E. and astaxanthin. The PUFA-cholesterol fraction of the booster emulsion plays the most important role, improving egg fertilization and promoting spawning. A positive role of the vitamin fraction can be obtained if combined with high levels of oxidative products (PUFA and cholesterol), but too high a vitamin level may cause oversaturation and, thus, negative effects. Therefore, including en- riched Artemia biomass daily in the broodstock feeding regime at a rate of 6% of the tank biomass (WWB) or higher can be recom- mended. ACKNOWLEDGMENTS This study was supported by the VLIR-Eigen Initiatieven Pro- gram of the Flemish Inter-University Council. The authors thank San Francisco Bay Brand Co. and INVE, Inc. for providing the Artemia products and Marcelo Hidalgo (CENAIM) and Karla Van Ryckeghem (ARC) for their technical assistance. We also thank the following students for their interest and participation in the experiments: Juan Valdivieso, Sylvia Alvarez, Alberto Torres. Juan Carlos Vasquez. and Jose Vergara. LITERATURE CITED Alava V. & A. Kanazawa. 1991. Effect of clam diets on ovarian develop- ment of Penaeus japonicus. Alava V. R.. A. Kanazawa, S. Teshima & S. Koshio. 1993a. Effect of dietary L-ascorbyl-2-phosphate magnesium on gonadal maturation of Penaeus japonicus. Nippon Suisan Gakkaishi 59:691-696. Alava V. R„ A. Kanazawa. S. Teshima & S. Koshio. 1993b. Effect of dietary vitamins A. E. and C on the ovarian development of Penaeus japonicus. Nippon Suisan Gakkaishi 59:1235-1241. Alfaro J. & X. Lozano. 1993. Development and deterioration of spermato- phores in pond-reared Penaeus vannamei. J. World Aquacult. Soc. 24:522-529. Bray W. A., A. L. Lawrence & L. J. Lester. 1990. Reproduction of eye- stalk-ablated Penaeus stylirostris fed various levels of total dietary lipid./ World Aquacult. Soc. 21:41-52. Browdy C. L.. A. Hadani. T. M. Samocha, and Y. Loya. 198. An evalua- tion of frozen Artemia as a dietary supplement for the stimulation of reproduction in penaeia shrimp. In: Aquaculture — A Biotechmology in progress. Bredene. Belgium. European Aquaculture Society:6 17-623. Browdy C. L.. K. McGovern-Hopkins, A. D. Stokes. J. S. Hopkins & P. A. Sandifer. 1996. Factors affecting the reproductive performance of the Atlantic white shrimp, Penaeus setiferus, in conventional and unisex tank systems. J. Appl. Aquacult. 6:1 1-25. Cahu C. L„ M. F. Gouillou-Coustans. M. Fakhfakh & P. Quazuguel. 1991 . The effect of ascorbic acid concentration in broodstock feed on repro- duction of Penaeus indicus. ICES 1991 Mariculture Committee Paper F:40. Cahu C. L., G. Cuzan & P. Quazuguel. 1995. Effect of highly unsaturated fatty acids, alpha-tocopherol and ascorbic acid, in broodstock diet on egg composition and development of Penaeus milieus. 1 12: 417—124. Castille F. L. & A. L. Lawrence. 19X9. Relationship between maturation and biochemical composition of the gonads and digestive glands of the 656 WOUTERS ET AL. shrimps Penaeus aztecus Ives and Penaeus setiferus (L.). J. Crustacean Biol 9:202-211. Chamberlain G. W. 1988. Stepwise investigation of environmental and nutritional requirements for reproduction of penaeid shrimp. PhD The- sis, Texas A&M University. Clarke A.. 1982. Lipid synthesis and reproduction in the polar shrimp Chorismus antarticus. Mar. Ecol. Prog. Ser. 9:81-90. Cowey C. J. Bell, D. Knox, A. Fraser, A. Jongson. 1985. Lipid and antioxidant systems in developping eggs of salmon (Salmo Salar). Lip- ids 20:567-572. Dall W. 1995. Carotenoids versus retinoids (vitamin A) as essential growth factors in penaeid prawns (Penaeus semisulcatus). Mar. Biol. 124:209- 213. Dall W.. D. M. Smith & L. E. Moore. 1995. Carotenoids in the tiger prawn Penaeus semisulcatus during ovarian maturation. Mar. Biol. 123: 435- 441. Harrison K. E. 1990. The role of nutrition in maturation, reproduction, and embryonic development of decapod crustaceans: a review. J. Shellfish Res. 9:1-28. Hansford S. & G. Marsden. 1995. Temporal variation in egg and larval productivity of Penaeus monodon eyestalk ablated spawners. J. World Aquacult. Soc. 26:396-405. Jeckel W. H., J. E. Aizpun de Moreno & J. Moreno. 1989. Biochemical composition, lipid classes, and fatty acids in the ovary of the shrimp Pleoticus muelleri Bate. Comp. Biochem. Physiol. 92B:27 1-276. Kanazawa A., L. Chim & Laubier. 1988. Tissue uptake of radioactive cholesterol in the prawn Penaeus japonicus Bate during ovarian matu- ration. Aquat. Living Resour. 1:85-91. King J. E. 1948. A study of the reproductive organs of the common marine shrimp, Penaeus setiferus (Linnaeus). Biol. Bull. 94:244-262. Lavens P. & P. Sorgeloos. 1991. Variation in egg and larval quality in various fish and crustacean species, pp. 221-222. In: P. Lavens, P. Sorgeloos, E. Jaspers, and F. Ollevier (eds.). Proceedings of the Larvi-91 — Fish and Crustacean Larviculture, Gent, Belgium. Lavens P. & P. Sorgeloos. 1996. Manual on the production and use of live food for aquaculture. FAO Tech. Paper 361. Leung-Trujillo J. R. & A. L. Lawrence. 1991. Spermatophore generation times in Penaeus setiferus. P. vannamei. and P. stylirostris. J. World Aquacul. Soc. 22:244-251. Leger P., D. A. Bengtson. K. L. Simpson & P. Sorgeloos. 1986. The use and nutritional value of Anemia as a food source. Oceanogr. Mar. Biol. Ann. Rev. 24:521-623. Lytle J. S., F. Lytle & J. T Ogle. 1990. Polyunsaturated fatty acid profiles as a comparative tool in assessing maturation diets of Penaeus van- namei. Aquaculture 89:287-299. Mantiri D. M. H„ G. Negresadargues, G. Charmantier. J. P. Trilles, J. Milicua & R. Castillo. 1996. Nature and metabolism of carotenoid pigments during the embryogenesis of the European lobster Homarus gammarus Linne. 1758). Comp. Biochem. Physiol. A— Physiol. 115: 237-241. Marsden G. E., J. J. McGurrenand & S. W. Hansford. 1997. A moist artificial diet for prawn broodstock: its effect on the variable reproduc- tive performance of wild caught Penaeus monodon. Aquaculture 149: 145-156. Mendoza R.. A. Revol. C. Fauvel. J. Patrois & J.-C. Guillaume. 1997. Influence of squid extracts on the triggering of secondary vitellogenesis in Penaeus vannamei. Aquacult. Nutri. 3:55-63. Middleditch B. S., S. R. Missler . D. G. Ward, J. B. McVey, A. Brown & A. L. Lawrence. 1979. Maturation of shrimp: dietary fatty acids. Proc. World Maricult. Soc. 10:472-476. Middleditch B. S.. S. R. Missler. H. B. Hines, J. B. McVey, A. Brown, D. G. Ward & A. Lawrence. 1980. Metabolic profiles of penaeid shrimp: dietary lipids and ovarian maturation. J. Chromatog. 195:359-368. Mourente G. & A. Rodriguez. 1991. Variation in the lipid content of wild-caught females of the marine shrimp Penaeus keralhurus during sexual maturation. Mar. Biol. 110:21-28. Naessens E.. P. Lavens, L Gomez., C. Browdy, K. McGovern-Hopkins. A. Spencer, D. Kawahigashi, P. Verstraete, M. Costero, B. Devresse & P. Sorgeloos. 1997. Maturation performance of Penaeus vannamei co-fed Artemia biomass preparations or a formulated pellet diet. Aquaculture 155:97-101. Nascimento I. A., W. A. Bray. J. R. Leung-Trujillo & A. Lawrence. 1991. Reproduction of ablated and unablated Penaeus schmitti in captivity using diets consisting of fresh-frozen natural and dried formulated feeds. Aquaculture 99:387-398. Palacios E.. A. M. Ibarra, J. L. Ramirez, G. Portillo & 1. S. Racotta. 1998. Biochemical composition of eggs and nauplii in white Pacific shrimp. Penaeus vannamei (Boone), in relation to the physiological condition of spawners in a commercial hatchery. Aquacult. Res. 29:183-189. Read G. H. L. & M. S. Caulton. 1980. Changes in mass and chemical composition during the moult cycle and ovarian development in im- mature and mature Penaeus indicus Milne Edwards. Comp. Biochem. Physiol. 66A: 431-437. Sagi A., M. Rise. K. Isam & S. Arad. 1995. Carotenoids and their deriva- tives in organs of the maturing female crayfish Cherax quadricarina- tus. Comp. Biochem. Physiol. 11233:309-313. Teshima S. & A. Kanagawa. 1983. Variation of lipid compositions during the ovarian maturation of the prawn. Bull. Jap. Soc. Sc. Fish. 49(6):957-962. Vincent M.. L. Ramos & L. Oliva. 1988. Variations qualitatives et quan- titatives des pigments carotenoides dans l'ovaire et l'hepatopancreas de Penaeus schmitti au cours de la maturation ovarienne. Arch. Internal. Physiologie Biochimie 96:155-164. Wouters R.. C. Molina, P. Lavens and J. Calderon. 1999. Contenido de h'pidos y vitaminas en reproductores silvestres durante la maduracion ovarica y en nauplios de Penaeus vannamei. Proceedings of the Fifth Ecuadorian Aquaculture Conference, 98-30 October 1999, Quayaquil, Ecuador. CD Rom Xu X. L.. W. J. Ji. J. D. Castell & R. K. O'Dor. 1994. Essential fatty acid requirements of the Chinese prawn, Penaeus chinensis. Aquaculture 127: 29^10. Journal of Shellfish Research, Vol. 18. No. 2. 657-662, 1949. SALINITY AND TEMPERATURE EFFECTS ON HEMATODINWM SP. IN THE BLUE CRAB CALLINECTES SAPIDUS GRETCHEN A. MESSICK,1 STEPHEN J. JORDAN,2 AND WILLIAM F. VAN HEUKELEM3 'National Ocean Service, NOAA Cooperative Oxford Laboratory Oxford, Maryland 21654-9724 'Maryland Department of Natural Resources Cooperative Oxford Laboratory Oxford, Maryland 21654-9724 'University of Maty I and. Center for Environmental Science Horn Point Laboratory Cambridge, Maryland 21613 ABSTRACT The parasitic dinoflagellate Hematodinium sp. infects and causes mortalities in blue crabs Callinectes sapidus Rathbun 1896 from high salinity coastal embayments. The seasonal infection cycle and apparent salinity and temperature requirements for infections reported from wild crab populations indicate that environmental factors influence the parasite's ability to proliferate within crab hemolymph. A series of experiments held crabs at various water temperatures and salinities to assay infection intensity and crab survival. There was a significant increase in mean intensity in infected crabs held in ambient 15-9 °C seawater for 32 days; at temperatures below 9 °C, mean intensity diminished. Mean intensity decreased significantly in infected crabs held in 10%r or 29%c artificial seawater at 9 °C for 73 days; the decrease was significantly greater at 10%o than at 29%o. Mean intensity increased in infected crabs held in 22%c seawater at either 12 or 16 °C. Presumably uninfected crabs held at 22 °C presented infections after 14 days. No infections were transmitted by exposure of uninfected crabs to infected crabs after 85 days. Low water temperature and salinity appear to limit the proliferation of Hematodinium sp. in blue crab hemolymph. Apparently uninfected crabs from endemic areas can carry pre-patent or latent infections. KEY WORDS: Hematodinium sp.. Callinectes sapidus, salinity, temperature, parasite INTRODUCTION Parasitic dinoflagellates have been reported in numerous crus- taceans including copepods, amphipods. and decapods (Chatton and Poisson 1931. Chatton 1952. Manier et al. 1971. Shields 1994). Hematodinium perezi Chatton and Poisson 1931 was origi- nally reported as a rare hemolymph-infecting parasitic dinoflagel- late in European crabs Carcinus maenas (Linnaeus 1758) and Lio- carcinus depurator (Linnaeus 1758) (Chatton and Poisson 1931). Crustaceans of commercial value reported with Hematodinium spp. infections include the Alaskan Tanner crab Chionoecetes bairdi Rathbun 1924 (Meyers et al. 1987), the Alaskan and New- foundland snow crab C. opilio (Fabricius 1788) (Meyers et al. 1990, Taylor and Khan 1995), the Norway lobster Nephrops nor- vegicus (Linnaeus) (Field et al. 1992), the American blue crab Callinectes sapidus (Newman and Johnson 1975). rock crabs Can- cer irroratus Say 1817 and C. borealis Stimpson 1859 (MacLean and Ruddell 1978). the Australian blue crab Portunus pelagicus Linnaeus (Shields 1992, Hudson and Shields 1994), Alaskan spot shrimp Pandalus platyceros Brandt 1851 (Bower etal. 1993. Mey- ers et al. 1994), and the velvet swimming crab Necora (Liocarci- nus) puber (Linnaeus) (Wilhelm and Mialhe 1996). Epizootics in some economically important species have had detrimental effects on their fisheries. Heavy mortalities in blue crabs from coastal bays of Maryland and Virginia have been associated with Hema- todinium sp. infections, with up to 100% infection prevalence dur- ing peak periods (Messick 1994). Seasonality of Hematodinium sp. infections has been reported in numerous crustaceans including C. bairdi (Meyers et al. 1987, Eaton et al. 1991, Love et al. 1993). the cancer crab Cancer pagu- rus Linnaeus 1758 (Latrouite et al. 1988). N. (L.) puber (Wilhelm and Boulo 1988; Wilhelm and Mialhe 1996). N. norvegicus (Field et al. 1992), and C. sapidus (Newman and Johnson 1975). Preva- lence of Hematodinium sp. infections in blue crabs from coastal bays of Maryland and Virginia is highest in autumn and early winter, but is reduced to zero by spring (Messick 1994). Hematodinium sp. infections are found only in blue crabs col- lected from waters with salinity higher than 119£r, although crabs from lower salinities have been sampled (Newman and Johnson 1975). The prevalence of this parasitic dinoflagellate in crusta- ceans has been attributed to several factors, including salinity, temperature, and molt cycle (Newman and Johnson 1975, La- trouite et al. 1988, Meyers et al. 1990. Eaton et al. 1991. Field et al. 1992). This paper reports results from a series of experiments conducted to assay salinity and temperature effects on Hemato- dinium sp. infections in blue crabs. Details on how salinity and temperature affect proliferation of this parasitic dinoflagellate within the hemolymph of its host will improve our understanding of its epizootiology and pathogenicity. METHODS Juvenile and adult blue crabs from coastal bays of Maryland were collected with an otter trawl. Each crab was tagged or placed in a numbered compartment to assure individual identification. Carapace width (CW) was measured as the longest distance be- tween epibranchial spines. Crabs were bled from the hemal sinus at the joint between the thorax and the 5th pleopod (swimmer fin) using a 1-cc insulin syringe equipped with a 0.5-in, 28-gauge needle. Expressed cells were allowed to adhere to an acid-cleaned, 0.1% w/v poly-L-lysine-coated microscope slide. Hemolymph preparations were fixed in Bouin's fluid and stained with Mayer's hematoxylin and eosin (H&E) (Luna 1968). To avoid puncture of 657 658 Messick et al. internal organs, some smaller crabs measuring less than 25 mm CW were induced to autotomize an appendage. Previous labora- tory experiments have shown that mortality was low when blue crab chelae were induced to autotomize (Costlow 1963). Autoto- mized appendages were squashed, smearing hemolymph and muscle tissue onto slides and processed as above. Crabs were assayed to determine whether they had patent Hematodinium sp. infections before they were utilized in experiments. Infection in- tensity was expressed as a percentage, derived by counting at least 300 cells per hemolymph preparation, dividing the number of para- sites by the total number of cells (parasites + hemocytes) counted and multiplying by 100. Due to concerns that infections were misdiagnosed, a compari- son was made to assay whether the parasite was undetected in hemolymph but present in other internal tissues. A sample of crabs was collected from a Maryland coastal bay in December 1997. In addition to hemolymph preparations, various tissues were dis- sected, placed in fixative, and processed for histology. Of the 24 crabs assayed, 16 (67%) had Hematodinium sp. parasites in the hemolymph. whereas 13 (54%) had parasites in other tissues. Based on this preliminary study, it was assumed that assaying hemolymph provided equivalent accuracy in diagnosis as other internal tissues, which require sacrificing animals and the addi- tional expense of histologic processing. Experiment I. Environmental influences Experiment 1 was designed to assay intensity of Hematodinium sp. infections and mortality in blue crabs exposed to ambient en- vironmental fluctuations in water temperature. Adult and juvenile crabs were held in a shallow tank measuring 3.7-m long x 0.6-m wide x 0.25-m deep. Smaller crabs were placed in a 6.42-mm mesh cage measuring 0.9-m long x 0.3-m wide x 0.2-m deep within the larger tank to avoid cannibalism by larger crabs. Raw. untreated coastal bay seawater flowed through tanks which were protected from direct sunlight and rain but exposed to ambient air tempera- ture. Sixty-seven infected crabs, 43 males and 24 females, were placed in the experiment. Mean CW was 70 mm, range 23-130 mm, and standard deviation (SD) = 35.7 mm. Crabs were fed Atlantic ribbed mussels Geukensia demissa Dillwyn 1817. Hemolymph was assayed for infection intensity on days 1,9, 12. 22. 32. 42, and 58 from 18 November 1993 to 1 1 January 1994. Over this period, temperature dropped from 15 to 1.1 °C; salinity ranged from 28-30%p. Experiment 2. Transmission Experiment 2 was designed to assay whether Hematodinium sp. could be transmitted to uninfected blue crabs via the water column, infected crab feces, or direct contact with infected crabs. Forty- three uninfected crabs were divided into 5 treatment groups: 3 uninfected crabs were exposed to raw seawater, 5 uninfected crabs were exposed to ultraviolet (UV) light-treated seawater. 10 unin- fected crabs were downstream from 10 infected crabs; 28 unin- fected crabs were in direct contact with 1 1 infected crabs; and 3 uninfected crabs were exposed to feces from 3 infected crabs. Infected and uninfected crabs were held in compartmentalized tanks measuring 15-cm long x 10-cm wide x 10-cm deep, open cages with 30-cm diameter x 46-cm high, or an open tank 3.7-m long x 0.61 -m wide x 0.25-m deep. Tank compartments isolated crabs from each other yet allowed seawater to flow between com- partments. Tanks were located in a boathouse and had flow- through coastal bay seawater at ambient temperature and salinity. The mean CW of 22 male and 21 female uninfected crabs which were assayed for disease transmission was 73.5 mm (range 28-128 mm). Crabs were fed Atlantic ribbed mussels, G. demissa, ad libitum. The mean CW of 10 male and 1 1 female infected crabs which were analyzed for disease progression was 72 mm (range 28-140 mm). From days 1 to 85 ( 1 November 1994 to 24 January 1995) temperature dropped from 16 to 6 °C. Salinity ranged from 29-30%f. Crabs were assayed for infection intensity on days 1.10, 21, 31,43, 59. and 85. Experiment 3. Salinity effects Experiment 3 was designed to compare parasite intensity be- tween infected crabs held in high or low salinity seawater. Crabs which were assayed and determined to be infected with Hemato- dinium sp. were placed in either 10 or 29%c non-circulated artifi- cial seawater maintained at 9 °C in a walk-in incubator at the University of Maryland's Center for Environmental Science Horn Point Laboratory near Cambridge. Maryland. Crabs were held in plastic boxes with 4-cm long x 5-cm wide x 5-cm deep watertight compartments which prevented water flow between sections. Sea- water was changed approximately every 10 days and freshly- hatched Artemia nauplii were fed to the crabs. Eighty-six infected crabs were divided into 2 groups: 41 crabs were held in salinity representative of the Maryland coastal bays (29%o). and 45 crabs were held in 10%r which was lowered from 29%o over a 2-day period. Mean CW of 47 female and 39 male crabs was 16 mm (range 10-30 mm). Hemolymph preparations were assayed blindly without knowledge of the salinity treatment. Due to lack of data from dead crabs. 71 of the original 86 crabs were statistically analyzed. Hemolymph samples were taken from crabs on days 0, 37. and 73. Crabs were held an additional 20 days after the last hemolymph assay to evaluate mortality. Experiment 4. Temperature and water type Experiment 4 compared the intensity of Hematodinium sp. in- fections in relation to temperature. Infection intensity was also compared between crabs held in either artificial or untreated raw seawater to investigate whether parasite proliferation is influenced by water type. Treatments included 22%o artificial seawater at 16 and 12 °C, and 22%c untreated, raw seawater at 16 and 12 °C. Crabs were isolated in plastic boxes with water-tight compart- ments which measured 4-cm long x 5-cm wide x 5-cm deep. The experiment was conducted in 2 walk-in incubators at the Horn Point Laboratory where temperatures were maintained at either 1 2 or 16 °C. Crabs were fed freshly hatched Artemia nauplii approxi- mately every 20 days; seawater within compartments was inter- mittently replaced. Seventy-two crabs. 37 females and 35 males, mean CW 18 mm (range 8-62 mm), were held in this experiment. All crabs were not assayed on the same day due to malfunctioning of temperature control equipment and other factors. Hemolymph samples were taken on day 1 and either days 32, 42, or 56. Experiment 5. Pre-patent infections Experiment 5 was conducted to determine if apparently unin- fected crabs canted pre-patent or latent infections which become patent at higher temperatures. Eighty-one apparently uninfected crabs were collected from Sinepuxent Bay. Maryland, in March when infection prevalence has historically been 0% (Messick 1994). Crab hemolymph was assayed to ensure no infections were apparent. Forty-two male and 39 female crabs, mean 28 mm CW (range 14-51 mm), apparently uninfected crabs were divided into Hematodinium sp. in Blue Crabs 659 2 groups: 53 crabs were placed in raw seawater and 28 crabs were placed in artificial seawater at 22 °C to assay whether seawater type in addition to temperature affected patency of infections. Crabs were isolated in plastic boxes with compartments which measured 4-cm long x 5-cm wide x 5-cm deep. Crabs were fed freshly hatched Anemia nauplii on day 7. Seawater within com- partments was replaced approximately every 5 days. Hemolymph samples were taken on days 1 and 14. Statistical methods Data from experiment 1 and experiments 3-5 were analyzed using a multivariate repeated measures analysis of variance (MANOVA) technique (SAS Institute Inc. 1988). The models tested the effects of class variables (sex, type of water, salinity, or temperature treatment) and a covariate (CW) on Hematodinium sp. infection intensity measured repeatedly on individual crabs. In addition to treating each repeated measure as a separate response variable, the method tested the effect of time on intensity, where each repeated measure represented a separate level of a derived class variable (time). Linear contrasts in the statistical models tested effects of time (= repeated measures) interactions (ex- ample: intensity at time 2 vs. intensity at time 4), and interactions of time ( = repeated measures) with other variables (time x size, time x sex, time x salinity, etc.). Observations with missing mea- surements of intensity were omitted from the analysis. Effects were considered significant if MANOVA tests or contrasts gener- ated probabilities

.05), which were higher than the smaller Uca uruguayensis (P < .05) and lower than the larger Cyrtograp- sus altimanus (P < .05; Fig. IB). In 1 cm/s"1. U. uruguayensis was still the slowest swimmer, but Chasmagnathus granulata was faster in terms of average speed (17 %, P < .05; Fig. 1C) than the larger Cyrtograpsus angulatus. In 3 cm/s"1 C. granulata was still faster than C. angulatus (25 %. P < .05; Fig. ID), but the speed was not different from that of Cyrtograpsus altimanus (P > .05: Fig. ID). Average swimming speed of C. altimanus did not differ between still water and 3 cm/s"1; whereas, the other species in- crease speed when increasing flow velocity. These differences may be related to the ecology of the species. Directional flow, a typical estuarine stimulus, is not expected to affect a mainly marine crab such as C altimanus to the extent that affect the other three strictly estuarine species studied here. Among these, C. granulata is re- stricted to estuarine habitats; whereas. C angulatus also inhabits seashores. Thus, it is possible that a megalopa that is restricted to estuaries during settlement may be more sensitive to particular factors (e.g., directional flow) of such habitat than a megalopa that could settle in other habitats. Comparing our results with the information available for other species, we found that maximum length of megalopa explained over 50% of the variance of the model for average speed and over 90% of the model for maximum speed (Fig. 2). The model of Size and Swimming Speed of Crab Megalopae 665 E ^o I- X o z LU CO £ Q LU UJ Q_ (/) O CO 0.3 £ 8 A 0.2 r^ a b c d 9 7 5 3 a b b B ~r T T l a _L 1 c SPECIES Figure. 1. Maximum length of megalopae (A) and average swimming speed of Vca uruguayensis (U.o), Chasmagnathus granulata (C.g), Cyr- tograpsus angulatus (C.ang), and Cyrtograpsus altimanus (C.alt) in still water (B), 1 cm/s"1 (C) and 3 cm/s "' (D). Different lowercase letters indicate significant differences (P < .OS). Box plots show median values (hlack points), quartiles (edges of central boxl, and non-outlier ranges (whiskers). average speed is strongly influenced by the departure of Varuna litterata (Studentized residual: 4.05. P < .01). This may be ex- plained by differences in the biology of this species. Varuna lit- terata undertakes long distance up-river migrations, including swimming against substantial currents and even climbing perpen- dicular waterfalls (Ryan and Choy 1990). Thus, it is expected that this species will have a higher relative average swimming speed 16 12 E 3 4 Q LU LU 0 Q_ CO 50 (j y = 47.58x + 0.21 Z 40 ^ = 0.92 P< 0.001 ♦ 6 y = 7.71x + 3.29 ♦ 7 r2 = 0 52 P<0.05 ♦ 5 ♦ 8 „ ♦ 9 23 «*4 ♦ 1 A ♦ 9 | 30 CO 20 5 • 10 3 ♦ 1 0 B 0 0.3 0.6 0.9 1.2 MAXIMUM LENGTH (cm) Figure. 2. Relationship between body length of megalopae and their average (A) and maximum swimming speed (B). Numbers indicate species: (1 ) Vca uruguayensis, (2) Chasmagnathus granulata. (3) Calli- nectes sapidus. (4) Cyrtograpsus angulatus, (5) Cyrtograpsus altimanus, (6) Cancer oregonensis, (7) Varuna litterata, (8) Pachygrapsus crassipes, and (9) Cancer magister. Data of 1, 2. 4, and 5 are from our work. Other information was obtained from (3) Luckenbach and Orth (1992), (6) O. Iribarne (unpublished), (7) Ryan and Choy (1990), (8) Shanks (1985) and (9) Fernandez et al. (1994). Dashed lines indicate the prediction of Peters et al. (1994). than the other studied species that do not face such extreme con- ditions. The relationship between size and swimming speed depends upon the range of size (Peters et al. 1994) and type of locomotion (Chia et al. 1984). In the range of size for crab megalopae. positive associations have been described between size and swimming speed for other marine invertebrate larvae (e.g., ascidian tadpoles: Chia et al. 1984). The model of Peters et al. (1994) fits the maxi- mum swimming speed of ascidian tadpoles (mean error = 42%; data reconstructed from Fig. 10 of Chia et al. 1984) but largely underestimates both average (mean error = 1061 %) and maxi- mum swimming speeds (mean error = 486 %) of crab megalopae (Fig. 2). Bias of such magnitude may have important implications in the interpretation of the role of swimming in settlement (Luck- enbach and Orth 1992; Fernandez et al. 1994) and dispersal of megalopae (Shanks 1995). General models describing the relationship between size and swimming speed may be useful in the analysis of the ecological and physiological general patterns (e.g.. Peters et al. 1994). De- spite this, caution should be made in the use of such models in 666 Valero et al. particular taxa. Distinctive characteristics of singular taxa may account for marked departures from the predictions of the general models and lead to biased conclusions. The results of our experiments together with the available in- formation show that, contrary to what is reported for most taxa of similar range of sizes, maximum length is a good indicator of swimming speeds of crab megalopae. We found departures from this relationship in our treatments of running water and an increase in the unexplained variance in the linear model of average swim- ming speed. Some departures may be explained by ecological differences among the species. Thus, the ecology of each species may be important in determining the average swimming speed of the species: whereas, maximum swimming speed seems to be more dependent upon physical constraints. Structural and physiognomic differences between species of similar size (Alexander and Chen 1990) and even intraspecific differences in swimming strategy (Rooney and Cobb 1991) or age (Valero et al. unpublished) may explain the variation within the models. ACKNOWLEDGMENTS This work was partially supported by Fundacion Antorchas (Grant # 1 30 1 6/ 1 -000 1 2 ) and the Uni versidad National de Mar del Plata (UNMDP; Grant # EXA 68). Fundacion Antorchas. UNMDP and Fulbright Commission supported J. V. and UNMDP supported T. L. We thank A. Bortolus, F. Botto, J. Gutierrez. L. Lucifora, G. Palomo, E. Schwindt. and R. Zenuto for their field assistance and helpful comments. S. Cobb provided helpful comments on an ear- lier version of the manuscript. LITERATURE CITED Anger. K., E. Spivak, C. Bas. D. Ismael & T. Luppi. 1994. Hatching rhythms and dispersion of decapod crustacean larvae in a brackish coastal lagoon in Argentina. Hellgolander Meeresunters 48:445-466. Alexander. D.E. & T. Chen. 1990. Comparison of swimming speed and hydrodynamic drag in two species of Idotea (Isopoda). J. Crust. Biol. 10:406—412. Boschi, E.E., M.A. Scelzo & B. Goldstein. 1967. Desarrollo larval de dos especies de Crustaceos Decapodos en el lahoratorio. Pachycheles haigae Rodriguez Da Costa (Porcellanidae) y Chasmagnathus granu- late Dana (Grapsidae). Bol. Inst. Biol. Mar. 12:1—16. Boschi, E.E. 1981. Larvas de Custacea Decapoda. pp. 699-758. In: D. Boltovskoy (ed.). Atlas del Zooplancton del Atlantico Sudoccidental. 1N1DEP. Mar del Plata, Argentina. Chia, F.S., J. Buckland-Nicks & M. Young. 1984. Locomotion of marine invertebrate larvae: a review. Can. J. Zool. 62:1205-1222. Fernandez, M., O.O. Iribame & D. Armstrong. 1994. Swimming behavior of Dungeness crab megalopae under still and flowing water. Estuaries 17:271-275. Luckenbach. M.W. & R.J. Orth. 1992. Swimming velocities and behavior of blue crab (Callinectes sapidus). Estuaries 15:186-192. Netter, J.. W. Wasserman & M.H. Kutner. 1990. Applied linear statistical models: regression, analysis of variance, and experimental designs. Richard D. Irwin. Burr Ridge, IL. 1181 pp. Nowell. A.R. & P.A. Jumars. 1987. Flumes: theoretical and experimental considerations for simulation of benthic environments. Ocean. Mar. Biol. Ann. Rev. 25:91-112. Peters, R.H.. E. Demers, M. Koelle & B.R. MacKenzie. 1994. XIV Ecol- ogy of aquatic organisms. 3. animals, the allometry of swimming speed and predation. Verh. Internal. Verein. Limnol. 25:2316-2323. Rieger. P.J. 1996. Desenvolmiento larval de Uca iCeluca) uruguayensis Nobili. 1901 (Crustacea. Decapoda. Ocypodidae). em lahoratorio. Nau- plius 4:73-103. Rooney. P. & J. S. Cobb. 1991. Effect of time of day, water temperature. and water velocity on swimming by postlarvae of the american lobster. Homarus americanus. Can. J. Fis. Aq. Sci. 48:1944-1950. Ryan. P. A.. & S. C. Choy. 1990. Observations on the mass upstream migration of Varuna litterata (Fabricius) megalopae (Decapoda. Brachyura. Grapsidae) in Fiji. Crustaceana 58:237-249. Scelzo M.A. & V. Lichtschein. 1978. Desarrollo larval y metamoforsis del cangrejo Cyrtograpsus altimanus Rathburn. 1914 (Brachyura, Grapsidae) en lahoratorio, con observaciones de la ecologia de la es- pecie. Physis 94:103-126. Shanks, A. L. 1985. Behavioral basis of intemal-wave-induced shoreward transport of megalopae of the crab Pachygrapsus crassipes. Mar. Ecol. Prog. Ser. 24:289-295. Shanks, A. L. 1995. Mechanisms of cross-shelf dispersal of larval inver- tebrates and fish. pp. 323-367. In: L. McEdward (ed.). Ecology of Marine Invertebrate Larvae. CRC Press, Boca Raton. FL. Spivak E., K. Anger, T. Luppi, C. Bas & D. Ismael. 1994. Distribution and habitat preferences of two grapsid crab species in Mar Chiquita Lagoon (Province of Buenos Aires, Argentina). Helgolander Meeresunters 48: 59-78. Zar, J.H. 1984. Biostatistical analysis, 2nd ed. Prentice-Hall. Upper Saddle River. NJ. pp. 1-718. Journal of Shellfish Research, Vol. 18. No. 2. 667-679, 1999. EVALUATION OF HARVEST STRATEGIES FOR TANNER CRAB STOCKS THAT EXHIBIT PERIODIC RECRUITMENT JIE: ZHENG AND GORDON H. KRUSE Alaska Department of Fish and Game Division of Commercial Fisheries P.O. Box 25526 Juneau. AK 99X02-5526. USA ABSTRACT Recruitment to most Tanner crab (Chionoecetes bairdi) stocks in Alaska is periodic, causing wide fluctuations in population abundance. We evaluated alternative management approaches for such Tanner crab stocks with a size-based computer simulation model. Our study focused on Bristol Bay Tanner crab, the largest Tanner crab stock in Alaska, for which a stock-recruitment relationship with recruitment periodicity has been estimated. Alternative management approaches include a 40% harvest rate on legal males (status quo), variable harvest rates based on reproductive biomass, and strategies based on mature abundance, gear selectivity, and shell condition. Under the apparent recruitment periodicity of 10-18 years, maximum mean yield is achieved with a legal harvest rate >60% with large variation in yield and a high probability of fishery closure. Because of weak density dependence, the yield curve is relatively insensitive to high harvest rates. No harvest strategies can prevent stock collapse when recruitment has long periodicity and high amplitude, although a conservative strategy reduces the probability of stock collapse. We propose a harvest strategy for Bristol Bay Tanner crab that is 0. 10%. or 20% of molting mature males when effective reproductive biomass is <7,030 1, 2:7,030 and < 15.400 t, or 2:15,400 t, respectively, with a 50% cap on harvest rate for exploitable legal crabs. The proposed strategy adjusts legal harvest rates according to changes in stock productivity indexed by recruitment strength: high legal harvest rates during the upward recruitment cycle and low rates that protect large-size crabs and reproductive potential during the downward recruitment cycle. As compared to the status quo harvest strategy, the new approach is easily implemented, has similar tradeoffs between high mean yield and relatively low variation in yield, while reducing shortages of mates for mature females and increasing fishing opportunities. KEY WORDS: Tanner crab, Chionoecetes bairdi, periodic recruitment, harvest strategies, fisheries management, Alaska INTRODUCTION Tanner crab (Chionoecetes bairdi) are widely distributed in the waters off Alaska, extending as far north as Norton Sound and as far south as Southeast Alaska. The stocks used to support some of the most important fisheries in Alaska. The fisheries have followed a boom and bust cycle. In the eastern Bering Sea. Tanner crab were first targeted by Japanese and Russian fleets in 1965. The eastern Bering Sea fishery expanded quickly in the late 1960s, and the catch reached 24.000 t in 1968. Foreign fishing for Tanner crab has been prohibited under the Magnuson Fisheries Conservation and Management Act since 1980. Directed fisheries for eastern Bering Sea Tanner crab by the U.S. fleet began in 1974. Catch peaked in 1978 at 31.300 t (Otto 1990). The population collapsed in the mid-1980s, and no fishing was allowed in 1986 and 1987. During 1990 to 1993, catches averaged 15,000 t and annual ex-vessel values averaged US$46 million. Catches dropped sharply after 1993, and the eastern Bering Sea fishery has been closed since 1997 because of the depressed stock condition. Most other Tanner crab fisheries in Alaska collapsed in the early to mid-1990s, and none of the depressed stocks have recovered. Wide fluctuations in catches are caused by fluctuations in population abundance for which highly variable recruitment dy- namics are responsible. Like many fish stocks (Koslow 1989), recruitment to most Tanner crab stocks in Alaska is periodic and strongly autocorrelated (Zheng and Kruse in press). Recruitment to the Bristol Bay stock was strong in the mid-1970s and late and early 1990s and weak during the mid-1980s and mid- and late 1990s; recruitment to the northern Gulf of Alaska stocks was strong in the mid-1970s and has been weak since the early 1990s (Zheng and Kruse in press). Although recruitment is likely to result from a combination of density-dependent and density-independent factors, it is difficult to separate the effects of density-dependent reproductive stock and autocorrelated environmental factors, as is typically the case (Deriso et al. 1986: Walters and Collie 1988). To date, stock-recruitment (S-R) relationships have been estimated only for Bristol Bay Tanner crab in the eastern Bering Sea (Zheng and Kruse 1998). For this stock, reproductive biomass explained only a small portion of recruitment variability, and residuals from the fitted S-R curve showed a strong cyclic trend (Zheng and Kruse 1998). Currently. Tanner crab fisheries in Alaska are managed by a size/sex/season approach; that is. harvest of only large males and no fishing during spring molting and mating periods. The size/sex/ season approach is based on economic consideration of market value and meat yield, protection of females for reproduction, and allowance of at least one mating season for males. In addition, commercial removals from assessed populations are based on a constant harvest rate strategy when abundance estimates are avail- able. For example, for the eastern Bering Sea stock, a harvest rate of 40% is applied to the abundance of legal-sized male crabs (>137 mm carapace width. CW). Optimal harvest rates have not formally been evaluated for any Tanner crab stock in Alaska. Fishery thresholds have not been established and evaluated either, despite the fact that many Tanner fisheries in Alaska are currently closed because of the depressed stocks. In 1999, the U.S. Secretary of Commerce ruled that Tanner crab were overfished in the eastern Bering Sea. National Standard 1 of the Magnuson-Stevens Fishery Conser- vation and Management Act requires that "conservation and man- agement measures shall prevent overfishing while achieving, on a continuous basis, the optimal yield from each fishery . . ." (NMFS 1996). For a Tanner crab stock exhibiting a strong periodic and autocorrelated recruitment pattern, what is the optimal harvest strategy to produce relatively high yield, low variation in yield, and minimum chance of stock collapses? Although harvest strat- 667 668 Zheng and Kruse egies for fish stocks with such recruitment patterns have been evaluated (e.g.. Koslow 1989; Parma 1990; Walters and Parma 1995), no such studies have been conducted for Tanner crab stocks. In this study, we constructed a size-based model, based on crab CW, to facilitate a computer simulation analysis of alternative harvest strategies for Tanner crab stocks that exhibit periodic re- cruitment. Our study focused on Bristol Bay Tanner crab, the largest Tanner crab stock in Alaska. Alternative harvest strategies include a 40% harvest rate on legal males (status quo), variable harvest rates based on reproductive biomass. and strategies based on mature abundance, gear selectivity, and shell condition. METHODS Population Model and Parameters The size-based population model constructed by Zheng et al. (1998) for Bristol Bay Tanner crab was used in this study and is summarized in the Appendix. We set the minimum CW at 93 mm for males and 70 mm for females and simulated crab abundance using width class intervals of 5 mm. The last width class included males >163 mm CW and females >115 mm CW. Population parameters from Zheng et al. ( 1998) were updated using data from 1975 to 1997 and are summarized in Table 1. Population abun- dances were simulated for June each year, after crabs have gen- erally completed annual molting and mating. Because fishing usu- ally occurred during November each year since 1993. we used a lag of 0.4 year between the abundance assessment and the No- vember fishery in our simulations. A constant natural mortality was used in our simulations. Han- dling mortality from the other crab fisheries was part of natural mortality. Handling mortality from the directed Tanner crab fish- ery and bycatch mortality from all nonpot fisheries were separated from natural mortality. Therefore, natural mortalities for both males and females were lower than those estimated by the size- based model of Zheng et al. ( 1998), in which all handling mortality was included in the estimates of natural mortality. To examine sensitivity of the alternative strategies to levels of natural mortal- ity, we compared evaluation criteria for low natural mortality and high natural mortality represented by 62.5% and 137.5% of the baseline natural mortalities (Table 1 ). The level of handling mortality from the directed pot fishery was determined by gear selectivities of sublegal males and females and handling mortality rate. We estimated the gear selectivities of sublegal male and mature female crabs from the observer data from 1990 to 1996 (Table 1) and assumed a 20% handling mor- tality rate for those crabs that are caught and returned to the sea (Zheng et al. 1998). To investigate sensitivity of results to handling mortality rate, we also simulated scenarios with 0% and 50% handling mortality rates that bracket the range of likely values. Bycatches were estimated for two kinds of nonpot fisheries: scallop and groundfish. Annual Tanner crab bycatch from the east- ern Bering Sea scallop fishery was assumed to equal the modeled population abundance times the current bycatch limitation rate of 0.1354% (J. Barnhart, Alaska Dept. of Fish and Game, Kodiak. Alaska, pers. comm.). The current limit of Tanner crabs in the eastern Bering Sea groundfish fisheries was a step function of total Tanner crab abundance estimated from the survey and was sepa- rately set for two zones (Witherell 1997). Zone 1 and part of Zone 2 are in Bristol Bay. The abundance of the modeled Bristol Bay population was about 40% of the total surveyed abundance of the eastern Bering Sea from 1988 to 1997; so. all bycatch limits for the groundfish fisheries were multiplied by 0.4 in the simulations. In addition, all bycatch limits from Zone 2 were also multiplied by 0.89, because an average of 89% of the observed bycatch in Zone 2 came from Bristol Bay from 1993 to 1997. The bycatch in Zone 2 rarely exceeded the limits; therefore, we set the maximum by- catch limit for the modeled population from the groundfish fish- eries in the Bristol Bay area of Zone 2 as 0.748 million of crabs (2.1*0.4*0.89) (Witherell 1997). Mortality rates for Tanner crab bycatches from the scallop fishery and groundfish fisheries were assumed as 40% and 80%, respectively (NPFMC 1996). Survey measurement error was assumed to follow a lognormal distribution. Simulated "true" values of effective reproductive bio- mass and crab abundance were multiplied by a measurement error to mimic the survey estimation process for each year. Effective reproductive biomass was defined as biomass of females >79 mm CW that can be mated by mature males (Zheng and Kruse 1998). The lognormal measurement errors were simulated with a standard deviation of 0.2 and a mean of zero. To prevent extremely large errors in estimated values of abundance, both ends of the mea- surement-error distribution were truncated to fall within its 98% confidence limits. S-R data for Bristol Bay Tanner crab were fitted to a normal Ricker model by Zheng and Kruse (1998), and this S-R relation- ship with cyclic residuals from a sine function was used to conduct our simulations (Fig. 1 ). Sensitivity of the harvest strategies to depensation was also examined by using a depensatory Ricker S-R curve with cyclic residuals. Sex ratio of recruits was assumed to be 55% males and 45% females based on the average ratio of recruit- ment estimates from 1976 to 1997 (Zheng et al. 1998). A lower proportion of female recruits is likely caused by a lower catch- ability by the trawl survey gear. The period length of recruitment cycles was randomly set from 10 to 18 years. Sensitivities to cycle period length and amplitude were investigated by varying cycle period length from 4 to 30 years and cycle amplitude from 0.4 to 2.5. Molting probabilities for males and maturity probabilities for females varied over time (Zheng et al. 1998). Although these prob- abilities were not strongly correlated with recruitment strengths, periods with higher molting probabilities for males and lower ma- turity probabilities for a given size for females generally occurred during good recruitment periods. To incorporate this dynamic fea- ture into the simulation model, we used two molting probability functions for males and two maturity functions for females based on the updated results by Zheng et al. (1998). The high molting probability function was used during periods with upward recruit- ment cycles whereas the low molting probability function was used during periods with downward recruitment cycles. Only a few years occurred when the 50% maturity for females were at large sizes (Zheng and Kruse 1998): thus, the low maturity probability function (becoming mature at large size) was used only during periods with the highest 50% of upward cycles. The high maturity probability function was used during the rest of a recruitment cycle. Alternative Strategies In this study, we examined three kinds of alternative harvest strategies to set guideline harvest levels (GHL; i.e.. annual catch quotas). These approaches ranged from a simple approach to a Evaluation of Tanner Crab Harvest Strategies 669 TABLE 1. Population parameters for a size-based model of Bristol Bay Tanner crab updated from Zheng et al. (1998), "New" and "Old" refer to shell condition of crabs. Male Crabs Weight Initial Ahund. Molting Probability Selecl Mid-CW New Old New- Old iivity (mm) (Kg) (million i High Low High Low New Old 95.5 0.260 1.202 0.613 0.944 0.827 0,13ft 0.000 0.104 0.057 100.5 0.304 1.449 0.975 0.922 0.777 0.13ft 0.000 0.103 0.056 105.5 0.354 1.371 1.201 0.893 0.717 0.13ft 0.000 0.164 0.057 110.5 0.408 1.070 1.295 0.855 0.648 0. 1 3ft 0.000 0.311 0.099 115.5 0.468 0.780 1.643 0.805 0.572 0.13ft 0.000 0.333 0.123 120.5 0.534 0.554 2. KM) 0.745 0.491 0. 1 3ft 0.000 0.488 0.12ft 125.5 0.606 0.366 2.233 0.673 0.410 0.136 0.000 0.706 0.139 130.5 0.684 0. 1 75 1.994 0.592 0.332 0.13ft 0.000 0.958 0.201 135.5 0.768 0.118 1.893 0.506 0.261 0.13ft 0.000 1.000 0.220 140.5 0.860 0.071 0.884 0.420 0.200 0.13ft 0.000 1.000 0.317 145.5 0.958 0.039 0.590 0.338 0.150 0.13ft 0.000 1.000 0.317 150.5 1 .064 0.019 0.398 0.265 0.111 0.136 0.000 1.000 0.317 155.5 1.177 0.008 0.242 0.203 0.080 0.136 0.000 1.000 0.317 160.5 1.298 0.003 0.145 0.152 0.058 0.136 0.000 1.000 0.317 165.5 1.428 0.000 0.204 0.113 0.041 0.136 0.000 1.000 0.317 Female Crabs Mid-CW Weight (Kg) Initial Abund. Mature Probability Selectivity (mm) New Old Low High New Old 72.5 0.215 1.359 2.158 0.024 0.368 0.007 0.030 77.5 0.256 1.520 2.816 0.056 0.488 0.024 0.042 82.5 0.300 1.255 2.893 0.126 0.632 0.059 0.080 87.5 0.349 0.842 2.822 0.260 0.768 0.170 0.102 92.5 0.402 0.476 2.409 0.462 0.867 0.323 0.142 97.5 0.461 0.227 1.624 0.677 0.928 0.333 0.205 102.5 0.524 0.091 0.873 0.837 0.961 0.333 0.359 107.5 0.592 0.021 0.335 0.926 0.978 0.333 0.359 112.5 0.665 0.007 0.115 0.968 0.987 0.333 0.359 117.5 0.743 0.003 0.043 0.987 0.992 0.333 0.359 Growth i Natural Mortality S-R Models Prop, by Width Para Male Female Level Male Female Para Normal Para Depensa Para Male Female a 15.75 25.60 mean 0.40 0.4; 1 a 2.0402 K 0.2031 ar 100.0 80.67 b 0.070 -0.134 low 0.25 0.2' P 0.0563 e 2.8031 p, 1.023 0.955 P 0.746 1.000 high 0.55 0.59 A 1.2676 P 10-18 yrs o" 0.4570 more complex approach incorporating gear selectivity and shell condition. Under the first harvest strategy, the status quo. GHL was set by legal harvest rate multiplied by legal male crab abun- dance. The current legal harvest rate is 40%. but we also evaluated nine other rates ranging from 10% to 60%. Under the second alternative. GHL was set by legal harvest rate multiplied by "ex- ploitable" legal male crab abundance. Because the fishery dispro- portionately harvests new-shell crabs over old-shell crabs, we de- fined exploitable legal males based on fishery selectivity param- eters. We estimated 100% selectivity for new-shell crabs and 32% selectivity for old-shell crabs based on comparison of catch and survey data from 1975 to 1997. Ten alternative harvest rates for exploitable legal males ranging from 15% to 65% were evaluated. Under the third approach. GHL was set by mature harvest rate multiplied by "molting mature males," but only legal males were allowed to be harvested with a catch cap of 50% of exploitable legal male abundance. In other words, the legal harvest rate is equal to the mature harvest rate multiplied by "molting mature male" abundance divided by legal male abundance. "Molting ma- ture males" were defined as 100% of new-shell males and 15% of old-shell males >1 12 mm CW. These mature males have a high probability of molting within a year. Ten alternative mature har- vest rates ranging from 10% to 35% were evaluated. Because the S-R relationship is weakly density dependent, we 670 Zheng and Kruse 10 20 30 40 Effective Reproductive Biomass (1000 t) Figure 1. Relationships between effective reproductive biomass (S) and total recruits at age 7 (i.e., 8-year time lag, upper plot) and re- siduals of logarithm of recruits per S from a normal Ricker curve (lower plot! for Bristol Bay Tanner crab. In the upper plot, numerical labels are brood year, solid line is a normal Ricker curve without autocorrelated component, and dotted line is an exponential S-R curve (depensatory curve). In the lower plot, open circles represent residuals estimated from deviations of S-R datapoints from the normal Ricker curve, and solid line represents residuals fitted to a sine function. did not attempt to estimate an optimal threshold in our simulations. Rather, we set a threshold based partly on past fishery management practice and partly on the S-R relationship. In the past, the effec- tive reproductive biomass was always below 7,030.0 t in the years when the fishery was closed. This level of effective reproductive biomass is slightly above the smallest effective reproductive bio- mass with an above average recruitment level (Fig. 1 ). A stair-step harvest rate schedule similar to that employed for the Bristol Bay red king crab (Paralithodes camtschaticus) fishery (Zheng et al. 1997) was also evaluated for each approach. We used 22.000 t of effective reproductive biomass as a base level, which is the average of simulated effective reproductive biomass under the status quo constant 40% legal harvest rate. When effective reproductive biomass was at or below 50%. 60%, or 70% of this base level, harvest rates would decrease by 50% or 40%. A com- bination of three levels of effective reproductive biomass and two levels of reduced harvest rates resulted in six alternative stair-step harvest rate schedules. We evaluated each stair-step schedule in combination with each of the three harvest strategy approaches. Simulations The alternative harvest strategies were evaluated by simulating the Bristol Bay Tanner crab stock and fishery with the population dynamic model and a standard set of population parameters. The simulation model was initialized with effective reproductive bio- mass from 1990 to 1997 (Zheng and Kruse 1998) and population abundance in 1997 (Table 1) so that year 1 corresponded to 1998. The simulated time horizon was set at 100 years. Each scenario was replicated 1000 times to ensure relative stability of statistics. Identical seeds for random number generators were used for all scenarios to compare different strategies under identical environ- mental conditions. We examined sensitivity of each strategy to changes in natural mortality, handling mortality, and S-R curve. The standard set of population parameters was used in each sensitivity analysis, except that both a normal Ricker S-R curve and depensatory S-R curve were used and that the parameter under consideration was assigned one of two opposite and extreme values. For sensitivity studies on recruitment cycles, we used 200 replicates, each for 1000 years. A longer simulated time horizon was needed to examine cycle period length. To evaluate the strategies, statistics were collected on effective reproductive biomass, probabilities of fishery closure, probabilities that the stock is below the overfished reference point as defined in the fishery management plan (NPFMC 1998), and yield. Probabili- ties of fishery closure are denoted as the proportions of replicates with estimated effective reproductive biomass below threshold so that the fishery is prohibited for a given year. The overfished level is defined for Tanner crab in the entire eastern Bering Sea. not just Bristol Bay. Based on the survey data from 1983 to 1997, we approximated the equivalent overfished level for Bristol Bay Tan- ner crab as 26,600 t of total mature male and female biomass. Results were averaged over the simulated time horizon and over all replicates. To assess optimality. an equal tradeoff value between increase in mean yield and decrease in standard deviation of yield was computed as 0.5*yield-0.5*standard deviation (Zheng et al. 1997) for each alternative strategy. RESULTS The tradeoff between mean yield and standard deviation of yield as a function of constant harvest rate (i.e.. without the stair- step) was similar among the three approaches (Fig. 2). Mean yield, standard deviation of yield, and proportion of years that mature population abundance was below the overfished reference point increased as a function of harvest rate, but the standard deviation increased at a faster rate than mean yield. The rate of increase in mean yield generally slowed down as harvest rate increased, es- pecially with legal harvest rate >40%, exploitable harvest rate >45%. and mature harvest rate >20%. Variations in yield, indexed by standard deviations of yield, were very high for all three ap- proaches. This is a direct result of the periodic recruitment feature of Tanner crab population dynamics. Even without a fishery, re- productive biomass fell below the overfished reference point in 9.4% of years. The legal harvest rate of 40% (status quo) is equiva- lent to an exploitable legal harvest rate of 45% and a mature harvest rate of 20%. Under equivalent harvest rates, both legal harvest rate and exploitable legal harvest rate approaches had simi- lar mean yield and standard deviation of yield, but the proportion of years at overfished levels was lower for the exploitable harvest rate approach than the harvest rate approach. Mean yield, standard Evaluation of Tanner Crab Harvest Strategies 671 -Mean Yield ■SD of Yield ■ Overfished 0 05 0 1 0.15 0.2 0 25 0.3 0.35 0 4 0 45 0 5 0 55 0.' Harvest Rale for Legal Crabs 0 39 0 34 TJ O r 0.29 T- 40%, exploitable legal harvest rate >45%, and mature harvest rate >20% (Fig. 2), we used these harvest rates as the high harvest rate levels in the stair-step functions. The legal harvest rate of 40% also happens to be the status quo harvest rate. For each approach, a decrease from 70 to 60 to 50% in cut-off levels of effective reproductive biomass or an increase in low harvest rates from 50 to 60% resulted in slightly higher trade-off values between in- crease in mean yield and decrease in standard deviation of yield but caused slightly higher percentages of years with fishery closure and with mature biomass being below the overfished reference point (Table 2). Overall, the mature harvest rate approach had slightly higher trade-off values between increase in mean yield and decrease in standard deviation of yield compared to the other two approaches. It also had slightly lower percentages of years with fishery closure and fewer years being overfished. The harvest strat- egy with a high mature harvest rate of 20% and a low rate of 10% with a cut-off of 15,400 t of effective reproductive biomass had the lowest percentages of years with fishery closure and at overfished levels among all the alternatives (Table 2). The trade-off value between increase in mean yield and decrease in standard deviation of yield was intermediate among the range in values among all harvest strategies (Table 2). In the context of National Standard I, we considered this strategy as the most attractive alternative to the status quo strategy. Sensitivity analyses of natural mortality, handling mortality, S-R curve, and recruitment cycle were conducted on the proposed new strategy and the status quo strategy. As expected, higher natu- ral mortality or handling mortality rate resulted in much lower catch and higher percentages of years with fishery closure and at overfished levels for all alternative strategies (Table 3). and vice versa for lower natural mortality or handling mortality rate. The depensatory S-R curve had a minor effect on the results of simu- lations except when depensation was combined with high natural mortality, which resulted in extremely low population abundances and few fishing opportunities (Table 3). Effective reproductive biomass rarely fell into the depensatory range under other circum- stances. Under the same conditions, the status quo harvest strategy had slightly higher mean yield, lower standard deviation of yield, higher percentages of years with fishery closure and at overfished levels than when the status quo harvest strategy included stair-step harvest rates (Table 3). The status quo harvest strategy also had higher mean yields than those for the proposed new strategy under the same conditions, but its standard deviations of yield and its percentages of years at overfished levels were much higher (Ta- ble 3). With the normal S-R curve, the status quo and proposed har- vest strategies were very sensitive to period length and amplitude of recruitment cycle, especially for a long period length and high amplitude (Fig. 3). Coefficient of variation of yield and propor- tions of years of fishery closure and at overfished levels increased substantially as period length and amplitude of recruitment cycle increased. For a given combination of period length and amplitude of recruitment cycle, the proposed harvest strategy resulted in only a minor improvement on coefficient of variation of yield and pro- portion of years of fishery closure over the status quo strategy (Fig. 3). The proposed harvest strategy reduced proportions of years at overfished levels considerably when the recruitment cycle period length was 18 years or less. The sensitivities of the status quo and suggested harvest strat- egies to the depensatory S-R curve depended on period length and amplitude of recruitment cycle (Fig. 4). For a period length ^10 years or an amplitude sl.0. effective reproductive biomass rarely fell below the depensatory range; thus, the simulation results be- tween the normal S-R curve and the depensatory S-R curve were almost identical for this region of parameter values (Figs. 3. 4). For combinations of period lengths £15 years and amplitudes a 1.5, coefficients of variation of yield and proportions of years of fish- ery closure and years at overfished levels were much higher with the depensatory S-R curve than with the normal S-R curve (Figs. 3, 4). For combinations of extremely long period length and high amplitude, effective reproductive biomass with the depensatory S-R curve was always below the threshold level (Fig. 4). Under likely ranges of 10-18 years of period length and amplitudes of 1 .0-1 .4, the depensatory S-R curve did not have a major impact on the simulation results. Similar to the results with the normal S-R curve, the proposed new harvest strategy reduced proportions of 672 Zheng and Kruse Cul-off (1000 t) TABLE 2. Comparisons of mean yield, standard deviation of yield (SD), equal trade-off between increase in mean yield and decrease in standard deviation of yield, mean effective reproductive biomass (SP), mean total mature biomass (TMB), percentage of years without fishing (Closure), and percentage of years below the overfished reference point (Overfished) for alternative harvest strategies. High Low Yield SD SP TMB Closure Overfished HR Low HR 1 1 1100 tl (1000 t) Tradeoff (1000 t) 1 1000 tl 7.030 11.000 13.200 15.400 11.000 13.200 15.400 7.030 11.000 13.200 15.400 11.000 13.200 15.400 7.030 11.000 13.200 15.400 11.000 13.200 15.400 0.400 0.000 0.400 0.200 0.400 0.200 0.400 0.200 0.400 0.240 0.400 0.240 0.400 0.240 0.450 0.000 0.450 0.225 0.450 0.225 0.450 0.225 0.450 0.270 0.450 0.270 0.450 0.270 0.200 0.000 0.200 0.100 0.200 0.100 0.200 0.100 0.200 0.120 0200 0.120 0.200 0.120 7.377 7.313 7.267 7.211 7.328 7.294 7.253 6.988 6.912 6.862 6.800 6.928 6.889 6.841 Matu 6.844 6.778 6.732 6.676 6.796 6.761 6.718 6.813 Harvest Rates Applied to Total Legal Crabs 7.999 -0.311 21.660 8.095 -0.391 8.138 -0.435 8.170 -0.480 8.071 -0.371 8.101 -0.403 8.122 -0.435 21.812 21.872 21.923 21.778 21.824 21.862 Harvest Rates Applied to Exploitable Legal Crabs 7.702 -0.357 21.971 7.801 -0.444 22.088 7.851 -"494 22.139 7.895 -0.548 22.185 7.780 -0.426 22.065 7.818 -0.464 22.106 7.850 -0.505 22.143 re Harvest Rates Applied to Molting Mature Male Crabs 7.258 -0.207 21.964 7.339 -0.281 22.049 7.381 -0 524 22.089 7.418 -0.371 22.125 7.316 -0.260 22.025 7.346 -0.292 22.054 7.372 -0.327 22.082 Estimated Historical Averages from 1975 to 1997 6.846 -0.016 23.586 62.133 62.676 62.957 63.245 62.552 62.763 62.978 63.944 64.436 64.701 64.989 64.335 64.545 64.772 64.078 64.462 64.682 64.922 64.355 64.520 64.703 65.205 (%) 14.521 14.034 13.848 13.721 14.150 14.018 13.919 13.598 13.291 13.167 13.059 13.351 13.251 13.158 13.366 13.121 13.000 12.875 13.188 13.087 12.994 13.043 (%) 31.847 30.917 30.162 29.385 31.146 30.579 30.004 28.480 27.483 26.772 26.019 27.706 27.122 26.524 27.333 26.464 25.754 24.986 26.714 26.192 25.620 26.087 ••Cut-off is a level of SP below which the low harvest rate (HR) is used and at or above which the high harvest rate is used. The status quo strategy is underlined and the proposed new strategy is shown in bold. Historical data were included for comparison. years of fishery closure and at overfished levels considerably when period length of recruitment cycle was 1 8 years or less and am- plitude was 1.8 or less (Fig. 4). Although the status quo harvest strategy is a constant legal harvest rate of 40%. legal harvest rates actually implemented dur- ing the last 23 years were quite different from this level and varied greatly over time (Fig. 5). Realized legal harvest rates were higher than 40% during 1977 to 1980 and 1989 to 1992 and much lower during 1983 to 1988 and 1994 to 1997. It seems that it is difficult to implement a constant legal harvest rate strategy. Preseason GHLs were generally slightly higher than actual yields for most years but much higher than actual yields when the GHLs were low (Fig. 5). The proposed harvest strategy leads to higher legal har- vest rates than the historical rates of the status quo strategy when population abundance is increasing and to lower rates when popu- lation abundance is decreasing. Historical harvest rates more closely match the proposed new harvest strategy than the status quo "constant" harvest rate strategy (Fig. 5). DISCUSSION Changing environments pose great challenges to fishery man- agers. Environmental shifts cause large changes in growth, mor- tality, and recruitment, making it difficult to design, evaluate, and implement optimal harvest strategies that are robust to wide swings in productivity. When trends in environmental effects on recruitment can be predicted, harvest strategies can be adjusted to maximize expected discounted yield. Escapement goals (Parma 1990) or harvest rates (Criddle et al. 1998) can be raised when favorable conditions are anticipated and lowered when poor con- ditions are expected. Even lacking knowledge of environmental effects, a constant harvest rate strategy still produces a long-term harvest close to the theoretical optimum for stocks with periodic or autocorrelated recruitment (Walters and Parma 1995). However, neither constant harvest rate nor escapement goal strategies can prevent collapse of stocks with high-amplitude, low-frequency re- cruitment variability, although a constant escapement strategy minimizes the risk (Koslow 1989). Harvest strategies for Tanner crab differ from those for many fish stocks, because they take into account differences in biology. Female Tanner crabs can store sperm for more than 1 year, and stored sperm from multiple matings may fertilize clutches for the subsequent 2 years (Paul 1984). A mature male Tanner crab can mate with a maximum of 8-10 females in a laboratory setting during a breeding season (Paul 1984). although the number of females a male can mate in the field may be less than this maxi- mum number because of low density or discrete spatial distribu- tions of the sexes and a limited mating window. Conceivably, the size/sex/season approach coupled with a suitable harvest rate on legal crabs of size one or two molts larger than mature males could adequately protect reproductive potential of Tanner crab. How- ever, molting probabilities of male crabs decrease sharply when they attain large claws (unpublished data), and periods of poor recruitment lead to depressed populations predominated by old "skip molt" or "terminal molt" crabs. Applying a constant harvest rate to total legal abundance when the abundance is low could result in a high discard rate of old-shell crabs and a very high Evaluation of Tanner Crab Harvest Strategies 673 TABLE 3. Comparisons of mean yield, standard deviation of yield (SD), mean effective reproductive biomass (SP), mean total mature biomass (TMBl, percentage of years without fishing (Closure), and percentage of years below the overfished reference point (Overfished) for three harvest strategies under low and high natural mortality (M), three levels of handling mortality iHMl and two S-R curves for the status quo, modified status quo. and proposed new strategy (see Table 2). Harvest Meld SD SP TMB Closure Overfished Strategy M HM (1000 t) (1000 t) (1000 1) (1000 ti (%) (%) Status quo Status quo Status quo Status quo Modif. status quo Modif. status quo Modif. status quo Modif. status quo New proposed New proposed New proposed New proposed Status quo Status quo Status quo Status quo Modif. status quo Modif. status quo Modif. status quo Modif. status quo New proposed New proposed New proposed New proposed Low High Normal Normal Low High Normal Normal Low High Normal Normal Normal S-R Curve and Harvest Rates Applied to Total Legal Crabs 89.759 41.418 67.118 55.236 90.272 42.721 67.632 57.198 12.157 4.837 8.933 6.687 12.227 4.841 9.067 6.895 Normal S-R Curve and Harvest Rates Applied to Molting Mature Male Crabs 0.2 9.404 9.865 34.374 97.521 0.2 3.868 4.753 13.872 42.442 0.0 7.267 8.085 23.694 69.373 0.5 5.757 6.396 19.886 58.441 Depensatory S-R Curve and Harvest Rates Applied to Total Legal Crabs 0.2 0.2 0.0 0.5 0.2 0.2 0.0 0.5 1 1 .080 4.226 8.213 6.172 11.046 3.857 8.004 6.077 33.494 13.709 23.675 19.000 33.673 13.923 23.682 19.614 Low 0.2 10.998 12.094 33.240 89.113 High 0.2 1.942 3.401 7.039 22.962 Normal 0.0 8.013 8.787 23.098 65.573 Normal 0.5 5.847 6.427 18.003 52.570 Low 0.2 10.962 12.165 33.420 89.640 High 0.2 1.756 3.263 7.686 25.569 Normal 0.0 7.790 8.924 23.109 66.134 Normal 0.5 5.803 6.697 18.834 55.173 Depensatory S-R Curve and Harvest Rates Applied to Molting Mature Male Crabs Low 0.2 9.334 9.812 34.117 96.845 High 0.2 1.765 3.178 7.655 25.373 Normal 0.0 7.060 7.942 23.123 67.870 Normal 0.5 5.524 6.221 19.228 56.738 2.26 33.55 11.81 18.43 2.15 32.88 11.78 16.44 1.94 32.78 11.75 14.79 2.26 63.97 11.82 18.76 2.16 60.41 11.79 16.60 1.94 60.43 11.77 14.921 13.92 43.45 28.84 35.65 12.44 40.90 27.05 32.48 4.07 40.37 23.01 28.38 13.78 69.56 29.16 36.51 12.25 64.24 27.30 32.89 4.05 64.09 23 L9 28.83 The modified status quo strategy is represented by a cut-off SP of 15.400 t. high harvest rate of 0.4 and low harvest rate SP below which the low harvest rate is used and at or above which the high harvest rate is used. of 0.2. "Cut-off' is a level of harvest rate on new-shell, relatively young crabs because of fishery selectivity to meet market demands. The proposed harvest strategy takes into account the relation- ship between shell condition and productivity levels of Tanner crab stocks. Strong year classes are dominated by new-shell crabs. Simulation results show that the proposed new strategy adjusts legal harvest rates according to recruitment strength, which is in- dexed by changes in shell condition. Contrary to the current har- vest strategy based on legal male abundance only, use of mature crab abundance and shell condition gives the proposed new strat- egy a forward-looking feature. When an increase in future legal crab abundance is expected because of increased recruitment to the mature segment of the stock, legal harvest rates are increased. Conversely, during a downward recruitment cycle, reduced legal harvest rates will forestall the decline of large, old-shell males that are most virile (Stevens et al. 1993: Paul et al. 1995). As a comparison to the status quo harvest strategy, the new approach had similar trade-off values between mean yield and variation in yield, but it led to fewer shortages of mates for mature females and reduced probability that population abundance falls below the overfished reference point over a long term. If repro- duction can be limited because of a shortage of mature males, it is most likely during periods of low population abundance. As abun- dance declines, spatial distribution becomes more patchy, thereby potentially reducing mating encounters. By incorporating a fishery threshold and stair-step harvest rates, the proposed new harvest strategy embodies a precautionary approach to fishery manage- ment (Restrepo et al. 1998). These features reduce mature harvest rates to protect reproductive potential during periods of low abun- dance when risks of overfishing or falling below the overfished reference point are high because of uncertainties in abundance estimates and population dynamics (i.e., depensation vs. compen- sation). Although we did not explicitly evaluate economic impacts of the management alternatives, the proposed new strategy compares favorably to the current strategy. Slightly greater mean yield im- plies higher average gross revenues under the status quo as com- pared to the proposed strategy. However, the proposed strategy results in greater fishery stability, as indicated by lower variability in yield and more fishing opportunities because of fewer fishery closures. At low population abundance, catch expectations (pre- season GHL) are much more indicative of actual harvests under the proposed new strategy than the status quo. Under the status quo harvest strategy, preseason GHL is set as 40% of legal male abun- dance, despite the fact that old-shell males predominate the popu- lation and that industry targets new-shell males, leading to more 674 Zheng and Kruse Status Quo Harvest Strategy Proposed Harvest Strategy >- >+— o > O CD CO o O CD co Ll -Q O ■o O O 0.75 Cycle Amplitude Figure 3. Contour plots of CV of yield, probability of fishery closure, and probability of total mature biomass below the overfished reference point by cycle amplitude and period length of the recruitment dynamics under the normal S-R curve for Bristol Bay Tanner crab. The plots are classified by the status quo harvest strategy and the proposed new strategy based on a 0, 107r , and 20<7r stair-step harvest rates of molting mature males. grounds prospecting and catch sorting. As a result, low in-season catch-per-unit-effort triggers fishery closures short of the GHL as a conservation measure. Not only do inflated catch expectations depress prices, they may attract more fishery participants, thus reducing average revenues and increasing aggregate costs. In our analysis of alternative harvest rate strategies, we at- tempted to consider total fishing mortality as the aggregate of landed catch, handling mortality of discards in the directed fishery, and bycatch mortality in ground fish and scallop fisheries. Land- ings are documented on transaction receipts between processors Evaluation of Tanner Crab Harvest Strategies 675 Status Quo Harvest Strategy Proposed Harvest Strategy 32 CD > > O CD CO o O CD -C CO O CD .C CO CD > O -Q O 28 18 8 28 D) 18 rz CD O O CD a. 28- 18 8 0.75 1.50 2.25 0.75 1.50 2.25 Cycle Amplitude Figure 4. Contour plots of CV of yield, probability of fishery closure, and probability of total mature biomass below the overfished reference point by cycle amplitude and period length of the recruitment d> namics under the depensatory S-R curve for Bristol Bay Tanner crab. The plots are classified by the status quo harvest strategy and the proposed new strategy based on a 0, 10%. and 20% stair-step harvest rates of molting mature males. and fishers called "'fish tickets." At-sea observers monitor bycatch aboard vessels fishing for other species. Typically, total bycatch of Tanner crabs by ground fish and scallop fisheries is a small per- centage of total crab abundance in the eastern Bering Sea. Whether a significant proportion of the Tanner crab population is adversely impacted by dredges and trawls, but not caught and observed, remains a matter of speculation. Large numbers of Tanner crabs are handled and discarded during crab fisheries because of restric- tions on size, sex, season, and target species. In our study of the red king crab fishery in Bristol Bay, increased handling mortality in 676 Zheng and Kruse 0.6 0.5 0.4 0.3 0.2 0.1 0 - Obs. Harvest Rate Pro. New Harvest Rate 35 30 25 20 15 10 5 4 -Preseason GHL -Actual Catch 75 77 79 97 85 87 Year Figure 5. Comparison of the historical harvest rates (solid linel and the harvest rates derived from the proposed new harvest strategy (dotted line) as a proportion of total legal crah abundance (upper plot) for Bristol Bay Tanner crab and comparison of preseason guideline harvest level (GHL) and actual yield for eastern Bering Sea Tanner crab (lower plot) from 1975 to 1997. our model resulted in lower optimal harvest rates and higher op- timal threshold levels (Zheng et al. 1997). For the Bristol Bay Tanner crab fishery, we found that handling mortality had similar, but less pronounced, effects. Female Tanner crabs have much lower catchability during the fishery than legal-sized males. Thus, the impact of handling mortality on female Tanner crabs is smaller than on sublegal male Tanner crabs or female red king crabs. In our sensitivity analysis, we bracketed handling mortality rate at 0 and 50% to span low rates from a study that attempted to emulate the fishing process (Macintosh et al. 1996) and high rates from a laboratory study (Carls and O'Clair 1995) that considered ex- tremely cold air temperatures during winter fisheries. An extensive bibliography of capture and handling effects was compiled by Murphy and Kruse ( 1995) and reviewed in some detail by Zheng et al. (1997). Additional research is needed to assess handling mortality rates experienced by Tanner crabs accurately during commercial fisheries in the Bering Sea. Results from ongoing studies of cold wind chill effects (Kruse 1998) may significantly affect our estimates of handling mortality rate during winter fish- eries. As this research is completed, the implications on crab fish- ery management need to be analyzed. Recruitment cycles are the most striking feature of the popu- lation dynamics of Bristol Bay Tanner crab. Recruitment cycles are common to many fish and crab populations with typical peri- odicity of lOto 26 years (Koslow 1989; Zheng and Kruse in press). At the short end of periodicity, strong year classes occurred every four years from 1976 to 1988 for Pacific herring (Clupea pallasi) stocks in Prince William Sound and Sitka Sound. Alaska (Zheng 1996). The recruitment periodicity of snow crab (Chionoecetes opilio) in the northwest Gulf of Saint Lawrence is 8 years (Sainte- Marie et al. 1996); whereas, periodicity for Dungeness crab (Can- cer magister) off Northern California is about 10 years (Higgins et al. 1997). At the long end of periodicity, some fish stock sizes had periodicity as long as 150 years (Koslow 1989). Although the time series is too short to estimate periodicity of the recruitment cycle by time series methods for Bristol Bay Tanner crab, it seems that the span between strong recruitment periods was about 13-14 years during the past 25 years. Because of the brevity of the available time series, the period length, and even the existence of repeatable cycles are not well established. The strong recruitment cycles may also be caused in part by age-class overlap in recruit- ment estimated by the size-based model. Nevertheless, despite uncertainty about the details, recruitment of Bristol Bay Tanner crab, as well as other Tanner crab stocks in Alaska, seems to be at least quasiperiodic. In our simulations for Bristol Bay Tanner crab, we set recruitment periodicity randomly from 1 0 to 18 years. We also examined the sensitivity of harvest strategies to recruitment cycles ranging from 4 to 30 years. We reached the same conclusion for Bristol Bay Tanner crab as Koslow (1989) did for fish stocks; no harvest strategies can protect a stock from collapse if the re- cruitment cycle is long. This is intuitive from crab biology. Be- cause Tanner crabs mature at about age 6 and few live longer than 12 years (Donaldson et al. 1981), significant numbers of mature Tanner crabs cannot be "banked" for more than 6 years for future reproduction. However, reducing harvest rates and saving some mature crabs for future reproduction when recruitment is in the downward cycle will reduce the chance of prolonged stock col- lapse. Cyclic or periodic recruitment of Tanner crab in Bristol Bay results in a weak density-dependent S-R relationship. So, recruit- ment strength depends partly on reproductive biomass but mostly on cycle phase. This weak density-dependent S-R relationship has important implications on Tanner crab harvest strategies. Our simulations showed that yield is maximized at legal harvest rates >60%. Because mature male Tanner crabs can annually mate with multiple females, and females can store sperm for future fertiliza- tion, harvesting large males does not have a proportional reduction on the reproductive stock. Because of this feature, our results are similar to Somerton's (1981) yield per recruit analysis, which showed that yield is maximized at legal harvest rates >70%. How- ever, this "catching them before they die or are too old" strategy may have problems, such as causing insufficient males for mating, leading to recruitment overfishing, or depleting the repro- ductive stock to such a low level that depensation may occur. In the Gulf of Alaska, many depressed crab stocks have had extended periods of poor recruitment — red king stocks for >20 years and Tanner crab stocks for >10 years (Zheng and Kruse in press). The Alaska Board of Fisheries (a regulatory body making fisheries management policies for the State of Alaska) policy on king and Tanner crab management does not strive to maximize yield (ADF&G 1998). Instead, other objectives are considered, such as maintaining multiple size classes in the stock, maintaining sus- tained and reliable yields, and minimizing risks of irreversible adverse effects on reproductive potential. For Bristol Bay Tanner crab, the yield curve is relatively flat at high harvest rates, but much lower harvest rates can attain just slightly lower mean yields. Evaluation ok Tanner Crab Harvest Strategies 677 Compared to other alternative strategies we considered, our pro- posed new harvest strategy produces slightly lower mean yield, significantly lower variation in yield, it adjusts harvest rates ac- cording to stock productivity and creates more fishing opportuni- ties while affording greater protection when the stock abundance is low. These features seem more consistent with the Board policy and provide a precautionary approach to fishery management. ACKNOWLEDGMENTS This paper was funded in part by cooperative agreement NA67FM0212 from the National Oceanic and Atmospheric Ad- ministration. The views expressed herein are those of the authors and do not necessarily reflect the views of NOAA or any of its subagencies. LITERATURE CITED Alaska Department of Fish and Game (ADF&G). 1498. Commercial shell- fish fishing regulations, 1998-1999 edition. Alaska Department of Fish & Game, Juneau, Alaska. 166 pp. Carls, M. G. & C. E. O'Clair. 1995. Responses of Tanner crabs, Chiono- ecetes bairdi, exposed to cold air. Fish. Bull. 93:44-56. Criddle. K.R.. M. Herrmann. J. A. Greenberg & E.M. Feller. 1998. Climate fluctuation and revenue maximization in the eastern Bering Sea fishery for walleye pollock. N. Am. J. Fish. Manage. 18:1-10. Deriso, R.B., S.H. Hoag & DA. McCaughran. 1986. Two hypotheses about factors controlling production of Pacific halibut. Int. N. Pacific Fish. Comm. Bull. 47:167-173. Donaldson, W.E., R.T. Cooney & J.R. Hilsinger. 1981. Growth, age, and size at maturity of Tanner crab, Chionoecetes bairdi M.J. Rathbun, in the northern Gulf of Alaska. Crustaceana 40:286-302. Higgins, K... A. Hastings, J.N. Sarvela & L.W. Botsford. 1997. Stochastic dynamics and deterministic skeletons: population behavior of Dunge- ness crab. Science 276:1431-1435. Koslow. J. A. 1989. Managing nonrandomly varying fisheries. Can. J. Fish. Aquat. Sci. 46:1302-1308. Kruse. GH. (ed.). 1998. King and Tanner crab research in Alaska: annual report from July 1, 1997 through June 30. 1998. Alaska Department of Fish and Game. Division of Commercial Fisheries Regional Informa- tion Rept. 5J98-07. Juneau, Alaska. Macintosh, R.A., BG. Stevens. J.A. Haaga & B.A. Johnson. 1996. Effects of handling and discarding on mortality of Tanner crabs (Chionoecetes bairdi). High-Latitude Crabs: Biology, Management, and Economics. Alaska Sea Grant Rept. 96-02. University of Alaska Fairbanks, Fair- banks, Alaska, pp 577-590. Murphy. M.C & GH. Kruse. 1995. An annotated bibliography of capture and handling effects on crabs and lobsters. Alaska Fish. Res. Bull. 2:23-75. National Marine Fisheries Service (NMFS). 1996. Magnuson-Stevens Fishery Conservation and Management Act as amended through Oc- tober 11, 1996. National Marine Fisheries Service, NOAA Tech. Memo. NMFS-F/SPO-23. North Pacific Fishery Management Council (NPFMC). 1996. Additional analysis for Amendment 37 (Bristol Bay red king crab savings area) and an environmental assessment/regulatory impact review/initial regu- latory flexibility analysis for Amendment 41: Management of red king crab. Tanner crab, and snow crab bycatch limits in Bering Sea ground- fish trawl fisheries and establishment of a trawl closure area in near- shore waters of Bristol Bay. A proposed amendment to the Fishery Management Plan for the groundfish fishery of the Bering Sea and Aleutian Islands. North Pacific Fishery Management Council. Anchor- age, Alaska. 268 pp. North Pacific Fishery Management Council (NPFMC). 1998. Fishery man- agement plan for Bering Sea/Aleutian Islands king and Tanner crabs. North Pacific Fishery Management Council, Anchorage. Alaska. 115 pp. Otto, R.S. 1990. An overview of eastern Bering Sea king and Tanner crab fisheries. Proceedings of the International Symposium on King and Tanner Crabs. Alaska Sea Grant Rept. 90-04. University of Alaska Fairbanks, Fairbanks. Alaska, pp. 9-26. Parma, A.M. 1990. Optimal harvesting of fish populations with nonsta- tionary stock-recruitment relationships. Nat. Res. Model 4:39-76. Paul, A.J. 1984. Mating frequency and viability of stored sperm in the Tanner crab Chionoecetes bairdi (Decapoda, Majidae). J. Crust. Biol. 4: 375-381. Paul, A. J., J.M. Paul & W.E. Donaldson. 1995. Shell condition and breed- ing success in Tanner crabs. J. Crust. Biol. 15:476-480. Restrepo. V.R., G.G. Thompson, P.M. Mace. W.L. Gabriel, L.L. Low. AD. MacCall, R.D. Methot. J.E. Powers, B.L. Taylor. P.R. Wade & J.F. Witzig. 1998. Technical guidance on the use of precautionary approaches to implementing national standard 1 of the Magnuson- Stevens Fishery Conservation and Management Act. NOAA Tech. Memo. NMFS-F/SPO-3 1 . 49 pp. Sainte-Marie. B„ J-M Sevigny. B.D. Smith & GA. Lovrich. 1996. Re- cruitment variability in snow crab (Chionoecetes opilio): pattern, pos- sible causes, and implications for fishery management. High-Latitude Crabs: Biology, Management, and Economics. Alaska Sea Grant Rept. 96-02. University of Alaska Fairbanks, Fairbanks, Alaska, pp. 451- 478. Somerton, D.A. 1981. Life history and population dynamics of two species of Tanner crab, Chionoecetes bairdi and C. opilio. in the eastern Bering Sea with implications for the management of the commercial harvest. Ph.D. dissertation. University of Washington, Seattle. 220 pp. Stevens, B.G.. W.E. Donaldson. J.A. Haaga & J.E. Munk. 1993. Morphom- etry and maturity of paired Tanner crabs, Chionoecetes bairdi. from shallow- and deepwater environments. Can. J. Fish. Aquat. Sci. 50: 1504-1516. Walters. C.J. & J.S. Collie. 1988. Is research on environmental factors useful to fisheries management? Can. J. Fish. Aquat. Sci. 45:1845- 1854. Walters. C. & A.M. Parma. 1995. Fixed exploitation rate strategies for coping with effects of climate change. Can. J. Fish. Aquat. Sci. 53: 148-158. Witherell, D. 1997. Summary of the Bering Sea and Aleutian Islands groundfish fishery management plan. North Pacific Fishery Manage- ment Council. Anchorage, Alaska. 17 pp. Zheng. J. 1996. Herring stock-recruitment relationships and recruitment patterns in the Northern Atlantic and Northeast Pacific oceans. Fish. Res. 26:257-277. Zheng. J. & G.H. Kruse. in press. Recruitment patterns of Alaskan crabs and relationships to decadal shifts in climate and physical oceanogra- phy. ICES J. Mar. Sci. Zheng. J. & G.H. Kruse. 1998. A stock-recruitment relationship for Bristol Bay Tanner crab. Alaska Fish. Res. Bull. 5:1 16-130. Zheng. J.. G.H. Kruse & M.C. Murphy. 1998. A length-based approach to estimate population abundance of Tanner crab, Chionoecetes bairdi. in Bristol Bay, Alaska, pp. 97-105. //;.' G S. Jamieson and A. Campbell (eds.). Proceedings of the North Pacific Symposium on Invertebrate Stock Assessment and Management. Can. Spec. Publ. Fish. Aquat. Sci. 125. Zheng, J., M.C. Murphy & G.H. Kruse. 1997. Analysis of harvest strategies for red king crab. Paralithodes camtschaticus. in Bristol Bay. Alaska. Can. J. Fish. Aquat. Sci. 54:1 121-1 134. 678 Zheng and Kruse APPENDIX. POPULATION MODELS Male Population Model The abundances by carapace width (CW) and shell condition in any one year result from abundances the previous year minus catch and bycatch, handling, and natural mortality, plus recruitment and additions to or losses from each width class attributable to growth. O i.i+\j+\ = (% N,„ 0„ ,.,+ l = 2 [PrnfM-lWrnJ-J*' + (Om,,e-Mm-CO, : (Nml. fi'Mm - CN, J'.i m.r.l' ""-CNmj./^[,"m)mnrj ^-lw,")mo,}} + RllU+iJ+l <-y-['M'")(\-,nnrj) + (0„ ' - co„ ■ /v',m'")(l-mo,) (Al) where Nml, and Om,, are new- and old-shell male (in) crab abun- dances in width class / and year r, Mm is instantaneous natural mortality for male crabs, inn,, and mo, are molting probabilities for new-shell and old-shell crabs, R,„,, is recruitment, y is lag in years between abundance assessment and the fishery, and P,„r , is pro- portion of molting crabs growing from width /' to width / after one molt. CN„, , , and Co,,,,, are combinations of bycatch mortality and catch (legal males) or bycatch and handling mortality (sublegal males) for new-shell and old-shell male crabs. Recruitment is de- fined as recruitment to the model and survey gear rather than recruitment to the fishery. We modeled male crabs 2 93 mm CW. P,„ v 1 is computed as follows. Mean growth increment per molt is assumed to be a linear function of pre-molt width. G, = a + ln. (A2) where a and b are constants. Growth increment per molt is as- sumed to follow a gamma distribution. g(.v|a,.P)=.v"'-'e-,p/(fi0,T(a/)). (A3) The expected proportion of molting individuals growing from width class /, to width class l2 after one molt is equal to the sum of probabilities with width range [i,. u) of the receiving width class L at the beginning of next year: that is. Pmhh = i !~VvKPW.v- (A4) where 1 is the mid-width of width class /,. Female Population Model Major differences between the male and female population models are molting probability and fishing mortality. Upon matu- rity, female crabs stop molting and growing. Female crabs are not allowed to be retained in catch and are returned to the sea. The growth of immature females was modeled by functions similar to males [Eqs. (A2-4)]. Because females mature at smaller sizes than males, we modeled female crabs 2 70 mm CW. New-shell females are either immature or primiparous (first- time spawners). and their abundance results from a combination of recruitment, growth, handling mortality, bycatch mortality, and natural mortality. Old-shell mature females are survivors of the mature females from the previous year: AWi = 2 [P,rM(Nt.r.,e-M<-CNfJ.Jel^)M%\ -m,,,)] /■=] + P r,i+ 1.1+ 1 CO, "fj+i-i^ „1>^1>A//N )m,+i„+(0, r.1+1. r -Mf (A5) where M, is instantaneous natural mortality for female (/) crabs, in,, is maturity probability for width class / and year, Rf,, is recruitment. CN, ,, and CO,,, are combinations of bycatch and handling mortality for new-shell and old-shell female crabs. Catch, Bycatch Mortality, Handling Mortality, and Recruitment Effective reproductive (or spawning! biomass was described in detail by Zheng and Kruse (1998). Annual effective reproductive biomass. SPr was estimated as SP, = ^[(N,.,.,'"; + OfJJ or,)W,]. l > 80 mm CW (A6) where Nfl, and Ofn is new-shell and old-shell female abundances in width class / and year t, W, is mean weight of female crabs in width class /, 1 is the mid-width of width class /, and nr, or or, are the ratios of male reproductive potentials TNM, and TOM, to new- shell and old-shell mature female abundances TNF, and TOF, (280 mm CW) in year t or year t - 1. respectively; that is. 111; = TNM, /TNF, or^maxlTOM^/TOF,^, TOM/TOF,] (A7) Because female Tanner crabs can store sperm for subsequent fer- tilization, the ratios in year t-\ are also used. If nr, or or, > 1, we set them equal to 1 : that is, there are sufficient mature males to mate with all mature females, and so the number of reproductive females is equal to the number of mature females. The male re- productive potentials for new-shell and old-shell mature females were defined as TWM,= 2[(0-3^m.,.f+Om.u)w»J, 113 mm < 1 < 137 mm CW TOM, = ^[(0.\N„,i, + 0,,,,,) on,], l 2 1 13 mm CW (A8) where Nm , , and O,,, , , are mature male crab abundances in width class / and year / with new-shell and old-shell conditions, respec- tively, and nn, and on, are the maximum average number of new- shell and old-shell females mated by a matable male (old-shell mature males: 100%; new-shell mature males: 30% for primipa- rous females or 10% for multiparous females) in year t and are computed as follows. nn, = i +j{TNF, - a\ )/(a2 - a\ ) and / < nn, < i +j on, = i +j{TOF, - b\ )/(b2 -b\) and /' < on, < i +j (A9) where 138 mm CW), EL,, is H, = H L,/EL, H, = H H, = mm[E{NM,/EL,)MH] for 1 st approach, for 2nd approach, for 3rd approach. (A 10) where H is legal harvest rate, L, is total legal crab abundance, £ is mature male harvest rate applied to NM,. molting mature male abundance (i > 113 mm CW. 100% of new-shell crabs and 15% of old-shell crabs), and MH is the maximum allowable legal male harvest rate (50%). Catch by width from the directed fishery is equal to the product of exploitable legal harvest rate, legal male abundance, and selectivity (snml for new-shell and som , for old- shell crabs), C/v;„ , , = H, NmJ, sn,nl i > 138 mm CW. CO',„l, = HlOmL,soml i>138mmCW. (All) and total yield, TC,, is obtained by multiplying by the correspond- ing weight and summing over all widths TC,= ^[(CN'mXt+ CO'^JWil i>138mmCW (A12) Handling mortality is incorporated in the size-based model for female and sublegal male crabs. The number of deaths from han- dling mortality is a function of harvest rate, gear selectivity, and handling mortality rate (HM). Handling mortality for sublegal males is PNmll = H,Nmll snmI HM, 93 < i < 1 37 mm CW POmlJ = H,0,nl,somlHM, 93 < i < 137 mm CW (A13) and handling mortality for females is PNfj, = H,N, j , snfJ HM. i > 70 mm CW POUj = H,Ollr sofJ HM. i > 70 mm CW ( A 1 4 ) To account for handling mortality of female crabs, effective reproductive biomass is updated after fishing by modifying Equa- tion (A6) to deduct handling mortality from female abundance. SP, = 2€(%i ^ PNuJ '"'< + {°i-'.' ~ POf.iJ orlW'}- i > 80 mm CW (A15) Catch from the directed pot fishery, male handling mortality, and male bycatch mortality are combined for CNmlr and COnlll in Equation (AD. and female handling mortality and bycatch mor- tality are summed for CNfl , and COfl, in Equation (A5). Recruitment is separated into a time-dependent variable. R,, and size-dependent variables, Uml and Ufl, representing the pro- portion of male and female recruits belonging to each width class. /?,„,, = 0.55 R,U,„, /?,,,, = 0.45 /?,{/,, (A16) where Um , and U, , are described by gamma distributions similar to Equations (A3) and (A4) with two sets of parameters a,, and pr Annual recruitment is described by a normal Ricker S-R model. R, =SP,_ (AI7) where k is recruitment age (8 years for males and 7 years for females), a and P are constants, and v, = 8, + A sin(2Trr/P) as environmental noise. 8, was assumed as a MO.a). We also as- sumed a depensatory S-R model as follows. R,= KSPl,ev; R, = SP,_k e"-p s/J'-*+u', where k and f) are parameters. Journal of Shellfish Research. Vol. 18. No. 2. 681-686. 1999. EVIDENCE OF DIARRHETIC SHELLFISH POISONING ALONG THE COAST OF MAINE STEVE L. MORTON,1 TOD A. LEIGHFIELD,1 BENNIE L. HAYNES,' DEBRA L. PETITPAIN,1 MARK A. BUSMAN,1 PETER D. R. MOELLER,1 LAURIE BEAN,2 JAY MCGOWAN,2 JOHN W. HURST, JR.,2 AND FRANCES M. VAN DOLAH1 3 Marine Biotoxins Program, Center for Coastal Environmental Health and Biomolecular Research, NOAA National Ocean Sennce, Charleston, SC 29412 " Maine Department of Marine Resources, McKown Point. W. Boothbay Harbor, ME 04575 ^NIEHS Marine and Freshwater Biomedical Sciences Center, Mount Desert Island Biological Laboratory, Salsbury Cove, ME 04672 ABSTRACT Following the occurrence of several unexplained incidents of shellfish-related gastroenteritis, field studies were con- ducted to determine if diarrhetic shellfish poisoning (DSP) toxins are present in Maine coastal waters. A protein phosphatase inhibition assay for DSP toxins revealed the presence of low levels of okadaic acid-like activity in blue mussels (Mytilus edulis) at sampling sites in the Frenchman Bay-Eastern Bay region. All other sites along the Maine coast were negative. Phytoplankton populations from this area were dominated by Dinophysis norvegica. a known toxic species. Two additional known toxic species of Dinophysis were also found: Dinophysis acuminata and D. rotunda. However, all plankton samples were negative for okadaic acid-like activity. Examination of the epiphytic communities from areas where mussels showed okadaic acid-like activity revealed the presence of the toxic di- noflagellate Prorocentrum lima in association with the brown alga, Ectocarpus sp. Epiphytic samples rich in P. lima were active in the phosphatase inhibition assay. Subsequent analysis of these samples using LC-MS/MS identified the presence of dinophysis toxin-1 (DTX-1). Empty P. lima thecae identified in the digestive tract of mussels from these areas indicate that P. lima is consumed by mussels. This is the first confirmation of P. lima in northern United States coastal waters and identifies DSP as a potential public health issue INTRODUCTION Diarrhetic Shellfish Poisoning (DSP) is one of several classes of seafood poisonings caused by naturally occurring marine mi- croalgae. primarily dinoflagellates. The state of Maine has annual blooms of the paralytic shellfish poisoning (PSP) producing di- noflagellate, Alexandrium tumarense, and has a comprehensive phytoplankton monitoring and shellfish testing program for PSP which successfully protects the public health (Shumway et al. 1988). Nonetheless, the occasional incidence of unexplained shell- fish associated gastroenteritis, as well as the rejection of a single lot of shellfish tested for DSP for import to the Netherlands, has raised the question of whether DSP is also an issue of public health significance for Maine. DSP causes severe cramping, nausea, vomiting and diarrhea. The syndrome is distinguishable from microbial food poisoning in its rapid onset (as early as 30 min) and generally lasts 2-3 days. The toxins responsible for DSP are a suite of polyethers containing transfused or cyclic ether rings (Wright and Cembella 1998), in- cluding okadaic acid and the dinophysis toxins (DTX1-^1). The first incidence of human shellfish-related illness identified as DSP occurred by Japan in the late 1970's (Yasumoto et al. 1978). where the dinoflagellate Dinophysis forth was identified as the causative organism, and the toxin identified as the causative agent was termed dinophysis toxin (DTX-1 ). Retrospective analysis of simi- lar disease outbreaks in the Netherlands (Kat 1985) and Scandi- navia (Kumagai et al. 1986) confirmed that these were also asso- ciated with Dinophysis. DSP is now a frequently encountered problem in Europe and Japan, where it significantly impacts ex- tensive aquaculture industries. In 1990, the first confirmed out- break of DSP in North American occurred in Nova Scotia, Canada (Quilliam et al. 1990). DSP outbreaks have also been reported from South America, South Africa, and Australia. Two specifics of Dinophysis present in Maine coastal waters, D. acuminata and D. norvegica, are found seasonally in high num- bers. Phytoplankton data collected by the Maine volunteer phy- toplankton-monitoring network in 1998 found peaks in the abun- dance of Dinophysis spp. in June and September (L. Bean pers. eomm.). An anecdotal report from the PSP testing station in La Moine. ME (J. McGowan, pers. coram.) indicated that Dinophysis was present in Salisbury Cove, ME in July 1998 at sufficient concentrations to discolor the water, prompting a "red tide" report from a local citizen. However, causative organism responsible for the DSP outbreak in Canada was not Dinophysis but an unrelated, benthic dinoflagellate, Prorocentrum lima (Marr et al. 1992). In this study, we surveyed the 3 1 sites along the Maine coast for okadaic acid-like activity in Mytilus using a rapid protein phos- phatase inhibition assay for DSP toxins. We next monitored for DSP producing dinoflagellates of the genus Dinophysis and Pro- rocentrum and tested for toxin production. DSP toxicity was found only in association with P. lima containing samples, which was identified as DTX-1 by LC-MS/MS. 681 682 Morton et al. MATERIALS AND METHODS Sample Collection Blue mussels {Mytilus edulis) were collected during low tide at 31 sites along the Maine coast (Fig. 1). The mussels were trans- ported to the laboratory and the digestive gland dissected. Ap- proximately 20 to 50 g of digestive gland were harvested from each location for toxin analysis. From select locations, the diges- tive gland from three individuals was preserved in 2% gluteralde- hyde for gut content analysis. Discrete water (500 ml) and net tow (35 p,m mesh, 3 min) samples were collected at both low tide and high tide at each of the sampling sites. Discrete water samples were concentrated using a 8 p.m filter and preserved in 1% gluteraldehyde for cell counts. From each net tow, a 25 mL aliquot was collected and preserved in 1% gluteraldehyde for species identification. The remainder of the plankton tow was concentrated by centrifugation for toxin analysis. During sample periods of low tide, the dominant mac- roalgae were collected at each site using the methods of Morton and Faust (1997). Phytoplankton populations of the discrete water samples and epiphytic dinoflagellate populations were estimated from counts using a 0.1 mL Palmer-Maloney Counting chamber. Each sample was counted at least four times. Because of the large errors asso- ciated with macroalgal surface area determination, dinoflagellate abundance in the epiphytic samples was expressed as cells per gram wet weight of host macrophyte. For positive identification of dinoflagellates, cells were examined using a Zeiss Axiovert SI 00 inverted microscope fitted with a Spot digital camera (Diagnostic Instruments, Inc.). Toxin Analysis Mussel and Algae Extraction For each mussel sample, 1 g of digestive gland was homog- enized in 4 mL 80% methanol for 2 min using a Polytron. The resulting homogenate was centrifuged and the supernatant was washed twice with 4 ml hexane. A 2 mL aliquot of the methanol extract was retained and analyzed directly or stored frozen at -20 °C. Algal samples were size fractionated by filtration through a 100 u.m and then a 20 (j,m screens. The 20-100 p,m fraction collected and concentrated by centrifugation. The cell pellet was extracted with 100% methanol. All extracts wre stored at -20 °C until analyzed. Colorimetric Phosphatase Inhibition Assay The protein phosphatase inhibition assay was carried out in a 96 well format using the procedure of Tubaro et al. (1996). The assay tests the ability of okadaic acid (OA) standard or unknown sample to inhibit activity of purified protein phosphatase 2A Cutler Bar Harbor * ■'■•■"'fT Jonesport Figure 1. Sample sites along the coast of Maine. Evidence oe Shelleish Poisoning 683 against a colorimetrie substrate, p-nitrophenyl phosphate (pNPP). All samples were diluted in reaction buffer (40 mM Tris HC1 pH 8.4. 34 mM MgCL, 4 mM EDTA. 4 mM DTT) at least 4-fold to reduce the methanol concentration to <5% in the assay in order to eliminate inhibition of the enzyme by methanol. For the assay, 50 p.1 samples and standards (0.1-1000 nM OA; LC Laboratories, Cambridge, MA). 100 p,L reaction buffer and 50 p.L purified PP2A enzyme (final concentration 0.1 U/mL; Upstate Biotechnol- ogy, Lake Placid, NY) were added to duplicate wells of a 96 well plate (Costar. Corning, NY). To start the reaction. 50 p.L pNPP (50 mM; Sigma. St. Louis, MO) was then added to each well and the reaction allowed to proceed for 1 h at room temperature. Protein phosphatase activity was determined by color development (405 nm) in the wells using a plate reader (Titertek Multiscan Plus, Hutitsville, AL). OA-like activity in the sample was quantified relative to the standard curve. The detection limit of the colori- metrie phosphatase assay was approximately 1 x 10"'" M okadaic acid equivalents. LC-MS/MS Samples that displayed protein phosphatase inhibition activity were analyzed by LC-MS/MS using a Finnigan LCQ mass spec- trometer. The methanolic extracts were injected on a C18 column (Zorbax 2.1 x 150 mm) and eluted with a gradient of 50-95% methanol/water containing 0. 1 % TFA at a flow rate of 0.2 mL/min. A splitter device was used to direct 109r of the column effluent to the electrospray source. The mass spectrometer was operated in positive ion mode. Analysis of the toxins was achieved by trapping |M + Na]+ species for each toxin and conducting selected ion monitoring experiments for distinctive fragment ions from the col- lisionally activated dissociation of the trapped parent ions. Chro- matographic traces were acquired for the detection of the fragment ions as well as undissociated parent ions. Limits of detection for LC-MS/MS of okadaic acid and DTX-1 were approximately 1 x 10~x M. RESULTS From the 31 sites where Mytilus edulis was collected, 4 sites from Eastern Bay and Frenchman Bay displayed protein phos- phatase inhibition activity in the colorimetrie assay for DSP toxins (Fig. 2). In all cases, the activity detected (6-20 ng/g) was close to the limit of detection of the assay (0.1 nM in-well or 1.6 ng/g hepatopancreas). All other sites from Maine were consistently negative. Subsequent analysis of the positive samples by LC-MS/ MS did not indicate the presence of okadaic acid or DTX- 1 . Given the low concentrations determined in the phosphatase inhibition assay, the absence of these toxins by LC-MS/MS may reflect differences in the respective detection limits of these methods of detection. Alternatively, it may indicate that the inhibitory activity Cutler Jonesport Bar Harbor Portland Figure 2. Sample sites from the Frenchman Bay-Eastern Bay sites where Mytilus edulis was found to have okadaic acid-like activity. 684 Morton et al. in these samples was due in part or entirely to the presence of DSP toxin congeners other than OA or DTX-1. Dinophysis species, primarily dominated by D. norx-egica, were observed throughout the sampling period from the Rockland. ME to Bar Harbor, ME areas. Two additional species, D. acumunata and D. rotundia, were also observed in low numbers. Highest concentrations of Dinophysis norvegica were routinely observed at Salisbury Cove. ME. All plankton tows that were dominated by Dinophysis, were negative for phosphatase inhibition activity and displayed no OA or DTX-1 when analyzed by the LC-MS/MS. Sites where Mytilis tested positive for okadaic acid-iike activity were therefore further studied in order to identify a causative or- ganism. At all sites where Mytihis tested positive, from the Lam- oine-Bar Harbor area and the Prospect Harbor area, the dominant macrophyte found was the brown alga, Ectocarpus sp. At these locations Ectocarpus had an epiphytic community that included the toxic dinoflagellate. Prorocentrum lima (Fig. 3). This di- noflagellate was not found at any other locations. A large sample (535 g wet weight) of the epiphytic microalgae associated with Ectocarpus was collected from the Lamoine Airport site, concen- trated, and extracted in methanol. Density of P. lima from this sample was approximately 200 cells/g wet weight. This extract displayed protein phosphatase inhibitory activity using the color- imetric assay. Subsequent analysis by LC-MS/MS showed the pro- duction of DTX-1 by the wild population of P. lima; however no OA was detected (Fig. 4). Microscopic analysis of the digestive gland from Mytilus col- lected from the Lamoine Airport site and Bar Harbor site showed empty thecae. consistent with the morphology of P. lima (Fig. 5). Intact cells of P. lima were not observed within the digestive gland of Mytilus. DISCUSSION Analysis of Mytilus digestive gland from 31 sites along the Maine coast resulted in one region, Eastern Bay and Frenchman Bay, located east of Mount Desert Island, that consistently showed low levels of okadaic acid-like protein phosphatase inhibition ac- tivity, indicative of DSP toxins. The range of okadaic acid-like activity found (6.4-20 ng/g) was at least 100-fold below the Eu- ropean and Canadian regulatory limits of 2 p.g/g digestive gland. This indicates that although DSP-like toxins were present in the Frenchman Bay-Eastern Bay region, they were not present at lev- els that represent a significant public health issue. No DSP-like toxin activity was found in Mytilus along any of Maine's penin- sulas, which are commonly sampled for PSP. Subsequent LC-MS/MS analysis of Mytilus samples that tested positive for OA-like activity did not identify the presence of OA or DTX-1, the two most common toxins associated with DSP. How- ever, the toxin concentrations found in the colorimetric assay were at or below the detection limits of the LC-MS/MS method for these two congeners. Samples were not analyzed for other DSP toxin congeners due to the lack of standards. Thus, it remains unknown if the phosphatase inhibitory activity present may be due in part to different DTX metabolites. For example. DTX-3 has been found to be a direct metabolite of DTX-1 in scallops (Patinopecten yes- soensis) (Suzuki et al. 1999). This metabolite has also been found in Irish and Spanish mussels (Marr et al. 1992; Fernandez et al. 1996). Thus, an okadaic acid metabolite such as DTX-3 could also be present, leading more phosphatase inhibition than can be attrib- uted to OA or DTX- 1 . Figure 3. Light micrograph of P. lima collected from Ectocarpus sp. During the field-sampling period, several plankton tows dis- played large populations of Dinophysis, primarily D. norvegica. Other species of known toxic Dinophysis. D. rotunda and D. acuminata, were also observed but in low abundance. The Gulf of Maine Phytoplankton Monitoring Network has found Dinophysis species along the entire coast of Maine between June through September (W. Norden, pers comm). However, in the present study, Dinophysis rich samples were consistently negative in both the protein phosphatase inhibition assay as well as LC-MS/MS. In some European and Canadian outbreaks of DSP. Dinophysis abun- dance has not always been correlated to these events. The locations where Mytilus tested positive for OA-like activity correlated with locations where the epiphytic dinoflagellute, P. lima, was found. This is the first study to show this toxic di- noflagellate in northern United States coastal waters. Outbreaks of DSP from the Atlantic coast of Canada have been attributed to P. lima, rather than Dinophysis (Marr et al. 1992, Lawrence et al. 1998). P. lima is one of the most wide spread toxic dinoflagellates, being found in both tropical areas as well as cold temperate areas of the Atlantic. Pacific, and Indian Oceans. However, its presence has not been documented in the intervening latitudes along the eastern seaboard of the United States. The cell densities found for P. lima in this current study are lower than that generally found in tropical areas (Bomber 1985: Carlson and Tindall 1985, Morton and Faust 1997). Collections of microalgae associated with the filamentous mac- rophyte. Ectocarpus. rich with P. lima, were found to possess protein phosphatase inhibitory activity. Extracts of these field col- lected P. lima contained DTX- 1 , but OA was not detected. DTX- 1 was also found to be the primary toxin produced by P. lima in Nova Scotia, Canada, with approximately 10-fold lower amounts of OA than DTX-1 (Marr et al. 1992). Environmental conditions such as temperature, light intensity, salinity, and N:P ratio have been shown to have a direct effect on the growth rates and toxin Evidence of Shellfish Poisoning 685 OA092999a-002/Extract DTX-002 • 9/29/99 - 1] 18 AM PL2 100 75 0 ii' i , i nuiiifiUiiiliy lil mmimjmypiuidi II il Ml IIUIHUIIIUl ilifJumkil I lil j^kiUik * 100 75- £ 50 « 25 0 lil hi il an |m lilt Um^MiImM na,ii mihi ,n I m . ;r""tt ir-'.rinih i idiittM 100 „ 75 E 50 0' 100-, _ 75- E SO- * 25- iip I. in ii M^iffM^t^m^mJK^Mim**'^*"' il ^-^-'■^■■■^ -m^-i.-Mj.. 25.850 0.0 1 7,5 547 22 5 1638 25.0 1820 27 5 2001 30.0 2183 Time (mini/Scan Figure 4. LC-MS/MS chromatograph of dinophvsis toxin-1 from P. lima. The traces reflect selected ion monitoring of distinctive fragments ions at M/Z of 765, 783. and 801 as well as the parent ion of 819. ratios of P. lima and other closely related toxic species of P. lima (Morton and Norris 1990, McLachlan et al. 1994, Morton et al. 1994). Clonal isolates of P. lima from field collections are cur- rently being grown in large scale laboratory cultures for further study on toxin production to identify which DSP-toxins are produced. The results of this study identify DSP-like activity in Maine coastal waters. P. lima and DSP toxins were found only in the Frenchman Bay-Eastern Bay region. In the current study, levels of OA-like activity found in mussels were not at levels that present a public health issue. However, the presence of DSP as a potential problem in Maine coastal waters must be considered in future management decisions regarding shellfish aquaculture. Maine's shellfish industry is based upon a publicly owned resource with limited aquaculture lease sites for the bottom culture of mussels and scallops. DSP became a problem in Canadian maritime waters when the practice of raft culture was introduced, creating a suitable substrate for the growth of Pilayella littoralis. This filamentous brown macrophyte species commonly fouls aquaculture lines in that region and serves as a host for the epiphytic dinoflagellate, P. lima (Lawrence et al. 1998). Thus, as aquaculture activity in Maine expands, the addition of DSP to the state shellfish-monitoring program may become important for protection of public health. The sampling sites where both P. lima and OA-like activity in Mytilus were protected embayments with muddy to gravel sub- strates which supported the growth of Ectocarpus. Other sites along the Maine coast with a similar substrate may also support the growth of Ectocarpus or other filamentous brown macrophytes. which may serve as a host for P. lima. Clearly, further ecological studies are required to determine the extent of P. lima along the coast of Maine. These sites differ substantially in character and location from the primary monitoring sites used by the State of Maine's PSP monitoring program. This, in addition to the benthic behavior of the causative organism, P. lima, indicates that a dif- ferent sampling strategy than that currently employed by the state for the long established PSP monitoring program maybe required to monitor for occurrences of DSP. ACKNOWLEDGMENTS This study was supported by funds from NOAA National Ocean Service and by a grant from the MDIBL NIEHS Marine and Freshwater Biomedical Sciences Center to FMVD. 686 Morton et al. Figure 5. Empty theca of P. lima from the digestive gland of Mytilus edulis. I 111 K\l( Kl (III I) Bomber, J. W. 1985. Ecological studies of benthic dinoflagellates associ- ated with ciguatera from the Floriday Keys. M.S. thesis. Florida Insti- tute of Technology. Melbourne. FL. 104 pp. Carlson, R. D. & D. R. Tindall. 1985. Distribution and periodicity of toxic dinoflagellates in the Virgin Islands. In D. M. Anderson. A. W. White. & D. G. Baden (eds) Toxic Dinoflagellates. Elsevier, New York, pp. 171-176. Fernandez. M. L., A. Miguez. E. Cacho, & A. Martinez. 1996. Detection of okadaic acid esters in the hexance extracts of Spanish mussels. Toxicon 34:351-359. Kat, M. 1985. Dinophysis acuminata blooms, the distinct cause of Dutch mussel poisoning. In. Anderson. D. M.. A. W. White, & D. M. Baden (eds) Toxic Dinoflagellates. Elsevier, Amsterdam, pp. 73-77. Kumagai. M. T. Yanagi, M. Murata. T. Yasumoto. M. Kat, P. Lassus, & J. A. Rodriguez-Vazquez. 1986. Okadaic acid as the causative toxin of diarrhetic shellfish poisoning in Europe. Agric. Biol. Chem. 50:2853- 2857. Lawrence, J. E.. A. G. Bauder. M. A. Quilliam. & A. D. Cembella. 1998. Prorocenlrum lima: a putative link to diarrhetic shellfish poisoning in Nova Scotia, Canada. In Reguera. B, J. Blanco, M. L. Fernandez & T. Wyatt (eds). Harmful algae. UNESCO, pp. 78-79. Marr. J. C. A. E. Jackson. & T. L. McLachlan. 1992. Occurrence of Pro- rocenlrum lima, a DSP toxin-producing species from the Atlantic coast of Canada. J. Appl. Phycol. 4: 1 7-24. McLachlan, J. L„ J. C. Marr. A. Conlon-Kelly. & A. Adamson. 1994. Effect of nitrogen concentrations and cold temperatures on DSP-toxin concentrations in the dinoflagellate Prorocenlrum lima ( Prorocentrales. Dinophyceae). Nat. Toxins 2:263-270. Morton. S. L. & M. Faust. 1997. Survey of toxic epiphytic dinoflagellates from the Belizean barrier reef ecosystem. Bull. Mar. Sci. 61:899-906. Morton. S. L. & D. R. Norris. 1990. Role of temperature, salinity, and light on the seasonality of Prorocentrum lima (Ehrenberg) Dodge. In Graneli, E. B. Sundstrom. L. Edler. & D. M. Anderson (eds) Toxic Marine Phytoplankton. Elsevier, New York. pp. 201-205. Morton. S. L.. J. W. Bomber. & P. T. Tindall. 1994. Environmental effects on the production of okadaic acid from Prorocentrum hoffinannianum Faust: temperature, light and salinity. J. Exp. Mar. Ecol. 178:67-77. Quilliam, M. A.. M. W. Gilgan, S. Pleasance. A. S. W. deFreitas, D. Doug- las. L. Fritz. T. Hu. J. C. Marr. C. Smyth. & J. L. C. Wright. 1990. Confirmation of an incident of diarrhetic shellfish poisoning in eastern Canada. In Gordon, D. C. (ed). Proc. 2nd Canadian Workshop on Harmful Marine Algae. Can. Tech. Rep. Fish. Aquat. Sci. No. 1799: 18-22. Shumway, S. E.. S. Sherman-Caswell. & J. W. Hurst, Jr. 1988. Paralytic shellfish poisoning in Maine: monitoring a monster. J. Shellfish Res. 7:643-652. Suzuki, T., H. Ota, & M. Yamasaki. 1999. Direct evidence of transforma- tion of dinophysistoxin-1 to 7-O-dinophysistoxin-l (dinophysistoxin-3) in the scallop Patinopecten yessoensis. Toxicon 37:187-198. Tubaro. A.. C. Florio. E. Luxich, S. Soso. R. Delia Loggia. & T. Yasumoto. 1996. A protein phosphatase 2A inhibition assay for a fast and sensitive assessment of okadaic acid contamination in mussels. Toxicon 34:743- 752. Wright. J. L. C. & A. D. Cembella. 1998. Ecophysiology and biosynthesis of polyethers marine biotoxins. In Anderson. D. M., A. D. Cembella. & G. M. Hallegraeff (eds) Physiological Ecology of Harmful Algal Blooms. NATO ASI Series. Vol. G41. Springer- Verlag. Berlin, pp. 427-J74. Yasumoto. T., Y. Oshima. & M. Yamaguchi. 1978. Occurrence of a new type of shellfish poisoning in the Tohoku district. Bull. Japan Soc. Sci. Fish. 44:1249-1255. Journal of Shellfish Research. Vol. 18, No. 2. 687-700. 1999. ABSTRACTS OF TECHNICAL PAPERS Presented at The Sixth International Littorinid Symposium Hofstra University Marine Laboratory Priory. Jamaica, W.I. July 24-31. 1999 Meeting Organizers: Dr. Joseph C. Britton Department of Biology P.O. Box 298930 Texas Christian University Fort Worth, Texas 76129, U.S.A. Dr. Robert F. McMahon Department of Biology Box 19498 The University of Texas at Arlington Arlington, Texas 76019. U.S.A. 687 Sixth International Littorinid Symposium Abstracts, July 24-31. 1999 689 CONTENTS T. C. Addy and L. E. Johnson The influence of spatial heterogeneity on the foraging of littorines 689 S. C. S. Andrade, V. N. Solferini and C. A. Magalhaes Genetic analysis of Brazilian populations of Nodilittorina lineolata (Prosobranchia: Gastropoda) 689 Robert Black and Michael S. Johnson Contrasting life histories and demographies of eight species of littorines at Ningaloo Reef, Western Australia 689 Elizabeth G. Boulding Regulation of wave-sheltered Littorina sitkana populations by pile perch, Rhacochilus vacca 690 Joseph C. Britton An introduction to Jamaican littorinids and littorinid habitats 690 M. Carballo, E. Roldn-Alvarez and C. Garcia Estimating genetic variances from wild data in Littorina saxatilis 690 J. T. Christensen, P. D. Jensen, P.-G. Saurian and P. Richard Tracing diets of three species of Littoraria using stable isotopes 690 S. Coughlan, J. Mercer, O. McClean, N. Connolly and G. Burnell An assessment of the potential for the sustainable development of the edible periwinkle [Littorina littorea L.) industry in Ireland 69 1 Raquel Cruz and Carlos Garcia Between-morphs comparisons of adaptive surfaces in a hybrid zone of Littorina saxatilis 69 1 Mark S. Davies and Peter Beckwith Winkle nutrition — the role of trail-following 692 Hans de Wolf, Ronny Blast, Machteld de Wit, Marcel Selens and Thierry Backeljau Possible effects of anthropogenic stress on the genetic and morphometric population structure of Littornia littorea 692 Antonio M. De Frias Martins Ellobiidae: Lost between land and sea 692 Deborah J. Gochfeld and Dwayne T. Minto When to move and where to go: Movement behavior of the Caribbean Littorinid, Cenchritis muricatus 693 Catriona Hassey Grazer biomass and biofilm standing stocks in the high intertidal on Hong Kong rocky shores 693 Paul A. Hohenlohe and Elizabeth G. Boulding Differentiating sibling species: Littorina scutulata and L. plena 693 Kerstin Johannesson Parallel evolution — Challenging taxonomy, cladistics and speciation theory 694 A'. Johannesson, M. Lejhall and N. Mikhailova DNA-allozyme coupling and shell dimorphism over microgeographic scale — further evidences of a complex habitat-related substructuring in the marine snail Littorina fabalis 694 Olive H. A. Lee Feeding ecology of the Hong Kong mangrove littorinids, Littoraria ardouiniana and L. melanostoma 694 C. A. Magalhaes Aggregation patterns of three species of periwinkles in the southeastern coast of Sao Paulo, Brazil 694 Robert F. McMahon Heat coma temperature and salinity relative to zonation pattern in Jamaica intertidal gastropods and chitons 695 C. D. McQuaid and A. Whittington-Jones Small scale variation in response of intertidal macroalgae to grazing by winkles and limpets 695 Peter J. Mill, Andrea P. Clarke, John Grahame and Dehnont C. Smith Aspects of shell shape in lagoonal littorinids 695 Dwayne Minton and Deborah Gochfeld Is life on a tropical rocky shore really so hard? 696 David G. Reid The mangrove littorinids: Evolution and adaptation in the genus Littoraria 696 Bronwen Scott and Kerry Neil Origin of Nerita atramentosa Reeve 1 855. a nerite of temperate rocky shores ( Neritopsina: Neritidae) 696 690 Abstracts, July 24-31, 1999 Sixth International Littorinid Symposium Maureen P. Small and Elizabeth M. Gosling Relationships and population structure of Littorina arcana Hannaford Ellis. L. compressa Jeffreys, and L. saxatilis (Olivi) in the British Isles using SSCPS of Cytochrome-B fragments 697 Delmont C. Smith Effect of temperature and desiccation on uric acid content of Littorina saxatilis 697 G. F. Warner Trans-zonal movements in winkles, Littorina littorea (L. ): Reasons and consequences 697 Craig S. Wilding, John Grahame and Peter J. Mill Correlation of morphological diversity with genetic diversity in the rough periwinkle Littorina saxatilis 697 R. F. Uglow and Gray A. Williams Variation in ammonia efflux rates with emersion of three Hong Kong Nodilittorina species 698 Brigitta Winnepenninckx, Thierry Backeljau and David Reid Nuclear ribosomal DNA sequences and the phylogeny of the Littorinidae 698 Sixth International Littorinid Symposium Abstracts, July 24-31. 1999 691 THE INFLUENCE OF SPATIAL HETEROGENEITY ON THE FORAGING OF LITTORINES. T. C. Addv and L. E. Johnson, G1ROQ, Universite Laval. Local 2056. Poste 2266, Pa- vilion, Vachon, Ste-Foy. Quebec G1K 7P4 Canada. The abundance and distribution of littorines often depends on the availability of topographic irregularities in the rock surface (e.g., holes and crevices). Presumably, the littorines use these mi- crohabitats as shelters during periods of unfavorable conditions (e.g., wave action, desiccation) which restricts their foraging ac- tivities to areas near shelters. This study examined several envi- ronmental factors that might influence the use of shelters by the littorine Littorina saxatilis along a gradient of wave exposure at Pointe Metis. Quebec (Canada). Specifically, we measured the percentage of animals inside and outside of both natural and arti- ficial irregularities (crevices and drilled holes, respectively) under a variety of climatic conditions experienced during low tide. The number of littorines outside of shelters was strongly correlated with wave exposure with only 5-10% outside of crevices in the most exposed locations compared with >70% outside at more pro- tected locations. Littorines were found outside of shelters mostly during nighttime low tides when the rocks remained wet for longer periods. During the daytime, higher proportions were found out- side during sunny conditions (the rock surface remained wet longer than under cloudy conditions, apparently due to the higher winds associated with cloudy days). Surprisingly, rainfall during the low tide had little influence suggesting that osmotic problems might negate any advantages of wetting the rock surface. Parallel laboratory experiments generally support these field patterns and together suggest that littorine foraging during low tide is largely determined by the climatic conditions. GENETIC ANALYSIS OF BRAZILIAN POPULATIONS OF NOD1L1TTOR1NA L1NEOLATA (PROSOBRANCHIA: GAS- TROPODA). S. C. S. Andrade, V. N. Solferini, Depto. de Gc- netica e Evolucao, IB, UNICAMP and C. A. Magalhaes, Depto. de Zoologia, IB, UNICAMP, Campinas. SP. Brazil. There are controversies about the number of species of the "Ziczac Complex" that inhabits Brazilian coast. Isozymic analyses (18 loci) were carried out on 13 different populations of those periwinkles to attempt to clarify the taxonomic question, and also to verify if there is correlation between their spatial distribution and variability patterns. Nodilittorina were sampled on one beach on the South coast (SO. 7 beaches on the Southeast coast (SP. RJ ). 2 on the Northeast (PE) and 3 on the North coast of Brazil, cov- ering a distance of about 6000 km. Genetic analysis showed that there is only one species in our samples, identified as Nodilittorina lineolata (D'Orbigny. 1840). The samples presented high variabil- ity (Hm = 0.17) and exhibited low genetic differentiation among them (Fst = 0.065). This is probably due to the fact that this species has long-living pelagic larvae. All samples presented heterozygote deficiency on at least 2 loci. These results suggest that these samples of Brazilian N. lineolata may be subject to strong selective pressures. The heterozygote deficiency and the high genetic variability could be related to high microhabitat het- erogeneity on the rocky shore, but new experiments are necessary to test this hypothesis. CONTRASTING LIFE HISTORIES AND DEMOGRAPHIES OF EIGHT SPECIES OF LITTORINES AT NINGALOO REEF, WESTERN AUSTRALIA. Robert Black and Michael S. Johnson, Department of Zoology, University of Western Aus- tralia. Nedlands, Western Australia, Australia 6907. This study considers four species of littorines from rocky shores. Nodilittorina australis, N. millegrana, N. trochoides, and Littoraria undulata, and four species of littorines from mangroves, Littoraria cingulata, L. filosa, L. scabra, and L. sulculosa over the period from July 1989 to June 1999. We conducted periodical censuses of the same replicate areas at four rocky shore sites spread over 50 km of shoreline, and at the same replicate man- grove trees at two sites at the opposite ends of a bay, about 2 km apart. This design allowed us to partition the variability in abun- dance of each species of snails into components associated with the sites, areas or trees within sites, years, areas x years, and the residual. The littorines of the rocky shores had greatest variability associated with differences among the four sites for the total popu- lations and among sites and year x site for recruits, while the littorines of the mangroves showed greatest variability associated with years, and sites x years. We also conducted shorter term mark and recapture studies to provide direct evidence about growth and survivorship of these snails. All three species of littorines from mangroves with suffi- cient recaptures showed rapid growth, reaching half their maxi- mum size in 0.94, 0.35, and 0.75 years for L. cingulata, L. filosa, and L. scabra, respectively, and attaining maximum lengths of 22 to 27 mm. The littorines of the rocky shores all had slower growth rates than the snails from the mangroves, and were all smaller in maximum size, from 10 to 17 mm. With the exception of N. millegrana, which took only 0.50 years to reach half maximum size, the snails on the rocky shores took much longer to reach half their maximal lengths than the littorines in the mangroves ( 1.23, 2.87. and 1.28 years for N, australis, N. trochoides, and L. undu- lata, respectively). Patterns in survival parallel the estimates of times to reach half maximal size. We never recaptured N. mil- legrana over intervals as long as 1 year, but some N. trochoides marked in 1988 were still alive in 1995. Between these extremes, the littorines of the mangroves seemed shorter lived than the rest of the littorines from the rocky shores. This long-term study of eight similar species in a single geo- graphic area highlights the variability in life histories and demog- raphy of littorines. 692 Abstracts, July 24-31, 1999 Sixth International Littorinid Symposium REGULATION OF WAVE-SHELTERED LITTORINA SIT- KANA POPULATIONS BY PILE PERCH, RHACOCHILUS VACCA. Elizabeth G. Boulding, Department of Zoology, Uni- versity of Guelph, Guelph, ON, NIG 2W1, Canada. The pile perch, Rhacochilus vacca (Embiotocidae) is abundant on the Pacific Coast of North America and is known to crush littorinid gastropods and other hard shelled prey with its heavy pharyngeal teeth. I investigated whether this predator had the po- tential to regulate the population density of Littorina sitkana, a direct-developing littorinid. on wave-sheltered shores. Laboratory feeding experiments at Bamfield Marine Station. B.C.. Canada showed that that pile perch were powerful and efficient predators on these snails. The fish required an average of only 19.1 (s.e. = 5.61, N = 20) seconds to crush and swallow a large snail and their laboratory consumption rates averaged 33.29 (s.e. = 6.27, N = 4) large snails per day. Even small adult fish (fork length: 21 cm) could crush the largest L. sitkana available (>1 1 mm shell width). Indeed, some fish showed a significant preference for large snails (shell width >6.3 mm) over small snails (3.35 mm < shell width <4.0 mm). Our observations with SCUBA during daytime high tides showed that the density of pile perch foraging in the intertidal averaged only 0.0592 (s.e. = 0.010, N = 20) fish per square metre (estimated fork lengths 5—10 cm). However, the intertidal distri- bution of fish which were actually consuming prey was highly aggregated. The fish swam parallel to shore and quickly located and consumed any high density patches of L. sitkana but were slower to locate and consume low density patches. This behaviour resulted in strongly density-dependent predation on the snails we deployed. I conclude that this predatory fish has the capability to regulate L. sitkana populations at least on this wave-sheltered shore. AN INTRODUCTION TO JAMAICAN LITTORINIDS AND LITTORINID HABITATS. Joseph C. Britton, Department of Biology, Box 298930, Texas Christian University, Fort Worth. Texas 76129. U.S.A. Jamaican rocky shores include natural, mobile, boulder, and cobble beaches, man-made boulder breakwaters, natural vertical sea cliffs, man-made concrete walls or ramps, and, especially, uplifted limestone platforms (microkarsted ironshore). all of which may harbor one to several species of littorinids, including Cen- chritis muricatus, Tectariits antonii, and at least five species of Nodilittorina. Littorinids dominate the upper rocky shore and share tidepools and the midlittoral with at least four species of nerites. two species of planaxids, two polyplacophorans, and at least one patellid, fissurellid, potamid, muricid and pulmonate. Unlike most temperate rocky shores, those of Jamaica, and the tropical Carri- bean in general, are noticeably deficient of chthamalid and bal- anoid barnacles. Jamaican mangroves are comprised of four distinctively zoned trees, the red mangrove. Rhizophora mangle, with prop roots usu- ally rising from the sea, the black mangrove, Avicennia germinans, with pneumatophores which are frequently inundated by tides, the white mangrove. Laguncularia racemosa, occasionally touched by tidal water and the buttonwood, Conocarpus erectus, occupying the highest position on the mangrove shore. There is only one species of littorinid, Littoraria angulifera, commonly found in the mangal, usually upon Rhizophora branches. The mangrove floor, however, is occupied by several species of ellobiids and melampids. ESTIMATING GENETIC VARIANCES FROM WILD DATA IN LITTORINA SAXATILIS. M. Carballo, E. Rolan- Alvarez, Facultad de Ciencias, Universidad de Vigo. 36200 Vigo, Spain, and C. Garcia, Facultad de Biologfa, Universidad de San- tiago de Compostela, 15706, Santiago de Compostela. Spain. On exposed Galician rocky shores, two ecotypes of Littorina saxatilis (RB versus SU) can be found at different shore levels. Parental ecotypes and intermediate individuals can be found in sympatry at a narrow mid-shore zone. These ecotypes differ in many morphological traits associated to different habitats, but the genetic basis of those differences are unknown. We assessed the inheritance of the morphological variability in these ecotypes by taking pregnant females from wild populations, and using off- spring-mother regression and correlation between full-sibs meth- ods. We sampled five levels of the shore in two localities (20 RB females from upper shore, 20 RB from upper mid-shore, 20 RB. 20 SU and 11-13 intermediate forms from mid-shore, 20 SU from lower mid-shore and 20 SU from lower shore). We took 1 1 shell measurements in every female and in three embryos per female. Genetic variances could be evidenced for some traits and samples. Genetic variance estimates were usually similar between full-sibs correlation and offspring-mother regression methods. Fur- thermore, the coefficients of genetic variance were similar, but somewhat smaller (range = 0.5-7%). to those published in the literature. However, correlation between full-sibs rendered more frequent significant estimates than offspring-mother regression, and there was a pattern which related the level of genetic variance with shore level. These results suggest possible bias in the esti- mates. The present methods may be a valuable tool for estimating genetic components from wild data, but some caution is needed during the interpretation of the results. TRACING DIETS OF THREE SPECIES OF LITTORARIA USING STABLE ISOTOPES. J. T. Christensen, P. D. Jensen, Dept. of Marine Ecology, Institute of Biological Sciences, Univer- sity of Aarhus. Finlandsgade 14, DK-8200 Aarhus N, Denmark, and P.-G. Sauriau, P. Richard, Centre de Recherche en Ecologie Marine et Aquaculture de L'Houmeau. CREMA (CNRS- IFREMER UMR 10). BP 5, F-17137 L'Houmeau. France. Analyses of carbon and nitrogen stable isotopes were used in an attempt to trace food sources of three Littoraria species in Rhizo- phora apiculata mangrove in Thailand. When occurring together Sixth International Littorinid Symposium Abstracts. July 24-31. 1999 693 Littoraria scabra, L. intermedia, and L. pallescens exhibit a ver- tical zonation with L. scabra on prop roots nearest the water. L. intermedia on roots and branches, and L, pallescens primarily on mangrove leaves. Ranges do, however, overlap and the animals move with the tide. Rhizophora leaves, scrapings from both leaf and prop-root surfaces and local particulate organic matter (POM) were well separated on the basis of their 8'3C and 815N values. In contrast, tissue from the three species of Littoraria showed con- siderable overlap and scatter in isotope ratio values suggesting that they are opportunistic feeders sharing similar food resources. The wide range of 813C values (-17.3%o to -26.3%o suggest carbon assimilation from multiple sources (epiphytes from leaves and prop roots, suspended POM and Rhizophora detritus). L. interme- dia and L. pallescens, the smallest species, had identical mean 813C values while L. scabra was significantly more 13C depleted. A diet of microalgae and cork cells from prop roots could explain this pattern with L. scabra, being larger, consuming relatively more cork cells. However, all three species were on average more 1:,N depleted than these food sources. Slightly more than one third of the L. scabra and L. intermedia individuals had S'^N values consistent with such a diet while the remaining and all L. pal- lescens were too depleted. L. pallescens had a significantly lower mean 8KN value than the other two species and must derive a significant amount of its food from a strongly l5N depleted source. Such a source was present in scrapings from leaf surfaces (815N = 0.3 ± 0.05%o, n = 2) and could represent N2-fixing microorgan- isms. Due to the wide scatter of Littoraria S'^N values from 6.1 down to -l%o, it is hypothesised that all three species assimilate microorganisms such as cyanobacteria utilizing ammonia as sub- strate, bacteria involved in denitrification and fungi. These find- ings were compared with microscopic analysis of gut contents. AN ASSESSMENT OF THE POTENTIAL FOR THE SUS- TAINABLE DEVELOPMENT OF THE EDIBLE PERI- WINKLE {LITTORINA LITTOREA L.) INDUSTRY IN IRE- LAND. S. Coughlan, J. Mercer, Shellfish Research Laboratory, Carna, Co. Galway. Ireland. O. McClean, N. Connolly, Coastal Resources Centre, National University of Ireland, Cork. Ireland, and G. Burnell, Dept. of Zoology, National University of Ireland, Cork, Ireland. A study covering biological and socio-economic aspects of the edible periwinkle industry in Ireland was undertaken as a reaction to widespread reports of a decline in periwinkles of marketable size. This research includes an investigation of the distribution, biomass and age/size frequency of periwinkles. Sites selected were those thought to hold exploitable quantities of winkles. Samples were taken from each site and retained for measurement to deter- mine biomass and to investigate geographic variation in morphom- etry. It is hoped this information together with results of extensive interviews with those involved in the industry, will be used to formulate a management strategy for the species. Consultations with wholesalers and harvesters suggest that a summer closed season would be the most widely supported of the proposed regu- lations. So far, seventy five sites have been visited; it appears that greater densities occur on the south coast on moderately exposed shores. In line with reports by other authors, morphometric ratios (i.e.. ratios between shell height, aperture height, shell width and aperture width) appear to be influenced by physical factors such as exposure. BETWEEN-MORPHS COMPARISONS OF ADAPTIVE SURFACES IN A HYBRID ZONE OF LITTORINA SAXATI- LIS. Raquel Cruz and Carlos Garcia, Departamento de Bioloxfa Fundamental, Facultade de Bioloxfa. Universidade de Santiago de Compostela. 15706 A Coruna, Galicia. Spain. Littorinid snails having direct development, such as Littorina saxatilis, can be very useful in experimental studies of evolution- ary problems, due to their very limited dispersal ability, their rela- tively high population densities, and to the fact that individuals are easy to find in the field. This species shows also many polymor- phisms of different kinds, which makes it an interesting subject to study the role of genetic variability in natural populations. We studied one of these polymorphic populations found on the exposed rocky shores of Galicia. NW Spain, where there is one morph adapted to the upper shore, and a another adapted to the more wave-exposed lower shore. We sampled 1 158 adult females on all shore levels, and counted the number of normal embryos in their embryo pouches, as an estimate of their fecundity, along with measurements of 26 morphological characters and nine environ- mental variables. We calculated two discriminant functions to separate the two pure morphs. The first was based on the morpho- logical measurements, and the second, on the environmental ones. Then we used the fecundity measurements to adjust a fitness sur- face to the bidimensional space defined by these two discriminant functions, to study the possible role of natural selection (acting through the variation in fecundity) in the differentiation observed between the pure morphs, and that between the pure morphs and their hybrids. We found some evidence for a fecundity depression in a section of the hybrid's surface, situated between the areas of the bidimensional space corresponding to each pure morph. The depression was not detected when the average fecundity of all hybrids was compared with those of the pure morphs. which would indicate that this average, as it is usually calculated, is not enough to describe the adaptive situation in a hybrid zone, because it does not take into account the position of the sampled individuals along the axis of differentiation between the pure morphs, and as our study shows, this axis does not need to coincide with the mere physical distance on the shore. The study of the interplay between a fecundity depression of hybrids, which would act to keep the pure morphs separate, and the gene flow between them, tending to join them in a single population, could perhaps improve our un- derstanding of the processes of parapatric speciation. In addition. 694 Abstracts, July 24-31. 1999 Sixth International Littorinid Symposium our results indicate that the variation in fecundity is not enough to explain the maintenance of this hybrid zone, because the sampled individuals were not distributed around the fecundity optimums expected for their morphs. Other fitness components, as viability, would be also important for this maintenance. WINKLE NUTRITION— THE ROLE OF TRAIL- FOLLOWING. Mark S. Davies and Peter Beckwith, Ecology Centre. University of Sunderland. Sunderland. SRI 3SD. U.K. Gastropod locomotion typically involves deposition of a mucus trail which is energetically costly. Post-deposition function of the trail would defray costs. We aimed to assess the role in nutrition of the mucus trail of Littorina littorea. Mucus trails adhered more microalgal cells from suspension than did a glass substratum: Am- phora coffeaeformis (pennate diatom) adhered in greater densities than did Tetraselmis suecica (flagellate prasinophyte). Trails con- taining microalgae of both species (50-100 cells mm-2, similar to in situ) induced more trail-following than bare trails. Conspecific trails were followed for longer than self-laid trails. Winkles moved significantly slower on bare trails (mean = 0.35 mm s"1) than on glass (0.68 mm s"1). though the addition of microalgae to trails increased the speed. The radular rasp rate was significantly in- creased on trails containing A. coffeaeformis (mean = 17.8 bites min~') and T. suecica (12.9) in comparison to control trails (4.3). Microalgae embedded in mucus entered the mouth. Following the passage of a winkle. A. coffeaeformis density was reduced by 38% and T. suecica by 43%. Winkles can clearly exploit food in trails and in doing so modify trail-following and feeding behaviour. Trail-following seems inextricably linked to nutrition. Since the intertidal is likely to be covered with mucus, or its degradation products, experiments on trail-following using clean substrata may have been erroneous. Distribution patterns of both L. littorea and benthic microalgae might be shaped by the ability of trails to adhere microalgae and by their subsequent exploitation by the grazer. POSSIBLE EFFECTS OF ANTHROPOGENIC STRESS ON THE GENETIC AND MORPHOMETRY POPULATION STRUCTURE OF L1TTORNIA LITTOREA. Hans de Wolf. Ronn.v Blust, Machteld De Wit. Marcel Selens, University of Antwerp (RUCA) Ecophysiology & Biochemistry. Groenenborg- erlaan 171, B-2020 Antwerp. Belgium, and Thierry Backeljau, Royal Belgian Institute of Natural Sciences. Vautierstraat 29, B-1000 Brussels, Belgium. Anthropogenic stressors introduced into the marine system, are primarely concentrated in coastal areas, from industrial and agri- cultural inputs and runoff from land. Recent ecogenetic studies have shown that anthropogenic stressors may affect the genetic constitution of a population, decreasing its variability and making it more vulnerable for extinction. In the present preliminary study we analysed the correlations of heavy metal pollution on the genetic and morphometric population structure of Littorina littorea, collected at seven sites along the highly polluted Western Scheldt estuary. Heavy metal levels (Ag, Al, As, Ca, Cd, Co, Cr, Cu, Fe, Hg, Mn, Ni, Pb, Sr and Zn) were determined in both the shells and the soft body parts of L. littorea, using Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES) and Inductively Coupled Plasma Mass Spectrometry (ICP-MS). Esterase variation was analysed using isoelectric focusing (IEF). while the morpho- metric population structure was determined by simple shell mea- surements using calipers. None of the analysed metals seemed to accumulate in the speci- men's shell matrix, but some metals accumulated in the soft body parts (i.e.. Fe, Cd, Mn, Zn, and Cu), showing a clinal concentration gradient towards the North Sea. Shell size and esterase variability increased with decreasing heavy metal pollution (i.e., increasing salinity). While the current results indicate a structuring of the genetic and morphometric variability in populations of L. littorea collected at the Western Scheldt, they do not indicate whether this structur- ing is caused by the existing heavy metal and/or the salinity gra- dient. ELLOBIIDAE: LOST BETWEEN LAND AND SEA. Antonio M. De Frias Martins, Departamento de Biologia, Universidade dos Acores. Ponta Delgada, Sao Miguel, Acores. Portugal. The Ellobiidae are a diverse group of archaeopulmonate snails living mostly near the sea, except for the Carychiinae which live in mountain forests and inside caves. The halophile ellobiids are commonly associated with the supratidal fringe of mangroves and salt marshes; they are, also, important components of the su- pratidal biota of the mobile rocky shore, namely under piles of rubble and shells. Their ecological requirements range from the muddy bottoms of stagnant mangrove embayments to the wave- swept crevices of exposed cobble shores, and they could be very important elements in the mineralization of the organic mud or algal detritus in such habitats. The various species of ellobiids occupy different portions of the shore, their distribution being loosely related to their taxonomic affiliation. The pythiinians (Pythia and. to a lesser extent, Myoso- tella) venture farther inland and may live in a nearly terrestrial environment, whereas the Pedipedinae prefer the upper intertidal. Within the Melampinae, some species of Melampus are found usually in mangroves and marshes, whereas others and Tralia prefer cobble shores. In rocky habitats, in addition to horizontal zonation, the ellobiids also partition their vertical distribution. The Ellobiidae range in size from barely 2 mm (Leuconopsis) to over 100 mm (Ellobium); the shell is variable in shape and can not be used as a reliable character for subfamilial allocation. The anatomy is also variable, but presents more reliable diagnostic Sixth International Littorinid Symposium Abstracts, July 24-31, 1999 695 characters. Six structural types of reproductive system set the boundaries for the subfamilia] division, which is supported by the existence of three structural types of nervous system. The internal morphology of the penial complex has proven to be a helpful associated character in the subfamilia] delimitation. WHEN TO MOVE AND WHERE TO GO: MOVEMENT BE- HAVIOR OF THE CARIBBEAN LITTORINID, CENCHR1- TIS MURICATUS. Deborah J. Gochfeld1 and Dwayne T. Minion,2 Hofstra University Marine Laboratory, Priory, Jamaica, West Indies. Intertidal molluscs display a variety of movement behaviors, and the type of behavior expressed by a species is related to the intensity of physical and biological stresses to which it is exposed. Although littorinids are believed to be non-homing and to display random movement patterns when foraging, few studies have ex- amined their behavior in detail. Tropical littorinids are exposed to extreme environmental conditions, and movement at night or on a tidal cycle may alleviate desiccation and heat stress, as would returning to a sheltered location following foraging. We examined the movement behavior of the Caribbean lit- torinid. Cenchritis muricatus, for cyclic migrations, homing, and non-random movement. Individual snails were marked and their locations on an uplifted limestone bench were monitored daily for 30 days. On six days selected from different parts of the lunar cycle, snail locations were monitored at three-hour intervals. Movement did not occur on a diurnal or tidal cycle, however, we did observe a mass down-shore migration in June and July. Al- though C. muricatus preferentially rested in crevices, there was no evidence of homing behavior. Snails resting on exposed rock sur- faces were four times more likely to move than snails resting in more sheltered microhabitats. In general, movement in C. murica- tus appears to be a direct response to wetting: >87% of movements occurred within 12 hours after rainfall or heavy dew, but move- ment was elicited by both fresh and salt water. We believe this behavior is primarily a response to desiccation stress, but it may also facilitate foraging. 'Present address: Department of Pharmacognosy. University of Mississippi, University, MS 38677, U.S.A. 2Present address: Department of Zoology, University of Ha- waii, Honolulu. HI 96822. U.S.A. GRAZER BIOMASS AND BIOFILM STANDING STOCKS IN THE HIGH INTERTIDAL ON HONG KONG ROCKY SHORES. Cairiona Hassey, Ecology Centre, University of Sun- derland, Chester Road, Sunderland, SRI 3SD. U.K. Bottom-up processes and resultant net primary production rates are known to influence grazer biomass and subsequent community structure on rocky shores. Variation in abiotic conditions (e.g., coastal water nutrient levels, rock type, etc.). have also been shown to influence algal production rates and, therefore, consumer bio- mass. High intertidal, splash zones worldwide support large den- sities of Littorinidae although these grazers often appear to be existing on bare rock. These zones do, however, support an epilithic biofilm which forms the base of the benthic food chain and maintains populations on the high shore. Refuge habitats are also known to be important for these snails in this physically harsh environment. The present study investigates the relationship be- tween grazer biomass and biofilm standing stock in the high in- tertidal/splash zone on Hong Kong rocky shores. The sampling zone (2.0 — 3.0 m above CD.) was defined according to the im- mersion:emmersion curve to lie above the 10% immersion line and to extend to the upper limit of the littorinids. Two littorinid species were common in this zone, Nodilittorina trochoides (Gray 1939) and N. radiata (Eydoux and Souleyet. 1852) which reached den- sities in excess of 500/nr and 300/m2, respectively. These species are microalgal grazers feeding on the epilithic biofilm. composed primarily of cyanobacterial species such as Gloeocapsa and Der- mocarpa. Chlorophyll a extraction of rock chips was used as an estimate of standing crop and the fractal dimension of the rock surface was used to estimate habitat complexity. These factors were correlated with littorinid biomass and abundance at a number of different spatial scales (both within and between shores) to investigate whether they had any predictive value and their spatial generality. DIFFERENTIATING SIBLING SPECIES: LITTORINA SCUTULATA AND L. PLENA. Paul A. Hohenlohe, Friday Har- bor Laboratories, University of Washington, Friday Harbor WA 98250. U.S.A.. and Elizabeth G. Boulding, Department of Zool- ogy, University of Guelph. Guelph. ON NIG 2W1, Canada. The sibling species Littorina scutulata and L. plena are difficult to distinguish nondestructively in the field. Here we present a new molecular tool and an analysis of the quantitative and qualitative characters that have been proposed as diagnostic. We collected 385 snails from 1 1 sites in Washington state and scored three qualita- tive shell characters (presence of basal ridge, presence of basal band, and size of tesselation pattern) as well as tentacle coloration and male penis morphology. Specimens were identified to species by amplification and restriction enzyme digestion of a 480 base pair fragment of the mitochondrial cytochrome b gene. Of the qualitative characters, penis morphology was infallible for differ- entiating the males and tentacle coloration was most effective for both sexes. The qualitative shell characters also showed significant differences between the species but are not likely to be effective for differentiation because of frequent damage to the shells from abrasion and fungus. We also took eight quantitative measurements of shell shape. L. scutulata had significantly larger shells than L. plena. After cor- recting for size. L. scutulata had significantly taller, narrower ap- ertures and spire angles, but the species did not differ in aperture 696 Abstracts, July 24-31, 1999 Sixth International Littorinid Symposium angle. Discriminant function analysis of these measures correctly assigned only 72.5% and 64.8% of the two species, respectively. Contrary to previous reports of habitat differences, only L. plena was found at exposed outer-coast rocky shores, while both species co-occurred at low and medium exposure rocky shores and cobble beaches within Puget Sound. PARALLEL EVOLUTION— CHALLENGING TAX- ONOMY, CLADISTICS AND SPECIATION THEORY. Ker- stin Johannesson, Department of Marine Ecology. Tjarno Marine Biological Laboratory, SE-452 96, Stromstad. Sweden. Under parallel evolution, similar traits arise independently in different areas of a species. They do so because they have the same evolutionary background and they respond to similar local selec- tion pressures. However, in heterogeneous environments this will result in parallel ecotypes among which phenotypic similarity and phylogenetic ancestry are incongruent. That is. we find common phenotypic traits in populations that live in similar but geographi- cally separated habitats, despite these populations being geneti- cally less related than phenotypically disparate populations living in adjacent but dissimilar habitats. Over the last two decades a multitude of illustrative examples of parallel evolution have been documented in, for example, di- atomes. higher plants, mollusks. fishes, and salamanders. In some of these cases, the originally related populations have diverged to such an extent as morphological or behavioural barriers to repro- duction have developed. However, because of phenotypic similar- ity, these barriers do not exist between the phenotypically similar populations evolved in parallel (coined "parallel speciation"). Par- allel species and parallel evolution raises conceptual problems of phylogenetic analysis, species definitions and theory of speciation. More problematic, however, may be the dramatic consequences this type of evolution may have on traditional taxonomy and sys- tematics based on external morphological characters. In this paper I will discuss these problems using data from species of Littorina, and from some other taxa, as examples. DNA-ALLOZYME COUPLING AND SHELL DIMOR- PHISM OVER MICROGEOGRAPHIC SCALE— FURTHER EVIDENCES OF A COMPLEX HABITAT-RELATED SUB- STRUCTURING IN THE MARINE SNAIL LITTORINA FA- BALIS. K. Johannesson, M. Lejhall, Department of Marine Ecology, Tjarno Marine Biological Laboratory. SE-452 96. Strom- stad. Sweden, and N. Mikhailova, Institute of Cytology RAS, St. Petersburg 194064. Russia. Genetic substructuring of species is probably a quite important factor in evolution but still we have too fuzzy data to test this hypothesis in most species. Obviously, we expect genetic substruc- turing primarily in species of poor dispersal, for example, marine benthic invertebrates having direct development. Mostly we ex- pect substructuring to be due to isolation by distance, but this is not always the case. In Wales, Littorina fabalis is dimorphic for size, colour and shell mierostructure between exposed and protected habitats. We assessed the differences in Swedish populations, and found that these were less pronounced, except for size. However, earlier stud- ies reveal strong allozyme differences in one locus. In this study we report DNA differences as well. Furthermore, the DNA and allozyme differences are strongly correlated which might be ex- plained by a chromosomal inversion involving both markers. The genetic substructuring of Littorina fabalis is also strongly habitat linked, but it remains to be tested if there are adaptive differences between different genotypes. An alternative may be two indepen- dent evolutionary lineages. FEEDING ECOLOGY OF THE HONG KONG MANGROVE LITTORINIDS, LITTORARIA ARDOUINIANA AND L. MELANOSTOMA. Olive H. K. Lee, Department of Ecology and Biodiversity, The University of Hong Kong. Pokfulam Road. Hong Kong. Littoraria ardouiniana and L. melanostoma are commonly found in Hong Kong mangroves where they occur on the leaves and trunks of mangroves such as Kandelia candel, Aegiceras cor- niculatwn, and Avicennia marina. Examination of the stomach and faecal contents showed that these two littorinid species mainly feed on the epidermal cells of leaves, while bark cells, fungal fragments and spores were also found. In contrast to studies from Thailand, fungal hyphae and spores were sparse in the diet. Monthly investigation revealed no seasonal variation in the diet nor in the availability of food items (particularly fungi) on the leaves. Laboratory and field-based observations showed that these littorinids were generally inactive when dry in the daytime but active at night and/or in rain. Activity could be stimulated by spraying with water. L. ardouiniana and L. melanostoma appear to be generalists in their diet, grazing on the leaf and bark surfaces of mangroves at night. Feeding activity can also, however, be trig- gered by environmental cues (e.g.. rain), suggesting that these species are opportunists, feeding whenever conditions are suitable. AGGREGATION PATTERNS OF THREE SPECIES OF PERIWINKLES IN THE SOUTHEASTERN COAST OF SAO PAULO, BRAZIL. C. A. Magalhaes, Depto de Zoologia, IB. UNICAMP. Campinas. SP. Brazil. Periwinkles tend to aggregate and remain inactive under dif- ferent stressful conditions for grazing. In this work I studied the frequency and duration of aggregation behavior, the size of the aggregations, the fidelity to a specific aggregation, and aggrega- tion arrangement in Nodilittorina lineolata, Littoraria flava. and Littoraria angulifera. N. lineolata aggregates in great numbers and for periods of until 25 days, usually associated to dry conditions of Sixth International Littorinid Symposium Abstracts, July 24-31, 1999 697 the rock surface or with excessive freshwater seeps during the rainy season. Aggregation is fortuitous in this species and there is no special arrangement of the individuals in the cluster. L. flava exhibits a tidal aggregation behavior, forming small groups com- posed of 10 to 30 snails, usually associated with depressions or crevices on the rocky shore. Aggregation is due to dry conditions, and some snails do present fidelity to the aggregation site, follow- ing back their mucus tracks back to their aggregation site after a foraging excursion. The smaller individuals occupy the central area of the cluster, surrounded by the large ones. L. angulifera, a mangrove inhabitant, presents an aggregation behavior similar, but less frequent and ubiquitous than L. flava. It aggregates over man- grove tree's stems, close to the low water tidal line during extreme spring low tides, or high in the trees on high tides. Snails of this species search for depressions or fissures to begin aggregates, al- ways on shadowed spots. Aggregation of individuals is random, with no apparent mucus following clues. HEAT COMA TEMPERATURE AND SALINITY RE- SPONSE RELATIVE TO ZONATION PATTERN IN JA- MAICAN INTERTIDAL GASTROPODS AND CHITONS. Robert F. McMahon, Department of Biology, Box 19498, The University of Texas at Arlington, Arlington, Texas 76019, U.S.A. Intertidal mollusc vertical distributions are assumed to be re- lated to species tolerances of temperature and desiccation. How- ever, species physiologic resistences exceed intertidal environmen- tal extremes, suggesting that they have little influence on zona- tion patterns. In contrast, day-to-day maintenance of zonation could be influenced by salinity variation which increases with shore height due to evaporation and to precipitation during periods of tidal emersion. The relationships between zonation, heat coma temperature (HCT) and behavioral response to varying salinity were tested among molluscs from a rocky shore near Priory, Ja- maica. HCT was determined as temperature of loss of pedal at- tachment on warming at 1 °C 5 min"1 while response to salinity was assessed by pedal re-attachment 1 hour after transfer into 0, 5, 10. 20, 30, 40, 50. 70. 100. 120, 160. 180. 200. or 220% seawater. Of 15 examined species, two low eulittoral chitons. Chi- ton squamosus and Acanthopleura granulata, had the lowest HTC's at 35.14 °C (s.e. = 0.35) and 37.03 °C (0.48), respectively. Among five low to mid-eulittoral nerites. HCT ranged from 41.5° (0.27) to 45.7 °C (0.15) and was not correlated with vertical dis- tribution. Batillaria minima, a mid-eulittoral tide pool cerithiid had a similar HCT of 43.04 °C (0.14). Among seven littorinid species, HCT ranged from 42. 1 °C (0.41 ) in the maritime, Cenxhritus mu- ricatus, to 47.0 °C (0.52) in the mid-eulittoral. Notilittorina ziczac. Thus, thermal tolerance was not correlated with zonation level among the rock dwelling species tested. In contrast, among six littorinid, five nerite and I cerithiid species, salinity tolerance was correlated with zonation and microhabitat. Among tide pool spe- cies, the highest zoned, Purperita pupa, pedally re-attached over the greatest salinity range. Among eight species inhabiting ex- posed rocks, those highest zoned pedally re-attached over the wid- est salinity ranges, and particularly under hyposaline conditions, with the highest-zoned Cenchritis muricatus reattaching at 01% salinity. Thus, salinity tolerance may influence maintenance of zonation. high-shore molluscs being capable of maintaining activ- ity over a greater range of salinities than low-shore species. Ob- servations on the shore supported this conclusion. SMALL SCALE VARIATION IN RESPONSE OF INTER- TIDAL MACROALGAE TO GRAZING BY WINKLES AND LIMPETS. C. D. McQuaid and K. Whittington-Jones, Dept. of Zoology & Entomology, Rhodes University. Grahamstown 6139. South Africa. A detailed exclusion experiment was set up to examine the effects of different size classes of grazers on algal biomass on a rocky shore in South Africa. Grazers were classified as "macro- grazers" (mainly limpets) and "mesograzers" (mainly winkles). There were seven treatments involving different designs of cages of 50 x 50 cm, with five replicates of each treatment. This allowed examination of the effects of each grazer type alone and in com- bination. The experiment was run in two seasons and at two shore heights. Repeated measurements were made at 2-A week intervals of the cover of several functional groups (foliose algae, algal turfs, encrusting algae, barnacles). The most important result was that variation among replicate treatments was enormous. Replicates which began with similar biomass of macroalgae responded completely differently to the same treatment. This produced large standard deviations which masked treatment effects. Analysis was by MANOVA, followed by ANOVA where this indicated a significant effect. Algal turfs and barnacles were affected by treatments in spring, but not winter, but meaningful biological interpretation of these results is difficult. Macrograzers seem generally to have had a more profound effect than mesograzers. More important is the conclusion that, while grazers very clearly had significant effects in some plots, different factors had overriding influences in others. Given the size of plots used, these factors must operate on scales of less than 1 m and appear to result iii patchiness on very small scales on these shores. Small scale variations in recruitment success, topography and the outcomes of species interactions are all likely contributors to heterogeneity. ASPECTS OF SHELL SHAPE IN LAGOONAL LITTORIN- IDS. Peter J. Mill. Andrea P. Clarke, and John Grahame, School of Biology, The University of Leeds, Leeds, LS2 9JT. England, and Delmont C. Smith, Department of Biological Sci- ences, State University of New York, College at Brockport, Brock- port. New York 14420 U.S.A. In the Littorina saxatilis complex, five morphs have been dis- tinguished, including two which at various times have been af- 698 Abstracts. July 24-31. 1999 Sixth International Littoiinid Symposium forded species status. One of these, L. tenebrosa, is sensu stricto. a lagoonal-dwelling morph which remains submerged and is found on, e.g., Chaetomorpha. It has a smooth, fragile high-spired shell which is black or dark brown and may be banded. There is another littorinid which occurs in coastal lagoons but which has a different shell shape, i.e., lagoonal L. saxatilis. In a survey of about 30 non-tidal, coastal lagoons over a stretch of some 150 km of coastline in eastern England, only one was found to contain L. tenebrosa with a further four containing lagoonal L. saxatilis. The one containing L. tenebrosa was largely surrounded by tall vegetation and was separated from the sea by a mature, tree- covered dune. Of the others, one was separated by a dune, the others by a shingle bank. In each case, inundation by the sea would be a very rare event. Shells from these populations were weighed and imaged, following which a number of shell variables were measured. The populations were compared with each other and with a sample of coastal L. saxatilis from the same region. Fur- thermore, comparisons were made with a population of L. teneb- rosa from Ireland. In this case, the lagoon probably had periodic influxes of sea water. L. saxatilis were obtained from its seaward end; L. tenebrosa was confined to the landward end, adjacent to a freshwater input. IS LIFE ON A TROPICAL ROCKY SHORE REALLY SO HARD? Dwayne Minton1 and Deborah Gochfeld2, Hofstra Uni- versity Marine Laboratory, Priory. Jamaica, W.I. Interactions of biotic and abiotic factors are considered the principle mechanisms controlling the dynamics of rocky shore communities. Unfortunately, little research has examined how these factors affect community structure of tropical rocky shore assemblages. We examined the effect of wave action and desic- cation on a rocky shore molluscan assemblage on the north shore of Jamaica. This assemblage exists entirely above mean high water (MHW) where physical factors were expected to be more impor- tant than biological factors. We compared the molluscan assem- blage along ten vertical transects exposed to different levels of wave action and desiccation potential. In all, nineteen species of mollusc were observed, with thirteen occurring on >5Q% of our transects. Five species occurred on >80% of the transects and were dominate at three different vertical heights: Cenchritus muricatus occupied the highest position on the shore (1.0-1.3 m above MHW); Nodolittorina dililata and N. ziczac occupied a middle region (approximately 0.4-0.8 m above MHW); and Acantho- pleura granulata and Notoacmaea antillarum were found lowest on the shore (approximately 0.05-0.15 m above MHW). We found no differences in species number, individual densities, or the ver- tical distribution of the species between transects with differing wave action or desiccation potential, suggesting that these physical factors are not operating at the spatial scale studied. However, gastropod molluscs preferentially occupied pit and crevice micro- habitats, which are believed to mediate physical stresses such as desiccation and high temperature. The distribution of these rocky shore molluscs may be the result of the availability of, and com- petition for. rare microhabitats. 'Present address: Department of Zoology. University of Ha- waii. Honolulu. HI 96822, U.S.A. ^Present address: Department of Pharmacognosy, University of Mississippi, University, MS 38677, U.S.A. THE MANGROVE LITTORINIDS: EVOLUTION AND AD- APTATION IN THE GENUS LITTORARIA. David G. Reid, Department of Zoology. The Natural History Museum. London SW7 5BD. U.K. The littorinids found largely or exclusively upon mangrove trees are members of the genus Littoraria, a well-defined mono- phyletic group of 36 species occurring throughout the tropics. Some, but not all. of these species are notable for unusual mor- phological characteristics: they may be ovoviviparous (releasing veliger larvae or crawling juveniles), their shells may show strik- ing color polymorphism (occurring in discrete yellow, brown, and pink morphs). and the central tooth of the radula exhibits a sup- posedly unique extra cutting edge (the so-called rachidian hood). In each case, workers have suggested that these attributes might be adaptations to the mangrove habitat, although this remains specu- lative in the absence of information about character evolution. In order to test these hypotheses in a phylogenetic context, a new phylogeny for the 36 species of Littoraria has been produced by cladistic analysis of morphological characters. By mapping habitat type and these three key morphological characters onto the phy- logenetic tree, it is possible to test whether the characters are indeed derived features that have been selected within the man- grove habitat (i.e.. adaptations), or whether they were ancestral fea- tures that permitted colonization of this habitat (i.e.. exaptations). This paper has been published in Phuket Marine Biological Centre Special Publication No. 19(1), ( 1999). ORIGIN OF NERITA ATRAMENTOSA REEVE 1855, A NERITE OF TEMPERATE ROCKY SHORES (NERITOP- SINA: NERITIDAE). Bronwen Scott, Conservation Biology Unit. School of Life Sciences and Technology. Victoria University of Technology S008, P.O. Box 14228, Melbourne City MC. Vic- toria 8001. Australia, and Kerry Neil, Department of Marine Bi- ology, James Cook University, Townsville. Queensland 4811, Australia. The gastropod Nerita is characteristic of tropical and subtropi- cal rocky shores and mangroves worldwide. At least twelve spe- cies occur in Australia, all but one confined to the warmer waters of the Indo-West Pacific. This exception is Nerita atramentosa Reeve 1855. which lives on rocky shores in temperate and warmer waters from southern Queensland to North West Cape, western Australia. It is also known from Norfolk Island, New Zealand and Sixth International Littorinid Symposium Abstracts. July 24-31. 1999 699 the Kermadec Islands. The shell of N. atramentosa resembles both that of the African N. senegalensis and of the central pacific N. picea, but the distribution of N. atramentosa in Australia is con- tiguous with that of the twelve Indo-West Pacific species. As the only temperate species. N. atramentosa poses a biogeographical problem: is its sister species from the tropical Indo-Pacific, the central Pacific or the western Indian Ocean? This study uses con- chological and anatomical characters to investigate the phylogeny of the Australian nerites as a basis for biogeographical studies. RELATIONSHIPS AND POPULATION STRUCTURE OF LITTORINA ARCANA HANNAFORD ELLIS, L. COM- PRESSA JEFFREYS, AND L. SAXATILIS (OLIVI) IN THE BRITISH ISLES USING SSCPS OF CYTOCHROME-B FRAGMENTS. Maureen P. Small and Elizabeth M. Gosling, School of Science. Galway-Mayo Institute of Technology, and Microbiology Department. National University of Ireland, Gal- way. Ireland. The saxatilis snail species complex are ubiquitous and impor- tant members of hard shore intertidal communities in the North Atlantic. The complex includes the recognized species Littorina arcana Hannaford Ellis. L. compressa Jeffreys and L. saxatilis (Olivi). We investigated the species and population structure of these snails from six locations in Ireland and Britain using a non- radioactive single strand conformational polymorphism (SSCP) analysis of a 375 base pair fragment of the cytochrome-b gene. Variability was high in this marker with 38 haplotypes found in 586 individuals. The most common haplotype in L. arcana and L. compressa was absent from L. saxatilis and the most common haplotype in L. saxatilis was found in low numbers in both L. arcana and L compressa. Haplotypes restricted to L. arcana and L. compressa formed a cluster separate from haplotypes restricted to L. saxatilis in a maximum-likelihood tree, minimum spanning tree and multidimensional scaling analysis. In the same type of analyses examining population relationships. L. arcana and L. compressa formed a group separate from L. saxatilis. We con- cluded that L. arcana and L. compressa are more closely related to each other than either is to L. saxatilis and suggest that this is a resolution to a previous trichotomy among these species. immersed in seawater (W) or dry (D). From an initial concentra- tion of 100.8 p.g g_l. animals maintained CAV showed a decline to 67.5 |xg g"1. Animals that were C/D increased to 157.1 p.g g"1. those that were HAV to 175.3 u.g g~\ and those H/D to 219.8 |xg g"1. If animals that had their uric acid raised by desiccation were then returned to seawater their uric acid concentration returned to initial levels within 48 h. During this time when uric acid concen- tration was declining, the animals also produced ammonia at rates nearly double those of snails kept in seawater. It therefore appears that nitrogenous wastes are stored as uric acid at times when water must be conserved, and that the uric acid is then eliminated when water again becomes available; most probably not as uric acid, but after conversion of uric acid to ammonia. TRANS-ZONAL MOVEMENTS IN WINKLES, LITTORINA LITTOREA (L.): REASONS AND CONSEQUENCES. G. F. Warner, School of Animal & Microbial Sciences, The University of Reading. Whiteknights. Reading RG6 6AJ. U.K. Two years of monthly samples of L. littorea at three levels on an estuarine shore at Southampton. U.K., have provided evidence of long-term movements up or down the shore. The zone occupied by L. littorea at Southampton is from about mid-tide level down into the sublittoral. Increases in population densities at the middle site of winkles in their second year of growth, indicate movements of young winkles towards the centre of their zonation range from both higher and lower levels. Increases in population densities of older winkles at the higher and lower levels indicate later disper- sion away from the centre of the zonation range. Direct evidence of movements between the middle level and the lower level is provided by the presence or absence in the shell of bore-holes made by the polychaete Polydora ciliata. which recruits on winkle shells mainly on the lower shore. Thus, young winkles at the middle site with bore-holes on the spire have moved up from a lower level while older winkles at the lower level which lack bore-holes have moved down from a higher level. At the upper level, fluctuating population densities and observations of feeding fronts in the spring indicate seasonal migrations, down in winter and up in spring. Advantages and disadvantages of living at upper or lower levels are discussed and related to differences in growth rates and mortality at the different levels. EFFECT OF TEMPERATURE AND DESICCATION ON URIC ACID CONTENT OF LITTORINA SAXATILIS. Del- inont C. Smith, Department of Biological Sciences, State Univer- sity of New York. College at Brockport, Brockport, New York 14420, U.S.A. I have previously reported uric acid content of Littorina saxa- tilis that changes according to season and location on the shore. Such variation could be due to temperature and/or desiccation effects. To test this, winkles were maintained in controlled tem- perature rooms at 10 °C (designated as C) or 25 °C (H). and either CORRELATION OF MORPHOLOGICAL DIVERSITY WITH GENETIC DIVERSITY IN THE ROUGH PERI- WINKLE LITTORINA SAXATILIS. Craig S. Wilding, John Grahame and Peter J. Mill, The School of Biology. The Univer- sity of Leeds. Leeds LS2 9JT, U.K. Both morphological and genetical studies of rough periwinkles have been used in a variety of studies to examine the structuring of populations. However, the complementarity of these two different approaches has not been directly estimated, despite the fact that 700 Abstracts, July 24-31. 1999 Sixth International Littorinid Symposium such an approach could lead to a better understanding of the basis of shape variability in these morphologically diverse animals. In the present study, variation in both the mitochondrial DNA and in the nuclear genome (measured via RAPDs and PCR-RFLP of anonymous nuclear loci) of Littorina saxatilis is compared to, and correlated with, measures of phenotypic variability using multi- variate approaches. Techniques involving Mantel matrix compari- sons, and AMOVA (analysis of molecular variance) are consid- ered. We show that in certain instances, a high degree of correla- tion between morphology and genetics can be uncovered, and this suggests that morphological approaches are of use for detecting popu- lation structure as well as environmental components of shell shape. VARIATION IN AMMONIA EFFLUX RATES WITH EMERSION OF THREE HONG KONG NODILITTORINA SPECIES. R. F. Uglow, Department of Biological Sciences. Uni- versity of Hull. Hull HU6 7RX. U.K.. and Gray A. Williams, Department of Ecology & Biodiversity and The Swire Institute of Marine Science, The University of Hong Kong, Hong Kong. On moderately-exposed Hong Kong shores, Nodilittorina tro- choides, N. radiata, and N. vidua are distributed from the splash zone to the mid-eulittoral respectively where they experience long emersion times and, in summer, high rock and air temperatures. Individuals of these species were collected from the shore after varying periods of emersion for (0 rain, awash and active controls, and 1. 4. and 22 h. inactive animals on natural rock). Ammonia efflux rates (n = 6) were measured at fixed intervals (30 mins, 1 and 2 h) on reimmersion. The awash groups, after 30 mins. had mean efflux rates of 3.91. 6.01, and 3.53 p,mol NH4 • g-' ■ IT1 which were high compared with the 1.66. 2.02 and 0.32 u.mol NH4 ■ g_1 ■ h"' rates after 2 h for N. trochoides, N. radiata, and N. vidua, respectively. These data reveal interspecific variations in efflux rates and that handling may evoke enhanced ammonia ex- cretion rates as a stress response. Post-emersion efflux rates were always lower than those of the comparable control group — apart from the 30 min, post-1 h emersion rates of N. vidua, which were unchanged from control rates. N. vidua also showed the least diminution of efflux rates following 4 and 22 h of emersion — an interspecific difference which correlates well with the generally greater activity and relatively lower tidal position of this species and the infrequency with which it may experience naturally 22 h of emersion, cf the other 2 species. The energetic implications of these differences in post- emersion efflux rates are discussed in terms of emersion tolerance, hence limitations to vertical distributions on the shore. No evi- dence was found of a switch from ammonotely to uricotely during the lengths of emersion tested here, as no animal excreted detect- able levels of urate during these experiments. NUCLEAR RIBOSOMAL DNA SEQUENCES AND THE PHYLOGENY OF THE LITTORINIDAE. Birgitta Win- nepenninckx, Thierry Backeljau, Royal Belgian Institute of Natural Sciences. Vautierstraat 29. B-1000 Brussels. Belgium, and David Reid, Department of Zoology, The Natural History Mu- seum. Cromwell Road, London SW7 5BD, U.K. Phylogenetic relationships within the family Littorinidae were exhaustively investigated in several papers by D. Reid. who in particular dealt with the generic relationships in his 1989 paper {Phil. Trans. R. Soc. Loud. B 324: 1-1 10) and with the intrageneric arrangement of the genus Littorina in his book of 1996 (System- atics and evolution of Littorina, Ray Society). Most of these analy- ses were based on morphological characters, yet the phylogeny of Littorina was recently also investigated by means of mtDNA se- quences. Despite all these efforts there still remain a number of unresolved or somewhat uncertain issues with respect to both the relationships within Littorina and the relationships among lit- torinid genera. We therefore attempted to investigate littorinid phylogeny at different taxonomic levels using complete 18S rDNA (± 1800 bp) and partial 28S rDNA (± 770 bp) sequences. We will present a first, restricted analysis of 18S rDNA data aiming at testing the suitability of this molecule to resolve relationships within the genus Littorina (with focus on the position of L. striata). while in a second analysis, we will present a very preliminary attempt to resolve generic relationships within the Littorinidae using both 18S and 28S rDNA sequences of representatives of 13 littorinid eenera. Journal of Shellfish Research. Vol. 18. No. 2. 701-731, 1999. ABSTRACTS OF TECHNICAL PAPERS Presented at INTERNATIONAL CONFERENCE ON SHELLFISH RESTORATION Cork. Ireland September 29-October 2. 1999 Conference Organizer: Dr. Gavin Burnell Department of Zoology University College Cork Lee Makings. Prospect Row Cork. Ireland 701 ICSR. Cork, Ireland Abstracts, Sept. 29-Oct. 2. 1999 703 CONTENTS E. Aloj Totaro, S. Costabile, S. Porco, and M. Totaro Monitoring eoastal shellfish as a sustainable resource in southern Italy 705 W. S. Arnold, D. C. Marelli, K. Hagner, M. Parker, P. Hoffman, and M. Harrison Assessing the effectiveness of bay scallop (Argopecten irradians) population restoration on the west coast of Florida, USA 705 W. S. Arnold, D. C. Marelli, P. Hoffman, and M. Humphrey Testing alternative strategies for population enhancement of hard clams Mercenaria spp. in the Indian River Lagoon, Florida 705 C. Backer The use of mathematical models to assess the carrying capacity of exploited ecosystems 706 B. Ball, B. Munday, and G. Fox The impact of a Nephrops otter trawl fishery on the benthos of the Irish Sea 706 M. D. Barry A mechanism to reduce plastic waste and increase mussel production 706 C. Bradshaw, L. O. Veale, A. S. Hill, and A. R. Brand The effect of scallop dredging on Irish Sea benthos: experiments using a closed area 707 R. Browne and J. P. Mercer Homarus gammarus "colourmorphs" their incidence in the indigenous population around the Irish coast and their potential use as biological markers 707 R. Browne Historical overview of the Irish lobster (Homarus gammarus) and inshore decapod crustacean fishery with reference to European (H. gammarus) and North American (H. americanus) landings 707 R. Browne and M. Norman A description of the fishery for Palaemon serratus (pennant) in the Connemara (west of Ireland) area and potential management methods 708 R. Browne and J. P. Mercer A review of innovations concerning Ireland" s lobster (Homarus gammarus) fishery ( 1992 to 1998) 708 C. A. Burton, J. T. MacMillan, and M. M. Learmouth Shellfish ranching in the UK 708 C. A. Burton The role of lobster hatcheries in ranching, restoration and remediation programmes 709 P. A. Byrne and J. O'Halloran The role of the Manila clam. Tapes semidecussatus as a tool in estuarine sediment toxicity assessment 709 R. B. Carnegie, B. J. Barber, D. L. Distel, and S. C. Culloty Development of PCR and in situ hybridization assays for detection of Bonamia ostreae in flat oysters, ostrea edulis. . . 709 L. D. Coen, R. E. Giotta, M. W. Luckenbach, and D. L. Breitburg Oyster reef function, enhancement, and restoration: habitat development and utilization by commercially- and ecologically-important species 710 P. A. Cook and N. A. Sweijd Some genetic considerations of shellfish ranching: A case-study of the abalone, Haliotis midae in South Africa 710 A. C. Crook and D. K. A. Barnes Seasonality of echinoid behaviour in a marine 'island.' Lough Hyne. Ireland 710 S. C. Culloty and M. F. Mulcahy Living with bonamiasis: Irish research since 1 987 711 V. Cummins, N. Connolly, and G. Burnell An assessment of the potential for the sustainable development of the edible periwinkle. Littorina littorea, industry in Ireland 711 N. Hankers Recovery of intertidal mussel beds in the Waddensea after large scale destruction 711 S. De Waal, N. Sweijd, B. Godfrey, P. Britz, and P. Cook Abalone ranching in South Africa: Hope for sustainable abalone production? 712 P. Dolmer, T. Kristensen, M. L. Christiansen, M. F. Petersen, P. S. Kristensen, and E. Hoffmann Short-term impact of blue mussel dredging (Mytilus edulis L.) on a benthic community 712 704 Abstracts. Sept. 29-Oct. 2. 1999 ICSR. Cork. Ireland D. L. Eslinger, M. E. Culver, P. Tester, M. Soracco, and K. Waters Integrating field and remote sensing data — an example from a harmful algal bloom event 712 D. Fagergren Tools for turning the tide of deteriorating water quality in shellfish growing areas: a decade of experience in Puget Sound, Washington State, USA 713 E. Fahy A new fishery for razor clams (Ensis siliqua ) on the East Coast of Ireland 713 E. Fahy Attempts to alleviate fishing pressure on stocks of brown crab (Cancer pagurus) caused by the Whelk fishery in the South Western Irish Sea 713 A. F. Flanagan, M. Kane, J. Donlan, and R. Palmer Azaspiracid; detection of a newly discovered Phycotoxin in vitro (poster) 714 P. G. Fleury, E. Goyard, J. Mazurie, S. Claude, J. F. Bottget, A. Langlade, and M. J. Le Coguic A monitoring tool for assessing oyster performances in different farming areas: The Ifremer Remora Network: Method and first results (1993-1998) in Brittany (France) 714 G. E. Flimlin, Jr. The hard clam task force (poster) 714 M. Gaspar, L. Chicharo, M. D. Dias, P. Fonseca, A. Campos, M. N. Santos, and A. Chicharo The influence of dredge design on the catch of Callista chione 715 R. Gidney and J. Hermse The affects that amnesic shellfish poisoning has on scallop processors and commercial fishermen 715 G. Hilgerloh, J. O'Halloran, T. Kelly, and G. Burnell The influence of oyster culture structures on birds in a sheltered Irish estuary 716 D. Hugh-Jones Breeding ponds as a basis for flat oyster (Ostrea edulis) culture and their use to develop resistance to the disease Bonamia ostreae 716 A. Jensen Artificial reefs for shellfish habitat: Results and ideas to date 716 M. S. Kelly, P. Pantazis, and P. Owen The commercial potential of the common sea urchin Echinus esculentus 716 H. S. Lenihan and G. W. Thayer Ecological effects of fishery disturbance to oyster reef habitat in eastern North America 717 D. L. Leonard The integration of remote sensing data with local and state monitoring data 717 A. Linnane, B. Ball, J. P. Mercer, G. van der Meeren, C. Bannister, D. Mazzoni, B. Munday, and H. Ringvold Understanding the factors that influence European lobster recruitment: A trans-european study of cobble fauna 717 U. Lobsiger and J. L. Manuel Applications of underwater imaging techniques in the monitoring and restoration of coastal habitats 718 M. W. Luckenbach, J. Harding, R. Mann, J. Nestlerode, F. A. O Beirn, and J. A. Wesson Oyster reef restoration in Virginia, USA: Rehabilitating habitats and restoring ecological functions 718 T. Malone HABSOS: A pilot project of the US global ocean observing system and The National Association of Marine Laboratories 719 P. McGinnity and K. F. Whelan The management of freshwater catchments 719 K. A. McGraw and M. Castagna A comparison of the Arkshell clams, Anadara ovalis and Noetia ponderosa, as potential mariculture species along the Atlantic Coast of the United States 719 E. Mc Knight and H. Nelson The Canadian Shellfish Sanitation Program — How shellfish closures are leading to improved water quality 719 T. McMahon, J. Silke, and B. Cahill Irish coastal dinoflagellate blooms and shellfish toxicity 720 ICSR. Cork, Ireland Abstracts, Sept. 29-Oct. 2. 1999 705 D. Minchin Exotic species: Implications for coastal shellfish resources 720 S. Munch-Petersen and P. Sand Kristensen On the dynamics of inter-tidal and sub-tidal stocks of blue mussels in the Danish Wadden Sea 721 B. Myrand, R. Tremblay, and J.-M. Sevigny Decrease in heterozygosity in suspension-cultured blue mussels following their transfer to grow-out sites and its potential consequence on mussel farm productivity 721 F. X. O'Beirn. J. A. Nestlerode, and M. W. Luckenbach Evaluating design criteria and recruitment levels in the restoration of oyster reef assemblages 721 T. O 'Carroll Co-ordinated Local Aquaculture Management Systems (C.L.A.M.S.) 722 E. O'Mongain and A. Collins Absolute determination of chlorophyll concentration and optical properties of water by airborne hyperspectral remote sensing 722 M. J. Orren Chemical effects of hypoxic and anoxic waters on shellfish 722 G. O' Sullivan and M. F Mulcahy Enigmas in the reproductive biology of Pacific oysters in Ireland 722 M. C. Paraso and M. D. Ford A national harmful algal bloom data management system 723 A'. T. Paynter and T. E. Koles Use of videography to assess differences between restored and non-restored areas in Chesapeake Bay 723 M. Pena, C. Gomes, and W. Hunte The application of Randomly Amplified Polymorphic DNA (RAPD) markers to stock discrimination of the White Sea urchin. Tripneustes Ventricosus. in the eastern Caribbean 723 R. Raine, S. O'Boyle, T. O'Higgins, M. White, J. Patching, T. McMahon, and B. Cahill A satellite and field portrait of a Gyrodinium aureolum bloom off south-western Ireland; August 1998 724 E. Rambaldi, G. Priore, G. Prioli, N. Mietti, T. Pagliani, and M. L. Bianchini Trials on clam (Chamelea Gallina) beds of an innovative hydraulic dredge with vibrating and sorting bottom 724 A'. /. Reitan, G. 0ie, Y. Olsen, and H. Reinerstsen Effect of increased primary production in a fjord on growth of blue mussels and scallops 724 L Righetti Nitrogen excretion by the Pacific Oyster. Crassostrea gigas, as a contributor to estuarine nutrient cycling in Tomales Bay, California 724 M. Robinson and O. Tally Mortality and dispersal in a benthic sub-tidal decapod community and of hatchery reared lobster Homarus Gammarus 725 S. M. C. Robinson, S. Bernier, and A. Maclntyre The impact of scallop drags on sea urchin populations and benthos in the Bay of Fundy. New Brunswick. Canada 725 A. G. Sellner The U.S. HAB ProgTam: One-half of a U.S. -Ireland HAB exchange 725 S. E. Shumway Harmful algal blooms and shellfish restoration: Permanent obstructions or temporary nuisances? 726 J. Silke and T. McMahon Dinoflagellate resting cysts in Cork Harbour: Implications for shellfish aquaculture 726 A. Smaal and M. van Stralen Shellfish carrying capacity and ecosystem processes 726 G. W. Thayer Opening remarks for gear impact and remediation session 727 O. Tully Restoration of lobster [Homarus gammarus) population egg production in depleted stocks 727 J. C. Wallace Clam farming and tourism — A difficult combination? The socio-economic role of Ruditapes decussatus cultivation in the Algarve 728 706 Abstracts. Sept. 29-Oct. 2, 1999 ICSR. Cork. Ireland N. W. White, L. E. Danielson, and M. V. Holmes Development of hydrologie modification indicators to support watershed-based restoration of shellfish resources impacted by fecal coliform contamination 728 C. A. Wilson, H. H. Roberts, J. Supan, and W. Winans The acquisition and interpretation of digital acoustics for characterizing Louisiana's shallow water oyster habitat 728 D. Jackson and T. O 'Carroll Co-ordinating shellfish and finfish aquaculture systems 729 ICSR. Cork. Ireland Abstracts, Sept. 29-Oct. 2. 1999 707 MONITORING COASTAL SHELLFISH AS A SUSTAIN- ABLE RESOURCE IN SOUTHERN ITALY. E. Aloi Totaro, S. Costabile, S. Porco, and M. Totaro, Cattedra di Ecologia, Facolta di Economia, Universita degli Studi del Sannio (Ben- evento). Italy. We consider the environmental characters and distribution of Gastropod and Lamellibranch culture along the south Italian coast. The goal is to consider the possibility that this activity is a typical sustainable management. It is necessary to manage and to consider the sea as the most complex system of the Biosphere. It is composed of two fundamental regions, the former is the coastal zone form, the latter is the ocean form. The coastal zone includes some coastal districts of variable width. Their develop- ment is linked, strongly, to the management capacity of coastal communities, so this development is also linked strongly to the correct type of government. Italian coastal districts show increasing alterations of their bio- logical resources such as the growing pollution caused by urban and industrial wastes, hydrocarbon inductions and the changes to the water temperature. Recently an interest in new sustainable models for coastal re- gions and for new sustainable utilizations of marine biological resources, based on the increase of water bio-farms has developed. It is important to advertise these new methods of managing coastal ecosystems. These ecosystems will be considered as "controllable ecosystems". Consequently, it will be necessary to safeguard, to equip, to improve them in order to increase rational production of clams. But first one must locate specific coastal areas suitable for this production with an appropriate development strategy. So to realize this program it is necessary: - to revalue the human resource in this specific professional sector - to restore the marginal areas and the non-used resources - to create a real interest in the economic opportunities of clam restocking in the light of the sustainable development. ASSESSING THE EFFECTIVENESS OF BAY SCALLOP {ARGOPECTEN IRRADIANS) POPULATION RESTORA- TION ON THE WEST COAST OF FLORIDA, USA. W. S. Arnold, D. C. Marelli, K. Hagner, M. Parker, P. Hoffman, and M. Harrison, Florida Marine Research Institute, 100 Eighth Av- enue SE, St. Petersburg. FL. 33701-5095, USA. Bay scallops (Argopecten irradians) were once abundant in the nearshore marine environment of western Florida. In recent years, however, many of the local populations that comprise the pur- ported bay scallop metapopulation in Florida have collapsed, re- sulting in closure of the commercial fishery and implementation of severe restrictions on recreational harvest. Hypothesized causes of that collapse include overfishing, habitat degradation, and in- creased prevalence of red tide, but the exact cause is unknown. We are attempting to restore scallop populations in Florida by planting cultured scallops in selected areas, but we are taking a stepwise approach to scallop restoration by first assessing the feasibility of this strategy before initiating a full-blown restoration effort. This experimental approach consists of first developing a thorough baseline of information concerning the local abundance and rela- tive rates of recruitment in natural populations. We then spawn small groups of scallops for which we identify a unique mitochon- drial marker. The resultant (genetically identifiable) broods are planted in cages within the location from which the parent stock was collected. We monitor the growth, survival, and reproductive development of each brood. Additionally, we employ artificial collectors to obtain juvenile scallops that are returned to the labo- ratory for genetic analysis. This allows us to estimate the propor- tional representation of cultured scallops in the recruiting popula- tion. Finally, we sample adult populations both at the planting site and at other sites throughout western Florida, to estimate the pro- portional representation of cultured scallops in the following adult year-class. If cultured scallops comprise a significant proportion of the subsequent year-class, or if we observe increased scallop abun- dance relative to background levels, then we will consider implement- ing a large scale restoration effort. If experimental enhancement fails, then we must assume that the causative factors of the collapse remain and must be mitigated before initiating full-scale restoration. TESTING ALTERNATIVE STRATEGIES FOR POPULA- TION ENHANCEMENT OF HARD CLAMS MERCENAR1A SPP. IN THE INDIAN RIVER LAGOON, FLORIDA. W. S. Arnold, D. C. Marelli, P. Hoffman, and M. Humphrey, Florida Marine Research Institute. 100 Eighth Avenue SE. St. Petersburg, FL. 33701-5095, USA. Clams of the genus Mercenaria support a commercial fishery in the Indian River lagoon on the east coast of Florida, but that fishery is characterized by unpredictable availability of clams. We are testing three alternatives (spawner transplants, seeding, and direct larval injection) to enhance the availability of clams in the Indian River in an attempt to stabilize the fishery and reduce economic hardships visited upon the fishermen during periods of low clam abundance. Both spawner transplants and seeding are designed to increase the local abundance of mature clams, thereby enhancing fertilization success and resultant recruitment into the local habitat. Larval injection is designed to circumvent the natural spawning process by introducing fertilized eggs directly into the lagoon. For the spawner transplant study, we are harvesting clams from open shellfishing waters and transplanting them into two closed shellfishing areas during each of four seasons. Each sea- sonal group is evaluated every three months for survival and re- productive development. Seeding provides an alternative approach to enhancing spawner concentrations. We have planted three size classes of seed clams (2 mm SL, 8 mm SL, 16 mm SL| under each of four protective treatments (no cover, mesh cover, oyster shell cover, mesh+oyster shell cover), and we periodically assess the 708 Abstracts, Sept. 29-Oct. 2, 1999 ICSR, Cork. Ireland survival, growth, and reproductive development of the planted clams. Ultimately, we hope to compare the relative biological and economic advantages of transplants versus seeding for enhancing hard clam spawner stocks in the lagoon. To evaluate the efficacy of direct larval injection, we have tracked introduced larval masses to monitor advection, survival, and settlement. The results of the larval tracking study will be used to evaluate the feasibility of this strategy for enhancement of the harvestable stock, and to identify modifications necessary to increase the proportion of introduced larvae that successfully set to the benthic environment. THE USE OF MATHEMATICAL MODELS TO ASSESS THE CARRYING CAPACITY OF EXPLOITED ECOSYS- TEMS. C. Bacher, CREMA. B.P 5, 17137 L'Houmeau. France. In carrying capacity assessments, the key notion is that there is an optimum standing stock, yielding a maximum production under some constraints due to the population dynamics, the culture prac- tice and the food limitation. The main objectives of this paper are therefore to present the background, the interest and the limitation of the mathematical models used in carrying capacity assessments. A simple model is presented first to illustrate how the dynamics of cultured popula- tion yields the stock/production relationship. This example is analysed more deeply to define the basic components underlying the carrying capacity modelling which are in turn illustrated with case studies. The first case study deals with the Thau lagoon (France) for which a mathematical model of the population dy- namics was built to predict the evolution of the standing stock and the annual production, taking into account several sources of vari- ability, the rearing strategy of the farmers, the fluctuation of the environmental conditions, and the growth variability between in- dividuals. In a second case study (Aiguillon Bay, France), it is shown how the growth of cultivated filter feeders is affected by the trophic conditions. Modelling the ecophysiology, the cultured population dynamics and the ecosystem dynamics are then com- bined to assess the carrying capacity of the main French produc- tion area (Marennes-Oleron Bay). As a conclusion, it is stressed that i) the carrying capacity assessment is of major importance in the scope of the sustainable development of aquaculture — with an example given of an EU/ China project, ii) a full understanding of the underlying biological and physical processes is needed, iii) the socio-economic compo- nent has not yet been taken into account. THE IMPACT OF A NEPHROPS OTTER TRAWL FISH- ERY ON THE BENTHOS OF THE IRISH SEA. B. Ball. B. Munday, and G. Fox, The Martin Ryan Marine Science Institute. National University of Ireland. Galway, Galway, Ireland. The fishery for the Norway lobster {Nephrops norvegicus) is concentrated on a mud patch in the north western part of the Irish Sea. The fishery commenced in the early '40's as a small-scale summer fishery, but the season has now extended to include most times of the year. Fishing intensity is estimated to be high (each square metre is trawled c. 5-10 times per year). No quanti- tative historical data on the benthos is available for the period prior to commencement of the fishery, although some limited qualitative data exists. This paper reports on studies of the benthos undertaken in the period 1994-1996. Short term effects of fishing on the benthos were investigated by means of samples taken both before and shortly after (c. 24 h) fishing activity. Studies of the medium to long term effects involved sampling the fauna of areas around wrecks (i.e. unfished pseudo-control sites) for com- parison with fished grounds. From the available data, attempts were made to calculate the short, medium and long-term impact of the fishery on the benthos and surrounding environment. Direct (short-term) effects were not quantifiable at a heavily fished off- shore site (75 m water depth), however, some changes were visible in a shallow (35 m water depth), low fishing intensity site. The medium to long-term effects are more easily detectable at the offshore site, while only minor changes are visible at the in- shore location. It would appear, therefore, that it is fishing intensity per se. rather than simply the direct impact from passage of the gear, that constitutes the major factor controlling long- term negative trends in the benthos of the Irish Sea Nephrops grounds. A MECHANISM TO REDUCE PLASTIC WASTE AND IN- CREASE MUSSEL PRODUCTION. M. D. Barry, All in a shell Ltd., Dooniskey, Lissarda, Co. Cork, Ireland. Ireland's suspended culture of mussels is almost exclusively based on the use of polyethylene tubular meshes. This is cheap material that is used once and discarded at harvesting. It has been a very successful system but has limitations. The yield of mussels at harvest is in the range of 3-5 kg/m. Disposal of used polyethylene meshes is a growing environmental problem. Ireland plans to double its suspended mussel production over the next 5 years. It will not be possible to double the existing area of mussel farms because of competition for coastal resources. Consequently. it is necessary to increase the yield/hectare on sites where this is possible. ALL IN A SHELL is testing the new "Mussel Ladder" patent of XPLORA PRODUCTS, of Glascow, Scotland using biodegrad- able cotton mesh and binding materials. The object is the development of a growing and handling sys- tem that will yield more mussels/m, be more suitable to new more exposed sites and solve the growing environmental problems of disposal of used polyethylene meshes. Early results in Ireland. Scotland and Norway are very encouraging. ICSR, Cork. Ireland Abstracts, Sept. 29-Oct. 2. 1999 709 THE EFFECT OF SCALLOP DREDGING ON IRISH SEA BENTHOS: EXPERIMENTS USING A CLOSED AREA. C. Bradshaw, L. O. Veale, A. S. Hill, and A. R. Brand, Port Erin Marine Laboratory. Port Erin, Isle of Man. IM9 6JA. British Isles. A 2 km2 area off the southwest coast of the Isle of Man ( Irish Sea) has been closed to commercial fishing with mobile gear since March 1989. This area was heavily fished for Pecten maximus prior to closure, and the seabed immediately surrounding the closed area is still one of the most heavily dredged in the Irish Sea. Two methods have been used to study the effect of scallop dredg- ing on the benthos in this closed area and adjacent fished areas. Firstly, divers have carried out visual transect surveys of the epib- enthos regularly since closure. Secondly, biannual fine-meshed dredge and grab sampling of experimental plots inside and outside the closed area since 1995 has enabled comparisons of the benthic infauna and epifauna of experimentally dredged plots, undredged control plots and plots exposed to commercial dredging. Since 1989. there have been consistent significant increases in the mean numbers of many species in the closed area, including Pecten maximus and Luidia ciliaris, and upward trends in numbers of hermit crabs, spider crabs and brittlestars have also been re- corded. Conversely, the common starfish. Asterias rubens, appears to be decreasing in abundance. Communities of experimentally disturbed plots have become less similar to adjacent undisturbed control areas and more similar to commercially dredged areas. At each sampling date, similarity between samples was greater out- side the closed area than inside. These results present strong evidence that scallop dredging al- ters benthic communities and suggest that the closure of areas to commercial dredging may allow the development of more hetero- geneous communities and allow the populations of some species to increase. A common problem with studying fishing disturbance is the lack of good control sites and this work also demonstrates the value of closed areas to scientific studies of bottom fishing. HOMARUS GAMMARUS "COLOURMORPHS" THEIR IN- CIDENCE IN THE INDIGENOUS POPULATION AROUND THE IRISH COAST AND THEIR POTENTIAL USE AS BIOLOGICAL MARKERS. R. Browne12 and J. P. Mercer,1 'Shellfish Research Laboratory, Cama. Co. Gal way; 2Taighde Mara Teo., Carna, Co. Galway. Ireland. The marking of Homarid lobsters is an important means of obtaining data on their ecology, biology, population dynamics, survival and fisheries management. To facilitate the recovery of marked individuals in the absence of a concerted sampling pro- gramme, they need to be highly visible for up to 10 years. Occasionally odd coloured (red. blue, white or yellow) H. gam- marus and H. americanus have been landed. These colour variants have a genetic basis. It is important however to distinguish be- tween genetically blue lobsters from those which can result from inadequate feeding and unnatural holding conditions in the labo- ratory or that result in the wild for various environmental factors. One of the defining features used to describe the appearance of colourmorphs in this study was the colour of the antennae. Blue colourmophs have blue antennae unlike the variants of normal coloured lobsters that have redish brown antennae. Biologists in the 1960's deemed that the genetic colour variants were potentially valuable as a biological marker for research. This poster presents the variants recorded in Irish waters and looks at their occurrence in wild stocks. Over a five year period (1993 to 1997) 8 '•electric" blue, 7 red, and 4 white H. gammarus (CL > 85 mm) have been reported or delivered to the Shellfish Research Laboratory (NUIG). This gives a very approximate estimate from the national catch records of natural occurrence of 1 : 566.000. 1 : 647.000 and 1/1,132.000 respectively. Before they can be used as markers fundamental issues need to be addressed. Their behaviour, survival, and potential impact on lobster stocks need to be evaluated. HISTORICAL OVERVIEW OF THE IRISH LOBSTER [HOMARUS GAMMARUS) AND INSHORE DECAPOD CRUS- TACEAN FISHERY WITH REFERENCE TO EUROPEAN {H. GAMMARUS) AND NORTH AMERICAN (H. AMERICANUS) LANDINGS. R. Browne, Taighde Mara Teo.. Shellfish Research Laboratory, Carna, Co. Galway. Ireland. The historical development of Ireland's (Homarus gammarus) fishery is reviewed and compared to the landings in Europe and the North American H. americanus fishery. Lobster fishing (H. gam- marus) has provided an essential form of income for over a century to many Irish coastal communities. There have been large fluctua- tions in landings over this time. In 1994 landings reached an all time high of 714 tonnes, valued at IR£6.3 million, in 1997 landings were 513 tonnes valued at IR£4.5 million. The Irish lobster fishery regulations are: a) Minimum carapace length of 85 mm b) Ban on the landing of lobsters that have been V-notched or which have a mutilated tail fan c) Prohibition of capture of lobsters by SCUBA diving, and d) Licensing of all fishing vessels. Included are chronological details on legislation and innovations relating to the inshore crustacean fishery sector. H. gammarus forms an integral part of a complex mixed inshore commercial fishery for decapod Crustacea. Landings of H. gammarus exhibit signs of an overexploited stock with declines in catch per unit effort. There is a general decline of total landings for H. gammarus throughout Europe. Historically the management practices used have not proved to be successful as is evident in the catches. In contrast with North America, where there is a plethora of regulations governing the H. americanus fishery, landings appear to be in a healthier state. To effectively manage the interrelated inshore crustacean fisheries fundamental biological issues need to be addressed. Enlightened measures based on factual information are required to sustain and develop the lobster [H. gammarus) fishery in Irish waters. 710 Abstracts. Sept. 29-Oct. 2. 1999 ICSR. Cork. Ireland A DESCRIPTION OF THE FISHERY FOR PALAEMON SERRATUS (PENNANT) IN THE CONNEMARA (WEST OF IRELAND) AREA AND POTENTIAL MANAGEMENT METHODS. R. Browne1 and M. Norman,2 'Taighde Mara Teo., Carna, Co. Galway. Ireland; "Shellfish Research Laboratory (NU1G), Carna. Co. Galway. Ireland. The caridean prawn Palaemon serratus are locally known as shrimp. For this reason they will be referred to as shrimp in this report. P. serratus is a valuable component of the Connemara inshore fishery. This survey was undertaken by Taighde Mara Teo. and the Shellfish Research Laboratory (NUIG) to investigate the effort expended into the fishery and yield of shrimp. The infor- mation obtained came from fishermen, shellfish merchants and experimental fishing. Aspects of P. serratus life cycle and biology are also discussed. The shrimp P. serratus fishery in Ireland was established in the mid 1970s and landings have steadily increased in volume up to the 1990s where they appear to have reached a plateau of over 300 tonnes. It has become an important ancillary activity for many inshore fishermen and a sole fishery occupation for some. In 1996. 398 tonnes of P. serratus were landed nationally valued at IR£2.03 million (Dept. of Marine). Commercial fishing predominantly oc- curs along the south and west coasts in suitable bays. The fishing season is weather and location dependent starting in July/August, ending in February/March. Approximately 22 tonnes of/5, serratus were landed by up to 41 boats between Slyne head and Barna (Connemara) in 1997. P. serratus were purchased by four mer- chants for IR£7 to £9 per kg. and exported to the continent. Man- agement strategies are discussed in the poster. These include the increase of pot mesh size and a closed season. A REVIEW OF INNOVATIONS CONCERNING IRELANDS LOBSTER {HOMARUS GAMMARUS) FISHERY (1992 TO 1998). R. Browne1-2 and J. P. Mercer,1 'Shellfish Research Laboratory, Carna, Co. Galway. Ireland; 2Taighde Mara Teo.. Carna. Co. Galway. Lobster (H. gammarus) are a valuable component of Ireland's mixed inshore fishery. Consistent fishing pressure due to buoyant market demand has established keen interest in their biology and ecology. It is generally perceived both by fishers and governmental bodies that the pressure on lobster stocks is increasing and stocks are likely to suffer. National landings have varied from 715 tonnes in 1994 to 513 tonnes in 1997. In the past seven years, a number of innovations have occurred in the Irish lobster fishery which are discussed in this paper. Due to the concerns expressed about declining catch per unit effort by lobster fishermen, two separate stock enhancement pro- jects with hatchery produced lobsters were initiated (SRL. Carna and Wexford Lobster Co-op. Came). The systems used to culture stage I larvae to postlarvae involved using two species of phy- toplankton and Artemia in vigorously aerated hoppers providing routine survival of over 40%. The techniques of postlarval nursery cultivation, transportation, release and potential marking methods are discussed. A summary of lobster production is provided. Management of Irish lobster stocks in 1992 focused on a mini- mum size 85 mm (carapase length) and a ban on the capture of lobsters by SCUBA diving. The voluntary conservation practise of protecting selected reproductively mature lobsters by "V" notching their uropods began in 1993. In 1994 the Irish Lobster Association (ILA) was formed to represent regional organisations. The nation- wide awareness campaign for management measures was strength- ened by support from North American industry and scientific rep- resentatives. Also in that year legislation and governmental finan- cial support served to give impetus to V notch programmes. SHELLFISH RANCHING IN THE UK. C. A. Burton, J. T. MacMillan, and M. M. Learmouth, Sea Fish Industry Authority. Seafish Aquaculture, Marine Fanning Unit, Ardtoe, Acharacle, Argyll PH36 4LD. UK. There are six species of shellfish of commercial importance within the UK, which can be thought of as 'ranched' to varying degrees. Seabed ranching on natural habitats is where aquaculture and fisheries interact and some of the distinctions blur. Seafish. as part of its role in promoting the sustainable management of re- sources, is examining the current state of the industry. Mussels (Mytilus edulis). Flat Oysters (Ostrea edulis) and, most recently. Cockles (Cerastoderma edule) are harvested as seed or part-grown animals and relayed on to controlled beds for on- growing to marketable sizes. This is the most extensive form of aquaculture. It is reliant upon natural spat fall and wild gathering for juveniles; the only intervention is to relocate the animals to more favourable growing areas. The move away from suspended culture in favour of on-bottom growing for King Scallops (Pecten maximus) has moved them from a semi-intensive to an extensively reared species. Hatcheries are in their infancy and most of the industry is reliant upon active collection of the wild spat as they settle from the plankton. These small juveniles are on-grown in rearing systems in the wild before being released on to suitable areas for growth to marketable sizes. Other than collection and protection of the juveniles, there is no other input. The species most often associated with the term 'ranching' is the European Lobster (Homarus gammarus). Egg-bearing females are gathered from the wild and transferred to controlled conditions in hatcheries where the larvae are hatched and the juveniles on-grown before they are released in to the wild. Until the recapture of the animals at market size as part of the wild fishery, there is no further human interference. Manila Clams (Tapes philippinarum) are a special case; as a non-native species, juveniles are produced in hatcheries and then seeded in to netted beds for on-growing under natural conditions. The netting protects the stock from predators and also provides containment. ICSR. Cork, Ireland Abstracts, Sept. 29-Oct. 2. 1999 711 THE ROLE OF LOBSTER HATCHERIES IN RANCHING, RESTORATION AND REMEDIATION PROGRAMMES. C. A. Burton, Sea Fish Industry Authority. Seafish Aquaculture. Marine Fanning Unit, Ardtoe. Acharacle. Argyll PH36 4LD. UK. Juvenile lobsters (Homarus spp) produced in a hatchery can be used for many purposes. Within the field of resource management the most likely uses are in ranching, restoration and remediation pro- grammes. The decrease in the cost of animals, resulting from the commercialisation of the hatcheries, makes this increasingly viable. Ranching: Successful trials in the UK proved that hatchery -reared juvenile lobsters could be released to the wild and that they would survive to market size. Their recapture in the commercial fishery and subsequent data demonstrated that these animals could have a positive effect on landings. The extension of UK legislation to confer protec- tion on seeded stocks opened the way for ranching initiatives to progress. Sustainable resource management is also promoted by these projects. Other programmes within Europe and elsewhere, were able to draw upon these results to support their own work. Restoration: Natural or anthropogenic environmental events can bring lobster stocks in an area to the brink of extinction. One location where this has occurred is the northern Adriatic Sea. Here, Italian researchers are using hatchery techniques to produce juve- niles from available native broodstock. Initially, the animals are being on-grown and then released in to protected areas. They will be monitored and, if successful, the programme may be extended. Ultimately, it may restore not only the lobster population but also the fishery to health. Genetically distinct or other unique popula- tions could be protected and restored in this way. This initiative has direct parallels with captive breeding and conservation efforts with terrestrial and avian species. Remediation: Where human activity, for example, major civil engineering projects, directly affects the marine environment, it runs the risk of damaging fisheries. Sessile or habitat dependant species such as molluscs and Crustacea are more frequently disad- vantaged than mobile fish. Creation of suitable 'new' or 'alterna- tive' habitat and stocking with hatchery derived juveniles may form part of the compensation package negotiated with fishery interests. Where appropriate, the design of the structure can be modified to produce the desired habitat. In other areas, where habitat has been destroyed (e.g.. by dumping or dredging), a pur- pose-designed artificial structure may be used. Economic consid- erations have been cited for the non-progression of schemes in the past, but as costs fall, the balance may alter. Lobsters are a prime species for use in this way. THE ROLE OF THE MANILA CLAM, TAPES SEMIDE- CUSSATUS AS A TOOL IN ESTUARINE SEDIMENT TOX- ICITY ASSESSMENT. P. A. Byrne and J. O'Halloran, Dept of Zoology & Animal Ecology. National University of Ireland. Cork. Increasing emphasis is being placed upon chemical analysis of sediments to determine the distribution and concentration of toxic chemicals in aquatic environments. The resulting data are often used to characterise chemical accumulations, including delineation of 'hotspots.' These data alone, however, provide no information regarding the possible biological associations. Therefore, direct ecotoxicity testing is required. Estuarine sediments frequently are repositories and therefore potential sources of anthropogenic contaminants. Many organic and metallic chemical compounds released into aquatic systems bind to particulates and so accumulate in the sediments, thus, sediments become repositories of contaminants in estuaries. These may also cause contamination through diffusion of porewater, re- suspension of particulates and dispersal of benthic fauna. There is a need to assess the biological affects of these anthropogenic con- taminants because they may be toxic to infauna and bottomfish. Sediment toxicity bioassays are a means for carrying out such an assessment and primarily provide data on toxicity by measuring the effects on the test organism. The Manila clam. Tapes setnidecussatus was used to assess the ecotoxicity of estuarine sediments field-collected from five estua- rine and coastal areas around the Irish Coast. The endpoints mea- sured during the study were survival, behaviour, condition indices, biochemical and histochemical. Of these endpoints, the most sen- sitive were survival, behaviour and histochemistry. The potential of this cultured bivalve species as a sediment toxicity bioassay organism is discussed. DEVELOPMENT OF PCR AND IN SITU HYBRIDIZATION ASSAYS FOR DETECTION OF BONAMIA OSTREAE IN FLAT OYSTERS, OSTREA EDULIS. R. B. Carnegie, B.J. Barber, D. L. Diste,' and S. C. Culloty,2 'School of Marine Sci- ences. 5735 Hitchner Hall, University of Maine, Orono. ME 04469. USA; 2Dept. of Zoology & Animal Ecology. University College. Lee Maltings, Prospect Row, Cork. Ireland. Rapid and sensitive methods for the detection of pathogens are needed for successful shellfish culture and restoration. Flat oysters (Ostrea editlis) infected with the microcell parasite Bonamia os- treae were used to develop a polymerase chain reaction (PCR) assay that will be more sensitive, faster, and less costly than stan- dard histology. Genomic DNA was extracted from hemolymph of a Maine oyster and the gill of an Irish oyster. Using the PCR and primers tuned to protistan rDNA. a single, identical amplicon of 528 bp was obtained from both samples. This product spanned 341 bp of 18S rDNA and 187 bp of ITS1, and was determined by BLAST search to closely resemble rDNA genes belonging to members of the Phylum Haplosporidia. A PCR reaction specific for this sequence was designed and used to assay gill tissue of 71 Irish oysters independently scored for B. ostreae infections using hemolymph smears. A product presumed to be B. ostreae sequence was generated in 100% of heavily infected oysters; 73% of mod- erately infected oysters; 50% of lightly infected oysters; 0% of 712 Abstracts, Sept. 29-Oct. 2, 1999 ICSR. Cork. Ireland scarcely infected oysters; and 3% of those apparently uninfected. Under these reaction conditions amplification was strongest when infection intensity was high, i.e., the relative amount of contami- nating oyster DNA was low. Fluorescent in situ hybridization was used to confirm that the PCR product belonged to B. ostreae. Improvement of the sensitivity of both PCR and in situ diagnostic methods is ongoing. OYSTER REEFFUNCTION, ENHANCEMENT, AND RES- TORATION: HABITAT DEVELOPMENT AND UTILIZA- TION BY COMMERCIALLY- AND ECOLOGICALLY- IMPORTANT SPECIES. L. D. Coen,' R. E. Giotta,1 M. W. Luckenbach,2 and D. L. Breitburg/ 'Marine Resources Re- search Institute. SCDNR.. Charleston. SC 29412. USA: Virginia Institute of Marine Science, Eastern Shore Lab. P.O. Box 350. Wachapreague. VA 23480, USA: 3The Academy of Natural Sci- ences, Estuarine Research Center, St. Leonard, MD 20685. USA. Marine and estuarine habitats (e.g.. seagrasses. mangroves, and salt marshes) have received a great deal of attention for their value as important habitats and hence of late have been protected. In contrast, oysters and the habitat they generate have been viewed historically to have value only as a resource. Consequently, virtually unregulated resource extraction has resulted in the significant destruction of both subtidal and intertidal oyster reefs. Traditional restoration efforts, therefore, have been directed at enhancing oyster landings depressed by either overharvesting or reduced environmental quality, via meth- ods ranging from improving water quality to substrate addition, but all with the goal of resource enhancement. Of late, new emphasis has been placed on their direct and indirect ecosystem services, including filtering capacity, benthic-pelagic coupling, nutrient dynamics, sedi- ment stabilization, provision of habitat, with a concomitant shift to restoring/enhancing ecological function. We discuss current restora- tion efforts in the U.S.. their methods and goals and briefly review the findings and recommendations. We also discuss some of the issues associated with realizing the broader goal of ecological restoration, particularly in light of evaluating of oyster reefs as critical or Essential Fish Habitats (=EFHs). including current work comparing oysters with marsh and mudflats tidal creek habitats in SC. Finally, we review the impact harvesting and die-offs have on oyster reefs using experi- ments examining recruitment, growth, and survival of oyster spat as indicators of restoration potential by following recovery of manipu- lated sites after simulated harvesting, repletion, and/or major popula- tions die-offs. SOME GENETIC CONSIDERATIONS OF SHELLFISH RANCHING: A CASE-STUDY OF THE ABALONE, HAL1- OTIS MIDAE IN SOUTH AFRICA. P. A. Cook and N. A. Svveijd, Department of Zoology. University of Cape Town, South Africa. Several recent studies, where molecular genetic techniques and markers have been applied to marine organisms, have revealed various levels of population structure that were not predicted from previous studies. Some results have suggested that recruitment patterns may be relatively localised, even in broadcast spawners with pelagic larval stages. This has important implications for stock recognition and the delineation of biogeographic boundaries and may become more important as efforts at shellfish restoration or enhancement increase. Abalone ranching and enhancement is being attempted in sev- eral countries, with varying degrees of success having been re- ported. These attempts have stimulated discussion on ecological and genetic implications and widely differing views have been put forward. A synthesis of this discussion will be presented and a specific case study will be reviewed. The South African abalone. Haliotis midae, is distributed along at least 2000 Km of coastline, spanning three recognised biogeographical zones, but is currently managed as a single stock. A mtDNA analysis of wild and hatchery populations revealed at least two (possibly three), major genetic sub-divisions of the popu- lation. This is in contrast to results of a previous study that used more traditional methods (allozymes), which did not reveal any sub-divisions. The two stocks of abalone identified (east and west of Cape Agulhus), differ in terms of both haplotype frequencies and genetic diversity. Hatchery cohorts from two farms (one using east coast broodstock and one using west coast broodstock) both revealed significantly reduced genetic diversity and skewed hap- lotype frequencies relative to their source populations. The implications of these results for mariculture and ranching of the South African abalone are discussed and their possible ap- plicability to wider shellfish restoration issues is reviewed. SEASONALITY OF ECHINOID BEHAVIOUR IN A MA- RINE ISLAND,' LOUGH HYNE, IRELAND. A. C. Crook and D. K. A. Barnes, Dept. Zoology & Animal Ecology. Univer- sity College. Lee Makings, Cork. Ireland. Many echinoid species have been observed to display migra- tory and so-called "covering' behaviours. The functional signifi- cance of these has yet to be quantified although some of the most popular theories have interpreted them as an adaptive response to avoid over exposure to light and/or as an anti-predator strategy. Predation pressure and light intensity may be seasonal as well as diurnal in nature but were not investigated in the context of echi- noid behaviour until the present study. The aim of our research was therefore to examine. /'/; situ, the potential seasonality of both diurnal migration and covering behaviour in a population of the purple sea urchin. Paracentrotus lividus at Lough Hyne. Co. Cork, Ireland. This is a species of commercial importance and one that has suffered severe declines in its natural population density through overfishing, particularly in the Mediterranean. At Lough Hyne. P. lividus has also undergone dramatic population fluctua- tions but these are not believed to be a response to overfishing. ICSR, Cork, Ireland Abstracts, Sept. 29-Oct. 2. 1999 713 Our results showed significant seasonal trends in the proportion of individuals displaying covering behaviour, the mean number and proportion of available items used to cover and the availability of covering items in P. lividus at Lough Hyne. In addition, a significant diurnal and seasonal pattern was found for P. lividus migratory behaviour. Predator density and light intensity were also shown to vary seasonally at Lough Hyne. A quantitative assess- ment of variables that may influence individual behaviours and thus subsequent growth patterns and survival may be of impor- tance to the aquaculture industry. LIVING WITH BONAMIASIS: IRISH RESEARCH SINCE 1987. S. C. Culloty and M. F. Mulcahy, Department of Zoology and Animal Ecology. National University of Ireland Cork. Lee Makings. Prospect Row. Cork, Ireland. The European flat oyster. Ostrea edulis, was the major species of oyster grown in Ireland until 1987. when, following significant mortalities, the parasite Bonamia ostreae was diagnosed in oysters in Cork Harbour. Since then the parasite has been detected in a number of other major oyster-growing areas also. The impact of the disease, together with the field- and lab-based investigations and trials since that time are reviewed, with particular reference to the biology and specificity of the parasite and the development of resistance in oysters bred from survivors. AN ASSESSMENT OF THE POTENTIAL FOR THE SUS- TAINABLE DEVELOPMENT OF THE EDIBLE PERI- WINKLE, LITTORINA LITTOREA, INDUSTRY IN IRE- LAND. V. Cummins,1 N. Connolly,' and G. Burnell,2 'Coastal Resources Centre, University College, Cork; 2Zoology Depart- ment, University College, Cork. The industry created from harvesting the common periwinkle Littorina littorea is valuable both in terms of employment and export earnings. Approximately £5 million worth of periwinkles is exported annually, primarily to France. Little information is available on the common periwinkle Lit- torina littorea in Ireland with no scientific information available on the quantities harvested per year, or the impact of this harvest on the populations of periwinkles around the coast. One of the biggest problems facing the industry is it's almost complete lack of regulation. In addition, many wholesalers have suggested a pos- sible over-harvesting of the resource. They are finding it increas- ingly difficult to obtain medium to large sized winkles (greater than 16 mm in height), which the continental market requires. There is clearly a need for a comprehensive study to review the current state of Irish periwinkle stocks. This project, currently in its second and final year, aims to assess the sustainable develop- ment of the edible periwinkle industry in Ireland. This is being achieved by conducting an industry review in conjunction with a national stock assessment programme. Surveys of harvested shores show correlations between factors such as substrate type and peri- winkle densities. For example, although occasionally found in sandy patches, periwinkles favour a stable substrate. Relationships also exist between shell height and distribution according to bio- logical zones on the shore. Other results show that winkle densities are effected by exposure, with the highest density of winkles oc- curring in moderate environments. The industry review indicates that ninety five percent of whole- salers would support a closed season to give the periwinkles a chance to grow, with the suggestion of a summer season receiving the most backing. The data accumulated throughout this project will be incorporated into a GIS. This will assist in creating an appropriate management strategy for the industry. RECOVERY OF INTERTIDAL MUSSEL BEDS IN THE WADDENSEA AFTER LARGE SCALE DESTRUCTION. N. Dankers, A.G. Brinkman, A. Meijboom, and J. Zegers, Institute for Forestry and Nature Research, P.O. Box 167. 1790 AD Den Burg, Texel, The Netherlands. Intertidal musselbeds almost disappeared from the Waddensea in the 1980s. The major cause was fisheries, combined with a period of low spatfall. In the three Waddensea countries different management options were adopted in order to reach the 'ECO- TARGET', 'a larger area and more natural distribution of inter- tidal mature musselbeds. ' In Denmark large areas were closed for fisheries, and musselbed area is increasing. In Schleswig Holstein, the whole intertidal was closed and development of beds looks promising. In Lower Saxony, beds are still under fishery pressure which gave the opportunity to study the fishery impact. In the Dutch sector a co-management option has been developed. Some areas are closed; in the rest of the inter- tidal, areas with high potential for development of stable beds are left undisturbed by fishermen. A number of developing beds has been studied over at least 6 years. Surface area of individual beds did not change much, horizontal and vertical structure increased, several age classes are present, stability of the beds increased and in all beds spatfall was a common phenomenon, also when spatfall on the surrounding flats was limited. Based on the locations of mature beds in the past, and physical characteristics of these areas (exposure time, wave action, sediment characteristics etc) a model was devel- oped in order to predict which areas have a high potential for the development of mature beds. The maps constructed on the basis of the model are used in sectoral management plans for closing areas for seed fisheries and restoration programs for mus- sel beds. 714 Abstracts, Sept. 29-Oct. 2. 1999 ICSR. Cork, Ireland ABALONE RANCHING IN SOUTH AFRICA: HOPE FOR SUSTAINABLE ABALONE PRODUCTION? S. De Waal, N. Svveijd,1 B. Godfrey,2 P. Britz,2 and P. Cook,1 'Department of Zoology, University of Cape Town. Rondebosch 7700, South Af- rica; 2Department of Ichthyology and Fisheries Science. Rhodes University. P.O. Box 94. Grahamstown 6140. South Africa. South Africa's commercial abalone fishery for the perlemoen (Haliotis midae) is in crisis due to over-exploitation caused by large scale poaching, increasing recreational fishing effort and ineffective law enforcement. This threatens the livelihood of people currently employed in the industry, as well as the effective implementation of South Africa's Marine Living Resources Act which aims to achieve equity in access rights to fishery resources. Land-based abalone mariculture has been successfully imple- mented at several locations along the South African coastline, and subsequently, abalone seeding with hatchery produced spat is be- ing considered both for stock enhancement and commercial aba- lone ranching. Abalone ranching offers an incentive-based mecha- nism for co-management of defined areas, and hope for sustainable abalone production. The assumption is that "ranchers' will have an incentive to protect the seeded areas against poaching. Pilot aba- lone ranching trials are being conducted in heavily poached areas on the South African East coast (Indian Ocean), as well as on the West coast (Atlantic Ocean) where abalone do not occur naturally. Effective protocols for seeding hatchery reared abalone have been developed, and results to date indicate that commercially viable growth and survival is achievable. Genetic studies (using mtDNA and microsatellite DNA markers) are being conducted on both wild and hatchery populations in order to assess the genetic im- plications of ranching. These studies aim to contribute to the de- velopment of appropriate hatchery genetic management pro- grammes and also to evaluate the potential of genetic tags as a tool to assess survival. SHORT-TERM IMPACT OF BLUE MUSSEL DREDGING (MYTILUS EDULIS L.) ON A BENTHIC COMMUNITY. P. Dolmer, T. Kristensen, M. L. Christiansen, M. F. Petersen, P. S. Kristensen, and E. Hoffmann, Danish Institute for Fisheries Research, Charlottenlund Castle, DK-2920 Charlottenlund, Den- mark. The short-term effect of mussel dredging in a brackish Danish sound was studied. A diver identified a commercial dredging track and an analysis of the species composition inside the track and at an adjacent control area showed that dredging changed the com- munity structure by reducing the density of small polychaetes. In order to investigate the extent and the duration of the dredging impact experimental dredging was conducted. The experimental dredging removed 50% of the mussels in the two dredged areas. Immediately after dredging, a significantly lower number of spe- cies was measured inside the mussel beds in dredged areas com- pared to control and boundary areas. This effect lasted for at least 40 days. The analysis of the species composition showed that the dredged area had a significantly lower density, particularly of small polychaetes compared to the boundary area. An increased number of species was recorded outside the mussel beds just after dredging, but this effect lasted for less than 7 days. After dredging, brown shrimps, C. crangon invaded the dredged areas. This spe- cies is an important predator of smaller invertebrates, and it is suspected that it was feeding on small vulnerable polychaetes ex- posed at the sediment surface after dredging. The dredging process was observed to form 2-5 cm deep furrows in the seabed, but the sediment texture and the organic content of the sediment were not affected. The biomass accumulation of individual blue mussels was significantly lower in the dredged area compared to the boundary area. This indicates that the disturbance of the mussel bed structure reduced growth and that the lowering of intraspecific food competition caused by a reduced density of mussels did not increase the accumulation of biomass in the mussels that remained in the dredged area. INTEGRATING FIELD AND REMOTE SENSING DATA- AN EXAMPLE FROM A HARMFUL ALGAL BLOOM EVENT. D. L. Eslinger,1 M. E. Culver,1 P. Tester,2 M. Sor- acco,'-3-4 and K. Waters,1 'NOAA NOS Coastal Service Center, 2234 South Hobson Ave.. Charleston. SC. 29405. USA; 2NOAA NOS Center for Coastal Fisheries Habitat Research. Beaufort Laboratory. 101 Pivers Island Rd„ Beaufort, NC. 28516, USA; 'Systems Engineering and Security, 7474 Greenway Center Drive, Greenbelt, MD, 20770 USA; 4NOAA NESDIS Office of Satellite Data Processing and Distribution, FB4, Suitland, MD, 20746. USA. Several National Oceanographic and Atmospheric Administra- tion (NOAA) offices are developing applications which integrate a variety of field observations with remotely sensed data, with a goal of increasing our understanding of biological and physical pro- cesses in the upper ocean. Although coastal managers have a need for timely, accurate information about events such as harmful algal blooms, they may not have the sophisticated computer technology needed to easily utilize new remote sensing data sources. We are developing methods of delivering integrated data sets over the world wide web and in geographical information system environ- ments. We will demonstrate some of these integrated data products, including the integration of surface wind, temperature, currents, and bathymetry data with remotely sensed sea surface temperature and chlorophyll fields. Examples from a harmful algal bloom will be shown. ICSR. Cork. Ireland Abstracts, Sept. 29-Oct. 2. 1999 715 TOOLS FOR TURNING THE TIDE OF DETERIORATING WATER QUALITY IN SHELLFISH GROWING AREAS: A DECADE OF EXPERIENCE IN PUGET SOUND, WASH- INGTON STATE, USA. D. Fagergren, Puget Sound Water Quality Action Team. P.O. Box 40900. Olympia. WA 98504- 0900. USA. Shellfish production and consumption in the Pacific Northwest has historic and economic significance. In the mid 1980s major acreage was downgraded due to non-point source pollution from a rapidly-growing, human population along the shores of Puget Sound. Local governments, tribes, state and federal agencies, shellfish growers and citizens combined forces to halt the decline of water quality. These watershed efforts focused primarily on correcting failing on-site (septic) systems, decreasing polluted stormwater runoff and improving animal keeping practices. Turning the tide didn't come easily: many of the planning efforts were funded through new state grants, but heavy competi- tion for these funds resulted in stalled efforts in implementing actions at the local watershed level. In 1992. shellfish protection legislation was passed to make mandatory the formation of programs and districts, funded by local fees, whenever a downgrade in growing waters occurred. A shell- fish closure response strategy was also employed. Progress was slow at first; in some instances well-organized citizen opposition fought the mandatory districts designed to identify and correct fecal coliform problems. Over time, many downgraded areas began to show signs of improvement, and in 1998 a record four major commercial grow- ing areas, approximately 4500 acres, were officially upgraded. This represents 2 1 % of the total acreage downgraded in the past decade ( 1989-1998), and 42% of the total acreage upgraded during that time. Another tool, the "early warning system', was developed to alert local governments, shellfish growers and citizens in ad- vance of a downgrade and was designed to jump-start corrective actions. In early 1999. three areas are threatened by downgrades, illus- trating the difficult and continuing challenge of protecting water quality. Changing priorities for natural resource protec- tion and competition for scarce monetary resources have also compromised our ability to deal with long-term corrective actions designed to preserve and restore important shellfish growing waters. A NEW FISHERY FOR RAZOR CLAMS (ENSIS SILIQUA) ON THE EAST COAST OF IRELAND. E. Fahy, Marine In- stitute. Fisheries Research Centre, Abbotstown, Castleknock. Dul- bin 15. An investigation of the biology and the fishery for razor clams (Ensis siliqua) on the most important bed to have been discovered to date in Ireland, provides basic information on which a manage- ment plan for sustainable harvesting of this resource might be devised. The clams occur with a number of common interstitial invertebrate species. The clams range in age between 0 and 19 years-old. Males grow at a faster rate than females and the largest animals on the bed are males. These findings are in general agree- ment with what has been discovered of the biology of the species elsewhere but there are some significant differences. The Gorman- stown clams appear to be slower growing than the species in Britain or in Portugal where it has been investigated in greatest detail. The characteristics of the gonadal cycle are fairly similar but Ensis siliqua spawns later in the year at Gormanstown than off the Portuguese coast. It is reckoned to reach first maturation at three to four years of age at Gormanstown. one in Portugal. Samples for the biological investigations were provided by commercial vessels whose skippers and the processors also co- operated in making documentation available. The average age of clams captured in 1998 was 9.34 years; the following year the average age had failed to 8.34 years and a cohort (?) of very small individuals were encountered for the first time. The bed at Gormanstown has been estimated from GPS data to be 2100 hectares in extent. It is situated between the 7 m depth isopleth and it runs into the intertidal area. The original clam biomass of the bed has been calculated using data on landings which incorporates figures on rejection rates provided by proces- sors and from breakage rates observed in the samples obtained from the commercial boats — at 1500 tonnes. To the beginning of July 1999 it is estimated that more than 1,000 tonnes of clams have been removed. This bed is believed to have supplied virtually the whole Spanish market for razor clams during the past two years. ATTEMPTS TO ALLEVIATE FISHING PRESSURE ON STOCKS OF BROWN CRAB (CANCER PAGURUS) CAUSED BY THE WHELK FISHERY IN THE SOUTH WESTERN IRISH SEA. E. Fahy, Marine Institute. Fisheries Research Centre. Abbotstown. Castleknock. Dublin 15. The fishery for brown crab has. in keeping with the trend for inshore fisheries generally, greatly intensified in recent years. The animals are harvested for their claws in the spring and whole meat in the autumn. An important by-product for this fishery is bait for the whelk fishery and this use of brown crab is general wherever whelk are fished in Britain and continental Europe. Crabs are sold to the processors who also buy whelk and the crab is exchanged in the ports for whelk supplied to the factory. Crab bait is estimated to weigh 1% of the whelk harvested, al- though where the stocks of whelk have been run down, the crab bait ration remains the same and it can amount to 20% of the whelk catch by weight. At the peak of the fishery it was estimated that some 470 tonnes of brown crab were going into the whelk fishery each year causing concern among crab fishermen and conservation authorities alike. 716 Abstracts. Sept. 29-Oct. 2, 1999 ICSR. Cork. Ireland Much of the crab is supposed to originate in the claw fishery which involves discarding the bodies of the animals. However, examination of the bait used revealed that a large proportion of the crab used as bait still had claws attached. There have been reports of undersized crab being used as whelk bait and of fishing effort directed on brown crab for use as bait only. The thrust of the work described here is to examine alterna- tives to the present combinations of crab and dogfish. Two lines of enquiry were pursued: one sought alternatives among fish offals of various kinds. Whitefish. particularly cod. proved almost as good. Pelagic species (herring, scad) did not yield ac- ceptable results. The other was to test an artificial bait incorpo- rating some crab meat among a list of other ingredients but reck- oned to reduce demand for brown crab by some 60% . Results from these trials did not differ statistically from trails using natural crab bait rations. The brown crab fishery will always produce a discard and it is desirable to utilize it in some way. There is a thin line between this and the abuse of the fishery leading to inevitable over- exploitation. AZASPIRACID; DETECTION OF A NEWLY DISCOV- ERED PHYCOTOXIN IN VITRO. A. F. Flanagan,1'2 M. Kane,2 J. Donlon,' and R. Palmer,2 'Department of Biochemis- try, National University of Ireland, Galway, Ireland; ^National Diagnostic Centre. BioResearch Ireland. National University of Ireland. Galway, Ireland. Azaspiracid is the name given to a previously unknown phy- cotoxin responsible for an outbreak of shellfish based diarrhetic food poisoning in Holland in 1996. Although azaspiracid was previously classified as a diarrhetic shellfish poison (DSP), aza- spiracid toxicity has since been classified into a class of its own, called azaspiracid poisoning (AZP). Animals exposed to aza- spiracid by intraperitoneal injection react differently than those exposed to other DSPs. The marine azaspiracid response is char- acterised by hopping, scratching and progressive paralysis leading to death of the animal within 60-90 min. The development of alternative diagnostic strategies for the de- tection of phycotoxin contamination in shellfish is driven by scien- tific, ethical and financial concerns. To address this, an assay has been developed based upon the cytopathological responses of cultured mammalian cells to okadaic acid type toxins. The primary response of these cells to any of the okadaic acid family of toxins is to 'round- up' and lose their distinctive morphology. This effect, which is evi- dent less than three hours after sample application, is due to the action of toxins as phosphoprotein phosphatase inhibitors. Cells treated with okadaic acid type toxins don't exhibit any significant decrease in viability until more than 48 h post toxin application. Azaspiracid does not cause the 'rounding up' effect on cultured cells. Cells exposed to azaspiracid exhibit a slight morphological change (the cells appear to 'crenate' slightly) but their cellular viability, as measured by an MTT assay, drops to less than 90% of the viability of control cells after 18-24 h. Combining cell mor- phology observation at three hours with 24-h viability measure- ment enable the detection of both okadaic acid type toxins and aza- spiracid in shellfish. A MONITORING TOOL FOR ASSESSING OYSTER PERFORMANCES IN DIFFERENT FARMING AREAS: THE IFREMER REMORA NETWORK: METHOD AND FIRST RESULTS (1993-1998) IN BRITTANY (FRANCE). P. G. Fleury, E. Goyard, J. Mazurie, S. Claude, J. F. Bouget, A. Langlade, and M. J. Le Coguic. IFREMER. Regional Shellfish Laboratory of Brittany. 12 rue des resistants. F-56470 La Trinite- sur-mer, France. The network IFREMER/REMORA has watched every year, since 1993. from February to December, mortality, growth and quality criteria of two oyster class-size ("juveniles" of the first- year class and "adults" of the second year class), distributed among various stations of the French oyster areas. This has provided a standard and simple annual assessment of the rearing results, gen- erally tallied with professional observations. From year to year, data series have been obtained, from which mean values (refer- ences), chronological trends or spatial differences can be analysed. Results in Brittany, from 1993 to 1998. show that, in a "normal" year. 30 g oysters grow to 60-80 g at the end of the year, with a mortality rate of 10-15%. Mortality occurs mainly in spring for adults and in summer for juveniles. Beyond annual variations, mainly connected to climatic and hydrological influences, differ- ences between monitoring stations remain high, which gives data for characterization of sites. On the other hand, unusual mortalities (in 1995), lack of growth (1998) or problem of quality (infestation by the worm Polydora) could be pointed out and quantified. More- over, unusual results could be observed, which might not be ex- plained only by annual trends or geographical specificities. This gives REMORA network a role of alert and advice in the collective oyster management. Lastly. REMORA data, especially when linked to climatological or hydrological data seriess, may allow studies in many areas: description of oyster quality, biological indicators in coastal water monitoring, explanatory models of the oyster-farming ecosystems. THE HARD CLAM TASK FORCE. G. E. Flimlin. Jr., Rutgers Cooperative Extension, NJ Sea Grant Marine Extension Program. 1623 Whitesville Rd., Toms River, NJ 08755 USA. Since 1996. Gef Flimlin has been the Organizer for a group of Industry. Research Scientists. Resource Managers, and Extension Personnel who have an interest in the wild harvest and aquaculture of clams. The Hard Clam Task Force (HCTF) was formed because there was a prevailing perception among many in the "Northern ICSR. Cork. Ireland Abstracts, Sept. 29-Oct. 2. 1999 717 Quaho" Mercenaria mercenaria industry that clam harvests coast wide have been in decline for several decades, yet researchers and industry felt frustrated that little money has been directed toward this organism. Initially the group comprised of representatives from Atlantic Canada to Florida, however it was decided to include West Coast representatives working on Manila clam (Tapes japonica) since many of their issues are common with those sur- rounding. The group has identified several focus areas where a positive impact could be felt: the need to attract research money, serve as an effective conduit to standardize sampling protocols and data collection efforts so that results can be comparable, identify a number of research priorities that are designed to fill the informa- tion voids, and to use the Internet to leverage the expertise of our members and use each other as resources for experimental design, information resources, and sources of ideas for rehabilitation, edu- cation and outreach in the field of hard clam research. Using the expertise of other members, the agent has begun a stock restoration program for hard clams in New Jersey with In- dustry assistance. There have also been several other groups formed from the membership which address topics such as a Stock Assess- ment Group. Hydroacoustic Study Group, an Interstate Transport Is- sues Group, and a decision to bring more industry to the table to participate in the process, probably through a large regional meeting. The Task Force aims to marry science and management with contri- butions from academia and industry, just as the ICSR does. THE INFLUENCE OF DREDGE DESIGN ON THE CATCH OF CALLISTA CHIONE. M. Caspar,1 L. Chicharo,2 M. D. Dias,'' P. Fonseca,4 A. Campos,4 M. N. Santos,1 and A. Chi- charo,2 'Institute de Investigacao das Pescas e do Mar, Centro Regional de Investigacao Pesqueira do Sul. Av. 5 de Outubro, 8700 n 305 Olhao. Portugal; 2Universidade do Algarve, UCTRA, Campus de Gambelas, 8000 Faro, Portugal; 'institute de Investi- gacao das Pescas e do Mar, Del. de Setubal. Av. Jaime Rebelo 29 A. 2900 fi 409 Setubal. Portugal; 4lnstituto de Investigacao das Pescas e do Mar. Av de Brasilia, 1400 Lisboa, Portugal. An important bivalve fishery takes place along the Southwest coast of Portugal. At present, in this part of the Portuguese coast, the most important commercial species is the clam Callista chione. This species constitutes the target of a specific fishing activity carried out by an artisanal fleet. Dredges are the gear used in this fishery. The shape and structure of the Portuguese dredge has been maintained constant throughout time and consists of a small, heavy semicircular iron structure, with a net bag and a toothed lower bar at the mouth. Welded to this iron structure are three metal shafts forming a kind of hen's foot where the towing cable is attached. This year, some fishermen from the region of Setubal. have been trying to introduce a new type of dredge in the fishery, debating, therefore, being a gear that maintaining the fishing efficiency rela- tively to the target species, carries less damages to the accompa- nying species. The basic difference between both dredges resides in the retention structure of the bivalves. In fact, in the new dredge. the net bag is replaced by a rectangular metallic grid. To evaluate a possible introduction of this new dredge in the fishery, IPIMAR has conducted a study with the objective to compare the efficiency of the gears and to evaluate the impact produced by both dredges in the benthic community. The experiments were carried out dur- ing March 1 999, on the Southwest coast of Portugal, from a site off Troia. Three tow duration 5. 10, and 20 minutes were investigated. A total of 24 hauls was completed. 4 for each tow duration and dredge. The experiments were conducted by attaching to the gear a cover bag with a 20 mm mesh. After each haul the catches in the bag and in the cover were sorted separately. All animals retained are allocated scores on a scale of 1-^4 in which 1 equate to good and 4 equate to dead. The results obtained show that the catches coming from the traditional dredge (TD) are composed of a great fraction of juveniles of C. chione. where the new dredge (ND) are composed of a quite great number by individuals with a superior size to the minimum legal length (50 mm). This result is an indi- cator that the mesh of the bag of the TD used in the exploitation of this resource is not adequate. For the 3 tow duration, the mean fishing yield obtained for the ND was always superior to the TD, due to its great efficiency in capture. For all 3 tow duration, the impact caused by both dredges on the target species and upon the macrobenthic community in general (percentage of damaged indi- viduals) was similar and low. The greatest advantage in the usage of the ND relatively to the TD, is to allow the smallest individuals (independently of the species) to escape rapidly through the me- tallic bars on the grid, increasing their probability of survival. THE AFFECTS THAT AMNESIC SHELLFISH POISON- ING HAS ON SCALLOP PROCESSORS AND COMMER- CIAL FISHERMEN. R. Gidney and J. Hermse, Scallop Dredg ing Association, 5 Boat Croft. Kemnay, Aberdeenshire. AB5 1 5GZ. Algal toxin monitoring in Scotland is earned out by the Fishery Research Services, Marine Laboratory in Aberdeen. The monitor- ing is required under EU directive (91/492). There have been sea areas closed to commercial scallop fishermen due to Amnesic Shellfish Poisoning (ASP) for some 10 weeks (July-September 1999) now on the Scottish West Coast. Smaller closed areas have also been in force around the Orkney Islands and East of Aber- deen. In some cases the levels of ASP toxins have been extremely high (231 ug/g) when the statutory maximum level is 20 ug/g. The ASP toxin is affecting the commercial fishing of king scallops (Pectin maximus) and queen scallops (Aquipecten oper- cularis). Fishermen and processors have been extremely diligent in observing and protecting the closed areas in order to preserve consumer confidence. However, the industry is becoming restless as the financial burden of such a lengthy ban begins to bite. The 718 Abstracts. Sept. 29-Oct. 2, 1999 ICSR. Cork, Ireland size of closed areas, the integrity of sampling procedures, the testing methodology and the causes of the high levels of toxins are amongst the questions now being asked. A meeting of the Scottish Parliament (at which the SDA will make a presentation) will shortly debate the problem and make recommendations. The Scallop Dredging Association (SDA) is suggesting a pooling of end product testing results and other resources with the scientific community in order to better co-ordinate the testing and sampling procedures. The SDA is requesting that research is required into the forecasting, causes and effects of Harmful Algal Blooms. THE INFLUENCE OF OYSTER CULTURE STRUCTURES ON BIRDS IN A SHELTERED IRISH ESTUARY. G. Hilger- loh, J. O'Halloran, T. Kelly, and G. Burnell, Dept of Zoology & Animal Ecology. University College Cork, Ireland. In an estuarine bird feeding area on the south coast of Ireland (Cork Harbour) a commercial company had installed oyster grow- ing tables (trestles), to which bags of oysters were attached. The tables were arranged in long rows separated by lanes left for the tractors. The study investigated birds in the shellfish culture area. The occurrence of the different species and their behaviour were compared to a nearby area free of aquaculture within the same estuary. First results showed that all species which occurred in the aquaculture free area could also be observed in the trestle-area. The most frequently observed species were oystercatcher Hae- matopus ostralegus. redshank Tringa totanus, dunlin Calidris al- pina, curlew Numenius arquata. black-headed gull Larus ridibun- dus and common gull Larus canus. The percentage of feeding birds per species did not differ in the two areas of comparison. However, there were significant differences in density in four of the six studied species. ARTIFICIAL REEFS FOR SHELLFISH HABITAT: RE- SULTS AND IDEAS TO DATE. A. Jensen, School of Ocean and Earth Science, University of Southampton Oceanography Cen- tre, Southampton. Artificial reefs are used worldwide in a variety of roles: pro- motion of fishery catch, habitat protection and recreation being the most common. In Europe the majority of artificial reefs have been placed for habitat protection (Posidonia and other seagrass mead- ows) and developing finfish fisheries yield. The use of artificial reefs for shellfish culture/ranching is in its infancy in Europe, existing as a reality in the Adriatic sea where mussel and oyster cultivation has become possible because of ar- tificial reef deployment and as a ranching proposal in Northern Europe where several strands of the knowledge needed to promote lobster ranching have been established but still require some re- search before development can occur. In the Adriatic Sea the combination of eutrophic. shallow water and significant natural mussel larval production have provided the conditions which have facilitated development of mussel and oys- ter cultivation in association with Italian 'pyramid reefs'. In northern Europe the deployment of artificial reefs, develop- ment of lobster hatcheries and research showing the effective sur- vival of hatchery reared lobsters and inclusion in a fishery are the basis for further work to make lobster ranching a reality. This paper summarises the work undertaken in Europe (and elsewhere) to date and speculates about the future possibilities of promoting shellfish cultivation by using artificial reef technology. Japanese workers have pioneered the use of artificial reefs for abalone habitat and developed coastal structures to influence larval settlement. In order to make artificial reefs an economic reality in Europe, we must learn from their example, particularly in the large scale of their operations. BREEDING PONDS AS A BASIS FOR FLAT OYSTER {OS- TREA EDULIS) CULTURE AND THEIR USE TO DEVELOP RESISTANCE TO THE DISEASE BONAMIA OSTREAE. D. Hugh-Jones, Loch Ryan Shellfish Ltd.. c/o the Thatched Cottage. Penberth St.. Buryan, Penzance, England TR 6HJ. Flat oyster production in Cork over the last quarter of a century has relied entirely on its own production of spat from shore-based breeding ponds. Over this period, selection for fast growth has led to the routine production of market-size oysters of 70-120 g in 3 years. Since the stock was almost totally destroyed by the disease. Bonamia, in 1987, the breeding objective has since concentrated on mass selection for resistance to the disease. This work, which in currently being evaluated in Ireland. France and Holland, appears to be showing promising results, with production in Cork Harbour, where oysters are exposed to the disease for their entire growing period, rising to as much as 80% of former levels. THE COMMERCIAL POTENTIAL OF THE COMMON SEA URCHIN ECHINUS ESCULENTUS. M. S. Kelly, P. Pan- tazis, and P. Owen, Scottish Association for Marine Science. PO Box 3. Oban, Argyll. PA34 4AD. Scotland, UK. As over fishing continues to delibitate world sea urchin popu- lations there is increased interest in echinoculture as a means of meeting market demand for urchin roe. Inherent in this is the investigation of the commercial potential of species previously considered unviable. The regular echinoid Echinus esculentus is widely distributed around the British Isles. While islanders and maritime communities once considered this species as a food re- source, the gonad is now generally considered to be of a poor quality and unsuited to the European palette. The skins of roe are however of a similar size to those of the Japanese species har- vested for the domestic market. Therefore an investigation of its commercial potential was initiated. A survey was conducted of gonad content and colour in E. esculentus from the Scottish west coast from June to August 1998. ICSR. Cork, Ireland Abstracts, Sept. 29-Oct. 2. 1999 719 The survey found the gonad content was variable in terms of weight and colour. A relatively small proportion of urchins (35%) had a gonad which was considered marketable in terms of both yield and colour. Urchins with a marketable yield of roe were found where algal growth or encrusting organisms were most pro- lific. Good roe colours and high roe yields were not necessarily coincident. Tank based diet trials were then conducted to assess the ability of this species to utilise pelleted feeds. E. esculentus were fed either a commercially available salmon feed, an artificial diet designed for sea urchins or a more natural macroalgal diet. The artificial diets enhanced gonad growth. The urchin diet also ben- eficially influenced gonad colour. The data suggest a fishery is likely to be impractical economically and could potentially be very destructive for the wild urchin population. A fully integrated cul- ture approach is therefore the most feasible option for commercial exploitation of this species. An artificial diet could be used to enhance gonad growth, colour and flavour. The aim would be to produce urchins of test diameter 45-50 mm. within 2 years, with a premium quality roe export to the far eastern markets. ECOLOGICAL EFFECTS OF FISHERY DISTURBANCE TO OYSTER REEF HABITAT IN EASTERN NORTH AMERICA. H. S. Lenihan1 and G. W. Thayer,2 'University of North Carolina at Chapel Hill. Institute of Marine Sciences. 3431 Arendell St., Morehead City, North Carolina. 28557: 2National Oceanic Survey, Beaufort Laboratory, 101 Piveris Island Rd. Beaufort, North Carolina. USA, 28516 Eastern oysters. Crassostrea virgirdca, are ecosystem engineers that create biogenic reef habitat important to estuarine biodiversity, benthic-pelagic coupling, and fishery production in North America. Oysters and oyster reefs have declined dramatically over the last century due to overfishing, habitat destruction, reduced water quality, and/or the introduction of diseases. We provide evidence from experiments conducted in the Neuse River. North Carolina, that oyster loss results from interactions among multiple environmental disturbances. Our results indicate that destructive harvest practices reduce the height of oyster reefs, causing flow speeds across reefs to decrease. Reduced flow speeds on reefs caused increased sedimentation and decreased the quality of sus- pended food material for oysters. In turn, these factors helped explain why oysters on harvest-disturbed reefs grew slowly, were in relatively poor condition, had relatively high incidences of para- sitism, and died at relatively high rates. Oysters on harvest- disturbed reefs located in deep water were exposed to bottom- water hypoxia/anoxia during estuarine stratification, killing oysters and other reef-associated invertebrates and demersal fishes. We also found that oyster reefs were utilized by 23 species of com- mercially valuable fishery organisms (fishes and the crab, Calli- nectes sapidus), many of which recruited to reefs and/or used them as foraging substrate. Fishes occupying harvest-damaged reefs in deep water were forced into shallow water during a hypoxic/ anoxic period. Escaping fishes accumulated in high densities on reefs in shallow water where they decimated prey populations. Therefore, indirect trophic interactions in undisturbed habitats re- sult from multiple human disturbances in remote, disturbed habi- tats. In general, interactions among environmental disturbances imply a need for the integrative approaches of ecosystem manage- ment to restore and sustain estuarine habitat. THE INTEGRATION OF REMOTE SENSING DATA WITH LOCAL AND STATE MONITORING DATA. D. L. Leonard, National Oceanic and Atmospheric Administration. National Ma- rine Fisheries Service, Habitat Conservation Office. 1315 East West Highway. Silver Spring, MD 21401. USA. The coastal states of the United States are all required under the National Shellfish Sanitation Program to prepare Marine Biotoxin Contingency Plans in anticipation of harmful algal events. In the last 15 years these HAE events have affected 20 of the 23 coastal states. The costs to the states and the industry are estimated at 50 million dollars annually. Emerging technologies such as remote sensing, spectral analysis, and internet access can provide an early warning system for states and industry while new laboratory meth- ods can expedite identification of toxin type and levels. On the local level the mussel watch programs maintained by states and plankton sampling supported by volunteers can be used to verify the data received from the federal government. In response to the growing need for support the US HAB National Contingency Plan will use a National Clearinghouse for direct transfer of information and access to the technical expertise of federal and state agencies. academia, and private industry on a day-to-day basis and an early alert system for most toxic effects. A National Event Response Coordinator will alert all federal agencies of an event and arrange for the specific expertise and technical assistance to be provided on a national and local level by appropriate agencies. UNDERSTANDING THE FACTORS THAT INFLUENCE EUROPEAN LOBSTER RECRUITMENT: A TRANS- EUROPEAN STUDY OF COBBLE FAUNA. A. Linnane,1 B. Ball,1 J. P. Mercer,2 G. van der Meeren,' C. Bannister,4 D. Mazzoni,' B. Munday,1 and H. Ringvold/ 'Martin Ryan Science Institute. National University of Ireland. Galway. Ireland; Na- tional University of Ireland, Galway, Shellfish Research Labora- tory. Carna, Co Galway. Ireland; 'Institute of Marine Research. Austevoll Aquaculture Research Station. N 5392 Storebo. Nor- way; 4CEFAS. Lowestoft laboratory. Pakefield Road. Lowestoft. Suffold NR33 OHT. UK; "DIPROVAL. via del Guasto 5/B. Uni- versita Degli Studi di Bologna, 40126, Bologna, Italy. The re-stocking or enhancement of Homarid populations has been the focus of several research projects in recent years. How- ever, despite being able to successfully rear lobsters in captivity, there are few reports of newly settled European lobsters [Horn- 720 Abstracts. Sept. 29-Oct. 2. 1999 ICSR, Cork, Ireland marus gammarus) in the wild. This is of serious concern for fish- ery managers when trying to ascertain if restocking or enhance- ment programmes are successful at the fishery level. Based on published work from the eastern Atlantic seaboard cobble and boulder substrata have been identified as an important nursery habitat for early benthic phase (EBP) American lobsters [Homarus americanus). In Europe, cobble and boulders are uti- lised by adult lobsters yet little is known of the associated fauna of these substrata, primarily due to the difficulties involved in sam- pling such sites. This paper, based on an E.U. funded research project (LEAR), describes the results of quantitative airlift suction sampling from cobble habitat in Norway, the UK Ireland and Italy. Overall, crustaceans and molluscs were the most abundant species in all countries. While the collective densities of animals per m" of cobble were similar to that of the United States, the species diver- sity in Europe was significantly higher. Among the Crustacea, the reptant decapods dominated at all sites. These findings tend to support the modern day "exclusion hypothesis" that competition and predation from other species limits the successful recruitment of EBP European lobster to the benthos. The results are being used to devise both models and experiments designed with the aim of providing a greater insight into the factors that influence the re- cruitment of Homarus gammarus. APPLICATIONS OF UNDERWATER IMAGING TECH- NIQUES IN THE MONITORING AND RESTORATION OF COASTAL HABITATS. U. Lobsiger and J. L. Manuel, Tris- Mar Research Inc.. One Research Drive. Dartmouth, NS, B2Y 4M9, Canada. Underwater imaging techniques are being applied in many ap- plications of coastal zone management. A review of such tech- niques is presented with specific case studies and two novel de- velopments that are particularly relevant to coastal shellfish re- sources will be discussed in some detail. The relative merits of tethered video with surface display vs. autonomous image acquisition will be demonstrated using a num- ber of habitat monitoring examples. They include: A winter sur- vey of benthic energy dynamics in a high-Arctic inlet: the "Mer- maid" remotely operated imaging vehicle: the "Video Grab", the "CAMPOD", as well as time lapse stereo and sediment-profiling photography of polychaete assemblages. Other examples include the direct observation of fishing gear in action with MANTA and in situ cameras on scallop rakes and fishing trawls. It will be further shown that quantitative sea-floor imaging can serve as ground-truthing for synoptic techniques, including side-scan sonar, RoxAnn, QT Seabed, and hyperspectral airborne imaging. Thanks to new technology developments in image acquisition, storage, communication and interpretation, novel applications with much better cost/benefit ratios of information value over logistical costs are now possible. Smaller size, lower cost, real-time options of satellite and internet communication and powerful PC-platform image analysis software render imaging as a powerful adjunct and as a primary tool of investigation for many coastal habitat surveys and monitoring programs. The MarineCanary™ (patent pending), a concept that uses the feeding behaviour of bivalve mollusks as keystone species of coastal ecosystem health will be discussed in detail. Behaviour patterns of a statistically relevant number of animals are monitored in situ with time lapse imaging over days and weeks, and behav- iour is correlated with growth and reaction to contaminants and toxins. A new ultra rugged and cost-effective digital time lapse camera system will be introduced. OYSTER REEF RESTORATION IN VIRGINIA, USA: RE- HABILITATING HABITATS AND RESTORING ECO- LOGICAL FUNCTIONS. M. W. Luckenbach, J. Harding, R. Mann. J. Nestlerode, F. X. O Beirn,1 and J. A. Wesson,2 'Vir- ginia Institute of Marine Science, College of William and Mary. Gloucester Point. VA 23062. USA: :VMRC, 2400 Washington Street, Newport News, VA, USA. Repletion efforts in response to declines in abundance of the east- ern oyster, Crassostrea virginica, have historically relied upon trans- planting of oyster seed and planting of a suitable settlement substrate. These efforts have generally failed to revitalize the fishery because they ( 1 ) failed to rehabilitate degraded reef habitat and (2) placed little emphasis upon reestablishing a population age structure capable of sustaining a self-supporting reef. More recently restoration efforts in Virginia have focussed on reconstructing 3-dimensional reef habitats and establishing brood stock sanctuaries with an emphasis on restor- ing lost ecological functions of reefs. Manipulative studies of reef placement, construction material and interstitial space have lead to the development of design criteria for maximizing oyster recruitment, growth and survival on constructed reefs. Further, we have charac- terized the successional development of resident macrofaunal com- munities on restored reefs and have begun to relate that development to specific habitat characteristics. Utilization of these restored reef habitats by transient species has been characterized through extensive field collections and underwater video observations: gut analyses of finfish are beginning to elucidate trophic linkages between the reefs and adjacent habitats. In addition, these structures appear important to the early developmental stages of juvenile fishes, some of which have considerable recreational and commercial importance. These studies are helping us to ( I ) clarify the ecological functions supported by oyster reef habitat, (2) define design criteria for reconstructing reefs and (3) establish success criteria for such restoration projects. While destructive fishing of oyster reefs appears inconsistent with meeting these goals, an emerging paradigm is that reef sanctuaries can be used to support desired ecological functions as well as supply recruits to adjacent areas which can be managed from a fisheries perspective. ICSR, Cork. Ireland Abstracts. Sept. 29-Oct. 2. 1999 721 HABSOS: A PILOT PROJECT OF THE US GLOBAL OCEAN OBSERVING SYSTEM AND THE NATIONAL AS- SOCIATION OF MARINE LABORATORIES. T. Malone and C. Horn, Point Laboratory, UMCES. PO. Box 775. Cambridge MD 21613 USA. Estuarine and coastal marine ecosystems are subject to convergent inputs of materials and energy from terrestrial, atmospheric, oceanic and anthropogenic sources that vary over a broad range of time-space scales. As a consequence, coastal ecosystems are experiencing un- precedented changes related to climate change, rapid increases in the number of people living near the coast, the exploitation of coastal resources, atmospheric deposition, and land-use practices in coastal watersheds. In this regard, there are some indications that the occur- rence of "harmful algal blooms" is increasing in both time and space. However, the evidence for such a trend is not compelling and the causes of HAB events are poorly understood. It has become painfully clear that the scarcity of observations on coastal ecosystems of sufficient duration, spatial extent, and resolution and the lack of real-time data telemetry, assimilation and visualization are major impediments to the documentation of pattern and to the development of a predictive understanding of environmental variability and change in coastal waters. This prob- lem is particularly evident in the case of HABs. This paper de- scribes an effort to network coastal marine laboratories and gov- ernment agencies to design and implement an observing system tor HABs (HABSOS) that will provide timely access to data and information on HAB events, the environmental conditions under which they occur, and their impacts. The goal is to systematically document the time-space dimensions of HAB events and their effects, to forecast their occurrence and issue alerts, and to mitigate deleterious effects. THE MANAGEMENT OF FRESHWATERCATCHMENTS. P. McGinnity and K. F. Whelan, Marine Institute. Salmon Man- agement Services, Furnace, Newport. Co. Mayo. Ireland. The basis of holistic catchment management can be traced back over twenty years, but it is only during the past five years that is true significance has been recognised by resource mangers. There is increasing recognition that the exploitation of natural resources such as the processes associated with agricultural production, for- estry and other forms of resource usage impact negatively on lands and waters. This is well recognised in fisheries management. How- ever, current responses to such phenomena are often poorly planned and lack a strategic dimension. The development of an alternative approach to resource management, a catchment or wa- tershed management approach, which better integrates the main- tenance of ecological values with resource exploitation and devel- opment, has been identified as an essential response. This paper traces the history of catchment management in Ireland and its evolving role as a fundamental strategy in the long-term manage- ment of fisheries and freshwater resources. Although intergrated catchment management has been become a buzzword amongst the resource management sector, to date it lacks a clear-cut definition. This paper describes in some detail the practical field programmes which has recently been undertaken to test the feasibility of develop- ing data acquisition technologies to support a clearly defined catch- ment management strategy. Finally the paper highlights the pragmatic and philosophical hurdles which must be quickly overcome if catch- ment management is to fulfil its initial promise as a novel, funda- mentally sound pillar in modern integrated resource management. A COMPARISON OF THE ARKSHELL CLAMS, ANA- DARA OVALIS AND NOET1A PONDEROSA, AS POTEN- TIAL MARICULTURE SPECIES ALONG THE ATLANTIC COAST OF THE UNITED STATES. K. A. McGraw1 and M. Castagna," 'State University of New York. P.O. Box 2000. Cort- land. NY. 13045-0900. USA; 2Virginia Institute of Marine Sci- ence. Eastern Shore Laboratory, Wachapreague, VA, 23480, USA. A new fishery for arkshell clams (Anadara ovalis and Noetia ponderosa) began along the coast of Virginia in 1991. Before that time, arkshell clams were considered a useless by-catch with the harvest of the hard clam. Mercenaria mercenaria, and simply dis- carded. Since no regulations exist for the harvest of arkshell clams in Virginia, fishermen quickly depleted stocks in some areas to meet a constantly increasing market for the clams. Our studies over the last several years on A. oralis and N. ponderosa indicate that the former is a good candidate species for mariculture. Although both species have been spawned and reared previously using standard hatchery techniques. Von Bertalanffy growth curves show that the growth rate of A. ovalis is about twice as fast as that of N. ponderosa in Virginia waters (2.5-3 years for A. oralis, compared to at least 6-8 years for N. ponderosa to reach average market size of 56 mm in shell height). In addition. A. oralis averages about 10% more meat for a given size category because it has a much thinner shell than N. ponderosa. We plan to pursue a pilot study using hatchery-reared A. ovalis seed clams and a conventional off-bottom culture technique to test the feasibility of large-scale mariculture endeavors. A good market for arkshell clams exists in the United States, and, if successful, mariculture of A. ovalis could provide a reliable source of seed clams and market-sized clams, help restock depleted arkshell stocks, and alleviate some of the overfishing of arkshell clams occurring in the oceanside lagoon sys- tems of Virginia. THE CANADIAN SHELLFISH SANITATION PRO- GRAM—HOW SHELLFISH CLOSURES ARE LEADING TO IMPROVED WATER QUALITY. E. McKnight1 and H. Nelson,2 'Marine Environment Division, Environment Canada, 351 St. Joseph Blvd.. Hull, Quebec, K1A OH3. Canada; Pollution Prevention and Assessment Division. Environment Canada. 224 West Esplanade. North Vancouver. British Columbia. V7M 3H7. Canada. The Canadian Shellfish Sanitation Program (CSSP) is a tripar- tite program administered by three federal agencies to ensure the 722 Abstracts. Sept. 29-Oct. 2. 1999 ICSR. Cork, Ireland safe consumption of molluscan shellfish. Based on the U.S. Na- tional Shellfish Sanitation Program (NSSP). the first critical con- trol point of the CSSP is the proper classification of shellfish growing areas. Classifications, based on sanitary shoreline evalu- ations and water quality surveys, are Environment Canada's re- sponsibility under the CSSP. Shellfish closures made under the CSSP have focused the need for improved water quality in Cana- da's coastal areas. Environment Canada's major ecosystem initiatives in its coastal regions empower local communities to address their envi- ronmental challenges, including those related to water quality. On Canada's east coast. Environment Canada provides seed funding to thirteen east coast communities to develop and implement com- prehensive environmental management plans. On the west coast of Canada, the major environmental stew- ardship program is the 5 year Georgia Basin Ecosystem Initiative (GBEI) which began in 1998. This initiative resulted from a grow- ing concern in the early 1990 is that the health of the marine environment was deteriorating rapidly in relationship to population growth and development along the coast. The degradation was clearly demonstrated by the closure to harvesting of an increasing number of commercial and recreational shellfish beds. Concerted efforts at the community level to identify and reme- diate non-point pollution are gaining momentum and some of the early initiatives from the mid 1990s are showing positive results. New programs and funding agreements through the GBEI and new regulations under the Canada Shipping Act to restrict sewage dis- charges from vessels will be discussed. Classification upgrades to shellfish growing areas on the west coast of Canada reflect a growing improvement in general marine environmental quality and shoreline amenities. IRISH COASTAL DINOFLAGELLATE BLOOMS AND SHELLFISH TOXICITY. T. McMahon,1 J. Silke,1 and B. Ca- hill,2 'Marine Institute. Fisheries Research Centre, Abbotstown, Dublin 15, Ireland; "Marine Institute. Irish Marine Data Centre. 80 Harcourt Street. Dublin 2. Ireland. Since the mid 1980s the Marine Institute's Fisheries Research Centre has carried out a monitoring programme in Irish Coastal waters involving both the identification and quantification of phy- toplankton species present in near surface waters and the detection in shellfish of toxins of algal origin. Recurrent blooms of the dinoflagellate Gyrodinium aureolum have been recorded, particularly along the southwest coast. In several cases mortalities of marine fauna, including farmed finfish and shellfish, have been associated with these blooms. Recent studies have shown that the blooms originate offshore and are subsequently advected into the bays under a particular set of me- teorological conditions. Details of the deployment of two moored instrument arrays, with near real-time data telemetry, designed to provide an early warning of these bloom events are described. Results from the algal toxin monitoring programme are pre- sented and discussed. The results show that Irish shellfish can have a very complex toxin profile. Toxins (Gonyautoxin-2 and Gon- yautoxin-3) causing Paralytic Shellfish Poisoning (PSP) associ- ated with the presence of the dinoflagellate Alexandrium tama- rense have been detected in one production area. Toxins (Okadaic acid and DTX-2) causing Diarrhetic Shellfish Poisoning (DSP) associated with the presence of the dinoflagellates Dinophysis acuta and Dinophysis acuminata have been detected in many of the main shellfish production areas in the country. More recently a novel toxin, azaspiracid as well as several of its analogs, of unknown origin have been also been detected in shellfish in Ire- land. Shellfish toxicity has occurred during both summer and win- ter months and therefore monitoring is now a year round activity. A. tamarense is known to produce a cyst stage as part of its life cycle. The cysts can remain viable in the sediments for several years. A survey of the distribution of cysts of A. tamarense in the surface sediments in Cork Harbour was carried out in order to determine if a "seed bed" of toxic cysts was present in the area. The results of the survey are presented and discussed. EXOTIC SPECIES: IMPLICATIONS FOR COASTAL SHELLFISH RESOURCES. D. Minchin, Marine Institute, Fish- eries Research Centre. Abbotstown. Dublin, Ireland. Introduced species form the basis for the development of many important economies in Europe and worldwide. Many of these species are globally in production and the successful development of a species in one part of the globe is followed by trials elsewhere. Unfortunately there is a dependence on a small number of species and it is likely that there are many other species that could be profitably used. The ICES Code of Practice forms the basis for a responsible procedure for the introduction of 'new' species for culture. Exotic species are in constant transit whether in with live-trade or inadvertently with the transport ships. Although the great ma- jority of these species are not known to have great impacts there are some that will have consequences that man affect human health or may compromise coastal shellfish resources. The trade of live species, and in particular the trade of half-grown oysters, continues to be implicated in the range expansions of molluscan pests. Mea- sures taken to reduce the spread and impacts of harmful species in the case of those species moved by shipping is difficult to achieve. Treatment measures are being actively researched. It appears that the continued expansion of phytoplankton toxins may in part be due to innoculations from ballast water discharges by ships. The close prox- imity of shellfish resources to ballast water discharge sites and ber- thing regions could compromise shellfish production. There are indi- cations that TBT may have influenced the populations of scallops in Cork Harbour and the improving water quality there may lead to a recovery of these populations. It may also provide more suitable conditions for invasive species. It is therefore predicted that once the ICSR. Cork. Ireland Abstracts, Sept. 29-Oct. 2. 1999 723 planned discontinuation of the usage of TBT takes place that there will be more frequent exotic species invasions. This paper examines some species that are likely tp become introduced to Ireland in the coming early century and the conse- quences of these introductions for the shellfish economy. ON THE DYNAMICS OF INTER-TIDAL AND SUB-TIDAL STOCKS OF BLUE MUSSELS IN THE DANISH WADDEN SEA. S. Munch-Petersen and P. Sand Kristensen, The Danish Institute for Fisheries Research (DIFRES), Charlottenlund Castle. DK-2920 Charlottenlund, Denmark. As part of a monitoring programme for the commercially ex- ploited stock(s) of mussels (Mytihis edulis L.) in the Danish Wad- den Sea. samples have been collected regularly from 1990 to 1998. both from sub-tidal and inter-tidal settlements. These samples are the basis for the estimated size composition of the mussels as well as estimates of total biomass by locality. The observed size distributions have been used for identifica- tion of cohorts. Cohorts are easy to identify and follow during the first year after settlement. Later, however, the size of the individu- als in the cohorts will vary to an extent such that the size distri- butions seemingly merge into one single group By slicing the size distributions into components representing cohorts using the avail- able growth parameters for Mytilus in the Wadden Sea and assum- ing certain values of standard deviation of the mean length-at-age this development of the composed size distributions is demon- strated. Estimates of annual production are compared to production figures for similar localities in the German Wadden Sea. DECREASE IN HETEROZYGOSITY IN SUSPENSION- CULTURED BLUE MUSSELS FOLLOWING THEIR TRANSFER TO GROW-OUT SITES AND ITS POTENTIAL CONSEQUENCE ON MUSSEL FARM PRODUCTIVITY. B. Myrand.1 R. Tremblay,2 and J.-M. Sevign.' 'DIT-MAPAQ. C.P. 658. Cap-aux-Meules, Canada, GOB 1B0; "GIROQ, Univer- site Laval, Quebec, Canada. G1K 7P4; Mnstitut Maurice- Lamontagne. MPO-Canada. 850 Route de la Mer. Mont-Joli. Canada, G5H 3Z4. In the Magdalen Islands (southern Gulf of St. Lawrence. Canada), spat is collected in a small lagoon, placed in mesh sleeves, and transferred to other sites for grow-out in suspension- culture. We observed a significant decrease in multilocus heterozy- gosity (MLH) estimated at 7 allozyme loci for mussels sleeved at two densities (usual density: 2.44 ±0.14 and high density: 2.14 ± 0.2 1 ) compared to that of the spat (2.95 ± 0.08) used for sleeving. This decrease in MLH occurred within the first year after the transfer to the grow-out site. It was not the result of adverse con- ditions at the grow-out site. We hypothesize that more heterozy- gous individuals are more active and thus get out of the mesh sleeves more rapidly. Doing so. they are more prone to fall-offs. mainly those resulting from turbulence created by heavy winds over the shallow lagoons while they are still weakly attached. This decrease in MLH may have important impacts on the productivity of the local mussel industry. First, we observed an inverse rela- tionship between MLH and basal metabolism, so that more het- erozygous individuals have lower metabolic needs. Further, we observed that more heterozygous individuals were significantly more resistant to various stressful conditions. Indeed, the MLH of survivors to stressful conditions was significantly higher than that of the controls during two successive experiments (August and September 1997). The LT50 of 50-60 mm suspension-cultured mussels (MLHf: 2.70 ± 0.16 in August and 3.08 ± 0.14 in Sep- tember) was also systematically lower than that of 50-60 mm wild mussels taken from the spat collection site (MLHt: 4.00 ± 0.18 in August and 3.82 ± 0.21 in September) under identical stressful conditions. It seems that the mussel growers are not taking full advantage of the spat potential as they are losing the more hetero- zygous individuals while in suspension-culture. EVALUATING DESIGN CRITERIA AND RECRUITMENT LEVELS IN THE RESTORATION OF OYSTER REEFAS- SEMBLAGES. F. X. O'Beirn, J. A. Nestlerode. and M. W. Luckenbach, Virginia Institute of Marine Science, College of William and Mary. Gloucester Point. VA 23062. USA. The construction of reef structures to promote shellfish habitat restoration represents a significant investment of public and/or private resources. Thus, it is essential that effective design and construction protocols be established and evaluated on sufficiently large spatial and temporal scales to ensure success. These scales often exceed those of standard manipulative field experiments. We describe findings from large-scale field experiments, begun in 1995-1996. addressing the interactive roles of substrate type and tidal elevation on the early development of oyster reef assem- blages. In addition to the development of oyster populations, we examine successional trends in macrobenthos associated with the reefs and relate these to reef design characteristics. We observed a particularly strong interaction between substrate type and regional recruitment level (i.e.. high vs. low recruitment years) on the de- velopment of reefs, adding emphasis to the importance of scale in restoration. The relationship between the development of oyster populations and a variety of community parameters we hope will resolve the temporal and spatial scales that must be applied to oyster reef restoration efforts. Concomitantly, we hope to deter- mine more accurately the appropriate success criteria for similar restoration endeavors. 724 Abstracts. Sept. 29-Oct. 1999 ICSR. Cork, Ireland CO-ORDINATED LOCAL AQUACLLTURE MANAGE- MENT SYSTEMS (C.L.A.M.S.). T. O'Carroll, BIM. PO Box 12, Crofton Road, Dun Laoghaire, Co. Dublin. As distinct from national aquaculture policy and development programmes the concept of Co-ordinated Local Aquaculture man- agement Systems (C.L.A.M.S.) focuses at the local bay level (while still taking on board relevant national policies). The concept is to amass relevant baseline data and in conjunction with the formation of a local C.L.A.M.S. group, to formulate an aquacul- ture development plan for the bay while incorporating and extend- ing the successful concepts of Single Bay Management to all farmed species (including extensive culture). C.L.A.M.S. will provide the following tangible outputs: — • A concise description of the bay/area in terms of physical char- acteristics, history, aquaculture operations, future potential, problems, etc • Integration of a series of codes of practice for current aquacul- ture operations and translation of those national codes to the specific circumstances of each bay or coastal region • Expansion of the concept of SBM to species other than salmon • A development plan for aquaculture in the bay • Information on other activities in the bay • A local and national communication network, with "bottom up" and "top down" dialog capacity In addition C.L.A.M.S. intends to incorporate the development plans of the local individuals as well as integrating the manage- ment practices of the various species sectors that may be operating in the same bay. Over the next few years the concept of Coastal Zone Manage- ment will become an applied legislative reality. It is envisaged that C.L.A.M.S. could form the basis for the aquaculture section of any new local Coastal Zone management initiatives. ABSOLUTE DETERMINATION OF CHLOROPHYLL CONCENTRATION AND OPTICAL PROPERTIES OF WA- TER BY AIRBORNE HYPERSPECTRAL REMOTE SENS- ING. E. O'Mongain and A. Collins, Spectral Signatures Limited. Roebuck, Belfield, Dublin 4, Ireland. The absorption characteristics of the pigments of chlorophyll modulate the observable optical properties of marine and inland waters. In clear open ocean waters the blue pigment of chlorophyll is used for chlorophyll estimation by remote sensing. In coastal and inland waters, the blue pigment is masked by other absorption processes. In these waters it is the red pigment which is observed by spectral remote sensing. Airborne measurements over Wexford and Cork estuaries show the distribution of near surface chloro- phyll concentration. The technique compares modelled and mea- sured hyperspectral reflectance in a way which utilises the known absorption characteristics of pure water to provide measurement calibration. An absolute measure of the total absorption in the 580 nm region of the spectrum, which is strongly related to visibility or sechi depth, is also obtained. Thus sediment and chlorophyll dis- tribution are independently monitored for validation of hydrody- namic and eutrophication modelling projects. For surface based validation measurements the ChloroFlow in- strument system similarly utilises the red pigment of chlorophyll to provide in-vivo photometric chlorophyll estimates in real time on a flow-through basis. This instrument also provides measurements of the full spectral absorption curve for investigation of the role of other pigments and sediments and for reflectance modelling exer- cises. Absolute measurements of spectral absorption (+ back- scattering) in a long pathlength cuvette is the basis of this system which has gone into service in U.K. marine waters this season to assist in the monitoring of sensitive waters. CHEMICAL EFFECTS OF HYPOXIC AND ANOXIC WA- TERS ON SHELLFISH. M. J. Orren, Oceanography Depart- ment, The National University of Ireland, Galway, Ireland. Normal, ("healthy") coastal sea waters are 90%. or more, satu- rated with dissolved oxygen (DO). The inflow of new, freshly- oxygenated seawater balances, or exceeds, the demand of DO required by aerobic (oxic) microbial decay of organic matter. However, if inflow is restricted, such as in a stagnant pool, or if the input of labile organic matter, say, from man-made inputs, exceeds the assimilative capacity of the DO to meet this demand, then DO declines rapidly. When DO levels fall to about 60 to 20 micromol DO/kg (1.5 to 0.5 ml DO/1), a hypoxic system, the chemistry changes dramatically. Nitrate converts to nitrite and some denitri- fication occurs, releasing nitrogen and nitrous oxide gas, while previously insoluble metal oxides begin to become solubilised. When DO becomes undetectable, an anoxic (anaerobic) system, there is no DO for animals to breathe and sulphate ion. relatively abundant in seawater, is microbially (anaerobically) converted to highly toxic sulphide chemical species. But what other chemical re- actions occur? Precipitated metal oxides re-dissolve, rapidly releasing their adsorbed load of highly toxic trace elements, including mercury, lead and cadmium, while adsorbed organics are also released. Par- ticulate sulphur compounds are produced and may be ingested, while other co-existing reduced organic and inorganic compounds may be extremely toxic to all shellfish and other living matter. Anoxic waters destroy shellfish beds since, unlike finfish, or mobile crustaceans such as lobsters, the sessile animals cannot take avoidance action. In this paper the onset of anoxia will be traced and the chemical mechanisms which occur will be discussed, with some comment on precautions to minimise shellfish mortalities from anoxia. ENIGMAS IN THE REPRODUCTIVE BIOLOGY OF PA- CIFIC OYSTERS IN IRELAND. G. O'Sullivan and M. F. Mulcahy, Department of Zoology & Animal Ecology, National University of Ireland, Lee Maltings, Prospect Row, Cork, Ireland. The study aims to assess the number of segments of an oyster which should be examined histologically to obtain an accurate ICSR. Cork. Ireland Abstracts, Sept. 29-Oct. 2. 1 399 725 conclusion on the total gonadal variation between male, female and hermaphrodite. Comparison of results from previous studies raised the possibility of variation in gonad from one part to an- other. (Steele 1998, Sato L998). The reason for the failure of oysters to spawn in Cork harbour when comparable oysters at comparable temperatures and phytoplankton levels in Dungarvan did spawn. (Steele 1998). is being examined. Oysters in Cork harbour became ripe but reabsorbed the gonad, without spawning. (Steele 1998). The spawning capability and larval survival of ripe oysters brought in from Dungarvan and from Cork are being assessed. The reproductive cycle and gonad development are being stud- ied over a 12 month period and examined histologically before being staged, using a modified version of the method devised by Mann (1979). Temperatures at both sites are being monitored daily, chlorophyll a levels are recorded on a regular basis. Mean condition index and shell index are calculated for each sample, as TBT is a suspected factor in Cork harbour, (Minchin et al. 1996). A NATIONAL HARMFUL ALGALBLOOM DATA MAN- AGEMENT SYSTEM. M. C. Paraso and M. D. Ford, Coastal Ocean Laboratory/NODC/NOAA, 1315 East-West Highway, Sil- ver Spring. MD. 20910. USA. The National Oceanographic Data Center (NODC) is develop- ing a system which will provide access to physical, chemical, and biological information acquired from various sources to assist in harmful algal bloom (HAB) management and research. Initially, a prototype system will be developed for the Gulf of Maine and the Gulf of Mexico. Later, the system will expand to all US coastal areas affected by HABs. Sources of data include routine monitor- ing efforts, event driven monitoring, topical research initiatives (ECOHAB), and the NODC archive. Routine monitoring data and biological data will be held in a new database. A common interface will link this database to other databases that hold HAB-related data such as NODC s Ocean Profile Database and Ocean Current Time Series Database. A large data management system like this will allow researchers and re- source managers to access pertinent data for a particular region for decision making, or to compile a historical perspective on events in a particular region. The combination of physical, chemical, and biological data from many sources in this system will provide researchers with a tool to assist in developing forecast models for harmful algal blooms. USE OF VIDEOGRAPHY TO ASSESS DIFFERENCES BE- TWEEN RESTORED AND NON-RESTORED AREAS IN CHESAPEAKE BAY. K. T. Paynter and T. E. Koles, Chesa- peake Biological Laboratory and Department of Biology. Univer- sity of Maryland, College Park. MD 20742 USA. Restoration of shellfish beds has at least two obvious applica- tions: commercial and ecological. While the commercial value of restored areas to the shellfish industry may be obvious, the eco- logical value of restored shellfish beds has been difficult to quan- tify. Oyster bar restoration in Chesapeake Bay has only recently begun in earnest. In 1995 and 1996 10 acres of oyster "bottom" was restored with 1 million hatchery produced oyster spat per acre. In 1997 and 1998. additional areas were restored in the Choptank. Patuxent and Chester Rivers. As the laboratory charged with moni- toring these efforts, we have experimented with various techniques of assessing oyster bar restoration. Initial sampling was conducted by scuba divers using quadrats. Surface shell was collected from within the quadrats and returned to the laboratory. Live oysters were counted, measured and tested for condition, gonadal maturity and parasite presence and abundance. Organisms associated with the shell collected including anemones, curved mussels and bar- nacles were also quantified. Although these samples provided an accurate measure of oyster abundance and health, they were inad- equate for ecosystem-level observations. Recently we have employed underwater videogTaphy to assess differences between restored and non-restored areas. This technique appears to hold great promise as a method for learning more about what restored oyster bars may produce at the ecosystem-level. Quad- rat sampling rarely represented motile species such as blennies. go- bies and crabs while videography not only showed those species were present but also showed how those organisms were related to the physical structure of the reef. We believe the use of quantitative videography will be an important and useful tool in demonstrating the ecological value of oyster and other shellfish reefs. THE APPLICATION OF RANDOMLY AMPLIFIED POLY- MORPHIC DNA (RAPD) MARKERS TO STOCK DIS- CRIMINATION OF THE WHITE SEA URCHIN, TRIP- NEUSTES VENTRICOSUS, IN THE EASTERN CARIBBE- AN. M. Pena,1 C. Gomes,2 and W. Hunte,2 'Marine Resource and Environmental Management Programme (MAREMP), Uni- versity of the West Indies, PO Box 64, Cave Hill. Barbados: ^Natural Resource Management Programme (NRM). University of the West Indies, PO Box 64. Cave Hill. Barbados. The white sea urchin, Tripneustes ventricosus, supports artisa- nal fisheries in several countries in the eastern Caribbean, but the implementation of sound management practices for the fisheries requires an understanding of population stock structure in the re- gion. Genetic differences among populations of urchins from An- guilla. St. Lucia. Barbados, Carriacou and Grenada were investi- gated in this study using randomly amplified polymorphic DNA (RAPD) analysis. Five arbitrarily designed primers detected 52 polymorphic loci in 182 individual sea urchins. Analyses of variance detected a high and significant level of population differentiation among all popu- lations, both when populations were simultaneously analysed (-ST = 0.872; P < 0.001), and when they were analysed in population pairs (P < 0.001 in all cases). Estimates of gene flow (Nm) ranged 726 Abstucrs. Sept. 29-Oct. 1999 ICSR, Cork. Ireland from 0.009 to 0.078. indicating limited gene flow between popu- lations. Cluster analyses of similarity and percent match indices revealed subdivision of the populations into five distinct genetic groups, which aligned perfectly with geographical location. Gene flow was significantly correlated with geographical distance (P < 0.05). but distance explained only 38.25% of the variation in gene flow. The implication of these results is that there are five unique genetic stocks of T. ventricosus in the eastern Caribbean. This is surprising, given the small geographical distances separating the stocks and the presence of a planktonic larval phase in T. ventri- cosus. We are currently attempting to verify these results using EPIC primers of the nuclear allelic system. A SATELLITE AND FIELD PORTRAIT OF A GYRO- DIN WM AUREOLUM BLOOM OFF SOUTH-WESTERN IRELAND; AUGUST 1998. R. Raine,1 S. O'Boyle,1 T. O'Higgins,1 M. White,1 J. Patching,1 T. McMahon,2 and B. Cahill,3 'Martin Ryan Institute. National University of Ireland. Galway; "Marine Institute. Fisheries Research Centre. Dublin; 'Marine Institute. Marine Data Centre. Dublin. An extensive surface bloom of the dinoflagellate Gyrodinium aureolum (Hulberti ( = Gymnodinium mildmotoi) occurred off southwestern Ireland during August 1998. The bloom was evident both from remotely sensed satellite ocean colour data, and as vis- ibly discoloured water. The bloom extended from the mouth of Bantry Bay around towards Cork, extending some 40 nautical miles offshore. The timing of the bloom co-incided with a field survey in the area. This paper compares the surface distributions of chlorophyll and G. aureolum concentrations with satellite ocean colour and thermal infra-red SST pictures, from which may be derived the origins of the bloom. It would appear that weak coastal upwelling transported a thermocline population of G. aureolum up to the surface in the region of the Fastnet Rock, from where it was wind-dispersed eastwards across the northern Celtic Sea. TRIALS ON CLAM {CHAMELEA GALLINA) BEDS OF AN INNOVATIVE HYDRAULIC DREDGE WITH VIBRATING AND SORTING BOTTOM. E. Rambaldi, G. Priore, G. Prioli. N. Mietti, T. Pagliani,1 and M. L. Bianchini,2 'Consorzio Medi- terraneo s.c.r.L. Via Nazionale 243, 00184 Roma, Italy: 2P. F. Raisa/CNR. Consiglio Nazionale delle Ricerche. Via Tiburtina 770. 00159 Roma, Italy To solve some of the productive and environmental problems related to the use of the hydraulic dredge in bivalve mollusc fish- ing, art experimental gear with vibrating bottom grid and other technical changes has been tested on clam (Chamelea gal Una) beds. Comparative fishing surveys have pointed out a significantly different selectivity of the vibrating dredge, with respect to a stan- dard gear: in fact, undersize clams are sieved out during the fishing process, and almost no juveniles were caught. Speaking of the product quality, laboratory analyses show that the internal sedi- ment is significantly lower in the catch from the modified dredge, thanks to a sort of "alarming" device. Nevertheless, the number of damaged clams suggests a greater mechanical stress of the vibrat- ing grid. As for the environmental effects, the vibrating bottom is selective for the associated fauna too, as it is shown by the mean weight of all the by catch species, which is higher in the experi- mental gear. Moreover, the riddling goes on continuously, allow- ing the immediate release of the sorted out organisms, which are repositioned in the area origin, thus avoiding a "contagious" dis- tribution. In conclusion, these preliminary indications suggest a positive evaluation of the modified dredge, especially when con- sidering its innovative design, still with wide margins for improve- ment. EFFECT OF INCREASED PRIMARY PRODUCTION IN A FJORD ON GROWTH OF BLUEMUSSELS AND SCAL- LOPS. K. I. Reitan, G. Oie, Y. Olsen, and H. Reinerstsen, SINTEF Fisheries and aquaculture. Group of Bio-resources. Trondheim. Norway. The growth of scallops (Pecten maximus) and blue mussels (Mytilus edulis is dependent on both the availability and the quality of the food, and physical factors as temperature and water current in the fjord. The primary production in a fjord normally fluctuates during the season. A controlled addition of nitrogen, phosphorous and silicon (N:Si:P = 16:8:1. where P = 0.4 mg m"3 day"1) to a closed small fjord (Hopavaagen in Norway) increased the primary production with 50% compared to control locations. However, this addition of nutrients gave no increase in biomass of phytoplankton in the fjord. The increased primary production in Hopavagen resulted in higher growth of scallops compared to the control. This increased growth of scallops was probably due to a higher growth rate of phytoplankton in the fjord when extra nutrients were added. The growth pattern of blue mussels differed from that of scal- lops. No significant difference in shell height was obtained for blue mussels with and without extra addition of nutrients to the water. NITROGEN EXCRETION BY THE PACIFIC OYSTER. CRASSOSTREA GIGAS, AS A CONTRIBUTOR TO ESTUA- RINE NUTRIENT CYCLING IN TOMALES BAY, CALI- FORNIA. L. Righetti, San Francisco State University. Ca: Rom- berg Tiburon Center. CA. Because of its importance as an aquaculture commodity, the filtering capacity and dietary requirements of the Pacific Oyster, Crassostrea gigas, have been studied in some detail. Most inqui- ries have focused on the question of nutrient up-take by the animal, and the portion of their intake that is converted to meat pro- duction. Excreted organic matter may be returned cycling sys- tems, in such forms as NH4. or removed from these systems. As 1CSR. Cork, Ireland Abstracts, Sept. 29-Ocl. 2. 1999 727 in estuarine systems primary production is influenced by biotic and abiotic factors. The activity of filter-feeders can alter the compo- sition of available nutrients in the water column, thus influencing phytoplankton food webs. This study examines the organic matter excreted by C. gigas, and considers the resultant impact of oyster aquaculture on the estuarine food web of Tomales Bay, California. MORTALITY AND DISPERSAL IN A BENTHIC SUB- TIDAL DECAPOD COMMUNITY AND OF HATCHERY REARED LOBSTER HOMARUS GAMMARUS. M. Robinson and O. Tully, Zoology Department. Trinity College Dublin. Dub- lin 2. IRELAND. Depletion of decapod stocks and increasing market prices has lead to a number of stock enhancement programs in Northern Europe. One such method utilised for the enhancement of European lobster Homanis gammanis involves the release of hatchery-reared juvenile lobster onto the seabed. Due to the scarcity of data pertaining to the habitat requirements of wild juvenile H. gammanis, stage V reared lobsters are commonly released into habitats similar in characteristics to those occupied by adults. The fate of hatchery-reared lobster, and their affect on the decapod communities resident in the release areas is unclear. Certain release methodologist result in localised, short- term high densities of juvenile lobster. Juvenile lobsters were released at high density into enclosed and unconfined experimental plots containing existing wild fauna within a commonly utilised release ground. Twelve percent of the initial seeded lobsters were recovered from enclosures after one month, compared to a one percent recovery from unconfined plots. Overall density. 4.8 individuals- 1 m2, may represent a rough ap- proximation to the saturation density of juvenile H. gammanis. although wild densities are unlikely to reach this level. Enclosure and/or the presence of juveniles reduced the abundance of young of the year porcellanid crab Pisidia longicornis, but did not affect any other species, or community structure as a whole. Dispersal can play an important role in regulating density- dependant processes acting on mobile benthic decapod Crustacea. The influence of dispersal may become less important for species below saturation density resident in physically complex habitats with a plentiful food supply. Movement in densely populated ar- eas, subject to losses to demersal and benthic predators, may result in increased mortality. The numbers of released lobsters that sur- vive to recruit to the fishery, and the resultant financial viability of this enhancement method, remains unclear. THE IMPACT OF SCALLOP DRAGS ON SEA URCHIN POPULATIONS AND BENTHOS IN THE BAY OF FUNDY, NEW BRUNSWICK, CANADA. S. M. C. Robinson, S. Bernier, and A. Maclntyre, Biological Station. Dept, Fisheries and Oceans. St. Andrews. New Brunswick, EOG-2XO, Canada. The fishery for the green sea urchin {Strongylocentrotus droe- bachiensis) began in New Brunswick in the late 1980's in response to increasing demand from Asian markets. Fishing was initiated by members of the scallop fishery and the harvesting practices were based on familiar gear and similar operations in Maine. Because of the potential harvesting impact on the shallow water benthic popu- lations a study was initiated in 1993 to document 1 ) the proportion of sea urchins damaged during the harvesting operation, 2) the impact n and subsequent recovery time of the associated benthic flora and fauna and 3) the impacts on the bottom substrate. Two representative sites were chosen for the study. At each site, a towing lane was created parallel to shore and divided into a treat- ment section and control section. Divers used a fixed line transect method to survey the control and treatment plots prior to and after the harvesting operation (2 m Digby drag). Further surveys were scheduled three and six months later. The results from the study showed a significant increase in the number of broken sea urchin tests after the harvesting operation from 0.05 m2" to 1 .4 m2~ at the Passamaquoddy Bay site. Similar trends were found in Grand Manan. On both sites in the experimental plot, there was an in- crease in the density of mobile predators such as hermit crabs, starfish, whelks and sculpins. While the lobster density declined to zero in the experimental plot, the lack of body parts suggested they moved out of the area. The dragging operation also adversely affected a number of macrophytes. THE U.S. HAB PROGRAM: ONE-HALF OF A U.S.- IRELAND HAB EXCHANGE. K. G. Sellner, Silver Spring, MD. USA, 20910. ECOHAB (Ecology and Oceanography of Harmful Algal Blooms) is the U.S. national program that provides support for research on blooms that impact living resources, coastal econo- mies, and public health along the U.S. coast. Five agencies, NOAA, NSF, EPA, ONR, and NASA, provide funding for re- search to 1 ) determine the linkages between physics, water quality, and HAB biology, ecology, physiology, behavior, and toxicity in order to develop regional forecasting capabilities for expression of HABs in U.S. waters, 2) expand general knowledge on HAB spe- cies, and 3) investigate and determine species- and region-specific prevention, control, and mitigation strategies for reducing coastal impacts. Results from the ECOHAB regional studies should be considered as resources for Irish goals of developing an Irish- specific program, where bathymetry, circulation, and HAB taxa characteristics of Irish coastal areas and species might fit with similar sets of conditions in the U.S.. permitting transfer of U.S. results (like models) to Ireland as a foundation to build upon. Irish results from mariculture and its own species would, in turn, expand U.S. knowledge on N. Atlantic taxa and impacts that conceivably might at some time develop in U.S. waters. Additionally, the U.S. ECOHAB Program is part of a larger U.S. effort, that includes other areas of likely interest to Irish HAB management commu- nities, including Federally-supported HAB monitoring and assess- ment programs, event response activities, education/outreach 728 Abstracts, Sept. 29-Oct. 2, 1999 ICSR. Cork. Ireland programs, and public health-related epidemiological and cohort studies, yielding a fairly comprehensive national HAB program. Cross-Atlantic exchange of HAB information, results, and man- agement approaches is an extremely important area for future U.S. -Irish HAB communities. HARMFUL ALGAL BLOOMS AND SHELLFISH RESTO- RATION: PERMANENT OBSTRUCTIONS OR TEMPO- RARY NUISANCES? S. E. Shumway, Natural Science Divi- sion, Southampton College of Long Island University, Southamp- ton, NY 11968 USA. It is now generally accepted that the number and frequency of harmful algal bloom (HABS) is increasing and they are increas- ingly blamed (sometimes wrongly) for the destruction or demise of shellfish beds and aquaculture operations, or pointed to as a source of major concern with respect to siting of aquaculture and resto- ration ventures. These HABS occur worldwide and. in some areas, they are a common and seasonal occurrence. In other areas they are sporadic or unique. These blooms have far-reaching impacts on ecosystem integrity, species interactions, aquatic animal health, population growth, human health, economy, industry, and ecology. For obvious reasons, algal species known to be associated with human health risks have received the most attention and commer- cially important fish and filter-feeding shellfish have been the primary species of concern. Algal species which cause human illnesses are not the only species of importance with regard to animal or environmental health or economics, and commercially important fish and shellfish are not the only organisms impacted. Current discussions frequently focus on mitigation and control of adverse effects of these harmful algae. This presentation will review our knowledge of harmful algal- shellfish (molluscan and crustacean) interactions worldwide and discuss ways in which shellfish restoration efforts may be under- taken successfully in the face of these imposing threats. The dis- cussion will stress science-management interaction and ways in which restoration efforts may be undertaken without imposing undue environmental stress and expense. DINOFLAGELLATE RESTING CYSTS IN CORK HAR- BOUR: IMPLICATIONS FOR SHELLFISH AQUACUL- TURE. J. Silke and T. McMahon, Marine Institute. Fisheries Research Centre, Abbotstown, Dublin 15. Ireland. In Ireland a monitoring programme for the detection of algal toxins in shellfish is carried out by the Marine Institute's Fisheries Research Centre. The monitoring programme is carried out under EU Directive 91/492. The North Channel area of Cork Harbour has, to date, been the only location in Ireland where toxins causing Paralytic Shellfish Poisoning (PSP) have been detected in shellfish above the regulatory limit. During the summer of 1996, 1997, and 1998 mussels (Mytilus edulis) from the North Channel area were found to contain PSP toxins above the regulatory limit for a short period and a ban on harvesting was imposed. Oysters (Crassostrea gigas) remained below the regulatory threshold. The dinoflagellate Alexandrium tamarense, a known vector of PSP toxins, was observed in the area during each of the toxic events. The exact origin of the populations of A. tamarense was unknown. SHELLFISH CARRYING CAPACITY AND ECOSYSTEM PROCESSES. A. Smaal and M. van Stralen, Netherlands Insti- tute for Fisheries Research RIVO-DLO. Centre for Shellfish Re- search. PO Box 77, 4400 AB Yerseke, The Netherlands. Shellfish culture in the Netherlands consists of bottom culture of mussels {Mytilus edulis) and oysters (Crassostrea gigas. Ostrea edulis) and fisheries and experimental relaying of wild cockles (Cerastoderma edule) Annual surveys show high densities of shellfish and conse- quently they are the most abundant functional group in the culti- vation areas Wadden Sea and Oosterschelde estuary. Annual yields of shellfish culture and fisheries are registrated by the fishery board: as all mussels are delivered to the auction, detailed long- term data are available of condition (% flesh) and annual yield from the various production areas. The "exploitation" carrying capacity of the Oosterschelde eco- system— defined as the standing stock of the exploited species at which the yield of the marketable cohort is maximised — was evaluated before and after completion in 1987 of a large scale coastal engineering project consisting of the construction of a storm-surge barrier. This project resulted in decreased current ve- locities, increased water residence time, decreased nutrient loads and increased water transparency. The phytoplankton population showed a resilient response by maintaining primary production while species composition adopted to changed light and nutrient conditions. Phytoplankton turnover increased significantly. It was demonstrated that the average annual condition of mus- sels delivered to the market showed a significant negative corre- lation with the annual shellfish standing stock in the pre-barrier phase ( 1980-1984) and in the first period of the post-barrier phase (1987-1990). There was also a significant correlation between mussel growth and the annual primary production in both periods. It was therefore concluded that the carrying capacity of the eco- system was fully exploited and that the primary production deter- mined the carrying capacity. In third period (1991-1997), how- ever, the correlation appeared to have vanished. Meanwhile, shellfish culture has adapted to the new conditions in the Oosterschelde estuary. Mussel lease sites were relocated in response to changed hydrodynamic conditions, cultivation tech- niques have evolved and lease sites are now used in a more ex- tensive way. Although the standing stock has maintained, the an- ICSR. Cork. Ireland Abstracts, Sept. 29-Oct. 2. 1999 729 nual yield has increased, residence time of mussels on the lease sites has decreased, hence the turnover has increased. Yet. the condition of market delivered mussels has not changed. Appar- ently, the mussel farmers made the choice to increase yield rather than quality. Total standing stock of shellfish has furthermore shown a decrease of cockle densities due to a lack of spatfall. and a dramatic increase of Pacific oysters. The latter has spread from lease sites to virtually everywhere in the estuary, and is now con- sidered a threat to other shellfish. The interaction between shellfish and the ecosystem can be de- scribed in terms of primary production, grazing pressure and water renewal. For the Oosterschelde it will be shown that phytoplankton turnover and shellfish filtration have increased, while water renewal has decreased. The interactions have intensified, and the conse- quences for carrying capacity estimation will be discussed. OPENING REMARKS FOR GEAR IMPACT AND REME- DIATION SESSION. G. W. Thayer, Beaufort Laboratory. Beau- fort, North Carolina, USA. The alteration of benthic habitat by fishing activities is not well understood, yet it is generally acknowledged that some fishing gears may influence species composition and diversity and alter habitat complexity. It is also considered that the potential effect of fishing activities on benthic habitats is a key issue facing the long-term sustainability of our coastal and marine living resources. In the United States the eight Fishery Management Councils (FMCs). which are responsible for managing fisheries, have been required to develop a comprehensive habitat plan for their respec- tive management areas. These plans are to identify Essential Fish Habitat (defined as "those waters and substrate necessary to fish for spawning, breeding, feeding, or growth to maturity"' where "waters" includes aquatic areas and their associated physical, chemical, biological properties that are utilized by fish). As part of the Essential Fish Habitat amendments to the Magnuson-Stevens Fishery Conservation and Management Act (December 1996) all FMCs were to identify adverse impacts of fishing activities on Essential Fish Habitat, and protect Essential Fish Habitat from fishing and gear impacts. As one might expect, this has been highly controversial. It is evident that the effects of fishing in estuaries and coastal waters can be diverse and far-reaching, but effects are obscured by uncertainty for a number of reasons because systems are dynamic, many systems already have a long history of harvesting, and im- pacts may not be seen due to inappropriate scales of observation (Blaber et al. Draft MS. "Effects of fishing on the structure and functioning of estuarine and near shore ecosystems." Special issue of ICES Journal of Marine Science). It also has proven difficult to separate the effects of fishing from other sources of variability. Currently, several environmentally-related agencies in the United States are developing joint funding initiatives directed at the ef- fects of fishing activities on benthic habitats. These initiatives are addressing the relations between the biological and physical ef- fects of gear to the geological characteristics of benthic habitats. Workshops have and are being convened not only to address gaps in our knowledge but also to generate research priorities and plans. This is a very hot topic. The workshop convened here at the International Conference on Shellfish Restoration has brought together a series of presen- tations and videos dealing with ( 1 ) impact of fishing gear in shal- low estuarine and nearshore waters, approaches to management and restoration of these areas and (2) with management consider- ations and impacts of gear to deeper water benthos and benthic habitats. Our (co-chairs Gordon Thayer. US Department of Com- merce, NOAA and Michael Kaiser. School of Ocean Sciences, Menai Bridge. UK) goal has been to bring together experts in the arena of gear impact to benthic shellfish habitats to provide an improved basis for environmental management decisions regard- ing not only gear types and potential impacts to benthic habitats but also restoration of habitats impacted by fishing gear. RESTORATION OF LOBSTER (HOMARUS GAMMARUS) POPULATION EGG PRODUCTION IN DEPLETED STOCKS. O. Tully, Zoology Department. Trinity College Dub- lin. Dublin 2. Ireland. Stocks of lobster (H. gammarus) in Ireland have declined sig- nificantly over the past 30 years as indicated by trends in CPUE. The current stock biomass may therefore be below the environ- mental carrying capacity as larger stocks previously existed. The decline has presumably occurred because of recruitment overfish- ing or because of environmental factors that reduced larval pro- duction and supply to the seabed or survival of juvenile lobsters. Restoration of population egg production is being achieved in Irish lobster fisheries by conservation of marked females (using a tail v-notch) which have lifetime legal protection from fishing. In one fishery this program has been successfully implemented over a 5 year period. Population egg production has been increased by 23 million eggs per annum and possibly by 25%. Based on esti- mates of growth, mortality and spawning frequency this contribu- tion will increase in the future without the addition of new marked females. Predicting the effects on recruitment is hampered by the lack of information on temperature and hydrodynamic effects on larval supply to the seabed and post settlement mortality rates (essentially the stock recruit relationship of a closed population). Fishery log book data is being used to monitor subsequent effects on recruitment into the fishery. This requires a long time series of data. CPUE from 1995 fi 1997 was stable. In 1998 a 15% increase in the catch rate of undersized lobsters was recorded compared with the previous year. This may have been due to favourable conditions for larval survival 3 years previously in addition to the added egg production potential in the population. The efficacy of this method relative to other technical measures is currently being assessed for all Irish lobster fisheries. 730 Abstracts, Sept. 29-Oct. 2, 1999 ICSR, Cork, Ireland CLAM FARMING AND TOURISM— A DIFFICULT COM- BINATION? THE SOCIO-ECONOMIC ROLE OF RUDI- TAPES DECUSSATUS CULTIVATION IN THE ALGARVE. J. C. Wallace, Centre of Marine Sciences (CCMAR). University of Algarve, Campus de Gambelas, 8000 Faro, Portugal. The oldest and most traditional, marine aquaculture in Portugal is clam farming. About 90% of the national cultivated bivalve production is produced in the Algarve, on the coastal wetlands known as the Ria Formosa. More than 1 00 hectares of these tidal wetlands are used for the production of 'ameijoa boa', Ruditapes decussatus ( = Tapes decussatus = Venerupis decussata), [eng. carpet-shell clam; fr. palourde; ger. teppichmuschel; esp. almeja fina]. The production involves, essentially, the tending of family plots, of various sizes, which have been sown with clam 'seed' collected from the natural population of the Ria. The Ria Formose, extending for several kilometers along the Algarvian coast, is also the recipient of sewage from several towns, which, during the tourist season, receive tourists in numbers far exceeding those of the resident populations. The peak tourist season also coincides with the period of maximum sea temperatures and. due to evapo- ration, highest salinities. These factors combine to produce a de- terioration in water quality in the Ria. It is, however unclear whether or not the clam production is affected directly, or simply stressed sufficiently to become weakened and more susceptible to parasitic diseases. The paper considers the recent history of this aquaculture, as well as the present situation and possible future developments. DEVELOPMENT OF HYDROLOGIC MODIFICATION IN- DICATORS TO SUPPORT WATERSHED-BASED RESTO- RATION OF SHELLFISH RESOURCES IMPACTED BY FECAL COLIFORM CONTAMINATION. N. W. White, L. E. Danielson, and M. V. Holmes, Brooks Hall. School of De- sign, Box 7701. NC State University, Raleigh. NC 27695-7701. USA. The Jump Run Creek Project in North Carolina is a multi- disciplinary, multi-agency, watershed-based restoration project that combines in-column water quality data, stormwater quality data, storm and stream flow measures, with spatial analysis, and community involvement to investigate causes and solutions of bac- terial loading contributing to shellfish closure management. Pre- liminary project efforts were reported at ISCR '98, Hilton Head showing data indicating that increased volume and velocity of stormwater flows may be the primary transport vector. However, the impervious surface area in this watershed is less than 5% — well below the threshold of 12%— 25% cited in the literature as the level at which nutrient loading causes water quality degradation. Very little investigation has been conducted to review the effects of land use modification on bacterial loading. This effort reports on one approach to measure impacts to watershed hydrology indicat- ing conditions contributing to excessive bacterial loading and bed closure. Using scanned images of aerial photos of the watershed from the 1960s through the 1990s, overlain with current parcel data, and linked to GIS classification ranks are assigned to each parcel indicating the degree of hydrologic modification. Param- eters being used are vegetative cover, ditching, and impervious surface area. Using regression techniques, spatial and temporal models of hydro-modification will be regressed against thirty years of bacterial data collected by North Carolina Division of Environ- mental Health — Shellfish Sanitation Branch to investigate rela- tionships. The aerial and GIS have been completed. The hydro- modification database is currently under construction. Regression analysis will be conducted during the summer months; so that a report on the results can be made to the conference in September. Over the coming year, this information will be used in two ways: 1 ) to help watershed citizens understand land use activities that contribute to bacterial loading; and 2) in conjunction with water- shed citizens, to design and locate BMPs for mitigation and res- toration. THE ACQUISITION AND INTERPRETATION OF DIGI- TAL ACOUSTICS FOR CHARACTERIZING LOUISI- ANA'S SHALLOW WATER OYSTER HABITAT. C. A. Wil- son, H. H. Roberts, J. Supan, and W. Winans, Oyster geophysics Program, LSU. Baton Rouge, Louisiana 70803, USA. Louisiana's oyster industry generally ranks first nationally in production, averaging 10-12 million pounds of shucked meat, with an average total economic value of approximately $56 million annually. However, coastal Louisiana, like many deltaic land- masses, faces continued landscape alteration from natural pro- cesses and anthropogenic impacts that are and will affect oyster production. Many steps are being taken at both State and Federal levels to slow/mitigate these changes. Most promising of these strategies is river diversions, which introduce freshwater and sedi- ment to river-flanking environments (lakes, bays and associated marshlands). Two such diversion projects planned by Louisiana Department of Wildlife and fisheries and US Army Corps of En- gineers. Caernarvon and Davis Pond, are designed to nourish marshes with water and sediment as well as to help establish ideal isohalines over historic oyster grounds. Critical to the success of these programs is a rapid and accurate means to qualify and quan- tify changes in oyster habitat. Digital high resolution acoustic instrumentation linked to state- of-the-art acquisition and processing software is available for building a baseline of information that can be used for evaluating future changes in shallow water bottoms with special application to oyster habitat. Application of digital side-scan sonar (100 and 500 kHz) and a broad-spectrum sub-bottom profiler (4-24 kHz) for rapidly acquiring both surficial and shallow subsurface data (average water depths 0.7 to 3.0 m) has now been accomplished. ICSR, Cork. Ireland Abstracts, Sept. 29-Oct. 2, 1999 731 These data sets "calibrated" with coring, surface sampling, and other "ground truthing" techniques have enormous potential tor understanding (a) distributions of bottom sediment types, (b) lo- cations of oyster reefs and distributions of scattered oyster clumps and shells, (c) fisheries habitats, (d) areas of active sedimentation and erosion, and (e) shallow subsurface configurations that influ- ence surface conditions. This approach of linking high-resolution acoustic data with various direct sampling techniques has been verified in pilot studies. Geo-referenced side scan sonar mosaics of oyster lease areas sur- veyed were incorporation into a GIS database. Using image- processing techniques to analyze mosaic reflectance patterns, we es- timated the percent and total acreage of several bottom types. Results were calibrated with field collected ground truth measurements. CO-ORDINATING SHELLFISH AND FINFISH AQUACUL- TURE SYSTEMS. D. Jackson and T. O'Carroll, As CLAMS (Co-ordinated Local Aquaculture Management Systems) is introduced to the various bays around the country a process will be undertaken to open dialogue with and between local aquaculture operators and to facilitate the development of an appropriate plan. The steps in this process are outlined and the key elements necessary to facilitate the preparation of an effective CLAMS document are identified. In a workshop setting the practicalities of establishing an ef- fective local CLAMS group are explored under the following headings: • Introduction to concept of CLAMS • The Paperwork • Information and Integration • Risk Assessment • Feedback to Regulatory agencies • Development Plan 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 care- fully prepared in the style followed in prior issues of the Journal of Shellfish Research (1991) before submission 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 or in italics 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 Publi- cations 16 and 17: Common and Scientific Names of Aquatic Invertebrates from the United States and Canada: Mollusks and CSNAIUSC: Decapod Crustaceans, or relevant publi- cations for other geographic regions. Abbreviations, Style, Numbers: Authors should follow the style recommended by the sixth edition (1994) of the Council of Biology Editors [CBE] Style Manual, distributed by the American Institute of Biological Sciences. All linear measurements, weights, and volumes should be given in metric units. r 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 or laser print and planned so that important details will be clear after reduction 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 prepared 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 repro- duced without reduction. Originals of graphic materials (i.e.. line drawings) are preferred and will be returned to the author. Each illustration 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 alpha- betically at the end of the paper. Abbreviations in this sec- tion should be those recommended in the American Stan- dard 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 a recent issue of the Jour- nal 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. Students (only if first author and a member of NSA) will not be assessed page charges. 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: Appropriate photographs may be submitted for consideration for use on the cover of the Jour- nal of Shellfish Research. Black and white photographs and color illustrations will 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@southampton.liunet.edu Membership information may be obtained from the Edi- tor or the Treasurer using the form in the Journal. Institu- tional subscribers should send requests to: Journal of Shell- fish Research, P.O. Box 465, Hanover, PA 17331. MBL WHOI LIBRARY IdH 1AAL 5 Islay D. Marsden Reproductive cycles of the surf beach clam Paphies donacina (Spengler. 1793) from New Zealand 539 Michael A. Rice Uptake of dissolved free amino acids by northern quahogs. Mercenaria mercenaria and its relative importance to organic nitrogen deposition in Narragansett Bay. Rhode Island 547 Nancy H. Hadley, Robert B. Baldwin, M. R. Devoe and R. Rhodes Performance of a tidal-powered upwelling nursery system for northern quahogs (hardclams) {Mercenaria mercenaria) in South Carolina 555 Paula A. Y. Maas, Stephen J. Kleinschuster, Michael J. Dykstra, Roxanna Smolowitz and Jason Parent Molecular characterization of QPX (Quahog Parasite Unknown), a pathogen of Mercenaria mercenaria 561 Melanie J. Leng and Nick J. G. Pearce Seasonal variation of trace element and isotopic composition in the shell of a coastal mollusk. Mactra isabelleana 569 H. Masski and J. Guillou The role of biotic interactions in juvenile mortality of the cockle {Cerastodenna edule L.): Field observations and experiment 575 Melbourne R. Carriker and Gregory L. Gruber Uniqueness of the gastropod accessory boring organ (ABO): Comparative biology, an update 579 Karl W. Mueller and Annette Hoffmann Effect of freshwater immersion on attachment of the Japanese oyster drill, Ceratosloma inomatum (Recluz 1851 ) 597 Kasim Cemal Giiven, Zeliha Yazici, Serap Akinci and Erdogan Okus Fatty acids and sterols of Rapana venosa (Valenciennes, 1846) 601 Sei-Ichi Okumura, Shoujiro Kinugawa, Aiko Fujimaki. Wataru Kawai, Hidetaka Maehata, Kazuhiro Yoshioka, Ryouko Yoneda and Kunio Yamamori Analysis of karyotype, chromosome banding, and nucleolus organizer region of Pacific abalone. Haliotis discus hamuli ( Archaeogastropoda: Haliotidae) 605 James O. Harris, Greg B. Maguire, Stephen J. Edwards and Stephen M. Hindrum Effect of pH on growth rate, oxygen consumption rate, and histopathology of gill and kidney tissue for juvenile greenlip abalone, Haliotis laevigata Donovan and blacklip abalone. Haliotis rubra leach 611 J. Jonasson, S. E. Stefansson, A. Gudnason and A. Steinarsson Genetic variation for survival and shell length of cultured red abalone {Haliotis rufescens) in Iceland 621 Jiann-Chu Chen and Won-Chung Lee Growth of Taiwan abalone Haliotis diversicolor supertexta fed on Gracilaria tenuistipitata and artificial diet in a multiple-tier basket system 627 Rodney D. Roberts, Tomohiko Kawamura, and Hideki Takami Morphological changes in the radula of abalone (Haliotis iris) during post-larval development 637 Gabriela Moura, Ricardo Guedes and Jorge Machado The extracellular mineral concretions in Anodonta cygnea (L. ): different types and manganese exposure-caused changes 645 R. Wouters, L. Gomez, P. Lavens and J. Calderon Feeding enriched Anemia biomass to Penaeus vannamei broodstock: Its effect on reproductive performance and larval quality 651 Gretchen A. Messick, Stephen J. Jordan, and William F. Van Heukelem Salinity and temperature effects on Hematodinium sp. in the Blue Crab Callinectes sapidus 657 Juan Valero, Tomas Luppi and Oscar Iribarne Size as indicator of swimming speed in crab megalopae 663 Jie Zheng and Gordon H. Kruse Evaluation of harvest strategies for tanner crab stocks that exhibit periodic recruitment 667 Steve L. Morton, Tod A. Leighfield, Bennie L. Haynes, Debra L. Petitpain, Mark A. Busman, Peter D. R. Moeller, Laurie Bean, Jay McGowan, John W. Hurst, Jr., and Frances M. van Dolah Evidence of diarrhetic shellfish poisoning along the coast of Maine 68 1 Abstracts of technical papers presented at the Sixth International Littorinid Symposium. July 24-31. 1999. Priory, Jamaica, W.I 687 Abstracts of technical papers presented at International Conference on Shellfish Restoration, September 29-October 2, 1999. Cork, Ireland. . 701 COVER PHOTO: New Zealand greenshell mussels, Pema canaliculus, courtesy of Rebecca Holland. National Institute of Water and Atmospheric Research Ltd., Auckland. New Zealand The Journal of Shellfish Research is indexed in the following: Science Citation Index®, Sci Search®, Research Alert*. Current Contents E/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. 18, No. 2 1999 CONTENTS John Kraeuter and Susan Ford Honored Life Member: Harold Haley Haskin 337 Susan Ford Honored Life Member: Carl James Sindermann 34 1 Susan Ford Honored Life Member: Aaron Rosenfield 343 A. G. Jeffs, R. C. Holland, S. H. Hooker and B. J. Hayden Overview and bibliography of research on the greenshell mussell, Perna canaliculus, from New Zealand waters 347 Gustavo Darrigran, Pablo Penchaszadeh and M. Cristina Damborenea The reproductive cycle of Limnoperna fortunei (Dunker, 1857) (Mytilidae) from a neotropical temperate locality 361 R. W. Penney and M. J. Hart Distribution, genetic structure, and morphometry of Mytilus edulis and M. trossulus w ithin a mixed species zone 367 Michele Pelc and Richard R. Alexander Salinity and sediment-mediated byssal thread production by Mytilus edulis Linnaeus and Geukensia demissa Dillwyn from New Jersey salt marshes 375 M. T. Sicard, A. N. Maeda-Martinez, P. Ormart, T. Reynoso-Granados and L. Carvalho Optimum temperature for growth in the catarina scallop (Argopecten ventricosus-circularis, Sowerby II. 1842) 385 Dan C. Marelli, William S. Arnold and Catherine Bray Levels of recruitment and adult abundance in a collapsed population of bay scallops (Argopecten irradians) in Florida 393 Stephen L. Estabrooks The telomeres of the bay scallop, Argopecten irradians (Lamarck) 401 I). Roddick, E. Kenchington, J. Grant and S. Smith Temporal variation in sea scallop (Placopecten magellanicus) adductor muscle RNA/DNA ratios in relation to gonosomatic cycles, off Digby, Nova Scotia 405 Noe A. Santamaria, Esteban F. Felix-Pico, Jose Luis Sdnchez-Lizaso, J. Ricardo Palomares-Garcia and Manuel Mazon-Sudstegui Temporal coincidence of the annual eelgrass Zostera marina and juvenile scallops Argopecten ventricosus (Sowerby II, 1842) in Bahi'a Concepcion. Mexico 415 Yantian T. Lu, Norman J. Blake and Joseph J. Torres Oxygen consumption and ammonia excretion of larvae and juveniles of the bay scallop. Argopecten irradians Concentricus (SAY) . . 419 Yantian T. Lu, Norman J. Blake and Joseph J. Torres Biochemical utilization during embryogenesis and metamorphosis in the bay scallop. Argopecten irradians Concentricus (SAY) 425 Quiyang Zhang, Gang Yu, Richard A. Cooper and Terrence R. Tiersch Chromosomal location by fluorescence in situ hybridization of the 28S ribosomal RNA gene of the eastern oyster 43 1 A. B. Strychar and B. A. MacDonald Impacts of suspended peat particles on feeding and absorption rates in cultured eastern oysters (Crassostrea virginica, Gmelin) 437 Mi Seon Park, Hyttn Jeong Lim, Qtae Jo, Jang Sang Yoo and Minjee Jeon Assessment of reproductive health in the wild seed oysters. Crassostrea gigas, from two locations in Korea 445 Kim J. Friedman and Paul C. Southgate Growout of blacklip pearl oysters. Pinctada margaritifera, on chaplets in suspended culture in Solomon Islands 45 1 Suifen Lyu and Standish A. Allen, Jr. Effect of sperm density on hybridization between Crassostrea virginica, Gmelin and C. gigas (Thunberg) 459 Gustavo W. Calvo, Mark W. Luckenbach, Standish A. Allen, Jr. and Eugene M. Burreson Comparative field study at Crassostrea gigas (Thunberg. 1793) and Crassostrea virginica (Gmelin. 1791) in relation to salinity in Virginia 465 Susan Ford, Eric Powell, John Klinck and Eileen Hofmann Modeling the MSX parasite in eastern oyster (Crassostrea virginica) populations. I. Model development, implementation, and verification 475 Michelle C. Paraso, Susan E. Ford, Eric N. Powell, Eileen E. Hofmann and John M. Klinck Modeling the MSX parasite in eastern oyster (Crassostrea virginica) populations. II. Salinity effects 501 Eric N. Powell, John M. Klinck, Susan E. Ford, Eileen E. Hofmann and Stephen J. Jordan Modeling the MSX parasite in eastern oyster (Crassostrea virginica) populations. III. Regional application and the problem of transmission 517 CONTENTS CONTINUED ON INSIDE BACK COVER