^°fcOA Fishery Bulletin SrATES & h r Vol. 84, No. 1 January 1986 THEILACKER, GAIL H. Starvation-induced mortality of young sea-caught jack mackerel, Trachurus symmetricus, determined with histological and morphological methods 1 RENAUD, MAURICE L. Hypoxia in Louisiana coastal waters during 1983: impli- cations for fisheries 19 LO, N. C. H., and T. D. SMITH. Incidental mortality of dolphins in the eastern tropical Pacific, 1959-72 27 MIDDLETON, ROBERT W., and JOHN A. MUSICK. The abundance and distribution of the family Macrouridae (Pisces: Gadiformes) in the Norfolk Canyon area 35 KEIRANS, WALTER J, SIDNEY S. HERMAN, and R. G. MALSBERGER. Differen- tiation of Prionotus carolinus and Prionotus evolans eggs in the Hereford Inlet estuary, southern New Jersey, using immunodiffusion 63 HUNT, JOHN H., WILLIAM C. LYONS, and FRANK S. KENNEDY, JR. Effects of exposure and confinement on spiny lobsters, Panulirus argus, used as attractants in the Florida fishery 69 BE ACHAM, TERRY D Type, quantity, and size of food of Pacific salmon (Oncorhyn- chus) in the Strait of Juan de Fuca, British Columbia 77 JONES, CYNTHIA. Determining age of larval fish with the otolith increment tech- nique 91 MOYLE, PETER B., ROBERT A. DANIELS, BRUCE HERBOLD, and DONALD M. BALTZ. Patterns in distribution and abundance of a noncoevolved assemblage of estuarine fishes in California 105 KRYGIER, E. E., and W G. PE ARCY The role of estuarine and offshore nursery areas for young English sole, Parophrys vetulus Girard, of Oregon 119 STEIMLE, FRANK W, PAUL D. BOEHM, VINCENT S. ZDANOWICZ, and RALPH A. BRUNO. Organic and trace metal levels in ocean quahog, A rctica islandica Linne, from the northwestern Atlantic 133 RALSTON, STEPHEN, REGINALD M. GOODING, and GERALD M. LUDWIG. An ecological survey and comparison of bottom fish resource assessments (submers- ible versus handline fishing) at Johnston Atoll 141 WILLASON, STEWART W, JOHN FAVUZZI, and JAMES L. COX. Patchiness and nutritional condition of zooplankton in the California Current 157 JOHNSON, P. T, R. A. MacINTOSH, and D A. SOMERTON. Rhizocephalan infec- tion in blue king crabs, Paralithodes platypus, from Olga Bay,-Kodiak Island, Alaska 177 (Continued on back cover) V Seattle, Washington U.S. DEPARTMENT OF COMMERCE Malcolm Baldrige, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION John V. Byrne, Administrator NATIONAL MARINE FISHERIES SERVICE William G. Gordon, Assistant Administrator Fishery Bulletin The Fishery Bulletin carries original research reports and technical notes on investigations in fishery science, engineering, and economic The Bulletin of the United States Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries in 1904 ar the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates were issued as documents through volume 46; the last docume was No 1103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared as a numbered bulleti A new system began in 1963 with volume 63 in which papers are bound together in a single issue of the bulletin instead of being issue individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued quarterly. In this fori it is available by subscription from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. It is al: available free in limited numbers to libraries, research institutions, State and Federal agencies, and in exchange for other scientif publications. SCIENTIFIC EDITOR, Fishery Bulletin Dr. William J. Richards Southeast Fisheries Center Miami Laboratory National Marine Fisheries Service, NOAA Miami, FL 33149-1099 Editorial Committee Dr. Bruce B. Collette National Marine Fisheries Service Dr. Edward D. Houde Chesapeake Biological Laboratory Dr. Merton C. Ingham National Marine Fisheries Service Dr. Reuben Lasker National Marine Fisheries Service Dr. Donald C. Malins National Marine Fisheries Service Dr. Jerome J. Pella National Marine Fisheries Service Dr. Jay C. Quast National Marine Fisheries Service Dr. Carl J. Sindermann National Marine Fisheries Service Mary S. Fukuyama, Managing Editor 'H-0656) is published quarterly by the Scientific Publications Office, National Marine Fisheries Service, nd Point Way NE, BIN C15700, Seattle, WA 98115. Second class postage is paid in Seattle, Wash., and additional offices. [MASTER send address changes for subscriptions to Superintendent of Documents, U.S. Government Printing Office, Washington, i402. Although the contents havr i ighted and may be reprinted entirely, reference to source is appreciated. - mined that the publication of this periodical is necessary in the transaction of the public business required by law of this Departmc I unds for printing of this periodica] has been approved by the Director of the Office of Manage- ment and Budget through 1 April 19J Fishery BulletirC^ CONTENTS L^22dlHole^Mass. Vol. 84, No. 1 January 1986 THE IL ACKER, GAIL H. Starvation-induced mortality of young sea-caught jack mackerel, Trachurus symmetricus, determined with histological and morphological methods 1 RENAUD, MAURICE L. Hypoxia in Louisiana coastal waters during 1983: impli- cations for fisheries 19 LO, N. C. H., and T. D. SMITH. Incidental mortality of dolphins in the eastern tropical Pacific, 1959-72 27 MIDDLETON, ROBERT W., and JOHN A. MUSICK. The abundance and distribution of the family Macrouridae (Pisces: Gadiformes) in the Norfolk Canyon area 35 KEIRANS, WALTER J., SIDNEY S. HERMAN, and R. G. MALSBERGER. Differen- tiation of Prionotus carolinus and Prionotus evolans eggs in the Hereford Inlet estuary, southern New Jersey, using immunodiffusion 63 HUNT, JOHN H., WILLIAM C. LYONS, and FRANK S. KENNEDY, JR. Effects of exposure and confinement on spiny lobsters, Panulirus argus, used as attractants in the Florida fishery 69 BEACH AM, TERRY D. Type, quantity, and size of food of Pacific salmon (Oncorhyn- chus) in the Strait of Juan de Fuca, British Columbia 77 JONES, CYNTHIA. Determining age of larval fish with the otolith increment tech- nique 91 MOYLE, PETER B., ROBERT A. DANIELS, BRUCE HERBOLD, and DONALD M. BALTZ. Patterns in distribution and abundance of a noncoevolved assemblage of estuarine fishes in California 105 KRYGIER, E. E., and W G. PE ARCY The role of estuarine and offshore nursery areas for young English sole, Parophrys vetulus Girard, of Oregon 119 STEIMLE, FRANK W, PAUL D BOEHM, VINCENT S. ZDANOWICZ, and RALPH A. BRUNO. Organic and trace metal levels in ocean quahog, Arctica islandica Linne, from the northwestern Atlantic 133 RALSTON, STEPHEN, REGINALD M. GOODING, and GERALD M. LUDWIG. An ecological survey and comparison of bottom fish resource assessments (submers- ible versus handline fishing) at Johnston Atoll 141 WILLASON, STEWART W, JOHN FAVUZZI, and JAMES L. COX. Patchiness and nutritional condition of zooplankton in the California Current 157 JOHNSON, P. T, R. A. MacINTOSH, and D A. SOMERTON. Rhizocephalan infec- tion in blue king crabs, Paralithodes platypus, from Olga Bay, Kodiak Island, Alaska 177 (Continued on next page) Seattle, Washington 1986 For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington DC 20402— Subscription price per year: $21.00 domestic and $26.25 foreign. Cost per single issue: $6.50 domestic and $8.15 foreign. Contents— Continued Notes WEBER, EARL C, and STEPHEN R. GOLDBERG. The sex ratio and gonad indices of swordfish, Xiphias gladius, caught off the coast of southern California in 1978 185 UCHIYAMA, JAMES H., RAYMOND K. BURCH, and SYD A. KRAUL, JR. Growth of dolphins, Coryphaena hippurus and C. equiselis, in Hawaiian waters as determined by daily increments on otoliths 186 FROST, KATHRYN J., and LLOYD F. LOWRY Sizes of walleye pollock, Theragra chalcogramma, consumed by marine mammals in the Bering Sea 192 VAN ENGEL, W. A., R. E. HARRIS, JR., and D. E. ZWERNER. Occurrence of some parasites and a commensal in the American lobster, Homarus americanus, from the Mid- Atlantic Bight 197 COLLETTE, BRUCE B. Resilience of the fish assemblage in New England tide- pools 200 JOHNSON, PHYLLIS T Parasites of benthid amphipods: ciliates 204 MASON, J. C. Fecundity of the Pacific hake, Merluccius productus, spawning in Canadian waters 209 SELZER, LAWRENCE A., GREG EARLY, PATRICIA M. FIORELLI, P. MICHAEL PAYNE, and ROBERT PRESCOTT Stranded animals as indicators of prey utiliza- tion by harbor seals, Phoca vitulina concolor, in southern New England 217 WARNER, JOHN, and BOYD KYNARD. Scavenger feeding by subadult striped bass, Morone saxatilis, below a low-head hydroelectric dam 220 RANCK, CAROL L., FRED M. UTTER, GEORGE B. MILNE R, and GARY B. SMITH. Genetic confirmation of specific distinction of arrowtooth flounder, Atheresthes stomias, and Kamchatka flounder, A. evermanni 222 The National Marine Fisheries Service (NMFS) does not approve, recommend or en- dorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to NMFS, or to this publication furnished by NMFS, in any advertising or sales promotion which would indicate or imply that NMFS ap- proves, recommends or endorses any proprietary product or proprietary material mentioned herein, or which has as its purpose an intent to cause directly or indirect- ly the advertised product to be used or purchased because of this NMFS publication. Best NMFS Publications for The Publications Advisory Committee of the National Marine Fisheries Service has an- nounced the best publications authored by the NMFS scientists and published in the Fishery Bulletin and the Marine Fisheries Review for 1983. Only effective and inter- pretive articles which significantly contrib- ute to the understanding and knowledge of NMFS mission-related studies are eligible, and the following papers were judged as the best in meeting this requirement: "Seasonal variation in survival of larval northern anchovy, Engraulis mordax, estimated from the age distribution of juveniles" by Richard D. Methot, Jr. appears in Fishery Bulletin 81:741-750.' Richard D. Methot, Jr., fishery biologist is from the Southwest Fisheries Center's La Jolla Labo- ratory, National Marine Fisheries Service, NOAA, 8604 La Jolla Shores Drive, La Jolla, California 92038. "To increase oyster production in the north- eastern United States" by Clyde L. MacKenzie, Jr. appears in Marine Fisheries Review 45(3): 1-22. Clyde L. Mackenzie, Jr., fishery biologist is from the Northeast Fisheries Center's Sandy Hook Laboratory, National Marine Fisheries Service, NOAA, Highlands, New Jersey 07732. AWAR STARVATION-INDUCED MORTALITY OF YOUNG SEA-CAUGHT JACK MACKEREL, TRACHURUS SYMMETRICUS, DETERMINED WITH HISTOLOGICAL AND MORPHOLOGICAL METHODS Gail H. Theilacker1 ABSTRACT Young jack mackerel, Trachurxis symmetricus, living offshore are starving while those living nearshore are healthy. These results for sea-caught jack mackerel were determined by using histological and mor- phological criteria that reliably diagnosed the viability of laboratory-raised jack mackerel. Both the histological and morphological indices indicated that 350 km offshore about 70% of the first-feeding jack i mackerel were starving. In contrast, 12% of the fish collected near islands and banks were starving. In both habitats, mortality rates decreased to zero for jack mackerel at 2 weeks of age The accuracy of the techniques for prediction of the nutritional state of wild larvae is discussed and evaluated. Jack mackerel, Trachurus symmetricus, hatch with yolk reserves that last for 5dat 15°-15.5°C. After the yolk is absorbed, they must eat within 3 d or die of starvation. In addition, growth is retarded in lar- vae that have experienced only 1 d of starvation, and resumption of normal growth does not occur until 2-3 d after the starvation period (Theilacker 1978, 1981). Thus, in the laboratory, availability of food at the time of first feeding affects growth and survival of young jack mackerel. In the field, the relative im- portance of starvation as a source of mortality of jack mackerel is unknown. It was first suggested by Hjort (1914) (reviewed by May 1974) that the strength of the year class is determined early in life by the availability of food for larvae at the time of first feeding (the critical period hypothesis). But only recently (O'Connell 1980) has the presence of starving ocean-caught larvae been documented. In this study I give evidence that starvation may be a major cause of natural mortality of young jack mackerel at sea. I use two techniques, developed in the laboratory, to determine the incidence of starva- tion (Theilacker 1978). The potential use of these techniques to monitor sea samples for larval survival is discussed. METHODS Collection In May 1980 a concentration of jack mackerel eggs and larvae was located 350 km off the coast of 'Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. Manuscript accepted February 1985. -mr""1"" ",TT 1 cmm t™ aj Ufl i mag California (lat. 31°00'N and long. 120°30'W). A 400 mi2 grid was established which contained 41 sta- tions, 4 mi apart; it took 4 d to sample all stations (Fig. 1). At each station, a standard oblique bongo net tow (Smith and Richardson 1977) and aim net sample were taken. The bongo samples will be used in another study to estimate growth and mortality of jack mackerel larvae (Hewitt et al. in press). The 1 m net (505 /urn mesh) was used to sample larvae qualitatively from the upper 50 m of water. Ahlstrom (1959) found that 88% of the larval jack mackerel collected off California were in the upper 50 m, and all the jack mackerel collected by Devonald (1983) were above 42 m. A special collection procedure was used for the samples taken for histological and mor- phological analyses. Immediately after the net tow, the sample was preserved in Bouin's solution to avoid autolysis of larval tissues (elapsed time was usually 8 min) (Theilacker 1978). The collecting net was not washed down (a procedure required for quantitative samples), and the cod end containing the sample was placed directly into Bouin's solution. The preserved sample was removed from the cod end within an hour. After 2 d, Bouin's solution was replaced by 70% alcohol. In addition to jack mackerel collections taken in the open ocean 350 km offshore, a few special tows (n = 24) for assessment of starvation were made dur- ing routine cruises in 1978, 1979, and 1980 near the Channel Islands (Anacapa, Santa Barbara, and San Clemente) and Tanner Bank. Preparation of Fish More than 2,000 jack mackerel were collected in 1 FISHERY BULLETIN: VOL. 84. NO. 1 Los Angeles Anacapa Santa Barbara- San Clemente Tanner Bank— o San Diego _ 31* N — Figure 1— Location of jack mackerel, Trachurus symmetricus, col- lections off the coast of California. Nearshore stations were at Anacapa, Santa Barbara, and San Clemente Islands and at Tan- ner Bank. The grid of open-ocean stations was 350 km offshore; stations were 4 mi apart. samples taken offshore; from 0 to 262 fish were caught per sample (Table 1). Larvae sorted from the samples (n = 445) were counted and five body measurements taken: standard length (SL, tip of up- per jaw to perpendicular at end of notochord); head length (HL, tip of upper jaw to cleithrum); eye diameter (ED); body depth at the pectoral (BD-1); and body depth at the anus (BD-2). After measure- ment, some larvae (n = 369) were prepared for histological examination. When samples contained fewer than 50 jack mackerel, most larvae were ex- amined, but when samples contained more than 100 jack mackerel, about 25% of the fish were examined histologically. Jack mackerel size distribution in the offshore study area (determined for 400 fish taken from stations 16, 23, 34, and 35) was similar among stations and ranged between 2.6 and 4.7 mm SL. lb ensure analysis of all ages in the larger samples, fish were taken equally from each of four length classes: <3.0; 3.0-<3.5; 3.5-<4.0; 4.0-<5.0 mm. These larvae were imbedded in paraffin, sectioned at 6 pm, and stained with Harris hematoxylin and eosinphloxine B (Theilacker 1978). In my analysis of histological data I combined the first two size classes because the size at first feeding was 3.2 mm. The prevalence of starvation was assessed for 371 jack mackerel selected from 20 of the 32 positive sta- tions (Table 1). In addition, I analyzed 41 jack mackerel taken in 14 hauls from the inshore stations near the Channel Islands and Tanner Bank. Histological Analysis The histological assessment of nutritional state is based on distinct cellular changes that occur in tissues when larval jack mackerel were deprived of food; these changes are well documented by Umeda and Ochiai (1975), O'Connell (1976), and Theilacker (1978). Tb determine the condition of individual ocean-caught jack mackerel, I used the histological criteria I developed in the laboratory by starving jack mackerel except I did not grade the pancreas. Grades were assigned to 11 histological characteristics of the brain, digestive tract, liver, and musculature (Theilacker 1978, 1981). Fish identities were unknown during this examination. I classified jack mackerel larvae into four categories (healthy, recovering, starving, and dying) according to their histological scores (the summation of the grades for each of the 11 histological characteristics). Tissues of jack mackerel from the sea which had tissues similar in appearance to the tissues of feeding, laboratory-raised fish were classified as healthy; sea-caught jack mackerel which resembled laboratory fish that had fasted before eating were classified as recovering (these fish showed signs of feeding and digestion, but also showed signs of star- vation); sea-caught larvae which were classified as starving resembled larvae that had been starved in the laboratory for 1-3 d (Theilacker 1978, 1981). I did not observe the dying category in laboratory- starved larvae; this category is described in Results. Morphological Analysis Tb detect starvation I used a set of morphological THEILACKER: MORTALITY OF SEA-CAUGHT JACK MACKEREL Table 1.— Number of jack mackerel collected and the condition of those that were ana- lyzed histologically. tation Number of fish s Starv- Recover- No. Sampled Analyzed Dying ing ing Healthy Offshore 1 0 0 2 2 1 0 0 1 0 3 0 0 4 2 0 5 0 0 6 0 0 7 2 1 0 1 0 0 8 2 2 0 0 2 0 9 1 1 1 0 0 0 10 0 0 11 3 3 3 0 0 0 12 0 0 13 1 0 14 1 0 15 >200 0 16 >200 0 17 20 13 0 8 5 0 18 >125 0 19 43 35 7 0 1 27 20 242 64 8 19 13 24 21 >250 0 22 >175 0 23 150 32 1 0 4 27 24 1 0 25 23 0 26 4 3 3 0 0 0 27 0 0 28 262 58 3 36 14 5 29 11 11 1 4 4 2 30 250 57 4 13 18 22 31 32 9 7 1 0 1 32 109 25 0 2 20 3 33 31 23 1 3 10 9 34 38 0 35 43 0 36 31 24 3 4 1 16 37 7 5 2 1 2 0 38 2 2 1 0 0 1 39 0 0 40 1 0 41 0 0 Total (Offshore) >2,264 369 45 92 95 137 Around Islands Anacapa 12 12 0 1 0 11 Santa Barbara 3 3 0 0 2 1 San Clemente 17 17 0 1 5 11 Tanner Bank 9 9 0 1 0 8 Total (Nearshore) 41 41 0 3 7 31 characteristics that successfully diagnosed the ex- tent of starvation in 85% of the laboratory-reared jack mackerel (Theilacker 1978). The technique is based on a stepwise discriminant analysis (SWDA) using 11 body part measurements. The analysis allowed me to distinguish between individuals belonging to fed and starved treatments, given a set of morphological measurements that describe the characteristics of the individuals in each feeding treatment. The 11 body part measurements used to distinguish between groups of fed and starved jack mackerel were 1) head length, 2) eye diameter, 3) body depth at the pectoral, 4) body depth at the anus, 5) head length/standard length, 6) eye diameter/standard length, 7) body depth at the pec- toral/standard length, 8) body depth at the anus/ standard length, 9) eye diameter/head length, 10) body depth at pectoral/head length, and 11) body depth at anus/head length. Standard length was used in the ratios but not as a unit to allow discrimina- tion between feeding and starving fish of the same length. FISHERY BULLETIN: VOL. 84, NO. 1 Adjustment for Shrinkage ® In order to use morphological measurements to diagnose starvation of jack mackerel, it is essential to adjust for shrinkage of body measurements. Both handling and preservation cause shrinkage of lar- val fishes, and the amount of shrinkage varies among body parts. Final fish size is dependent not only on initial size but also on the handling time (which is different for the laboratory and the field) and the type of preservative used (Blaxter 1971; Theilacker 1980a; Hay 1981). The shrinkage of laboratory speci- mens of jack mackerel preserved in Bouin's solution is known (Theilacker 1980a), but for field-collected specimens the shrinkage caused by the net tow and the subsequent effect of Bouin's preservative must be evaluated. I conducted laboratory experiments to estimate the amount of shrinkage caused by handling (net retention) and by preservation. Live jack mackerel were pipetted individually (time = 0) onto a slide, and four body measurements were taken before placing the fish into a net container through which 15°C seawater circulated. Standard length, head length, eye diameter, and body depth at the anus were measured. Body depth at the pectoral fin was not measured because it was difficult to measure quickly on live jack mackerel. During net treatments, I usually remeasured each fish four more times at 5-7 min intervals, replacing the fish in the net be- tween each set of measurements. After 25-30 min, the fish were preserved in either Bouin's fixative (used for histological analyses) or 5% buffered Formalin2 (as per shipboard procedures; Smith and Richardson 1977). Remeasurements after preserva- tion were taken in 3-4 wk. (a\ Shrinkage of net-captured larval fish has been shown to decrease with increasing fish size For ex- ample, shrinkage of northern anchovy decreased from 19% at 4 mm SL to 8% at 18 mm SL (Theilacker 1980a). The jack mackerel tested in this study ranged between 3.35 and 4.10 mm SL, and within this restricted length group shrinkage was proportional to size Thus for the shrinkage analy- ses, all jack mackerel were combined into one group. For the combined size group, length of the jack mackerel body (Fig. 2) and the head continued to shrink for the duration of the net treatment. Width of the body (Fig. 3) and the eye shrank initially, and then remained relatively constant during additional treatment. To account for positive correlation be- tween body parts, a multivariate analysis (Table 2) was used to relate the ratio of net-treated size to live size (for each body part) with treatment time In- dividual shrinkage was highly variable; for example, shrinkage of body depth varied between 0 and 23% for treatment times between 5 and 20 min (Fig. 3). However, since these were the best estimates of average shrinkage for body parts, the regressions (Table 2) were used to calculate the adjustment fac- tors needed for this study. Factors for each body part ui > Q UI < 111 CE H I »- UI a z UI -I a K < o z < l.U • 1 •• • ••• • • • • • • 1 ••• • • • • 0.9 • • • •• ••• ••• 1 • • • • •1 • • • • • • • • • • • • • • « • • • • 0.8 • • • • • • • • • • • • t 0.7 _ 0.6 " i i 1. .. ...1 . _L.. j 0.0 6.0 12.0 18.0 TIME (MIN) 24.0 30.0 Figure 2— Shrinkage of standard length, shown as the ratio of net- treated size to live size, of individual Trachurus symmetricus lar- vae as a function of net-treatment time; estimated parameters are in Table 4. 1.0 | t l. .... UI > 3 ... . ... • a HI < UI a 0.9 - 7 * .\..| i..i.|. «|..|t |. • • * t- Ul 0.8 . . . • • •• z • z a. UI a 0.7 • • > a o a 0.6 li i i i | J 0.0 6.0 12.0 18.0 24.0 TIME (MIN) 30.0 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Figure 3— Shrinkage of body depth, shown as the ratio of net- treated size to live size, of individual Trachurus symmetricus lar- vae as a function of net-treatment time; estimated parameters are in Thble 4. THEILACKER: MORTALITY OF SEA-CAUGHT JACK MACKEREL $ Table 2.— Shrinkage of jack mackerel larvae. Parameters estimated from multi- variate linear equations relating the ratio of the net-treated size of a mackerel body part to its live size (y) with the net-treatment time (x). Net-treated size/ live size1 (SE) (SE) pz Standard length (SL) Head length (HL) Eye diameter (ED) Body depth (BD-2) 1.0109 (0.0117) 0.9281 (0.0157) 0.9360 (0.0168) 0.8980 (0.0177) -0.0105 (0.0008) -0.0038 (0.0011) -0.0027 (0.0012) -0.0014 (0.0013) <0.001 0.66 0.001 0.12 0.031 0.06 0.280 0.02 1n = 89. Probability that slopes differ from zero. were calculated by 1) combining the shrinkage ratio at 8 min (average elapsed time for field collections, see Methods) with 2) the average shrinkage in Bouin's preservative after the net treatment, and 3) comparing the combined shrinkage with results from shrinkage determined in the laboratory study (Theilacker 1980a; Table 3). Also given in Table 3 are average shrinkage ratios calculated for specified time intervals. Adjustment factors for standard length, head length, and eye diameter (Table 3) support the view that shrinkage of field-collected fishes is greater than shrinkage of fishes preserved in the laboratory. Shrinkage of BD-2 was an exception to this pattern, however, as less shrinkage occurred under simulated field conditions (20-23%) than in the laboratory (25%). I (Theilacker 1980a) reported a similar paradox for northern anchovy where simulated-field net treatments caused 8% shrinkage of BD-2 as com- pared with 10% shrinkage for standard laboratory preservation. Jack mackerel shrinkage was greater in Bouin's solution than in Formalin, results which are consistent with studies on northern anchovy. Also, as with northern anchovy, Formalin preserva- tion caused a slight increase in the size of the jack mackerel eye (Table 3). I adjusted the body measurements of the ocean- caught jack mackerel with the shrinkage factors (ratio R8, Table 3). Use of these adjustments should equate the morphology of preserved, ocean-caught jack mackerel (this study) with the morphology of preserved, laboratory-raised jack mackerel that were used to develop the morphological SWDA (see Methods: Morphological Analysis). It was necessary to reestimate the SWDA function for this study, although nearly the same analysis was made previously (Theilacker 1978). A new estimate was re- quired because pectoral body depth was not included in the shrinkage measurements in this study; hence, an SWDA function that excluded this measurement was needed. Elimination of pectoral body depth from the analysis reduced the level of predictability from 85% to 78%. This new function was used here to classify the condition of ocean-caught jack mackerel Table 3.— Shrinkage of jack mackerel larvae1. Treatment ratio (R) is treated size divided by previous size (1.00 = no shrinkage). Ratios Treatment Mean Standard Head Eye Body ratio R n time length length diameter depth 8 min net/live size 2«1 89 8 0.93 0.90 0.91 0.89 5-10 min net/live size Ro 36 7.3 0.94 0.90 0.92 0.89 11-15 min net/live size R, 22 12.6 0.87 0.88 0.88 0.86 16-28 min net/live size a, 27 19.4 0.81 0.86 0.89 0.88 Bouin's fixative/ net-treated size 3fls 15 — 0.91 0.84 0.93 0.91 Formalin fixative/ net-treated size 3«« 13 — 0.96 0.93 1.08 0.91 Laboratory-preserved in Bouin's fixative live size 4*7 45 — 0.92 0.82 0.90 0.75 Calibration factor = R7IR,xR5 5*8 — — 1.09 1.08 1.06 0.93 1Range in standard length 3.35-4.10 mm. Calculated from regression (Table 2); ocean-caught fish preserved within 8 min; see text. Shrinkage in fixative after net treatment. "Data from Theilacker (1980a). Adjustment factor to equate measurements of field-collected mackerel (this study) with measurements of laboratory-raised mackerel (Theilacker 1978). FISHERY BULLETIN: VOL. 84, NO. 1 after the size of their body parts was adjusted for shrinkage. RESULTS Habitat Conditions A larval-density gradient was apparent in the open ocean study area. High densities of jack mackerel larvae (100-<300/sample) were found in the central stations and in stations near the western boundary of the grid; lower densities (20-50) were found to the north and east, and densities of larvae approached zero at the southern stations that were occupied at the beginning and again at the end of the 4-d obser- vation period (Fig. 4). Larval densities in the south did not change during this period. The study area was chosen because temperature, viewed on satellite thermal image of the sea surface, corresponded to the temperature range (15°-16°C) associated with jack mackerel spawning (Farris 1961). Surface temperature in the study area in- creased from 15.2°C in the north to 16.8°C at the southern stations, with the majority of jack mackerel found in water temperatures of 16.1°-16.6°C. Water temperatures inshore of the grid were about 14°C. A temperature-salinity curve obtained at station 19 (Fig. 1) agreed well with the curves obtained from inshore stations with the exception of the warm- water portion of the curve, which appeared to be a thin, warm lens of open ocean water intruding coast- ward over deeper coastal water. Histological Assessment of Fish Condition I used the tissue characteristics of laboratory fish (raised at 15.0° -15.5° C) of known feeding history as the criteria to determine the nutritional condition of the sea-caught jack mackerel. Photomicrographs of the diagnostic tissue characteristics were documented by Theilacker (1978). Many of these characteristics are shown also for wild fish (Fig. 5, see also Figures 6-14). In addition, the wild fish ex- hibited four tissue conditions that were not observed in the laboratory: lesions in the brain; luminal vacuoles in the midgut; total degeneration of the midgut mucosal cells; and a wavy configuration of the muscle fibers. Each of these conditions will be considered in the following section that describes the tissues of ocean-caught fish. My emphasis will be on those tissue characteristics that diagnose starvation in young jack mackerel. Brain The brain of an ocean-caught jack mackerel was considered normal when the neurons were distinct, round, and closely spaced. In these fish, brain cell division was common, but it was not graded. One percent of the jack mackerel examined had brain le- sions of the type (Fig. 6) induced by ultraviolet light in larval northern anchovy, Engraulis mordax, and Pacific mackerel, Scomber japonicus (Hunter et al. 1979). The grading system classified these jack mackerel (n = 3) into the healthy category. In a ^Pt. Conception 'O^s? ~"*a*fcl_os Angeles 'San Diego 31» N • — 120.5* W NUMBER OF LARVAE 0-5 20-50 100-<300 Figure \.—Trachurus symmetricus larval density gradient shown as number of larvae collected per sample (not quantitative). Sta- tion grid located 350 km off the coast of California. THEILACKER: MORTALITY OF SEA-CAUGHT JACK MACKEREL single specimen, lesions were present not only in the brain but throughout the spinal cord (Fig. 7) as well. In addition, the gut and associated glands had deteriorated to the extent that this fish was con- sidered starving. An abnormal central nervous system of a jack mackerel larva consisted of vacuolar degeneration and shrinkage of neurons. The degenerating neurons exhibited increased staining (Fig. 8). Digestive Track and Associated Glands The midgut mucosa of young jack mackerel is com- posed of a single layer of columnar epithelial cells. Older fish (3.7-4.0 mm) showed increased mitotic ac- tivity in the basal layer. Microvilli bordered the midgut lumen only in fish that appeared healthy. Mucosal cells were closely united in the fish con- sidered to be normal (Figs. 9, 10). Basal separations between these cells were common, not only in fish that appeared to be starving but also in fish that showed signs of feeding and digestion (Fig. 11). O'Connell (1980) also reported that sea-caught north- ern anchovy exhibited basal separations between mucosal cells while the apical portions were well joined. All wild jack mackerel categorized as recovering had basal separations between midgut mucosal cells. Laboratory fish that were artificially starved for 1-2 d before feeding showed these separations for several days after feeding resumed. In the laboratory, lar- vae did not grow while their tissues were regen- erating (Theilacker 1981). Many sea-caught jack mackerel of all ages had intracytoplasmic vacuoles in the midgut epithelium. Basal and membrane lined, these vacuoles resem- bled the vacuolar condition found in some recover- ing, laboratory fish (Theilacker 1981). In addition, many sea-caught larvae had smaller, luminal vacuoles that were found in the laboratory fish (Fig. 12). These luminal vacuoles may indicate a degen- erative condition. In higher vertebrates a metabolic imbalance can cause vacuolar degeneration. Vacuola- tion appears first as numerous small, clear vacuoles dispersed throughout the cytoplasm. As the condi- tion becomes more severe, these minute vacuoles coalesce to form large (sometimes single) clear spaces that displace the nucleus (Anderson 1971). On the other hand, the numerous luminal vacuoles can secrete mucous into the lumen or store fat. Use of a routine mucicarmine staining was negative for the presence of mucous cells. Unfortunately, the presence of fat in the vacuoles could not be tested because fat is removed during tissue preparation by clearing agents. Neither vacuolar condition was graded. Another unusual condition of the midgut occurred in many of the smaller wild jack mackerel. In these fish, the margin of the lumen had lost its integrity, microvilli were absent, and the sloughing of the mucosal cells into the lumen (a condition common in starved laboratory jack mackerel) appeared to have progressed until the lumen contained masses of undefinable, cellular material (Fig. 13). O'Connell (1980) described a comparable condition which he found in the midgut of a single, northern anchovy specimen, the smallest examined. All jack mackerel exhibiting this condition were smaller than the size attained at first feeding, indicating shrinkage had occurred. The hindgut also contained necrotic debris, and other diagnostic tissues were in poor condition. These jack mackerel were classified as dying. Hindgut mucosal cells of wild jack mackerel typically showed eosin-staining inclusions that are reported to be sites of intracellular digestion (Iwai 1968, 1969; Iwai and Tanaka 1968; Watanabe 1981). Inclusions in the wild jack mackerel varied in inten- sity; in healthy specimens the intensity appeared to be related to time of day (feeding period), increasing during daylight hours and decreasing during the night. Although the presence and intensity of hind- gut inclusions were noted, they were not graded. Inclusions were not present in larval teleosts de- prived of food in the laboratory (Theilacker 1978; Umeda and Ochiai 1975; O'Connell 1976). However, in many wild jack mackerel showing signs of starva- tion the presence of pale inclusions indicated that the fish had eaten at some time in the past. The key diagnostic characteristics of the pancreas were obscure in ocean-caught jack mackerel because of the intensity of staining. In laboratory fish, the pancreas was very sensitive to food deprivation. For example, a breakdown in the symmetry of the acinar secretory unit was detectable after 1 d of food deprivation (Theilacker 1978). In the wild fish, the intensity of the staining of the pancreas was difficult to control (see Fig. 12), and I was not able to obtain consistent results, hence the condition of the pan- creas was not evaluated. The jack mackerel liver was considered normal when hepatocytes had clear, distinct nuclei (Fig. 9). The appearance of the cytoplasm was quite variable; in some larvae very few intracellular spaces existed in the cytoplasm of the hepatocytes whereas in others extensive intracellular spaces existed. Presumably these spaces are areas where glycogen and fat are stored within the cell. This presumed in- corporation of stores was most marked in healthy FISHERY BULLETIN: VOL. 84, NO. 1 mitotic notochord swim brain muscle bladder Figure b— Trachurus symmetricus larva, 3.75 mm SL. All 11 histological criteria graded as healthy. Bar = 281 ^m. Figure 6— Head of Trachurus symmetricus larva graded healthy. Mitotic activity and the location of brain lesions are indicated. Bar = 47 \im. B = brain. Figure 1— Trachurus symmetricus larva graded as starving. Lesions present throughout brain and spinal cord. Bar = 47 ^m. B = brain, N = notochord. Figure 8— Pectoral area of a dying Trachurus symmetricus larva showing darkly stained primitive nerve cells, wavy muscle fibers, necrotic and atrophied liver, and loss of integrity of midgut mucosal cells. Bar = 47 ^m. FG = foregut, L = liver, m = muscle, MG = midgut, N = notochord, SB = swim bladder, SP = spinal cord. Figure 9— Pectoral area of healthy Trachurus symmetricus larva collected offshore showing parallel muscle fibers and abundant inter- muscular tissue, distinct nuclei in liver and midgut, and good cellular integrity. Note deflating swim bladder. Bar = 47 ^m. FG = foregut, IM = intermuscular tissue, L = liver, M = muscle, MG = midgut, N = notochord, P = pancreas, SB = swim bladder. Figure 10— Pectoral area of healthy Trachurus symmetricus larva collected near San Clemente Island showing abundant glycogen reserves in the liver. Bar = 47 y.m. FG = foregut, L = liver, M = muscle, MG = midgut, N = notochord, P = pancreas, SB = swim bladder. 8 THEILACKER: MORTALITY OF SEA-CAUGHT JACK MACKEREL jack mackerel collected near islands and banks (Fig. 10) whereas healthy jack mackerel collected offshore showed moderate to little storage (Fig. 9). At the other end of the grading scale, the shrunken livers of jack mackerel considered to be starving con- tained darkly stained hepatocytes composed of even- ly stained cytoplasm with indistinct, irregular nu- clei. Musculature Healthy muscle tissue in jack mackerel had the following characteristics: few spaces between the muscle fibers; distinct and parallel, striated myo- fibrils; and abundant, basophilic and nucleated intra- muscular tissue (Fig. 9). Nourishment was con- sidered inadequate in fish exhibiting separated (Figs. 11, 14) and hyaline muscle fibers (Fig. 13) and a reduction (Figs. 11, 14) or absence (Fig. 13) of intra- muscular tissua In some sea-caught jack mackerel, muscle fibers were wavy (Fig. 8). Presence of wavy muscle fibers in wild fish was considered abnormal because it was always associated with the poor con- dition in the other diagnostic tissues, but this charac- teristic was not used in classification. In starved laboratory fish, nonparallel fibers were reported (Theilacker 1978, 1981), but the wavy pattern was unusual. There were fish with intermediate spaces between muscle fibers that, according to the scores of the other diagnostic tissues, appeared healthy. The Figure 11.— Trachurus symmetricus larva graded recovering. Prominent separations between midgut and hindgut epithelial cells, slight muscle fiber separation and intermediate intermuscular tissue containing distinct nuclei. Bar = 47 \im. HG = hindgut, IM = inter- muscular tissue, M = muscle, MG = midgut, N = notochord. Figure 12— Healthy Trachurus symmetricus larva showing luminal vacuoles in the midgut. This histological characteristic was not graded. Bar = 47 ^m. M = muscle, MG = midgut. Figure 13.— -Trachurus symmetricus larva graded dying. No intermuscular tissue; hyaline muscle fibers; total degeneration of midgut mucosa. Bar = 34 ^m. HG = hindgut, M = muscle, MG = midgut. Figure 14— Recovering Trachurus symmetricus larva showing slight muscle fiber separation and slight reduction of intermuscular tissue Bar = 47 \im. HG = hindgut, IM = intermuscular tissue, M = muscle, MG = midgut, N = notochord. FISHERY BULLETIN: VOL. 84, NO. 1 grading system usually classified these fish into the recovering category. General Histological Observations In jack mackerel that were considered healthy, swim bladder inflation was first noted at 3.4 mm. Swim bladders were inflated in larvae taken at night whereas they were deflated in those taken in the day. The swim bladders of 72% of the fish were deflated by 0700 (n = 81) except for fish scored in the starving category where inflation was common at any time of day, which was possibly a symptom of starvation or an additional energy-sparing func- tion of the swim bladder (Hunter and Sanchez 1976). Theilacker (1978) pointed out that the gallbladder was always enlarged in jack mackerel that were deprived of food in the laboratory, and this condi- tion occurred in sea samples of starved larvae taken in the day. On the other hand, gallbladder enlarge- ment was also found in the healthy fish as well as starved fish collected at night. According to Love (1970), the gallbladder discharges its contents when stimulated by food. Jack mackerel do not eat at night, so the gallbladder of healthy fish may remain distended during the night. Thus enlargement of the gallbladder was not used to diagnose starvation. Theilacker's (1978) samples of fed and unfed fish were taken only during the day, when feeding oc- curs. Mitotic figures in the brain of jack mackerel oc- curred in fish collected at all times of day and night. On the other hand, mitosis of mucosal cells in the midgut was restricted to the night. It seems that mucosal cells of northern anchovy also divide late at night, when the digestive tracts are empty (O'Con- nell 1981). - Evidence for Starvation in the Sea s Results of the histological analysis showed that starvation was a major source of mortality for the smallest jack mackerel larvae (<3.5 mm) as 59% ap- peared to be dying of starvation, 23% were eating but had fasted previously, and only 19% were class- ed as healthy. The incidence of starving larvae decreased to 16% in the 3.5-4.0 mm size class and was 3% in the older larvae (Table 4). The numbers of fish used for the histological assessment of star- vation was adequate for the smallest (<3.5 mm SL) larval size class (coefficient of variation ranged be- tween 0.09 and 0.15 for the four condition categories), but larger samples would be needed to give a reliable estimate of the fraction starving for the older larvae (>3.5 mm SL) because of the low incidence of starvation. Despite the fact that jack mackerel abundance decreased from west to east and north to south (Fig. 4), I found no consistent differences in the incidence of starvation between fish taken from areas of high larval density and those taken from areas of low lar- val density (Fig. 15). Therefore, to estimate mortality due to starvation, I combined all samples collected in the offshore area. To estimate mortality rates on a daily basis, the observed number of fish belong- Table 4. — Histological condition of jack mackerel collected 350 km off the coast of California. cc < - < 50% 0% 1950 I960 YEAR CLASS 1970 Figure 17— Relative recruitment strengths of jack mackerel year classes in southern California. Virtual year-class strength is measured by the sum of percentage contributions to seasonal land- ings over the lifetime of the year class. The dashed line indicates average strength (from MacCall and Stauffer 1983; Fig. 4). feeding (Theilacker 1980a), and 4) a growth rate of 0.37 mm/d for healthy sea-caught northern anchovy (Methot and Kramer 1979). Although the number of first-feeding larvae was low in O'Connell's data (n = 23), I calculated a starvation-induced mortal- ity rate of between 35 and 46%/d. Thus my calcula- tions indicate that substantial numbers of northern anchovy larvae as well as jack mackerel larvae are dying at the time of first feeding. This loss rate for northern anchovy is similar to estimated total mor- tality rate at this stage, 39%/d (Lo in press; 1978 data), which suggests that starvation is the major source of mortality at first feeding. This conclusion for northern anchovy could not be drawn at the time that O'Connell did his work because the data on net shrinkage were not known. The average rates estimated by O'Connell were much lower because he combined larval size classes. Attempts to assess larval starvation in the sea using morphological criteria are more common (Shelbourne 1957; Honjo et al. 1959; Nakai et al. 1969; reviewed by May 1974; Ehrlich et al. 1976), but they have seldom been successful, probably because of the biases introduced by failure to correct ade- quately for shrinkage (see next section). Recently Devonald (1983) used a morphometric index with shrinkage adjustments to assess jack mackerel feeding regimes off California. She found good correspondence between jack mackerel condition and prey availability and concluded that feeding con- ditions were better near islands than in the area between islands. Several of her samples and my samples were taken concurrently (San Clemente and Tanner Bank; Table 1), and I found that 92% of the jack mackerel from the island habitat were healthy. Thus, my results obtained using histological criteria confirm Devonald's conclusion. Other techniques used in the past to assess food availability include RNA/DNA (Buckley 1980), food in gut (Rojas de Mendiola 1974; Ciechomski and Weiss 1974; Arthur 1976; Ellertsen et al. 1981), and otoliths. Of course otolith work is critical because estimates of growth rates are essential for assess- ment of mortality, but it is of no value for assessing growth at the onset of feeding (Methot 1981). Arthur (1976) conducted the only other study on the feeding of jack mackerel off the coast of Califor- nia. He found, after examining the stomach contents of 750 specimens from 65 offshore samples, that 60% of the first-feeding jack mackerel and 10% of the older larvae (7 mm) had empty stomachs. This obser- vation lends additional credence to my histological evaluation of jack mackerel collected offshore that shows 59% of the first-feeding fish and 3% of the older fish (>4 mm) were starving. I believe my estimates of jack mackerel mortality due to starvation are conservative The assumptions I made about the persistence of starvation and the duration of growth were based on extensive laboratory studies (Theilacker 1978, 1981). Because the majority of jack mackerel were collected at sites warmer (16.1°-16.6°C) than the culture temperature (15°-15.5°C), the durations for growth and starvation may be altered, but the final estimate of mortality due to starvation is higher after the appropriate changes to the durations are made Furthermore, if net retention of robust fish is greater than reten- tion of thin fish of the same length, starvation may be underestimated. In addition, the selection of unhealthy larvae by predators would also increase the starvation estimate Previous evidence supporting the occurrence of starving fish larvae in the ocean has been mainly cir- cumstantial (reviewed by May 1974; Jones and Hall 1974; Lasker 1975). Evidence from this study and O'Connell's (1980) study shows that starvation does occur and that the young stages of jack mackerel and northern anchovy are highly vulnerable Comparison of Morphological and Histological Criteria for Starvation Diagnosis The incidence of starvation based on mor- 14 THEILACKER: MORTALITY OF SEA-CAUGHT JACK MACKEREL phological criteria was essentially the same as that based on histological criteria. Owing to the relative ease, and low cost of measuring fish compared with a histological examination, the morphological analysis is an attractive approach. On the other hand, histological analysis defines a cause and effect rela- tion between structure and starvation whereas gross morphological measurements provide an index of starvation which is highly vulnerable to errors and biases in calibration and interpretations. Because of the importance of these measurements in recruit- ment studies, it is appropriate to consider the merits of and potential errors in these techniques in some detail. (2/ The morphometric approach relies on measure- ments of fish to compare reared and wild animals at the same developmental stage Thus shrinkage ad- justments are needed to intercalibrate laboratory measurements and field measurements. Fish shrink when collected in a net and preserved, and shrinkage of the size of all body parts is dependent on the time in the net, size of fish, and type of preservative used (Blaxter 1971; Theilacker 1980a; Hay 1981). In this study, tow time was controlled at 5 min and samples were preserved within 8 min. Thus damage to the fish and shrinkage were minimal, but the samples were not quantitative It is doubtful that the morpho- metric technique will work with jack mackerel taken in standard, quantitative collections. Quantitative net tows are 20 min, and they include an additional hosing down of the nets before sample preservation (Smith and Richardson 1977). The procedure damages the larvae, causing extensive shrinkage which makes accurate measuring difficult. Further, a long tow time decreases confidence in time-specific shrinkage estimates because fish can be collected at any time during the towing period. Increasing the tow time also causes both the magnitude of the shrinkage correction factor and the standard error of its estimate to increase For example, in this study, standard length of jack mackerel shrank by an average of 6.0 ± 0.6% in 8 min and 19.0 ± 1.0% in 20 min. While laboratory calibration is absolutely essen- tial for the morphometric analysis, no shrinkage calibration is needed for the histological analysis, and it might be possible to use the histological observa- tions on other fishes. Diagnostic criteria for the starving condition of jack mackerel (Theilacker 1978), northern anchovy (O'Connell 1976), and yellowtail, Seriola quinqueradiata, (Umida and Ochiai 1975) were similar. In addition, important biological information is gained while using the histological approach whereas gross morphological indices provide no such information. For example, histological analysis of jack mackerel has revealed a pattern of diel swim bladder inflation and a disrup- tion of this rhythm, accumulation of glycogen reserves, and brain lesions presumably produced by UV radiation (Hunter et al. 1979). There is just no substitute for this extensive biological information. On the other hand, population work requires large samples, and morphological indices are probably the only practical means for working with very large samples. Thus, the optimal experimental design for population work on starvation is probably the use of morphological criteria (calibrated for shrinkage) combined with a smaller subsample of fish which are graded histologically. All work requires special net tows, preservation, procedures, and laboratory calibration. Caution needs to be exercised when transferring information obtained in the laboratory to the field. Raising larval jack mackerel in small containers is known to affect growth, nutritive condition, and possibly activity (Theilacker 1980b). Additionally, there is evidence that wild fish tend to be thinner than their laboratory counterparts (larval herring, Blaxter 1971; juvenile herring, Balbontin et al. 1973; larval northern anchovy, Arthur 1976). My use of the morphometric SWDA assumes that the morpho- metric criteria I developed in the laboratory for lar- val jack mackerel raised in large tanks are applicable to ocean-caught jack mackerel. ACKNOWLEDGMENTS Many thanks to Brian Rothschild who suggested research on the nutritive condition of larval fish and to William T (Tosh) Yasutake who offered me a per- sonalized course in teleost histology. The offshore collections were made possible by Roger Hewitt's ef- fective planning, the crew of the RV David Starr Jordan, and the assistance of Jack Metoyer and Carol Kimbrell. Miguel Carrillo sorted the mackerel, Richard Kiy measured them, and Jack Metoyer prepared them for histological analyses. Metoyer also helped with the shrinkage study. Nancy Lo assisted with all statistical applications. I appreciate John Hunter's and Martin Newman's constructive reviews of the manuscript. Many thanks to the Technical Sup- port Group for typing services. LITERATURE CITED Ahlstrom, E. H. 1959. Vertical distribution of pelagic fish eggs and larvae off California and Baja California. U.S. Wildl. Serv., Fish. 15 FISHERY BULLETIN: VOL. 84, NO. 1 Bull. 60:107-146. Anderson, W. A. D. 1971. Degenerative changes and disturbances of metabolism. In W. A. D. Anderson (editor), Pathology, p. 68-95. The C. V. Mosby Co., St. Louis, 6th ed. Arthur, D. K. 1976. Food and feeding of larvae of three fishes occurring in the California Current, Sardinops sagax, Engraulis mordax, and Trachurus symmetricus. Fish. Bull., U.S. 74:517-530. 1977. Distribution, size, and abundance of microcopepods in the California Current system and their possible influence on survival of marine teleost larvae Fish. Bull., U.S. 75: 601-611. Balbontin, F, S. S. de Silva, and K. S. Ehrlich. 1973. A comparative study of anatomical and chemical charac- teristics of reared and wild herring. Aquaculture 2:217- 240. Beers, J. R., and G. L. Stewart. 1967. Micro-zooplankton in the Euphotic zone at five locations across the California Current. J. Fish. Res. Board Can. 24: 2053-2068. 1970. Numerical abundance and estimated biomass of micro- zooplankton. In J. D. H. Strickland (editor), The ecology of the plankton off La Jolla, California, in the period April through September, 1967. Vol. 17 (Part VI), p. 67-87. Bull. Scripps Inst. Oceanogr. Blaxter, J. H. S. 1971. Feeding and condition of Clyde herring larvae Rapp. P.-v. Reun. Cons. int. Explor. Mer 160:128-136. Buckley, L. J. 1980. Changes in the ribonucleic acid, deoxyribonucleic acid, and protein content during ontogenesis in winter flounder, Pseudopleuronectes americanus, and effect of starvation. Fish. Bull., U.S. 77:703-708. ClECHOMSKI, J. D, DE, AND G. WEISS. 1974. Estudios sobre la Alimentacion de larvas de la merluza, Merluccius merluccius Hubbsi de la anchoita, Engraulis an- choita, en la mar. Physis. Seer. A Buenos Aires 33:199-208. Devonald, K. F. 1983. Evaluation of the feeding success of first-feeding jack mackerel larvae off southern California, and some con- tributing factors. Ph.D. Thesis, Univ. California, Scripps Inst. Oceanogr., 227 p. Ehrlich, K. F, J. H. S. Blaxter, and R. Pemberton. 1976. Morphological and histological changes during growth and starvation of herring and plaice larvae Mar. Biol. (Berl.) 35:105-118. Ellertsen, B., E. Moksness, P. Solendal, T. Str0mme, S. TlLSETH, T WESTGARD, AND V. 0IESTAD. 1981. Some biological aspects of cod larvae (Gadus morhua L.). ICES Symposium on Early Life History of Fish, Woods Hole Mass., April 1979. Rapp. R.-v. Reun. Cons. int. Explor. Mer 178:317-319. Farris, D. A. 1961. Abundance and distribution of eggs and larvae and sur- vival of larvae of jack mackerel (Trachurus symmetricus). U.S. Fish Wildl. Serv., Fish. Bull. 61:247-279. Hay, D. E. 1981. Effects of capture and fixation on gut contents and body size of Pacific herring larvae Rapp. P.-v. Reun. Cons. int. Explor. Mer 178:395-400. Hewitt, R., G. H. Theilacker, and N. C. H. Lo. In press. Causes of mortality in young jack mackerel. Mar. Ecol. Prog. Ser. Hjort, J. 1914. Fluctuations in the great fisheries of northern Europe viewed in the light of biological research. Rapp. P.-v. Reun. Cons. Perm. int. Explor. Mer. 20:1-228. HONJO, K., T KlTACHI, AND H. SUZUKI. 1959. On the food distribution and survival of post larval iwashi-1-Distribution of food organisms, the food of the an- chovy and ecologically related species along the southwest- ern Pacific coast of Honshu, Sept.-Nov. 1958. Reports on the major coastal fish investigations, and the investigations for forecasting of oceanographic conditions and fisheries (preliminary report), February 1959. (Engl. Transl. by S. Hayaski.) Hunter, J. R. 1981. Feeding ecology and predation of marine fish larvae In R. Lasker (editor), Marine fish larvae morphology, ecology and relation to fisheries, p. 33-77. Wash. Sea Grant Pro- gram, Univ. Wash. Press, Seattle Hunter, J. R., and C. Sanchez. 1976. Diel changes in swim bladder inflation of the larvae of the northern anchovy, Engraulis mordax. Fish. Bull., U.S. 74:847-855. Hunter, J. R, J. H. Taylor, and H. G Moser. 1979. Effects of ultraviolet irradiation on eggs and larvae of the northern anchovy, Engraulis mordax, and the Pacific mackerel, Scomber japonicus, during the embryonic stage Photochem. Photobiol. 29:325-338. Iwai, T. 1968. The comparative study of the digestive tract of teleost larvae— V. Fat. absorption in the gut epithelium of goldfish larvae Bull. Jpn. Soc. Sci. Fish. 34:973-978. 1969. Fine structure of gut epithelial cells of larval and juvenile carp during absorption of fat and protein. Arch. Hist. Jpn. 30:183-199. Iwai, T, and M. Tanaka. 1968. The comparative study of the digestive tract of teleost larvae— III. Epithelial cells in the posterior gut of halfbeak larvae Bull. Jpn. Soe Sci. Fish. 34:44-48. Jones, R., and W B. Hall. 1974. Some observations on the population dynamics of the larval stage in common gadoids. In J. H. S. Blaxter (editor), Early life history of fish, p. 87-102. Springer-Verlag. Berl. Lasker, R. 1975. Field criteria for survival of anchovy larvae: the rela- tion between inshore chlorophyll maximum layers and suc- cessful first feeding. Fish. Bull., U.S. 73:453-462. Lasker, R., H. M. Feder, G H. Theilacker, and R. C. May. 1970. Feeding, growth, and survival of Engraulis mordax lar- vae reared in the laboratory. Mar. Biol. (Berl.) 5:345-353. Lo, N. C. H. In press. Egg production of the central stock of northern an- chovy, 1951-1982. Fish. Bull., U.S. 83. Love, R. M. 1970. The chemical biology of fishes. Acad. Press (Lond.), p. 222-257. MacCall, A. D, and G. D Stauffer. 1983. Biology and fishery potential of jack mackerel (Trachurus symmetricus). Calif. Coop. Oceanic Fish. Invest. Rep. 24:46-56. May, R. C. 1974. Larval mortality in marine fishes and the critical period concept. In J. H. S. Blaxter (editor), The early life history of fish, p. 3-19. Springer-Verlag, N.Y. Methot, R. D, Jr. 1981. Growth rates and age distributions of larval and juvenile northern anchovy,, Engraulis mordax, with inferences on larval survival. Ph.D. Thesis, Univ. California, San Diego, 328 p. 16 THEILACKER: MORTALITY OF SEA-CAUGHT JACK MACKEREL Methot, R. D., Jr., and D. Kramer. 1979. Growth of northern anchovy, Engraulis mordax, larvae in the sea. Fish. Bull., U.S. 77:413-423. Nakai, Z., M. Kosaka, M. Ogura, G. Hayashida, and H. Shimozono. 1969. Feeding habit, and depth of body and diameter of digestive tract of Shirasu, in relation with nutritious condi- tions. J. Coll. Mar. Sci. Technol, Tbkai Univ. 3:23-34. O'CONNELL, C. P. 1976. Histological criteria for diagnosing the starving condi- tion in early post yolk sac larvae of the northern anchovy, Engraulis mordax Girard. J. Exp. Mar. Biol. Ecol. 25:285- 312. 1980. Percentage of starving northern anchovy, Engraulis mordax, larvae in the sea as estimated by histological methods. Fish. Bull., U.S. 78:475-489. Rajos de Mendiola, B. R. 1974. Food of the larval anchoveta, Engraulis ringens J. In J. H. S. Blaxter (editor), The early life history of fish, p. 277-285. Springer-Verlag, N.Y. Shelbourne, J. E. 1957. The feeding and condition of plaice larvae in good and bad plankton catches. J. Mar. Biol. Assoc U.K. 36:539-552. Smith, P. E., and S. L. Richardson. 1977. Standard techniques for pelagic fish egg and larva surveys. F.A.O. Fish. Tech. Pap. 175, 100 p. The i lacker, G. H. 1978. Effect of starvation on the histological and morpho- logical characteristics of jack mackerel, Trachurus sym- metrica, larvae Fish. Bull., U.S. 76:403-414. 1980a. Changes in body measurements of larval northern an- chovy, Engraulis mordax, and other fishes due to handling and preservation. Fish. Bull., U.S. 78:685-692. 1980b. Rearing container size affects morphology and nutri- tional condition of larval jack mackerel, Trachurus sym- metricus. Fish. Bull., U.S. 78:789-791. 1981. Effect of feeding history and egg size on the mor- phology of jack mackerel, Trachurus symmetricus, larvae. ICES Symposium on Early Life History of Fish, Woods Hole, Mass., April 1979. Rapp. P.-v. Reun. Cons. int. Explor. Mer 178:432-440. Theilacker, G. H., and K. Dorsey. 1980. Larval fish diversity. In Workshop on the effects of en- vironmental variation on the survival of larval pelagic fishes. Intergov. Oceanogr. Comm. Rep. 28:105-142. UNESCO, Paris. Umeda, S., and A. Ochiai. 1975. On the histological structure and function of digestive organs of the fed and starved larvae of the yellowtail, Seriola quinqueradiata. [In Jpn., Engl, summ.] Jpn. J. Ichthyol. 21:213-219. Watanabe, Y. 1981. Ingestion of horseradish peroxidase by the intestinal cells in larvae or juveniles of some teleosts. Bull. Jpn. Soc. Sci. Fish. 47:1299-1307. 17 HYPOXIA IN LOUISIANA COASTAL WATERS DURING 1983: IMPLICATIONS FOR FISHERIES Maurice L. Renaud1 ABSTRACT Hypoxic bottom water (<2.0 ppm dissolved oxygen) was present in shallow (9-15 m) waters south of cen- tral Louisiana in June and July 1983. It was patchy in distribution from south of Barataria Pass to south and west of Marsh Island. Data suggested that bottom water hypoxia did affect the abundance and distribu- tion of shrimp and bottomfish. Offshore bottom water dissolved oxygen was significantly correlated with 1) combined catches of brown and white shrimp (r = 0.56, P < 0.002), 2) fish biomass (r = 0.56, P < 0.001), and 3) vertical density gradient (r = -0.73, P < 0.001). Several hypoxic stations were in regions designated as potentially hypoxic through a posteriori analysis of satellite data. Micrapogonius undulatus was the dominant fish species nearshore and offshore Penaeus aztecus and P. setiferus were sparsely distributed throughout the study area. The presence of bottom water hypoxia (<2.0 ppm dissolved oxygen) in the nearshore Gulf of Mexico is a common, recurring, and mostly seasonal (June- August) event. It is generally thought to be associ- ated with temperature and salinity stratification ini- tiated by freshwater runoff and with phytoplankton blooms during hot, calm weather (Fotheringham and Weissberg 1979; Bedinger et al. 1981; Comiskey and Farmer 1981; Turner and Allen 1982a, b; Boesch 1983; Leming and Stuntz 1984). Phytoplankton respiration and decomposition of sinking organic matter are major oxygen consuming processes. High oxygen demand of the organic load in freshwater runoff (Gallaway 1981) and lack of a direct oxygen replenishing mechanism (strong winds) in the pres- ence of vertical stratification contribute to hypoxia formation (Harris et al. 1976; Ragan et al. 1978; Swanson and Sindermann 1979; Harper et al. 1981). Christmas (1973) and Boesch (1983) discussed possi- ble nitrate pollution in rivers and coastal hypoxia. Boesch (1983) presented a brief history of hypoxia in the Gulf of Mexico and evaluated its causes and consequences. The extent to which any factor is in- volved with hypoxia formation is unknown. Hypoxia in the Gulf of Mexico has been most noticeable in shallow (<20 m) Louisiana waters. It has been reported infrequently on the Texas shelf (Harper et al. 1981; Gallaway and Reitsema 1981). Low oxygen levels have also been measured east of the Mississippi River Delta inshore of barrier islands 'Southeast Fisheries Center Galveston Laboratory, National Marine Fisheries Service, NOAA, 4700 Avenue U, Galveston, TX 77550. and in inland bays (May 1973; Christmas 1973) and offshore of Mobile Bay, AL (Turner and Allen 1982b). Abnormally high concentrations of moribund fish and crustaceans near the shoreline ("jubilees") in Alabama have also been linked to hypoxia (May 1973). Considerable interest in hypoxia has been renewed by a less than average shrimp harvest in 1982 (Klima et al. 1983) and 19832. In this paper I report the loca- tions and extent of Louisiana coastal hypoxia in 1983 and discuss the interrelationships of fish and shrimp abundance and distribution with environmental parameters. METHODS Nearshore data were collected in a 7.3 m Aqua- Sport at a total of 56 stations from nine transects west of the Mississippi River Delta (long. 89°33'W to 90°14'W) from 1 to 16 June 1983 (Fig. 1). The transects, perpendicular to shore, ranged from 5 to 8 km in length and 1 to 16 m in depth. The six east- ernmost transects were sampled twice, with a sam- pling interval of 14 d. Shrimp and bottomfish were collected at 23 of 56 stations in 15-min tows with a 3.0 m box trawl. Towing speed was about 3 kn. Before each tow, water temperature, salinity, and dissolved oxygen concentration were recorded at 1 m depth intervals with a Hydrolab 8000. Hydro- graphic profiles were made at the remaining 33 stations. An offshore study area extending from long. 21983 Gulf Coast Shrimp Data, NOAA, NMFS. Manuscript accepted January 1985. FTSHFRV RT1I T FTTN- VOT 84 MO 1 1 Q«fi 19 FISHERY BULLETIN: VOL. 84, NO. 1 a o c o a. o v/ C K o -a P C _o s-, o a. -o -a CO oo iH >> 3 i c 3 1-5 E a o o c a! 3 O c a o c i i -a > w M D O 20 RENAUD: HYPOXIA IN LOUISIANA WATERS 90°47'W to 93°02'W was sampled with a 24.4 m steel-hull commercial shrimp trawler from 30 June to 6 July 1983 (Fig. 1). Depth varied from 4 to 20 m and distance from shore ranged from 8 to 54 km. Shrimp and bottomfish were collected at 34 of 65 stations in 20-min tows with a 12.2 m semiballoon trawl. The same trawl was used as a midwater shrimp sampler above previously identified hypoxic water. Surface and bottom measurements of water temperature, salinity, and dissolved oxygen concen- tration were recorded before each tow. Water samples were collected with a Kemmerer bottle Salinities were measured with a refractometer. Temperature and dissolved oxygen concentration were measured with a YSI Model 51-B. Surface and bottom hydrographic data were collected at the re- maining 31 stations. The Southeast Area Monitor- ing and Assessment Program (SEAMAP)3 person- nel collected similar data off Louisiana in June 1983. SEAMAP dissolved oxygen data were included in the contour analyses. The Harvard SYMAP program (Dougenik and Sheehan 1975), a Northwest Alaska Fisheries Center Contour Subroutine, and the Galveston Laboratory Generalized Mapping system were utilized to pro- duce a map of dissolved oxygen contours off Loui- siana. Koi4 presents an indepth explanation of these contour mapping programs. Vertical density gra- dient of the water column, shrimp catch, and fish catch were regressed with bottom water dissolved oxygen concentration. A "best fit" line through the data was determined using the least squares concept. Surface water temperature (°C) and chlorophyll content (mg/m3) were measured off Louisiana by the Coastal Zone Color Scanner (CZCS) aboard the Nimbus-7 satellite Personnel from the Mississippi Laboratories of the Southeast Fisheries Center, working at the National Space Technology Labora- tories, Mississippi, used CZCS and "ground truth" field data to predict potentially hypoxic areas in coastal Louisiana waters. RESULTS AND DISCUSSION Regions of hypoxic bottom water have been detected along portions of the Texas-Louisiana coastline every summer from 1972 to 1983 (Harris 3Southeast Area Monitoring and Assessment Program: a State- Federal cooperative research effort organized to assess the distribu- tion and abundance of shrimp and bottomfish in the Gulf of Mexico. 4Koi, D. 1985. Generalized geographic mapping system. Un- publ. manuscr., 47 p. Southeast Fisheries Center Galveston Laboratory, National Marine Fisheries Service, NOAA, 4700 Avenue U, Galveston, TX 77550. et al. 1976; Ragan et al. 1978; Bedinger et al. 1981; Harper et al. 1981; Reitsema et al. 1982; Boesch 1983). Hypoxia was noted from 16 June to 6 July 1983. It was patchy in distribution and found main- ly in 9 to 15 m depths from south of Barataria Pass to south and west of Marsh Island (Fig. 1). A total of 34 fish and 11 invertebrate species were collected offshore The Atlantic croaker, Micropo- gonius undulatus, and the Atlantic threadfin, Poly- dactylies octonemus, were the dominant bottomfish at 58% and 30% of the stations, respectively; Atlantic bumper, Chloroscombrus chrysurus, was the com- mon pelagic. Brown shrimp, Penaeus aztecus; white shrimp, P. setiferus; mantis shrimp, Squilla empusa; and broken-back shrimp, Trachypenaeus sp., were the most common invertebrates collected, but in small quantities. Total crustacean catch was always <5.0 kg/h. Bottom water dissolved oxygen concentration was significantly correlated with 1) fish biomass (r = 0.56, P < 0.001) (Fig. 2) and the number of brown and white shrimp present (r = 0.56, P < 0.002) (Fig. 3). Shrimp and bottomfish were generally absent from hypoxic stations. Atlantic croaker were not at stations with hypoxic bottomwater, and shrimp catches never exceeded 2 kg/h in the areas. Sea cat- fish, Arisus felis; butterfish, Peprilus paru; and Atlantic bumper were common in trawls at hypoxic sites. These were also the most abundant fish in mid- 4.0r- x Ul 2 i CM 25 2.o < 2 o m x to o o • " • y= a + bLogx r=0.56 0.0 J. I _L _L _L 2.0 4.0 BOTTOM WATER DISSOLVED OXYGEN CONCENTRATION (PPM) 6.0 FIGURE 2— Offshore fish biomass in relation to bottom water dissolved oxygen concentration. 21 FISHERY BULLETIN: VOL. 84, NO. 1 2.0r- a P oc fflS I U-CJ °« oc t- uj >> m q. 2S = 5 OX 1.0 0.0 2.0 4.0 BOTTOM WATER DISSOLVED OXYGEN CONCENTRATION (PPM) 6.0 Figure 3.— Offshore shrimp abundance in relation to bottom water dissolved oxygen concentration. water trawls above previously identified hypoxic areas. Therefore, it was concluded that they were captured from the upper water column as the trawl passed through it. Four brown shrimp, three lesser blue crabs, Callinectes similus, and one mantis shrimp were the only crustaceans captured in five midwater trawls. The relationship between shrimp and bottomfish abundance and distribution indicates that they do not pass through or over hypoxic water masses. Actual avoidance behavior in the field has not been documented. Nearshore, a total of 20 fish and 5 invertebrate species were collected. Atlantic croaker was the dominant species. Brown shrimp were present in low numbers at most stations. White shrimp; blue crabs, Callinectes sapidus; lesser blue crabs; and sea bobs, Xiphopenaeus sp., were the only other crustaceans collected. A high variability in fish and shrimp abun- dance was probably due to the low fishing efficiency of the small net at the deeper nearshore stations. As a result, no significant correlation was present at nearshore stations between bottom water dis- solved oxygen concentration and fish or shrimp abundance Vertical density stratification was present at both nearshore and offshore stations. Dissolved oxygen concentration and vertical density gradient were negatively correlated (r = -0.73, P < 0.001) (Fig. 4). This agrees with Leming and Stuntz (1984) who found a high correlation between bottom dissolved oxygen content and surface to bottom density gra- dients off Louisiana in 1982 (r = -0.74, P < 0.001). Offshore, the mean difference between surface and bottom dissolved oxygen was 6.4 ppm (standard er- ror = 0.40) in hypoxic areas and 1.6 ppm (standard error = 0.08) in nonhypoxic areas. Temperature generally did not vary more than 2°C between the surface and bottom regardless of the area. During the first week of July, 92% of the hypoxic stations were in areas predicted as potentially hypox- ic through a posteriori analyses of remote sensing data. Hypoxic areas were characterized by surface water temperatures near 30 °C, which agrees with Leming and Stuntz (1984). They discussed satellite data acquisition, its value in identifying and forecasting hypoxic regions in the Gulf of Mexico, z III a > x o ^ > | W O Q < « s e z t m *2 2 ° I- H o m 8. Or* 6.0 -' 4.0 - 2.0 - 0.0 Figure 4.— Bottom water dissolved oxygen concentration in relation to vertical density gradient of the water column. Den- sity gradient is expressed as (bottom sigma-t minus surface sigma-t)/depth. 0.5 1.0 1.5 DENSITY GRADIENT 22 RENAUD: HYPOXIA IN LOUISIANA WATERS and its implications regarding shrimp management. The effect of hypoxia on shrimp is not completely understood. It is possible that an extensive area of hypoxic bottom water can act as a physical barrier to juvenile shrimp migration offshore and to post- larval migration into nursery grounds. Limited in- direct evidence supports this hypothesis. Gazey et al. (1982) described a shrimp mark-release study in Louisiana. Extensive longshore and offshore move- ment occurred before the recapture of the shrimp during 1979, when hypoxia was not reported off Louisiana (Fig. 5). In 1978, when hypoxia was wide- spread along the Louisiana coastline (Fig. 6), shrimp did not move comparable distances. It was possible that hypoxia reduced shrimp movement into offshore waters. The most extensive occurrence of hypoxic bottom water recorded in Louisiana coastal waters occurred from May 1973 to May 1974 (Flowers et al. 1975; Ragan et al. 1978). It was widespread between Barataria and Timbalier Passes and extended up to 30 km offshore in some regions. Ragan et al. (1978) reported several areas to be anoxic The duration and severity of this hypoxic condition may have had an impact on the offshore brown shrimp fishery in 1973. Total brown shrimp catch and CPUE (catch per unit effort) in 1973 were significantly lower (paired £-test, P < 0.05) than in 1972 (fn. 2). Catch declined 36% (2.8 million kg) and the mean CPUE was reduced by 120 kg/vessel per d. Movement of juvenile brown shrimp to the offshore fishery occurs from May to August (Cook and Lindner 1970). Monthly catch and CPUE of brown shrimp from January through April 1973 did not differ from the same time period in 1972; however, catch and CPUE from May through December were significantly lower (paired £-test, P < 0.01) in 1973. Postlarval recruitment of brown shrimp occurs from January to May (Baxter and Renfro 1966). An interaction between hypoxia and postlarval recruitment in 1974 might have been responsible for the continued poor harvest of brown shrimp that year. Catch and CPUE were still sig- nificantly lower than in 1972 (paired i-test, P < 0.05). It was not until 1976 that brown shrimp catch sur- passed the 1972 levels (Table 1). A decline in total shrimp catch of Louisiana in 1982 may have been related to a large region of hypoxic bottom water reported by Stuntz et al. (1982). Although hypoxia has not been directly linked to declines in annual catch, its presence during critical LOUISIANA C~-30 ^ f1 Figure 5— Movement of tagged juvenile brown shrimp from Caillou Lake and Barataria Bay expressed as days at large before recapture (from Gazey et al. 1982). Shrimp were released in July 1979. Hypoxia was not documented off this coastal area in 1979. 23 FISHERY BULLETIN: VOL. 84, NO. 1 LOUISIANA Figure 6.— Movement of tagged juvenile brown shrimp from Caillou Lake, expressed as days at large before recap- ture (from Gazey et al. 1982). Shrimp were released in June 1978. Regions of hypoxic bottom water, noted from June to August, are overlaid onto this map (Fotheringham and Weissberg 1979; Bedinger et al. 1981; Comiskey and Farmer 1981). Table 1. — Louisiana brown shrimp catch data. 1972 1973 1974 1975 1976 5-yr average Catch per unit effort (kg/vessel per d) Jan .-Apr. May-Aug. Sept.-Dec. 190 383 376 216 225 233 180 249 328 196 296 346 208 348 268 198 300 310 Annual average 344 1223 '256 288 302 284 Catch (millions of kg) Jan. -Apr. May-Aug. Sept.-Dec. Total 0.831 4.529 2.293 7.653 1.478 2.630 0.822 M.930 0.633 2.702 1.578 14.913 0.645 2.112 1.414 4.171 1.020 5.966 2.601 9.587 0.921 3.588 1.742 6.251 Effort (24-h days fished) Jan. -Apr. May-Aug. Sept.-Dec. Total 4,379 1 1 ,828 6,361 22,568 6,870 11,722 3,528 22,120 3,509 10,852 4,805 19.166 3,288 7,128 4,083 14,499 4,903 17,127 9,715 31,745 4,590 11,731 5,698 22,020 'CPUE and catch data in 1973 and 1974 were significantly lower than that in 1972 (paired Mest, P < 0.05) portions of the shrimp life cycle implicate it as a prob- able source of variation in annual shrimp yield. Sup- port for this viewpoint has been documented in laboratory experiments which indicate that brown and white shrimp detect and avoid water with low oxygen levels.5 Brown shrimp were the least tolerant of the two species. They avoided dissolved oxygen concentrations up to and including 2.0 ppm. White shrimp did not avoid oxygen levels higher than 1.5 ppm. Variable behavior was exhibited by both species at higher treatment levels. Total time (TT) spent in water with 1.5 ppm did not differ between species, 5Renaud, M. 1985. Detection and avoidance of oxygen depleted water by Penaeus setiferus and Penaeus aztecus. Unpubl. manuscr., 16 p. Southeast Fisheries Center Galveston Laboratory, National Marine Fisheries Service, NOAA, 4700 Avenue U, Galveston, TX 77550. 24 RENAUD: HYPOXIA IN LOUISIANA WATERS nor did their response time (RT), i.e, time taken to retreat into normal seawater. However, these measurements were significantly (£-test, P < 0.001) shorter for brown shrimp (TT = 6.2, RT = 3.8 min) versus white shrimp (TT = 20.0, RT = 6.2 min) when tested at 2.0 ppm. Behavioral responses of brown and white shrimp exposed to hypoxic water included 1) an initial increase in activity, 2) walking or swim- ming retreat, and 3) rapid eye movements. White shrimp also exhibited notable abdominal flexing, periods of exhaustion, and sometimes death. These three latter behaviors were not observed with brown shrimp. Dissolved oxygen levels tested are common along Louisiana's Gulf Coast during the summer and early fall. Therefore it is not unreasonable to assume that similar behavioral responses occur in nature Hypoxia in the New York Bight (Swanson and Sindermann 1979) had a severe impact on the com- mercial fisheries of sedentary species. Surf clam, Spisula solidissima; ocean quahog, Arctica islan- dica; and scallop, Placopectin magellanicus, abundance was reduced by 92%, 25%, and 12%, respectively, in the affected area. The response of recreational fish species, summer flounder, Paralichthys dentatus, and bluefish, Pomatomus saltatrix, to low oxygen levels was noted by changes in their distribution patterns during the hypoxic event. Temperature stratification, phytoplankton blooms, spoil deposition, and sewage treatment outflow were alleged major contributors to hypoxia formation in the New York Bight. It was concluded, however, that abnormal climatological and hydrological phenomena were responsible for this hypoxic event. Swanson and Sindermann (1979) stated that effective regulation of waste disposal into riverine and oceanic environments may control or restrict bottom water hypoxia formation. Future research on the phenomenon of hypoxia should be centered on its predictability; remote sensing has potential in this area. Timely informa- tion dissemination on the extent and location of hypoxic areas would help fishermen to avoid areas where low catches might be anticipated or to harvest a crop before it dies or migrates. ACKNOWLEDGMENTS The author expresses his sincere appreciation to 1) the Lousiana Wildlife and Fisheries Department for providing NMFS personnel with services at their Grand Terre Island Marine Laboratory and at their Field Station in Caillou Lake; 2) the Gulf States Marine Fisheries Commission for access to the 1983 SEAMAP data; 3) David Trimm for this major con- tribution to data collection; 4) Dennis Koi for com- puter services and related software analyses, especially those relevant to contour mapping; 5) Frank Patella for acquisition and transformation of several years of Gulf coast shrimp data; 6) Tom Lem- ing for satellite data; and 7) Beatrice Richardson for typing the manuscript. LITERATURE CITED Baxter, K. N., and W. C. Renfro. 1966. Seasonal occurrence and size distribution of postlarval brown and white shrimp near Galveston, Texas, with notes on species identification. U.S. Fish Wildl. Serv., Fish. Bull. 66:149-158. Bedinger, C. A., R. E. Childers, J. W. Cooper, K. T. Kimball, and A. Kwok. 1981. Pollution fate and effect studies. In C. A. Bedinger (editor), Ecological investigations of petroleum production platforms in the central Gulf of Mexico, Vol. 1, Part 1, 53 p. Report to the Bureau of Land Management, New Orleans, LA, Contract No. AS551-CT8-17. Boesch, D. F. 1983. Implications of oxygen depletion on the continental shelf of the northern Gulf of Mexico. Coastal Ocean Pollut. Assess. News 2:25-28. Christmas, J. Y. (editor). 1973. Cooperative Gulf of Mexico estuarine inventory and study, Mississippi: Phase I, area description, Phase II, hydrology, Phase II, sedimentology, Phase IV, biology. Gulf Coast Research Laboratory, Ocean Springs, MS, 434 p. Comiskey, C. E., and T. A. Farmer (editors). 1981. Characterization of base-line oceanography for the Tex- oma region brine disposal sites. Vol. I. Final Report to U.S. Department of Energy, Strategic Petroleum Reserve Office, Wash., D.C., Contract No. DEAC01-774508788, 130 p. Cook, H. L., and M. J. Lindner. 1970. Synopsis of biological data on the brown shrimp Penaeus aztecus aztecus Ives, 1891. InM. N. Mistakidis (editor), Pro- ceedings of the World Scientific Conference on the Biology and Culture of Shrimps and Prawns, p. 1471-1497. FAO Fish. Rep. 57(4). DOUGENIK, J. A., AND D. E. SHEEHAN. 1975. SYMAP User's Manual. Camera Stat of Bedford, Cambridge, MA, 187 p. Flowers, C. W., W. T. Miller, and J. D. Gann. 1975. Water chemistry. In J. G. Gosselink, R. H. Miller, M. Hood, and L. M. Bahr (editors), Environmental assessment of a Louisiana offshore port and appertinent pipeline and storage facility. Vol. II, App. V, Sect. 1, 86 p. Final Report to Louisiana Offshore Oil Port, New Orleans, LA. Fotheringham, N., and G. H. Weissberg. 1979. Some causes, consequences and potential environmen- tal impacts of oxygen depletion in the northern Gulf of Mex- ico. Proc. 1 1th Annu. Offshore Tech. Conf ., April 30-May 3, 1979, 3611:2205-2208. Gallaway, B. J. 1981. An ecosystem analysis of oil and gas development on the Texas-Louisiana continental shelf. U.S. Fish Wildl. Serv., Off. Biol. Serv., Wash., D.C., FWS/OBS-81-27, 88 p. Gallaway, B. J., and L. A. Reitsema. 1981. Shrimp spawning site survey. In W. B. Jackson and E. P. Wilkens (editors), Shrimp and redfish studies; Bryan Mound brine disposal site off Freeport, Texas 1979-1981. 25 FISHERY BULLETIN: VOL. 84, NO. 1 NOAA Tech. Memo. NMFS-SEFC-67, Vol. IV, 84 p. Available from National Technical Information Service, Springfield, VA 22151. Gazey, W. J., B. J. Gallaway, R. C. Fechhelm, L. R. Martin, and L. A. Reitsema. 1982. Shrimp mark release and port interview sampling survey of shrimp catch and effort with recovery of captured tagged shrimp. In W. B. Jackson (editor), Shrimp popula- tion studies: West Hackberry and Big Hill brine disposal sites off southwest Louisiana and upper Texas coasts, 1980-1982, Vol. II, 306 p. NOAA/NMFS Final Report to Department of Energy. Harper, D. E., L. D. McKinney, R. R. Salzer, and R. J. Case. 1981. The occurrence of hypoxic bottom water off the upper Texas coast and its effects on the benthic biota. Contrib. Mar. Sci. 24:53-79. Harris, A. H., J. G. Ragan, and R. H. Kilgen. 1976. Oxygen depletion on coastal waters. La. State Univ. Sea Grant Summ. Rep., Proj. No. R/BOD-1, 161 p. Klima, E. F, K. N. Baxter, F. J. Patella, and G. A. Matthews. 1983. Review of 1982 Texas closure for the shrimp fishery off Texas and Louisiana. NOAA Tech. Memo. NMFS-SEFC-108, 22 p. Available from National Technical Information Service, Springfield, VA 22151. Leming, T. D., and W. E. Stuntz. 1984. Zones of coastal hypoxia revealed by satellite scanning have implications for strategic fishing. Nature (Lond.) 310: 136-138. May, E. B. 1973. Extensive oxygen depletion in Mobile Bay, Alabama. Limnol. Oceanogr. 18:353-366. Ragan, J. G., A. H. Harris, and J. H. Green. 1978. Temperature, salinity and oxygen measurements of sur- face and bottom waters on the continental shelf off Louisiana during portions of 1975 and 1976. Nicholls State Univ., Prof. Pap. Ser. (Biol.) 3:1-29. Reitsema, L. A., B. J. Gallaway, and G. S. Lewbel. 1982. Shrimp spawning site survey. In W. B. Jackson (editor), Shrimp population studies: West Hackberry and Big Hill brine disposal sites off southwest Louisiana and upper Texas coasts, 1980-1982, Vol. IV, 88 p. NOAA/NMFS Final Report to Department of Energy. Stuntz, W E., N. Sanders, T D. Leming, K. N. Baxter, and R. M. Barazotto. 1982. Area of hypoxic bottom water found in northern Gulf of Mexico. Coastal Ocean. Climatol. News 4:37-38. Swanson, R. L., and C. J. Sindermann (editors). 1979. Oxygen depletion and associated benthic mortalities in New York Bight, 1976. NOAA Prof. Pap. No. 11, 345 p. Rockville, MD. Turner, R. E., and R. L. Allen. 1982a. Bottom water oxygen concentration in the Mississippi River Delta Bight. Contrib. Mar. Sci. 25:161-172. 1982b. Plankton respiration rates in the bottom waters of the Mississippi River Delta Bight. Contrib. Mar. Sci. 25:173-179. 26 INCIDENTAL MORTALITY OF DOLPHINS IN THE EASTERN TROPICAL PACIFIC, 1959-72 N. C. H. Lo1 and T. D. Smith2 ABSTRACT The estimates of the number of dolphins killed annually from the beginning of the U.S. tuna purse seine fishery in the eastern tropical Pacific are used by the National Marine Fisheries Service in developing management advice for the U.S. purse seine fleet. We estimated the annual number of dolphins killed incidentally in the tuna purse seine fishery for 1959-72. Kill data were available for only a few years prior to 1970. Because no obvious trend was shown with the existing data, kill rates were averaged over those years and stratified by various categories: large and small vessels, sets with large catch of tuna and small catch of tuna, sets which used backdown (a dolphin-releasing procedure), and sets which did not use backdown. These kill rates, combined with estimated number of sets, produced the estimated annual kills. Because data were available only for some of the years, they had to be pooled to obtain annual estimates. As a result, the annual estimates were highly correlated. Because the total as well as the annual estimates are of interest, it is necessary to compute the variance-covariance of the estimated annual kills. The an- nual kill from 1959 to 1972 varied from 55,000 in 1959 to 534,000 in 1961. There were three distinct maxima of 534,000, 460,000, and 467,000, corresponding to peaks in number of sets made on dolphins in 1961, 1965, and 1970. The total kill from 1959 to 1972 was estimated to be about 4.8 million, with a coefficient of variation of 17%. The eastern tropical Pacific tuna purse seine fleet began to develop rapidly in the late 1950's and has grown to over 100 U.S.-registered vessels and a substantial number of non-U.S.-registered vessels in recent years. This fleet fishes primarily for yellow- fin tuna, Thunnus albacares, and skipjack tuna, Kat- suwonus pelamis. Majority of the yellowfin tuna are taken while the tunas are schooling with dolphins primarily of the species Stenella attenuata and S. longirostris. Birds and dolphins are frequently used as cues in finding the tuna. During the capture of the tuna, some of the dolphins are killed or drown- ed by becoming tangled in the net webbing (Perrin 1969). The number of dolphins killed has been estimated to have been greater than one-half million in some of the years in the 1960's (Smith 1983). Cur- rently, fewer animals are killed each year due to im- provements in the fishing gear and in procedures to release dolphins. Estimates of the total number of dolphins killed each year in this fishery are used as a basis for management advice by the National Marine Fisheries Service (NMFS). In this paper we describe in detail the method used in Smith (1983), including estimation of the variances and covariances of the annual kill estimates so that the variance of the total kill for the period can be estimated. Additionally, we reexamine the data used in previous estimates (Per- rin 1970; Perrin and Zweifel 19713; Perrin et al. 1982; Smith 1983; Smith and Lo 1983), and we present revised estimates of the total numbers of dolphins killed. MATERIALS AND METHODS The model used to estimate the total annual in- cidental kill of dolphins (Tt) in the eastern tropical Pacific tuna purse seine fishery is Tt = RtXt (1) where t denotes the year (1959 to 1972), R denotes the number of dolphins killed per set, and X denotes the number of sets made involving dolphins. The rate of kill (R) varies between larger and smaller vessels, and in dolphin sets where fewer and greater amounts of yellowfin tuna are caught (Lo et al. 1982). In addi- tion, the rate of dolphin kills is generally less if Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, 8604 La Jolla Shores Drive, La Jolla, CA 92038. 2Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. Manuscript accepted February 1985. FISHERY BULLETIN: VOL. 84, NO. 1, 1986. 3Perrin, W. F, and J. R. Zweifel. 1971. Porpoise mortality in the eastern tropical tuna fishery in 1971. Unpubl. manuscr., 22 p. Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, 8604 La Jolla Shores Drive, La Jolla, CA 92038. 27 FISHERY BULLETIN: VOL. 84, NO. 1 backdown, a dolphin-release procedure, is used (Green et al. 1971; Barham et al. 1977; Smith and Lo 1983). lb account for these factors affecting rates of dolphin kill, Equation (1) can be reexpressed with the rates and numbers of sets stratified by vessel tuna carrying capacity, catch of fish, and use of back- down procedure: ft-ZIZ 1=1 y=i a-=i Ktijk Xfijk (2) where t i J k year 1 for vessel capacity >600 tons; 2 for vessel capacity <600 tons 1 for yellowfin tuna catch >lk ton; 2 for yellowfin tuna catch lk ton of yellowfin tuna; unsuccessful, 600 tons for unsuccessful sets. For successful sets the optimal vessel class stratification was not clear; either 400, 600, or 800 tons can be used as division points for stratification. For consistency, we adopted the same stratification used for unsuccessful sets. (The results were similar with alternative stratifica- tion schemes.) Other factors such as the age of the vessel and the experience of the captain could af- fect kill rates but were not considered in the stratification because these factors could not be isolated for analysis. The mean number of dolphins killed varied markedly over the years but without any obvious trends (Table 1). A two-way analysis of variance with the data pooled over years showed statistically significant differences in kill rates in sets made by small and large vessels (P < 0.01) and in successful and unsuccessful sets (P < 0.01). Thus Equation (2) was simplified by eliminating the time stratification for kill rates, whereas the vessel size and catch strata were retained. Few observations are available for sets where backdown was not used. In successful sets, backdown was used more than 90% of the time; thus, we have observations on kill rates in only 20 sets where back- down was not used. Thirteen of these sets were made by large vessels and seven by small vessels, and the mean kill rates within vessel size class are highly variable and not significantly different. The overall ratio of the kill rates, pooled over vessel size, when backdown was not used and when it was used is significantly greater than unity, and the annual Table 1. — Average numbers of dolphins killed (M) in purse seine sets in the eastern tropical Pacific by year, for small and large vessels making successful (>1A ton tuna) and unsuccessful (<1A ton tuna) net sets. Standard deviation (SD), number of sets (A/), and number of trips are given. Successful sets Unsuccessful sets and No. of No. of year M SD N trips M SD N trips Data source Small vessels (<600 tons 1964 60 47 1965 26 28 1968 130 114 1971 117 180 carrying capacity) 20 1 35 1 13 1 19 3 60 3 4 13 8 4 10 1 11 2 3 1 1 1 2 Smith and Lo (1983)1 Smith and Lo (1983) Smith and Lo (1983) Unpubl. NMFS 1972 57 110 103 6 4 10 16 5 Unpubl. NMFS Total 62 108 190 12 6 13 33 10 Large vessels (>600 tons 1971 41 56 carrying capacity) 16 2 0 Unpubl. NMFS 1972 37 123 117 6 0.4 1.4 12 5 Unpubl. NMFS Total 37 119 133 8 0.4 1.4 12 5 'From table 5 of Smith and Lo (1983), omitting incomplete data collected in 1966. 28 LO and SMITH: INCIDENTIAL MORTALITY OF DOLPHINS ratios vary without a consistent trend over time (Table 2). In unsuccessful sets the use of the backdown pro- cedure was more variable because the conditions of the set are more diverse For example, only a few or no dolphins may be captured, and the net may not be retrieved in the usual manner. Because of this diversity and because so few observations are avail- able, we consider one kill rate for all unsuccessful sets. Reexpressing Equation (2) to account for a constant ratio of kill rates for successful sets when backdown was used and when it was not used, and for no difference in kill rates for unsuccessful sets, yields Mil i=i y_i k-1 ■K'ijk-X-tijk = 2. l^nll V^till + CXH12) + R.i2.Xtl2.) (3) 1=1 where C = R..l2IR,.n and the subscript . is used when that stratifying variable is not considered. For example, R.^ is the kill-per-set not stratified by year t, and XH2. is the total number of sets not stratified by use of backdown. Estimates of the total number of sets involving dolphins from 1959 to 1972. with approximate variances, are given by Punsly (1983). He also gives partial estimates of the numbers of successful and unsuccessful sets, but does not provide estimates of the numbers of sets by vessel size Punsly's data did not indicate the use of the backdown procedure The coefficients of variation (CV) of Punsly's estimates are <1% in all years except 1959 and 1960, when it was 8%. The percentage of unidentified sets in 1959-61 was higher than subsequent years because set type was not recorded systematically (Hammond4). We assume these estimates are in fact constants, because in most years, and in the absence of additional information in 1959-61, the CVs are small compared with the CVs of the kill rates (0.13-1.0, Table 1). By applying the proportions of successful and un- successful dolphin sets from Punsly's partial estimates to his totals, we obtained numbers of suc- cessful and unsuccessful dolphin sets. We further prorate these estimated numbers of successful and unsuccessful sets to large and small vessels by multiplying by the estimates of proportions from NMFS (Anonymous 19765) of sets made by vessels of each size class (Table 3). The slight differences between the totals for each year given by Punsly are due to rounding. The number of sets during which backdown was used can be estimated from the estimated total number of sets involving dolphins (Table 3) and the observed proportion of successful sets in which back- down was used (Table 2). The observed proportions increase from 0.79 in 1964-65 to almost unity (0.96) by 1972. The backdown procedure was reportedly 4P. S. Hammond, Sea Mammal Research Unit, British Antarctic Survey, Cambridge, England, pers. commun. 1983. 6Anonymous. 1976. Report of the workshop on stock assess- ment of porpoises involved in the eastern Pacific yellowfin tuna fishery (La Jolla, July 27-31, 1976). Southwest Fish. Cent., Ad- min. Rep. LJ-76-29, 54 p. + app. Table 2. — Mean number of dolphins killed (R) during purse seine sets in the eastern tropical Pacific Ocean when the backdown dolphin-release procedure was and was not used. Also given are the ratio of numbers killed with and without backdown (C), the proportion of suc- cessful sets where backdown was used (P), the number of sets (A/), number of trips, and stan- dard error in parentheses. Backdown used Yes No No. of No. of Year "mi N trips °M2 N trips C P 1964' 44 16 1 128 4 1 3.0 0.79 19651 48 6 1 24 2 1 0.50 19661,2 — 17 1 — 2 1 — 0.89 19681 142 11 1 92 1 1 0.65 19713 81 30 5 111 4 3 1.40 19723 41 193 12 169 9 6 4.10 0.96 Total 50 256 21 131 20 12 2.62 "(0.80) 0.93 1From Smith and Lo (1983). 2Kill rates tor 1966 omitted because incomplete data were collected. 3NMFS records. 600 tons) vessels, and for successful (>1/» tons tuna) and unsuccessful (<1/4 tons) sets, modified from Punsly (1983). Successful sets Unsuccessful sets small large small large Year PW (*m.) (*f22») (*fl2.) 1959 326 0 265 0 1960 3,170 0 2,303 0 1961 3,888 32 3,928 0 1962 1,773 5 1,942 19 1963 2,291 10 2,092 23 1964 4,444 45 3,089 64 1965 5,346 27 2,418 29 1966 4,948 44 1,835 25 1967 3,363 2 841 3 1968 2,956 175 982 41 1969 5,365 1,401 1,402 192 1970 4,936 2,313 957 412 1971 1,871 2,602 652 409 1972 2,704 4,982 855 846 developed on one vessel in 1959-60 (Barham et al. 1977) and used by at least three vessels in 1961 (Anonymous 1962). If 79% of the sets in 1964-65 were made using this procedure, as suggested by the very limited available data, a rather rapid increase in usage must have occurred in 1962 and 1963. This is possible because, if properly used, the procedure reduces the amount of handling time of dead dolphins, thus speeding up the fishing operation. As an approximation, we assume that usage increased from 0 to 0.79 linearly from 1959 to 1964-65, and was 0.89 for 1966-71 and 0.96 for 1972. Denoting the interpolated and extrapolated esti- mates of the proportion of successful sets using the backdown dolphin release procedure by Pt gives Xtill - Pistil, Xtii2 = (1 - Pt) Xti\»- Substituting these relationships into Equation (3), with the assumption that the estimated numbers of sets given by Punsley (1983) are constants, the following equations result when the terms are rearranged: Tt = X {P'in[xtii*Pt + C(l - PtV^tii'] + P'i2»Xti2»} i = Z \R'illXtil»[Pt + C(l - PJ\ + P'i2'Xti2.\. i (4) The time series of estimated annual kill (tt) from 1959 to 1972 was obtained by pooling the available data over years and strata, resulting in estimates that are not statistically independent. Thus in order to estimate the variance of the total kill of dolphins for the period in addition to the variances it is necessary to determine the covariances among the annual estimates. We denote the estimates of the total kill of dolphins (ft) for each year from 1959 to 1972 by the vector f, and denote the estimates of the variances of the elements of f by the symmetric matrix If. The estimate of the kill in each year (Equation (4)) can be expressed in matrix form as the product of a vec- tor of the numbers of sets in each of the four com- binations of the vessel size and fishing success classifications (Xt), and a vector of the four corre- sponding kill rates (Qf). Each element of T then can be expressed as a matrix product Tt = X\ Qt (5) where X't = (Xtn„ Xm„ Xnz„ XtZ2.) Qt = Qn Qt2 Qt3 Qa R.in[Pt(i -Q + C] R.2n[Pt(l -Q + C] R»\2» R. 22. P'inft P»21lft K*\2* R*22* and /, = P,(l - 6) + C. Then the variance-covariance matrix of T is 30 LO and SMITH: INCIDENTAL MORTALITY OF DOLPHINS Zr = V(T59) Cov(T59, f60) V(f 60) Cov(f 59, f 72) Cov(f60, f72) . . . V(f 72) V(X'59 Q59) Cov(Z'59 Q59, X'60 Q60) V(X'60 Q60) Cov(X'59 Q59, X'12 Q72) Cov(X'60 Q60, X'72 Q72) . . . V(X'72 Q72) with VCfy = X'flQtXt as the diagonal elements of If (6) where 1qt = V(R.inft)Cov{R.ulft,R.2nft) 0 0 V&.211 ft) o o ^.21.) 0 V(R.22.) (7) The diagonal elements of If can be computed by noting that R,i2. is uncorrelated with R.in, Pt, or C, and the covariance of Pt and C is zero because one C value is used for all years in 1959-72 and Pt can be different between years. The off-diagonal elements of If are Cov(t „ tj = CovPT'A, X'mQm) 4 4 = 11^ Cov(QMJ, Qmj) X„ i=i i=i (8) ,mj) ■"■my Expressions for each of the terms in If are given in the Appendix. RESULTS AND DISCUSSION The estimates of the total number of dolphins killed incidentally in the tuna purse seine fishery from 1959 to 1972 (Table 4, from Equation (4)) vary from a low of 55,000 in 1959 to a high of 534,000 in 1961. Three distinct maxima of 534,000, 460,000, and 467,000 are apparent (Fig. 1), corresponding to peaks in numbers of sets made on dolphins in 1961, 1965, and 1970 (Table 3). A total of about 4.8 million dolphins is estimated to have been killed in the whole period (Table 4). The CVs of the annual estimates decline rapidly Table 4.— Estimated number of dolphins killed by year (Equation (4)), with standard errors (SE) and coefficient of variations (CV). Year Number killed SE CV 1959 55,000 18 0.32 1960 478,000 146 0.31 1961 534,000 149 0.28 1962 216,000 54 0.25 1963 240,000 54 0.22 1964 390,000 77 0.20 1965 460,000 92 0.20 1966 374,000 58 0.15 1967 257,000 39 0.16 1968 229,000 35 0.15 1969 461,000 68 0.15 1970 467,000 70 0.15 1971 254,000 43 0.17 1972 380,000 61 0.16 1959-72 4,790,000 857 0.18 31 FISHERY BULLETIN: VOL. 84, NO. 1 800 r 700 - 600 c a M 3 o £ 500 - z 400 O a "- 300 O K Ul CD 2 z 200 100 J l_ _l_ _l_ JL _1_ 1959 60 61 62 63 64 65 66 67 68 69 70 71 72 YEAR Figure 1.— Estimated numbers of dolphins killed in the east- ern tropical Pacific tuna purse seine fishery from 1959 to 1972. Standard errors of the estimates shown as vertical bars. From Table 4. from 32% in 1959 to 15% from 1966 to 1970, and then increase only slightly in 1971 and 1972. The covariances are large (upper triangular matrix, Table 5). They are all positives, and tend to be smaller for pairs of estimates widely spaced in time The covariances can be examined more easily in terms of correlation coefficients (lower triangular matrix, Table 5). The correlations range from 0.31 to 0.99. The CV of the estimated total is 18%. This is substantially higher than the corresponding value of 6% obtained when the covariances are ignored. Because the total is the sum of 14 numbers, an ap- proximate 95% confidence interval, obtained by add- ing and subtracting two standard errors, is 3.1-6.5 million dolphins. The variation in the estimated numbers of dolphins killed over the period 1959-72 is due to several fac- tors: 1) The number of sets made involving dolphins varied from year to year depending on the number of sets of tuna schooling in the absence of dolphins; such tuna are apparently preferred when available 2) The use of the backdown dolphin-release pro- cedure increased rapidly from 1959 to 1964. How- ever, the development of the backdown dolphin- release procedure is not well known. The available data reflect the tendency of captains to use the technique once it was known. There is little infor- mation on how rapidly the procedure became known to other captains and no information on how rapid- ly they learned to use it effectively. Our assumption of a linear increase probably overestimates the use of backdown initially, but may or may not overesti- mate its subsequent use 3) The proportion of suc- cessful sets made by small vessels increased from about 50% from 1959 to 1964, to >75% from 1965 to 1972 (Table 1). The higher dolphin kill rate for suc- cessful sets results in an increase in estimated dolphin kills as the proportion of successful sets in- creased. 4) The increase in the proportion of sets which were made by large vessels starting in 1968 results in a decrease in estimated dolphin kill rates due to the lower dolphin kill rate of these vessels. Several factors which may have affected the numbers of dolphins killed in this period have not been accounted for because of the assumptions made by incomplete data. Chief among these assumptions were 1) the relatively small samples are represen- tative of the fleet as a whole 2) the kill rates on un- successful sets are not affected by the use of back- down, 3) the ratio of kill-per-set in successful sets without backdown to that with backdown is constant Table 5.- -Covariances (upper triangular matrix, x1010) and correlation coefficients (lower triangular matrix) for the estimated total dolphins killed by year, from 1959 to 1972. 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1959 0.41 0.42 0.14 0.13 0.16 0.15 0.11 0.07 0.06 0.13 0.13 0.06 0.04 1960 0.99 3.49 1.24 1.15 1.37 1.32 0.92 0.62 0.56 1.10 1.08 0.54 0.40 1961 0.99 0.99 1.27 1.19 1.45 1.38 0.95 0.64 0.57 1.13 1.11 0.55 0.42 1962 0.97 0.98 0.99 0.44 0.55 0.52 0.34 0.23 0.21 0.41 0.40 0.19 0.15 1963 0.92 0.94 0.96 0.98 0.58 0.53 0.33 0.22 0.20 0.39 0.38 0.18 0.15 1964 0.76 0.79 0.83 0.88 0.95 0.75 0.43 0.29 0.26 0.50 0.49 0.23 0.20 1965 0.79 0.81 0.84 0.88 0.92 0.93 0.56 0.34 0.27 0.51 0.50 0.23 0.20 1966 0.65 0.66 0.67 0.67 0.67 0.62 0.86 0.30 0.21 0.39 0.38 0.17 0.16 1967 0.75 0.76 0.77 0.78 0.78 0.71 0.89 0.93 0.16 0.30 0.29 0.13 0.10 1968 0.74 0.75 0.76 0.77 0.76 0.70 0.77 0.69 0.90 0.31 0.30 0.14 0.10 1969 0.73 0.74 0.75 0.76 0.75 0.68 0.75 0.66 0.87 0.98 0.63 0.33 0.27 1970 0.71 0.72 0.73 0.73 0.72 0.65 0.70 0.62 0.83 0.94 0.98 0.36 0.38 1971 0.58 0.59 0.59 0.59 0.57 0.50 0.54 0.46 0.63 0.75 0.85 0.92 0.25 1972 0.34 0.35 0.36 0.36 0.36 0.34 0.37 0.34 0.39 0.43 0.55 0.65 0.83 32 LO and SMITH: INCIDENTIAL MORTALITY OF DOLPHINS for both large and small vessels for all years, and 4) the kill rate itself for sets with backdown did not change over the years. Although each of the unaccounted for factors could have an effect on the estimated numbers of dolphins killed, the magnitude of such effects is probably smaller than the magnitude of the effects of vessel size, set success, and use of backdown described in this study. For example, although the kill rate data available are few, there are some additional data which are not available to us, but which are reported- ly similar (Smith and Lo 1983). The last three assumptions noted above deal with the dolphin kill rates with and without backdown, and would tend to both increase and decrease the estimates, if they could be taken into account. Our estimates of the total number of dolphins killed (Table 4) are slightly lower than previous esti- mates made using the same method (Smith 19796, 1983). The previously estimated total number of dolphins killed from 1959 to 1972 was 5.1 million (total of Smith's [1983] table 4, divided by 0.96 for other species and by 1.048 for injured animals). The difference between the two estimates resulted from the revision of the estimated number of sets that cap- ture tuna associated with dolphins (Punsly 1983) and of the numbers of dolphins killed per set (Smith and Lo 1983). There are alternate approaches to estimating the numbers of dolphin killed. For example, estimates could be made from data on the numbers of fishing trips made (kill-per-trip), or the number of tons of tuna caught (kill-per-ton). These approaches make different assumptions about the fishing process (Lo et al. 1982; Hammond and Tsai 1983), and require data which are not as precise as are data on the total numbers of sets. For example, fishing trips are dif- ficult to count consistently because they may not be completed within the calendar year and may be ex- 6Smith, T. D. (editor). 1979. Report of the Status of Porpoise Stocks Workshop, August 27-31, 1979, Southwest Fisheries Center, La Jolla, California Southwest Fish. Cent., Admin. Rep. LJ-79-41, 120 p. tended by partial unloading of the catch. There are fewer such problems with the data for kill-per-set estimators on the number of dolphins killed, and the problems that exist have already been resolved (Punsley 1983). LITERATURE CITED Anonymous. 1962. How tuna seining paid off for the U.S. fleet in 1961. Fish Boat, Feb., p. 19-30. Barham, E. G., W. K. Taguchi, and S. B. Reilly. 1977. Porpoise rescue methods in the yellowfin purse seine fishery and the importance of Medina panel mesh size Mar. Fish. Rev. 39(5): 1-10. Green, R. E., W. F. Perrin, and B. P. Petrich. 1971. The American tuna purse seine fishery. In Hilmar Kristjonsson (editor), Modern Fishing Gear of the World, Vol. 3, p. 182-194. Fish. News (Books) Ltd., Lond. Hammond, P. S., and K. T. Tsai. 1983. Dolphin mortality incidental to purse-seining for tunas in the eastern Pacific Ocean, 1979-81. Rep. Int. Whaling Comm. 33:589-597. Lo, N. C. H., J. Powers, and B. E. Wahlen. 1982. Estimating and monitoring incidental dolphin mortality in the eastern tropical Pacific tuna purse seine fishery. Fish. Bull, U.S. 80:396-401. Perrin, W. F. 1969. Using porpoise to catch tuna. World Fishing 18(6): 42-45. 1969. The problem of porpoise mortality in the U.S. tropical tuna fishery. Proceedings of the 6th Annual Conference on Biology, Sonar, and Diving Mammals, p. 45-48. Stanford Research Institute Perrin, W. F, T. D Smith, and G. T. Sakagawa. 1982. Status of populations of spotted dolphin, Stenella at- tenuate/,, and spinner dolphin, S. longirostris, in the eastern tropical Pacific In FAO, Mammals in the seas, Vol. IV. Small cetaceans, seals, sirenians, and otters, p. 67-83. Punsly, R. G. 1983. Estimation of the number of purse-seiner sets on tuna associated with dolphins in the eastern Pacific Ocean dur- ing 1959-1980. Inter-Am. Trop. Tuna Comm. Bull. 18:229- 299. Smith, T D. 1983. Changes in size of three dolphin (Stenella spp.) popula- tions in the eastern tropical Pacific Fish. Bull., U.S. 81:1-14. Smith, T. D, and N. C. H. Lo. 1983. Some data on dolphin mortality in the eastern tropical Pacific tuna purse seine fishery prior to 1970. U.S. Dep. Commer., NOAA Tech. Memo. SWFC-TM-NMFS-34, 26 p. 33 FISHERY BULLETIN: VOL. 84, NO. 1 APPENDIX In Equation (7), the first and second terms on the main diagonal are V(R.tllft) = V(R.lU)V(f<) + R2.mV(ft) + fMR.m) (A-l) for i = 1 and 2, noting that Cov(R.!Uft) = 0. The variance of ft is given by V(A) = V(Pt) (1 + V(Q) + Pf V(Q (A-2) + &V(Pt) + V(C) - 2V(P,)C - 2V(C)Pf + 2 Cov(Pt, Q. This last term is assumed to be zero, as noted above. The off-diagonal element in Equation (7) is Cov(R.lUft, R.jnft) = R.m R.JU V(ft) (A-3) for i ¥= j = 1 and 2. In Equation (8), based upon Equation (5) Cov(Qm, Qmj) = CovtR.,! fui R.fl i/J for i = and j = - 1,2 = 1,2 0 i ±j for i = = 3,4 VCR.*,.) i = j and j = = 3,4 where Cov(R.m fu, R.ju fm) (A-4) [R%u + V(R.jU)]Cov(fu, fm) + fufmV(R.m) i = j R-iuR.ju Cov(/M, /J i # j assuming Cov(R.ai, R.jn) = 0 and Cov(/M, /J = Cov(PM, Pm) [V(Q + C2] (A-5) + V(Q.[1 + AA -Pu ~Pml 34 THE ABUNDANCE AND DISTRIBUTION OF THE FAMILY MACROURIDAE (PISCES: GADIFORMES) IN THE NORFOLK CANYON AREA1 Robert W. Middleton2 and John A. Musick3 ABSTRACT The Norfolk Canyon off Virginia and the adjacent slope areas were sampled with 13.7 m otter trawls in June 1973, November 1974, September 1975, and January 1976. Trawl depths ranged from 75 to 3,083 m, and 22 species of macrourids were captured during the study. Coryphaenoides rupestris demonstrated seasonal movement to shallower water (ca. 750 m) during winter. Nezumia bairdii, N. aequalis, and Cory- phaenoides carapinus exhibited a significant positive correlation between head length and depth (r2 = 0.47, 0.37, and 0.35, respectively). Nezumia bairdii apparently spawns in July or August, and reaches an age of about 11 years. New size records were established for Nezumia aequalis (64 mm head length (HL)) and N. bairdii (66 mm HL). New depth records were established for Coelorinchus c. carminatus and N. aequalis (884 and 1,109 m, respectively). The known geographic ranges for Coelorinchus carib- beus, C. occa, Nezumia cyrano, Coryphaenoides colon, Hymenocephalus gracilis, H. italicus, Bathygadus macrops, Macrourus berglax, and Gadomus dispar were extended to the Norfolk Canyon area. The Macrouridae (Pisces: Gadiformes) includes some of the most abundant archibenthic deep-sea fish species (Marshall 1965, 1971; Marshall and Iwamoto 1973; Iwamoto and Stein 1974) and attains greatest abundance and diversity on the continental slopes of the world oceans (Marshall and Iwamoto 1973). Present knowledge of the life history and ecology of macrourids has been accrued piecemeal from faunal lists and taxonomic works (Gunnerus 1765; Gunther 1887; Gilbert and Hubbs 1920; Farron 1924; Iwamoto 1970; Okamura 1970; Marshall and Iwamoto 1973; Iwamoto and Stein 1974), or from studies on physiology, anatomy, and life history (Kulikova 1957; Marshall 1965; Phleger 1971; Ran- nou 1975; Rannou and Thiriot-Quiereaux 1975; Haedrich and Polloni 1976; McLellan 1977; Merrett 1978; Smith et al. 1979). The meager literature on reproduction and growth of macrourids and other deep-sea anacanthine fishes has recently been reviewed by Gordon (1979). With the advent of in- creasing expertise in deepwater trawling, some macrourid species, such as Coryphaenoides rwpestris and Macrourus berglax, have become commercially •Contribution No. 1226 from the Virginia Institute of Marine Science. 2Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, VA 23062; present address: Minerals Management Service, U.S. Department of the Interior, 1951 Kidwell Drive, Vienna, VA 22180. 3Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, VA 23062 important in the western North Atlantic. Experi- mental commercial trawling was initiated by the Soviet Union in 1962, and many studies directly related to the commercial fishing of macrourids have been subsequently published by Soviet workers (Podrazhanskaya 1967, 1971; Sawatimskii 1971, 1972; Grigor'ev 1972) and to a lesser extent by Polish researchers (Stanek 1971; Nodzinski and Zukowski 1971). The present study examines the seasonal distribu- tion and abundance of the macrourid species cap- tured in the Norfolk Canyon area. In addition aspects of age, growth, and reproduction of selected domi- nant species are also described. MATERIALS AND METHODS Gear The data presented in this paper were obtained on four cruises to Norfolk Canyon and the adjacent open slope to the south (Fig. 1) conducted by the RV Columbus Iselin (June 1973) and RV James M. Gillis (November 1974, September 1975, January 1976). On all cruises a 13.7 m semiballoon otter trawl with 1.3 cm (stretched) mesh in the cod end liner and 5.1 cm (stretched) mesh in the wings and body was employed. Steel "china V" doors at the end of 22.9 m bridles were used to permit spreading of the net from a single warp (Musick et al. 1975). Manuscript accepted March 1985. FISHERY RIILLETIN: VOL. 84. No. 1. 1986. 35 FISHERY BULLETIN: VOL. 84, NO. 1 1 L, 1 -1 , " i i ■T 1 / ,j * ■• 37*30' i / ■; 1 / ' ; *** ) / / / i i * :S~ - / / > .— •*■ / / i ! ' «» — J t / / \ 1 A- * ' rf», . _"-'' NORFOLK • . li O: ftl; * — +-*♦ ■ " *-«-r~,«» CANYON • 37*00' 2-' t a V,'X. 'ex c <• • • £:•'"•' • O ,* / • 8 1 I00> c / I- }o/ _ < n; 6/ mi • /Oi • *, • / ■ fOl •,' / '*. • \ J I N OV-. • ,'''' i / / $v ■ •••;• • • / m • V • s / 1 ;> • <• < !\ 1* J 36*30' j i i • iff 2> \ •; II '151 / 1 . > i ' : ■ ■ i ■ ■ i 75*00' 74*30' 74*00' Figure 1— Map of the Norfolk Canyon study area with station 73*30' locations indicated. 73*00' Sampling Design Norfolk Canyon and an adjacent open slope were divided into five sampling strata: 75-150 m, 151-400 m, 401-1,000 m, 1,001-2,000 m, and 2,001-3,000 m. Six stations were then randomly assigned in each depth stratum. The duration of all tows in depths of <2,000 m was 0.5 h (bottom time). Where the depth exceeded 2,000 m, the tow times were ex- tended to 1 h. Station depth was determined from a sonic precision depth recorder when the net was set and then every 3 min for the duration of the 0.5 h tows (every 6 min for the 1-h tows). Mean station depth was then determined by averaging the 11 resultant values. Data Collection and Analysis Head lengths instead of total lengths were measured because macrourids have slender whiplike tails that are easily damaged during trawling. The head length (HL) was measured to the closest millimeter, from the tip of the snout to the posterior edge of the opercle using Helios4 dial calipers. The fish were weighed with an Ohaus dial-a-gram scale Calibration showed the scale to be accurate within ■•Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 1.0-1.5 g under all typical shipboard conditions. The sex and gonadal conditions of freshly captured specimens were noted. Gonadal samples for histo- logical processing were stored in Davidson's preser- vative and later mounted using standard paraffin techniques. Sections (5 mm) were stained with Mayer's hematoxylin and eosin counterstain. Saccular otoliths and a scale sample were removed from all Nezumia bairdii and stored dry. Represen- tative otolith samples were chosen randomly from individuals over the entire size range of fish captured. The length-weight relationships for Nezumia bair- dii, Coryphaenoides armatus, and C. rupestris were analyzed using log transformed weights regressed against head length (Fig. 2). Regression analysis of head length on depth of cap- ture was performed for each species to determine any significant change in head length with change in depth. Testing of the hypothesis that fi = 0 for the regression line ascertained whether there was a significant change of size with changing depth. The coefficient of determination (r2) was also calculated to determine what proportion of the variance of head length could be attributed to change in depth. The a posteriori Student-Newman-Keuls analysis of means was used as a second method for inter- preting the size/depth relationship. This method calculated the mean depth of capture of each head- 36 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE length interval, combined the head lengths in subsets whose mean depths did not differ significantly from each other, and defined the constituents of each subset. Due to the large size and thickness of the macrou- rid otoliths, standard age determination techniques proved unsuccessful (Christensen 1964; McEachran and Davis 1970). Therefore, a thin section was re- moved from each otolith and, using a dissecting microscope, the number of bands presumed to be an- nual were counted and recorded. Gonads of the specimens were classified into repro- ductive stages for analysis. The criteria for these stages were as follows: Stage 1— Undeveloped. The gonads were immature and no development was evident. The reproduc- tive organs were difficult to distinguish within the body cavity. Stage 2— Early Immature The reproductive organs had enlarged slightly. The sex could be determined, but no vascularization of the ovaries was apparent. The organs of both sexes had a highly translucent appearance. Stage 3— Immature The ovaries were enlarged and vascularization had begun. The testes had become discernibly "sausage shaped". The organs of both sexes were opaque. Stage 4— Late Immature The reproductive organs of both sexes were full size The ovaries were about 90% vascularized. The testes had become milky white in color. Stage 5— Mature The reproductive organs were developed completely. Ovaries were fully vascular- ized and had a granular appearance. Stage 6— Ripe. Advanced spermatogenesis or oogenesis was evident. The oocytes were fully developed in the females and the male testes con- tained milky-white seminal fluid. Stage 7— Spent. The testes and ovaries were spent. The reproductive organs were flaccid and had recently released sperm or eggs. RESULTS AND DISCUSSION Species Accounts Coelorinchus c. carminatus (Goode 1880) Coelorinchus c. carminatus is a relatively shallow water macrourid reported from depths of 89-849 m (Marshall and Iwamoto 1973). In the study area this species was captured in depths of 210-884 m (Fig. 3). Marshall and Iwamoto (1973) reported C. c. car- minatus from northern Brazil to the Grand Banks, but absent in the Bahama Island chain. The largest specimen captured in our study had a head length of 70 mm, while Marshall and Iwamoto (1973) reported specimens with 73 mm HL. During our study, a maximum of 188 individuals and 4 kg of C. c. carminatus were captured in a 0.5-h trawl. This species also contributed as much as 34.2% of the number and 27.8% of the biomass of benthic fishes captured in individual samples. Figure 4 shows the depth distribution of C. c. car- minatus incremented by 2 mm size groups. A slight increase in head length with increase of depth was apparent. The slopes of the regression lines were shown to be significantly different from zero. The coefficient of determination (Table 1) also showed a correlation between head length and depth. There was variability among cruises, but this may be at- _ 2 S o o"n = 305 ? n = 422 Coryphaenoides armatus -i 1 1 r -i r ~r d*n = 156 $ n = 279 Coryphaenoides rupestns 0 10 20 30 40 SO 60 O 10 20 30 40 SO 60 70 80 90 100 0 HEADLENGTH (mm) -i — i — i — i — i — 30 50 100 -I ISO 200 Figure 2.— The log (wt) versus head length regressions for Nezumia bairdii, Coryphaenoides armatus, and Coryphaenoides rupestris. 37 FISHERY BULLETIN: VOL. 84, NO. 1 Figure 3— Minimum and maximum depth of capture, with minimum, maximum, and modal temperatures of capture for each species and each cruise A Coelonnchus C carminatus <* A* A*3 Afc D Nezumia bairdii fi ,* .o* . f> j* v5 A«, SN Depth M'n 'm > Man 2 52 260 210 226 ..Min Depth (m.) Moi 270 315 277 310 776 750 828 884 1350 1525 1644 1470 _ Mm Temp (°C) Ma, Mode 47 49 43 45 _ Mm Temp t°C) Ma< Mode 4 1 4 1 37 37 II 3 106 II 1 110 IQO 96 92 7. 1 90 7 4 8 7 7 8 5 1 5.1 4.4 53 w Nezumia aequalis >° o* o* Jo jo' 0s U Coryphaenoides rupestris <5 A» J? Jo Depth (m) Moi 367 578 330 452 Depth Min tmj Man 636 578 616 828 986 912 1109 884 1591 1525 1108 1698 Mm Temp (°C) Moi Mode 45 45 43 4 3 Mm Temp (°C> Mo. Mode 4 1 45 37 4 1 57 7.1 78 80 49 57 43 5.0 5 1 62 4 5 49 46 5.3. 4.0 48 t Coryphaenoides carapinus ,-e o" ,o* ,os f> J* *{3 A«= i Coryphaenoides armatus ,n° S>* ,o* . ... Min Depth (m) Mat 1194 1403 1189 1108 Depth Min (m) Mai 2100 2257 2250 1876 2642 2767 2679 2642 3083 2920 Mm Temp (°C) Ma. Mode 3.5 2 9 2 5 2 9 Mm Temp (°C) Mo. Mode 2.5 2 3 24 4.1 4 2 4 0 3 9 3.3 2 8 3 2 3 8 3 9 3 7 3 8 29 24 28 Table 1.— The coefficient of determination (r2) for the change in head length with change in depth regression lines. Cru ses Jan. June Sept. Nov. Combined Species 76-01 73-10 75-08 74-04 cruises Coelorinchus c. carminatus 0.006 0.23 0.13 0.44 0.23 Nezumia aequalis 0.45 0.15 0.62 0.14 0.37 Nezumia bairdii 0.12 0.50 0.44 0.49 0.47 Coryphaenoides rupestris 0.04 0.19 0.08 0.11 0.02 C. carapinus 0.59 0.005 0.30 0.37 0.35 C. armatus 0.000 — 0.05 0.000 0.14 tributed to sampling artifacts and the relatively nar- row depth range (674 m) of this species. The analysis of variance showed a significant dif- ference in mean depths of the head length groups (F = 35.9, F(table; a . 0.01) = 1.79). The Student- Newman-Keuls test divided the group into two significantly different subsets; one 10-50 mm HL and the other 51-70 mm HL. Other macrourids (N. bairdii and N. aequalis) had high biomass but low numerical abundance at the deep end of their ranges, indicating the presence of a few large specimens there This was not the case for C. c. carminatus (Fig. 5). The occurrence of fish distributing by size can be obscured if the larger members of the population traverse the entire range The biomass of the species would be elevated at the shallower depths so that a consistent biomass level is present throughout the depth range Comparison of Figure 4 with Figure 5 shows that although the mean depth of capture for this species increased with head length, the larger fish occurred over the en- tire depth range This pattern is important because it shows that for some fishes the "bigger-deeper" phenomenon described by Polloni et al. (1979) may really be a "smaller-shallower" phenomenon. A plot of mean fish weight against depth as used by Polloni 38 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE UJ 3 00 ~3 in CM o 1 ii II ro c «- ( I I I I I I I I I I I I I I I I I I I I I I I I I I I I I OOOOOOOOOOQOOO o oooooooooooooo o SCDsO*CviO«,sO»pJOoOi*» CM NNNNN"--"- 0- UJ o o ■CO o in o o o > o z o I *■ cm OJ — II II c — _o I I I I I I I I I I I I I I I I I I I oooooooooc oo ooo ooo O CO <0 * (SJ O 00 10 fO (SJ (VI (SI W (SI — — o I I I I I I I I I I 1 1 ° ooo o o in o CM -o E E X I- o z UJ -I o < UJ X o <0 8 (Si X H 0. UJ Q c 5 -a c CS ca c a) CD c CD c 2 M o "o c CD u CtJ CD w co 03 _ < S S ~ s- CD ~ -w O I 1 < CD -8 m £ cm o MM " ir> I •-< i a — i o •CH CBK. 4B4 _o MO •O- cn _o (SI II c II rr i i i i ' i i i i i i" I 8 § 8 § | § 1 e * ♦ N O • » N N N N N - I I | I I I I I I I I I I I ' o o o o o o o o o 5 5 o o o <, v o _l cr UJ CD r> 2 Coelorinchus c. carminatus o.u - o O - 73-10 = 74-04 o * A - 75-08 2.0- o°AA • X a o X D*ofi D D X - 76-01 1.0- A X rP° o 0 x o o X Ao e o A o o D A 1 ! A 1 I o D 1 r i X X T 1 4.0-1 3 0- CD UJ ^ o 2.0- 1.0- Cs c£ □ xa o° D Ao % a A o o a a ° o — I 1 1 1 1 r- 100 200 300 400 500 600 DEPTH (m ) — j— 700 800 900 1000 Figure 5— The distribution of log transformed (log (x + 1)) abundance and weight of Coelorinchus c. carminatus at each station, plotted against depth. et al. (1979) may have a highly positive slope, but these data are impossible to interpret without infor- mation about length-frequency patterns with depth. The temperatures at which C. c. carminatus were captured varied from 4.3° to 11.3°C (Fig. 6). The average temperature of collection was 7.6 °C. Nezumia aequalis (Gunther 1878) Nezumia aequalis is a closely related congener of 40 N. bairdii and is found primarily south of the study area (Marshall and Iwamoto 1973). Nezumia ae- qualis attains a head length of at least 53 mm and has a depth distribution of 200-1,000 m. Its Figure 6— The temperature range for each species, by cruise Th dot designates the modal temperature, Ccc. - Coelorinchus carminatus, N.b. - Nezumia bairdii, N.a. - Nezumia aequalis, C.i - Coryphaenoides rupestris, Gc. - Coryphaenoides carapinus, C.s - Coryphaenoides armatus. MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE 12 I I JANUARY 76-01 10 H 9 8- 7- 6- 2- I - 1 T i i T -L i 1 1 1 1 1 C.cc N.b. N.o C.r Cc Co SPECIES 12 - JUNE 73-10 II - 10 - 9 - i > 8- 7- 6- 5- 4- • - 1 J 1 3- 2- 1 - 0- C.cc Mb. N.O. C.r Cc. Co. SPECIES 12 II - 10 - 9 - 8- 7 6- 5- 4 - 3- 2 I- SEPTEMBER 75-08 XI — >^ «^ — C.cc N. b — I — Nfl — r~ C.r - J— Cc - 1 Co 12-1 II 10 - 9- 8- 7- 6- NOVEMBER 74-04 3- I - 5- -L 4- 1 — Ccc T 1 l l T T X — i — Nk. - J— No — r- c# — T— Cc Co. SPECIES SPECIES 41 FISHERY BULLETIN: VOL. 84, NO. 1 geographic range is listed as from the Faroe bank to northern Angola in the eastern Atlantic, the Mediterranean, and from Davis Straits to northern Brazil in the western Atlantic (Marshall and Iwamoto 1973). In the Norfolk Canyon area the depth of capture of N. aequalis was from 330 to 1,109 m. The greatest number in a trawl was 40 in November of 1974, and the highest biomass per trawl was 300 g in Septem- ber 1975. Nezumia aequalis comprised up to 8.9% of a trawl catch by number and 3.1% by weight. The analysis of variance of the mean depths of the head length groups gave a F value of 3.32 (F(table; a = 0.01) = 2.11). The Student-Newman-Keuls analysis showed only one subset, probably because of the low sample size Examination of Figure 7 suggests head length increased with depth, and the slope of the line was significantly different from zero. Although its bathymetric range was extensively sampled, densities were low and few mature speci- mens were captured (Fig. 8). These findings are in contrast to the distribution and abundance of its cogener, N. bairdii, suggesting competitive exclu- sion. Alternately, Norfolk Canyon populations ofN. aequalis may represent expatriation from denser populations in the Gulf of Mexico or on the Blake Plateau. The temperature range for N. aequalis captured in the Norfolk Canyon area was from 4.3° to 8.0 °C (Fig. 6). The average temperature of collection was 5.3°C. Nezumia bairdii (Goode and Bean 1877) Nezumia bairdii is a relatively small macrourid with a reported head length of up to 60 mm (Mar- shall and Iwamoto 1973). During our study the head lengths varied from 12 to 66 mm with the weight of the largest specimen being 295 g. The geographic range of N. bairdii extends from the Straits of Florida north to the Grand Banks (Marshall and Iwamoto 1973). Nezumia bairdii is captured com- monly between 90 and 183 m in the northern part of its range and appears to undergo tropical sub- mergence because it is found primarily between 548 and 731 m in the southern parts of its range The inclusive depth range is 90-2,285 m (Goode and Bean 1885; Marshall and Iwamoto 1973). One anomalous catch at a depth of 16.5 m was recorded in Vineyard Sound (Bigelow and Schroeder 1953), but this was most likely a discard from a commercial fishing vessel. Within the study area the depth of capture ranged from 270 to 1,644 m (Fig. 3). The largest catch in a half hour tow was 76 fish and the greatest biomass per half hour tow was 5.7 kg. Nezumia bairdii com- prised up to 30% of the demersal fish catch in number and up to 15% of the biomass. In the January plot (Fig. 9), the head length in- creased slightly with depth. The regression line of the mean depth of each head length class showed a positive slope significantly different than zero. By June (Fig. 9) the regression line showed a highly significant positive slope and three distinct size groups separated by depth were evident. The first group included those fish <30 mm HL, the second group was from 30 to 42 mm HL, and the third group was >43 mm HL. The head lengths at the start of maturity for females (27 mm) and males (32 mm) cor- respond well with the dividing line between size groups one and two, as defined by depth distribu- tion. Also, N. bairdii females and males can be fully mature at 44 and 45 mm HL, respectively (Fig. 10). These values are close to the division between the second and third size groups noted above The three size groups appear to reflect maturity stages as well as size differences, and this may contribute to the bathymetric differences. The first group consisted of all immature fish that were not found in deep water in June The second group could be termed the transitional group because it included fish that were just starting to mature and those more highly developed. Since this group included such a diverse spectrum of maturity, it encompassed portions of the depth ranges of both immature and mature fish. The third group consisted of all mature fish and was not found in water shallower than approximately 600 m in June In September, the larger fish had reached their deepest limit, and immature N. bairdii were virtually absent deeper than 1,000 m. By November (Fig. 9), the largest fish were returning to shallower water to complete what appears to be a seasonal migration cycle Examination of histological sections of gonads showed that the only spent N. bairdii were captured on the September cruise Although no ripe fish were caught on any cruise, these spent fish suggest that N. bairdii spawns in July or August, coincident with the time when the mature fish are inhabiting their deepest level. Marshall's (1965) hypothesis concerning reproduc- tion of certain macrourids states that fertilization takes place at the bottom. Subsequently the eggs, which are buoyant, develop and hatch on their way upward to the seasonal thermocline The larvae then maintain themselves just below the thermocline, in order to take advantage of the plankton that tends to accumulate there in the density gradient. In con- 42 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE O . O UJ z ID ~3 I to o _ oo — ii n r i i i i i i i i i i i i o o o O O o o o o o o o o o Sep vo « eg o • N N N N N ~ ! I I I I I I I I o o o p o o o o o o W ♦ CM O 00 Mill! o o o o o o o z * eg O ec <© csi eg eg eg - - X »- a. UJ O TT O O eg O 0) P O -8 _o CL UJ CO <0 CO "* 00 o 1 ii II if) c ^ ( 1 I I I I I I I I I I I I I I II I T ooooooooo Oo OOO oooo Ocotf>vegOeo<0« •n eg eg eg eg eg — — — O O eg I I I I I I I I I § 8 § p ° O CD 10 O a. UJ o t. 0) -O o £ to c 0) 5 o ii r> _ -t-> E e E 5 CO ^^ C o a> «• £ X CJ 1- o ft CO o z UJ o _J 0) o < S 3 UJ C o X 1) eg II C O u ho C s- (1) J3 O e- the lines enclose t d of trawls. O « 3 (0 O Nezur inten r» t « <2 c E E co <« ■a C o "*— ' 5 « «■ b ° X »- o 05 o z (O UJ ft +j -1 < us de gle is UJ > o o X IM a> j= ^ "& J> o c of head the mea o Figure 7 — Graph qualis. The dot is 43 Nezumia a equal is FISHERY BULLETIN: VOL. 84, NO. 1 CO 2.0-1 O UJ a. _ co — -♦- x U. — ' 1.0 O a> o _J cr UJ CD Ao o a A A £^o° ° n a £*r o oD o ax a A e ; 73- •10 a : 74- 04 a : 75- ■08 X : 76- ■01 3.0-1 — 2.0- - ^ ^ O 1.0- D o a o o x °D * g X oa a A A — I 1 1 1 r— 100 200 300 400 500 — I 1 1 1 1 1 1 1 1 1 600 700 800 900 1000 1 100 1200 1300 1400 1500 DEPTH ( m ) Figure 8— The distribution of log transformed (log (x + 1)) abundance and weight of Nezumia aequalis at each station, plotted against depth. junction with Marshall's hypothesis, the advantages of the type of seasonal migration suggested by our data are twofold. First, the migration concentrates the reproductively mature fish in a limited area thereby increasing the probability of a sexual en- counter. Second, it allows additional time for develop- ment of eggs on their rise to the upper layers, and concurrently lessens the chance that the egg will travel through the thermocline and be removed from the area by the more aqtive surface currents (although egg density could be such that neutral buoyancy occurs at the thermocline). If these sug- gestions hold true, it would be expected that the lar- vae would benefit from the high productivity and warmer temperatures of the surface waters and have enhanced growth. As productivity declines in the late 44 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE UJ -3 CD O O O in CVJ i ii II rO c - f ! I I I I I I I I I I I I I I I I I I I I I I I I I I I ! OOOOOOOOOOOOOO o oooooooooooooo o 8coiO«-<\JO o z fO If) ID *■ ^- o 1 il II c - I I I I I I I I I I I I I I I I I 1 I I I I I I I I I I I I ooooo oooocqoooq oo ooo oooooooooo O 0 <0 * N OCDlOVCSlOatf^N k> eg eg eg eg eg — — — — — a. UJ Q -6 5 s § in 1 II II I — I ■ X Is- I I I I I I I I I I I I I I I I I I I I I I I 1 I I I ' oooooooooooooo oooooooo O CO <0 * ** ° * IO eg eg eg o <0 o o t o ro o eg _o Q. UJ CO 00 o m o pj eg eg eg eg - — — ~ ~ X H a. UJ Q c nj o 10 C CD S 3 CD o a. to m ^-* U-, E o h 0) _Q ^"* £ ¥ 3 C X V h- 4= o z II o UJ c IO _l n aJ < ho a UJ 3 O X t-c eg +-> co bo O C cS 5 a> >- 0) 45 O C c« V 6 CD CO il CO t. ^ 6 3 •*; c O ~$ 3> <0 •h5 -o II O o 6 E s £ -l a < UJ O x eg _o if CO $ CV T5 CO 3 CO u % to C CD - C •- 7-1 < 2 6H Irh i Irhl -EEZB- H=£3 — I ■*=fc=H -EE& — i 1 1 1 1 1 1 O 10 20 30 40 50 60 70 HEADLENGTH Figure 10.— The gonadal maturity stages plotted against head length for Nezumia bairdii. recruitment of young occurred between the months of November and January. No small N. bairdii were captured benthically between the proposed deep- water spawning time and the shallower January recruitment spike. The larger N. bairdii occurred deeper than the small ones (Figs. 9, 12) demonstrating the "larger- deeper" phenomenon. The age and growth analysis of N. bairdii presented many problems. Due to the thickness of the sacculus otolith a thin cross section had to be removed from each. After examination of the thin sections, two problems became apparent. First, all of the smaller specimens had two hyaline zones. Because the specimens were obtained on the winter (January; 76-01) cruise, all had hyaline zones around the perimeter as expected. There was, in addition, a well-defined hyaline zone in the interior of all the otoliths obtained from the smallest fishes available (<27 mm HL). Subsequently two hypotheses were proposed: 1) a period of hyaline zone formation (slow growth) occurred between June-July (spawning) and January, and 2) young N. bairdii were not available to our trawl until the second winter hyaline zone was forming (age about 1.5 yr). The first hypothesis was discarded because a period of slow growth within the first 6 mo would have no apparent selective advantage It should be noted, however, that since the larvae of N. bairdii were probably pelagic, a change from planktonic feeding to benthic feeding would have occurred dur- ing that time. Such an ontogenetic change occurs in related gadid fishes. Musick (1969) described the 70-1 60 50- 40 30 20- 10- Nezumia bairdii J ^ H.L. (mm) Figure 11— Head length frequency distribution for Nezumia bairdii by cruise The number above each cruise indicates the number of specimens. 46 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE ontogenetic transition for Urophycis chuss and sug- gested that the transition from pelagic to demersal adaptations in morphology and behavior occurred within a period of 12-24 h. This short time span would be unlikely to be reflected in macroscopic hyaline band formation. Therefore, the second hy- pothesis appeared more likely, and led to the con- clusion that the juvenile N. bairdii remained pelagic until the second winter and then descended from the water column to the bottom where they were captured. Nezumia bairdii The second problem was that in the older fish (>4 yr) the outer bands were very difficult to define with any degree of confidence The percentage of unreadable otoliths increased from about 5% in fish <4 yr to about 50% in fish >4 yr. The mean head length of N. bairdii with four bands was 42.7 mm, the size at the onset of sexual maturity. Growth may have slowed down to compensate for the energy needed for reproduction, and produced spatially close and obscure hyaline zones. Therefore spawn- ing checks may have had considerable influence on co 3 0n a. ^ co - 2. OH O o> o _l £T UJ CD o- ho o X X oo D ■x o X loxo X ° hfr ttfV ° a x a e = 73-10 a = 74-04 A • 75-08 X -- 76-01 4.0-1 3.0 ® *2.0H £ o 1.0- P*h xxX a «& o x \ xS°_ 6 a ^ a a 4> *$ oo 0 A a A x — I 1 1 1 l 1 1 1 1 1 200 400 600 800 1000 1200 1400 1600 1800 2000 DEPTH (m) Figure 12— The distribution of log transformed (log (x + 1)) abundance and weight of Nezumia bairdii at each station plotted against depth. 47 the interpretation of the hyaline zones. Using the length at age data, a Walford growth transformation graph was plotted (Beverton and Holt 1957). Instead of calculating the L^, we used our largest specimen (66 mm HL ). The estimate of Brady's coefficient (K) obtained from this graph was 0.276. Using the Walford graph, the head lengths for those presumed ages >4 yr could be iteratively generated. This method gave a maximum age of ap- proximately 11 yr. The von Bertalanffy growth equa- tion for length was Lt = 66 (L - e -0.276 (T+0. 16)). Rannou (1976) studied the age and growth of a congener (N. sclerorhyncus) that occupies a similar depth range in the western Mediterranean. He calculated a K coefficient of 0.16 and an L^ of 42.31 mm HL. Thus, although this species is smaller than N. bairdii, it has a much slower growth rate, prob- ably attributable to lower productivity in the western Mediterranean compared with the slope off the mid- Atlantic coast of the United States (Koblentz-Mishke et al. 1970). The length-weight regression for N. bairdii (Fig. 2) was analyzed. The solution of the line for N. bair- dii males was log (weight) = 0.038 (head length) + 0.083, r2 = 0.810, and for females it was log (weight) = 0.035 (head length) + 0.216, r2 = 0.760. These length-weight relationships are not unlike those summarized by Gordon (1979) for other small macrourids (Coelorinchus coelorinchus, C. occa, and Nezumia aequalis). In summary, larger N. bairdii were captured deeper and the minimum and maximum depths of capture off the mid-Atlantic coast were 270 m and 1,644 m. The fish seasonally migrated to deeper water with the mature fish occurring deeper than immature fish. The males matured at about 45 mm HL and the females became mature at 44 mm HL. Nezumia bairdii probably spawned pelagic eggs in July and August and the young apparently remained pelagic until the second winter (January), when they first appeared in bottom trawls. The maximum age of N. bairdii was presumed to be 11 yr. The temperature range for N. bairdii was from 3.7° to 10.0°C, with the average temperature of capture being 5.3°C (Fig. 6). Coryphaenoides rupestris (Gunnerus 1765) Coryphaenoides rupestris is a large macrourid that reaches a total length of about 100 cm (Sawatim- skii 1971; Nodzinski and Zukowski 1971; Marshall FISHERY BULLETIN: VOL. 84, NO. 1 and Iwamoto 1973), and is found on both sides of the North Atlantic. In the eastern North Atlantic it ranges from the Trondhjem area to the Bay of Biscay. In the western North Atlantic it is reported to occur from Davis Strait to ca. lat. 37°N (Marshall and Iwamoto 1973), although two specimens (81 and 100 mm HL) were captured by C. Richard Robins5 at lat. 23°29.8-32.0'N, long. 77°05.5'W. The depth distribution of C. rupestris varies from about 180 to 2,200 m (Leim and Scott 1966) with highest abun- dance occurring between 400 and 1,200 m (Marshall and Iwamoto 1973). Coryphaenoides rupestris is rarely used as a food fish in the United States, but the German Democratic Republic, the Soviet Union, and Poland fish commercially for it in the western North Atlan- tic In 1968, the Soviets recorded a harvest of 30,000 tons of C. rupestris off Labrador, Baffin Island, and Greenland (Nodzinksi and Zukowski 1971). The catches of this macrourid were reported to increase during the second half of the year as the catches of redfish and cod decreased (Sawatimskii 1971). Coryphaenoides rupestris was captured in the Nor- folk Canyon area at depths of 578-1,698 m (Fig. 3). Sawatimskii (1971) reported that C. rupestris is known to form dense aggregations off the coast of Labrador. In November 1974 an anomalous catch of over 6,000 C. rupestris with a total weight >1,000 kg was obtained in a half hour tow in the Norfolk Canyon area. A random subsample of 1,000 speci- mens was examined and no sexually mature fish were found. Although the head length ranged from 59 to 110 mm, the length-frequency curve was strongly unimodal at 76 mm. The greatest number and biomass of C. rupestris caught in "normal" half hour tows was 128 fish comprising 39% of the in- dividuals and 68 kg, and representing 65% of the total catch by weight. The largest specimen captured had a head length of 155 mm. The head length distribution by depth and by cruise (Fig. 13) suggested a mass movement of C. rupestris toward deeper water during the summer months, and a reciprocating movement to shallower water in the winter. In January, the majority of C. rupestris was captured between 700 and 800 m, while in June and September there appeared to be a movement toward deeper water. By November the depths of capture decreased and were similar to those of January, and the slope of the head length- depth regression for C. rupestris was significantly 5C. Richard Robins, Rosenstiel School of Marine and Atmospheric Science, Division of Biology and Living Resources, 4600 Ricken- backer Causeway, Miami, FL 33149. 48 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE UJ z z> -3 ro 0 00 00 1 11 11 to IV- c •~ I I I I I I I I I I I I I I I 00000000 00000000 a OB UJ ♦ <\l O <*> UJ cm cm cm cm cm — — I I I I I I I I I I I I I I OOOOOO o OOOOOO o rt T CM m O <3> '00 ID r» > m O ID z (0 0 O to m in I fO c «_ * 1 1 1 t 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 oooooooooc o o to o CD CM o CO CM o CM CM O O O CO CM — o o CM I 1 1 I I I I ! M §0000 o O o o O CO UJ * CM c* — » E E X (9 z jo uj -J o < UJ M X (0 in Q. UJ o or to I <*r> < —3 5 5j" CVJ S- 1 C0 11 c 11 rr~r 3 2 3 O ■> 00 f| CM T I I I I I I I I I I I I I I I I T I I I I I I 000000900000 o UJ eg o CM o O o CO O CO O CM O o o CO o TT O o CM X r- a. UJ Q O in 00 m m '43 lO o rnr ■ ii 1 1 ii ■! i Q- UJ (/> 00 0 0 h- 10 If) 11 11 r- c •♦- II w I) O lO 9> 00 X o a UJ _«"> X -8 X h- 0. UJ Q r I I 1 I 1 1 I I I I I 1 I 1 I I 1 1 1 1 1 I II I I I I ' I 1 00 000 0000000000 Oo OOO OOOOOOOOOO Ocmcmcm s 3 C fl> J5 c to C 0) o a> 0, w O S-, 01 £ 3 c 01 5 01 J3 _o "o C 01 0> C Oi "S 5 1 C <4-c CO o ~ OS E E ?> c o o or ° .22 # 3 10 O ^ *> J3 *J 0) bj) ^> C 23 5 > o> a -a cs 0) -c ■s.1 O * o 9- 49 FISHERY BULLETIN: VOL. 84, NO. 1 different from zero. There was no apparent seasonal size segregation evident as in Nezumia bairdii, but the graph of numerical abundance against depth also indicated a general seasonal movement down slope in September (Fig. 14). Similar seasonal movements have been shown by Savvatimskii (1971) off Newfoundland. Females may be mature from about 104 mm HL and males from 71 mm HL (Fig. 15). Podrazhanskaya (1971) supported Zarkharov and Mokanu's (1970) theory that C. rupestris spawns in Icelandic waters. She stated that C. rupestris spawn near Iceland and the Irminger Current could transport the eggs and larvae to Greenland. From Greenland the western branch of the West Green- land Current would transport larvae to Baffin Island where the Labrador Current would move the fish down to the Newfoundland banks. When the fish in the Newfoundland area attain a size of 40-50 cm total length (TL), they start to migrate back to Iceland. Podrazhanskaya gave the modal lengths for C. rupestris in each area. The smallest fish (modal TL of 45-47 cm) were found on the Northern Newfound- land bank and the largest (modal TL of 98-100 cm) were found around Iceland. Fish from between Baf- fin Island and West Greenland had modal lengths 200 400- 600 800 1 000-1 1200 1400- 1600 1800 2000- 2200- 2400- 2600- 2800- 3000 LOO 0 I I > JAN. JUNE NOV. SEPT. Coryphaenoides rupestris abundance - log(**« I ) FIGURE 14.— Diagram of depth plotted against the log transform- ed (log (s + 1)) numerical abundance, by cruise, for Coryphaenoides rupestris. of 60-62 and 78-80 cm, respectively. Podrazhanskaya's (1971) modal-length data for each area in conjunc- tion with Savvatimskii's (1971) age and growth data reveal that the modal-length fish off the Newfound- land banks are about 6 yr old, off Baffin Island they are 9-10 yr, around Greenland they are 15-16 yr, and at Iceland they are over 20 yr. If a spawning migra- tion occurs, it does not preclude spawning by some members of the population not undergoing migra- tion, thereby accounting for the small percentage of ripening fish to be found outside of their primary spawning area. If Podrazhanskaya's migration theory is valid, some interesting observations can be made First, the C. rupestris found on the east coast of the United States may be derived from the larvae that failed to metamorphose by the time they reached the New- foundland banks and continued to drift southwest. The predominant currents move south and west from Newfoundland to Cape Hatteras (Worthington 1964; Webster 1969; Gatien 1976), thereby affording a means of transport for unmetamorphosed larvae (Wenner and Musick 1979). Additionally, the modal length for the 7,011 C. rupestris caught in the Nor- Coryphaenoides rupestris 3 - i- 10 a. EH 4=h ^B-\ I 6 5 - 2 - -«=E 3H I I I I — i 1 1 1 i 1 1 1 i p i SO 60 70 80 90 100 MO 120 130 140 ISO 160 HEAOLENGTH FIGURE 15.— The gonadal maturity stages plotted against head length for Coryphaenoides rupestris. 50 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE folk Canyon area was 46 cm, exactly that which was found for C. rupestris in the Newfoundland bank area. However, no small C. rupestris were captured in the Norfolk Canyon area. We found only 2 fish with a head length <40 mm (24 cm TL) and only 10 fish with head length <50 mm (30 cm TL). The regression line for head length against log (weight) (Fig. 2) was analyzed. The solution for C. rupestris males was log (weight) = 0.023 (head length) + 0.82, r2 = 0.898, and for females it was log (weight) = 0.018 (head length) + 1.16, r2 = 0.885. Unfortunately these length-weight data cannot be compared directly with those summarized by Gor- don (1979) because we measured head lengths in our study and he gave standard lengths. We do not have the data at present to compute the regression for head length on standard length for this species. Temperatures at which C. rupestris were captured near Norfolk Canyon ranged from 3.7° to 5.7°C (Fig. 6). The average temperature was 4.9°C. Coryphaenoides rupestris does not follow the "larger-fewer-deeper" pattern shown for N. bair- dii in Norfolk Canyon because it migrates seasonally (Fig. 16) and the larger specimens traverse the en- tire bathymetric range (Fig. 13). In summary, C. rupestris migrated seasonally to shallower water in the fall and early winter. Catch per unit effort increased in the fall and winter, and a dense aggregation was found in the fall. Podra- zhanskaya's (1971) spawning and migration theory appears feasible but further intensive study is need- ed. No ripe, running, or spent fish were captured in the Norfolk Canyon area out of 7,011 individuals examined. There was a trend for the larger C. rupestris to range deeper but not to the degree that was found in N. bairdii. It appears that the distribu- tion of C. rupestris was more closely related to temperature than to depth, the species being found mostly within the 4°-5°C range. Coryphaenoides carapinus (Goode and Bean 1883) Coryphaenoides carapinus is another small macrourid which grows to about 390 mm TL, and is found on the lower slope and abyss from 1,000 to 3,000 m (Haedrich and Polloni 1976). In the western North Atlantic it has been found between Nova Scotia and Cape Hatteras (lat 37°N) and in the eastern Atlantic from lat. 50°N to the Equator. Cory- phaenoides carapinus has also been reported from the mid-Atlantic ridge (Marshall and Iwamoto 1973). In the Norfolk Canyon area C. carapinus was cap- tured at 1,108-2,767 m (Fig. 3). The largest number caught in one trawl was 37 (total weight 550 g). These were captured in September 1975 at a depth of 1,803 m. Coryphaenoides carapinus comprised up to 23.4% of a catch in number, but only 4.3% in biomass. The maximum size captured was 90 mm HL. Coryphaenoides carapinus tended to be larger at the lower end of its depth range (Fig. 17). The slope of the regression line for head length with depth was significantly different than zero. The coefficient of determination was 0.346. Figure 18 displays low numbers and high vari- ability in the capture of C. carapinus in relation to depth. The phenomenon of fewer, larger fish at the deeper part of the bathymetric range was evident but obscured because of the relatively small size of C. carapinus, low numbers, and contagious distribution. Coryphaenoides carapinus was taken at temper- atures of 2.5°-4.2°C with the average temperature being 3.7°C (Fig. 6). Some overlap in distribution with depth and temperature occurred among C. carapinus, C. armatus, and C rupestris. Because C carapinus is a small species and mostly a ben- thic feeder (Haedrich and Polloni 1976) and C. ar- matus and C. rupestris are large species that forage into the water column (Podrazhanskaya 1971; Haedrich and Henderson 1974; Smith et al. 1979), competitive interaction is probably low. Coryphaenoides armatus (Hector 1875) Coryphaenoides armatus is cosmopolitan in distri- bution, being found in all oceans except the Arctic. It commonly is found from 2,200 to 4,700 m, with a few specimens being captured as shallow as 282 m (Marshall and Iwamoto 1973). Larger individuals have been shown to forage off the bottom for pelagic prey (Haedrich and Henderson 1974; Pearcy 1975; Smith et al. 1979). Coryphaenoides armatus attains a size of 165 mm HL and over 870 mm TL (Iwamoto and Stein 1974). The largest specimen captured in Norfolk Canyon was 146 mm HL. Although C. ar- matus is one of the deepest living macrourids, it is rather well-known biologically because of its broad distribution and availability to deepwater trawls (Haedrich and Henderson 1974; Pearcy and Ambler 1974; McLellan 1977; Smith 1978). Coryphaenoides armatus was taken in every suc- cessful trawl from 2,100 m to our deepest trawl of 3,083 m in the Norfolk Canyon area and virtually was confined to below the 3°C isotherm (Fig. 3). In 51 FISHERY BULLETIN: VOL. 84, NO. 1 2 LU 5 3.0- O u X O o> o -J i. OH cr UJ GQ Coryphaenoides rupestris o* D x • = 73-10 a = 74-04 a = 75-08 X - 76-01 a a X t r o A t 1 1 1 1 r A a x a a ~i 1 1 1 -i r A o a T 6.0-1 5.0- 4 0- x — (S) x 3.0- > o 2 0- 1.0- x ,a a ° - . x a d X - 1 1 1 1800 2000 — i 1 1 1 1 1 1 1 1 r- — i 1 1 1 1 1 r 200 400 600 800 1000 1200 1400 1600 DEPTH (m.) Figure 16— The distribution of log transformed (log (x + 1)) abundance and weight of Coryphaenoides rupestris at each station, plotted against depth. one trawl C. armatus comprised 92.7% of the bentho- pelagic fish numbers and 93.4% of the biomass. In a 1-h trawl the maximum number captured was 76 and the maximum biomass was 21.2 kg. No increase in fish size with increased depth was evident in the data (Fig. 19) (Table 1), and the slope of the regression line for head length with depth was not significantly different from zero. However, known depth range of C. armatits was incompletely sam- pled in this study, and further samples from greater depth may lead to other conclusions. The distribution of numerical abundance and weight with depth are shown in Figure 20. Cor- yphaenoides armatus increased in abundance from 2,100 to 2,600 m, beyond which its abundance re- mained constant. The regression lines for head length against log (weight) were analyzed (Fig. 2). The solution for males was log (weight) = 0.017 (head length) + 0.956, r2 = 0.967, and for females it was log (weight) = 0.016 (head length) + 1.029, r2 = 0.972. The maturity stages of C. armatus against head 52 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE UJ z 3 ~i 10 O CD r^- i II n to r- c *- t I I I I I I I I I II I I I I I I 1 I I I I I I I I I I I oooooooooooooo o oooooooooooooo o «<0<0«CMOCO O Z o 1 I*- II c o CM o fO II 8 I 1 I I I I II I ! I I I I I I I I I I I OOOOO QOOOC OO OOO OOOOo o e « t n o ct> to * cm o K) N N N n cm — — — — — Q. UJ Q I I I I I I I I §§88 • <0 « CM o w o « — E E o X I- o Q < UJ O X (VI _o CO £ c 01 -a c S3 CO c u CO ft o s- 0) s 3 c a> II c bo c 2 _o o C 01 to 01 c . ;s to a* > ^ § -o -" c3 o to I o -8 o _ * CM o _o o o CM -O a. UJ en 00 o I m S 2 II II E E o 2 UJ -J o < UJ 9 x l i i i i i I ; i i i i i i I I i i l i i l l I i I ' I ' ' § oooooooooooooo oooooooooooo o co <0«cm occ(c-» ft ^> 2 « X a. UJ o a. UJ o w D o 53 Coryphaenoides carapinus FISHERY BULLETIN: VOL. 84, NO. 1 3. On 2 UJ o UJ Q- ^_ if) — 2.0- O o» o tr i UJ QD 3 Z .0- X o t r a o OD ° a" & x X o°.° , az ~i 1 1 r x x i r D ~i r x 0 2 73- ■10 O = 74- -04 & : 75- ■08 X : 76- -01 -i 1 1 r t r 3.0-1 X — 2.0- S- o 1.0- o ° a f ^ XO On a a a X u o a «^ X ° X X t 1 1 1 1 1 1 1 r- — i 1 1 1 \ 1 1 1 1 1 1 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 DEPTH (m) Figure 18— The distribution of log transformed (log (x + 1)) abundance and weight of Coryphaenoides carapinus at each station plotted against depth. lengths are shown in Figure 21. No mature males were found, but the females matured at about 78 mm HL. Coryphaenoides armatus was captured in temperatures ranging from 2.3° to 3.3°C (Fig. 6). The majority of individuals, however, were caught between 2.4° and 2.9°C during the study and the average temperature was 2.6°C. Distribution of Macrourids With Temperature Depth distribution has been used commonly throughout the literature to delineate the habitat of various fishes, including macrourids (Macpherson 1981). The temperature ranges for each species in 54 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE UJ -3 I rO if) UJ < I I I I I I I ! I I 1 I I I I 00000000 00000000 0U3 * W o O o <7> o - o 0KJ rtn -8 «D _ o o <0 o > o o 1^- 11 'i C t- CVI CM CM X I- UJ Q * I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 00000 OOOO0OO0OO OOOOO OOOOOOOOOO O CD IS ^ N OOOtf>VC\lOeO(C*CVJ rocMCMCMCVJCM — — — — - Q. UJ Q o ff> — . E E o z *" _i o < UJ O I IS O o £ 3 c 0) 4= T3 C CO- CO c O) o> a co 01 3 C 0> J3 II C a1 ho C S3 o> J3 u C a> . to i2 o> ~ is o o o -8 -1 ■ I 1 I < o I CD Is- !2 10 c r t 1 1 ' 1 ' ' 1 1 1 t o o o o o o o o O CD <£ * CM CM CM o o (VI CM O o o CM I 1 I I I I 0000 o CD I I I I I I I I I I I I 00000 o 0- UJ » *- «5 O C «h o or ° .9 £ b <*> •^ £ .m aT _o> T3 M co C 3 03 >- K o> o> > «- J3 c ~C cy o is _C CO Vi o L.S w . as 2 55 FISHERY BULLETIN: VOL. 84, NO. 1 UJ o UJ Q_ ^^ O) — Coryphaenoides armatus 3.0-1 2.0- O o> o -1 i. oH a: UJ CD a : 73-10 D : 74-04 A : 75-08 X - 76-01 CP x X A A A X t 1 1 1 r t r ~\ 1 1 1 1 o 5.0-1 4.0- 3.0- ^ o -1 2 0- o- « ^ X A A A 1600 — I 1800 ~T T 2000 2200 2400 DEPTH (m ) 2600 2800 T 1 3000 Figure 20— The distribution of log transformed (log (x+1)) abundance and weight of Coryphaenoides armatus at each station plotted against depth. the present study showed some overlap, but the temperatures at which the population modes were found were fairly discrete except for Nezumia ae- qualis, Nezumia bairdii, and Coryphaenoides rupestris. In Figure 6 the relationship of species with temperature is more clearly defined. The minimum temperature of each species remained fairly constant as did the maximum and modal temperature for those species in which there was no indication of seasonal migratory patterns (Coelorinchus car- minatus, Coryphaenoides carapinus, C. armatus). The 3.5°C minimum temperature found for C. carapinus in June was probably not accurate since the deepest trawl of that cruise did not encompass the entire range of C. carapinus. Similarly, the minimal temperatures for C. armatus may not be representative Competition Among Macrourids Competition among macrourids in the Norfolk Canyon region is probably minimal because the species differ in body size and feeding strategies or, if feeding strategies are similar, the species have dif- ferent distributions with temperature and depth. Close congeners such as Nezumia bairdii and N. aequalis might be expected to occupy similar depth p;r MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE 6-1 5 4 3 2 UJ O < V) £ < s Coryphaenoides armatus HE£ EH -I 1- 6 -i 5- 4 3- -EEtEEH "^ 40 SO 60 70 80 90 100 110 HEADLENGTH Figure 21— The gonadal maturity stages plotted against head length for Coryphaenoides armatus. and temperature ranges; however, the N. aequalis in this area were at the northern limit of their geographic range, occurred in small numbers, and may have been in direct competition with TV. bair- dii. Although C. rupestris also occupied the lower section of the two Nezumia spp. temperature and depth regimes, direct competition was probably low because of their dissimilarity in mouth size and morphology and related differences in diet (Podrazhanskaya 1971; Geistdoerfer 1975; McLellan 1977). Abundance and Density of the Family Macrouridae In the study area the abundance of macrourids, in water shallower than 2,000 m, was fairly constant with respect to other bottom fishes. The average per- cent of macrourids by number in each cruise was 16.6% in cruise 73-10 (June), 15.0% in 74-04 (December), 14.6% in 75-08 (September), and 18% in 76-01 (January). The major peaks of abundance were found between 300 and 400 m, where Coelorin- chus c. carminatus was present, and around 800 m where the complex comprised of Nezumia aequalis, N. bairdii, and Coryphaenoides rupestris dominated (Fig. 22). In depths of over 2,000 m the numerical dominance of C. armatus was evident. Some of the minor inflections can be attributed to the contagious distributions displayed by these fishes. The graph of macrourid biomass (Fig. 23), as per- cent of the catch, was similar to that for numerical abundance except for a shift in biomass from 800 m to below 1,000 m between January and June. This was probably because of the seasonal movement of the larger macrourid Coryphaenoides rupestris. Be- tween about 1,400 and 2,200 m, macrourids made up a very small portion of the biomass, although their percent by number was comparable with lesser depths. The dominant macrourid in this area, C. carapinus, was small, and Antimora rostrata, a large morid, was the most abundant member of the benthic fish community from 1,300 to 2,500 m (Wen- ner and Musick 1977). In depths >2,200 m the biomass of C. armatus steeply increased with depth, until it was the predominant member of the benthic community. All the macrourid species, with the exception of C. rupestris, maintained a fairly constant numerical distribution from cruise to cruise There was ap- parent variability for C. carapinus and C. armatus, but this was due to the small number of samples from deeper areas. Distribution of macrourids as the percent of catch revealed a gradual replacement of species with depth, and the predominance of C. ar- matus in depths >2,500 m. Macrourids made up a major numerical portion of the benthic fish community from 300 m to the deepest station at 3,083 m. Macrourids were also a main component of the biomass of the commu- nities from 300 to 3,083 m, excluding the 1,300- 2,500 m range where the morid, A. rostrata, dominated. Although Macrouridae is a dominant family in the Norfolk Canyon area, the potential for a fishery is essentially nonexistent. Coryphaenoides rupestris is the only species which attains an appreciable size in the mid-Atlantic area; a modal length of 46 cm TL. However, this size is much smaller than typically found in the North Atlantic and the density of organisms is generally low (normally <0.86 in- dividuals/1002). In addition, C. rupestris demon- strates a tropical submergence, being found deeper in lower latitudes. The depth range of this species in the Norfolk Canyon area (578-1,698 m), combined with smaller size and lower density of organisms, in- dicate that a commercial fishery would not be economically feasible 57 FISHERY BULLETIN: VOL. 84, NO. 1 O O X Q- O 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 0 76-01 2 4 h 6 8 10 I 2 14 16 18 20 22 24 26 28 30 . 75-08 JAN. 73-10 JUNE SEPT. . 74-04 NOV. 0 50 00 PERCENT Figure 22— Depth versus relative abundance (as percent, by number, of total capture) for the family Macrouridae, by individual cruise Comparison With Other Studies The comparison of this study with others in the North Atlantic lends support to Marshall and Iwamoto's (1973) hypothesis that the greatest diver- sity of macrourids is in the bathyal tropical regions. The number of macrourid species declines from tropical to boreal regions. Marshall and Iwamoto (1973) reported 32 macrourid species from the Carib- bean and Gulf of Mexico, but only 22 species were captured during our study (Table 2). Bullis and Struhsaker (1970) found that Macrouridae was one of the dominant families on the western Caribbean slope between 201 and 400 fathoms (368-732 m). The deepest stratum sampled was 451-500 fathoms (825-914 m), and macrourids (9 species) comprised about 67% of the individuals captured within these depths. Within the same depths in the Norfolk Can- yon area the dominant macrourids (4 species) con- tributed about 31% to the total catch. Merrett and Marshall (1981) remarked on the high diversity (and apparent resource partitioning) of macrourids from a tropical upwelling area off north- west Africa and reported 26 species from there They found 18 species on the slope (< 1,600 m), including four species of Nezumia. Bathygadine macrourids were important off Africa but virtually absent in our study area. Thus macrourid diversity is probably highest on the continental slope in the tropics, par- ticularly in areas of higher productivity. In addition, high diversity is manifested there at several tax- onomic levels, from the species to the subfamily. Haedrich et al. (1975) reported the capture of 121 macrourid specimens (3 species) in 29 trawls off Southern New England. Their trawl depths ranged from 141 to 1,928 m. Their findings were similar to 58 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE O O X »- a. bJ o u 2 76-01 JAN. 4 4 6 - e 10 12 14 16 IS 20 22 24 26 28 2 ■ 4 ■ 6 8 10 12 14 16 18 20 22 24 26 28 30 73-10 JUNE 75-08 SEPT. 74-04 NOV. 0 50 OO 1 i I PERCENT Figure 23— Depth versus relative abundance (as percent, by biomass, of total capture) for the family Macrouridae, by individual cruise Table 2. — Species captured during study, with total number and total weight. Total Total Total weight Total weight Species number (g) Species number (g) Coelorinchus c. carminatus 1,827 38,597 Coryphaenoides colony 1 20 Coelorinchus caribbaeus'1 10 419 Coryphaenoides leptolepis 12 4,922 Coelorinchus occay 1 2 Ventrifossa occidentalis 60 1,449 Nezumia aequalis 285 4,041 Ventrifossa macropogon 1 8 Nezumia bairdii 2,222 72,865 Hymenocephalus gracilis^ 1 1 Nezumia longebarbatus2 12 1,299 Hymenocephalus italicus^ 1 12 Nezumia sclerorhyncus 1 8 Bathygadus favosus 2 — Nezumia cyrano* 1 — Bathygadus macropsy 1 22 Coryphaenoides rupestris 7,120 1,229,304 Sphagemacrurus grenadae2 4 30 Coryphaenoides carapinus 213 4,703 Macrourus bergiax3 2 4,470 Coryphaenoides armatus 391 120,456 Gadomus dispart 1 — 'Range extension from the Gulf of Mexico-Caribbean area. 2Also reported by Haedrich and Polloni (1974). 3Range extension from Boreal Northwest Atlantic. 59 FISHERY BULLETIN: VOL. 84, NO. 1 those in the present study within the 350-1,100 m depth interval. Respectively, the family Macrouridae accounted for 21% and 22.4% of the fishes captured in these depth intervals. Haedrich and Krefft (1978) studied the fish fauna in the Denmark Strait and Irminger Sea. In the five fish assemblages that they reported, macrourids were abundant in the 2,026-2,058 m assemblage (22.4%) and very dominant in the 763-1,503 m (48.3%) and 493-975 m (55.4%) assemblages. Macrourids were conspicuously absent from their group three assemblage, although it was well within macrourid depth and temperature range (280-776 m, 1.4°-7.4°C). An interesting aspect of Haedrich and Krefft's (1978) study was evident in their group two assemblage Coryphaenoides rupestris was the highly dominant fish (48.3%) in this group, and the temperature range of this group (3.9°-5.6°C) corre- sponded closely to the temperature range we found for C. rupestris in the present study (3.7°-5.7°C). Pearcy et al. (1982) summarized data on deep-sea benthic fishes collected over several years off Oregon (Day and Pearcy 1968; Pearcy and Ambler 1974). Iwamoto and Stein (1974) reported 11 species of macrourids from the northeast Pacific and Pearcy et al. (1982) recorded 8 of these off Oregon. A com- parison of these data with ours shows that the greatest contrast in the two areas is on the upper and middle slope (500-1,000 m) where five common species are regularly encountered in the western Atlantic (Coelorinchus c. carminatus, Nezumia bair- dii, C. aequalis, Coryphaenoides rupestris, and Ven- trifossa occidentalis), but Pearcy et al. (1982) record- ed no macrourid as common. This faunal difference may be due to the high density off Oregon of scor- paeniform and lycodine fishes, many of which may fill niches on the upper slope occupied by macrourids elsewhere The macrourid fauna in depths >2,000 m have many similarities to our study. Coryphaenoides armatus becomes increasingly dominant below this depth and often is the only species captured deeper than 3,000 m in both areas (see also Musick and Sulak 1979). Among other macrourid species Cory- phaenoides leptolepis is usually second or third in abundance at abyssal depths in both regions (Musick and Sulak 1979). This distribution pattern is very different from that reported for the continental rise in the tropics off west Africa (Merrett and Marshall 1981) where C. armatus and other large rat tails were very rare Marshall and Merrett (1981) speculated that the rari- ty of large predatory scavengers in the upwelling area they studied might be because of the com- petitively superior fishes of small size which were better adapted to use the constant abundant food supply there This speculation is not supported by data from the southern Sargasso Sea and Bahamas (Musick and Sulak unpubl. data), a tropical region quite low in productivity, in which large rat tails, such as C armatus, are also very rare The virtual absence of C. armatus from tropical abyssal areas may be due instead to some restriction on the life history of the species. Musick and Sulak (1979) have sug- gested that this species (along with some other large species of predator/scavenger such as C. rupestris and Antimora rostrata) may migrate to boreal areas to spawn. The tropics may be too far removed from such spawning areas for individuals to successfully return. ACKNOWLEDGMENTS We wish to thank all colleagues formerly or pres- ently with the Virginia Institute of Marine Science for their enthusiastic participation in the deep-sea program, and particularly to Charles Wenner, Richard Carpenter, Douglas Markle, George Sedberry, and Kenneth Sulak. Daniel Cohen of the Los Angeles County Museum of Natural History kindly contributed cogent comments on early stages of this manuscript. Drafts and final copy of this report were prepared by the Virginia Institute of Marine Science Report Center. LITERATURE CITED Beverton, R. J. H., and S. J. Holt. 1957. On the dynamics of exploited fish populations. Fish. Invest. Minist. Agric. Fish. Food (G.B.) Ser. II 19, 533 p. BlGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Gulf of Maine U.S. Fish Wildl. Serv., Fish. Bull. 53, 557 p. Bullis, H. R., Jr., and R T. Struhsaker. 1970. Fish fauna of the Western Caribbean upper slope Q. J. Fla. Acad. Sci. 33(l):43-76. Christensen, J. M. 1964. Burning of otoliths, a technique for age determination of soles and other fish. J. Cons. Int. Explor. Mer 29:73-81. Day, D. S., and W. G. Pearcy. 1968. Species associations of benthic fishes on the continen- tal shelf and slope off Oregon. J. Fish. Res. Board Can. 25:2665-2675. Farron, G. R 1924. Seventh report on the fishes of the Irish Atlantic Slope The macrourid fishes (Cory-phaenoididae). Proc R. Ir. Acad., 36 Sect. B. Gatien, M. G. 1976. A study in the slope water region south of Halifax. J. Fish. Res. Board Can. 33:2213-2217. Geistdoefer, P. 1975. Ecologie alimentaie des Macrouridae Teleosteens Gadi- formes. Ph.D. Thesis, Univ. Paris, Paris, France 315 p. 60 MIDDLETON and MUSICK: ABUNDANCE AND DISTRIBUTION OF MACROURIDAE Gilbert, C. H., and C. L. Hubbs. 1920. The macrourid fishes of the Philippine islands and the East Indies. U.S. Natl. Mus. Bull. 100, l(pt. 7):369-582. Goode, G. B., and T. H. Bean. 1885. Descriptions of new fishes obtained by the United States Fish Commission mainly from deep water off the Atlantic and Gulf Coast. Proc U.S. Natl. Mus. 8:589-605. Gordon, J. D. M. 1979. Phenology of deep-sea anacanthine fishes. In P. J. Miller (editor), Fish phenology: anabolic adaptiveness in tele- osts, p. 327-359. Symp. Zool. Soc Lond., Acad. Press, N.Y. Grigor'ev, G. V. 1972. Orazmnozhenii tuporylogo makrurusa severnoi Atlan- tiki (Reproduction of Macrurus rupestris (Gunner) of the northern Atlantic). Tr. Polyarn. Nauchno-Issled. Proektn. Inst. Morsk. Rybn. Khoz. Okeanogr. (PINRO) 289:107- 115. (Engl, transl., Fish. Res. Board Can., Transl. Ser. 2529.) Gunnerus, J. E. 1765. Efterretning OM berglaxen, en rar norsk fishk, som kunde kaldes: Coryphaenoides rupestris. K. Nor. Vidensk. Selsk. Skr. Trondhjem 3(4):40-58. GUNTHER, A. 1887. Report on the deep-sea fishes collected by H.M.S. Challenger during the years 1873-76. Report on the scien- tific results of the voyage of H.M.S. Challenger. Zoology 22:1-335. Haedrich, R. L., and N. R. Henderson. 1974. Pelagic food of Coryphaenoides armatus, a deep ben- thic rattail. Deep-Sea Res. 21:739-744. Haedrich, R. L., and G Krefft. 1978. Distribution of bottom fishes in the Denmark Strait and Irminger. Deep-Sea Res. 25:705-720. Haedrich, R. L., and P. T. Polloni. 1976. A contribution to the life history of a small rattail fish, Coryphaenoides carapinus. Bull. South. Calif. Acad. Sci. 75:203-211. Haedrich, R. L., G. T. Rowe, and P. T. Polloni. 1975. Zonation and faunal composition of epibenthic popula- tions on the continental slope south of New England. J. Mar. Res. 33:191-212. IWAMOTO, T. 1970. The R/V "Pillsbury" deep sea biological expedition to the Gulf of Guinea, 1964-65. Macrourid fishes of the Gulf of Guinea. Stud. Trop. Oceanogr. Miami 4:316-431. Iwamoto, T, and D L. Stein. 1973. A systematic review of the rattail fishes (Macrouridae: Gadiformes) from Oregon and adjacent waters. Occas. Pap. Calif. Acad. Sci. No. Ill, 79 p. Koblentz-Mishke, O. J., V. V. Volkovinsky, and J. G. Kabanova. 1970. Plankton primary production of the world ocean. In W. S. Wooster (editor), Scientific exploration of the South Pacific, p. 183-193. Natl. Acad. Sci., Wash., DC. Kulikova, E. B. 1957. Growth and age of deep water fishes. In B. N. Nikitin (editor), Marine biology, p. 284-290. Tr. Inst. Okeanol. Akad. Nauk. SSSR 20. [Engl, trans., Am. Inst. Biol. Sci. 1959.] Leim, A. H., and W. B. Scott. 1966. Fishes of the Atlantic Coast of Canada. Fish. Res. Board Can., Bull. 155, 485 p. Marshall, N. B. 1965. Systematic and biological studies of the Macrourid fishes (Anacanthini-Teleostii). Deep-Sea Res. 12:299-322. 1971. Explorations in the life of fishes. Harv. Univ. Press, Camb., 204 p. Marshall, N. B., and T. Iwamoto. 1973. Family Macrouridae In Fishes of the western North Atlantic Mem. Sears Found. Mar. Res., No. 1, part 6, p. 496-662. Macpherson, E. 1979. Ecological overlap between macrourids in the western Mediterranean Sea. Mar. Biol. (Berl.) 53:149-159. McEachran, J. D, and J. Davis. 1970. Age and growth of the striped sea-robin. Trans. Am. Fish. Soc 99:343-352. McLellan, T. 1977. Feeding strategies of the macrourids. Deep-Sea Res. 24:1019-1036. Merrett, N. R. 1978. On the identity and pelagic occurrence of larval and juvenile stages of rattail fishes (family Macrouridae) from 60°N, 20°W and 53°N, 20°W. Deep-Sea Res. 25:147-160. Merrett, N. E., and N. B. Marshall. 1981. Observations on the biology of deep-sea bottom-living fishes collected off northwest Africa (08°-27°N). Prog. Oceanogr. 9:185-244. Musick, J. A. 1969. The comparative biology of two American Atlantic hakes, Urophycis chuss and U. tennuis (Pisces: Gadidae). Ph.D. Thesis, Harvard Univ., Cambridge, 150 p. Musick, J. A., and K. J. Sulak. 1979. Characterization of the demersal fish community of a deep-sea radioactive dump site Contract Report submitted to the U.S. Environmental Agency. Va. Inst. Mar. Sci., 61 p. Musick, J. A., C. A. Wenner, and G. R. Sedberry. 1975. Archibenthic and abyssobenthic fishes. In May 1974 baseline investigations of deep water dumpsite 106 NOAA evaluation. Rep. 75-1, 388 p. NODZYNSKI, J., AND C. ZUKOWSKI. 1971. Biological and technological characteristics of grenadier family fishes (Macrouridae) of the Northwestern Atlantic Studia I Materially Series D, No. 5, p. 1-45. Okamuro, O. 1970. Fauna Japonica. Macrourina (Pisces). Acad. Press, Tokyo, 179 p. Pearcy, W. G 1975. Pelagic capture of abyssobenthic macrourid fish. Deep- Sea Res. 23:1065-1066. Pearcy, W. G, and J. W. Ambler. 1974. Food habits of deep-sea macrourid fishes off the Oregon Coast. Deep-Sea Res. 21:745-759. Pearcy, W. G., D. Stein, and R. Carney. 1982. The deep-sea benthic fish fauna of the northeastern Pacific Ocean on Cascadia and Tufts Abyssal Plains and ad- joining continental slopes. Biol. Oceanogr. 1:375-428. Phleger, G. F 1971. Biology of macrourid fishes. Am. Zool. 11:419-423. Pokdrazhanskaya, S. G. 1967. Feeding of Macrurus rupestris in the Iceland area. Ann. Biol. 24:197-198. 1971. Feeding and migrations of the roundnose grenadier, Macrourus rupestris, in the Northwest Atlantic and Icelandic waters. Int. Comm. Northwest Atl. Fish., Redb. Part III, p. 115-123. Polloni, P., R. Haedrich, G. Rowe, and C. H. Clifford. 1979. The size-depth relationship in deep ocean animals. Int. Rev. Gesamten Hydrobiol. 64:39-46. Rannou, M. 1975. Donnees nouvelles sur 1-activite reproductrice cyclique des poissions benthiques bathyaux et abbyssaux. Acad. Sci. Paris, T281, Ser. D, p. 1023-1025. 1976. Age et croissance d'un poisson bathyal: Nezumia sclerorhynchus (Macrouridae Gadiformes) de la mer 61 FISHERY BULLETIN: VOL. 84, NO. 1 d'Alboran. Cah. Biol. Mar. 17:413-421. Rannou, M., and C. Thiriot-Quievreuax. 1975. Structure des otolithes D'un Macrouridae (poisson Gadi- forme) Bathyal. Etude au microscope Electronique a Balayage Ann. Inst. Oceanogr., Paris 51:195-201. Sawatimskii, P. I. 1971. Studies of the age and growth of roundnose grenadier (Macrourus rupestris) in the North Atlantic 1967-1970. Int. Comm. Northwest Atl. Fish., Redb. Part III, p. 125-138. 1972. 0 vozraste tuporylogo makrurusa severo-zapadnoi Atlantiki i vozmozhnom vliyanii promysla no ego chislennost' (The age of the rock grenadier in the northwest Atlantic and a possible influence of fisheries on its population numbers). Tr. Polyarn. Nauchno-Issled. Proektn. Inst. Morsk. Rybn. Khoz. Okeanogr. (PINRO) 28:116-127. (Engl, transl., Fish. Res. Board Can., Transl. Ser. 2491.) Smith, K. L., Jr. 1978. Metabolism of the abyssopelagic rattail Coryphaenoides armatus measured in situ. Nature (Lond.) 274:362-364. Smith, K. L., Jr., G. A. White, M. B. Laver, R. R. McConnaughey, and J. P. Meador. 1979. Free vehicle capture of abyssopelagic animals. Deep- Sea Res. 26A:57-64. SOKAL, R. R., AND F. J. ROHLF. 1969. Biometry: The principles and practice of statistics in biological research. W. H. Freeman and Co., San Franc, 776 P- Stanek, E. 1971. Badania zasobow rybnych wod Labradora i Nowej Fundlandii (Studies on the fish stocks in the waters off Labrador and Newfoundland). Pr. Morsk. Inst. Rybackiego, p. 121-146. (Engl, trans., Fish. Res. Board Can., Transl. Ser. 2754.) Webster, F. 1969. Vertical profiles of horizontal ocean currents. Deep- Sea Res. 16:85-98. Wenner, C. A., and J. A. Musick. 1979. Biology of the morid fish Antimora rostrata in the western North Atlantic J. Fish. Res. Board Can. 34:2362- 2368. WORTHINGTON, L. V. 1964. Anomalous conditions in the slope water area in 1959. J. Fish Res. Board Can. 21:327-333. Zakharov, G. P., and I. D. Mokanu. 1970. Distribution and biological characteristics of Macrourus rupestris of the Davis Strait in August-September 1969. Reports of Pinro Marine Expeditionary Investigations, 3rd cruise of R/V Perseus III. 62 DIFFERENTIATION OF PRIONOTUS CAROLINUS AND PRIONOTUS EVOLANS EGGS IN HEREFORD INLET ESTUARY, SOUTHERN NEW JERSEY, USING IMMUNODIFFUSION Walter J. Keirans,1 Sidney S. Herman,2 and R. G. Malsberger2 ABSTRACT Immunochemical techniques were used to classify the planktonic eggs of Prionotus carolinus (northern searobin) and Prionotus evolans (striped searobin) collected from a southern New Jersey estuary. Results of immunochemical identifications were compared with identifications based upon the commonly used morphological character of egg oil globule distribution. An average identification error of 22.3% was found when results using this conventional morphological characteristic were compared with immunodiffusion results. Improved accuracy of searobin egg identification can be achieved in future ichthyoplankton studies by using immunochemical techniques. A similar application of immunochemical identification techniques should also better resolve classification uncertainties among other morphologically similar co-temporal and co-spatial planktonic fish eggs. The accuracy of ichthyoplankton analysis is often limited by the lack of reliable, distinguishing, mor- phological characteristics that are useful for identi- fying fish eggs and larvae. Conventional character- istics used to identify fish eggs include egg and oil globule diameters; number, distribution, and pigmen- tation of oil globules; and pigmentation patterns on developing embryos. However, overlapping diameters of eggs and a similar if not identical number of oil globules with comparable pigmentation and size among closely related species impose a relatively high degree of uncertainty concerning the identity of planktonic fish eggs from many areas. Increased accuracy has been more recently achieved through the analysis of fish eggs using biochemical, im- munological, and ontogenetic methods. Morgan (1975) examined electrophoretic patterns of white perch and striped bass egg extracts and found dif- ferentiation was possible on this basis. Orlowski et al. (1972) differentiated cunner, Tautogolabrus ad- sperus, from tautog, Tautoga onitis, eggs using monospecific antisera in microimmunodiffusion analyses. The technique was especially useful with early stage eggs which were morphologically iden- tical. Ontogenetic methods allow careful study of laboratory-reared eggs and larvae of known paren- tage to document species-specific developmental histories. These studies may provide new distin- 1 Department of Biology, Lehigh University, Bethlehem, PA; pres- ent address: E. I. du Pont de Nemours Co., Inc., Glasgow Research Laboratory, Wilmington, DE 19898. department of Biology, Lehigh University, Bethlehem, PA 18015. guishing morphological features for future egg iden- tifications. However, additional means are required where well-documented features shared with other species do not provide adequate differentiation of field-collected eggs. This paper is a report on the results obtained from a microimmunodiffusion analysis which successful- ly differentiated the planktonic eggs of the north- ern searobin, Prionotus carolinus, from those of the striped searobin, Prionotus evolans, which were col- lected from the Hereford Inlet estuary, southern New Jersey, between May 1973 and September 1974 (Keirans 1977). Identifications based separately upon immunochemical and morphological evidence were also compared to evaluate the reliability of differen- tiations based entirely upon conventional mor- phology. Prionotus spp. were selected in our study first because the searobins represent a large breeding population which appears co-temporally and co-spatially near shore to provide an abundant source of gravid adults. Eggs of known parentage became readily available for preparation of ex- perimental reagents and specimens. Secondly, this study would expand the application of microimmuno- diffusion analysis to species differentiation as an ex- tension of the study of Orlowski et al. (1972), which documented differentiation of eggs from two genera. Finally, the identification of Prionotus spp. ova has never been properly resolved. Prionotus carolinus ova were described by Kuntz and Radcliffe (1918) as highly transparent but slight- ly yellowish spherical eggs ranging from 1.0 to 1.15 mm in diameter. The yolk sphere contained a Manuscript accepted March 1985. FISHERY BULLETIN: VOL. 84. NO. 1. 1986. 63 variable number of 10 to 25 unequal-sized oil globules scattered over the yolk surface which showed some tendency toward aggregation with progressing development. The diameter range was extended from 0.94 mm to 1.20 mm by Bigelow and Schroeder (1953) and Wheatland (1956), respectively. The up- per diameter limit extension was verified by Her- man (1963). Prionotus evolans ova have never been positively identified. Perlmutter (1939) made a ten- tative identification, later accepted by Marshall (1946), from ripe ova stripped from gravid females collected in Long Island Sound and described as having similar appearance and diameter as northern searobin eggs, but with oil globules clustered at one pole rather than dispersed across the yolk sphere surface This singular observed morphological dif- ference of oil globule distribution pattern has beer used as the primary distinguishing characteristic between ova of Prionotus carolinus and Prionotus evolans. MATERIALS AND METHODS Conventional Identifications Field-collected, buffered Formalin3-preserved plankton samples were physically sorted for all ichthyoplankton using forceps under a dissecting microscope, and the criterion of oil globule distribu- tion differences established by Perlmutter (1939) was used to tentatively separate P. carolinus from P. evolans eggs. The annual cycle and species composi- tion aspects of the field-collected samples using con- ventional means for egg and larval identifications have been submitted elsewhere for publication. 3Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. FISHERY BULLETIN: VOL. 84, NO. 1 Immunochemical Identifications Antigens and Immunizations Antigen preparations from both species of searobin eggs were generated using the techniques developed by Orlowski et al. (1972) with ovarian tissue from ripe adults and immature individuals. The four antigen preparations presented in detail in Table 1 were each used to elicit immune responses in at least two New Zealand white rabbits to improve the probability of obtaining useful antisera. Preim- mune serum samples were obtained from each animal to establish that no reactivity with antigen existed prior to immunization. The soluble protein antigens of Prionotus evolans (PeSP) and Prionotus carolinus (PcSP) were injected intravenously in 4.7 and 4.8 mg protein doses (stan- dard biuret analysis), respectively, to begin the im- munization program. Maintenance injections of 2 mg protein followed on a weekly basis. Blood samples were obtained by cardiac puncture 3 wk following the first injection and the presence of precipitating antibody was demonstrated by the standard precip- itin ring test (Abramoff and LaVia 1970). Additional monthly cardiac puncture samples were monitored by quantitative double diffusion (Feinberg 1957) until after about 12 wk; a titer of 32 was reached in all animals receiving soluble antigens when sera were tested with 40 ^g homologous antigen. Particulate protein antigens from macerated ovarian tissue of northern (PcPP) and striped (PePP) searobins were prepared in a 1:1 emulsion with Freund's complete adjuvant (Cappell Laboratories). PcPP (8 mg) and PePP (10 mg) protein preparations were injected subcutaneously along several bilateral dorsal sites on New Zealand white rabbits. Rabbits injected with Freund's complete adjuvant developed Table 1. — Antigen characterization and nomenclature. Antigen source and Range protein concentration Method of Immunization route and dose Titer Double- Complement Species designation (mg/mL) determination Initiation Maintenance diffusion fixation Prionotus carolinus Mature ova 8-15 Biuret Intravenous Intravenous 32 Northern searobin (PcSP) (4.7 mg) (2mg) Immature 15-40 Microkjeldahl Subcutaneous Intravenous 1,280 follicular (8mg) (2mg) material (PcPP) Prionotus evolans Mature ova 8-15 Biuret Intravenous Intravenous 32 Striped searobin (PeSP) (4.8 mg) (2mg) Immature 15-40 Microkjeldahl Subcutaneous Intravenous 1,280 follicular (10 mg) (2mg) material (PePP) 64 KEIRANS ET AL.: DIFFERENTIATION OF PRIONOTUS EGGS Arthus reactions following a single dose Subsequent injections were accomplished intravenously using Millipore (0.45 /mi) filtrates of PcPP and PePP. Titers were monitored utilizing the standard complement fixation assay because of the particulate consisten- cy of the macerated antigen preparation (Kabat and Mayer 1961). Maximum titers of 1,280 were obtain- ed after about 10 wk with immunizations using PcPP or PePP. Antiserum Specificity Antisera elicited in response to both soluble and particulate antigens were multicompetent and ex- hibited cross-reactions with heterologous antigens. The presence of common antigens between the northern searobin and striped searobin ovarian material preparations required the specific adsorp- tion of antisera with these shared antigens to render a given antiserum monospecific (Eisen 1974). Although antisera elicited in response to particulate protein antigens exhibited precipitation reactions in agar with both soluble antigens and extracts of par- ticulate antigens from the two species under con- sideration, they were not competent in reactions with homologous fish eggs. Therefore, since the selected method for analysis of planktonic eggs was immuno- diffusion, only antisera elicited in response to solu- ble antigens were used in all analyses of unknowns. Specific adsorption of common antigens shared by northern and striped searobins was accomplished by adding PcSP to antisera elicited in response to PeSP and vice versa. Adsorption lots of 1.5 mL anti-PeSP antisera combined with 70 \xL PcSP (0.65 mg pro- tein) were incubated at 4°C for 48 h prior to use. This adsorption eliminated all reactivity of anti-PeSP antisera with both PcSP and known ova of P. carolinus, without significantly reducing activity with ova of P. evolans. This specifically adsorbed anti- PeSP, which reacted solely with known homologous ova of P. evolans under controlled conditions, was used as the basis for differentiation of northern and striped searobin eggs. Species-specific anti-PeSP antisera capable of 100% accuracy in differentiating known ova of both searobins was the reagent selected for use in all immunodiffusion analyses. Microimmunodiffusion Analysis Unknown planktonic fish eggs were analyzed with monospecific anti-PeSP antiserum in a micromodi- fication of the immunodiffusion technique (Ridgeway et al. 1962). Microscope slides (2.5 x 8 cm) were washed, rinsed first in distilled water and then methanol, and wiped dry. Two milliliters of 1% No- ble Agar (Difco) in FA-Bacto buffer (Difco), pH 7.2, were applied across each slide on a leveling table and allowed to harden. Slides were then placed over a template and wells cut using a Brewer needle with beveled inner surface (Ridgeway et al. 1962). Agar plugs were removed from wells by aspiration. Reagents were applied with either 1 mL syringes (Burron) or sterile capillary pipettes, and 0.005 to 0.01 mL was required to fill each well. A typical testing array appears in Figure 1, where corner wells contain unadsorbed antiserum, the central well con- tains adsorbed or monospecific antiserum, and re- maining wells contain individual fish eggs which have been broken using jeweler's forceps. FA-Bacto buf- fer was applied to each well following egg disrup- tion, and slides were allowed to incubate in moist chambers for 18 h at 20°C. Slides were then washed for 24 h in FA-Bacto buffer, and stained according to the method of Crowle (1958). Results were always recorded at a fixed time interval following slide preparation to insure comparability from one deter- mination to another. RESULTS AND DISCUSSION A total of 732 searobin ova were recovered from plankton samples collected in the 1973-74 period. The combined morphological characteristics of egg diameter, number, color, and distribution of oil globules, and embryo pigmentation when present, allowed the separation of searobin eggs from those of other species with reasonably high confidence Preliminary classifications of Prionotus ova into either evolans or carolinus species was based upon differential oil globule distribution patterns reported by Perlmutter (1939). Striped searobin, P. evolans, eggs were placed into one grouping based upon a polar or clustered oil globule distribution, and north- ern searobin, P. carolinus, eggs placed into a second group having oil globules generally dispersed across the yolk sphere Each tentatively classified egg was then analyzed in the microimmunodiffusion method illustrated in Figure 1, to establish the immunochemical reactivity of soluble egg antigens with adsorbed and unadsorb- ed anti-PeSP antisera. When soluble P. evolans egg antigens were sufficiently concentrated, a classical line of identity was observed with fusion of precipitin bands between adsorbed and unadsorbed anti-PeSP wells. Identification of P. carolinus eggs was based upon reactivity with unadsorbed anti-PeSP anti- serum and no reactivity with adsorbed anti-PeSP. Previously established reactivity of unadsorbed anti- 65 FISHERY BULLETIN: VOL. 84, NO. 1 Figure 1— Testing array (lOx). C: Prionotus carolinus ovum (1.00 mm); E: Prionotus evolans ovum (1.00 mm); AA: Anti-PeSP antiserum (adsorbed: 0.20 mL antiserum: 0.11 mg PcSP protein); AN: Anti-PeSP antiserum (unadsorbed). Specific adsorption of cross-reactive antibodies has occurred with PcSP rendering anti-PeSP antiserum (AA) incompetent to react with antigens of Prionotus carolinus ova (C), indicated by the lack of precipitin bands about the central well adjacent to (C) egg wells. Corner wells contain multicompetent, unadsorbed anti-PeSP antisera. PeSP with known P. carolinus eggs was considered sufficiently definitive for its use in differentiating P. carolinus from P. evolans ova. The immunochemical classifications derived from this analysis indicated that an average 22.3% mis- classification error had been made when eggs were differentiated solely on the basis of oil globule distributions. An approximately equal number of both northern and striped searobin eggs had been mistakenly identified, based upon oil globule distribution patterns. The final classification based upon immunochemical data was 406 ova of P. carolinus and 32G ova of P. evolans. It was confirmed that egg diameters could not serve as a reliable characteristic for species classi- fications by retrospectively analyzing diameters of immunochemically classified eggs according to the period of field collection. The data presented in Table 2 illustrate that no statistical difference exists in the diameter ranges of P. carolinus and P. evolans eggs for the collection period of this study. However, the trend of declining egg diameters over the spawning season previously documented by other workers is Table 2.— Immunochemical classification of Prionotus spp. eggs collected in plankton samples. Average diameter Range Date (mm) (mm) n Prionotus carolinus 1973 May 1.16 1 .02-1 .24 4 June 1.06 1.00-1.21 10 July 1.08 1.05-1.10 3 August 1.02 0.92-1.18 312 September 0.99 0.90-1.05 32 1974 July 0.98 0.95-1.02 13 August 0.96 0.92-1.02 3 September 0.99 0.92-1.02 29 Prionotus evolans 1973 May 1.12 1.00-1.25 10 June 1.06 1.00-1.12 35 July 1.08 1.00-1.15 2 August 1.03 0.95-1.12 225 September 0.98 0.90-1.08 26 1974 July 0.97 0.95-1.00 6 August 1.02 1.02 1 September 0.99 0.92-1.02 21 66 KEIRANS ET AL.: DIFFERENTIATION OF PRIONOTUS EGGS confirmed. The data also show that in 1973 and 1974, the ratios of eggs collected in plankton samples and identified based upon morphology and immuno- chemical reactions for nothern and striped searobins were 1.1:1 and 1.6:1, respectively. These ratios are similar in magnitude to the ratio of northern and striped searobin adults observed by Marshall (1946). Finally, the data indicate that egg diameter and oil globule distribution cannot serve to reliably dis- tinguish northern from striped searobin eggs. An immunochemical distinction can be made that sug- gests morphology alone is inadequate to provide a positive identification of P. evolans eggs. The course of future research in immunochemical taxonomy of fish eggs should emphasize an increase in sensitivity, as well as automation of the analysis. At present, the utility of the immunodiffusion method is limited by its labor-intensive nature. Ini- tial stages of the analysis require manual sorting of ova from plankton samples that is tedious, time- consuming, and subject to error. Bowen et al. (1972) initiated studies in which a moderate degree of suc- cess was achieved in sorting fish ova from pelagic plankton samples on sucrose density gradients. However, estuarine plankton samples that contained a wide range of particulate materials characterized by different sizes, densities, and shapes, and that also included high levels of detrital materials, disturbed the gradients sufficiently to destroy separation potential. Despite the recognized limitations, there is currently no practical alternative to manual sort- ing of plankton samples. Immunodiffusion analysis requires that individual fish eggs be subjected to several manual manipula- tions, with the final determination in solid media re- quiring the careful applications of reagents. Screen- ing large numbers of planktonic ova with several dif- ferent antisera becomes impractical on a large scale. A more rapid and potentially more specific approach to immunochemical ichthyoplankton identifications might employ monoclonal antibodies coupled to fluorescent indicator molecules. The antibody prod- ucts of fused mouse lymphocytes and myeloma cells may be screened and selected for exquisite specificity to single antigenic determinants or epitopes using egg antigens of known origin, preferably those associated with the chorion surface, to procure a reagent that would specifically label ova without re- quiring that each egg be mechanically ruptured. Identifications might be based upon the differential fluorescence characteristic of a particular fluo- rescent label associated with a selected antibody and labelled eggs might be isolated using a fluorescence- activated cell sorter. The utility of immunochemical identifications with demonstrably superior accuracy to conventional methods has been established with both intergeneric and interspecific differentiations. Several systems remain which might benefit from immunochemical differentiations, such as the complete elucidation of several sciaenid and clupeid species which occur in complex estuarine systems, such as the Chesapeake Bay and Potomac River estuary. Relationships be- tween scombrids, bothids, and pleuronectids with more southerly distributions would serve to delineate adult ratios, population distributions, and spawning seasons. Finally, the capability of the immune system to differentiate among epitopes with relatively small structural difference (Karush 1962) might eventually be applied to the detection of racial differences or subpopulation distinctions among fish ova of the same species. ACKNOWLEDGMENTS The Noyes Foundation provided fellowship funds for the senior author. Michael Criss and Marian Glaspey assisted in collecting the samples. LITERATURE CITED Abramoff, P., and M. LaVia. 1970. Biology of the immune reponse McGraw-Hill, Inc., N.Y., 492 p. BlGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Gulf of Maine U.S. Fish Wildl. Serv., Fish. Bull. 53:1-577. Bowen, R. A., J. M. St. Onge, J. S. Colton, Jr., and C. A. Price. 1972. Density-gradient centrifugation as an aid to sorting planktonic organisms. I. Gradient materials. Mar. Biol. (Berl.) 14:242-247. Crowle, A. J. 1958. A simplified micro double-diffusion agar precipitin tech- nique J. Lab. Clin. Med. 52:784-787. Eisen, H. N. 1974. Immunology: An introduction to molecular and cellular principles of the immune responses. Med. Dep., Harper & Row, Inc., Hagerstown, MD, 624 p. Feinberg, J. G. 1957. Identification, discrimination, and quantification in Ouchterlony gel plates. Int. Arch. Allergy 11(3-4):129-152. Herman, S. S. 1963. Planktonic fish eggs and larvae of Narragansett Bay. Limnol. Oceanogr. 8:103-109. Rabat, E. A., and M. M. Mayer. 1961. Complement and complement fixation. Experimental immunochemistry. 2d ed. Chase Thomas Publ., Spring- field, IL, 905 p. Karush, F. 1962. Immunologic specificity and molecular structure Adv. Immunol. 2:1-40. Keirans, W. J., Jr. 1977. An immunochemically assisted ichthyoplankton survey with elaboration on species specific antigens of fish egg 67 FISHERY BULLETIN: VOL. 84, NO. 1 vitellins; southern New Jersey Barrier Island - Lagoon Com- plex. Ph.D. Thesis, Lehigh Univ., Bethlehem, PA 169 p. KUNTZ, A., AND L. RADCLIFFE. 1918. Notes on the embryology and larval development of twelve teleostean fishes. Bull. U.S. Bur. Fish. 35:89-134. Marshall, N. 1946. Observations on the comparative ecology and life history of two sea robins, Prianotus carolinus, and Prionotus evolans strigatus. Copeia 1946:118-144. Morgan, R. P., II. 1975. Distinguishing larval white perch and striped bass by electrophoresis. Chesapeake Sci. 16:68-70. Orlowski, S. J., S. S. Herman, R. G. Malsberger, and H. N. Pritchard. 1972. Distinguishing cunner and tautog eggs by immunodiffu- sion. J. Fish. Res. Board Can. 29:111-112. Perlmutter, A. 1939. A biological survey of the salt waters of Long Island, 1938. Section I. An ecological survey of young fish and eggs identified from tow-net collection. Suppl. 28th Annu. Rep. N.Y. Conserv. Dep. Part 11:11-71. RlDGEWAY, G. J., G. W. Klontz, and C. Matsumcto. 1962. Intraspecific differences in serum antigens of red salmon demonstrated by immunochemical methods. Int. North Pac Fish. Comm. 8:1-13. Wheatland, S. B. 1956. Oceanography of Long Island Sound. 1952-1954. VII. Pelagic fish eggs and larvae. Bull. Bingham Oceanog. Coll., Yale Univ. 15:234-314. 68 EFFECTS OF EXPOSURE AND CONFINEMENT ON SPINY LOBSTERS, PANULIRUS ARGUS, USED AS ATTRACTANTS IN THE FLORIDA TRAP FISHERY John H. Hunt,1 William G. Lyons,2 and Frank S. Kennedy, Jr.2 ABSTRACT Traps in the south Florida spiny lobster fishery are baited with live sublegal-sized lobsters (shorts), many of which are exposed for considerable periods aboard vessels before being placed in traps and returned to the sea. Average mortality rate of lobsters exposed Vz, 1, 2, and 4 hours in controlled field tests was 26.3% after 4 weeks of confinement. About 42% of observed mortality occurred within 1 week after ex- posure, indicating exposure to be a primary cause of death. Neither air temperature during exposure nor periodic dampening with seawater had significant effects on mortality rate Mortality among confin- ed lobsters increased markedly in the Atlantic oceanside but not in Florida Bay during the fourth week of confinement following exposure, probably because more natural food organisms entering traps from nearby seagrass beds delayed starvation at the latter site. Mortality caused by baiting traps with shorts may produce economic losses in dockside landings estimated to range from $1.5 to $9.0 million annually. The fishery for spiny lobster, Panulirus argus, in south Florida utilizes a method of baiting traps that is apparently unique among fisheries worldwide Sublegal [<76 mm carapace length (CL)] lobsters, locally called "shorts", are confined in traps as living attractants for legal-sized lobsters. Shorts have been demonstrated to be effective attractants of other lobsters (Yang and Obert 1978; Lyons and Kennedy 1981; Kennedy 1982). Some use of shorts as bait in the Florida fishery occurred as early as the 1950's (Cope 1959), but use increased appreciably after 1965 when the minimum legal size was reduced from 1 lb (about 79-80 mm CL) to 3 in (76 mm) CL, and the fishery expanded from Atlantic oceanside reefs and flats into Florida Bay where availability of shorts is considerably greater (Lyons et al. 1981). The practice was widespread but illegal during early years of its use (Wolff erts 1974) and only received legal sanction in 1977. Today, bonded fishermen are allowed to possess as many as 200 shorts aboard a vessel for use as bait. Shorts are customarily kept in wooden boxes on deck until replaced in traps, and exposure times vary from several minutes to 1 h or more As many as 1 million shorts may be confined in traps as bait during peak portions of the harvest season (Lyons and Kennedy 1981). 'Florida Department of Natural Resources, Bureau of Marine Research, Marathon, FL 33050. 2Florida Department of Natural Resources, Bureau of Marine Research, St. Petersburg, FL 33701. During 1979, the Florida Department of Natural Resources initiated a study in which baiting prac- tices in the fishery were mimicked under controlled conditions to determine whether starvation occurred among lobsters confined in traps for long periods. So much mortality occurred among tested lobsters during the first 2 wk of confinement that the study was redirected toward causes of that mortality. Ex- posure was strongly implicated by preliminary results (Lyons and Kennedy 1981). Spokesmen for the fishing industry suggested that observed mor- tality was caused by other factors related to ex- perimental design, prompting expansion of the pro- gram to test those factors. This report presents results and conclusions from that expanded program. The relationship between exposure and mortality is examined, including in- fluences of season and location. Mortality rates of lobsters held dry or periodically dampened prior to placement in traps are also compared. Results from this study are used in a model which estimates the relative importance of baiting mortality to economics of the fishery. METHODS Mortality rates of spiny lobsters used to bait traps were measured in Florida Bay 3 km north of Vaca Key and in the Atlantic Ocean 6 km south of Vaca Key. The Florida Bay site was located in shallow water (~3 m) with a muddy sand substrate overlain by seagrass beds. The ocean site was located in Manuscript accepted March 1985. FISHERY BULLETIN: VOL. 84, NO. 1. 1986. 69 FISHERY BULLETIN: VOL. 84, NO. 1 deeper water (~8 m) just inside the reef tract; the bottom consisted of a mosaic of scattered seagrasses, small patch reefs, and open areas of coarse sand. Salinities at both sites ranged from 34%o to 36%o and water temperature ranged seasonally from 17° to 29°C. The effect of exposure was examined at both sites. Lobsters were held in shaded boxes for lk, 1, 2, and 4 h and then placed in traps. Entrances were sealed, and no lobsters were added after treatments were established. Each treatment utilized 5 standard wooden slat lobster traps; each trap contained 3 lobsters (total 15 lobsters/ treatment) for each ex- posure period. Control treatments (minimum ex- posure) also consisted of 5 traps each containing 3 lobsters, but these lobsters remained in traps in which they were originally captured and were ex- posed only for the time required to clean, seal, and return a trap to the water. Intent was to place sublegal lobsters in all traps, but use of some larger lobsters was necessary to conduct experiments. Traps in oceanside experiments were reinforced with wire mesh sides to reduce damage by loggerhead turtles, Caretta caretta; traps in Florida Bay were not reinforced with wire sides. In Florida Bay, all lobsters exposed >1 h were dampened every xk h by pouring a bucket of seawater into the porous holding box, whereas equal numbers of lobsters exposed >1 h in oceanside tests were always treated with and without seawater dampen- ing every V2 h to test the effect of dampening. Con- trol and V2-h treatments were the same in dampened (wet) and undampened (dry) tests because their total exposure periods were less than or equal to the period between dampenings. After initiation, all experiments were sampled at 1-wk intervals for 4 wk by pulling each trap and counting remaining live lobsters. The mortality estimate is a combination of missing lobsters and those observed to be dead. Several lines of evidence indicate that missing lobsters died and did not escape Only lobsters too large to fit between trap slats were used in experiments, and trap entrances were boarded shut to seal the ordinary avenue of departure Additionally, observations made during frequent dives at traps where lobsters died during other experiments indicated that carcasses could be broken up sufficiently by scavengers within 24 h after death to wash through slats when traps were pulled. All original data, taken as number of living lobsters remaining in a trap each week, were con- verted to weekly mortality rates calculated as the number of lobsters that died during that week divid- ed by the initial density during that week. This method provided the only independent, non- cumulative estimate of mortality. All other methods biased the data by either increasing the weight given to deaths later in the experiment or altering mor- tality estimates because of trap losses. Although this method provided unbiased estimates of mortality, data still were not normally distributed, so all testing of treatment means used nonparametric Wilcoxon Two Sample Tests (Sokal and Rohlf 1969) to deter- mine where the differences of significance occurred. Standard notations are used to designate signi- Table 1.— Average weekly spiny lobster mortality (%) for each location, exposure period, and wet or dry treatment. N = number of traps; x = mean; SE = standard error; W = wet; D = dry. Initial N Week after initial exposure Week 1 Week 2 Week 3 Week 4 Cumulative mortality % Treatment N X SE N X SE N X SE N X SE Florida Bay Control 15 15 0.0 0.0 15 0.0 0.0 15 2.2 2.2 15 0.0 0.0 2.2 V2 h 20 20 8.3 5.3 19 3.5 3.5 18 0.0 0.0 17 0.0 0.0 11.8 1 h W 20 17 7.8 3.5 17 3.9 3.9 16 6.2 3.4 16 6.2 6.2 24.1 2 h W 20 18 14.8 5.5 18 1.8 1.8 18 1.8 1.8 18 3.7 2.5 22.1 4 h W 20 20 15.0 5.6 19 5.3 2.9 19 5.3 2.9 18 0.0 0.0 25.6 Atlantic Reef Control 29 28 4.8 2.8 23 1.4 1.4 23 0.0 0.0 27 7.4 3.2 13.6 1/2 h 29 29 8.0 3.6 24 1.4 1.4 23 4.3 4.3 27 12.3 4.8 26.0 1 h W 29 29 16.1 4.8 24 9.7 3.7 19 7.0 4.1 24 12.5 5.2 45.3 D 29 29 11.5 3.8 24 9.7 5.1 22 4.5 2.5 27 11.1 5.3 36.8 2 h W 29 29 13.8 5.1 17 3.9 2.7 15 4.4 3.0 20 5.0 2.7 27.1 D 29 29 16.1 5.4 23 5.8 2.7 22 4.5 2.5 24 5.6 3.3 32.0 4 h W 29 29 12.6 3.8 23 4.3 3.2 19 8.8 6.2 22 6.1 2.8 31.8 D 29 29 11.5 4.1 21 7.9 4.5 18 1.8 1.8 23 1.4 1.4 22.6 70 HUNT ET AL.: EXPOSURE AND CONFINEMENT ON SPINY LOBSTERS ficance at probability levels of 0.05, 0.01, and 0.001. Weighted cumulative average mortality values were obtained by multiplying the relative effort (%) in each treatment (eg, site, exposure period >Vz h) by the cumulative mortality for that treatment and then summing those values. RESULTS The mortality experiment was conducted four times between January and September 1980 in Florida Bay and six times between May 1981 and June 1982 near Atlantic reefs. Wet vs. dry tests were conducted with each oceanside replicate The un- weighted average cumulative mortality calculated from Table 1 for all lobsters exposed lk, 1, 2, and 4 h, both sites combined, was 26.3% at the end of 4 wk. Average weighted cumulative mortality in Florida Bay was 20.8%, and that near Atlantic reefs was 31.9%. When weighted for relative effort at each site, the overall mortality rate increased to 28.5%. No tests were established at oceanside stations during December, January, or February, so effects of air and water temperatures on mortality during exposure were tested only in Florida Bay. Of four tests conducted there, two were established during cool months (January, February; air 15.2°-21.0°C, water 17.0°-17.5°C during initiation), and two were established during warm months (May September; air 27.6°-33.5°C, water 29.3°-29.5°C). Mean week- ly mortality rates of lobsters during these tests (winter x = 4.4%; summer x = 4.6%) were not sig- nificantly different. Average mortality rates obtained in wet vs. dry treatments (Table 1, Fig. 1) were not significantly different for any exposure or subsequent confine- ment period. Furthermore, neither wet nor dry treat- ments consistently caused greater mortality. Because all Florida Bay lobsters were dampened when exposed >1 h, comparisons of bay vs. ocean mortality rates were made using wet treatments only. All five treatments (Control, V2, 1, 2, and 4 h) were combined and overall mean weekly mortality rates were compared. The average weekly mortality rate of lobsters in bay tests (x = 4.5%) differed significantly (Z = 2.51, P < 0.05) from that of lobsters tested in the ocean (x = 7.6%). 45 « 35 o 0 25 J2 E o 15 - Figure 1.— Cumulative mortality rates (%) for exposure tests: A. Florida Bay, wet only; B. Atlantic reefs, wet only; C. Wet (W) vs. dry (D), Atlantic reefs only. C = controls; exposure periods = Vz, 1, 2, and 4 h. 71 FISHERY BULLETIN: VOL. 84, NO. 1 Comparisons of each exposure period within a treatment with every other exposure period within that treatment are shown in Table 2. In the bay, mor- tality rates experienced by controls were significant- ly different than those of lobsters exposed 1, 2, or 4 h. Additionally, lobsters exposed V2 h suffered a significantly lower mortality rate than did those ex- posed 4 h. However, some of these differences were not significant among lobsters exposed at the Atlan- tic reef site Among dampened lobsters tested there, only the mortality rate of those exposed 1 h differed significantly from that of controls and from that of lobsters exposed V2 h. Among undampened lobsters tested at the ocean site, mean mortality rates of con- trols differed significantly only from those exposed 1 or 2 h. Differences between controls and 1 h ex- posures were significant in every treatment, but mean mortality rates never differed significantly among lobsters exposed 1, 2, or 4 h. The mean mortality rate of all tested lobsters dur- ing the first week following exposure was 11.2%, which represents about 42% of all mortality; 54% of all mortality in Florida Bay and 38% of all which took place near Atlantic reefs occurred during the first week (Table 1, Fig. 1). High mean weekly mor- tality rates which occurred during week 1 decreas- ed to much lower levels during week 2 (4.7%) and week 3 (3.9%) in both bay and ocean (Fig. 2). Com- parisons of mean mortality rates incurred during week 1 with those of weeks 2 and 3 revealed signifi- cant differences in every instance (Table 3). During week 4, the overall rate increased to 6.1% (Fig. 2), but this combined value masked highly divergent changes in rates of mortality at bay and ocean sites. Table 2.— Results of Wilcoxon Two Sample Tests (Z values) from comparisons of mean weekly mortality rates from dif- ferent exposure periods for various treatments at Florida Bay (Bay) and Atlantic Reef (Ocean) locations. C = con- trols; exposure = hours. Tests Exposure C Vz 1 2 4 Bay wet C — 1/2 1.14 — 1 2.48* 1.62 — 2 2.52* 1.68 0.02 — 4 2.93** 2.17* 0.51 0.49 — Ocean wet c — Vz 1.10 — 1 3.07** 2.02* — 2 1.87 0.81 1.17 — 4 1.93 0.85 1.16 0.03 — Ocean dry C — Va 1.10 — 1 2.20* 1.12 — 2 2.12* 1.03 0.10 — 4 1.17 0.08 1.01 0.92 — Bayside mortality rates actually decreased slightly, whereas oceanside rates increased dramatically. Statistical comparisons between mean mortality rates during weeks 1 and 4 demonstrate significant differences in the bay but not in the ocean (Table 3). Graphic depictions of cumulative weekly mortality rates (Fig. 1) reveal a decrease in slope after week 1 at both bay and ocean sites. These decreases in- dicate reduced rates of mortality which persist through the end of the experiment in the bay and through week 3 in the ocean. However, the slope in- creases sharply during week 4 in most oceanside tests, indicating an additional period of high mor- tality there. DISCUSSION Exposure unquestionably causes mortality among Panulirus argus used to bait traps. Increasing ex- Week B A D T 1 1 pg|:p:;::jjj:j::::::::::::::| B A D T 2 I!!!!!!!!!!!!!!l B A D T 3 |"E: ::J[r- B A D T 4 ::-:E=:::::::::::::::i:i:::3 , , .... 4 8 12 Percent Mortality • = P < 0.05; P*S 0.01; P< 0.001 Figure 2— Average weekly mortality rates (%) per treatment type during weeks 1-4, all exposures combined. A = oceanside (Atlan- tic Ocean) wet; B = bay (Florida Bay) wet; D = oceanside dry; T = all treatments combined. 72 HUNT ET AL.: EXPOSURE AND CONFINEMENT ON SPINY LOBSTERS posure periods up to 1 h resulted in corresponding increases in mortality. Similar mortality has been observed in the Western Australia spiny lobster (Panulirus cygnus) fishery (Brown and Caputi 1983; Brown et al. in press). In that fishery, undersize lobsters are not used as bait but are often retained aboard vessels for varying periods during the sort- ing process. Tb test effects of that practice, Austral- ian lobsters were tagged, held aboard vessels for 0, Vi, V2, 1, and 2 h, and then released. Recapture rates were markedly lower in exposed groups than in con- trols. As in our experiments, results from exposure times >1 h were similar to those of 1 h exposures. The greatest rate of mortality to Panulirus argus in our tests occurred during the first week follow- ing exposure (Fig. 2). Although physiological causes of mortality have not been determined, several fac- tors may be involved. Dehydration due to desicca- tion may affect survival, but lobsters dampened at V2 h intervals died at rates similar to those left un- attended. One effect of exposure is to dry gills (Anonymous 1980), which may result in respiratory problems. Dehydration and gill damage may cause mortality directly, but more likely are contributory factors to physiological stress caused by buildup of toxic compounds in the blood. Handling stress has been demonstrated to cause temporary acidic con- ditions in the blood of European lobsters, Homarus vulgaris (McMahon et al. 1978). After reimmersion in seawater, lobsters should rehydrate fairly quick- ly, but effects of physiological stress are likely to linger. Contrary to prior expectations, mortality rates of dampened lobsters did not differ significantly from those left unattended (dry). Dampening also failed to enhance survival of the northern lobster, Homarus americanus (McLeese 1965). McLeese suggested Table 3.— Results of Wilcoxon Two Sample Tests (Z values) from comparisons of mean weekly (1-4) mor- tality rates for various treatments at Florida Bay (Bay) and Atlantic Reef (Ocean) locations. Tests Week 1 2 3 4 Bay wet 1 2 3 4 2.86** 2.40* 3.58*** 0.55 0.94 1.48 — Ocean wet 1 2 3 4 2.72** 3.04** 0.66 0.59 2.08* 2.50* — Ocean dry 1 2 3 4 2.40* 3.33*** 1.31 1.02 1.14 2.13* — P *S 0.05; * * = P « 0.01 ; * * * = P < 0.001 . that a relationship existed between metabolic rate and mortality. An increase in metabolic rate and con- current more rapid depletion of reserves may have offset advantages of increasing moisture by dampen- ing during our experiments as well. Exposure was probably the principal cause of mor- tality among bait lobsters during our tests in Florida Bay. However, a small but distinctly greater level of mortality among all lobsters, including controls dur- ing weeks 1-3 and a marked increase in mortality during week 4 at the ocean site, suggest that other factors in addition to exposure were responsible for mortalities there (Figs. 1, 2). When average mortality rates of controls (Table 1) are subtracted from overall average mortality rates of exposed lobsters, resul- tant values (18.6%, Florida Bay; 18.3%, Atlantic reefs) are nearly equal and probably represent the rates of mortality actually ascribable to exposure at each site Thus, effects of exposure were similar regardless of where traps were placed. Mortality due to other effects related to confine- ment evidently do vary depending upon locations where traps are placed, especially if confinement periods are lengthy. Increased mortality rates such as those we observed during week 4 at the Atlantic reef site may result from starvation. Lyons and Ken- nedy (1981) presented evidence of weight loss and starvation among lobsters confined at densities of 3 and 5/trap in Florida Bay for 8 wk. Rate of weight loss increased during week 4 among lobsters at den- sities of 5 but did not increase rapidly until week 6 among lobsters confined at densities of 3. Those tests were conducted in the same portion of Florida Bay where present exposure tests were conducted, an area characterized by muddy sand overlain by sea- grass beds. A disparity in available food organisms between this area and that where oceanside tests were conducted may explain differences in mortal- ity during week 4. Seagrass beds in Florida Bay are lush and heavi- ly covered with epibionts (J. H. Hunt, pers. obs.). These epibionts serve as food for larger organisms which in turn are appropriate food for Panulirus argus. Snails in the genera such as Modulus, Turbo, Astraea, and Cerithium and crabs in the genera Mithrax and Pitho are abundant in these grass beds and are frequently seen within or clinging to sides of lobster traps. All of these also occur commonly in stomach contents of P. argus in south Florida (W G. Lyons, pers. obs.). At the ocean site, grass beds are sparse and patchily distributed, and fewer organisms enter traps from the surrounding sand. It seems reasonable to suppose that the weight loss observed to occur among lobsters confined near lush 73 FISHERY BULLETIN: VOL. 84, NO. 1 grass beds (Lyons and Kennedy 1981) might occur at accelerated rates in the relatively more sparse ocean environment. If food is sufficiently scarce, ac- celerated weight loss may lead to starvation and in- creased mortality within the observed 4-wk period. Traps in these experiments had their entrances boarded over to prevent escape, whereas lobsters that escape from traps used in the fishery are likely to recover from effects of starvation. Escape rates, though, are quite low, ranging from 0.8 to 1.8%/d (Yang and Obert 1978; Davis and Dodrill 1980; Lyons and Kennedy 1981). We offer no explanation for our observation that highest mortality rates are associated with 1-h ex- posures nor for the persistent background mortality among oceanside controls. Nevertheless, neither seem to be artifacts of experimental design and, in- stead, probably represent other yet-to-be understood physiological reactions to stress caused by exposure, handling, or confinement. If so, they represent other effects of baiting with shorts and are justly included among estimates of total fishery-induced mortality. Economic Effects of Mortality Baiting traps with shorts results in significant economic loss to the fishery. Although use of shorts is an effective means of attracting other lobsters without requiring out-of-pocket expenses for bait, each bait lobster that dies is one that potentially will not enter fishery landings. In addition, repair of broken legs, antennae, and other injuries caused by handling may retard growth by as much as 40% (Davis 1981), increasing the time required for a lobster to attain legal size and extending the time during which it may be used as bait. An injured lobster that escapes from a trap where it was placed will direct energy toward repair, not growth, thereby reducing the probability that it will attain legal size during its next molt. If the lobster does not attain legal size, it is again vulnerable to capture and to use as bait. Confinement itself also results in reduced lobster growth rate (Kennedy 1982), which doubt- lessly extends the period during which a lobster may be vulnerable to use as bait. The hidden costs of baiting with shorts needs to be considered in future management efforts. The following model, based only upon observed mortali- ty rates, estimates that cost: Y = AxBxCxD where Y = seasonal mortality of shorts used as bait; A = number of traps in the fishery; B = average number of shorts per trap; C = season length (in months); D = average monthly mortality rate. Because the actual allocation of fishery traps among Florida Bay and Atlantic sites is unknown but believed to be relatively equal, we selected the unweighted average cumulative 4-wk mortality rate to estimate monthly mortality throughout the fishery. By using a range of values for other variables, several estimates of the average number of shorts that die seasonally because of fishery bait- ing practices may be obtained (Table 4). Thus, if each trap in the fishery is baited with only 1 short/mo and all fishermen leave the fishery after only 4 mo, more than 600,000 sublegal lobsters may die as a result of their use as bait. If all traps are deployed for the full 8 mo and each trap uses 3 shorts as bait, more than 3.6 million shorts may die as a result of that use Both examples probably represent extreme cases, and actual fishery-induced mortality probably lies somewhere between these estimates. The problem is really more complex. Some lobsters that die because they were used as bait would prob- ably fall victim to other causes, but natural mortal- ity among lobsters of sizes appropriate for use as bait (65-75 mm CL) may be low, particularly since incidence of their principal predators, large ser- ranids, has been greatly reduced in the fishery area. Furthermore, not all traps are baited with shorts because shorts are not readily available in some peripheral areas of the fishery. Both of these factors suggest that the model may overestimate fishery- induced mortality. However, values used in the model for numbers of shorts per trap are probably low. Fishermen prefer to use 3-5 shorts/trap (Gulf of Mex- Table 4. — Estimates of the economic effect of baiting with shorts in the south Florida spiny lobster fishery. Average Seasonal monthly No. of No. of mortality mortality traps in Season shorts/ of shorts rate1 fishery2 length3 trap4 as bait 0.263 573,000 4 1 602,796 0.263 573,000 4 3 1 ,808,338 0.263 573,000 6 1 904,194 0.263 573,000 6 3 2,712,582 0.263 573,000 8 1 1,205,592 0.263 573,000 8 3 3,616,776 'Unweighted average cumulative 4-wk mortality rate from this study. 2Number of traps in 1981 (E. J. Little, Jr., Southwest Fisheries Center Resource Statistics Office, National Marine Fisheries Service, NOAA, P.O. Box 269, Key West, FL 33041, pers. commun. November 1982). 3The season is 26 July-31 March, 8+ mo; some fishermen begin removing their traps after November, and many have left the fishery by the end of January, causing a considerable reduction in the number of traps fished during February and March. ^Conservative estimates; fishermen try to put as many shorts as available into traps. 74 HUNT ET AL.: EXPOSURE AND CONFINEMENT ON SPINY LOBSTERS ico and South Atlantic Fishery Management Coun- cils 1982), and it seems probable from fishermen's comments that virtually no shorts are intentionally released. Similarly, the model only allows one input of bait per month, whereas in reality additional shorts are continually introduced, typically at 1-2 wk intervals, to replace others lost because of death or escape. These factors suggest that the model may underestimate fishery-induced mortality. Regardless of which values are applied, the model indicates that resultant losses to the fishery are con- siderable Since a lobster weighs slightly <1 lb at legal size, fishery-induced mortality may cause losses ranging from 0.6 to 3.6 million lb. At recent ex-vessel prices of $2.50 per pound, this represents a poten- tial loss to the fishery of $1.5-$9.0 million annually. In 1981, total reported commercial lobster har- vest was 5.9 million lb valued at $14.5 million3, so the hidden cost of baiting with shorts is consider- able This loss may be viewed as a necessary cost, albeit large, of doing business in the fishery or as a prob- lem that may be alleviated by alternative manage- ment strategies. If the latter course is deemed necessary, use of other baits and installation of escape gaps that allow shorts to escape while retain- ing legal lobsters in traps (Bowen 1963) are poten- tially effective strategies to increase harvest of legal lobsters without adversely affecting the popu- lation. ACKNOWLEDGMENTS This project was partially funded by a research grant (2-34 1-R) from PL 88-309 (Commercial Fisheries Research and Development Act) through the Fisheries Management Division, National Marine Fisheries Service, NOAA, U.S. Department of Commerce, and was administered by the Florida Department of Natural Resources (FDNR) Bureau of Marine Research. R. S. Brown, Western Australia Department of Fisheries and Wildlife, provided unpublished manu- scripts of related recent studies of Panulirus qjgnus. Field assistance was provided by D. G. Barber, S. F. Barber, G. F. Bieber, S. E. Coleman, J. W Lowry R. H. McMichael, Jr., G. K. Vermeer, and M. A. Winter, all presently or formerly FDNR employees. G. K. Vermeer, M. A. Winter, and R. G. Muller Statistical Surveys Branch. 1983. Florida landings 1981. Southeast Fisheries Center National Statistical Office, National Marine Fisheries Service, NOAA, 75 Virginia Beach Drive, Miami, FL 33149. (FDNR) provided valuable discussion and other assistance during manuscript preparation. All are gratefully thanked. LITERATURE CITED Anonymous. 1980. The fate of undersized rock lobsters returned to the sea. West. Aust. Dep. Fish. Wildl, Fish. Ind. News Serv. (F.I.N.S.) 13:10-12. Bowen, B. K. 1963. Management of the western rock lobster, (Panulirus longipes cygnus George). Proa Indo-Paa Fish. Counc. 14: 139-154. Brown, R. S., and N. Caputi. 1983. Factors affecting the recapture of undersize western rock lobster Panulirus qjgnus George returned by fishermen to the sea. Fish. Res. 2:103-128. Brown, R. S., J. Prince, N. Caputi, and J. Jerke. In press. Fishery induced mortality of undersize western rock lobster. West. Aust. Dep. Fish. Wildl. Bull. Cope, C. E. 1959. Spiny lobster gear and fishing methods. U.S. Fish Wildl. Serv., Fish. Leafl. 487, 17 p. Davis, G. E. 1981. Effects of injuries on spiny lobster, Panulirus argus, and implications for fishery management. Fish. Bull., U.S. 78:979-984. Davis, G. E., and J. W. Dodrill. 1980. Marine parks and sanctuaries for spiny lobster fishery management. Proa Gulf Caribb. Fish. Inst. 32:194-207. Gulf of Mexico and South Atlantic Fishery Management Councils. 1982. Fishery management plan, environmental impact state- ment and regulatory impact review for spiny lobster in the Gulf of Mexico and South Atlantic. Gulf of Mexico and South Atlantic Fishery Management Councils, Tampa, Fla., var. p. Kennedy, F. S., Jr. 1982. Catch rates of lobster traps baited with shorts, with notes on effects of confinement. In W. G. Lyons (editor), Proceedings of a workshop on Florida spiny lobster research and management, p. 20. Fla. Dep. Nat. Resour. Mar. Res. Lab., St. Petersburg. Lyons, W. G, D. G. Barber, S. M. Foster, F. S. Kennedy, Jr., and G. R. Milano. 1981. The spiny lobster, Panulirus argus, in the middle and upper Florida Keys: population structure, seasonal dynamics, and reproduction. Fla. Mar. Res. Publ. 38, 38 p. Lyons, W. G, and F. S. Kennedy, Jr. 1981. Effects of harvest techniques on sublegal spiny lobsters and on subsequent fishery yield. Proa Gulf Caribb. Fish. Inst. 33:290-300. McLeese, D. W. 1965. Survival of lobsters, Homarus americanus, out of water. J. Fish. Res. Board Can. 22:385-394. McMahon, B. R., P. J. Butler, and E. W. Taylor. 1978. Acid base changes during recovery from disturbance and during long term hypoxic exposure in the lobster, Homarus vulgaris. J. Exp. Zool. 205:361-370. SOKAL, R. R., AND F. J. ROHLF. 1969. Biometry: the principles and practice of statistics in biological research. W. H. Freeman and Co., San Franc., 776 P- 75 FISHERY BULLETIN: VOL. 84, NO. 1 WOLFFERTS, R. C. YANG, M. C. K., AND B. OBERT. 1974. Fishermen's problems in the spiny lobster fishery, hi 1978. Selected statistical analyses of Key West spiny lobster W. Seaman, Jr., and D. Y. Aska (editors), Conference pro- data. In R. E. Warner (editor), Spiny lobster research ceedings: Research and information needs of the Florida review; proceedings of a conference held December 16, 1976 spiny lobster fishery, p. 3 [abstr.], 8, 9. Fla. Sea Grant Rep. in Key West, Florida, p. 4-7. Fla. Sea Grant Tech. Pap. No. 4. SUSF-SG-74-201, Gainesville, FL. 76 TYPE, QUANTITY, AND SIZE OF FOOD OF PACIFIC SALMON (ONCORHYNCHUS) IN THE STRAIT OF JUAN DE FUCA, BRITISH COLUMBIA Terry D. Beacham1 ABSTRACT The volume, numbers, and size of prey of sockeye, Oncorhynchus nerka; pink, 0. gorbuscha; coho, 0. kisutch; and chinook, 0. tshawytscha, salmon were investigated for troll-caught salmon in the Strait of Juan de Fuca off southwestern Vancouver Island during 1967-68. Sockeye salmon was the least piscivorous species with only 7% of the stomach volume comprised of fish, while chinook salmon was the most piscivorous species at 56%. Sand lance, Ammodytes hexapterus, and euphausiids were the most important fish and invertebrate prey, respectively. As predator size increased, mean size of fish prey increased, and predators shifted to species of larger mean size Similar results were found for the invertebrate prey, with mean number of prey consumed per predator increasing for the larger invertebrate species as predator size increased. Rate of increase in mean length of fish prey was proportional to increasing predator length. The observed increase in invertebrate size with increasing predator length was not statistically signifi- cant. Although chinook and coho salmon had similar diets, they were caught at significantly different water depths. Oncorhynchus species with fewer, shorter, and more widely spaced gillrakers have higher proportions of fish in their diet than species with numerous, long, and narrow set gillrakers. The life history of Pacific salmon is quite variable among species, with fry of pink salmon, Oncorhyn- chus gorbuscha, and chum salmon, 0. keta, migrating to sea soon after emergence from the gravel, while those of sockeye salmon, 0. nerka, coho salmon, 0. kisutch, and chinook salmon, 0. tshawytscha, may spend up to 2 yr in freshwater. Once in the ocean they can migrate a considerable distance from their natal streams and feed on a variety of organisms (Godfrey et al. 1975; French et al. 1976; Major et al. 1978; Takagi et al. 1981). Salmon thus move through a number of habitats during their life cycle and con- sume a diverse array of prey. Food preferences of salmon in the range of habitats that they occupy have been an area of con- tinuing investigation (Allen and Aron 1958; Prakash 1962; LeBrasseur 1966; Parker 1971; Eggers 1982). Relative amounts of different prey types eaten in varying environments have been examined, as well as preferences by different sizes of predators in rela- tion to the size and abundance of prey. Oncorhyn- chus species differ considerably in their size, mor- phology, and ocean distribution (Hikita 1962; Neave et al. 1976; Takagi et al. 1981; Beacham and Mur- ray 1983). Morphological differences and diet parti- tioning have been reported for many fish species (Keast and Webb 1966; Hyatt 1979), and diet parti- tioning may thus be expected among Oncorhynchus species. Prey size is related to predator size (O'Brien 1979; Gibson 1980), and differential prey selection among Oncorhynchus species may also be apparent. Stomach contents of sockeye, pink, coho, and chinook salmon were investigated in a research troll- ing program conducted off southern Vancouver Island in the Strait of Juan de Fuca during 1967-68. The relative importance of different prey types, in- cluding fish and invertebrates, in the diet of the four species was studied with respect to prey size, preda- tor size, predator morphology, and diet partitioning in relation to salmonid habitat and morphology. MATERIALS AND METHODS The salmon were obtained by test trolling in the Strait of Juan de Fuca during 19 June-11 October 1967 and 1 May-12 July 1968 (Fig. 1). Detailed methodology of the program has been outlined by Graham and Argue (1972). For each salmon sampled, date, fork length (mm), round weight, and sex were recorded. Stomachs were removed, placed in num- bered cloth sample bags along with any food organisms in the mouth cavity, and preserved in 10% Formalin2 solution. 'Department of Fisheries and Oceans, Fisheries Research Branch, Pacific Biological Station, Nanaimo, British Columbia V9R 5K6, Canada. 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Manuscript accepted March 1985. FISHERY BULLETIN: VOL. 84, No. 1, 1986. 77 FISHERY BULLETIN: VOL. 84, NO. 1 I24°00' Figure 1.— Location of study area in Strait of Juan de Fuca off southwestern Vancouver Island. Laboratory analysis involved sorting the contents into the classifications outlined in Table 1 by using a low-power binocular microscope. Numbers of organisms in each classification were recorded, if possible, for each individual salmon. Once individuals were counted, displacement volumes (mL) were determined separately for fish contents, for crusta- cean contents, and for miscellaneous organisms. If organisms were too digested to assign to individual classifications but could be identified as fish or crustaceans, their volumes were included in either the unidentified fish volume or unidentified crusta- cean volume classification. Two techniques of data analysis were used. Table 1.— Percentage of salmon sampled with empty stomachs and average number of prey per fish with non-empty stomachs. N % empty Prey type Class CO o c ffl "O c CO CO en c CO X "ST .C CO ij -q o CO rr<2- CO H— I— CO o CO ;g (0 Q. UJ 2 CO E CO CO CO £ CO O CO ;g CO >. 2 * CO ■o o Q. jc Q. E < <0 5 6 CO O CO C CO 18 CO CO 2 o Sockeye <55 cm 22 46 — — — 3.7 13.2 5.0 0.3 — — 0.1 Sockeye >55 cm 117 41 0.2 — — — 13.5 8.6 0.4 0.1 — — 0.4 Total 139 42 0.2 — — — 12.1 9.3 1.1 0.1 — — 0.4 Pink <55 cm 301 26 0.7 — — — 9.7 13.7 1.0 0.3 — 0.1 0.3 Pink >55 cm 498 32 0.4 — — 0.1 15.3 13.1 2.4 0.1 0.1 0.1 0.4 Total 799 30 0.5 — — 0.1 13.1 13.3 1.9 0.2 0.1 0.1 0.4 Coho <40 cm 1,045 49 0.4 — — 0.2 6.3 1.2 0.4 1.3 0.1 0.2 0.3 Coho 40-60 cm 1,039 28 5.8 — 0.1 0.3 29.8 0.3 0.9 0.3 — 0.2 0.4 Coho >60 cm 130 32 0.5 0.2 — — 51.0 0.4 0.6 — — — 0.6 Total 2,214 38 3.3 — — 0.2 22.1 0.7 0.7 0.6 — 0.2 0.4 Chinook <40 cm 607 39 1.1 — — 0.1 5.4 0.1 0.7 0.4 — — 0.7 Chinook 40-60 cm 786 36 1.6 0.1 — 0.1 15.3 0.2 0.2 0.6 — — 0.1 Chinook >60 cm 83 47 0.8 0.3 — — 62.4 — 0.4 0.1 — — — Total 1,476 38 1.4 0.1 — 0.1 13.6 0.1 0.4 0.5 — — 0.3 'Other than Parathemisto. 78 BEACHAM: FOOD OF PACIFIC SALMON OFF BRITISH COLUMBIA Methodology for the first, percent occurrence of each of the prey types, has been outlined by Hynes (1950). All chi-square tests in the analysis for frequency of occurrence of prey types have one degree of freedom. The second technique involved determining percent- age by volume of total stomach contents for fish, crustaceans, miscellaneous organisms, and also for the individual prey classifications. Fish, crustaceans, and miscellaneous organisms were recorded by volume, and thus determining percentage of total stomach volume for each classification was direct. For individual prey types, it was necessary to con- vert numbers of individual organisms to volumes by calculating the volume displaced by a single organism of each prey type This was done by selec- ting individual salmon of each species with only one fish and/or one crustacean prey type in the stomach. The unit volumes for each prey type were then calculated as the sum of the fish or crustacean volumes for the selected fish divided by the number of the prey type under consideration. If there was only one unknown in the stomach contents with prey of known (calculated) volumes (the number of prey types multiplied by their unit volumes), the total volume of known prey was subtracted from the total fish or crustacean volume until only one unknown prey class remained. Then the volume of the prey class in question was obtained and its unit volume calculated. Comparisons of prey size among the species were analyzed by analysis of variance For an individual salmon with more than one fish or one crustacean prey class in its stomach, volume of each prey class was determined by multiplying the number of organisms by their unit volume This total volume obtained was scaled proportionately so that individual components when summed equalled the total known fish or crustacean volume RESULTS Volume and Frequency of Food Items For each species, over 30% of the individuals had empty stomachs (Table 1). In comparing fish with non-empty stomachs, sockeye salmon was the least piscivorous, with a mean 7% fish component in the diet (Fig. 2). In sockeye salmon <55 cm fork length (FL), only 2% of the stomach volume was comprised of fish. At 17% of total food volume, fish was a greater dietary component of pink salmon than of sockeye (Fig. 2). However, the fish component of the diet of sockeye and pink salmon was considerably less than that of coho (46%) and chinook (56%) salmon. Fish comprised 30% of the stomach content volume of coho <40 cm FL, but almost 50% of the stomach content volume of larger coho. Chinook salmon was the most piscivorous of the four species, and the 56% fish component of the diet was constant for the three size classes of chinook salmon inves- tigated, although the species composition of the fish prey changed. The relative importance of individual prey types was investigated for the four salmon species. Sand lance, Ammodytes hexapterus, was virtually the sole fish component of the diet of sockeye salmon, oc- curring in 4% of the 81 non-empty sockeye salmon stomachs sampled (Fig. 3). Euphausiids were the most important prey for sockeye, occurring in 58% of non-empty stomachs and comprising 71% of the total volume of food eaten. The hyperiid amphipod Parathemisto comprised over 11% of the volume of food eaten. Of the fish prey species, sand lance was again the most important for pink salmon, occurring in 9% of 562 non-empty stomachs and comprising 10% of total stomach contents (Fig. 4). There was no significant difference between sockeye and pink salmon in the frequency of occurrence of sand lance in their diets (x2 = 2.65, P > 0.05). Fish species other than sand lance (herring, Clupea harengus, and rockfish, Sebastes sp.) comprised less than 1% of stomach contents of pink salmon. As in sockeye salmon, the dominant invertebrate prey types were euphausiids at 62% of stomach content volume and Parathemisto at 14%. Frequency of occurrence of euphausiids (x2 = 1.63, P > 0.05) and Parathemisto (x2 = 3.54, P > 0.05) were similar for sockeye and pink salmon. Fish species were a significant food for coho and chinook salmon. For example, sand lance occurred in 27% of 1,364 non-empty stomachs of coho salmon, and also comprised 27% of total stomach volume (Fig. 5). Herring comprised <1% of the stomach con- tent volume of coho <40 cm FL, but 25% of the volume for coho >60 cm FL. The dominant inverte- brate prey type was euphausiids, comprising 51% of total stomach contents, while all invertebrate prey types combined comprised only 54%. The relative importance of fish as a prey type was greatest in chinook salmon, with sand lance again the dominant prey species, occurring in 34% of 914 non-empty stomachs, and comprising 35% of total volume of contents (Fig. 6). Sand lance occurred in the diet of chinook and coho salmon at similar frequencies (x2 = 0.80, P > 0.05), as did herring (x2 = 0.08, P > 0.05). Herring comprised 9% of the stomach contents for chinook salmon <40 cm FL, but 33% of the stomach contents for chinook salmon >60 cm FL. 79 FISHERY BULLETIN: VOL. 84, NO. 1 UJ O > X o < o h- ^ 100- 80- 60- 40 20 -I 0 100 80 60 40- 20- 0 100 80 60 40 20J 0 100 80H 60 40^ 20 <55cm >55cm Total SOCKEYE <55cm <55cm PINK <40cm 40-60cm >60cm COHO <40cm 40-60cm >60cm CHINOOK O !2 in in II <=> o — a> o O in c o o o O 2 I! c c D O = o> ^^ <->o in c D o o tn 3 O <2 si c c o o zz o> 01 a> z. Figure 2— Percentage volumes of stomach contents of the fish, crustacean, and miscellaneous organism component for sockeye, pink, coho, and chinook salmon sampled in Strait of Juan de Fuca during 1967-68. Coho ate greater numbers of fish than did chinook salmon (Table 1), but chinook had a greater volume of the stomach contents composed of fish (56% chinook, 46% coho). This result suggests chinook eat larger fish than coho (Table 2). As with coho, euphausiids were the dominant invertebrate prey type of chinook salmon, comprising 40% of a total invertebrate volume of 44% of stomach contents. However, euphausiids occurred significantly more often in the diet of coho salmon than in chinook salmon (x2 = 4.73, P < 0.01). Fish were a more significant dietary component of chinook and coho salmon than of sockeye and pink salmon. Sand lance occurred significantly more often in the diet of chinook and coho salmon than in the diet of sockeye and pink salmon (x2 = 152.9, P < 0.01). Similar results were also found for herring (x2 = 18.1, P < 0.01), rockfish (x2 = 7.2, P < 0.01), and mixed fish species (x2 = 39.0, P < 0.01). Inverte- brate prey were more significant in the diet of sockeye and pink salmon than in that of chinook and coho. Euphausiids occurred more frequently in the diet of sockeye and pink salmon (x2 = 199.3, P < 0.01), as did Parathemisto (x2.= 619.5, P < 0.01), crab larvae (x2 = 171.1, P < 0.01), and amphipods (x2 = 9.2, P < 0.01). There was no difference in frequency of occurrence of crabs in the diet (x2 = 0.01, P > 0.05) which occurred only at low levels or not at all, but mysiids occurred more frequently in the diet of chinook and coho salmon than in 80 BEACHAM: FOOD OF PACIFIC SALMON OFF BRITISH COLUMBIA UJ CL CT Z> O O o o z UJ O UJ rr u. >5 UJ o > X o < o r- CO 100 80 60 40 20 0 100^ 80 60 40 20 0 100 80- 60- 40 20 0 S0CKEYE 1 1 1 1 1 <55 cm III — 1 1 1 - >55 cm 1 1 100- 80 60- 40- 20 0 <55cm >55 cm o c D D CO X 2C o o rr a> O to '55 3 O .c Q. UJ D Q_ a> v) if> D T3 "O ^1 > •— o a o >. •= O D Q. E < o o — O o o >- o z: UJ O UJ tr ;s UJ _l O > X o < o h- co 100 80 H 60 40- 20- 0- 100 80 60 40- 20 0 PINK I00n 80 60 40 20 0 100 80 60 40 20 o c X) c o CO <55cm I 1 r >55cm <55cm >55cm O) o o £ co o to ■D *— m H— k_ to b a> XT D XT xt O Q. 3 o UJ o a. 0) in 10 10 to to o T3 T3 -O 3 c > O o O O CO Q. L. 0) 0) D >> •— o C O O s XT Q. O D E ^^ < Figure 4— Percentage frequency of occurrence and percentage stomach volume of prey types for pink salmon. less Parathemisto, were eaten per individual predator. The difference in predator response to euphausids and Parathemisto may be examined in relation to the size of the prey. The unit volumes of an individual euphausiid were about four times larger than those of an individual Parathemisto (Table 2). In each salmon species examined, as the predators increased in size, they switched from the smaller Parathemisto to the larger euphausiids and also crab larvae, consuming greater numbers of the larger prey and decreasing numbers of the smaller prey. Chinook and coho salmon also consumed significantly larger Parathemisto than did sockeye and pink salmon (F = 4.9; df = 3,98; P < 0.01). For the invertebrate prey, an increase in predator size resulted in greater numbers of larger prey being consumed. As predator size increased, there was an increase in the size of the prey consumed (Table 2). Larger predators consumed larger sand lance and herring. Chinook and coho salmon consumed larger sand lance (F = 3.7; df = 3,613; P < 0.05) and mixed fish species (F = 2.9; df = 2,128; P < 0.05) than did sock- eye and pink salmon. In coho and chinook salmon, there was also a tendency for larger salmon to switch prey types from the smaller sand lance to the larger herring and rockfish. Increasing predator size pro- duced shifts in both the type, number, and size of the prey consumed. Changes in size of prey and predators were in- 82 BEACHAM: FOOD OF PACIFIC SALMON OFF BRITISH COLUMBIA 60-, 40 20-| 0 60 404 o 20-1 UJ o -z. Ul en rr Z> o o o Z> o UJ DC ^ 0 60 40 20 0 C0H0 <40cm I 1 i 1 40-60 cm l 1 >60 cm UJ o > x o < o I- cn 60 40-1 20 0 60- 40- 20 0 60^ 40 20 0 o a o D CO <40 cm I — I 40-60 cm >60cm <1> X -2£ Ul F o o 0) 3 o> en -C O Q. D 3 Ul D 0_ d) D > .O O o Ul ■o Ul O Cl D O D - Figure 5.— Percentage frequency of occurrence and percentage stomach volume of prey types for coho salmon. vestigated for the two most frequently occurring fish prey (sand lance, herring) and crustacean prey (euphausiids, Parathemisto). Size classes for sock- eye and pink salmon were below and above 55 cm FL, and those for chinook and coho salmon below and above 60 cm FL. I assume that the value of the cube root of the volume ratio of the prey is propor- tional to the prey length ratio, and thus changes in prey size can be compared with changes in predator size Mean size of the fish component of the prey in- creased as predator size increased (Table 3). As the size of pink, coho, and chinook salmon increased by 13%, 65%, and 69%, respectively, the size of the sand lance consumed increased by 16%, 83%, and 83%, respectively. The size of herring eaten also increased as predator size increased, and for pink and chinook salmon it was about equal to the increase in size of the predator species. When the predator responses to increase in size of both prey species are pooled, there is a weak correlation between increasing predator length and increasing prey length (r = 0.69, n = 6, P > 0.05); but if the coho salmon response to increasing herring size is deleted, the relationship is much stronger between increasing predator and prey size (r = 0.98, n = 5, P < 0.01). Apparent trends of invertebrate prey size with predator size were not statistically significant. For sockeye and pink salmon, mean size of individuals in the two invertebrate prey classes decreased as 83 FISHERY BULLETIN: VOL. 84, NO. 1 60n 0 3 40 u 20 60cm i — i UJ o > X o < o I- 60n 40- 20- 0 60 40^ 20 0 60 40- 20- 0 <40cm I 1 40-60 cm >60cm in ;g '35 D -C Q. 3 UJ O 0- 0) o > D in T3 in in ■D O Q. .C a. E < in D O in in 3 c O O a> c o o o Figure 6.— Percentage frequency of occurrence and percentage stomach volume of prey types for chinook salmon. predator size increased, but not significantly (Table 3) (r = -0.24, n = 4, P > 0.05). For chinook and coho salmon, mean size of the invertebrate prey increased as predator size increased (r = 0.42, n = 4, P > 0.05). However, the increase in prey size was considerably less than the increase in predator size (Table 3). The results of the previous analyses are sum- marized as follows. As predator size increased, in- dividual predators selected larger fish prey of one species, but not a greater number of the prey. There was also a shifting from smaller prey species (sand lance) to larger ones (herring, rockfish). As predator size increased, there was a tendency to shift from smaller invertebrate prey (Parathemisto) to larger types (euphausiids, crab larvae). Greater numbers of the larger prey were consumed by an individual predator, while numbers of smaller prey consumed declined. Although larger invertebrate prey types were preferred as predator size increased, larger in- dividuals of each prey class were not necessarily selected by larger predators. Species Comparisons The dietary components of the four species of salmon investigated are different, and there is more than one possible reason for the apparent partition- ing of diet among the salmon species. Perhaps because the salmon occupied different depth zones, the differences in diet are attributable simply to dif- 84 BEACHAM: FOOD OF PACIFIC SALMON OFF BRITISH COLUMBIA CO 0) CD E c >^ 0) r CD > C o o a> CO 3 a) 5 co CD 3 CO . CD § S3 CD f co c CD CD iE s r~ Q. CO c I fo c « CO 05 CD > T3 CD a. CO CD E O > c CO CD E CD J3 CO O CO LU m .< sueaoBisnjo SnO8UB||90S!l^| sqejo spodjqdujv spnsAy\| 8EAJB| qBJQ ojsiLuaqieJBd spnsnBqdng qs|i jsijio (sajsegss) 6uuj8h 83UB| PUBS CO CO TO o 1 CM 1 £«•, j^ co 21 2 CM^ o o~ S 1 CO 1^ o i o CO o CO o> 00 CO o q CM CM CO ^ I CO q d d d d d *~ d d •* ,_ 1 i I * jfr CO CO - 1^ CO CM O ■ s 1 1 1 o CO o CO i^ CO 00 CO CO o o q 1 o CO d d d d d d d ,_ d 1 i ^ ^ Si- CM t^_ CO CM CM, CO C § 1 1 o CD t*- <3> CD o a> CO CO CO ■* o q 1 CM *~ T— T— d *~ d t — CM CM c\i 1 ~ ~ 2, 2^ CO CO 00 o CO — i-» CO 5- CM- 00 CO 1 CO 00 CO 00 CO 05 o CO CM o CD 1 — CO o q CM ^_ CD CM CO q cvi Cvj d d d T— CM CM ,_ ,_ CM CM T~ c. CM^ G- 00 00 CO CM^ CO r^ CM_ CM CO 00 00 c i^ r^ o CD C\J 1^ o CO 1^ CO CO CO CO o q CO CO CO CO oo CO d -,— d d d d d d ,~ d d d d d S 5 p 00 3^ 00 CO CO CO- o CM s 5? CO s R p o 00 CO ■18 m. Coho and chinook salmon have similar diets, but are found at significantly different depths (x2 = 714.7, P < 0.01). Thus partitioning of the diets among salmon species is not related simply to water depth. Morphological characters of the salmon species were compared with their food preferences. Chinook and coho salmon have fewer, shorter, and more wide- ly spaced gillrakers than those of sockeye and pink salmon (Table 5). As gillrakers are used to strain food organisms from water passing over the gills (Lagler et al. 1962), I expected salmon species feeding on planktivorous prey to have more gillrakers that are longer and more closely set than those in primarily piscivorous salmon species. Similar arguments could be made for tooth size (Table 5). Partitioning of the diet among the species of salmon investigated is clearly a reflection of morphological differences among the species. DISCUSSION The calculation of unit volumes for individual prey classes is an important component of the analysis. Prey types were assumed to be in a similar state of FISHERY BULLETIN: VOL. 84, NO. 1 digestion for the different size classes of each species of salmon so that calculated unit volumes would be comparable Violation of this assumption may ac- count for the inverse predator-prey size relationship found for sockeye and pink salmon with euphausiids and Parathemisto. The analysis of relative sizes of the species eaten assumes that different prey types were not more or less digested than others. This is unlikely to be strictly true, but it was assumed that differential digestability of the prey species did not significantly alter their relative sizes. Previous work on diet description of Oncorhynchus species has indicated that there can be considerable variability in dietary components of a particular species. However, some general conclusions can be drawn. Sockeye salmon are the least piscivorous of ' the northeast Pacific Oncorhynchus species (Allen and Aron 1958; LeBrasseur 1966; Foerster 1968). Euphausiids have been reported consistently as a major contributor to the diet of pink salmon (Maeda 1954; Ito 1964; Takagi et al. 1981). The fish compo- nent reported has been variable, ranging from <1% to over 90% of stomach volume (Takagi et al. 1981). Chinook and coho salmon tend to be the most piscivorous (Allen and Aron 1958; Prakash 1962; Reimers 1964; LeBrasseur 1966; Machidori 1972). For chinook salmon, fish were reported to provide Table 4. — Number of salmon caught with non-empty stomachs and depth of water (m) in Strait of Juan de Fuca, British Columbia. Salmon were caught by troll gear. Numbers in parentheses are per- cent of each species caught in each depth zone. Depth (m) Sockeye Pink Coho Chinook <9.1 8 (9.9) 41 (7.3) 385 (28.1) 20 (2.2) 9.1-18.3 10 (12.3) 95 (16.9) 360 (26.3) 60 (6.6) 18.3-27.4 26 (32.1) 159 (28.3) 269 (19.6) 134 (14.6) 27.4-36.6 23 (28.4) 151 (26.9) 211 (15.4) 267 (29.1) 36.6-45.7 7 (8.6) 65 (11.6) 86 (6.3) 119 (13.0) 45.7-54.8 7 (8.6) 50 (8.9) 58 (4.2) 316 (34.5) Total 81 561 1,369 916 Table 5. — Comparisons of morphometric and meristic characters of Pacific salmon whose dietary components were investigated in this study. Gillraker Tooth Species No.1 Spacing2 • Length3 size4 Sockeye Pink Coho Chinook 33.7 30.4 21.2 20.7 close moderate wide wide 2.6 3.4 2.1 2.0 smallest small moderate large 'From Hikita (1962). 2From Morrow (1980). 3Gillraker length as percent of postorbital-hypural length. Gillraker length is from Hikita (1962), postorbital-hypural length from Beacham and Murray (1983). "From Vladykov (1962), Hikita (1962). 86 BEACHAM: FOOD OF PACIFIC SALMON OFF BRITISH COLUMBIA a larger proportion of the diet of larger chinook salmon than of smaller ones (Milne 1955; Reid 1961). In my study, the fish component of the diet was similar for all size classes of chinook salmon. This may be due to differences in availability of inverte- brate prey to the smaller chinook salmon among the studies. For example, Ito (1964) found that squid were the largest dietary component of chinook and coho salmon caught in drift nets in high seas fisheries. Variability in diets of the different species may be due in part to prey abundance, selection by the predator, and possible selectivity by the sampling gear used. Hook and line sampling may select fish of different diets than would perhaps gill nets. Salmon caught by trolling may have a higher com- ponent of fish in the diet than those caught by gill nets. In my study, fish did constitute a larger pro- portion of the diet in larger coho salmon than in smaller ones, as noted for chinook salmon. My study has examined the distribution of prey types and sizes for salmon caught from June to October only. Although the relative proportions of fish and inverte- brate prey could change seasonally for the salmon species examined, the relative ranking of the species in terms of proportion of fish in their diet should re- main constant. Availability of prey types can alter markedly the proportions in a predator's diet. Herring comprised over 70% of the stomach contents of troll-caught chinook and coho salmon caught off the east and west coasts of Vancouver Island in 1957 (Prakash 1962). My study showed that during 1967-68, her- ring comprised <20% of the stomach contents of chinook and coho salmon in the same area. Stock abundances of herring declined rapidly in the late 1960's in British Columbia (Hourston 1978), in- dicating that during a period of low herring abun- dance, sand lance became an important dietary com- ponent of chinook and coho salmon in this area. Pink salmon in southern British Columbia and Washington State show 2-yr cycles of abundance, with returns absent in even-numbered years. This pattern of abundance has been suggested to be a result of predation by returning adults of the domi- nant brood year on fry of the alternate brood year (Ricker 1962). In my study, fish other than sand lance, herring, or rockfish comprised <1% of the stomach contents of pink salmon sampled in 1967. These results suggest that predation by the domi- nant broodline on the alternate broodline may be neither necessary nor sufficient to account for cycles in pink salmon abundance. The effect of prey size on selection by planktiv- orous fish has been examined by Werner and Hall (1974), O'Brien et al. (1976), O'Brien (1979), Gibson (1980), and Eggers (1982). Eggers found that juvenile sockeye salmon prefer large nonevasive prey, but will eat small and/or evasive prey when the former is not available I found that as predator size increased, prey size increased also, both in terms of size of individuals within a prey type, and a shifting from smaller to larger prey types. The predators presumably decrease the amount of time and energy needed to ingest a given amount of food by switch- ing from smaller to larger prey, given that the large prey types are sufficiently abundant. Werner and Hall (1974) attributed a preference by predators for only a part of the prey types available as a method for increasing foraging efficiency. These results sug- gest that the salmon species examined do select prey both for size and availability, presumably to increase foraging efficiency. Morphological differences and diet partitioning have been previously noted for many fish species (Keast and Webb 1966; Hyatt 1979). As outlined by Hyatt (1979), many planktivorous feeding fish tend to have numerous, well-developed, close-set gill- rakers. My study indicated that the more piscivorous chinook and coho salmon have fewer gillrakers than the more planktivorous sockeye and pink salmon. Lake trout, Salvelinus namaycush, populations that are more planktivorous also have more and longer gillrakers than less planktivorous ones (Martin and Sandercock 1967). Oncorhynchus masou (masou or cherry salmon), found in the western Pacific Ocean, has fewer gillrakers than either chinook or coho salmon (Hikita 1962) and, as an adult, feeds largely on fish (Tanaka 1965). Chum, 0. keta, salmon have an average of 2-3 more gillrakers than chinook and coho (Hikita 1962), and the diet of chum salmon sam- pled in the spring and summer during 1956-63 in the North Pacific comprised between 10 and 35% fish (Neave et al. 1976). In the genus Oncorhynchus, as gillraker number declines, the proportion of fish in the diet increases. Morphological differences among the species account for a greater partition- ing of the diet than do differences in water depths in which the individual species are located. Pacific salmon are adaptable in their diet, shift- ing their preferred prey species in relation to prey size and abundance It seems unlikely that salmon abundance is affected by the abundance of any one type of prey. For example, the decline in abundance of British Columbia herring stocks was not followed immediately by declines in salmon abundance Growth rates of salmon may be affected by changes in diet and this could have an impact on stock popula- tion dynamics. 87 FISHERY BULLETIN: VOL. 84. NO. 1 ACKNOWLEDGMENTS I am indebted to those people who collected and sampled the salmon for stomach contents that were analyzed in this paper. Sharon Henderson and Bruce Bernard were invaluable for their assistance in data analysis and computer programming. J. G. McDonald provided the initial suggestion for the study. Clyde Murray and two referees offered many valuable criticisms of the manuscript. Lauri Mackie drafted the figures. The manuscript was prepared with the help of the staff of the Publications Unit of the Pacific Biological Station. LITERATURE CITED Allen, G. H., and W. Aron. 1958. Food of the salmonid fishes of the western North Pacific Ocean. U.S. Fish Wildl. Serv., Spec Sci. Rep. Fish. 237, 11 p. Beacham, T. D., and C. B. Murray. 1983. Sexual dimorphism in the adipose fin of Pacific salmon (Oncorhynchus). Can. J. Fish. Aquat. Sci. 40:2019-2024. Eggers, D. M. 1982. Planktivore preference by prey size Ecology 63:381- 390. Foerster, R. E. 1968. The sockeye salmon, Oncorhynchus nerka. Bull. Fish. Res. Board Can. 162:1-422. French, R., H. Bilton, M. Osako, and A. Hartt. 1976. Distribution and origin of sockeye salmon (Oncorhyyi- chus nerka) in offshore waters of the North Pacific Ocean. Int. North Pac. Fish. Comm. Bull. 34, 113 p. Gibson, R. M. 1980. Optimal prey-size selection by three-spine sticklebacks (Gasterosteics aculeatus): a test of the apparent-size hypothe- sis. Z. Tierpsychol. 52:291-307. Godfrey, H., K. A. Henry, and S. Machidori. 1975. Distribution and abundance of coho salmon in offshore waters of the North Pacific Ocean. Int. North Pac. Fish. Comm. Bull. 31, 80 p. Graham, C. C, and A. W. Argue. 1972. Basic data on Pacific salmon stomach contents and ef- fort for 1967 and 1968 test trolling catches in Juan de Fuca Strait, British Columbia. Can. Dep. Environ. Fish. Serv., MS Rep. 1972-76, 405 p. Hikita, T. 1962. Ecological and morphological studies of the genus On- corhynchus (Salmonidae) with particular consideration on phytogeny. Sci. Rep. Hokkaido Salmon Hatchery 17, 60 p. Hourston, A. S. 1978. The decline and recovery of Canada's Pacific herring stocks. Can. Fish. Mar. Serv. Tech. Rep. 784, 17 p. Hyatt, K. D. 1979. Feeding Strategy. In W. S. Hoar and D. J. Randall (editors), Fish physiology, Vol. VIII, Bioenergetics and growth, p. 71-119. Acad. Press, N.Y. Hynes, H. B. N. 1950. The food of freshwater sticklebacks (Gasterosteus aculeatus and Pygosteus pungitius) with a review of the methods used in studies of the food of fishes. J. Anim. Ecol. 19:36-58. Ito, J. 1964. Food and feeding habits of Pacific salmon (genus On- corhynchus) in their oceanic life Hokkaido Reg. Fish. Res. Lab. Bull. 29:85-97. (Engl, transl., Fish. Res. Board Can. Transl. Ser., 1309.) Keast, A., and D. Webb. 1966. Mouth and body form relative to feeding ecology in the fish fauna of a small lake, Lake Opinicon, Ontario. J. Fish. Res. Board Can. 23:1845-1874. Lagler, K. F, J. E. Bardach, and R. R. Miller. 1962. Ichthyology. John Wiley and Sons, Inc., N.Y., 545 p. LeBrasseur, R. J. 1966. Stomach contents of salmon and steelhead trout in the northeastern Pacific Ocean. J. Fish. Res. Board Can. 23: 85-100. Machidori, S. 1972. Observations on latitudinal distribution of offshore coho salmon in early summer, with reference to water tempera- ture and food organisms. Jpn. Far Seas Fish. Res. Lab. Bull. 6, p. 101-110. Maeda, H. 1954. Ecological analyses of pelagic shoals— I. Analysis of salmon gill-net association in the Aleutians. (1) Quantitative analysis of food. Jpn. J. Ichthyol. 3:223-231. Major, R. L., J. Ito, S. Ito, and H. Godfrey. 1978. Distribution and origin of chinook salmon (Oncorhyn- chus tshawytscha) in offshore waters of the North Pacific Ocean. Int. North Pac Fish. Comm. Bull. 38, 54 p. Martin, N. V, and F K. Sandercock. 1967. Pyloric caeca and gill raker development in lake trout, Salvelinus namaycush, in Algonquin park, Ontario. J. Fish. Res. Board Can. 24:965-974. Milne, D J. 1955. Selectivity of trolling lures. Fish. Res. Board Can. Prog. Rep. Pac. Coast Stn. 103:3-5. Morrow, J. E. 1980. The freshwater fishes of Alaska. Alaska Northwest Publ. Co., Anchorage AK, 248 p. Neave, F, T Yonemori, and R. G Bakkala. 1976. Distribution and origin of chum salmon in offshore waters of the North Pacific Ocean. Int. North Pac Fish. Comm. Bull. 35, 79 p. O'Brien, W. J. 1979. The predator-prey interaction of planktivorous fish and zooplankton. Am. Sci. 67:572-581. O'Brien, W. J., N. A. Slade, and G. L. Vinyard. 1976. Apparent size as the determinant of prey selection by bluegill sunfish (Lepomis macrochirus). Ecology 57:1304- 1310. Parker, R. R. 1971. Size selective predation among juvenile salmonid fishes in a British Columbia inlet. J. Fish. Res. Board Can. 28: 1503-1510. Prakash, A. 1962. Seasonal changes in feeding of coho and chinook (spring) salmon in southern British Columbia waters. J. Fish. Res. Board Can. 19:851-866. Reid, G M. 1961. Stomach content analysis of troll-caught king and coho salmon, Southeastern Alaska, 1957-58. U.S. Fish Wildl. Serv., Spec Sci. Rep. 379, 8 p. Reimers, P. E. 1964. A modified method of analyzing stomach contents with notes on the food habits of coho salmon in the coastal waters of Oregon and southern Washington. Fish. Comm. Oreg. Res. Briefs 10, p. 46-56. RlCKER, W. E. 1962. Regulation of the abundance of pink salmon populations. 88 BEACHAM: FOOD OF PACIFIC SALMON OFF BRITISH COLUMBIA In N. J. Wilimovsky (editor), Symposium on Pink Salmon, 16, p. 75-135. p. 155-211. Inst. Fish., Univ. Br. Columbia, Vancouver, B.C. Vladykov, V. D. Takagi, K., K. V. Aro, A. C. Hartt, and M. B. Dell. 1962. Osteological studies of Pacific salmon of the genus On- 1981. Distribution and origin of pink salmon (Oncorhynchus corhynchus. Bull. Fish. Res. Board Can. 136, 172 p. gorbuscha) in offshore waters of the North Pacific Ocean. Werner, E. E., and D. J. Hall. Int. North Pac. Fish. Comm. Bull. 40, 195 p. 1974. Optimal foraging and the size selection of prey by the Tanaka, S. bluegill sunfish (Lepomis macrochirus). Ecology 55:1042- 1965. A review of the biological information on masu salmon 1052. (Oncorhynchus masou). Int. North Pac Fish. Comm. Bull. 89 DETERMINING AGE OF LARVAL FISH WITH THE OTOLITH INCREMENT TECHNIQUE Cynthia Jones1 ABSTRACT Aging of larval fish from otoliths rests on the assumption that increments are formed daily. Indeed, proper validation of the relationship between increment deposition and age is fundamental to accurate age deter- mination of field-captured fish, lb evaluate the universality of daily deposition of otolith increments, the literature was reviewed and exceptions discussed. Laboratory studies under optimal conditions generally (17 species out of 20) show that larvae deposit daily increments. However, in studies that examined increment deposition under suboptimal or extreme conditions, deposition was not daily in over half of the species. Nondaily deposition caused by extreme conditions (eg, total starvation, abnormal photoperiod) may not invalidate the otolith increment tech- nique if those conditions do not occur in the field. Nondaily deposition under suboptimal conditions (eg., low temperature, intermittent starvation) that larvae may face in nature cause concern about this tech- nique for aging field-captured larvae Deposition in many species has not been examined under suboptimal conditions, nor has the effect of suboptimal conditions been shown on the age at first increment forma- tion. The literature shows that the technique should be validated under both optimal conditions and those that mimic nature Otoliths have been used to age fish since Reibisch (1899) first observed annular ring formation in Pleuronectes platessa (as reported in Ricker 1975). Assessing age by counting annular rings works well in adults of temperate species where pronounced seasonal changes in growth result in bands (formed from tightly spaced growth increments deposited in the winter) in the otolith which correspond to each year of life Discovery of fine increments, analogous to annual rings, but instead formed daily, has per- mitted the age of larval fish to be determined. While studying temperate water species, Pannella (1971) observed that about 360 fine increments oc- curred between annular rings and suggested that these were deposited daily. He used this knowledge when reading the otoliths of adult tropical fish (whose otoliths also had fine increments) to show pat- terns of growth that were grouped into 14- and 28-d cycles (Pannella 1974). The initial application of the otolith aging tech- nique to larval fish was done by Brothers et al. (1976). Daily increment deposition was verified for northern anchovy, Engraulus mordax, and California grunion, Leuresthes tenuis, which were reared from eggs in the laboratory. Since this initial application, the otolith increment technique has been used widely to Graduate School of Oceanography, University of Rhode Island, Kingston, RI 01882-1197; present address: Department of Natural Resources, Cornell University, Ithaca, NY 14853. estimate age in at least 29 species of larval fish. It has been used in freshwater and marine species, and applied to field-captured species, at times without adequate validation. The ultimate purpose in developing the otolith aging technique for application to young fish is the ability to accurately age field larvae and juveniles. If the technique is to be applied directly to the field, based on conclusions drawn from rearing larvae in the laboratory, then the deposition of increments must be daily under conditions experienced in the field during these early life stages. The applicability of this technique relies on the assumption that 1) either surviving larvae (or sampled larvae) are those that grew under moderately good conditions (few lar- vae under suboptimal conditions survive) or 2) lar- vae can encounter suboptimal conditions, a propor- tion of these larvae will survive, and increment deposition is not affected by these suboptimal con- ditions. The first assumption is difficult to evaluate without using the hypothesis that increments are daily. The second assumption has been tested and the results can be summarized. The second assump- tion is based on increment deposition being triggered by a zeitgeber, an external factor that entrains a diel cycle within the larvae Validation of daily increment deposition under con- ditions within the natural range of experience of the larvae is fundamental to accurate estimation of age in field-captured fish. When the estimation technique Manuscript accepted March 1985. EISHEEY ptit T itttm. wm Q/i wn i iqoc 91 FISHERY BULLETIN: VOL. 84, NO. 1 used to age larvae is inaccurate, estimates of growth and mortality, which rely on knowledge of age, will also be inaccurate The purpose of this paper is to discuss the use of the otolith increment technique to age larval fish. The published literature is used to evaluate the hypothesis, H0: Larval age is equal to otolith incre- ment count (plus age at first increment deposition) under conditions that are encountered in the field. An additional idea can be evaluated: That time of initial increment deposition is influenced by incuba- tion time The paper will discuss the factors which affect deposition of increments, validation studies that have been performed, and application of the technique in the field. Factors which are likely to affect increment deposition in the field must be assessed by the valida- tion procedure In addition, the adequacy of valida- tion that has been performed is evaluated, and ramifications in field applications are discussed. FACTORS AFFECTING DEPOSITION RATES Mechanisms that have been postulated as initiators of differentiation of otolith microstructure are photo- period, feeding, and temperature Increment deposi- tion has been tested in the literature under two conditions: 1) tests within the natural range of experience of the fish which could be optimal (non- stressful) and suboptimal (stressful), and 2) abnor- mal conditions that are wholly outside of their experience Taubert and Coble (1977) stated that photoperiod entrained a diel clock that resulted in daily forma- tion of otolith increments. Tanaka et al. (1981) stud- ied the formation of increments in Tilapia nilotica using scanning electron microscopy and found that the fast growth (incremental) zone started a few hours after light stimulus and that the slow growth (discontinuous) zone was formed immediately after light stimulus. Neither change in photoperiod length nor feeding time affected increment initiation. Brothers and McFarland (1981), however, reported that the discontinuous zone began near midnight. These results are contradictory, and without further investigations force the conclusion that the temporal formation of increments is species-specific Abnormal photoperiods have been shown to dis- rupt daily increment formation in Fundulus hetero- clitus (Radtke and Dean 1982) and in Tilapia mossambica (Taubert and Coble 1977). Constant light, however, did not disrupt daily increment forma- tion in Oncorhynchus tshawytscha (Neilson and Geen 1982) or in Scophthalmus maximus (Geffen 1982). Unlike photoperiod changes, which are regular and gradual in nature, feeding times can occur at irre- gular intervals and might cause deviations in daily increment deposition. Two studies have tested the effects of feeding within the normal range experi- enced by fish larvae Neilson and Geen (1982) found that subdaily increments could be induced through frequent discrete feedings: feeding four times a day resulted in formation of more than one increment in Oncorhynchus tshawytscha. Daily and subdaily in- crements were not distinguished in counts. Tanaka et al. (1981) found conversely that feeding time had no effect on the initiation of increment formation in Tilapia nilotica. Larvae were fed once a day, but the times of feeding were changed. Perhaps multi- ple feeding during the day results in the subdaily in- crements that sometimes appear in otoliths. The ef- fect of starvation (an extreme circumstance in the field) on increment deposition has been tested in only three species: Scophthalmus maximus (Geffen 1982), Morone saxatilis (Jones 1984), and Oncorhynchus nerka (Marshall and Parker 1982). Geffen raised the turbot larvae on rotifers and Artemia until they were 10 d old. Larvae were then starved for 23 d. Jones did not supply exogenous food from hatch onward. Both Geffen and Jones found that starvation disrupted increment formation. Marshall and Parker fed their sockeye salmon larvae for the first 3 wk of life, and then starved them for 2 wk. Marshall and Parker found that starvation over 2 wk had no ef- fect on increment deposition. It is possible that the difference might reflect different age-specific sen- sitivity to starvation, rather than species-specific responses. Brothers (1978) has linked temperature as a prime factor in increment deposition. Working with tem- perate stream populations, he has found that diel temperature changes result in daily increment for- mation. Brothers (1978) stated that "six or more in- crements per day may be formed as the result of short term, . . .relatively minor. . . temperature fluc- tuations." Other investigators (Radtke and Dean 1982; Geffen 1982) found that small temperature changes had no effect on the rate of increment deposition. Apparently, temperature response is also species-specific. LABORATORY STUDIES OF INCREMENT DEPOSITION Initial Ring Deposition When fish are raised in the laboratory from eggs 92 JONES: DETERMINING AGE OF LARVAL FISH through the larval stages, two parameters fun- damental to application of the increment technique to field populations can be determined: 1) age at first increment deposition and 2) testing of daily incre- ment deposition under artificial conditions. Age at initial increment deposition for 18 species of fish is listed in Table 1. Radtke (1978) speculated that in species having slowly developing embryos, initial deposition occurs at, or before, hatch; in species having rapidly developing embryos, initial increment deposition does not occur until yolk-sac absorption or first feeding. This hypothesis is not substantiated in the currently published literature. Information for nine species of laboratory-reared fish larvae (Table 2) shows no such trend for data currently reported in the literature Even for the same suborder, Clu- peoidei, opposite development and initial increment deposition patterns exist for herring (Clupea haren- gus) and the northern anchovy. The Case for Daily Increment Deposition Seventeen species have shown consistent daily deposition of increments under what are presumed to be good conditions for growth. The species that have shown daily increment deposition come from both freshwater and marine habitats and encompass a wide variety of lifestyles. In addition, six species held in the laboratory and sampled over known periods of time demonstrated daily increment deposition (Table 3). Four investigation groups (Struhsaker and Uchiyama 1976 for Stolephorus pur- pureas, Taubert and Coble 1977 for Lepomis macro- chirus, Campana and Neilson 1982, Wilson and Larkin 1980 for Oncorhynchus nerka) brought lar- vae and juveniles into the laboratory, reared them for a period of time, then correlated increment counts to days of captivity. Schmidt and Fabrizio (1980) took consecutive samples from a field popula- tion of Micropterus salmoides, which had a short spawning period and correlated the time between samples to the change in mean increment count. Lack of Daily Deposition Rates The most controversial results obtained so far come from studies of increment deposition in larval Clupea harengus (Table 1). Agreement for daily in- crement deposition has not been obtained. Studies that observed daily deposition by Gjosaeter2 and and Gj«isaeter and 0iestad (1981) indicate that 2Harold Gjdsaeter, Institute of Marine Research, P.O. Box 1870 5011 Bergen, Norway, pers. commun. February 1983. increments are deposited with roughly daily periodicity and that initial increment deposition begins at first feeding (4-5 d). Gjosaeter and 0iestad (1981) found that 99 increments were formed in 97-d-old larvae. Gjosaeter, however, cau- tioned that these results were based on small sam- ple sizes. Lough et al. (1982) reported on larval her- ring reared in the laboratory that lived until age 18 d. They did confirm that increment deposition began at yolk-sac absorption, but did not find that the in- crements were daily. In fact, only three increments were laid down within 18 d. Lack of confirmation of daily deposition is easy to dismiss, since the lar- vae did not survive past 18 d. However, Geffen (1982) has demonstrated an inter- action between growth rate and increment deposi- tion rate. Only under circumstances of very fast growth, 0.42 mm/d (a rate which is faster than growth rates postulated for field animals) did incre- ment deposition approach daily periodicity (0.92 in- crements/d). It is noteworthy that the growth rates in her study were related to container size; faster growth occurs in bigger containers. The variance of increment count at age is small and homogeneous only under the fastest growth condition (Norway Pond). The increasing variance with age in the other conditions leads to the speculation that some of these larvae were unknowingly starving. However, since the slope of the regression line for the Norway Pond condition is significantly different than 1 incre- ment/d, this result cannot be dismissed. There would be obvious value in repeating these experiments. Gef- fen also found that increment formation did not begin before yolk-sac absorption and was in agree- ment with the other investigators on this point. The literature (Table 1) shows only one case (Oncorhyn- chus nerka) where independent investigators have confirmed daily increment deposition (Wilson and Larkin 1980; Marshall and Parker 1982). Geffen (1982) found that increment deposition was also a function of growth in Scophthalmus maximus (Table 4) under various conditions of temperature and photoperiod. Under two conditions— 1) 20°C, constant light, and 2) 24°C, 12L:12D— increments were deposited daily For all other conditions in- crements were not daily. Under all conditions, deposition rate was a function of length. Although Geffen did not point this out, comparisons of growth at different temperatures can also be drawn from the data. Larvae were grown under 20°C and 24°C, both under a 12L:12D cycle. Larvae grew faster and deposited more increments at 24°C. Such differences in temperature might be used to explain differences in increment deposition except that the other case 93 FISHERY BULLETIN: VOL. 84, NO. 1 TJ o c o co -O [0 o CO to o 03 CO >, d) c TJ 0) CD CD CD CO c g o c to TJ CD i_ 03 OJ h_ >> k_ O o 03 g to o o. OJ TJ OJ E OJ i_ o c o 6 I CD to c o ■5 c o o CD £ \_ 03 CD cr cT CD a O CD Zi ToS O CD g ® E m els- < OJ CO c CD CD O CD O- CO CO c c o c TO c o o l_ to CD CO 3 c CD o N CO CO CO > k_ CD ro g CO to OJ c to CO c 03 w 3 CO C o o CO cn cn CD to cn i c cd -j; O 03 CD TJ cd g.01 ^ II cn 03 CO II c to 111 Ol., CD C 22 > - cn S tx < CD CO CD CO ^ >- "O £ co CD to CO CD o o TJ If) 4 o co o' c_ CX) Q CO _l eg oo t_ TJ o n o in oo n> *, cd CD CD co - to «- >. to CD O O CO D) c: CO CD CD C Q. CO 5"-C CD Ol CVI Z CO cd (!) T~ CD r CO (1) e CD o O to "^ CD £ CO CM ID CM CO 1— TO CO £ ^ to co cn c ~" g 2 ro -■ o 2 ~ o = £ -1 9 ° in t- CO CO c c c o C i5 to Q. C O CO c Q o to ■"3- CO OJ CD r^ cd CO 00 o CO CO E oo o o ^ o CM CO to CD 75 E TJ c CO cd CD o CO CO cn cn CD c o je- ez _co Q. o id c o CO to cn cn CD CD E o.2 o o oo r-- tM CM o § E = 1 JO ^ o o r-- CM C\J Q CD O h- CO CD cn CD • CD 5J ■— ■C t t t! o o o ro a> CD OJ ^ ID ID CD S 00 CD CO cn c -c o ■^ g co cn c CD CD T3 C O Q. to CD i— o o c O to cn is CD CD CL O TJ ID "* CO cn to c >> '\— i— *- <^ to ct: t- CO CD O "53 j= jd o CO cn i c CD -C »? O CO CD CD •o cn C CO CD ^ o CJ o 1o 0) to >- CO TJ CO CD c: CD CD CD TJ CDTJ CD ^ o ■" u CJ. 3 g CO CD c CD .C &g ?2 to CD • C OJ? ^ CO CD CD tj cn C CO ° cr o o CD c !cz o To TJ lo CD-*- c o fc OJ °^ TJ O c „ °> ■= II CL~ OJ g CD TJ O g-E o) CO CO CO CO ^- to c to c g CO cn CD O ID CD CJ TJ CD 5 "° CO c c o i& to CD co S- 2 O CD O CD CO o CD c_ >. CM co CO CO q3 0) co . c ®se cn £ CD j O 33. o _l CD o -I 2 o g>E Uj co CO CO CD a> CD >. >> o CO CD n o ii~T3 OJ CD O CD" « C CD co CD TJ C eg TJ O O T3 C CO cd ±; C CD 5 r r- ^ T3 CO ^ CO CD Is" 5- to CD X3 00 a: "a CO cr § 3 ZZ. CO CO 3 3 -C «•* o to 3 Is a cd E CO CO .to 15 P c o §> a tj CD U. CD -J CO Q> CO , ■B tt o ^ m 00 r-- CD C CO E CO CO co CD CO CO O CO a ■Q c- CD t- (11 3 S C (11 E CU -J § , — , f ^r CD co CD CD T— ^^ CD CO J£ CD C TJ o CO ~3 tr to 3 15 ■c .eg Q. CD CD C S CO O O CO 3 s § CM CO CO c CO c g to CO CD k_ cn CD £ 1o ^ ^: CO to CO CO TJ -C CJ « ,_ CO CO CD E CD E CO to o CD CD to >s >> 00 CD CO CD 3 CO 3 ■C O I- O "ir o S> c ■* O 94 JONES: DETERMINING AGE OF LARVAL FISH c o ■5 c o o O) c 1— to CO CC CO J* CO c J£ ra C (0 _i *"* _l o o 00 CM CM cT co 0) o oi 0) c V) c fed 3 wk fed versus starved excess 1x/d 4x/d ambient >10°C <10°C o o CM o 2 C Q CM 5qj CM ■* ""t t- CM CM o c o to o " CD c CO to C_ s s « _1 CO — o Tt *~ l_ £ ■Jo 5 to >-o >« >^ 11 (0 5 ~x CO to CO to CM > > > o o o T- O CO 09 cb CD CD CD o «l CO O) . c *-a CO c O CO CO CO "O CD C CO 9 c II o o o o> "O c o o c « c?S Q. co : CO o CM o CM r™ CO Q Q Q CM O CM i — -1^1 n Zj ■* CM ■>* CM CM r- T— •o "0 T3 O CO * co CM CO co J) to CD O) o» O) CD CD CD CD CO CO CD if en Q. CD o ° c E CD t: . k_ .Q e CD C CO o CO CO 5 to CD CO ■o c CO c/o c c Q CM 5 CM > i O) co _ CD O CD C CD o o o ■* CM CM o c O CO o — cm" co c "O ?i co aj -o « CD to c •o o o 3 o c CO o CO to c ^ 9 • CM ff CO ■* T- CM Q CM ^ -J CM CM CO t- CO CM J> CD CD CD C CD r> e CO CM CM : O- CD O C o CD O 2 S , c CD CD ° © CO "O -Q CD CO 0 <- <= 5"° «? to a? ° 2 2 a. oi CD CO CO CD CD CO E "o "c: c CO JZ II a. e CD "O O Q. C co CD , c CO '^ *-a CD C O CO CD CD ■O CD C CO ° C 11 , c CD CO V CD O 5 o ^5 -o -a c CO III I CO CO CO CO 2 O c CO > O) o c CM I--- CDS CO "O CD c :CD g« o CD U .2 | £ a. CD CO a> CO c g CO CO CD CO CD CO O) . c c CD C n CO o co CO CD O) CD O O CO 211 T3 c CO CO CD O c S co ■o to c O o LO CM t^ CO 00 ^^CO 00 CM to O) CD CD 111 o o CM 00 CO CD . C CD C CO c o ro CD CD T3 CD C <0 II o CJ T3 CO a m ,< — c c O CD O • ^ ra E £~ CO CD Ci CD c >" co O II CO CO to -c o c ^, -c CD C !c 0 fo E o CO CD T3 CO O co -O k_ m 2 •<*■ CO O o CO O SI I- CD T3 8| O c CD CD O •o c CO C CM O 00 to a> co to •c t-> o <2 5 =*• ^^ t: co o -c o CO c — O CO E CO >. o < C CM CO 00 ,__o> .CO CO E CO 2 CD CO CO CD CD O >. CM CO c O) CO CO "O "-* co m c >- 0 Laroche 1 (1982) CO CO C- CD CD ^ CJ 0 rr 0 CO ■ 3 ^_ 1 2 fs CD S C CO CT 3 8 § CO u c CO E CD "to w 0 CO CO > o 00 CD CD co rr Q. Q. CO co OJ CO to -Q CD CO c o co JC Is c5 2 9- CD CD o c o SI CO 3 CD t S CD > c CO CD CD CJ co co co CD CD CD CM CO o> l — — ' c CO CO 0 to 3 E CO "5 3 E s X a CO 0 0 E CO w and (1982) O O CO co ,_ C >- c CD CD CO » ^^ .0 O) tr r~- co rr C 3 CD f^ ■* 2 CD CO 13 C7> CO en 5? X 3 C c ZZ* O CO CO rr 3 5 H E CO CO CO 0 CO CO S-i co fc |1 CO C 95 FISHERY BULLETIN: VOL. 84, Na 1 Table 2.— Relationship between incubation time, egg size, and initial increment deposition: Determing whether species with long incuba- tion and large eggs initiate increment deposition on or before hatch, while species with short incubation and small eggs initiate increments at first feeding or yolk sac absorption, ysa = yolk sac absorption. Initial Egg Temper- Incubation increment size Species ature Source time Source deposition Source (mm) Source Clupea harengus =10°C - Blaxter (1969) =18 d Blaxter (1969) 4-5 d ysa See Table 1 0.9-1.7 Blaxter (1969) Engraulis 11°-21°C Lasker (1964) 1-5d Lasker (1964) =5 d Brothers et al. ^2 mordax (1976) Fundulus 24°-30°C Radtke (1978) 14 d Radtke (1978) Before hatch Radtke (1978) 2 Armstrong and heteroclitus Child (1965) Gadus morhua 4°C Radtke and Waiwood (1980) 19 d Radtke and Waiwood (1980) 1 d Radtke and Waiwood (1980) 1.1-1.6 Blaxter (1969) Menidia menidia 19.4°- Barkman 7-10 d at Barkmann Before hatch Barkman 1.2 Barkmann and 21.6°C (1978) 23°-25°C and Beck (1976) from regression (1978) Beck (1978) Morone saxatilis 18°C Jones (1984) 2 d Jones (1984) 6-9 d Jones (1984) Parophrys 20°C Laroche et al. 3-3V2 d Laroche et al. 4-5 d Laroche et al. vetulus (1982) (1982) (1982) Pseudopleu- 5°-8°C Radtke and 14 d at McPhee' 9-10 d Radtke and 0.8 Smigielski and ronectes Scherer 8°C Scherer Arnold americanus (1982) (1982) (1972) Tilapia nilotica 27°C Tanaka et al. (1981) 4 d Tanaka et al. (1981) At hatch Tanaka et al. (1981) 'Grace McPhee, P.O. Box 210972, Auke Bay, AK 99821, per. commun. summer 1983. Table 3.— Otolith increment deposition for larval fish maintained in the laboratory over a known time span. Are Known-age increments Number Species Source span < daily? Validation of fish Lepomis Taubert and Coble =6-176 d yes Correspondence be- gibbosus (1977) after swim up tween age and rings Lepomis Taubert and Coble =6-125 d yes Correspondence be- macrochirus (1977) after swim up tween age and rings Micropterus Schmidt and Fabrizio Between 47 and 81 yes Correlation between 98 salmoides (1980) rings change in ring count and time interval Oncorhynchus Wilson and Larkin Between 14 and 26 yes Slope = 1 ring/d 100 nerka (1980) rings Platichthys Campana and Neilson 8-10 mo old yes Slope = 1 ring/d 13 (in situ) stellatus (1982) 81 (temp and light) Stolephorus Struhsaker and yes Correspondence be- 174 purpureus Uchiyama (1976) tween rings and days of daily deposition (24L, 20°C) would be an anomaly under this hypothesis. Ten studies have investigated deposition rates under suboptimal, extreme or varying conditions (Table 4). These studies are important to the under- standing of the underlying mechanisms causing in- crement deposition. Two studies, one by Radtke and Dean (1982) and one by Taubert and Coble (1977), demonstrated disruption of daily increment forma- tion under extreme or abnormal changes in photo- period. Taubert and Coble (1977) found that in simulated winter conditions, cold temperature and shorter photoperiod resulted in cessation of incre- ment formation in Lepomis cyanellus. At and below temperatures of 10°C, growth and increment deposi- tion ceased. If such changes occurred gradually, as occurs in the normal lifetime of fish, acclimation to these temperature changes might be expected through most of the temperature range. Within nor- mal physiological limits (especially where some growth continued), increment deposition would be assumed to continue regularly. However, Marshall and Parker (1982) also found that temperatures below 10°C resulted in cessation of increment deposi- tion in sockeye salmon. Hence two studies have shown that increment deposition is not maintained 96 Table 4.— Otolith increment deposition for known-age larval fish under experiments where various culture conditions were tested. ■ Source Conditions of growth Species Light Food Temp Other tank size 120 L, 500 L, Effect on increment deposition Clupea Geffen (1982) Increment deposition rate was re- lated to growth rate. Also, larvae harengus 310 m3 4,440 m3 grew faster in bigger container and deposited more rings. Fundulus heteroclitus Radtke and Dean (1982) Multiple L/D con- ditions 24°C 30°C Temperature affects growth rate, but not increment deposition. Increment deposition rate disrupted under con- stant dark or under <24-h photo- period. Lepomis cyanellus Taubert and Coble (1977) 15L/9D 10L/14D 4°-25°C Fewer hours of light and lower temperature resulted in cessation of ring deposition. At 10°C or less, growth ceased, as did increment formation. Morone saxatilis Jones (1984) 14L/10D Fed, starved, intermittent 18°C Increment deposition rate was dis- rupted during periods of starvation. starved, then Increments not daily in sagittae dur- fed ing 2-3 mo under optimal conditions. Oncorhynchus nerka Marshall and Parker (1982) Fed Starved <10°C >10°C Starvation for 10 d did not affect in- crement deposition. Temperatures <10°C resulted in cessation of incre- ment formation. Oncorhynchus tshawytscha Neilson and Geen (1982) 24D 24L 12L/12D 4x/d 1x/d 11°C 5.2°C Formation of increments was related to feeding frequency. Temperature affected width of increment, not deposition rate. Photoperiod had no effect. Salmo salar Geffen (1983) 24D 6L/6D 12L/12D 8°C 10°C 15°C Rate of ring deposition increased with increased light and temperature. Scophthalmus maximus Geffen (1982) 24L 6U6D 12U12D Fed Starved 20°C 24°C Daily increments formed under 24L-20°C and 12L/12D-24°C. Starva- tion and 6L/6D interrupted increment formation. Increment formation related to growth rate. Tilapia mossambica Taubert and Coble (1977) 24L 24U12D 15U9D Every 3 h Every 6 h Intermittent Daily increments formed under 24-h photoperiod, not under 36-h cycle nor constant light. Subdaily incre- ments induced. No effect from feeding cycle. Tilapia nilotica Tanaka et al. 12U12D 3 h before dark Formation of increment triggered by (1981) 18L/6D 6U18D 3 h after light light stimulus. Feeding time had no effect under 12L/12D. below certain temperatures. In two other studies where temperatures ranged from 24°C to 30°C (Radtke and Dean 1982) and from 5.2°C to 11°C (Neilson and Geen 1982), these temperatures af- fected thegrowth rate and width of increments, but did not alter the increment deposition rate Six studies looked at the relationship between feeding and daily increment deposition. Jones (1984), Geffen (1982), and Marshall and Parker (1982) showed opposite effects of starvation on increment deposition. Jones (1984) found that starvation of young larvae for 2 wk resulted in deposition of only one increment every other day. However, in addition to lengthy starvation, the effect of short-term, in- termittent periods of starvation was also studied and resulted in nondaily increment formation. Geffen (1982) found that starvation interrupted deposition in larval turbot, while Marshall and Parker (1982) found that starvation for 2 wk had no effect on daily deposition in sockeye salmon. Long-term starvation experiments test for interruption of increment deposition under extreme conditions. lb age larvae in the field, it is important to determine the mini- mum number of consecutive days of starvation need- ed to affect increment deposition. Once these values are known, it is important to determine whether field larvae actually experience these levels of deprivation. Three studies looked at feeding time or frequen- cy on increment deposition. Neilson and Geen (1982) found that feeding frequency could induce forma- 97 FISHERY BULLETIN: VOL. 84, NO. 1 tion of subdaily increments in Oncorhynchus tshawytscha. Both Tanaka et al. (1981) and Taubert and Coble (1977) found that feeding time had no ef- fect on increment deposition in larval mouthbrooders (Tilapia nilotica and T. mossambica). Little agreement has been reached in these studies concerning the effect of light, temperature, or feeding on increment formation. The effects of variability in temperature, food, salinity, and other factors (extreme photoperiods would not be en- countered) relate directly to the problems of ac- curately aging larvae from the field. At the moment, environmental effects appear to be species-specific. Indeed, specific tests of the effect of suboptimal con- ditions (which are likely to occur in the field) on in- crement deposition have rarely appeared in the literature Such analyses, conducted for more species, might confirm the conventional wisdom that deviation from daily deposition rate is abnormal. However, the questions raised by the studies re- viewed here (Table 4) remain to be fully addressed or dispelled. APPLICATION IN THE FIELD Current Applications The ability to age larval fish precisely provides more accurate estimates of growth, mortality, and the ability to discern the effects of environmental variables on the first year of life Rapid growth in the first months of life has commonly been thought to be critical to survival. Evidence in support of this hypothesis (Brothers et al. 1983) and contrary to it (Methot 1983) exists. The otolith increment aging technique has been Table 5. — Application of the otolith increment aging technique in field grown larvae. Species Source Based on prior validations (validations in Table 1) Validation source Sample size Application Ammodytes Scott (1973) no 71 dubious Clupea Graham and Joule controversial See Table 1 for 545 harengus (1981)- Geffen (1982) found deposition details Townsend and depended on 300 Graham (1981) growth rate. Gjdsaeter and Lough et al. (1982) 0iestad (1981) found deposition 311 was daily. See Table 1 for Jones (1985) details. 481 Engraulis Methot and yes Brothers et al. 587 mordax Kramer (1979) (1976) Fundulus Radtke and Dean yes Radtke and Dean not heteroclitus (1982) (1982) given Gadus mgrhua Gjdsaeter and Tilseth (1981) yes Radtke and Waiwood (1980) 30 - Steffenson (1980) yes Radtke and Waiwood (1980) 138 Haemulon Brothers and no, but refers to =306 flavolineatum McFarland (1981) data as otolith age Halichoeres Victor (1982) yes marked juveniles 10 bivittatus Lepomis Taubert and Coble yes Taubert and Coble = 150 macrochirus (1977) (1977) Back-calculated growth. Determine hatching dates and de- lineate cohorts which are followed through time. Determine hatching dates and as- sess growth rates of larval cohorts. Noted cessation of growth in winter. Use age to delineate growth. Fit Gompertz function of length-at-age data. Determination of within-season growth differences based on uncer- tainty in otolith aging. Fit Gompertz function to length-at- age data to obtain growth rates. Also mention that starvation slowed incre- ment deposition. Compare length-frequency histo- grams with increment-frequency his- tograms. Show relationship between hatching and lunar cycle. Regression of age estimated from morphologic development versus in- crement counts. Back-calculated hatch date from in- crements. Compare these to field observations of spawning time. Correspondence between otolith microstructure and events in the life history. Derive "otolith" growth rates. Determine daily deposition of incre- ments and use to determine settling pattern. Allometric relationship between oto- lith length and fish length tested for 2 lakes. 98 V JONES: DETERMINING AGE OF LARVAL FISH applied to larval field populations of many species of fish (Table 5). Most applications have been based on laboratory validation of daily increment deposi- tion for the individual species studied. Some have not. Methot and Kramer (1979), based on validation of daily increment deposition by Brothers et al. (1976), obtained growth rates for wild populations of Engraulis mordax by fitting a Gompertz function to length-at-age data. Various other field applications of the increment aging technique are listed in Table 5. Of special interest is a comparison of growth estimates for Parophrys vetulus from modal progres- sion of length frequencies and otolith increments (Laroche et al. 1982). Growth based on the increment count method was 2-3 times faster. If the increment count method proves to be accurate, then mortality estimates could be considerably changed. For at least four species listed in Table 5, labora- tory validation was not conducted. These applica- tions assume a given age at initial deposition and daily increment deposition thereafter. The validity of these assumptions depends on the species and on the sensitivity of the application to inexactness in the age estimation. For example, controversial results have been obtained for larval herring, Clupea harengus. Geffen (1982) showed that growth rates could be overestimated by as much as three times the actual rate However, analysis of Gulf of Maine herring data (Jones 1985) showed that differences in growth between larvae hatched early and late in the season could be drawn. Until sensitivity analyses, laboratory verification, or other evidence exists to assure daily increment formation as a universal phenomenon under suboptimal conditions, there will be some doubt about the accuracy of aging field- captured larvaa Transition from the Laboratory to the Field A question that remains to be answered when applying laboratory-derived increment deposition Table 5.— Continued. Species Source Based on prior validations (validations in Table 1) Validation source Sample size Menidia Barkman et al menidia (1981) Morone Brothers et al. saxatilis (1976) yes no Barkman (1978) 105 (lab) Application Oncorhynchus nerka Wilson and Larkin (1982) yes Wilson and Larkin (1980) 64 Parophrys vetulus Laroche et al. (1982) yes Laroche et al. (1982) 331 Rosenberg and Laroche (1982) yes Laroche et al. (1982) 233 Pseudopleu- ronectes Radtke and Scherer (1982) yes Radtke and Scherer (1982) 120 amencanus Stolephorus purpureus Struhsaker and Uchiyama (1976) yes Struhsaker and Uchiyama (1976) 213 Thalossoma bifasciatum Victor (1982) Victor (1983) yes Victor (1982) marked juveniles 68 103 28 species of coral reef fish Brothers et al. (1983) no 210 Compare growth in lab and field. Calculate hatching dates. Compare growth between early and late hatched larvae. Correspondence between increment estimated age and spawning season. Growth through lifetime of juvenile. Relationship between fish weight and otolith size. Use daily in- crements as time marker. Determine growth of aged field lar- vae and fit Gompertz and von Ber- talanffy functions. Compare length- frequency and otolith techniques. Growth during metamorphosis. Re- late to age and transformation in morphology. Comparison of length-frequency and increment-frequency histograms for field larvae. Daily growth rate calcu- lated. Compare growth rates over time. Built growth curves based on age. Discussion of relationship to feeding. Preliminary study of growth rate dif- ference between areas. Determine daily increment deposi- tion. Calculate pattern of settlement based on age estimate. Determine length of larval life prior to recruitment. Examine otoliths for marker between postlarvae to juvenile. 99 1 IOilL.ni UULiLiLillll. »VU. Ol, l*KJ. 1 rates to field populations is the constancy of deposi- tion rates between these environments. Most labora- tory studies have occurred under constant tempera- ture and salinity and under conditions of artificial food types and densities and low light intensities compared with the field. Often, increments from otoliths of laboratory-grown larvae are much fainter than those from otoliths of field-captured larvae Since field conditions can fluctuate to extents that have been shown to cause increment disruption in laboratory situations, a way to verify daily deposi- tion in the field would be an important contribution. A transitional step between the laboratory and the field has been made by Laurence et al. (1979) and 0iestad (1982). Laurence et al. (1979) raised known- age larvae in a flow through enclosure This study was designed to measure the growth and survival of fish larvae exposed to varying prey concentrations in the field. Modifications of this system could be used to study increment deposition in known-age lar- vae exposed to field conditions. 0iestad (1982) pre- sented a review of larval fish studies performed in enclosures. Gjrisaeter and 0iestad (1981) reared known-age larvae in large enclosures and determined increment deposition rates (Table 1). Few inves- tigators have used such enclosures for validation of otolith increment deposition rates for field simulated studies. Enclosures should prove particularly valuable for validation and simulation of suboptimal field conditions on growth and increment deposition. Statistical Applications Once the veracity of daily increment deposition is established, a wide variety of statistical methods can be used in otolith studies. Statistical methods that have been employed in larval otolith studies have been linear regressions to establish increment deposition rates and curve fitting techniques to es- tablish growth rates from length-at-age data. Linear regression has also been applied regardless of whether it actually fits the data. It is important to check for lack of fit, selection of the appropriate model, and weighting before applying linear regres- sion blindly. It is recommended that, when possible, confidence intervals and standard deviations be in- cluded in the data presentation. Investigators are beginning to relate increment widths, as indicators of growth, with environmen- tal conditions (Methot and Kramer 1979; Lough et al. 1982). When increment widths are correlated directly with environmental factors, either no correlations are seen (Neilson and Geen 1982) or correlations may be spurious. Problems exist in measuring the physical conditions to which the lar- vae have been exposed, especially since larvae may move from one area to another. In addition, there are questions concerning food availability and its concentration and patchiness. Another consideration in relating growth to environmental conditions is that, as the fish grows, the width of the outer incre- ments decreases proportionately to decreases in length. Better results might be obtained either with covariance analysis or by fitting a growth function to data then using the residuals in correlation tests. Investigations of residuals with exploratory tech- niques such as principal component analysis or canonical correlation might prove fertile Comparison of Scanning Electron and Light Microscopy Scanning electron microscopy (SEM) has been used to confirm otolith structure (Dunkelberger et al. 1980; Watabe et al. 1982) and to compare incre- ment counts with those obtained by transmitted light microscopy (Radtke and Waiwood 1980; Campana and Neilson 1982; Neilson and Geen 1982; Radtke and Dean 1982; Tsuji and Aoyama 1982; Ralston and Miyamoto 1983). Under optimal conditions, counts using both methods were equivalent except for lar- val cod. Radtke and Waiwood (1980), using SEM, determined that cod produced daily increments from hatch onward, while Gj«teaeter (1981), using a light microscope, did not observe increment formation un- til 4-5 d after hatch. Most investigators did not verify deposition seen with the light transmission microscope with SEM studies. Confirmation with SEM is highly desirable when increments are nondaily. However, extensive use of the technique for field surveys is prohibited by the additional cost and preparation time when compared with light microscopy. In cases where suboptimal or abnormal field conditions may result in nondaily increment formation (Jones 1984), SEM, used in conjunction with ancillary techniques, may assist identification of the proportion of larvae for which age is underestimated with light micros- copy. CONCLUSIONS The report of the otolith workshop held in Bergen, Norway (Anonymous 1982) stated that the ap- pearance of increments in otoliths of larval fish living in diverse habitats and representing many families, argues strongly for the universality of this phenom- enon. Validation that these increments are indeed, 100 JONES: DETERMINING AGE OF LARVAL FISH deposited daily has been reported in 17 out of 20 species (Table 1) grown under optimal laboratory conditions. However, evidence exists that daily deposition can be interrupted under suboptimal and abnormal conditions, or can be dependent on growth rate (Table 6). When the effect of photoperiod is ig- nored (changes in photoperiod are very gradual in the field), more than 50% of the tests under subop- timal and extreme conditions have shown nondaily increment deposition rates. For other species, tests under suboptimal conditions were not conducted and the effect of these conditions on increment deposi- tion rate is undetermined. The effect of varying con- ditions on the age at initial increment deposition has also not been addressed. To apply the otolith aging technique to fish from the natural environment, the scientist must either assume that larvae sampled grew under optimal conditions (those exposed to suboptimal conditions died) or verify that the species almost always deposit daily increments under field encountered conditions, or establish the error bounds for the relationship between age and incre- ment count. Attempts to clarify the natural phenomena that drive daily increment formation have given con- flicting results. Photoperiod, feeding periodicity, and temperature fluctuations have all been cited as causing daily increment formation. When these fac- tors are within normal ranges, it is likely, for most larvae, that deposition is daily. However, for larvae experiencing conditions outside tolerable ranges or abnormal conditions, the period of formation is likely to deviate from daily deposition. It is important to determine whether the minimum exposure to subop- timal conditions which result in nondaily deposition is actually experienced by larvae in the field. These hypotheses are amenable to further testing. More basic research on the causation of increment deposi- tion or more extensive testing under a variety of con- ditions for a given species will yield more informa- tion. In situ testing with known-age larvae in enclosures which closely mimic field conditions could yield valuable results. The Bergen otolith workshop report (Anonymous 1982) has recommended that in- crement deposition be verified for each new species, under a variety of test conditions. Two issues, cost effectiveness and accuracy, are im- portant in determining whether the otolith incre- ment technique is preferable to length-frequency analysis. Recommendations made in the report from the Bergen otolith workshop (Anonymous 1982) are that "the precision of an age determination ... be tested against other available methods ... by a cost benefit analysis (i.e is enough precision gained by Table 6— Incidence of nondaily increment deposition for species reared under suboptimal and extreme conditions. Stars (*) indicate nondaily deposition caused by exposure to suboptimal conditions; triangles (A) indicate nondaily deposi- tion caused by exposure to extreme conditions; circles (O) in- dicates no interruption of daily deposition. Tank Species Light Food Temp size Clupea harengus • Fundulus heteroclitus • ,A O. Lepomis cyanellus • • Morone saxatilis *,A Oncorhynchus nerka O • 0. tshawytscha O O • O Salmo salar A * Scophthalmus maximus O.A A Tilapia mossambica A O T. nilotica * o using this method to pay the costs and effort in preparation)". A good example would be the results shown in Laroche et al. (1982) when the otolith method was compared with modal progression of length frequencies, estimated growth rates differed by a factor of 2-3. Benefits should also include non- monetary considerations, such as decrease in error which will propagate through estimates based on age determinations (i.e, growth and mortality). Sensi- tivity analyses can be used to show situations where more accurate estimates are necessary. Specific recommendations for improving reliability and replicability are discussed in the Bergen otolith workshop report (Anonymous 1982). In addition to these, Brothers3 has suggested that other otoliths, ( such as the lapillus, be used in analysis. Aging by the otolith increment technique is a powerful tool. Not only can population estimates of growth and mortality be refined, but growth of in- dividuals can be obtained. Issues such as the impor- tance of environmental factors to survival, the pro- portion of fast growing larvae to recruitment, and demonstration of compensation in field larvae may become easier to address with the availability of this technique However, it is equally important to make sure that the technique is based on good scientific technique ACKNOWLEDGMENTS I thank David Bengtson, John Forney, Saul Saila, Ann Durbin, and Bernard Skud for their thoughtful review of this manuscript and Walter Berry for his many helpful suggestions and discussions. 3Edward Brothers, 3 Sunset West, Ithaca, NY 14850, pers. com- mun. September 1983. 101 LITERATURE CITED Anonymous. 1982. Report on an Otolith Workshop in Bergen, January 18-29, 1982. Dep. Fish. Biol., Univ. Bergen and Inst. Mar. Res., Bergen, Norway, 28 p. Armstrong, P. B., and J. S. Child. 1965. Stages in the normal development of Fundulus hetero- clitus. Biol. Bull. (Woods Hole) 128:143-168. Barkman, R. C. 1978. The use of otolith growth rings to age young Atlantic silversides, Menidia menidia. Trans. Am. Fish. Soc 107: 790-792. Barkman, R. C, D. A. Bengtson, and A. B. Beck. 1981. Daily growth of the juvenile fish (Menidia menidia) in the natural habitat compared with juveniles reared in the laboratory. Rapp. P.-v. Reun. Cons, int Explor. Mer 178: 324-326. Barkmann, R. C, and A. Beck. 1976. Incubating eggs of the Atlantic silverside on nylon screen. Prog. Fish-Cult. 38:148-150. Blaxter, J. H. S. 1969. Development: eggs and larvae In W. S Hoar and D. J. Randall (editors), Fish physiology, Vol. Ill, p. 177-252. Acad. Press, N.Y. Brothers, E. B. 1978. Exogenous factors and the formation of daily and sub- daily growth increments in fish otoliths. [Abstr.] Am. Zool. 18:631. Brothers, E. B., E. D. Prince, and D. W. Lee. 1983. Age and growth of young-of-the-year bluefin tuna, Thun- nus thynnus, from otolith microstructure In E. D. Prince and L. M. Pulos (editors), Proceedings of the International Workshop on Age Determination of Oceanic Pelagic Fishes: Tunas, billfishes, and sharks, p. 49-59. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 8. Brothers, E. B., C. P. Mathews, and R. Lasker. 1976. Daily growth increments in otoliths from larval and adult fishes. Fish. Bull., U.S. 74:1-8. Brothers, E. B., and W N. McFarland. 1981. Correlations between otolith microstructure, growth, and life history transitions in newly recruited french grunts [Haemulonflavolineatum (demarest) Haemulidae]. Rapp. R- v. Reun. Cons. int. Explor. Mer 178:369-374. Campana, S. E., and J. D. Neilson. 1982. Daily growth increments in otoliths of starry flounder (Platichthys stellatus) and the influence of some environmen- tal variables in their production. Can. J. Fish. Aquat. Sci. 39:937-942. DUNKELBERGER, D. G., J. M. DEAN, AND N. WATABE. 1980. The ultrastructure of the otolithic membrane and otolith in the juvenile mummichog, Fundulus heteroclitus. J. Mor- phol. 163:367-377. Geffen, A. J. 1982. Otolith ring deposition in relation to growth rate in the herring (Clupea harengus) and turbot (Scophthalmus max- imus) larvae Mar. Biol. (Berl.) 71:317-326. 1983. The deposition of otolith rings in Atlantic salmon, Salmo solar L., embryos. J. Fish Biol. 23:467-474. Gj^saeter, H. 1981. Dagsonelesing som metode i aldersstudier pa fisk, med eksempler pa anvendelse pa tropiske og boreale arter. Ph.D. Thesis, Univ. Bergen, Bergen, Norway, 172 p. GJOSAETER, H., AND V. 0IESTAD. 1981. Growth patterns in otoliths as an indication of daily growth variations of larval herring (Clupea harengus) from an experimental ecosystem. ICES CM. 1981/H:31. GJOSAETER, H., AND S. TlLSETH. 1981. Primary growth increments in otoliths of cod larvae (Gadus morhua L.) of the arcto-norwegian cod stock. Fisk. Dir. Skr. Ser. Havunders. 17:287-295. Graham, J. J., and B. J. Joule. 1981. Preliminary evaluation of the 1980 larval year class strength of coastal Maine herring. N.A.F.O. (Northwest Atl. Fish. Organ.), Sci. Counc Rep. Doc 81/IX/140, 22 p. Jones, C. M. 1984. The otolith increment technique: Application in larval fish. Ph.D. Thesis, Univ. Rhode Island, Kingston, 123 p. 1985. Within-season differences in growth of larval herring, Clupea harengus. Fish. Bull., U.S. 83:289-298. Laroche, J. L., S. L. Richardson, and A. A. Rosenberg. 1982. Age and growth of a pleuronectid, Parophrys vetulus, during the pelagic larval period in Oregon coastal waters. Fish. Bull., U.S. 80:93-104. Lasker, R. 1964. An experimental study of the effect of temperature on the incubation time, development, and growth of Pacific sar- dine embryos and larvae Copeia 1964:399-405. Laurence, G. C, T. A. Halavik, B. R. Burns, and A. S. Smigielski. 1979. An environmental chamber for monitoring "in situ" growth and survival of larval fishes. Trans. Am. Fish. Soc 108:197-203. Lough, R. G., M. Pennington, G. R. Bolz, and A. A. Rosenberg. 1982. Age and growth of larval Atlantic herring, Clupea harengus L., in the Gulf of Maine-Georges Bank region based on otolith growth increments. Fish. Bull., U.S. 80:187-199. Marshall, S. L., and S. S. Parker. 1982. Pattern identification in the microstructure of sockeye salmon (Oncorhynchus nerka) otoliths. Can. J. Fish. Aquat. Sci. 39:542-547. Methot, R. D, Jr. 1983. Seasonal variation in survival of larval northern an- chovy, Engraulis mordax, estimated from the age distribu- tion of juveniles. Fish. Bull., U.S. 81:741-750. Methot, R. D, Jr., and D. Kramer. 1979. Growth of the northern anchovy, Engraulis mordax, lar- vae in the sea. Fish. Bull., U.S. 77:413-423. Neilson, J. D, and G. H. Geen. 1982. Otoliths of chinook salmon (Oncorhynchus tshawytschay. Daily growth increments and factors influencing their pro- duction. Can. J. Fish. Aquat. Sci. 39:1340-1347. 0IESTAD, V. 1982. Application of enclosures to studies on the early life history of fishes. In G. D. Grice and M. R. Reeve (editors), Marine Mesocosms, p. 49-62. Springer-Verlag, N.Y. Pannella, G. 1971. Fish otoliths: daily growth layers and periodic patterns. Science 173:1124-1127. 1974. Otolith growth patterns: an aid in age determination in temperate and tropical fishes. In T B. Bagenal (editor), The Proceedings of an International Symposium on the Aging of Fish, p. 28-39. Unwin Brothers, Surrey, Engl. Radtke, R. L. 1978. The formation and growth of otoliths in the estuarine fish, Fundulus heteroclitus. Ph.D. Thesis, Univ. South Carolina, Columbia, 57 p. 1980. The formation and growth of otoliths from redfish (Sebastes spp.) larvae from the Flemish Cap (Division 3M). N.A.F.O (Northwest Atl. Fish. Organ.), Sci. Counc Rep. Doc 80/IX/153 Ser. N225, 5 p. 1984. Formation and structural composition of larval striped 102 JONES: DETERMINING AGE OF LARVAL FISH mullet otoliths. Trans. Am. Fish. Soc. 113:186-191. Radtke, R. L., and J. M. Dean. 1982. Increment formation in the otoliths of embryos, larvae, and juveniles of the mummichog, Fundulus heteroclitus. Fish. Bull., U.S. 80:201-215. Radtke, R. L., and M. D. Scherer. 1982. Daily growth of winter flounder (Pseudopleuronectes americanus) larvae in the Plymouth Harbor estuary. In Pro- ceedings of the Fifth Annual Larval Fish Conference, p. 1-5. Louisiana Cooperative Fishery Research Unit, Louisiana State University, Baton Rouge, LA. Radtke, R. L., and K. G. Waiwood. 1980. Otolith formation and body shrinkage due to fixation in larval cod {Gadus morhua). Can. Tech. Rep. Fish. Aquat. Sci. No. 929, 10 p. Ralston, S., and G. T. Miyamoto. 1983. Analyzing the width of daily otolith increments to age the Hawaiian snapper, Pristipomoides filamentosus. Fish. Bull., U.S. 81:523-535. Ricker, W. E. 1975. Computation and interpretation of biological statistics offish populations. Fish. Res. Board Can., Bull 191, 382 p. Rosenberg, A. A., and A. S. Haugen. 1983. Individual growth and size-selective mortality of larval turbot (Scophthalmus maximus) reared in enclosures. Mar. Biol. (Berl.) 72:73-77. Rosenberg, A. A., and J. L. Laroche. 1982. Growth during metamorphosis of English sole, Paro- phrys vetulus. Fish. Bull., U.S. 80:150-153. Schmidt, R. E., and M. C. Fabrizio. 1980. Daily growth rings on otoliths for aging young-of-the- year largemouth bass from a wild population. Prog. Fish- Cult. 42:78-80. Scott, J. S. 1973. Otolith structure and growth in northern sand lance, Ammodytes dubius, from the Scotian shelf. Int. Comm. Northwest Atl. Fish., Res. Bull. 10, p. 107-115. Smigielski, A. S., and C. R. Arnold. 1972. Separating and incubating winter flounder eggs. Prog. Fish-Cult. 34:113. Steffensen, E. 1980. Daily growth increments observed in otoliths from juvenile East Baltic cod. Dana 1:29-37. Struhsaker, P., and J. H. Uchiyama. 1976. Age and growth of the nehu, Stolephorus purpureus (Pisces: Engraulidae), from the Hawaiian Islands as indicated by daily growth increments of sagittae Fish. Bull., U.S. 74:9-17. Tanaka, K., Y. Mugiya, and J. Yamada. 1981. Effects of photoperiod and feeding on daily growth pat- terns in otoliths of juvenile Tilapia nilotica. Fish. Bull., U.S. 79:459-466. Taubert, B. D, and D. W. Coble. 1977. Daily rings of otoliths of three species of Lepomis and Tilapia mossambica. J. Fish. Res. Board Can. 34:332-340. Townsend, D. W., and J. J. Graham. 1981. Growth and age structure of larval Atlantic herring, Clupea harengus harengus, in the Sheepscot River estuary, Maine, as determined by daily growth increments in otoliths. Fish. Bull., U.S. 79:123-130. TSUJI, S., AND T. AOYAMA. 1982. Daily growth increments observed in otoliths of the lar- vae of Japanese Red Sea bream Pagrus major (Temminch et Schlegel). Bull. Jpn. Soc Sci. Fish. 48:1559-1562. Victor, B. C. 1982. Daily otolith increments and recruitment in two coral- reef wrasses, Thalassoma bifasciatum and Halichoeres bivit- tatus. Mar. Biol. (Berl.) 71:203-208. 1983. Recruitment and population dynamics of a coral reef fish. Science 219:419-420. Volk, E. C, R. C. Wissmar, C. A. Simenstad, and D. M. Eggers. 1984. Relationship between otolith microstructure and the growth of juvenile chum salmon (Oncorhynchus keta) under different prey rations. Can. J. Fish. Aquat. Sci. 41:126-133. Watabe, N., K. Tanaka, J. Yamada, and J. M. Dean. 1982. Scanning electron microscope observations of the organic matrix in the otolith of the teleost fish Fundulus heteroclitus (Linnaeus) and Tilapia nilotica (Linnaeus). J. Exp. Mar. Biol. Ecol. 58:127-134. Wilson, K. H., and P. A. Larkin. 1980. Daily growth rings in the otoliths of juvenile sockeye salmon (Oncorhynchus nerka). Can. J. Fish. Aquat. Sci. 37:1495-1498. 1982. Relationship between thickness of daily growth in- crements in sagittae and change in body weight of sockeye salmon (Oncorhynchus nerka) fry. Can. J. Fish. Aquat. Sci. 39:1335-1339. 103 PATTERNS IN DISTRIBUTION AND ABUNDANCE OF A NONCOEVOLVED ASSEMBLAGE OF ESTUARINE FISHES IN CALIFORNIA Peter B. Moyle,1 Robert A. Daniels,2 Bruce Herbold,1 and Donald M. Baltz3 ABSTRACT The patterns of distribution and abundance of the fishes of Suisun Marsh, a portion of the Sacramento- San Joaquin estuary in central California, were studied over a 54-month period. Tbtal fish abundance in the marsh exhibited strong seasonality; numbers and biomass were lowest in winter and spring and highest in late summer. Freshwater inflow was highest in the winter and lowest in late summer, when salinities and temperatures were highest. Twenty-one species were collected on a regular basis; the 10 most abundant were Morone saxatilis, Pogonichthys macrolepidotus, Gasterosteus aculeatus, Hysterocarpus traski, Cottus asper, Spirinchus thaleichthys, Acanthogobius fl.avimanus, Catostomus occidentalis, Lep- tocottus armatus, and Platichthys stellatus. Another 21 species occurred in small numbers on an irregular basis. Twenty of the 42 species had been introduced to California since 1879. Of the 21 common species, 14 were residents, 4 were winter seasonals, and 3 were spring/summer seasonals. The resident species fell into two groups: a group of native species that were concentrated in small dead-end sloughs and a group of native and introduced species that were most abundant in the larger sloughs. The seasonal species were also a mixture of native and introduced species. Tbtal fish abundance and species diversity declined through the study period, which seemed to be related to strong year classes of some species early in the study and the prevalance of freshwater conditions late in the study. The structure of the fish assemblage was fairly consistent over the study period but changes are expected in the near future The structure of the Suisun Marsh fish assemblage was similar to that found in other river-dominated estuaries, despite the mixture of native and introduced species. The Sacramento-San Joaquin Estuary system is the largest estuary on the west coast of North America. It has been highly modified by surrounding urban, industrial, and agricultural development and by ex- tensive diversion and pollution of the freshwater that flows into it (Conomos 1979). It supports a diverse fish fauna of native and introduced species, but most previous studies have concentrated on species impor- tant to sport and commercial fisheries, especially striped bass, Morone saxatilis, and, to a much lesser extent, white sturgeon, Acipenser transmontanus; chinook salmon, Oncorhynchus tshawytscha; Ameri- can shad, Alosa sapidissima; and white catfish, Icta- lurus catus (Skinner 1972; Moyle 1976). Studies of other species have been few (Ganssle 1966; Turner and Kelley 1966; Baltz and Moyle 1982; Stevens and Miller 1983; Daniels and Moyle 1983), and there have been no community-level analyses equivalent to those conducted on estuarine fish communities in other 'Wildlife and Fisheries Biology, University of California, Davis, CA 95616. zWildlife and Fisheries Biology, University of California, Davis, CA; present address: Biological Survey, New York State Museum, Albany, NY 12230. 3Wildlife and Fisheries Biology, University of California, Davis, CA; present address: Coastal Fisheries Institute, Louisiana State University, Baton Rouge, LA 70803. parts of the world (e.g., Dahlberg and Odum 1970; Livingston 1976; Sheridan and Livingston 1979; Meeter et al. 1979; Blaber and Blaber 1980; Quinn 1980; Thorman 1982). The fish assemblage of the Sacramento-San Joaquin Estuary system is unusual because few of its component species are likely to have evolved together; it is composed of a mixture of introduced and native freshwater, estuarine, and euryhaline marine species (Table 1). The introduced species come from a number of geographic areas, while most of the native species have their centers of abundance in either the rivers upstream or the saltwater bays downstream from the estuary. There are no really comparable estuaries on the Califor- nia coast, although some of the much smaller and more saline estuaries south of the Sacramento-San Joaquin Estuary do have fish assemblages composed in part of introduced species (Allen 1982). We began in January 1979 systematic sampling of the fishes in Suisun Marsh on a monthly basis. Suisun Marsh was chosen as a study site because of its central location on the estuary, its proximity to the University of California, Davis campus, and the availability of earlier data from sporadic sampling by the California Department of Fish and Game The data indicated that the fish fauna was typical of the Manuscript accepted March 1985. FISHERY BULLETIN: VOL. 84, NO. 1, 1986. 105 FISHERY BULLETIN: VOL. 84, NO. 1 Table 1— Fishes collected in Suisun Marsh, Solono County, CA, in decreasing order of numerical abundance in our trawls. The principal environment of each species is coded as follows: A = anadromous, E = estuarine, F = freshwater, M = marine. Species Numbers Origin Striped bass, Morone saxatilis Splittail, Pogonichthys macrolepidotus Threespine stickleback, Gasterosteus aculeatus Tule perch, Hysterocarpus traski Prickly sculpin, Cottus asper Yellowfin goby, Acanthogobius flavimanus Sacramento sucker, Catostomus occidentalis Common carp, Cyprinus carpio Threadfin shad, Dorosoma petenense Staghorn sculpin, Leptocottus armatus Starry flounder, Platichthys stellatus Longfin smelt, Spirinchus thaleichthys Delta smelt, Hypomesus transpacificus American shad, Alosa spadissima Sacramento squawfish, Ptychocheilus grandis Chinook salmon, Oncorhynchus tshawytscha Hitch, Lavinia exilicauda Inland silverside, Menidia beryllina Goldfish, Carassius auratus Northern anchovy, Engraulis mordax Sacramento blackfish, Orthodon microlepidotus Pacific herring, Clupea harengeus White catfish, Ictalurus catus Bluegill, Lepomis macrochirus Mosquitofish, Gambusia affinis Black crappie, Pomoxis nigromaculatus Bigscale logperch, Percina macrolepida White sturgeon, Acipenser transmontanus Fathead minnow, Pimephales promelas Brown bullhead, Ictalurus nebulosus Rainwater killifish, Lucania parva Green sunfish, Lepomis cyanellus Pacific sanddab, Citharichthys sordidus Pacific lamprey, Lampetra tridentata Surf smelt, Hypomesus pretiosus Channel catfish, Ictalurus punctatus Black bullhead, Ictalurus melas Shiner perch, Cymatogaster aggregata Golden shiner, Notemigonus crysoleucus Warmouth, Lepomis gulosus Rainbow trout, Salmo gairdneri Longjaw mudsucker, Gillichthys mirabilis 24,154 E. North America (E) 11,250 Native (E) 9,956 Native (F-E) 7,693 Native (F-E) 4,639 Native (F-E) 1,786 Japan (E-M) 1,703 Native (F) 1,573 Asia (F) 1,088 E. North America (E) 985 Native (M) 849 Native (M) 650 Native (E) 450 Native (E) 218 E. North America (A) 140 Native (F) 96 Native (A) 56 Native (F) 50 E. North America (F-E) 45 Asia (F) 34 Native (M) 25 Native (F) 24 Native (M) 23 E. North America (F) 16 E. North America (F) 15 E. North America (F) 14 E. North America (F) 10 Texas (F) 10 Native (E) 9 E. North America (F) 6 E. North America (F) 5 E. North America (E) 4 E. North America (F) 4 Native (M) 4 Native (A) 3 Native (M) 3 E. North America (F) 3 E. North America (F) 3 Native (M) 3 E. North America (F) 1 E. North America (F) 1 Native (A) 1 Native (M) freshwater dominated portions of the estuary. The marsh is also of considerable interest because it is the largest brackish-water marsh in California. It is managed primarily as a wintering area for migratory waterfowl, but its importance as a nursery area for striped bass, salmon, and other fishes is being in- creasingly recognized (Baracco 1980). The purpose of this paper is to analyze the distribution and abun- dance of the fishes of the marsh in relation to each other, major environmental factors, and major crustacean species, during a 54-mo period. STUDY AREA Suisun Marsh is a large (ca 34,000 ha) tidal marsh located just downstream of the confluence of the Sacramento and San Joaquin rivers (Fig. 1). About 11,000 ha of the marsh consist of sloughs that are influenced by tidal action. The remainder consists of diked wetlands managed to attract wintering waterfowl (Baracco 1980) and for pasturage The sloughs are shallow (most are <2 m deep) and may fluctuate in depth as much as 1 m during extreme tides. Salinities have ranged from 0 to nearly 17 ppt in recent years, with the highest salinities occurring in late summer of drought years and the lowest salinities occurring annually in winter and spring when river outflows are highest (Baracco 1980). Because increased upstream diversion of water is threatening water quality in the marsh, major modifications to the water distribution system within the marsh are being made to ensure that salinites do not become too high for production of the plants that attract waterfowl. 106 MOYLE ET AL.: NONCOEVOLVED ASSEMBLAGE OF ESTUARINE FISHES Suisun City* Pey tonia Study Area Figure 1.— Locations of sample sites (*) in Suisun Marsh, Sacramento-San Joaquin Estuary, CA. During this study, two major habitat types were sampled: 1) small dead-end sloughs that were 7-10 m wide and 1-2 m deep and 2) Suisun Slough, which connected all the dead-end sloughs and was 100-150 m wide and 2-4 m deep. A third habitat, Montezuma Slough, was also sampled, but the data were not used here because our methods did not sample it ade- quately. This slough is deep (3-4 m), wide, and riverlike; it is the marsh's main source of freshwater. METHODS Sampling was conducted monthly at seven loca- tions throughout the marsh (Fig. 1), from January 1979 through June 1983, with the exception of December 1979 and October 1980. Four of the loca- tions were in dead-end sloughs (Peytonia, Boynton, Mallard, and Goodyear), one was a small slough open at both ends (Cutoff), and two were in Suisun Slough. Sampling was conducted biweekly from January 1980 through June 1981, but the samples for each month were lumped together for analysis, as the samples within months were comparable All samples were taken during the day, as 24-h studies conducted in April 1979 and 1980 did not exhibit any signifi- cant differences between day and night samples. The principal means of sampling was a four-seam otter trawl with a 1 x 2.5 m opening, a length of 5.3 m, and mesh sizes that tapered down to 6 mm stretch in the bag. At each location, the trawl was towed for either 5 min (small sloughs) or 10 min (Suisun Slough) at about 4 km/h. The longer periods were necessary in large sloughs because of the small catches that prevailed there Each location was sam- pled at least twice on each date This method of sampling was biased because large fishes probably avoided the trawl, and fishes that favor the emergent vegetation were undersampled, as were fishes in the upper part of the water column (Kjelson and Colby 1977). However, these problems were minimized by the narrowness and shallowness of most of the sampling sites; in any case such biases were consis- 107 FISHERY BULLETIN: VOL. 84, NO. 1 tent across the course of this study, so that com- parisons should be unaffected. In addition, two loca- tions on the marsh were sampled with a 10 x 1 m, 6 mm mesh, seine, on an irregular basis. An effort was made to seine every month but it was often not possible, as the sites were difficult to seine at ex- treme high or low tides. Fishes from each trawl were placed in washtubs of water to minimize mortality and then identified, measured to the nearest millimeter (standard length), and returned to the water as quickly as possible If more than 100 fish of any one size class of a species were captured, only the first 100 were measured; the rest were counted. Early in the study, samples of all fishes were weighed (wet weight, in gram), and a length/weight relationship developed for each species. This was later used to estimate the biomass of fish in each trawl. The shrimps Crangon franciscorum and Palaemon macrodactylus in each trawl were also counted. For the oppossum shrimp, Neomysis mercedis, an index of abundance was used, based on a l-to-5 scale, where "1" represented <3 individuals; "2", 3-50 shrimp; "3", 50-200, "4", 200-500, and "5", >500. The index was necessary because most N. mercedis probably passed through the net due to their small size (3-5 mm). Neverthe- less, they were present seasonally in most hauls, at times in enormous numbers. At each location, salinity and temperature were taken with a YSI S-C-T meter and transparency was measured with a Secchi disk. Tidal height was deter- mined from a tide tabla An index of monthly fresh- water outflow from the combined Sacramento and San Joaquin Rivers at Chipps Island was obtained from the California Department of Water Resources (unpubl. data). For analysis, all the data were summarized by site and month. A Spearman rank correlation analysis using data ranked by month (N = 52) was used for the initial analysis because many of the variables did not conform to a normal distribution. Because no single transformation could be applied to all the variables, nonparametric statistics were used as the most conservative method. We used 13 variables for the analysis (Table 2). In addition, rank abundance (by numbers) by month for the following species categories was used: 1) total striped bass, 2) year- ling and older striped bass, 3) young-of-year striped bass, 4) total splittail, 5) yearling and older split- tail, 6) young-of-year splittail, 7) total tule perch, 8) tule perch adults, 9) tule perch young-of- year, 10) total prickly sculpin, 1 1) yearling and older prickly sculpin, 12) prickly sculpin young-of- year, 13) carp, 14) longfin smelt, 15) delta smelt, 16) staghorn sculpin, 17) starry flounder, 18) threadfin shad, 19) Sacramento sucker, 20) yellowfin goby, and 21) threespine stickleback. Because only minor differences were found among the correlations associated with adult and juvenile striped bass, tule perch, splittail, and prickly sculpin, only the results for the totals for these species will be presented. Analyses were also run using the data from each trawl separately. Species were analyzed using both numbers and grams. Because these data were all of species abundances, a log-normal transformation was used to normalize them. The results were similar in most respects to the analyses using ranks so are not presented here However, because we were uncer- tain as to the validity of using ranked data for prin- cipal components analysis (PCA), we based our discussion on cautious inspection of the correlation matrix as generated. A principal components analysis was run using the correlation matrix (Dix- on and Brown 1977) of 1„ numbers of fish per trawl (N = 1,238), to produce groups of species that presumably were responding to the environment in the same general ways. Table 2. — Environmental variables used in the correlation analyses. Variable Units Notes Month series 1-54 January 1979 to June 1983 Water year 1-5 Begins in October of each year Salinity ppt Temperature °C Secchi depth cm Neomysis mercedis 1-5 index abundance Mean monthly 0-11 index California Department of outflow Water Resources Crangon franciscorum No./trawl Palaemon macrodacytlus No./trawl Fish species No./trawl Total fish numbers No./trawl Total fish biomass Biomass/ trawl Wet weight Species diversity Index Shannon-Weiner (H) RESULTS Environmental Variables Salinity and temperature were negatively corre- lated with river outflows (Table 3, Fig. 2). Salinity had a strong (P < 0.01) positive correlation only with Secchi depth. River outflows generally peaked in February, March, or April, as the result of run-off from melting snow in the Sierra Nevada. Lowest 108 MOYLE ET AL.: NONCOEVOLVED ASSEMBLAGE OF ESTUARINE FISHES Table 3. — Spearman rank correlation coefficients between fish species ranked by month by numbers and other variables ranked by month. Underlined values are significant at P > 0.05. CO CO CO sz o 3 CO o O O) c Q. 3 O CO o CD CD E ■o CO sz CO c Q. 3 U CO CD ■o c 3 n o ai C CO E c o -o := CD E U— CO c •D o a> a. s a. a> CD U o a. o o £ c CO CD SZ >> k_ ^-. a. CO CO fc_ ^ CD o sz CO CO £ CO ¥ O Q. CO a _l 1- CO CO Month series -0.42 -0.72 -0.51 -0.38 -0.53 -0.58 0.16 -0.10 -0.29 -0.09 -0.26 -0.15 -0.21 Temperature 0.54 0.28 0.08 0.21 0.49 0.49 0.41 -0.33 -0.41 -0.28 -0.55 -0.03 0.01 Salinity 0.62 0.24 0.53 0.14 0.43 0.38 -0.36 -0.14 0.18 0.17 0.24 0.13 -0.06 Secchi depth 0.09 -0.09 0.29 -0.18 -0.08 -0.04 -0.54 -0.09 0.33 0.31 0.52 0.06 -0.28 Outflow -0.74 -0.36 -0.49 -0.27 -0.62 -0.44 0.06 0.04 0.07 -0.12 0.16 -0.06 0.14 Neomysis mercedis -0.42 -0.02 -0.24 0.09 -0.45 -0.05 0.28 0.23 0.07 -0.25 -0.10 0.07 0.08 Crangon franciscorum 0.27 -0.01 0.05 0.06 0.46 -0.18 -0.03 -0.10 0.01 -0.59 -0.15 0.29 -0.23 Palaemon macrodactylus 0.43 -0.10 0.06 -0.10 0.34 0.20 0.16 -0.18 -0.30 -0.21 -0.32 0.14 0.23 No./trawl 0.67 0.64 0.72 0.53 0.46 0.45 0.07 0.21 0.16 -0.07 0.06 -0.18 -0.10 g/trawl 0.39 0.71 0.53 0.57 0.39 0.81 -0.06 -0.13 0.04 -0.24 0.13 0.04 0.04 Species/trawl 0.42 0.74 0.48 0.56 0.60 0.52 0.24 0.10 0.21 0.15 0.00 0.21 0.43 Diversity (H) -0.10 0.45 0.23 0.45 0.28 0.35 0.31 0.21 0.26 0.09 0.12 0.36 0.43 flows occurred from August through October. Sali- nity, temperature, and Secchi depth were generally lowest (0-1 ppt, 8°-ll°C, and 17-18 cm, respective- ly) when outflows were highest, and highest (4-9 ppt, 19°-23°C, and 25-40 cm, respectively when outflows were lowest. There is, however, considerable year- to-year variation in these cycles. When outflows were comparatively low (1979, 1981), salinities, temper- atures, and turbidities peaked at higher levels than they did in high outflow years. Because 1982 and 1983 were exceptionally wet years, virtual freshwater conditions prevailed throughout both years. Invertebrates Neomysis mercedis became very abundant in the marsh from April to June, but the population de- clined rapidly through the summer, reaching a low in October (Fig. 2). This pattern fits with previous studies of this species, which showed that its popula- tions generally followed the mixing zone up and down the estuary and were reduced at temperatures higher then 22°C and salinities >7 ppt (Orsi and Knutson 1979). In this study, N. mercedis abundance showed a significant positive correlation with out- flows and significant negative correlations with tem- perature, salinity, and turbidity (Table 4). It also showed a significant negative correlation (Table 3) with two of its major predators in the marsh, striped bass and yellowfin goby (Herbold 19854). Palaemon macrodactylus and Crangon francis- corum also showed seasonal patterns of abundance (Sigfreid 1980), but the patterns were much less marked than those of N. mercedis. Palaemon macro- dactylus were most abundant during July through October and least abundant during January and February, while C. franciscorum were most abundant in November and December and least abundant in January through March. Palaemon macrodactylus abundance therefore showed strong positive corre- lation with temperature and salinity and a negative correlation with outflows. Crangon franciscorum abundance was also negatively correlated with outflows, but had a positive correlation only with salinity. Fishes A total of 42 species, represented by about 67,000 individuals, were collected in the 1,238 trawl hauls made during the study. The four measures of overall fish abundance and diversity showed negative cor- relations with month series and with years, in- dicating a general decline through the study period (Table 4, Fig. 3). Numbers, biomass, and number of species had positive correlations with temperature and/or salinity and negative correlations with out- "Herbold, B. 1985. Resource partitioning within a non-co- evolved assemblage of fishes. Unpubl. Ph.D. Thesis, Univ. Califor- nia, Davis. 109 FISHERY BULLETIN: VOL. 84, NO. 1 DELTA OUTFLOW INDEX SALINITY TEMPERATURE UATER TRANSPARENCY 1979 Figure 2.— Trends in abiotic factors and Neomysis mercedis abun- dance within Suisun Marsh. Temperature is in °C. Average outflow in 10,000 cubic feet per second of the Sacramento River was calculated by the California Department of Water Resources. Salini- ty is given in parts per thousand. Neomysis mercedis abundance rankings are described in text. 20- 1979 flow, indicating that catches were highest in late sum- mer and lowest in early spring. However, when the patterns of occurrence of the 12 most abundant species were examined, three groups appeared: resi- dent species, winter seasonals, and spring/summer seasonals. The "resident species" included the native split- tail, tule perch, Sacramento sucker, prickly sculpin, and threespine stickleback as well as the introduced striped bass, carp, and yellowfin goby. Two additional species, native white sturgeon and introduced American shad, probably also belonged in this category, as they were caught at all times of the year but too infrequently to draw any firm conclusions. Splittail, striped bass, tule perch, Sacramento sucker, carp, and yellowfin goby had similar patterns of abundance (Figs. 4, 5) and were correlated (P < 0.05) with each other and with total biomass, numbers, and species (Tables 3, 4). All six species usually became more abundant in our catches as the sum- 110 MOYLE ET AL.: NONCOEVOLVED ASSEMBLAGE OF ESTUARINE FISHES Table 4.— Spearman rank correlation among species ranked by month (lower matrix) by numbers and among environmental and other variables ranked by month (upper matrix). Underlined values are significant at P > 0.05. 1 8 10 11 12 13. Striped bass 14. Splittail 15 Tule perch 16. Sacramento sucker 17. Yellowfin goby 18. Carp 19. Prickly sculpin 20. Stickleback 21. Delta smelt 22. Longfin smelt 23. Threadfin shad 24. Staghorn sculpin 25. Starry flounder 0.19 0.68 0.46 -0.14 -0.41 -0.11 -0.07 -0.17 -0.24 -0.09 0.15 - 0.50 0.44 0.38 0.51 0.58 0.53 0.13 ■0.09 0.08 •0.01 ■0.02 ■0.13 0.30 0.54 0.27 0.54 0.32 0.38 •0.13 0.05 0.21 0.09 0.38 0.05 0.03 ■0.10 0.19 0.46 0.22 0.46 0.34 0.07 0.03 0.12 0.00 0.06 0.30 0.43 0.14 -0.04 -0.04 -0.51 -0.79 -0.80 0.62 -0.29 0.19 0.44 0.33 0.30 0.37 0.78 -0.55 0.42 0.32 0.48 0.37 0.39 0.13 -0.43 0.04 -0.24 0.01 0.01 -0.11 0.51 -0.52 -0.41 -0.58 -0.34 -0.51 0.35 -0.56 -0.32 -0.14 -0.03 -0.10 0.06 0.11 0.34 0.09 -0.19 0.25 0.22 -0.17 0.09 0.12 0.04 0.15 0.05 -0.15 -0.15 0.40 0.55 0.56 0.21 -0.38 -0.15 -0.04 0.43 0.69 0.24 -0.11 -0.47 0.42 0.36 0.21 0.06 -0.02 0.20 0.18 0.17 0.30 0.01 0.18 -0.14 0.25 0.22 0.10 0.37 0.00 -0.64 0.05 0.07 -0.19 -0.06 0.16 0.01 -0.05 -0.08 0.51 0.74 0.07 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Month series Temperature Salinity Secchi depth Outflow Neomysis Crangon Palaemon Numbers/ trawl Grams/trawl Species/trawl Diversity (H) 13 14 15 16 17 18 19 20 21 22 23 24 1O0O 500 MEAN BI0MASS PER TRAUL MEAN NUMBER OF FISH PER TRAWL I 979 I 98 1 I 982 I 983 Figure 3.— Trends in mean numbers and grams of fish per trawl. mer progressed although the two introduced species, striped bass and yellowfin goby, tended to peak later than the other species. Consequently, they all showed significant (P < 0.05) negative correlations with outflow. All except Sacramento sucker and tule perch had significant positive correlations with salinity and temperature. There was a general decline in fish abundance throughout the 5-yr period. This was reflected in that four of the six species showed a positive correlation with species diversity, and all had a negative correlation with month series. Prickly sculpin seemed to peak in abundance earlier in the year than the first six species (Fig. 4) but the pattern was obscured by the considerable year-to-year variation in abundance of young-of-year fish. Adults were resident in the marsh but appeared in the trawls on an irregular basis because of their tendency to hide under logs and other objects (Moyle 1976). Overall, prickly sculpin had negative corre- lations with salinity and Secchi depth, but positive correlations with temperature, N. mercedis abun- dance, and species diversity (Table 3). Threespine stickleback abundance had a negative correlation only with temperature, presumably because their reproductive behavior obscured our ability to catch them. They were most abundant in the trawls in February through May, and the catch consisted primarily of gravid females and schools of young- of-year fish. The males were apparently defending their nesting territories in emergent vegetation. By late summer sticklebacks were rare in the trawls but could be taken in seine hauls made through weedy areas. The "winter seasonals" were three plankton- Ill 10t PRICKLY SCULPIN SACRAMENTO SUCKER I 982 I 983 FISHERY BULLETIN: VOL. 84, NO. 1 TULE PERCH 1988 Figure 4— Capture rates of native resident species within Suisun Marsh. Mean catch per effort is described as percent of the total catch for each species. feeding species, delta smelt (native), longfin smelt (native), and threadfin shad (introduced). All three species tended to be most abundant in November through January, although the pattern was not always consistent (Fig. 6). Threadfin shad were the most erratic of the three species in abundance; they were especially abundant in the summer of 1981. Longfin smelt were largely absent from our samples in 1979 and 1981. Delta smelt abundance was positively correlated (P < 0.05) with that of the other two species, although the correlation between long- fin smelt and threadfin shad was not significant. All three species had negative correlations with tem- perature, and positive correlations with Secchi depth. The "spring/summer seasonals" were staghorn sculpin and starry flounder, both euryhaline marine species that were represented mainly by young-of- year. Their patterns of abundance were not consis- tent (Fig. 6) and the peaks occurred anytime from March through September. Consequently, staghorn sculpin did not show any significant correlations with the environmental variables, although starry flounder did show negative correlation with Secchi depth. Both species had a positive correlation with species diversity, presumably because they were rare in our samples during the last 2 years when the marsh was dominated by freshwater. In addition to the 12 species that appeared regular- ly in our trawls, there were a number of other species of potential importance to the fish community that were either not sampled adequately by the trawl or were absent because of the effects of the 1976-77 drought. Five species that were not sampled ade- quately were inland silverside, chinook salmon, Sacramento squawfish, mosquitofish, and rainwater killifish. The silversides were abundant year around in the shallow, sandy or weedy areas found in some sloughs. Silversides appeared in seine hauls in 20 of the 22 mo in which seining was done; they were generally the most abundant fish in these hauls. Juvenile chinook salmon and squawfish were com- 112 MOYLE ET AL.: NONCOEVOLVED ASSEMBLAGE OF ESTUARINE FISHES mon in the marsh in February, March, and April (times of high outflows) and were taken mainly in seines. The tendency of the salmon to remain close to the banks and vegetation and to get sucked into YELLOWFIN goby CARP STRIPED BASS diversions of marsh water consequently has led to the screening of one major diversion in the marsh. Squawfish were abundant in the Sacramento River and juveniles are known to disperse widely during high flows (Smith 1982). Mosquitofish and rainwater killifish were present in ponds adjacent to the sloughs, along with silversides and sticklebacks; mos- quitofish were planted in some areas for mosquito control. Principal Components' Analysis The PCA using the numbers per trawl matrix resulted in four components that explained 47% of the variance in the matrix (Table 5). The first com- ponent loaded most heavily on tule perch, Sacra- mento sucker, and splittail, native resident species most abundant in dead-end sloughs, and to a lesser extent on carp and threadfin shad, introduced species common in such sloughs. The second com- ponent loaded heavily on striped bass, yellowfin goby, and carp, three introduced species resident through- out the marsh but most frequently captured in the main sloughs; all reached peaks of abundance in late summer. The third component loaded most heavily on prickly and staghorn sculpins, two benthic species that peaked in abundance during the summer months but were relatively scarce during the last 2 Table 5.— Loadings (rotated) of major fish species on four com- ponents produced by a principal components analysis of numbers of fish per trawl (n = 1,238). Values over 0.500 are underlined. Figure 5.— Capture rates of introduced species within Suisun Marsh. Mean catch per effort is described as percent of the total catch for each species. Compo- Compo- Compo- Compo- nent nent nent nent 1 2 3 4 Splittail adults 0.487 0.300 -0.024 -0.149 Splittail juveniles 0.549 0.100 0.318 0.149 Striped bass adults 0.058 0.701 0.078 - 0.073 Striped bass juveniles 0.183 0.631 -0.157 0.046 Longfin smelt -0.124 0.029 -0.032 0.747 Delta smelt 0.022 -0.061 -0.027 0.734 Threadfin shad 0.447 -0.286 -0.121 0.319 Common carp 0.403 0.403 0.166 -0.084 Yellowfin goby -0.011 0.660 0.023 0.016 Tule perch adults 0.827 0.049 0.085 -0.102 Tule perch juveniles 0.833 -0.036 - 0.045 -0.017 Sculpin adults 0.254 -0.038 0.377 -0.263 Sculpin juveniles 0.090 0.043 0.780 - 0.077 Starry flounder 0.117 0.256 0.107 -0.030 Staghorn sculpin 0.043 0.047 0.727 0.118 Sacramento sucker 0.637 0.197 0.341 0.102 Threespine stickleback 0.039 -0.296 0.486 -0.147 Eigenvalue 2.826 1.874 1.829 1.391 Cumulative proportion of variance explained 0.200 0.304 0.396 0.472 113 THREADFIN SHAD DELTA SMELT 1979 1980 1982 Figure 6— Capture rates of seasonal species within Suisun Marsh. Mean catch per effort for each month described as percent of total for each species. FISHERY BULLETIN: VOL. 84, NO. 1 LONGFIN SMELT STAGHORN SCULPIN 1979 years of the study. A similar pattern was shown by threespine stickleback, which also had a relatively large positive loading on this component. The fourth component loaded heavily on delta and longfin smelt, and to a lesser extent on threadfin shad. This is the winter seasonal group identified in the previous analysis. DISCUSSION During the 5-yr study period, the fish assemblage of Suisun Marsh had the following character- istics: 1. There was a strong seasonal pattern of total fish 114 MOYLE ET AL.: NONCOEVOLVED ASSEMBLAGE OF ESTUARINE FISHES abundance with numbers and biomass lowest in winter and spring and highest in late summer. Fishes were least abundant when river outflows were highest and most abundant when salinities and tem- peratures were highest. 2. There was an overall decline in fish abundance and species diversity through the study period. 3. Of the 21 species that occurred in the marsh on a regular basis, 14 were residents, 4 were winter seasonals, and 3 were spring/summer seasonals. Another 21 species occurred sporadically, in small numbers. These were mainly marine and freshwater species that presumably could become established in the marsh if environmental conditions changed significantly. 4. The abundant resident species fell into two groups, one made up of native species that concen- trated in the small dead-end sloughs and the other a mixture of introduced and native species that were widely distributed in the marsh, but most abundant in the larger sloughs. 5. The structure of the fish assemblage (i.e., the pattern of distribution and abundance) was fairly consistent over the 54-mo period. The seasonal pattern of fish abundance was due to a number of factors, most importantly 1) varia- tion in sampling efficiency, 2) influxes of young-of- year fish, 3) favorable environmental conditions for most fish species in late summer, and 4) abundance of Neomysis mercedis. When outflows were high, water levels in the marsh were high and showed lit- tle tidal fluctuation. Therefore trawling was less ef- ficient because there was more water and more flooded vegetation available as cover for fish. How- ever, even under these conditions most of the sam- pling areas were rarely more than 2 m deep, so our trawl covered at least half the water column, and large catches were common, especially early in the study. Therefore, variation in sampling efficiency may have exaggerated the peaks and valleys of the catch curves (Figs. 4, 5) but was unlikely to obscure the general trends in abundance Probably the most important contributor to the seasonal patterns was the increase in young-of-year striped bass, splittail, prickly sculpin, and tule perch, in June through August. These species (and others, to a lesser extent) became vulnerable to our trawl at 30-40 mm SL, and catches of several hundred individuals in a 5-min tow were made on occasion. The rapid growth of these species during summer (Daniels and Moyle 1983; Herbold and Moyle, unpubl. data) indicated that en- vironmental conditions, including warm tempera- tures and moderate salinities, were favorable for them and for other euryhaline species (ag., staghorn sculpin, starry flounder). These same conditions also favored N. mercedis, a small shrimp that is an im- portant food item in summer diets of most of the fishes (Herbold fn. 4). It is possible that the summer peak in fish abundance may be due also in part to fishes moving in to take advantage of an abundant food resource The decline in N. mercedis abundance in late summer may be related in part to fish preda- tion, although it is presumably related mainly to their seasonal movements within the entire estuary (Orsi and Knutson 1979). The overall decline in fish abundance over the study period seemed to be due to two factors: varia- tion in reproductive success of major species and the fact that 1982 and 1983 were years of unusually high precipitation and runoff, so freshwater conditions prevailed throughout the summer months of both years. Splittail showed an unusually strong year class in 1978, which dominated the 1979, and, to a lesser extent, 1980 samples (Daniels and Moyle 1983). Catches of splittail in 1979 were typically 2-5 times greater than in subsequent years. Striped bass, tule perch, and carp also showed peaks of abundance in 1979 and had low abundances in 1982-83, with one or two peaks of abundance in between. Except for carp, the peaks were largely due to influxes of young- of-year fish. The reason for the abundance of the 1978 year class of fish was presumably related to 1978 being a year of high, but not excessive, outflows. Increased reproductive success during high outflow years has been documented for striped bass (Stevens 1977), splittail (Daniels and Moyle 1983), American shad, chinook salmon, and longfin smelt (Stevens and Miller 1983). However, under extreme outflow con- ditions (such as existed in 1982 and 1983), young- of-year fish are apparently carried downstream to areas below the marsh (San Francisco and San Pablo Bay) where chances of survival may be less (Stevens 1977). Drought also contributed to the variation in the fish fauna. During 1976 and 1977, severe drought reduced freshwater inflows to the marsh, resulting in sustained high salinities. Freshwater fishes de- clined dramatically during the drought period (Herr- gesell et al. 1981) and the fishery for catfish (main- ly white catfish and black bullhead) was greatly reduced (Baracco 1980). The catfish populations did not recover during the study period, but the regular appearance of young-of-year white catfish in our trawls in late 1983 indicated a recovery may be in progress. Other freshwater fishes found in the marsh (Table 1) showed no signs of increasing. Most were represented in our samples by <10 individuals that 115 FISHERY BULLETIN: VOL. 84, NO. 1 had presumably been washed into the marsh from freshwater habitats upstream. However, black crap- pie and perhaps other centrarchids contributed to the local fishery prior to the drought, mainly in the upper ends of the larger sloughs, so a recovery can be expected. Despite the decline in freshwater fishes during the drought, there was no corresponding major increase in the abundance of euryhaline marine species characteristic of nearby San Francisco Bay (Herr- gesell et al. 1981). Marine species (such as northern anchovy, Pacific herring, and shiner perch) general- ly appeared in our samples in late summer when salinities were highest, in parts of the marsh closest to Suisun Bay. Considering the annual and long-term variations in fish abundances and the fact that the fish assem- blage is made up of a mixture of native and intro- duced species, the consistency of the assemblage structure during the study is surprising. Coevolution has obviously little role in an assemblage in which the most abundant species (striped bass) entered in 1879 and other abundant species entered in the 1960's (yellowfin goby) and 1970's (inland silversides) (Moyle 1976). The apparent consistency in structure seemed to be the result of 1) two introduced species, striped bass and carp, that were consistently abun- dant in the marsh, 2) the group of native resident fishes that was persistent in deadend sloughs, and 3) the native fishes that moved in and out of the marsh on a seasonal basis. This does not mean that the structure observed during this study will persist indefinitely. A number of changes in the fish fauna may already be occur- ring. For example, the presence of young-of-year white catfish in 1983 and 1984 may signify a shift of the assemblage towards catfishes and centrar- chids, such as existed before the 1976-77 drought. Striped bass are presently in a long-term decline in abundance, a trend which seems to be continuing (Kelley et al. 1982). Past history indicates that new introductions of fishes into the system are likely: specifically, the white bass, Morone chrysops, has recently become established in part of the San Joa- quin drainage and may become a major new predator in the Sacramento-San Joaquin Estuary if planned eradication attempts fail (California Department of Fish and Game unpubl. data). Furthermore, addi- tional diversions of freshwater from the estuary are planned (Herrgesell et al. 1981), and major modifica- tions to the marsh channels are planned or under- way (Baracco 1980), so the environment, especially in the dead-end sloughs, may change significantly. It is difficult to predict what the combined effects of all these changes will be on the present fish assemblage, but extinctions of both native and intro- duced species in the estuary have occurred in the past (Moyle 1976) and could occur again in the future The structure of the fish assemblage of Suisun Marsh is similar in may respects to the structure of the fish assemblages of other large estuaries (e.g., Markle 1976; Meeter et al. 1979), despite the impor- tance of recently introduced species and the stabi- lizing influence humanity has had on the pattern and amount of freshwater inflow (Kahrl 1978). In most such estuaries, as in the Sacramento-San Joaquin, the assemblages are dominated by juvenile fishes, and most species have substantial populations out- side the estuary. As in Suisun Marsh, the fish assem- blages of such estuaries are made up of a relatively small number of the species available in nearby marine and freshwater environments. Presumably, the species composition of an estuarine assemblage is determined in large part by the ability of the species to tolerate the particular set of environmen- tal conditions that exist there Since these conditions may change with short-term climatological changes, the fish assemblages may change as well (Meeter et al. 1979; Marais 1982). Thus coevolution is given lit- tle chance to operate in estuarine systems in general. In this context, it is not surprising that the fish assemblage of the Suisun Marsh behaves ecologically in a way similar to fish assemblages in most other estuarine systems. Because resource partitioning is commonly observed among estuarine fishes (Sheri- dan and Livingston 1979; Whitfield 1980), competi- tion may be an important process in determining the structure of estuarine fish assemblages (Thorman 1982), a hypothesis we are currently investigating in the Suisun Marsh. ACKNOWLEDGMENTS This project was supported by the California Department of Water Resources (DWR) and by the Agricultural Experiment Station, University of California (Project No. 3930-H). It would not have been possible without the support and encourage- ment of Randall L. Brown, Central District, DWR. Numerous volunteers assisted the sampling effort, but especially Larry Brown, Sonia Cook, Bart Daniel, Lynn Decker, Tim Ford, Bret Harvey, Ned Knight, Tim Takagi, Bruce Vondracek, Eric Wikra- manayake, and Wayne Wurtsbaugh. The manuscript was reviewed in various drafts by Larry Brown, Beth Goldowitz, Ned Knight, and Eric Wikramanayake. The manuscript was "processed" by Donna Raymond. 116 MOYLE ET AL.: NONCOEVOLVED ASSEMBLAGE OF ESTUARINE FISHES LITERATURE CITED Allen, L. G. 1982. Seasonal abundance, composition, and productivity of the littoral fish assemblage in Upper Newport Bay, Califor- nia. Fish. Bull., U.S. 80:769-790. Baltz, D. M., and P. B. Moyle. 1982. Life history characteristics of tule perch (Hysterocar- ■pus traski) populations in contrasting environments. En- viron. Biol. Fishes 7:229-242. Baracco, A. 1980. Aquatic resources of Suisun Marsh with an analysis of the fishery effects of a proposed water quality maintenance plan. Calif. Dep. Fish Game, Anad. Fish. Branch Admin. Rep., 26 p. Blaber, S. J. M., and T. G. Blaber. 1980. Factors affecting the distribution of juvenile estuarine and inshore fish. J. Fish. Biol. 17:143-162. Conomos, T. J. (editor). 1979. San Francisco Bay: the urbanized estuary. Pac, Div. AAAS, San Franc., 493 p. Dahlberg, M. D., and E. P. Odum. 1970. Annual cycles of species occurrence, abundance, and diversity in Georgia estuarine fish populations. Am. Midi. Nat. 83:382-392. Daniels, R. A., and P. B. Moyle. 1983. Life history of the splittail, (Cyprinidae: Pogonichthys macrolepidotus), in the Sacramento-San Joaquin estuary. Fish. Bull., U.S. 81:647-654. Dixon, W. J., and M. B. Brown (editors). 1977. Biomedical computer programs, Pseries. Univ. Calif. Press, Berkeley, 880 p. Ganssle, D. 1966. Fishes and decapods of San Pablo and Suisun Bay. In D. W. Kelley (editor), Ecological studies of the Sacramento- San Joaquin estuary, Part I. Zooplankton, zoobenthos, and fishes of San Pablo and Suisun Bays, Zooplankton and zoo- benthos of the Delta, p. 64-94. Calif. Dep. Fish. Game, Fish Bull. 133. Herrgesell, P. L., D W. Kohlhorst, L. W. Miller, and D. S. Stevens. 1981. Effects of freshwater flow on fishery resources in the Sacramento-San Joaquin estuary. In R. D Cross and D. L. Williams (editors), Proceedings of National Symposium on Freshwater Inflow to Estuaries, p. 71-87. U.S. Fish and Wildlife Service OBS 81-4. Kahrl, W. J. (editor). 1978. The California water atlas. Calif. Dep. Water Resour., 118 p. Kelley, D. W, and The Striped Bass Working Group. 1982. The striped bass decline in the San Francisco Bay- Delta Estuary. Calif. State Water Res. Control Board, 58 p. Kjelson, M. A., and D. R. Colby 1977. The evaluation and use of gear efficiencies in the estima- tion of estuarine fish abundance In V. F Kennedy (editor), Estuarine processes, Vol. 2, p. 416-424. Academic Press, N.Y. Livingston, R. J. 1976. Diurnal and seasonal fluctuations of organisms in a North Florida estuary. Estuarine Coastal Mar. Sci. 4:373- 400. Marais, J. F K. 1982. The effects of river flooding on the fish populations of two eastern Cape estuaries. S. Afr. J. Zool. 17:96-104. Markle, D. F. 1976. The seasonality of availability and movements of fishes in the channel of the York River, Virginia. Chesapeake Sci. 17:50-55. Meeter, D. A., R. J. Livingston, and G. C. Woodsum. 1979. Long-term climatological cycles and population changes in a river-dominated estuarine system. In R. J. Livingston (editor), Ecological processes in coastal and marine systems, p. 315-338. Plenum Press, N.Y. Moyle, P. B. 1976. Inland fishes of California. Univ. Calif. Press, Berkeley, 405 p. Orsi, J. J., and A. C. Knutson, Jr. 1979. The role of mysid shrimp in the Sacramento-San Joa- quin estuary and factors affecting their abundance and distribution. In T. J. Conomos (editor), San Francisco Bay: the urbanized estuary, p. 401-408. Pac. Div. AAAS, San Franc Quinn, N. J. 1980. Analysis of temporal changes in fish assemblages in Serpentine Creek, Queensland. Environ. Biol. Fishes 5:117- 133. Sheridan, P. F, and R. J. Livingston. 1979. Cyclic trophic relationships of fishes in an unpolluted, river-dominated estuary in North Florida. In R. J. Living- ston (editor), Ecological processes in coastal and marine systems, p. 143-160. Plenum Press, N.Y. Siegfried, C. A. 1980. Seasonal abundance and distribution of Crangonfran- ciscarum and Palaemon macroadactylus (Decapoda, Caridea) in the San Francisco Bay-Delta. Biol. Bull. (Woods Hole) 159:177-192. Skinner, J. E. 1972. Ecological studies of the Sacramento-San Joaquin Estuary. Calif. Dep. Fish, Game Delta Fish Wildl. Prot. Study Rep. 8, 94 p. Smith, J. J. 1982. Fishes of the Pajaro River System. Univ. Calif. Publ. Zool. 115:83-170. Stevens, D. E. 1977. Striped bass (Morone saxatilis) year class strength in relation to river flow in the Sacramento-San Joaquin Estuary, California. Trans. Am. Fish. Soc. 106:34-42. Stevens, D. E., and L. W. Miller. 1983. Effects of river flow on abundance of young chinook salmon, American shad, longfin smelt, and delta smelt in the Sacramento-San Joaquin River system. N. Am. J. Fish. Manage 3:425-437. Thorman, S. 1982. Niche dynamics and resource partioning in a fish guild inhabiting a shallow estuary on the Swedish west coast. Oikos 39:32-39. Turner, J. L., and D. W Kelley (editors). 1966. Ecological studies of the Sacramento-San Joaquin Delta. Part II. Fishes of the Delta. Calif. Dep. Fish Game, Fish Bull. 136, 168 p. Whitfield, A. K. 1980. Distribution of fishes in the Mhlanga estuary in rela- tion to food resources. S. Afr. J. Zool. 15:159-165. 117 THE ROLE OF ESTUARINE AND OFFSHORE NURSERY AREAS FOR YOUNG ENGLISH SOLE, PAROPHRYS VETULUS GIRARD, OF OREGON E. E. Krygier1 and W. G. Pearcy2 ABSTRACT Our trawling studies confirm that age group 0 English sole are common in shallow waters along the open coast as well as in estuaries of Oregon. Both areas appear to be important nursery areas for this species. Metamorphosing English sole were recruited to Yaquina Bay over many months between November and June during the 5 years studied. Seasonal trends in abundance of these transforming fish were rather similar to both Yaquina Bay and open coastal stations. Transforming individuals, however, were found earlier in the fall and later in the spring and summer along the open coast than in Yaquina Bay. Based on catch curves, the densities (no. m"2) of juvenile English sole were much higher in Yaquina Bay than along the open coast. Transforming sole (20-25 mm) were an exception. They were sometimes most abundant at the open coast location. Increasing densities of 20-40 mm length fish in the Yaquina Bay catches were accompanied by decreased catches of this size group at the open coast site This sug- gests immigration of a broad size range of both transforming and fully transformed individuals into Yaquina Bay. English sole, Parophrys vetulus Girard 1854, is a ma- jor component of the catches in the northeastern Pacific trawl fishery, usually ranking second only to Dover sole, Microstomas pacificus, in annual land- ings off Oregon (Barss 19763; Demory et al. 19764). It ranges from Baja California to Unimak Island in western Alaska, with commercial quantities at depths of 128 m or less (Hart 1973). Tagging studies have revealed a series of relatively discrete stocks of English sole off California, Oregon, Washington, and British Columbia (Ketchen 1956; Forrester 1969; Jow 1969; Pattie 1969; Barss 1976 fn. 3). Spawning of English sole is protracted, usually ex- tending from September through April, and is often variable in seasonal intensity within and among spawning seasons (Budd 1940; Ketchen 1956; Harry 1959; Jow 1969; Laroche and Richardson 1979). Much of this variability among years may be related to upwelling and bottom temperatures (Kruse and Tyler 1983). Spawning concentrations of adult English sole were found in the fall off the central Oregon coast at depths of 70-110 m (Hewitt 1980). 'College of Oceanography, Oregon State University, Corvallis, OR; present address: Alaska Trailers Association, 130 Seward Street, Juneau, AK 99801. 2College of Oceanography, Oregon State University, Corvallis, OR 97731. 3Barss, W. H. 1976. The English sole Oreg. Dep. Fish Wildl., Inf. Rep. 76-1, 7 p. 4Demory, R. L., M. J. Hosie, N. Ten Eyck, and B. O. Forsberg. 1976. Marine resource surveys on the continental shelf off Oregon, 1971-74. Oreg. Dep. Fish Wildl., 49 p. English sole are fecund, producing 327,600- 2,100,000 eggs, depending on the size of female (Ket- chen 1947; Harry 1959). Eggs are pelagic and hatch in about 4V2 d at 10° C (Alderdice and Forrester 1968). Larvae are often abundant during late winter and early spring in coastal waters of Oregon (Rich- ardson and Pearcy 1977; Mundy 1984). Larval abun- dance may fluctuate greatly among years, possibly due to annual differences in ocean conditions (Laroche and Richardson 1979; Mundy 1984). The pelagic phase lasts 8-10 wk (Ketchen 1956; Laroche et al. 1982), and most individuals complete metamor- phosis and acquire the morphology of benthic pleu- ronectids at 20 mm SL and 120 d of age (Ahlstrom and Moser 1975; Rosenberg and Laroche 1982). While early larval stages are rarely found in estu- aries (Misitano 1970; Pearcy and Myers 1974), trans- forming larvae and early juvenile stages of English sole are common in estuaries (Westrheim 1955; Smith and Nitsos 1969; Olsen and Pratt 1973; Pearcy and Myers 1974; Misitano 1976; Toole 1980; Bayer 1981) and shallow protected bays (Ketchen 1956; Kendall 1966; Van Cleve and El-Sayed 1969). Young English sole are known to utilize 13 estuaries along the Oregon coast and were absent in only 3 small estuaries surveyed along the southern Oregon coast.5 Villadolid (1927, as cited by Misitano 1970) captured 6Report of estuary surveys, July-August 1972. Fish Comm. Oreg. Intern. Rep. GS-73-1, 14 p. Manuscript accepted March 1985. _119_ FISHERY BULLETIN: VOL. 84, NO. 1 0-age English sole in San Francisco Bay but not off the coast. Based on the incidence of a parasitic infection, ap- parently acquired only in estuaries, and the absence of 0-age English sole in Demory's (1971) surveys off the northern Oregon-southern Washington coast, Olsen and Pratt (1973) concluded that estuaries are likely the exclusive nursery for English sole on the Oregon coast. Laroche and Holton (1979), however, captured 0-age English sole in shallow waters along the open Oregon coast, indicating that estuaries may not be the only nursery area for English sole off Oregon. The main objective of our study is to evaluate the relative importance of estuarine and open coastal nursery grounds for young English sole off Oregon. METHODS AND MATERIALS Bottom trawl collections provided most of the in- formation on the distribution and abundance of juvenile English sola Collections were made in Ya- quina Bay and along the open coast outside the bay. These were supplemented with extensive trawl col- lections farther to the north and south along the open coast and collections in other estuaries. Fish were collected using a 1.52 m wide, 56 cm high beam trawl (see Krygier and Horton 1975) from the RV Paiute and from a 7.3 m dory. Additional col- lections with a 2.72 m beam trawl (Carey and Heya- moto 1972) were made on the RV Cayuse. To retain small, settling fish, fine-mesh (1.5-3.5 mm stretch) liners were used in the trawls. The 1.52 and 2.72 m beam trawls were fitted with a 1.0 or 2.0 m circum- ference wheel, respectively, and a revolution counter to estimate the area sampled (Carey and Heyamoto 1972; Krygier and Horton 1975). Tows were made at 0.7-1.0 m s_1. Tow duration was normally 5-10 min on the bottom in estuaries and 10-20 min along the coast, usually at a 4:1 scope Most tows were dur- ing daylight hours. Collections for juvenile English sole were made in five different study areas (Fig. 1, Table 1): ESTUARINE 1) Yaquina Bay: 1.52 beam trawl collections were made in lower Yaquina Bay from January 1970 through February 1972 by Krygier and Johnson (un- publ. data) and Krygier and Horton (1975) and sup- plemented by collections in 1977-79. Additionally, we used collections made by Myers (1980) with a 100 m beach seine (11.0 mm stretch mesh in the inner wing and bunt (Sims and Johnsen 1974)). 2) Other estuaries: The 1.52 m beam trawl was towed from a 7.3 m dory in four estuaries north and south of Yaquina Bay (Tillamook and Siletz Bays, 107.5 and 35.2 km to the north of Yaquina Bay and Alsea Bay and Umpqua River estuary, 21.3 and 105.6 km to the south). Each estuary was divided into seven equal-area portions from which we planned to take three random trawl collections (2 of the 21 trawls in the Umpqua River estuary were not com- pleted). COASTAL 3) Moolack Beach: 1.52 m beam trawl collections were made on a monthly or bimonthly basis in shallow (3-31 m depth) nearshore waters in a 1.0 km2 area just north of Yaquina Head during 1977, 1978, and 1979. Moolack Beach is semiprotected by headlands to the north and south and offshore by a reef that rises from 15 m to 6 m. 4) Grid stations: Collections were taken with a 2.72 m beam trawl, approximately monthly, during 1978 at 1.9, 5.6, and 9.3 km (1, 3, and 5 nmi) offshore along lat. 44°41.6'N, 44°36.6'N, and 44°31.6'N. Thir- teen collections were also made in this area with the 1.52 m beam trawl. Table 1.— Summary of collections used in this study. Net No. Dates (sampling Area type trawls frequency) Yaquina Bay M.52 m 178 16 Jan. 70-25 Jan. 71 (weekly or biweekly); 17 Feb. 71-25 Feb. 72 (bimonthly) 21.52 m 26 26 Apr.-28 June 77 (bi- monthly) 21.52 m 96 1 Dec. 77-14 Sept. 79 (monthly to bimonthly) 22.72 m 8 16 Nov. 77, 1 Feb. 78, 27 Nov. 78 beach 196 12 July 77-11 Nov. 78 seine (various: daily, biweekly, weekly, bimonthly) Moolack 1.52 m 16 28 Apr. 77-23 June 77 (bimonthly) 1.52 m 76 11 Jan. 78-24 Sept. 79 (bimonthly of monthly) Grid 1.52 m 13 21 Apr. 77-27 June 77; 15 June 78-28 Sept. 78 2.72 m 106 17 Nov. 77-25 Oct. 78 (monthly) North-South 1.52 m 40 2 June 77-13 June 77, 15 June 78-21 July 78 2.72 m 83 15 May 78, 27 June 78, 25 Oct. 78 Estuaries 1.52 m 82 8-12 May 78, 21 trawls each in Tillamook, Siletz and Alsea; 19 trawls in Umpqua 1Net liners 3.5 mm and cod end liner of 1.5 mm stretch mesh, 1970-72. 2Net liners 3.2 mm stretch mesh, 1977-79. 120 KRYGIER and PEARCY: NURSERY AREAS FOR YOUNG ENGLISH SOLE 44° 40' 44° 30' ^ COLUMBIA -RIVER -46c TILLAMOOK •^BAY SILETZ BAY -45c Xyaquina bay TTalsea bay .kUMPQUA • VF^ RIVER 'COOS BAY 44< Figure 1.— Location of sampling stations in the North-South coastal survey (right) and at Moolack Beach, the grid stations I, II, III, and within Yaquina Bay (left). In Yaquina Bay the numbers 1-4 indicate locations of stations for sampling in 1970-72, the solid dots loca- tions in 1977-79, and the arrows indicate seine stations in 1977. 5) North-south coastal survey: 1.52 m beam trawl collections were made from 111 km to the north (lat. 45°37.5'N) and 111 km to the south (lat. 45°36'N) of Yaquina Bay at 9.3 km intervals (Fig. 1) at depths of 9-18 m in June 1977 and May-October 1978. Most samples were preserved in 5% Formalin6 and seawater. In the laboratory fish were identified, sorted, and standard length (SL) measured to the 6Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. nearest millimeter. Nearly all English sole captured in Yaquina Bay were 150 mm SL or less and included 0- and I-age fish (Rosenberg 1982). We call these fishes "juveniles" in this paper. RESULTS Variability of Catches The variability of the number of juvenile English sole caught per m2 in repeated trawls within the same area was low. Coefficients of dispersion (s2/x) 121 FISHERY BULLETIN: VOL. 84, NO. 1 were usually <0.1, indicating uniform distributions within the small areas (10-100 m2) and short inter- vals of time (1-2 h) sampled. Variability was higher and coefficients of dispersion sometimes differed sig- nificantly (chi-square, <0.05) from a random (Pois- son) distribution among different sampling depths at the same date (s2fx = 0.36-1.65) and among dif- ferent sampling dates within a single depth at Moo- lack Beach (s2fx = 1.2-2.31). Coefficients of disper- sion did not significantly differ from randomness either among the grid stations for the same sam- pling dates (s2/x = 0.87-1.82) or among different sampling dates at the same station (0.94-1.97). In general, at the scale of sampling we used, juvenile English sole had even, nonpatchy distributions. Gear Comparisons Tb compare the relative efficiencies of the 1.52 m beam trawl from the Paiute and the 2.72 m beam trawl from the Cayuse, 14 pairs of trials were made at the same time, while the vessels trawled on par- allel courses within 30 m of each other. No signifi- cant differences (P > 0.05; Mann-Whitney "U" tests, Tate and Clelland 1957) were found in the catch/m2 of juvenile English sole <150 mm for any paired trawl comparison. No significant differences were found in length- frequency distributions of P. vetulus captured in 10 of the 14 comparisons [Kolmogorov-Smirnov (K-S) test, late and Clelland 1957]. In the four pairs of tows that were significantly different (October 1978) the 2.72 m trawl caught more small (~20 mm SL) English sole per m2 than the 1.52 m trawl, while both trawls caught similar proportions in the 46-100 mm size range Comparisons were made between the sizes of English sole in beach seine samples and midchan- nel trawl samples in Yaquina Bay on six different dates. Differences were significant (K-S test; P < 0.05) for all comparisons because the beam trawl caught a much broader size range of fish, including individuals >40 mm which were rare or absent in the beach seine catches. Trends in Catches and Sizes of Fish Significant (H-test, P < 0.05) differences in catches/m2 at different depths at Moolack Beach and the grid stations show that in general the abun- dance of juvenile English sole in offshore waters was greatest in shallow water and decreased with in- creasing depth. Average catches/103m2 (+1 stan- dard deviation) of English sole <150 mm were 16 (±20), 61 (±14), 43 (±75), and 10 (±12) at the 9, 9-17, 12-18, and 18-31 m stations off Moolack Beach, com- pared with only 3 (±3) and 2 (±3) at the 40 and 64 m 1-3 and 1-5 grid stations at about the same latitude Newly transformed, benthic English sole (<24 mm) were found at all depths sampled in the Moolack Beach area, but the highest proportion of these recently metamorphosed fish was found at depths <18 m. Within the depth zones sampled the propor- tion of small English sole <30 mm decreased with depth and fish >150 mm were only captured at depths deeper than 18 m (Fig. 2). Juvenile English sole <150 mm were found along the entire 222 km coast sampled (Fig. 3). They were usually moderately abundant (^0.01 m2) between Siletz Bay and Alsea Bay, and near the Umpqua River and Tillamook Bay. Average catches, however, were higher off Moolack Beach than any other area, averaging 0.21 juvenile English sole/m2, an order of magnitude greater than most other offshore areas or the grid stations. Moolack Beach was apparently a region of the open coast with exceptionally high densities of English sole Juvenile English sole were generally most abun- dant at the shallowest depths in these collections, corroborating more intense sampling off Moolack Beach and at the grid stations (Fig. 3). Average catches at depths of 18 m and 36 m decreased about an order of magnitude between May (0.026/m2; SD 0.049) and October (0.003/m2; SD 0.003). Variations in Abundance of Settling Fish In our samples, metamorphosis or transformation, as indicated by migration of the left eye and by body pigmentation, occurred between 14-26 mm. Most fish had completed metamorphosis by 23 mm. In Yaquina Bay, the metamorphosing individuals first appeared in November of 1971 and 1978 (the 1972 and 1979 year classes) and in January of 1971 and 1978 (1971 and 1978 year classes) (Fig. 4). (In this paper we designate year classes by the year that most juveniles settled to the bottom; eg., products of spawning dur- ing the fall 1978-winter 1979 are called the 1979 year class.) Metamorphosing fish were present in Yaquina Bay until June (1970, 1978, 1979) or July (1971), but none was found after July during the four summer periods sampled. Maximum densities of these metamorphosing fish were observed between March and May in 1970, 1971, and 1978, but between November and January in 1978-79. Densities were variable Low densities 122 KRYGIER and PEARCY: NURSERY AREAS FOR YOUNG ENGLISH SOLE _ MOOLACK BEACH STATIONS <4m ■ . ■ ■■ r- _■ ,11 ■ m 30r GRID STATIONS § 20- IOh 0 30- 20- 10 0 30r 20 10 0 18m _a ■»■■! m ■ r « ■ ■■ ■■ u_ 40 m ■■■■ 64 m 20 40 60 80 _«» — «-_e r*^ 100 120 LENGTH (mm) ^--fh ,4 T-SJ- ^" ^M-ff 140 160 >I70 Figure 2— Length-frequency distributions of juvenile English sole caught at different depths at the Moolack Beach (above) and grid stations (below). 123 FISHERY BULLETIN: VOL. 84, NO. 1 :S/LETZ BAY °&fik. TILLAMOOK eiS.BAY YAQUINA BAY ALSEA BAY -45< O/m^ O.OOI-0.003 0.004 - 0.009 > 0.010 ' UMPOUA R. R 44< Figure 3— Catches of juvenile English sole (<150 mm) along the open coast during May, June, July, and October 1978. Hatched areas indicate untrawlable grounds due to crab pots or rocky outcrops. 124 KRYGIER and PEARCY: NURSERY AREAS FOR YOUNG ENGLISH SOLE 0 ' N ■ D Figure 4— Abundances of settling (<20 mm SL) English sole in Yaquina Bay for 1970-79 (solid line) and Moolack Beach for 1970-79 (dashed line). occurred during March 1970, January and February 1971, 1972, and April-May 1979, suggesting sea- sonal variation in spawning activity of adults (see Kruse and Tyler 1983), mortality of planktonic stages, or movement of young into or out of the estuary. Seasonal trends in catches of transforming English sole in Yaquina Bay and at Moolack Beach for 1978 and 1979 shows that fish <20 mm were found 1-2 mo earlier at Moolack Beach than in Ya- quina Bay during both years (Fig. 4). Moreover, tinued at Moolack Beach from 18 to 50 d after settling fish were no longer found within the estuary. Tb our surprise, similar densities of settling fish were caught in both areas. Seasonal trends were some- times similar, suggesting a common source of lar- vae and similar processes affecting variations in recruitment of metamorphosing fish at both the open-coast and estuarine areas. The catches/m2 of age groups 0 and I English sole (20-150 mm) are plotted as catch curves for each 5 mm size group (Fig. 5) where no. m2 = 2 of the number of individuals in each 5 mm size group total area sampled in m2 during sampling periods in which year class occurred recruitment of the 1978 and 1979 year classes con- Trends in the abundance of English sole were often 125 FISHERY BULLETIN: VOL. 84, NO. 1 1.0 00 £ b bfe*S£^^ o.i- 0.01 0.001 25 "Qi. Y^ \ .A 1971 \/970 V. /£ P' ---o-_ IBEACH i SEINE i^z: ••A 1 L B 1977 25 50 125 150 75 100 LENGTH (mm) Figure 5— Abundances of young English sole year classes as a function of length. (A) 1969-72, (B) 1977, similar for the four year classes sampled between 1969 and 1972 in Yaquina Bay (Fig. 5 A). Abundances of recently recruited individuals 20-45 mm in length were similar among the 1970, 1971, and 1972 year classes. The 1969, 1970, and 1971 year classes also increased in numbers/m2 between 75 and 90 mm before declining to low catches at larger sizes. Abun- dances of small fish of the 1969 year class are low because this year class was only sampled in 1970, when most fish were >75 mm. Catches/m2 of the 1977 and 1978 year classes in Yaquina Bay were generally larger than the 1969, 1970, 1971, 1972, and 1979 year classes (Fig. 5A, B, C). The 1977 cohort differed from other year 126 KRYGIER and PEARCY: NURERSY AREAS FOR YOUNG ENGLISH SOLE E d O.OI 0001 25 50 YAQUINA BAY I MOOLACK b— o <^^>^> BEACH _L 75 100 LENGTH (mm) 125 150 CO £ o 0.01 0.001 YAQUINA BAY l 25 50 75 100 LENGTH (mm) 1979 125 150 (C) 1978, (D) 1979 year classes. Note that some curves are based on incomplete sampling of all seasons. classes by having a large peak of abundance for 30-70 mm individuals, and the 1978 year class had much higher abundance of large (100-140 mm) individuals than other year classes. Obviously the trends shown by these catch curves cannot be explained by mortality alone Immigration of young benthic English sole into our sampling area of Yaquina Bay is suggested by the increased catches of 75-100 mm individuals of the 1970 and 1971 year classes and increased catches of 20 to 40-45 mm in- dividuals of the 1978 and 1979 year classes. Beam trawls catches at Moolack Beach for the 1977, 1978, and 1979 year classes and beach seine catches in Yaquina Bay for part of the 1977 year class 127 FISHERY BULLETIN: VOL. 84, NO. 1 and the 1978 year class indicate that the abundance of newly recruited, settling fish (<24 mm) of the 1977 and 1978 year classes was higher at Moolack Beach than in Yaquina Bay (Fig. 5B, C). These high catches at Moolack Beach were followed by a steep decline in catches to the 41-44 mm size class. English sole larger than 30 mm were consistently less abundant at Moolack Beach than in Yaquina Bay. Densities in- creased in Yaquina Bay concurrent with the steep decline of 20-44 mm individuals at Moolack Beach. These trends suggest immigration of young fish from the shallow waters of the open coast to Yaquina Bay over a range of sizes, from 20 to 40 mm. Two peaks occurred in the beach seine catches of the 1978 year class: at 20-25 and 40-45 mm. The first peak coincides with the sizes that decreased marked- ly in abundance at Moolack Beach. The second peak coincides with low abundance of 40-45 mm fish at Moolack, and with a decrease in catches of these sizes of fish at the trawl stations in Yaquina Bay. These trends of trawl-caught fish suggest that im- migration from Moolack Beach first occurred to the shallow waters of the bay and then to the deeper trawl stations. The peak in the catches of 40-45 mm fish at seine stations may be caused by immigration into these shallower waters of metamorphosed in- dividuals from either the offshore areas or deep areas of Yaquina Bay. Abundances and Sizes in Five Estuaries Age-0 English sole were present in all five estu- aries sampled with trawls during May and June 1978. The mean abundance of young English sole, which ranged from 0.7/m2 in Tillamook Bay to 0.02/m2 in the Umpqua estuary, generally de- creased from the northern to the southern estuaries (Table 2). The exception was Yaquina Bay. It was latitudinally the middle estuary, yet abundance of English sole there ranked above that in Siletz Bay. No consistent relationship was observed between mean abundances and the area of estuaries, river flows, tidal prisms, or flushing times using the data of Choi (1975) or Starr (1979)7. A broad size range of fish was caught in Tillamook, Siletz, and Alsea Bays, while we caught few in- dividuals larger than 36 mm in the Umpqua River estuary (Table 3). In Yaquina Bay, a higher propor- tion of large individuals (>65 mm) was found than in the other estuaries. A much broader range of sizes Table 2.— Mean abundance and standard deviation of 0-age English sole in five estuaries north and south of Yaquina Bay and along the open coast between 9 and 37 m, April-June 1978. No. of SD Location Date: 1978 hauls No./m2 (S) Estuary Tillamook Bay 8 May 21 0.715 0.916 Siletz Bay 9 May 21 0.184 0.206 Yaquina Bay 10 April; 12 July 6 0.332 0.251 Alsea Bay 10 May 21 0.059 0.075 Umpqua River 12 May 19 0.016 0.037 estuary Ocean Off Tillamook Bay 15 May; 17 June 12 0.005 0.013 Siletz Bay 16 May; 29 June 9 0.019 0.020 Alsea Bay 22, 29 June Umpqua River 28 June 3 0.001 0.001 estuary North of 16, 23 May; 14 0.006 0.011 Newport 29 June South of 18 June 9 0.003 0.004 Newport was captured in these estuaries than in open coastal areas on the dates sampled. Growth Despite prolonged recruitment of young English sole in Yaquina Bay (Fig. 4) distinct length modes were usually present for each sampling date Growth rates in Yaquina Bay, estimated by following the pro- gression of length modes of cohorts over time, were generally greatest (0.46-0.49 mm/d) during the late spring to early fall, while growth rates in winter were lower (0.26-0.32 mm/d) (Table 4). The growth rate from January to July 1970 was 0.47 mm/d, similar to the spring-fall estimates. Growth rates were estimated only for the spring-fall period off Moolack Beach. These were similar to those for Yaquina Bay fish but more variable, ranging from 0.28 to 0.42 mm/d. DISCUSSION Larvae of English sole are abundant in coastal waters off Oregon, ranking first among the flatfishes in some years (Richardson 19778; Richardson and Pearcy 1977; Mundy 1984). Young larvae (<10 mm) of English sole are rare in estuaries of the Oregon- California coast as evidenced by plankton samples 7Starr, R. M. 1979. Natural resources of Siletz esturary. Oreg. Dep. Fish Wildl., Estuary Inventory Rep. 2(4):l-44. 8Richardson, S. L. 1977. Larval fishes in Ocean waters off Ya- quina Bay, Oregon: Abundance, distribution and seasonality, January 1971 to August 1972. Oreg. State Univ. Sea Grant Publ. ORESU-T-77-003. 100 KRYGIER and PEARCY: NURSERY AREAS FOR YOUNG ENGLISH SOLE of only 6 larvae in 393 tows in Yaquina Bay (Pearcy and Myers 1974), 22 larvae in 84 tows in the lower Columbia River (Misitano 1977), and 4 larvae in 89 tows from Humboldt Bay (Eldridge 1970; Misitano 1970, 1976). However, young larvae are common in offshore collections (Porter 1964; Pearcy and Myers 1974; Laroche and Richardson 1979), and transform- ing larvae (19-22 mm) are frequent in collections from Humboldt Bay and the Columbia River estuary (Eldridge 1970; Misitano 1970, 1976). Thus young P. vetulus that enter estuarine nurseries do so as large transforming larvae or after completion of metamorphosis. Our data confirm the above findings. We found that settlement of metamorphosing English sole to the bottom was common both in the Yaquina Bay estuary and at Moolack Beach along the open coast. Transforming individuals along the coast were caught in largest numbers/m2 at depths of 16 m or less, but they were also captured at the deepest sta- tions sampled (Fig. 2). Since small larvae were rare in Yaquina Bay (Pearcy and Myers 1974), these trends suggest movement into the bay of transform- ing larval stages. Boehlert and Mundy (in prep.)9 have subsequently confirmed that small juveniles as well as transforming larvae of English sole recruit to Ya- quina Bay. Although densities of transforming larvae were sometimes higher at Moolack Beach than in Yaquina Bay, densities of juvenile fish >30 mm were usually over an order of magnitude higher in Yaquina Bay than at Moolack Beach, indicating either immigra- Table 3.— Length distribution of English sole caught in the five estuaries, Moolack Beach and grid stations, 10 April-12 June 1978. No. of < Standard lengths (mrr i) Location fish 14-20 21-25 26-30 31-35 36-40 41-45 46-50 51-55 56-60 61-65 66-70 71-75 76-80 81-85 86-90 8:V:78 Tillamook 2,979 904 1,619 296 48 19 23 26 31 13 4 4 2 9:V:78 Siletz Bay 673 242 256 72 36 13 13 21 14 5 1 10:V:78 Alsea Bay 306 41 98 49 25 20 15 19 27 9 1 1 1 12:V:78 Umpqua River estuary 54 30 12 5 4 1 1 1 10:IV:78 Yaquina Bay 163 46 16 1 6 11 11 11 23 18 11 6 2 1 12:VI:78 Yaquina Bay 156 2 6 9 9 18 6 6 12 17 23 18 16 8 3 3 10:IV:78 Moolack Beach 221 209 9 3 12:VI:78 Moolack Beach 24 5 12 5 1 1 23:V:78 Offshore grid 47 42 5 Table 4.— Growth of juvenile English sole esti- mated from modal progression of size-fre- quency histograms from catches in Yaquina Bay and Moolack Beach, 1970-79. mm/d Area and date (slope) ? Yaquina Bay Jan. 1970-July 1970 0.46 0.98 Dec. 1971 -Feb. 1972 0.26 0.92 Jan. 1972-Feb. 1972 0.32 0.91 Jan. 1978-Apr. 1978 0.31 0.91 Apr. 1970-Oct. 1970 0.46 0.96 May 1971-Oct. 1971 0.47 0.98 Mar. 1979-Sept. 1979 0.49 0.96 Moolack Beach Aug. 1978-Oct. 1978 0.41 0.98 May 1978-Oct. 1978 0.28 0.93 Apr. 1979-Sept. 1979 0.38 0.96 May 1979-Aug. 1979 0.42 0.99 June 1979-Sept. 1979 0.36 1.00 tion into the bay from the open coast during or after metamorphosis, or dispersal or higher mortality rates of young along the open coast than in the estu- ary. Increasing densities in Yaquina Bay, concurrent with decreasing densities at Moolack Beach, suggest immigration into the bay over an extended range of sizes from 25 to 40 mm. The mechanisms for such movements are not fully understood, but vertical movement of young fish off the bottom during periods of flood tide has been shown to effect transport into estuaries in several 9Boehlert, G. W., and B. C. Mundy. Recruitment dynamics of the English sole, Parophrys vetulus, to a west coast estuary. Unpubl. manuscr., 16 p. Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA. P.O. Box 3830, Hono- lulu, HI 96812. 129 FISHERY BULLETIN: VOL. 84 NO. 1 flatfish species. Cruetzberg et al. (1978) suggested that immigration of plaice, Pleuronectes platessa, lar- vae is based on such a "selective tidal transport," and that starvation induces the swimming behavior re- sulting in transport by currents. De Veen (1978) con- cluded that juvenile sole (Solea soled) use tidal trans- port to enter the Wadden Sea in the spring. Meta- morphosing larvae of the stone flounder, Kareius bicoloratus, also immigrate into estuarine nurseries with tidal currents; they were most abundant in plankton net collections during flood tides at night in an estuary of Sendai Bay, Japan (Tsurata 1978). Misitano (1976) captured metamorphosing English sole in a 1 m midwater trawl, especially after dark, in Humboldt Bay. Boehlert and Mundy (fn. 9) found that transforming English sole larvae were usually most abundant during flood tides at night in the moored plankton net that was nearest the bottom in the lower portion of the Yaquina Bay estuary and that recruitment to the bay was correlated with on- shore Ekman transport. Our estimates of growth from modal progressions length-frequency histograms [averaging 0.40 mm/d (s = 0.10) for Yaquina Bay and 0.37 mm/d (s = 0.06) for Moolack Beach] were considerably higher than Rosenberg's (1982) estimates even for the same years (Table 4). Rosenberg studied growth of 0-age English sole using fortnightly otolith rings as an aging tech- nique. He calculated that fish, 140-480 d of age, col- lected during 1978 and 1979 in Yaquina Bay and at Moolack Beach grew about 0.28 mm SL/d. Estimates of growth rates of juvenile English sole from length data by Westrheim (1955) in Yaquina Bay, as well as by Smith and Nitsos (1969) in Monterey Bay, and Van Cleve and El-Sayed (1969) and Kendall (1966) in Puget Sound were more similar to our estimates than those of Rosenberg (1982, table 2). The differ- ences in apparent growth rates between length fre- quency and otolith measurements are difficult to ex- plain. Avoidance of nets by larger sole (e.g., Kuipers 1975), emigration of larger fish out of the sampling area in the late summer, and prolonged immigration of small fish into the estuary, are likely. Any of these would result in an underestimates of growth by the length-frequency method (see Rosenberg 1982 for opposite explanations). Differential mortality of small fish (Rosenberg 1982) or methodological diffi- culties in analyzing otolith growth increments may also help explain the differences. Our study confirms the observations of Laroche and Holton (1979) that small 0-age English sole are not found exclusively in estuaries along the Oregon coast, and that average sizes of English sole increase with depth at Moolack Beach. Laroche and Holton (1979) suggested that even low density or localized utilization of the extensive unprotected offshore areas along the coast could be an important factor in determining the English sole production off Ore- gon. Tb evaluate this possibility, we determined total areas within the range of our sample depths in the lower reaches of the five estuaries and multiplied these areas by the average catch/m2 of 0-age English sole (<90 mm) to obtain an estimate of total number of young English sole in each estuary. The average catch was also determined from 47 collec- tions between 9 and 36 m where we found highest catches of 0-age fish, along 448 km of the open coast from our May-June catches (Table 2). The average catch/m2 of 0-age sole in the five estuaries usually was many times that along the open coast. But be- cause of the large differences in areas, the estimate for total abundance of 0-age sole during the May- June period on the open coast was about 643 x 105, considerably higher than the estimate for the five estuaries, 140 x 105. Most of the fish caught during this period, however, were transforming or recently metamorphosed juveniles that could have entered estuaries later in the year. This may in part explain the 17-fold decrease in average abundance of small sole along the open coast between 16-23 May (x = 0.039, n = 18, s = 0.11) and 28-29 June (x = 0.002, n = 29, s = 0.004) in the vicinity of Tillamook and Siletz Bays. Our estimate of total abundance along the coast in June is 70 x 105, about half the estimate for the five estuaries about a month and one-half earlier. Because of our small sample sizes, lack of sampling in some estuaries and open coast areas, and temporal differences (and associated mor- tality) among samples, these estimates must be con- sidered crude. Nevertheless, they suggest that shal- low waters of the open coast are important initial settling areas for English sole and that both estu- aries and the open coast are nursery grounds for fully transformed 0-age sole We need data on the growth and survival from estuarine and open coastal areas to evaluate their importance as nursery grounds and to assess their relative contributions to the commercially harvested and spawning population. Olsen and Pratt (1973) used parasites as indicators of English sole nursery grounds. The incidence of Echinorhynchus lageni- formis, an acanthocephalan that they considered was acquired only in estuaries, averaged 29.9% in 0-age English sole <117 mm SL captured in Yaquina Bay and 28.5% in 0-age fish collected offshore at depths of 10-80 m near the entrance of Yaquina Bay dur- ing November and December, a period after most 0-age fish had emigrated from the bay. They con- 130 KRYGIER and PEARCY: NURSERY AREAS FOR YOUNG ENGLISH SOLE eluded from these similar incidences of infection that there was no sizable influx of 0-age English sole to their offshore study area other than from estuarine nursery grounds. Their results imply that any 0-age fish that reside along the open coast during the spring and summer have much higher mortality rates than estuarine residents and do not contribute significantly to the offshore population of 0-age fish. Growth rates of 0-age English sole from Moolack Beach and Yaquina Bay, however, do not support this hypothesis. They appear to be similar (Rosenberg 1982; Table 4). Our catch curves (Fig. 5C, D) also pro- vide no evidence for grossly higher mortality rates at Moolack Beach. The total declines in abundances per m2 are fairly similar for English sole 50-100 mm, presumably a size range that occurs after immi- gration into the estuary but before emigration of larger sizes out of the estuary in the fall. The fact that 0-age English sole immigrate from offshore into estuaries where they are found in high concentrations suggests that this behavior is adap- tive Standing stocks and productivity of small ben- thic food organisms are undoubtedly higher in estu- aries than along the open coast, but because of the higher concentrations of young flounder in Yaquina Bay than Moolack Beach (Fig. 5), competition for food probably results in similar growth rates in these two habitats. The rapid decreases in the estuarine densities of 0-age English sole during the fall and winter months are evidence of emigration out of estuaries to offshore areas. In Yaquina Bay, we found a decrease in density of 0-age fish in the late fall as well as a decrease in average size at this time. Fre- quently age-0 (20-55 mm) and age-I (75-115 mm) fish were both present in the winter, with the age-I fish disappearing entirely from catches in the spring. Westrheim (1955) and Olsen and Pratt (1973) also found decreases in catch per effort and average sizes of young English sole that indicated definite emi- gration from Yaquina Bay after October. Forsberg et al. (1975)10 reported emigration of English sole from Tillamook Bay in early fall with few individuals remaining in November. According to Bayer (1981), small English sole were common at intertidal stations in Yaquina Bay most of the year, but they were absent during November and were less common during other fall months. Toole (1980) also found that English sole disappeared from intertidal areas in early fall at an average size of 68 mm SL and subsequently resided in subtidal 10Forsberg, B. O., J. A. Johnson, and S. M. Klug. 1975. Identi- fication and notes on food habits of fish and shellfish in Tillamook Bay, Oreg. Fish Comm. Oreg. Contract Rep., 85 p. channels until they were about 120 mm SL in Hum- boldt Bay. He associated these different distributions with changes in feeding habits, and possibly with a reduction in intraspecific competition among small and large 0-age English sole Indeed, emigration out of bays and estuaries in the fall may be related to limitations in the carrying capacity for high densities and standing stocks of young English sola We conclude that estuarine and offshore nursery grounds combine to significantly increase the sur- vival and total population size of 0-age fish. Utiliza- tion of these two diverse habitats may also improve the chances for good survival of young fish from at least one habitat even when adverse conditions af- fect the other. LITERATURE CITED Ahlstrom, E. H., and H. G. Moser. 1975. Distributional atlas of fish larvae in the California Cur- rent region: flatfishes, 1955 through 1960. Calif. Coop. Oceanic Fish. Invest., Atlas 23:1-207. Alderice, D. F, and C. R. Forrester. 1968. Some effects of salinity and temperature on early development and survival of the English sole (Parophrys vetulus). J. Fish. Res. Board Can. 25:495-521. Bayer, R. D. 1981. Shallow-water intertidal ichthyofauna of the Yaquina estuary, Oregon. Northwest Sci. 55:182-193. Budd, P. L. 1940. Development of the eggs and early larvae of six Califor- nia fishes. Calif. Dep. Fish Game, Fish Bull. 56, 53 p. Carey, A. G., Jr., and H. Heyomoto. 1972. Techniques and equipment for sampling benthic or- ganisms. In A. T. Pruter and D. L. Alverson (editors). The Columbia River estuary and adjacent ocean waters, bioen- vironmental studies, p. 378-408. Univ. Wash. Press, Seattle Choi, B. 1975. Pollution and tidal flushing predictions for Oregon's estuaries. M.S. Thesis, Oregon State Univ., Corvallis, 163 p. Creutzberg, F, A. Th. G. W. Eltink, and G. J. Van Noort. 1978. The migration of plaice larvae, Pleuronectes platessa, into the western Wadden Sea. In D. S. McLusky and A. J. Berry (editors), Physiology and behavior of marine or- ganisms, p. 243-252. Proc. 12th Eur. Symp. Mar. Biol., Stir- ling, Scotl. Pergamon Press, N.Y. Demory, R. L. 1971. Depth distribution of some small flatfishes off the north- ern Oregon-southern Washington coast. Res. Rep. Fish. Comm. Oreg. 3:44-48. De Veen, J. F 1978. On selective tidal transport in the migration of North Sea plaice (Pleuronectes platessa) and other flatfish species. Neth. J. Sea Res. 12:112-147. Eldridge, M. 1970. Larval fish survey of Humboldt Bay. M.S. Thesis, Hum- boldt State Coll., Areata, 52 p. Forrester, C. R. 1969. Results of English sole tagging in British Columbia waters. Pac. Mar. Fish. Comm. Bull. 7:1-10. Harry, G. Y., Jr. 1959. Time of spawning, length at maturity, and fecundity of 131 FISHERY BULLETIN: VOL. 84. NO. 1 the English, petrale, and Dover soles (Parophrys vetulus, Eopsetta jordani, and Microstomia pacificus, respectively). Oreg. Fish Comm., Res. Briefs 7(1):5-13. Hart, J. L. 1973. Pacific fishes of Canada. Fish. Res. Board Can. Bull. 80, 740 p. Hewitt, G. R. 1980. Seasonal changes in English sole distributions: An anal- ysis of the inshore trawl fishery off Oregon. M.S. Thesis, Oregon State Univ., Corvallis, 59 p. Jow, T. 1969. Results of English sole tagging off California. Pac Mar. Fish. Comm. Bull. 7:15-33. Kendall, A. W., Jr. 1966. Sampling juvenile fishes on some sandy beaches of Puget Sound, Washington. M.S. Thesis, Univ. Washington, Seattle, 77 p. Ketchen, K. S. 1947. Studies on lemon sole development and egg production. Fish. Res. Board Can., Prog. Rep. Pac. 73:68-70. 1956. Factors influencing the survival of the lemon sole (Paro- phrys vetulus) in Hecate Strait, British Columbia. J. Fish. Res. Board Can. 13:647-694. Kruse, G. H., and A. V. Tyler. 1983. Simulation of temperature and upwelling effects on the English sole (Parophrys vetulus) spawning season. Can. J. Fish. Aquat. Sci. 40:230-237. Krygier, E. E., and H. F. Horton. 1975. Distribution, reproduction, and growth of Crangon nigricauda and Crangon franciscorum in Yaquina Bay, Oregon. Northwest Sci. 49:216-240. Kuipers, B. 1975. On the efficiency of a two-meter beam trawl for juvenile plaice (Pleuronectes platessa). Neth. J. Sea Res. 9:69-85. Laroche, J. L., and S. L. Richardson. 1979. Winter-spring abundance of larval English sole, Paro- phrys vetulus, between the Columbia River and Cape Blanco, Oregon during 1972-75 with notes on occurrences of three other pleuronectids. Estuarine Coastal Mar. Sci. 8:455-476. Laroche, J. L., S. L. Richardson, and A. A. Rosenberg. 1982. Age and growth of a pleuronectid, Parophrys vetulus, during the pelagic larval period in Oregon coastal waters. Fish. Bull., U.S. 80:93-104. Laroche, W. A., and R. L. Holton. 1979. Occurrence of 0-age English sole, Parophrys vetulus, along the Oregon coast: An open coast nursery area? North- west Sci. 53:94-96. Misitano, D. A. 1970. Aspects of the early life history of English sole (Paro- phrys vetulus) in Humboldt Bay, California. M.S. Thesis, Humboldt State Coll., Areata, 57 p. 1976. Size and stage of development of larval English sole, Parophrys vetulus, at time of entry into Humboldt Bay. Calif. Fish Game 62:93-98. 1977. Species composition and relative abundance of larval and post-larval fishes in the Columbia River estuary, 1973. Fish. Bull., U.S. 75:218-222. Mundy, B. C. 1984. Yearly variation in the abundance and distribution of fish larvae in the coastal upwelling zone off Yaquina Head, Oregon from June 1969 to August 1972. M.S. Thesis, Ore- gon State Univ., Corvallis, 158 p. Myers, K. W W 1980. An investigation of the utilization of four study areas in Yaquina Bay, Oregon, by hatchery and wild juvenile sal- monids. M.S. Thesis, Oregon State Univ., Corvallis, 234 p. Olsen, R. E., and I. Pratt. 1973. Parasites as indicators of English sole (Pa rophrys vetu- lus) nursery grounds. Trans. Am. Fish. Soc 102:405-411. Pattie, B. H. 1969. Dispersal of English sole, Parophrys vetulus, tagged off the Washington coast in 1956. Pac Mar. Fish. Comm. Bull. 7:11-14. Pearcy, W G., and S. S. Myers. 1974. Larval fishes of Yaquina Bay, Oregon: A nursery ground for marine fishes? Fish. Bull., U.S. 72:201-213. Porter, P. 1964. Notes on fecundity, spawning and early life history of petrale sole (Eopsetta jordani) with descriptions of flatfish larvae collected in the Pacific Ocean off Humboldt Bay, California. M.S. Thesis, Humboldt State Coll., Areata, 98 p. Richardson, S. L., and W. G. Pearcy. 1977. Coastal and oceanic fish larvae in an area of upwelling off Yaquina Bay, Oregon. Fish. Bull., U.S. 75:125-145. Rosenberg, A. A. 1982. Growth of juvenile English sole, Parophrys vetulus, in estuarine and open coastal nursery grounds. Fish. Bull., U.S. 80:245-252. Rosenberg, A. A., and J. L. Laroche. 1982. Growth during metamorphosis of English sole, Parophrys vetulus. Fish. Bull., U.S. 80:150-153. Sims, C. W, and R. H. Johnson. 1974. Variable-mesh beach seine for sampling juvenile salmon in Columbia River estuary. Mar. Fish. Rev. 36(2):23-26. Smith, J. G., and R. J. Nitsos. 1969. Age and growth studies of English sole, Parophrys vetu- lus, in Monterey Bay, California. Pac Mar. Fish. Comm. Bull. 7:73-79. Tate, M. W, and R. C. Clelland. 1957. Nonparametric and shortcut statistics in the social, biological, and medical sciences. Interstate Printers and Publishers Inc, Danville, IL, 171 p. Toole, C. L. 1980. Intertidal recruitment and feeding in relation to optimal utilization of nursery areas by juvenile English sole (Paro- phrys vetulus: Pleuronectidae). Environ. Biol. Fishes 5: 383-390. , Tsuruta, Y 1978. Field observations on the immigration of larval stone flounder into the nursery ground. Tohoku J. Agric Res. 29: 136-145. Van Cleve, R., and S. Z. El-Sayed. 1969. Age, growth, and productivity of an English sole (Paro- phrys vetulus) population in Puget Sound, Washington. Pac Mar. Fish. Comm. Bull. 7:51-71. Westrheim, S. J. 1955. Size composition, growth and seasonal abundance of juvenile English sole (Parophrys vetulus) in Yaquina Bay. Oreg. Fish. Comm. Res. Briefs 6(2):4-9. 132 ORGANIC AND TRACE METAL LEVELS IN OCEAN QUAHOG, ARCTICA ISLANDICA LINNE, FROM THE NORTHWESTERN ATLANTIC Frank W. Steimle,1 Paul D. Boehm,2 Vincent S. Zdanowicz,1 and Ralph A. Bruno1 ABSTRACT Chemical contamination of biological resources is an important problem for resource managers. This study reports on body burden levels of several contaminants of concern: polychlorinated biphenyls (PCB), poly- nuclear aromatic hydrocarbons (PAH) of both petroleum and combustion sources, total petroleum hydrocar- bons, and seven trace metals (Ag, Cd, Cr, Cu, Ni, Pb, and Zn) in a resource species, the ocean quahog, collected between Virginia and Nova Scotia. Organic and trace metal contaminants were detected, at low levels, in all samples examined, with highest levels being generally found in samples from the inner New York Bight and Rhode Island Sound. The highest PCB and PAH values were 27 and 55 ppb, respectively; Ag, Cd, and Cr values were generally <5 ^g/g dry weight; Cu, Ni, and Pb generally <15 ^g/g dry weight with a few exceptions; and Zn ranged from 50 to 153 uglg dry weight. The ocean quahog, Arctica islandica Linne, is a large, bivalve mollusc found on both sides of the North Atlantic In the northwestern Atlantic, it oc- curs from just north of Cape Hatteras, NC, to New- foundland, Nova Scotia, being most abundant on the middle to outer continental shelf at depths between about 30 and 150 m (Merrill et al. 1969). The species is edible and some commercial harvesting has oc- curred since 1943 in the Rhode Island area; however, intensive fishing for this species did not begin until the 1970s when surf clam, Spisula solidissima (Dill- wyn), stocks, an inshore species, were drastically reduced by overfishing (Ropes 1979). Arctica islandica generally inhabit silty sand sedi- ments of the middle to outer continental shelf that are less influenced by waves and strong currents than shallower areas. Areas of silty sand are thought to be at least partially depositional in nature, i.e, fine organic-rich particles tend to accumulata It is gen- erally agreed that many chemical pollutants, intro- duced to the marine environment via impacted estu- aries and coastal areas, ocean dumping, and atmo- spheric sources, often are bound to and associated with fine organic and inorganic particle aggregates, both in the water column and at the sediment sur- face These aggregates ultimately can accumulate in these natural depositional areas as the results of some recent studies show that contaminants ap- 'Northeast Fisheries Center Sandy Hook Laboratory, National Marine Fisheries Service, NOAA, Highlands, NJ 07732. 2Battelle, New England Marine Research Laboratory, 397 Wash- ington Street, Duxbury, MA 02332. parently are accumulating in silty areas relatively remote from most possible sources, eg., organic con- taminants found south of Cape Cod, MA, in the mid- dle to outer continental shelf (Boehm 1983a). Some authors have also reported a trend of increasing sedi- ment trace metal levels with depth on the Middle Atlantic shelf (Harris et al. 1977), but the specific sources of these contaminants are still unknown. Because A. islandica is a common, sedentary, long- lived (Thompson et al. 1980) inhabitant of these sil- ty sands that frequently contain higher levels of con- taminants than coarser sands, the species may be particularly susceptible to contamination. Wenzloff et al. (1979) reported "greater average concentration of silver, arsenic, cadmium, copper, and zinc ... in ocean quahogs than in surf clams" for the Mid- dle Atlantic Surf clams are generally found in shallower, medium sand areas. Thus, A. islandica may be a good offshore "indicator" species to moni- tor for trends in marine chemical pollution. Although some studies on contaminant body burdens of A. islandica have been reported (ERCO 19783; Sick 1978, 1981; Wenzloff et al. 1979; Reynolds 1979; Payne et al. 1982), these studies have been limited generally to a particular restricted area, have not examined both types of contaminants or only a few components of each contaminant class, or have ex- amined only certain tissues, not whole body levels. The present study provides body burden data over Manuscript accepted April 1985. FTSHF.RY RTTT.T.F.TTN- VDT, 84 NO 1 IftSfi 3ERCO (Energy Resources Company). 1978. New England OCS Environmental Benchmark. Draft Final Rep., Vol. II, to U.S. Dep. Inter., Bur. Land Manage, Miner. Manage Serv., 628 p. 133 FISHERY BULLETIN: VOL. 84, NO. 1 a wide range of this species' occurrence in the north- western Atlantic and includes information on or- ganic, La, polychlorinated biphenyls (PCB), polynu- clear aromatic hydrocarbons (PAH) from combustion and petroleum sources, and bulk levels of the petro- leum hydrocarbon (PHC) class, and seven trace metal contaminants. The study includes the first known set of PCB data for this species. MATERIALS AND METHODS Ocean quahog samples were obtained at random stations from wide areas on the continental shelf of the northwestern Atlantic (Fig. 1). These were col- lected from annual, summer hydraulic dredge shell- fish surveys of NOAA's Northeast Fisheries Center from 1981 and 1982. At most stations, 10-12 medium-sized clams were selected, as available Half of the collection was prepared for organic analysis by wrapping them in aluminum foil that had been prewashed with spectral grade acetone followed by dichloromethane; the remaining half for trace metals were placed in polyethylene plastic bags. All were quickly frozen at -20°C. In certain areas where there were not sufficient samples at a particular sta- tion to provide material for both organics and trace metal analyses, samples were collected at a nearby station, with similar environmental characteristics, to complete the collection for the area. These paired station samples were not intermixed. Chemical Analysis - Organics In the laboratory, the thawed whole meats of each of the five or six individual clams in each station sam- ple were removed from the shells, pooled, and homo- genized in a high-speed blender. A 100 g (wet weight) aliquot was removed from the homogenate and pro- cessed according to the extraction, fractionation, and analytical methodology described by Warner (1976), as modified by Boehm et al. (1982). After aqueous caustic (0.5N KOH) digestion of the tissue for 12 h, the digestate was back-extracted three times with hexane The hexane extract was concentrated by ro- tary evaporation, then fractionated on a 5% deacti- vated alumina/activated silica gel column. The first eluting fraction from the alumina/silica column (fj) contained the saturated PHC; the second fraction (f2) contained the PCB and PAH. Quantitation pro- cedures closely followed those by Boehm (1983b). PHC factors were quantified using the internal standard method whereby all peaks are quantified relative to androstane in the fj fraction and 0-ter- phenyl in the f2 fraction. PCBs were quantified relative to the internal standard tetrazene (2, 3, 5, 6 tetrachloronitroben- zene). The average relative response factors of two or three isomers in each of the di-, tri-, tetra-, penta-, hexa-, hepta- and octachlorobiphenyls groups were applied to the sum of the peaks in each grouping. Thus, PCBs were quantified by isomer group rather than according to the Aroclor4-type quantification (Duinker et al. 1980, 1983; Boehm 1983b). PHCs were determined by the total of f: and f2 fractions, as analyzed by high resolution (fused silica capillary) gas chromatography with flame ionization detection (GC2/FID). A Hewlett Packard model 5840A gas chromatograph was used for all GC2 deter- minations. A 30 m fused silica SE-30 (0.25 mm i.d.; J and W Scientific) column was used to analyze the saturated hydrocarbon (ft) fraction. A 30 m SE-52 fused silica column was used to analyze the aroma- tic/olefinic (f2) fraction by GC2/FID and the same fraction by gas chromatograph/mass spectrometer (GC/MS) (see below). The f2 fractions were analyzed by GC2/ECD (electron capture detection) to obtain the PCB concentrations. PCBs were analyzed on a 30 m SE-52 fused silica column. The f2 fraction was also analyzed by a Finnegan MAT model 4530 computer-assisted GC/MS system for PAH deter- minations. GC/MS conditions were as follows: ioniza- tion voltage, 70 ev; electron multiplier voltage 2,000 volts; scan conditions 45-450 amu at 400 amu/s. Chemical Analysis - Trace Metals Whole clams, 5 or 6 per station, were thawed, and the whole body removed from the shells. Each indivi- dual clam was weighed in Pyrex beakers and dried for 16-20 h at 105°C. Twenty mL of 70% trace metal grade nitric acid were added to each beaker, which was covered with a Pyrex watch glass and heated (70° -75°C) on a ceramic hot plate until dry. After cooling to room temperature, another 20 mL of con- centrated nitric acid were added and the dissolution continued. After 3 or 4 repeated acid additions and evaporations, 10 mL of 30% hydrogen peroxide were added, the solutions evaporated to near dryness and removed from the heat. When cooled, samples were filtered through Whatman #4 filter paper and brought to a final volume of 25 mL in a Pyrex glass- stoppered graduated cylinder by adding 5% (w/v) ni- tric acid. Analysis was performed on a Perkin Elmer model 5000 atomic absorption (AA) spectrophotom- eter employing an air-acetylene flame and conven- 4Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 134 STEIMLE ET AL.: ORGANIC AND TRACE METALS IN OCEAN QUAHOG 44" 76° 74 !£ NARRAGANSETT BAY, NEW YORK BIGHT APEX "Mud Patch' "NEW BEDFORD HARBOR AND BUZZARDS BAY GEORGES BANK ®379 245 LEGEND • HEAVY METALS o ORGANICS ® BOTH 50 100 150 200 KILOMETERS 44" ^ 42- 40° 38°J 36 ■ ■ ■ ■-.- •_■ NMFS SanOv Hoc* 70° 68° 66" Figure l.-Station locations' collections oi Arctic a islandica. Stations 367, 335 and 349 are on the Scotian Shelf at the following coor- dinates: Station 367 (lat. 43°44'N, long. 61°08'W), Station 335 (43°25'N, 61°42'W) and Station 349 (43°21'N, 61°23'W). tional AA techniques. Reagent blanks were carried through the same procedure All reagents used were of trace metal analytical grade Deionized water was of 18 megohm purity. The National Bureau of Stan- dards (NBS) SRM 1566, freeze-dried oyster homog- enate, was used as the tissue standard. Recoveries were at least 80% of this standard in all cases. RESULTS The analytical results for organic contaminants are presented in Tables 1 (PHC and PCB) and 2 (PAH). PHC values are given as total saturated and aromatic hydrocarbons as determined by GC2. PCB values are given as total tri-, tetra-, penta-. hexa-, and hepta- 135 FISHERY BULLETIN: VOL. 84, NO. 1 Table 1.— PHC (petroleum hydrocarbon) and PCB (polychlorinated biphenyl) levels in northwestern Atlantic Arctica islandica. Area and station PHC O^g/g wet weight) Saturated Aromatic Total PCB (ng/g wet weight) Cl3 Cl4 Ci5 Cl6 Cl7 Total Inshore New York Bight 22 0.2 1.3 1.5 4.8 5.1 1.6 4.4 0.2 16.1 26 6.0 0.9 6.9 0.4 0.3 0.2 0.5 0.1 1.5 27 0.4 0.8 1.2 7.2 2.8 1.6 1.8 <0.1 13.4 28 0.2 0.9 1.1 6.7 4.2 2.6 6.5 0.1 20.2 32 3.2 4.1 7.3 5.4 5.1 4.1 10.7 0.7 26.8 47 <0.1 0.2 0.2 1.1 0.4 0.1 0.3 — 1.9 59 3.1 1.5 4.6 3.4 3.6 3.0 5.7 0.5 16.4 Offshore NJ-VA 77 1.7 0.7 2.4 5.6 1.2 2.9 2.3 0.3 12.2 107 0.2 0.8 1.0 8.0 1.7 0.8 2.8 0.2 13.3 168 0.2 1.1 1.3 2.3 1.3 1.0 3.5 0.5 8.5 173 0.4 0.5 0.9 4.4 4.5 1.3 3.4 — 14.0 174 1.0 0.3 1.3 2.6 0.8 0.4 1.5 <0.1 5.5 224 1.7 0.2 1.9 2.3 1.5 0.5 0.5 <0.1 4.9 Inshore S. New England 237 0.1 0.8 0.9 1.0 0.6 <0.1 <0.1 — 1.7 246 1.2 0.4 1.6 2.0 2.8 2.2 1.5 — 8.5 244 2.2 1.3 3.5 4.9 9.5 2.3 3.8 — 20.4 261 2.0 0.4 2.4 3.4 3.5 2.1 3.1 <0.1 12.1 239 2.9 1.1 4.0 1.4 2.1 1.6 1.9 <0.1 7.0 240 2.8 0.9 3.7 2.2 2.6 2.7 3.5 <0.1 11.0 241 2.3 1.6 3.9 4.1 6.6 6.2 6.3 0.2 23.2 242 1.9 1.8 3.7 2.9 4.6 5.9 6.5 0.2 20.1 Georges Bank 379 0.8 1.1 1.9 0.7 1.2 0.9 1.0 <0.1 3.8 Scotian Shelf 367 0.6 0.2 0.8 0.8 0.5 0.4 0.3 0.3 2.2 335 1.1 0.1 1.2 0.9 0.4 0.5 2.3 0.1 4.2 349 4.5 0.6 5.1 0.8 0.9 0.4 0.1 — 2.2 chlorobiphenyls (C13-C17), as well as total PCB. PAH values are presented as individual compounds (e.g, napthalene) or as homologous series (SN). Table 3 lists the mean trace metal concentrations and stan- dard deviations; data are presented on a dry weight basis to simplify comparisons with other studies. DISCUSSION PCB levels observed in this survey ranged from 2 to 30 ng/g (ppb) wet weight (Table 1). These values are in general agreement with other data reported for PCB levels in other coastal bivalves (Giam et al. 1976; Goldberg 1978; Gadbois 1982), but are lower than those (to 400 ppb) reported for estuarine spe- cies (Goldberg 1978; MacLeod et al. 1981; O'Connor et al. 1982; ERCO 1983). However, we have found little data on PCB levels in offshore molluscs nor any other data on PCB levels in A. islandica for compari- son. None of the A. islandica levels approach the cur- rent 2 ppm (= 2,000 ppb) U.S. Food and Drug Ad- ministration (FDA) "seafood action limit" for human consumption. In spite of the wide geographical range sampled, PCB levels were relatively uniform with only an order of magnitude difference between the high and low values. Clearly the Georges Bank (station 379) and remote Nova Scotia (stations 367, 335, 349) ocean quahogs were minimally contaminated, with their levels (2-5 ppb) reflecting the global PCB trans- port phenomena. The ocean quahogs in the near- shore New York Bight, Rhode Island Sound, and Buzzards Bay were more contaminated, with values up to 25 ppb. It is not surprising as previous biogeo- chemical studies in the western North Atlantic have clearly shown that several major urban pollutant sources influence the nearshore environment. For example, inputs of PCBs are specifically known to occur in the New York Bight, from esturine fluxes and via direct ocean dumping (Boehm 1983b) and in Buzzards Bay, MA, from industrial inputs to the New Bedford Harbor region (Weaver 1982). Some- what surprising were the elevated levels at some sta- tions on the outer New Jersey shelf (12-16 ppb) and in the Hudson Canyon area (20 ppb). Offshore trans- port of PCB material towards these stations via riverine fluxes followed by southerly transport along the New Jersey shore and down-canyon transport of ocean-dumped material are possible modes of transport to these stations (Boehm 1983b). 136 STEIMLE ET AL.: ORGANIC AND TRACE METALS IN OCEAN QUAHOG Table 2.— PAH (polynuclear aromatic hydrocarbon) levels in northwestern Atlantic Arctica islandica (ng/g wet weight). Area and station N IN P IP IDBT IF 1202 1228 1252 B(a)P IPAH PPI1 Inshore New York Bight 22 nd nd 4.0 11.9 2.0 1.2 5.5 1.1 <1 <1 23 40 26 nd nd 1.1 1.1 nd nd 1.1 nd nd nd 3.3 0 27 1.0 4.5 2.1 9.1 2.7 1.2 2.7 <1 <1 nd 22 54 28 9.1 12.0 1.3 5.2 <1 nd 1.8 nd nd nd 20 72 32 nd nd 3.9 12.4 nd nd 11.1 14.1 17.3 6.0 55 7 47 4.3 5.1 2.9 2.9 nd nd 3.1 3.0 4.0 2.0 18 28 59 1.0 5.3 1.0 11.5 <1 nd 1.5 <1 nd nd 20 77 Offshore NJ-VA 77 <1 3.7 3.3 9.2 <1 1.0 2.4 <1 nd nd 18 65 107 <1 6.7 2.5 10.0 2.5 3.5 1.5 <1 <1 nd 26 77 168 nd nd 1.8 1.8 nd nd 2.4 nd nd nd 4.2 0 173 1.3 5.9 1.8 6.2 1.3 2.0 2.3 <1 nd nd 19 72 174 <1 4.0 1.0 5.0 <1 nd 4.0 1.0 1.0 nd 16 56 224 1.4 6.1 2.0 7.8 2.1 1.5 1.3 <1 <1 <1 21 74 Mud Patch 237 <1 <1 2.8 5.0 <1 nd 5.7 3.7 5.4 2.5 19 31 246 nd 11.9 2.2 11.5 1.0 1.3 3.3 nd nd nd 29 81 Inshore S. New England 244 nd nd nd nd 2.4 nd 1.7 <1 <1 <1 6.1 39 261 nd nd 3.6 9.2 <1 <1 3.3 nd nd nd 15 51 239 nd nd 1.6 1.9 nd nd 2.9 <1 1.2 <1 7.0 4 240 <1 3.3 1.8 5.6 <1 <1 2.8 <1 <1 <1 16 51 241 nd nd nd 5.0 nd nd 4.0 1.0 1.0 <1 12 42 242 nd nd <1 <1 nd nd 1.5 nd nd nd 2.5 40 Georges Bank 379 nd nd <1 <1 nd nd <1 nd nd nd <1 0 Scotian Shelf 367 1.0 1.0 1.5 1.5 nd nd 1.1 nd nd nd 3.6 28 335 nd nd nd nd nd nd nd nd nd nd nd 0 349 4.3 5.1 2.9 2.9 nd nd 3.1 3.0 4.0 2.0 18 28 Wet weight concentrations = dry weight concentration 4- 7. N = naphthalene. IN = total naphthalenes (C0-CJ. P = phenanthrene. IP = total phenanthrenes (Cq-CJ. IDBT = total dibenzothiophenes (C0-C3). IF = total fluorenes (C0-C3). 1202 = fluoranthene + pyrene. 1228 = benzanthracene + chrysene. 1252 = benzofluoranthenes + benzopyrenes. B(a)P = benzo(a)pyrene. nd = not detected (<1 ng/g wet weight). dd, . . , *N + IDBT + (IP-P) + IF PPI = percent petroleum index = IPAH =3M + IP + IDBT 'From Boehm (1983a). + IF + 1202 IPAH 1252 + 1228 TV-ends in the PHC and PAH data reveal large-scale homogeneity in the concentrations observed. PAH levels ranged from nondetectable to 55 ppb, the high- est values occurring at the station 32 samples from the New York Bight, where the highest PCB level (27 ppb) was also observed. Although our sampling on Georges Bank consisted of only one station, results were similar to those of a more extensive study by Payne et al. (1982), the only other study of A. islandica we could locate that includes PHC data. If the entire Northeast region is considered a sample set, then the PAH values were 16.7 ± 12.0. However, the composition of the PAH which com- prises the total PAH number varied considerably, ranging from 0 to 81% "petroleum" PAH (Table 2). The percent petroleum index (PPI), developed by Boehm (1983a, b), estimates the relative contribu- tions of uncombusted fossil fuels, eg, petroleum, and from combustion sources to the total PAH assem- blage This indice, presented in Table 2, is based on the relative abundance of petroleum constituents, such as naphthalene, flourenes, dibenzothiophenes, and alkylated phenanthrenes, to the total PAH mix. The differences in PPI values for the various samples cannot, at this time, be ascribed to specific trans- port or selective uptake factors. However, a knowl- 137 FISHERY BULLETIN: VOL. 84. NO. 1 Table 3—Arctica islandica trace metal body burdens (mean and standard deviation, ^glg-6ry weight) in areas of the northwest Atlantic; N = number of individual clams examined at each site. Results of analysis of SRM 1566 are also included; 5-8 NBS (National Bureau of Standards) samples were ex- amined for each metal (nd = nondetectable). Area Ag Cd C r Cu N i Pb Zn station N X ±SD X ±SD X + SD X ±SD X ±SD X + SD X ±SD Georges Bank 379 6 0.79 0.25 1.36 0.33 3.07 1.38 10.30 2.22 3.46 1.17 4.08 2.07 61.8 11.4 Nantucket 245 5 0.96 0.09 2.75 0.66 2.98 0.83 7.25 1.41 9.54 3.81 5.02 2.21 88.3 21.3 S. New England 237 6 2.65 2.08 3.22 0.65 2.72 0.65 12.76 3.30 27.19 8.18 6.90 1.87 153.9 87.6 181 6 1.14 0.95 3.49 1.39 2.24 0.23 11.70 2.97 21.84 7.22 11.03 4.48 124.7 30.8 244 6 0.56 0.14 1.36 0.47 2.19 1.02 6.49 3.29 4.47 1.75 2.99 1.33 84.1 25.7 Rhode Island Sol nd 239 6 0.79 0.25 1.36 0.33 3.07 1.38 10.30 2.22 3.46 1.17 4.08 2.07 61.8 11.4 240 6 1 .76 0.65 1.39 0.48 4.02 1.26 11.47 2.92 6.28 1.61 6.71 2.51 87.4 12.8 241 6 1.59 0.93 0.96 0.14 3.96 2.23 12.47 3.80 5.83 2.42 4.61 1.71 126.3 55.4 Block Island Sound 261 6 1 .53 1.82 1.94 0.68 4.56 0.33 10.22 1.55 11.64 3.28 10.17 2.48 101.9 32.6 S. Long Island 189 6 0.74 0.60 2.48 0.63 1.88 0.59 8.78 0.94 18.73 4.75 3.30 0.80 128.4 35.2 23 6 1.18 0.53 2.17 0.67 1.09 0.19 10.31 3.27 17.28 6.07 3.41 1.12 120.2 19.6 26 6 0.84 0.45 1.43 0.56 5.47 1.22 15.78 5.83 9.87 2.73 8.66 3.25 117.6 35.7 29 6 5.25 1.64 1.06 0.29 4.78 3.35 13.65 3.20 8.93 5.06 9.67 4.11 100.7 49.9 New Jersey Shelf 32 6 0.52 0.26 0.67 0.35 2.38 0.20 8.16 5.38 4.47 2.90 3.46 2.24 50.2 22.0 47 5 0.53 0.36 1.20 0.42 1.46 0.83 8.37 2.49 9.87 5.09 3.11 0.87 84.1 31.8 59 3 0.44 0.15 0.23 0.05 0.90 0.09 6.01 1.37 6.01 1.68 1.64 0.28 62.2 18.8 174 6 1.50 0.91 2.19 0.93 1.87 0.65 4.08 0.82 7.79 2.70 4.16 1.60 50.8 6.0 224 6 0.46 0.13 3.06 0.91 1.78 0.43 5.63 1.02 14.91 6.92 5.60 2.23 91.9 33.9 Delmarva Shelf 107 6 0.39 0.25 1.87 0.62 2.44 0.90 4.16 1.02 10.27 2.52 4.80 2.77 58.9 14.2 173 5 0.44 0.30 2.34 1.54 1.62 0.48 4.46 1.76 11.14 4.94 4.35 2.32 61.4 26.2 171 6 0.51 0.12 1.66 0.56 1.71 0.44 5.25 1.57 10.91 3.21 3.55 1.09 75.8 31.8 167 6 0.52 0.29 1.59 0.48 1.98 0.56 7.04 4.36 9.45 4.20 3.54 1.21 76.2 43.4 168 5 2.22 1.36 3.08 1.03 2.38 0.68 5.19 1.79 13.74 5.34 6.51 2.57 74.6 14.8 123 6 2.40 1.68 2.53 0.58 3.34 1.01 4.98 1.16 14.13 3.95 5.91 1.58 77.9 20.6 165 5 0.54 0.36 2.09 0.68 2.36 0.78 6.44 2.39 11.48 4.78 4.35 2.21 74.4 28.5 NBS SRM 1566 — 8 0.71 0.24 2.86 0.19 1.07 0.52 49.50 4.00 1.72 0.35 nd — 772.0 53.0 edge of a baseline PPI value can be important for discerning the source of any change in contaminant levels in benthic animals. In a similar manner, the PCB value has been separated, by virtue of the use of capillary GC, into isometric groupings (Table 1). Again, there were dif- ferences in PCB composition between samples. For example, samples from stations 22, 28, 32, and 244 were largely comprised of tri-, tetra-, and hexachloro PCB isomers, while those from stations 107 and 27 contained significantly greater quantities of the tri- chlorobiphenyls. Aroclor 1016 and 1242 contain pro- portionately more of the C\l to Cl4 isomers while Aroclor 1254 contains a greater abundance of Cl4 to Cl6 isomers. In the future, it may be possible to ascribe the differences in the PCB composition in animals to possible sources through capillary GC/ECD measurements. Highest trace metal concentrations in A. islandica varied from metal to metal (Table 3); however, high- est mean Ag, Cr, Cu, and Pb concentrations were found in New York Bight (stations 26 and 29), while Ni and Zn were highest in the "Mud Patch" (stations 181 and 237) with the highest Cd values off Dela- ware (Table 3). Lowest concentrations, overall, were observed at midshelf stations off New Jersey and Maryland (with the exception of stations 167, 168, and 123 that could have been influenced by dump- ing at a nearby dumpsite) and station 379, on Georges Bank. Comparison of these data with those of Wenzloff et al. (1979), who analyzed metals in ocean quahogs from the New York Bight to an area off Chesapeake Bay, was attempted for temporal trends. Unfortunately, the Wenzloff et al. (1979) data were obtained from only foot muscle composites of 5 or 6 quahogs at each station, reported as means of all composites per half degree of latitude; hence, a direct comparison was not possible. The geogra- phic pattern, a decrease in metal concentrations with latitude believed present in the Middle Atlantic Bight 138 STEIMLE ET AL.: ORGANIC AND TRACE METALS IN OCEAN QUAHOG by Wenzloff et al. (1979), was not apparent from the present data or from the studies summarized in Table 4. Results of other studies involving whole body anal- ysis (Table 4) suggest that Cd, Ni, and Zn could also be high on Georges Bank; otherwise, the values pre- sented do not support any consistent latitudinal trends. Results indicate, however, on a local level, elevated trace metal levels were also usually associated with known areas of inputs, eg., waste dumpsites or ad- jacent to heavily industrialized coastal areas, such as the New York Bight apex (station 29), or natural depositional areas where trace metals from unknown sources are apparently accumulating, ag, the "Mud Patch" (stations 181, 237). The uptake and accumulation of trace metals by marine organisms are known to be affected by a number of variables. These variables include season, age, size, temperature, and interactive effects of several metals (Phillips 1977), and can be sources of some of the variability shown between the results of studies in the same area. Methodology is another source of variability between the results of each study, especially when intercalibrated results with standards are not available It is interesting to note that an expected close correlation between trace metal levels in the sediment and in A. islandica tissues was not evident in at least one study (Rey- nolds 1979), suggesting that the water and food or other suspended material could be the primary source of contaminants to this filter-feeding species. In conclusion, a set of measurements of several organic and seven trace metal contaminant levels in the commercially valuable ocean quahog have been obtained from a wide range of northwestern Atlan- tic locations. This set can be used as a base to moni- tor long-term changes in the assimilated levels and distributions of these compounds in this species and the risk to its health of future use as food. The levels found were well below the FDA seafood action limit, but elevated values were associated with impacted coastal habitats and possibly waste dumpsites. ACKNOWLEDGMENTS This study would not have been possible without the generous cooperation of the Northeast Fisheries Center's Resource Survey Group, specifically Thom- as Azarovitz, Charles Byrne, Donald Fletcher, Mal- colm Silverman, and others, who supplied us with the samples from annual clam assessment surveys. We also express our thanks to John B. Pearce and John O'Reilly for their support, and to Catherine Noonan, Maureen Montone, and Michele Cox for their assistance in preparing the manuscript. The paper was improved significantly from the comments of Donald Gadbois, Richard Greig, Carl Sindermann, Robert Reid, and unidentified reviewers. Funding for Table 4.— Comparison of mean trace metals levels {^g g 1 dry wt.) in Arctica islan- dica of the northwest Atlantic. Area and reference Ag Cd Cr Cu Ni Pb Zn Tissue type Georges Bank-Nantucket Sick (1978) 0.1 1.1 0.9 3.5 12.4 0.35 252 Whole body Erco (1978) 5.1 3.9 7.6 21.0 1.00 260 Whole body Payne et al. (1982) 4.5 1.7 5.4 27.0 3.50 150 Whole body Present study - stn. 379 0.8 1.4 3.1 10.3 3.5 4.1 62 Whole body Block Island Sound Steimle et al. (1976) 1.8 31 18.0 18.0 183 Whole body Rogerson and Galloway (1979)1 1.4 8.1 23 11.8 10.2 138 Whole body Present study - stn. 261 1.9 4.6 10 11.6 10.2 102 Whole body Southern Long Island Guarimo et al. (1979)1 3.0 5.6 17.4 27.9 14.1 122 ? Present study - stn. 189 2.5 1.9 8.8 18.7 3.3 128 Whole body New York Bight Wenzloff et al. (1979)1 15.8 3.5 <7.5 43.2 <5.0 9.8 107 Foot muscle Sick (1981) 0.7 7.9 5.3 "muscle" Present study - stn. 23, 26, 29, 32, 47 1.7 1.3 3.0 11.3 10.1 5.7 95 Whole body Off Delaware Reynolds (1979) 2.4 7.7 9.0 Whole body Present study - stn. 123, 167, 168 2.4 5.7 12.4 Whole body Chesapeak Bight Wenzloff et al. (1979)1 9.3 3.3 <8.0 34.6 <4.7 8.5 98 Foot muscle Present study - stn. 107 and south 1.0 2.2 2.3 5.4 11.6 4.7 71 Whole body 'Original wet weight data converted into dry weight by multiplying by 8. 139 FISHERY BULLETIN: VOL, 84, NO. 1 chemical analyses was provided, in part, by NOAA's Northeast Monitoring Program. LITERATURE CITED BOEHM, P. D. 1983a. Chemical contaminants in Northeastern United States marine sediments. U.S. Dep. Commer., NOAA Tech. Rep. NOS 99, 82 p. 1983b. Coupling of organic pollutants between the estuary and the continental shelf and the sediments and water column of the New York Bight region. Can. J. Fish. Aquat. Sci. 40(Suppl. 2):262-276. Boehm, P. D., J. E. Barak, D. L. Fiest, and A. A. Elskus. 1982. A chemical investigation of the transport and fate of petroleum hydrocarbons in littoral and benthic environments: The Thesis oil spill. Mar. Environ. Res. 6:157-188. DUINKER, J. C, M. T. J. HlLLEBRAND, AND J. P. BOON. 1983. Organochlorines in benthic invertebrates and sediments from the Dutch Wadden Sea; Identification of individual PCB components. Neth. J. Sea. Res. 17:19-38. DUINKER, J. C, M. T. J. HlLLEBRAND, K. H. PALMORK, AND S. WlLHELMSEN. 1980. An evaluation of existing methods for quantitation of polychlorinated biphenyls in environmental samples and sug- gestions for an improved method based on measurement of individual components. Bull. Environ. Contam. Toxicol. 25: 956-964. ERCO (Energy Resources Company). 1983. Levels of selected organic pollutants in soft clams, My a armaria, from the New Bedford area. Final Rep., Contract No. NA-81-FA-C-00013, to Natl. Mar. Fish. Serv., Northeast Fish. Cent, Highlands, NJ 07732, 16 p. Gadbois, D. 1982. PCBs and PAH in biota. In R. N. Reid, J. E. O'Reilly, and V. S. Zdanowicz (editors), Contaminants in New York Bight and Long Island Sound sediments and demersal species, and contaminant effects on benthos. Summer 1980, p. 37-46. US. Dep. Commer., NOAA Tech. Memo. NMFS-F/ NEC-16, Woods Hole, MA. Giam, C. S., H. S. Chan, and G. S. Neff. 1976. Concentrations and fluxes of phthlates, DDTs, and PCBs to the Gulf of Mexico. In H. L. Windom and R. A. Duce (editors), Marine pollutant transfer, p. 375-386. E. D. Heath and Co., Lexington, MA. Goldberg, E. D. 1978. The mussel watch. Environ. Conserv. 5:101-125. Guarino, C. F, M. D. Nelson, and S. S. Almeida. 1979. Ocean dispersal as an ultimate disposal method. J. Water Pollut. Control Fed. 51(4):773-782. Harris, R., R. Jolly, R. Huggett, and G. Grant. 1977. Trace metals. In Chemical and Biological Benchmark Studies on the Middle Atlantic Outer Continental Shelf, Ch. 8. Final Rep. to Bur. Land Manage, U.S. Dep. Inter., by Va. Inst. Mar. Sci., Gloucester Point, VA 23062. MacLeod, W. D, L. S. Ramos, A. J. Friedman, D. G. Burrows, P. G. Prohaska, D. L. Fisher, and D. W. Brown. 1981. Analysis of residual chlorinated hydrocarbons, aromatic hydrocarbons, and related compounds in selected sources, sinks, and biota of New York Bight. U.S. Dep. Commer., NOAA Tech. Memo. OMPA-6, Boulder, CO, 128 p. Merrill, A. S., J. L. Chamberlin, and J. W. Ropes. 1969. Ocean quahog fishery. In F. E. Firth (editor), The en- cyclopedia of marine resources, p. 125-129. Van Nostrand Reinhold Co., NY. O'Connor, J. M., J. B. Klotz, and T. J. Kneip. 1982. Sources, sinks, and distribution of organic contaminants in the New York Bight ecosystem. In G. F. Mayer (editor), Ecological stress and the New York Bight: Science and management, p. 631-654. Estuarine Research Foundation, Columbia, SO Payne, J. R., J. L. Lambach, R. E. Jordan, G. D. McNabb, Jr., R. R. Sims, A. Abasumava, J. G. Sutton, D. Generro, S. Gagner, and R. F. Shores. 1982. Georges Bank monitoring program, analysis of hydro- carbons in bottom sediments and analysis of hydrocarbons and trace metals in benthic fauna. Final Rep. July 1981-May 1982 to U.S. Dep. Inter., Miner. Manage Serv., Wash., DC, 189 p. Science Application, Inc., La Jolla, CA 92039. Phillips, D. J. H. 1977. The use of biological indicator organisms to monitor trace metal pollution in marine and estuarine environments— a review. Environ. Pollut. 13:281-317. Reynolds, B. H. 1979. Trace metal monitoring at two ocean disposal sites. U.S. Environ. Prot. Agency, ERL, Narragansett, RI, EPA 600/3-79-037, 64 p. Rogerson, P. F, and W. B Galloway. 1979. Preparation and characterization of a marine reference material for trace element determinations. In F. Jacoff (editor), Advances in marine research— Proceedings of a sym- posium, p. 1-8. U.S. Environ. Prot. Agency, ERL, Narragan- sett, RI, EPA 600/9-79-035. Ropes, J. W. 1979. Biology and distribution of surf clams (Spisula solidis- sima) and ocean quahogs (Arctica islandica) off the north- east coast of the United States. In Proceedings of the Northeast Clam Industries: Management for the future P 47-66. Mass. Coop. Ext. Serv., Univ. Mass., Amherst, MA. Sick, L. V. 1978. Trace metal analysis of zooplankton and macrobenthic tissues from Georges Bank. Sea Grant Rep. CMS-1-78, Coll. Mar. Stud., Univ. Del., Lewes, DE, 57 p. 1981. Seasonal and geographical dynamics of trace metal com- position in the benthic community of the New York Bight. Final Rep. to NOAA, MESA, New York Bight Proj., Stony Brook, NY, 128 p. Steimle, F. W, C. J. Byrne, R. N. Reid, and T. R. Azarovitz. 1976. Hydrology, sediments, macrofauna and demersal finfish of an alternate disposal site (East Hole in Block Island Sound) for the Thames River (Conn.) dredging project. Final Rep. to U.S. Navy, New London, CT, Informal Rep. #110, 63 p. NOAA, NMFS, MACFC, Sandy Hook Laboratory, Highlands, NJ. Thompson, I., D. S. Jones, and D. Dreibelbis. 1980. Annual internal growth banding and life history of the ocean quahog Arctica islandica (Mollusca:Bivalvia). Mar. Biol. (Berl.) 57:25-34. Warner, J. S. 1976. Determination of aliphatic and aromatic hydrocarbons in marine organisms. Anal. Chem. 48:578-583. Weaver, G. 1982. Status report on PCB pollution in New Bedford, Massa- chusetts. Mass. Executive Off. Environ. Aff., Boston, MA, 69 p. Wenzloff, D. R., R. A. Greig, A. S. Merrill, and J. W. Ropes. 1979. A survey of heavy metals in the surf clam, Spisula solidissima, and the ocean quahog, Arctica islandica, of the Mid-Atlantic coast of the United States. Fish. Bull., U.S. 77:280-285. 140 AN ECOLOGICAL SURVEY AND COMPARISON OF BOTTOM FISH RESOURCE ASSESSMENTS (SUBMERSIBLE VERSUS HANDLINE FISHING) AT JOHNSTON ATOLL Stephen Ralston,1 Reginald M. Gooding,1 and Gerald M. Ludwig2 ABSTRACT The deep slope (100-365 m) environment at Johnson Atoll in the central Pacific was surveyed with a submer- sible and the standing crop of commercially important bottom fishes (i.e, lutjanids, serranids, and carangids) estimated by visual quadrat censusing. Results are compared with an assessment made by hook-and-line fishing. Overall, 69 species of fish were recorded from the submersible and 10 from fishing. Well over half of the sightings from the submersible were new locality records. Bottom fish abundance estimates (fish/hec- tare and fish/line-hour) varied by site but agreed broadly with one another. Tbgether they are used to estimate catchability (0.0215 hectare/line-hour), which is shown to vary through the day. Bottom fish were contagiously dispersed along both vertical and horizontal dimensions, with increased numbers of the snapper Pristipomoides filamentosus in upcurrent localities. On a finer scale this species and Etelis coruscans were aggregated near underwater promontories and headlands, but at different depths. Numerous observations concerning the deep slope environment of this central Pacific Ocean atoll are included. Perhaps the most widespread precept in fisheries today is the supposition that catch rate is propor- tional to stock abundance (Gulland 1974; Ricker 1975; Pitcher and Hart 1982). Even so, there are numerous studies which demonstrate exceptions to this assumption (see for example MacCall 1976; Ban- nerot and Austin 1983). A departure from linearity in the relationship of these two variables reflects varying catchability. This variation may be due to schooling behavior, gear saturation, or any number of other factors which affect catch per unit effort (CPUE) in addition to stock abundance (Rothschild 1977). It is often difficult, if not impossible, to evaluate the validity of the linearity assumption in most practical situations. A multiple approach to stock assessment has often been suggested as a means of circumventing this problem, including the use of hydroacoustics (Barans and Holliday 1983; Thorne 1983), underwater television-diver surveys (Powles and Barans 1980), and submersibles (Uz- mann et al. 1977) to corroborate CPUE data. Con- sistency in results among a set of independent assessment techniques is necessary for validation and verification of data. 'Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, Honolulu, HI 96812. 2U.S. Fish and Wildlife Service, Honolulu, HI 96850; present ad- dress: Florida Fishery Research Station, U.S. Fish and Wildlife Ser- vice, P.O. Box 1669, Homestead, FL 33030. Submersibles in particular have also proven useful in studying the distribution of fishes in various deep- water habitats (Brock and Chamberlain 1968; Stras- burg et al. 1968; Colin 1974; Shipp and Hopkins 1978), in identifying nursery grounds of commercial- ly important rockfish species (Carlson and Straty 1981), and in assessing the effectiveness of baited longline gear (High 1980; Grimes et al. 1982). In many situations submersibles provide an ideal means of independent assessment (Uzmann et al. 1977) if questions concerning bias in visual surveys can be adequately addressed (Colton and Alevizon 1981; Sale and Douglas 1981; Brock 1982). The purpose of this study was to examine the distribution and abundance of tropical deep slope bottom fishes (i.e, lutjanids, serranids, and carangids) at Johnston Atoll in the central Pacific Ocean with a research submersible and to compare the results with an assessment made by fishing. This compari- son provides not only a basis for testing the validity of a CPUE statistic, but also for estimating the catchability coefficient. Both are important issues because of the widespread use of hook-and-line catch and effort statistics in resource assessments of bot- tom fish stocks worldwide (Moffitt 1980; Ralston 1980; Ivo and Hanson 1982; Ralston and Polovina 1982; Munro 1983; Forster 1984). Of special interest was determining the relationship between CPUE and visual estimates of bottom fish standing stock. Manuscript accepted April 1985. UlCirfDV DITI T ITTTTM. VOI O A M(l 1 1QOC 141 FISHERY BULLETIN: VOL. 84, NO. 1 In addition, a variety of observations made from the submersible substantially improved our understand- ing of factors controlling the distribution and abun- dance of the entire deep slope fauna at Johnston Atoll. DESCRIPTION OF THE STUDY AREA A National Wildlife Refuge since 1926, Johnston Atoll is located 1,250 km southwest of Oahu, HI. The atoll's physical environment has been reviewed by Amerson and Shelton (1976) and is summarized here. Located between lat. 16°40'-16°47'N and long. 169°24'-169°34'W (Fig. 1), Johnston Atoll lies in the North Pacific central water mass, where salinities range from 34.8 to 35.3°/00. Surface water temper- atures show little seasonality, ranging from 25° to 27 °C. The atoll is directly in the path of the wester- ly flowing North Equatorial Current, with surface currents typically 0.5 kn (0.25 m/s). Deeper layers flow smoothly past the atoll, but an island wake forms in lee surface waters, with effects evident up to 600 km downstream (Barkley 1972). The atoll is composed of a coral platform, encom- passing over 130 km2 of reef under water <30 m deep. A narrow lagoon lies between the northwest barrier reef and Johnston and Sand Islands to the southeast (Fig. 1). The atoll is unusual in that the main outer reef extends only about one quarter of the way around its perimeter (Fig. 1). A large por- tion of the atoll lies exposed to prevailing easterly weather conditions without benefit of barrier reef protection. Evidence suggests that subsidence and tilting of the reef platform to the southeast created this unusual condition. The climate is tropical marine, i.a, there is little seasonal variation in temperature and windspeed, but substantial variation in rainfall. A 4-mo "winter" season extends from December to March, when temperatures drop slightly, winds become more vari- able, and precipitation increases. The mean annual air temperature is 26.3 °C, with a daily range of 4.0° -4.5° C. Daily maximum and minimum temper- 169°30' W I ^ 1 16°45'N A-J Makalii dive sites 1-6 Cromwell fishing stations Figure 1— Map of Johnston Atoll. The lines encircling the atoll are isobaths of constant depth (fathoms). The four shaded areas at the upper left are emergent lands (Johnston, Akau, Hikina, and Sand Islands). Letters (A-J) indicate the 10 dive sites of the Makalii during the study. Numbers (1-6) indicate fishing stations of the Townsend Cromwell. 142 RALSTON ET AL.: BOTTOM FISH RESOURCE AT JOHNSTON ATOLL atures vary little throughout the year, as do sea sur- face temperatures, which are in near equilibrium with the air. Strong easterly trade winds prevail all year but increase during the summer period. Annual mean wind speed at Johnston Island is 13 kn (7.5 m/s) and monthly means range from 11 to 14 kn (5.5-7.0 m/s). METHODS Makalii The Makalii is operated by the National Undersea .esearch Laboratory at the University of Hawaii. It is a two-man, battery powered, 1-atmosphere sub- mersible which is 4.8 m long, with a pressurized cap- sule 1.5 m in diameter. When carrying a pilot and one observer, its normal operating speeds range from 1 to 3 kn (0.5-1.5 m/s). Maximum dive duration is 4-5 h and depth capability is 365 m. Equipment carried in this study included hydraulic manipulator, internal and external color video cameras, 2 video monitors, video recorder, video flood lights, Photo- sea3 35 mm still camera with strobe, current and temperature meters, and a dictaphone tape recorder. In addition, the Makalii is equipped with an environ- mental monitoring system for continuous recording of temperature, salinity, conductivity, oxygen, solar radiation, and depth. All three authors participated as observers dur- ing a series of dives at Johnston Atoll over the 2-wk period between 22 September and 5 October 1983. Once on station, a launch-recovery-transport plat- form was submerged to 20 m and divers released the Makalii, usually in 120 m of water. The submersible descended until encountering the bottom and locating the atoll's shelf break. Observations made on fishes during the dives were voice and video re- corded for later analysis. Slope angle was periodi- cally measured with a hand-held inclinometer. Visual estimates of the density of commercially im- portant bottom fishes (sensu Ralston and Polovina 1982) were made by a series of "quadrat" samples. These fishes included Cookeolus boops, Epinephelus quernus, Aphareus furcatus, A. rutilans, Etelis car- bunculus, E. coruscans, Pristipomoides auricilla, P. filamentosus, P. zonatus, Carangoides orthogram- mus, Caranx lugubris, Seriola dumerili, and Pon- tinns macrocephalus. During quadrat sampling the observer would look out his port and count the total number of bottom 3Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. fish, without regard to species, over an area of the bottom judged to be 30 m2. Quadrat areas always lay on the oblique planar surface of the slope face and were away from the immediate vicinity of the submersible A sampling period consisted of four counts systematically performed, one every 15 s. To the extent possible, each count was made at an in- stant in time. All bottom fish seen in the water column above the sample area were included in counts. The submersible progressed stepwise down the slope (100-365 m) in a clockwise direction around the atoll, with the observer's starboard port always oriented to the slope face. Upon reaching the Makalii' s depth limit, a slow stepwise ascent would begin to 100 m, where the dive would end. Descents generally lasted 2.5 h and ascents 1.5 h. Thus the entire vertical distribution of the deep slope was sampled more or less equally (i.e, observations were not concentrated in any particular depth zone). Townsend Cromwell The National Oceanic and Atmospheric Adminis- tration's (NOAA) RV Townsend Cromwell is 50 m long and when rigged for bottom handline fishing carries four hydraulic fishing gurdies (Charlin motors and Pacific King fishing reels), each with 365 m of braided prestretched 90 kg Dacron line The terminal rig is composed of four No. 28 Tonkichi round fishing hooks and a 2 kg weight. Stripped squid was used for bait and fishing was conducted only during the day. The vessel spent 3 d (3-5 November 1983) at John- ston Atoll sampling deep slope bottom fish by drift fishing. After wind and current directions had been determined, the vessel was positioned over the desired depth and fishing lines were dropped. Fish- ing continued until the vessel drifted over an un- suitable water depth, when lines were retrieved and the Townsend Cromwell repositioned. Single drifts were the fundamental sampling unit by which catch and effort statistics were summarized. Six fishing stations were occupied (Fig. 1), one during the morn- ing and afternoon of each day. Fork length to the nearest millimeter and depth of capture were re- corded for all fish landed. RESULTS Makalii Ten dives were completed at Johnston Atoll (Fig. 1). Due to precipitous dropoffs which occur through- 143 out the study area (100-365 m), the length of the atoll's 183 m (100 fathom) isobath (64 km) provides a convenient measure of total deep slope habitat (Ralston and Polovina 1982). The average point-to- point distance covered by the submersible during one 4-h dive was 2.27 km (s = 0.56 km). An aggregate 22.7 km were thus surveyed during this study, com- prising 35% of the deep slope habitat at the atoll. Temperature Ambient temperature and depth were recorded often during dives, from which temperature-depth profiles were later constructed. The results are sum- marized in Figure 2. The solid line represents me- dian temperatures at depth, with the shaded area encompassing the range of temperatures observed among all 10 dives. Surface water temperature was typically 27°C and the mixed layer 100 m deep. A second weak thermocline was found around 240 m. Although its depth varied somewhat (220-245 m), it was present around the entire atoll, i.e., both wind- ward and leeward exposures, and was observed as a shimmering layer below the submersible as it descended. This effect is believed due to refraction of light passing through variable density water, a result of the thermocline in association with a de- crease in salinity.4 Ambient water temperature usual- ly had dropped to 8.5 °C at a depth of 350 m. Slope Angle The relationship between the bottom's slope and depth was also measured. These data were sum- marized after each dive and bottom contours plot- ted. Overall, there was little variation in slope angle around the atoll, i.e., the general pattern was one of uniformity at all sites visited. Figure 3 presents pool- ed results for all slope angle-depth determinations. In the figure, horizontal and vertical scales are equal and the composite contour of the bottom (100-365 m) at Johnston Atoll is shown in profile. The slope was stratified into three 50-fathom depth zones for later analysis.5 The slope angle between 50 and 100 fathoms averages Q1 = 25° (Table 1). Similarly, 02 = 47° and 03 = 59°. There is a definite trend at Johnston Atoll for the slope to steepen with in- 4E. Chave, Hawaii Undersea Research Laboratory, University of Hawaii, Honolulu, HI 96822, pers. commun. June 1984. Stratification of depth into zones was performed using units of fathoms (1 fathom = 1.83 m) because nautical charts, hydrogra- phic surveys, and fathometers are so measured. For the sake of brevity and clarity, isobaths and depth strata will henceforth be given only in this unit of measure FISHERY BULLETIN: VOL. 84, NO. 1 Temperature - °C 10 15 20 25 30 50 100 150 E i Q. 0 Q 200 250 300 350 Figure 2.— The pooled relationship (w = 10) between temperature and water depth at Johnston Atoll. Solid line = median values; shaded area = range of values. Table 1.— Total habitat areas stratified by depth zones at Johnston Atoll. Digitized Depth horizontal Oblique planar stratum planar areas Slope habitat areas (fathoms) (ha) angle (ha) Emergent lands 305(1%) — 0-10 15,012(60%) — — 10-50 6,123(24%) — — 50-100 1 ,624 (7%) 25° 1,785 100-150 964 (4%) 47° 1,418 150-200 1 ,020 (4%) 59° 1,962 Total 25,048(100%) — 5,165 creasing depth, at least between 100 and 365 m. In the shallowest regions surveyed (<125 m) the bottom was a monotonous sandy plain in the shore- ward direction, but at 125 m it began to slope steeply 144 RALSTON ET AL.: BOTTOM FISH RESOURCE AT JOHNSTON ATOLL downward. Although not easily seen in the figure, a small but prominant ledge 5-10 m high encircled the atoll between 130 and 140 m. Somewhat deeper, between 180 and 275 m, the bottom was uniform in slope and its surface relatively smooth and devoid of features. Slope angles approached the vertical at most sites in the 300-350 m depth range, with over- hanging caves formed by subaerial dissolution.6 At the deepest points visited (360 m) the bottom became less precipitous, and in some areas a sediment-laden terrace had formed along the base of the deep dropoff. Based on estimates of slope angle, existing charts, and a hydrographic survey by the Townsend Crom- well, habitat areas for the three depth zones were determined. The positions of the 10 and 100-fathom isobaths were already known, but they were refined and the locations of the 50-, 150-, and 200-fathom isobaths estimated. Figure 1 is a simplified repre- sentation of a much larger chart which was digital- ly analyzed to determine the horizontal (i.e., level) areas bounded by isobaths (Table 1). The results show that emergent lands (Johnston, Akau, Hikina, and Sand Islands) account for only 1% (305 ha) of the level planar area of the atoll. The largest area (60%) lies between sea level and 10 fathoms. The 6Keating, B. H. Geologic history and evolution of Johnston Island: Submersible dive results. Manuscr. in prep. University of Hawaii, Honolulu, HI 96822. total horizontal extent of the atoll is about 25,000 ha. These results can be misleading, however, because a vertical slope provides no horizontal habitat area, and yet both reef fish diversity and standing crop are known to be positively correlated with topo- graphic relief (Luckhurst and Luckhurst 1978; Glad- felter et al. 1980; Carpenter et al. 1981). At John- ston Atoll the structural complexity of the sub- stratum frequently increased with slope angle. A better estimate of total habitat is the area of bot- tom irrespective of slope angle, estimated by dividing the horizontal planar area of a depth stratum by the cosine of the slope angle within it. This adjustment almost doubles the estimate of total habitat area in the 150-200 fathom zone, simply due to the precipi- tous dropoff found there. A composite 5,165 ha of habitat occurs between 50 and 200 fathoms. General Observations While this study focused primarily on the deep- water ichthyofauna of Johnston Atoll, many obser- vations were made on the oceanographic, geologic, and biotic characteristics of the study area. These are briefly recounted here. Currents running in directions parallel to the slope were frequently encountered. They were generally weak and did not exceed 0.3 kn (0.15 m/s). They sometimes exhibited reversals with depth. During 100- 150- i 200- a s a 250- 300- 350- — i 50 100 E o x: O N 150 200 Figure 3— Composite reconstruction of the deep slope at Johnston Atoll. Horizontal and vertical scales equal. Average slope angles (0) were measured for each of three 50-fathom depth strata. 145 FISHERY BULLETIN: VOL. 84, NO. 1 dive F, for example (Fig. 1), a 0.2 kn (0.10 m/s) cur- rent was observed at 125 m running south (i.e, counterclockwise when viewed from above). There was no current between 180 and 275 m. At 300 m, however, a 0.1 kn (0.05 m/s) current was observed, traveling in a northerly direction (i.e, clockwise). A similar depth-related current reversal was observed during dive C, although on this occasion the shallower current (170 m) ran clockwise and the deeper one (290 m) counterclockwise. In contrast, a weak downslope current (0.1 kn or 0.05 m/s) was observed but once (dive E at 305 m). No upwelling currents were encountered. Geologically, the deep slope of Johnston Atoll was grossly similar at all points visited. The low escarp- ment at 130 m was most likely due to erosion of an ancient limestone reef. This feature was character- ized by mounds of coral rubble, boulders, small undercut caves, and a profusion of fishes. Below it the slope angle was remarkably uniform, with low topographic relief. The bottom was still composed of limestone and showed severe biological and chem- ical weathering (i.e, dissolution) along the slope gra- dient, being pitted and striated with numerous shallow depressions. Few sediments or boulders were observed. At a depth of 240 m topographic relief in- creased, as large slab boulders became increasingly prominent. Subaerial dissolution had produced low shallow limestone caves, and fine sediments were more common. Between 290 and 335 m the slope was very steep, with a well-developed system of sharp ridges and deep erosional channels. The substratum had the superficial appearance of dark basalt but was composed of thin manganese crusts overlying an- cient limestone reef materials (Keating see footnote 6). Fine sediments spilled down the channels in the slope and piled up at the base of the deep dropoff (350 m). More limestone boulders were arrayed along this deep terrace and fine sediments covered the bottom. As expected, few fleshy macroalgae were seen. The only algae encountered regularly were two coral- lines, Halimeda sp. and an unidentified crustose species. The former occurred in small scattered clumps between 100 and 200 m, with loose remnant exoskeletal "sands" found in sediment pockets as deep as 290 m. The crustose form was abundant be- tween 150 and 250 m where it covered much of the slope face Otherwise, an unidentified species of brown algae seen on dive H between 150 and 250 m was the only other algae seen. A more detailed description of the algal biota at Johnston Atoll is in preparation.7 In contrast to the depauperate flora, the inverte- brate fauna was rich. Listed here are those forms seen often enough to constitute indicator species for particular depth strata. In addition to these a great many others were observed and photographed. In the Cnidaria, three stoney corals were especially plentiful: Leptoseris hawaiiensis (115-165 m), Stylaster sp. (135-245 m), and Madracis sp. (140-200 m). Several species of black corals (Order An- tipatharia) were also common. Of the crustaceans, a single large Panulirus marginatum, previously known only from one specimen (Brock 1973), was observed in a small hole during dive A at 122 m, and at least two types of galatheid crab were very abundant in small holes pitting the reef slope between 230 and 350 m. In deep water the remain- ing attached valves of dead rock oysters were seen in patches along the base of the deep dropoff (350 m), as was an unidentified species of solitary tunicate (335-365 m). Echinoids were particularly abundant immediately below the shelf break; eg, Diadema cf. savignyi (110-170 m), Chondrocidaris gigantea (120-160 m), and heart urchins (Brissidae, 130-200 m). Other than galatheid crabs, the 220-310 m zone was largely barren and devoid of mega- benthos. Ichthyofauna A total of 69 fish species in 29 families were ob- served during Makalii dives (Table 2). Overall, the proportional representation of different families was similar to that of the shallow water community (Gosline 1955; Randall et al. in press), although the representation of genera was grossly different. Ser- ranid species were most numerous with nine species observed (eight in the anthiine subfamily). Lutjanids were also abundant (eight species), but no members of the ubiquitous genus Lutjanus were seen. Forty of the species listed in Table 2 (58%) are new records for Johnston Atoll (Randall et al. in press). Photo- graphs of several fishes observed during dives are presented in Figure 4. An indication of species' depth distributions is given in Table 2. Because no observations were made in <100 m, upper limits can be misleading. This is particularly true of shallow-water species which penetrated to the 135 m escarpment but not beyond, including: Triaenodon obesus, Parapercis schau- inslandi, Aphareus furcatus, Chromis verater, Paru- peneus cyclostomus, P. multifasciatus, Forcipiger flavissimus, Holacanthus arcuatus, Bodianus bilu- 7C. Agegian, University of Hawaii, Honolulu, HI 96822, pers. com- mun. June 1984. 146 RALSTON ET AL.: BOTTOM FISH RESOURCE AT JOHNSTON ATOLL Table 2— Fishes encountered during dives (100-365 m) of the Makalii at Johnston Atoll. Included for each species are the minimum and maximum depths (m) of observation as well as the median and range of the depth distribution. Under the sighting column a value of 1 indicates a species was seen repeatedly (>5 times) during each dive of the submersible, 2 means the species was occasionally seen on each dive (<5 times), 3 signifies sightings on most dives but not all (i.e., species seen on several occasions), and 4 indicates rarity (see only once or twice during all dives). An asterisk to the left of a species name signifies a new record for Johnston Atoll (Randall et al. in press). Median Median Family-species Min-max (range) Sighting Family-species Min-max (range) Sighting Carcharhinidae Carangidae Carcharhinus amblyrhynchos 90-275 185(185) 1 Carangoides orthogrammus 105-170 135(65) 2 Carcharhinus sp. Caranx lugubris 105-355 190(250) 1 (probably galapagensis) 225-250 225(25) 3 C. melampygus 130-230 135(100) 2 Triaenodon obesus 120 4 Decapterus sp. 100 4 Mobulidae *Elagatis bipinnulata 90-150 120(60) 3 Manta sp. 120 4 'Seriola dumerili 120-335 215(215) 1 Muraenidae Apogonidae 'Gymnothorax berndti 220-260 260(40) 3 'Epigonus sp. 330-365 355(35) 2 *G. nudivomer 120-205 1 79(85) 2 Pomacentridae *G. nuttingi 185-300 250(115) 3 Chromis verater 120-140 130(20) 3 Ophichthidae Mullidae Myrichthys maculosus 150-215 185(65) 4 Parupeneus cyclostomus 125 4 Synodontidae P. multifasciatus 125 4 Unidentified synodontid 240 4 Chaetodontidae Holocentridae 'Chaetodon modestus 125-255 190(130) 2 'Myripristis chryseres 135-240 155(105) 2 'C. tinkeri 105-160 145(55) 3 "Neoniphon aurolineatus 150 4 Forcipiger flavissimus 125-145 130(20) 4 'Pristilepis oligolepis 165-345 230(180) 3 'Heniochus diphreutes 120-215 135(95) 2 Ophidiidae Pomacanthidae Brotula sp. Geniacanthus sp. 150 4 {multibarbata or townsendi) 230 4 * Holacanthus arcuatus 130-150 135(20) 3 Priacanthidae Labridae 'Cookeolus boops 165-260 220(95) 1 Bodianus bilunulatus 130-135 130(5) 3 Serranidae Cheilinus unifasciatus 120 4 'Anthias fuscinus 135-280 215(145) 1 'Polylepion russelli 245-280 275(35) 3 'A. ventralis 105 4 Acanthuridae Callanthias sp. 240-330 285(90) 4 "Acanthurus dussumieri 130 4 'Epinephelus quemus 135-350 230(215) 1 *Naso hexacanthus 120-165 150(45) 2 'Grammatonotus laysanus 310-350 335(40) 3 *Naso sp. 120-175 135(55) 1 'Holanthias elizabethae 155-260 230(105) 1 Zanclidae *H. fuscipinnis 160-215 170(55) 1 Zanclus comutus 125 4 Luzonichthys sp. Scorpaenidae (perhaps earlei) 105 4 'Pontinus macrocephalus 200-365 305(165) 2 'Plectranthias helenae 215-220 215(5) 3 "Scorpaena colorata 272 4 Mugiloididae Scorpaena sp. 225-355 290(130) 2 'Parapercis roseoviridis 215-270 245(55) 2 Triglidae "P. schauinslandi 105-170 145(65) 1 'Satyrichthys engyceros 355-365 365(10) 4 Lutjanidae Bothidae Aphareus furcatus 105-145 135(40) 2 Bothus mancus 270-350 310(80) 4 "A. rutilans 190-250 220(60) 3 Balistidae 'Etelis carbunculus 245-365 310(120) 3 'Sufflamen fraenatus 105-170 140(65) 1 *£. coruscans 250-355 270(105) 3 Xanthichthys auromarginatus 115-155 135(40) 1 * Pristipomoides auricilla 215-250 230(35) 3 Monacanthidae 'P. filamentosus 120-260 205(140) 3 Unidentified monacanthid 125 4 'P. zonatus 205-295 240(90) 1 Tetraodontidae * Symphysanodon maunaloae 230-365 300(135) 1 "Canthigaster sp. Emmelichthyidae (likely inframacula) 260-270 265(10) 4 'Erythrocles scintillans 295-320 300(25) 4 Unidentified tetraodontid Ostraciidae Ostracion sp. Diodontidae Diodon hystrix 135-150 145(15) 135 135 4 4 4 nulatus, Acanthurus dussumieri, Zanclus comutus, Xanthichthys auromarginatus, and Diodon hystrix. These fishes accounted for an increase in diversity at the 135 m dropoff. Similarly, due to the submer- sible's 365 m depth limit, lower bounds for some species are likely in error (eg., Symphysanodon mau- naloae, Epigonus sp., Pontinus macrocephalus, and Satyrichthys engyceros). Nonetheless, due to the large depth range sampled (100-365 m), the data still provide useful estimates of the depth distributions for most of the species listed. The data suggest that large species have great 147 FISHERY BULLETIN: VOL. 84, NO. 1 148 RALSTON ET AL.: BOTTOM FISH RESOURCE AT JOHNSTON ATOLL depth ranges. For example, all species with depth ranges exceeding 200 m are large (i.e, Caranx lugu- bris, Epinephelus quernus, and Seriola dumerili). Moreover, among extensively observed species, a significant Spearman correlation exists between ranked average weight and depth range (rs = 0.52, df = 25, P < 0.01). This finding should be viewed with caution because of potential biases in depth distributions (see above). The last column in Table 2 gives sighting scores for all species. Those assigned a value of 1 indicate species dominating the deep slope fish community in terms of species sightings. Note that some species were seen infrequently, but when encountered they were observed in large numbers (eg, Elagatis bipin- nulata, Fig. 4). Similarly, Pristipomoides filamen- tosus was not seen on every dive and was thus as- signed an abundance score of 3. In spite of this, when seen, it was abundant and it was the most frequent- ly caught while fishing (see next section). Sighting scores therefore do not indicate relative species' con- tributions to total standing crop biomass of the deep slope fish fauna. Quadrat Sampling A total of 974 quadrat sample counts were made during the 10 submersible dives. No attempt was made to estimate abundance separately for each species. Rather, the total number of bottom fish was recorded, regardless of species composition. Al- though severely reducing the detail of the data base, this did have the desirable effect of averaging biases due to attraction or repulsion of fishes to and from the Makalii. It was evident, for example, that some species were attracted to the submersible and follow- ed it about (e.g., Seriola dumerili and Caranx lugu- bris), whereas others were repelled and actively avoided the submersible's lights (eg., Pristipomoides filamentosus and Etelis coruscans). Still others did not seem to be greatly influenced (eg., Cookeolus boops, Epinephelus quernus, Pristipomoides zonatus, and Pontinus macrocephalus). By pooling species quadrat counts, the abundance of some species was overestimated, some underestimated, and some estimated without bias. Due to averaging, we believe that pooled counts provide the best available Figure 4— Johnston Atoll deep slope fishes. A. Caranx lugubris with wire coral; B. Epinephelus quernus peering out of cave; C. Seriola dumerili (foreground) and Caranx lugubris (background); D. school of Elagatis bipinnulata with Carangoides orthogrammus (above); E. Heniochus diphreutes with black coral; and F. aggrega- tion of Myripristis chryseres and Neoniphon aurolineatus. estimates of total bottom fish density along the deep slope of Johnston Atoll. Some 367 bottom fish were counted in quadrat samples, resulting in a mean encounter rate of 0.38 fish/quadrat. The data were fitted to the Poisson distribution to ascertain the dispersion pattern. A chi-square goodness of fit test yielded x2 = 325.32, df = 3, P « 0.005, demonstrating nonrandom dis- persion. The variance to mean ratio calculated from the frequency distribution of bottom fish/quadrat observations was 4.64 and was significantly greater than 1 (P « 0.005), indicating strong contagion. One of the principal explanations for this result is shoaling by Pristipomoides filamentosus and Ete- lis coruscans. Both are large species, which formed aggregations of up to 100 individuals well off the bot- tom (20 m) in the vicinity of underwater headlands and promontories. These monospecific groups ap- peared to feed in open water on plankton, consis- tent with previous dietary studies of P. filamentosus (Kami 1973; Ralston8). When either was observed, there was an increased likelihood of encountering conspecifics. As a consequence 10 or more P. fila- mentosus were seen in one quadrat on 7 occasions. Another factor contributing to clumping was non- random distribution with depth (Fig. 5). This figure presents the relationship between mean number of bottom fish per count and depth (vertical bars = standard errors). Note the two abundance peaks, the first at about 170 m and the second at 250 m. The former was due primarily to large numbers of Caranx lugubris and P. filamentosus. The location of the second peak was just below the second thermocline and was largely the result of local in- creases in numbers of Epinephelus quernus and P. zonatus. The mean numbers of bottom fish per quadrat, stratified into 50-fathom depth intervals, are also shown in Figure 5 (i.e, 0.57, 0.47, and 0.06 fish/count). These data were converted to densities (1 quadrat = 0.003 ha) such that from 50 to 100 fathoms an average of 190 bottom fish are estimated to occur per hectare of habitat. Similarly, in the two deeper strata, estimated densities of 156 and 20 bottom fish/ha occur. Given estimates of bottom fish density and depth- specific estimates of total available habitat (Table 1), estimates of the total standing crop of bottom fishes at Johnston Atoll indicate that about 339,000 fish occurred in the 50-100 fathom zone, 221,000 between "Ralston, S. Unpubl. data. Southwest Fisheries Center Hono- lulu Laboratory, National Marine Fisheries Service, NOAA, Hono- lulu, HI 96812. 149 1.1-1 1.0- 0.9- ~ 0.8 •o ^ 0.7 i °6 o o 0.5 n o 0.4 c a T3 % 0.3 < 0.2 0.1 H 0.0 FISHERY BULLETIN: VOL. 84, NO. 1 200 50 — I — 100 — I — 150 T 200 Depth 250 — I — 300 350 400 [ml Figure 5— The abundance of bottom fish (see text) in relation to depth. Solid line represents fish densities with changing depth (measured in meters or fathoms). Error bars are standard errors of means. Three 50-fathom depth zones are indicated, and mean fish densities within these are shown as circled points. 100 and 150 fathoms, and only 39,000 in the deep- est (150-200 fathom) zone. Roughly 600,000 com- mercially exploitable bottom fish are estimated to comprise the deep-sea hook-and-line resource at Johnston Atoll. Because the fish are spread over a total habitat of 5,165 ha (Table 1), this corresponds to average densities of 118 bottom fish/ha. Townsend Cromwell Anywhere from 2 to 4 lines were deployed while fishing, resulting in an aggregate 41.8 line-h of fishing effort spread over 23 vessel drifts. A catch of 133 fishes (Table 3) produced an overall CPUE of 3.18 fish/line-h. Another 12 fish were hooked but lost to sharks before landing. All species caught while fishing were observed from the submersible with the exception of the bramid, Eumegistus illustris. Deep- water lutjanids predominated (69%), but substantial numbers of serranids (22%) and carangids (8%) were caught, a composition typical of tropical deep slope fisheries worldwide (Talbot 1960; Ralston and Polo- vina 1982; Munro 1983; Forster 1984). Species Composition By Location Examination of catch data suggested a difference in species composition between upcurrent (sites 5 Table 3. — Species composition of the bottom fish catch from the Townsend Cromwell at Johnston Atoll. Family-species Catch Percent Average size (cm FL) Lutjanidae (snappers) Pristipomoides filamentosus P. zonatus P. auricilla Etelis carbunculus E. coruscans 43 35 5 5 4 32 26 4 4 3 54.4 40.8 34.6 51.2 72.7 Subtotal 92 69 Serranidae (groupers) Epinephelus quernus 29 22 69.8 Carangidae (jacks) Caranx lugubris Carangoides orthogrammus Seriola dumerili 7 2 2 5 2 2 48.1 43.5 79.5 Subtotal 11 9 Bramidae (pomfrets) Eumegistus illustris 1 1 70.3 Grand total 133 101 and 6) and downcurrent (sites 1-4) locations (Fig. 1). Landings were pooled into these two classes, and al- so by species category into Pristipomoides filamen- tosus, P. zonatus, Epinephelus quernus, and "others". The resulting 2x4 contingency table showed a lack of statistical independence between locations and species (x2 = 22.36, df = 3, P « 0.005). Examin- 150 RALSTON ET AL.: BOTTOM FISH RESOURCE AT JOHNSTON ATOLL ing individual contingency table cells showed that the greatest contribution to the total chi-square was for P. filamentosus (58% of total). Specifically, under the hypothesis of independence, 16.5 were expected downcurrent but only 5 were caught, while 26.5 were expected upcurrent where 38 were landed. The ap- parent surplus of P. filamentosus along the eastern exposure, where trade winds prevail and oceanic cur- rents first impact the atoll (Barkley 1972), may relate to this fish's habit of feeding on large deepwater plankton, especially salps (genus Pyrosoma). Bray (1981) has shown that small resident planktivores will travel to the upcurrent edge of a reef to access pelagic plankton. The distribution of P. filamentosus at Johnston Atoll may represent a similar situation on a much larger scale. Bottom Fish Catch Rate One-way analysis of variance (ANOVA) of CPUE data was used to examine whether geographical dif- ferences exist in bottom fish abundance, i.e, the two treatment classes were upcurrent and downcurrent regions (see above). The ANOVA was insignificant (F = 1.62, df = 1, 21, P = 0.21), although the mean catch rate along the eastern exposure (5.6 bottom fish/line-h) was 60% greater than downcurrent (3.5 bottom fish/line-h). This result suggests the lack of significance may have been due to small sample size The CPUE data were analyzed by time of day to determine if catchability fluctuates through the day. The results in Figure 6 show that fishing was distinctly better during the morning than afternoon. In this figure individual values of drift CPUE (n = 23) have been plotted against the midpoint of the drift time interval. The solid line represents aggre- gate catch rates, calculated by pooling both catch and effort statistics from all areas into 1-h intervals and then forming CPUE ratios. Different symbols repre- sent each of six separate fishing locations (Fig. 1). Note that catch rates were highest when fishing began each day and consistently declined to a low during the midafternoon. The data further indicate that catch rates may increase again with the onset of the evening crepuscular period, although the data are meager. This pattern was evident both within and among the six sites fished and, when averaged out, resulted in morning catch rates 2.07 times greater than afternoon rates. Catchability Having the Makalii and Townsend Cromwell at Johnston Atoll at similar times prompts comparison 15 -i 10 Q. (J 5- O-1 I 1 1 1 1 1 1 1 1 0800 1000 1200 1400 1600 Time of Day Figure 6.— The effect of time of day on the catch rate of bottom fish at Johnston Atoll. Catch rates calculated for each drift of the vessel and presented for each of six different fishing stations (see Figure 1). of the assessment techniques. We assume that in the 1-mo interim between visits no changes occurred in overall levels of abundance, because Johnston Atoll is a National Wildlife Refuge where no fishing is per- mitted and the fishes are typically long lived (Ralston and Miyamoto 1983; Ralston see footnote 8). Any differences in assessment are then likely due to dif- ferences in method. To compare surface estimates of bottom fish abun- dance with those derived from submersible surveys, we matched fishing stations (numbers) with submer- sible dives (letters) which occurred nearby (Fig. 1). Specific pairings were F-l, E-2, B-3, H-4, 1-5, and D-6. For each dive the overall abundance of bottom fish was estimated by forming the ratio of total fish counted to total number of quadrat counts, and then converting to density measured in bottom fish/ha. The CPUE statistics were used to estimate abun- dance for each fishing station, after correcting for fluctuating catchability (Fig. 6). The result is pre- sented in Figure 7. There is a positive correlation between CPUE and bottom fish density (r = 0.54), although it is insignificant. One means of estimating catchability, q, is to deter- mine the slope of the regression of CPUE on stock density. We estimated the functional regression (Ricker 1973) of the data presented in Figure 7 (solid line) and determined that q = 0.0215 ha/line-h. A second estimate of q is obtained by forming the ratio of the average catch rate of bottom fish at the atoll 151 FISHERY BULLETIN: VOL. 84, NO. 1 8-1 6- a, 5_ 4 - 3 al 2 - 1 - ~ I 1 1 1 I 50 100 150 200 250 Abundance I f ish / ha I 300 Figure 7— The relationship between Townsend Cromwell CPUE and Makalii abundance estimates. Line fitted by functional regres- sion. See text for further discussion. (3.18 fish/line-h) to the average density of bottom fish viewed from the submersible (118 bottom fish/ha). The resulting estimate of q is 0.0269 ha/line-h. DISCUSSION The most enlightening aspect of this study was our ability to perform an in situ assessment of factors controlling the distribution and abundance of the deep slope biota at Johnston Atoll. Organisms showed not only distinct zonational patterns with depth but clumped dispersion along horizontal gradients as well. The fish fauna of Johnston Atoll is often con- sidered a depauperate outlier of the Hawaiian fauna (Gosline 1955; Randall et al. in press). In a later paper, Gosline (1965) examined vertical zonation in Hawaiian fishes, arguing that depth zonation pat- terns are often sharply demarcated in intertidal and shallow-water habitats, but these become increasing- ly attenuated with depth. The results of our study and Randall et al. (in press) support his conclusion (see also Forster 1984). Some deep slope species have extremely broad depth ranges (exceeding 200 m), yet few representatives of the shallow-water communi- ty extend appreciably beyond the 130 m escarpment encircling the atoll. Other investigators have noted that many Hawaiian species, which are commonly thought of as strictly associated with coral reefs, penetrate to depths well in excess of those favoring the growth of scleractinian corals (Brock and Cham- berlain 1968; Strasburg et al. 1968; Clarke 1972). Yet the distributions of these fishes are limited largely to areas near the shelf break or shallower, while a true deep slope ichthyofauna, comprised largely of anthiids and lutjanids, exists along outer reef drop- offs at both Johnston Atoll and in the Hawaiian Islands. Distributional patterns of fishes were nonrandom along horizontal gradients as well, as was readily ap- parent in the atoll-wide distribution of Pristipo- moides filamentosus . Based simply on catch totals, 60% more P. filamentosus were expected to occur on the upcurrent exposure of the atoll than down- current, although 760% more were observed there, illustrating the clumped dispersion pattern which characterized this species during fishing surveys. Contagion was also evident in quadrat samples. Future studies would be well advised to incorporate statistical models consistent with these findings, in- cluding use of the negative binomial distribution to describe spatial patterns. On a more local scale, it was clear from submer- sible observations that P. filamentosus and Etelis cor- uscans were concentrated near underwater head- lands. Brock and Chamberlain (1968) made similar observations on deepwater populations of Chaetodon miliaris, attributing the very localized distribution of this species to increased accessibility of its food (plankton) in the vertical turbulence plumes formed by the impact of currents on underwater prom- ontories. Because of its known planktivorous food habits, this hypothesis could explain abundance pat- terns of P. filamentosus. Moreover, fishermen empha- size the importance of currents in locating feeding aggregations of both P. filamentosus and E. cor- uscans. These two species taken together comprise the most important species landed in the Hawaiian deep-sea hook-and-line fishery, both in terms of yield and economic value. The relative abundance of these species in the deepwater bottom fish community may be due to their utilization of an allochthonous plank- ton resource transported to neritic waters from the open sea. Bottom Fish Abundance Certain methodological problems hindered this study and should be reviewed before comparing the abundance estimates from the two surveys. Any technique, including those used here, has its own spe- cific combination of advantages and disadvantages. There is ample reason to suspect bias in assess- ments based on underwater visual surveys. Sale and Douglas (1981) have shown that a single visual fish 152 RALSTON ET AL.: BOTTOM FISH RESOURCE AT JOHNSTON ATOLL census seldom records all individuals present at the time of the census. Similarly, Colton and Alevizon (1981) showed that a quarter of the community they studied was characterized by significant diurnal changes in abundance. They concluded that unless sampling time is carefully controlled and standard- ized, results from visual abundance surveys may be seriously biased. Standardization was achieved in this study because all 10 dives started between 0840 and 0950 in the morning and each lasted 4 h. Further- more, Brock (1982) showed that large, conspicuous, diurnally active species are accurately censused with visual assessment techniques, although the most abundant are often underestimated. With the excep- tion of Cookeolus boops, which, although nocturnal, shelters in the open along the slope face, all of the species included in the quadrat sampling fit these criteria. Biases which frequently accompany visual assessments have thus been considered and mini- mized here Another factor which may have affected the results of Makalii surveys is attraction and repulsion of cer- tain species to and from the submersible Previous investigators have typically ignored this problem (Uz- mann et al. 1977; High 1980; Powles and Barans 1980; Carlson and Straty 1981), while at the same time acknowledging that some species are attracted (ag, black sea bass, southern porgy Pacific halibut, sculpin, and yelloweye rockfish) or repelled (eg, squid, herring, mackerel, butterfish, and wolf eel) to submersibles and divers. Nevertheless, as pointed out by Uzmann et al. (1977), one can at least observe the reactions of species to the submersible's presence, giving the viewer the opportunity to evaluate poten- tial sources of error. We have attempted to address this problem by pooling counts for all species. While admittedly this procedure may not remove all bias, it is our feeling that in the absence of more quan- titative information, little else can be done to im- prove the data. Studies are now being implemented to specifically evaluate the degree of attraction or repulsion of different species to the Makalii. Provided an awareness of these concerns, the results presented here support the contention that the catch of bottom fish/line-h is a suitable CPUE statistic This conclusion is based on the data pre- sented in Figure 7, where CPUE generally increases with fish density and the regression intercept passes close to the origin. Although the relationship is statistically insignificant, this is likely due to small sample size (n = 6). Moreover, differences in bottom fish abundance between upcurrent and downcurrent locations were shown to result largely from the con- tagious dispersion of Pristipomoides filamentosus along the eastern side of the atoll, where its primary food resource first becomes available for consump- tion. The estimation of catchability for deep-sea hook- and-line gear is a useful application of the dual sam- pling program presented hera The results suggest relatively great sensitivity of bottom fish stocks to exploitation pressure, a finding consistent with pre- vious and ongoing studies (Ralston 1984). If we use q = 0.0215 ha/line-h as an estimate of catchability, we conclude that 1 line-h of Townsend Cromwell fish- ing effort removes about 2.2% of the bottom fish inhabiting 1 ha of habitat. A similar finding was reported by Polovina9, who estimated q from the same vessel for a Mariana stock of bottom fish. Re- movals such as this are not insubstantial and under- score the importance of developing methods of stock assessment which can be used early in the develop- ment of a fishery and in the absence of conventional data sources. A combination of surface platform surveys with submersible ground-truthing is certain- ly a promising assessment technique to pursue (Uz- mann et al. 1977). ACKNOWLEDGMENTS We would like to thank the U.S. Army Corps of Engineers Pacific Ocean Division, the U.S. Army Toxic and Hazardous Materials Agency, and the Na- tional Undersea Research Program at the Univer- sity of Hawaii for making this study possible Special thanks go to the staff of the Hawaii Undersea Re- search Laboratory Program and the Makalii opera- tions crew for help in coordinating the dive program and in meeting our needs for logistical support. LITERATURE CITED Amerson, A. B., Jr., and P. C. Shelton. 1976. The natural history of Johnston Atoll, central Pacific Ocean. Atoll Res. Bull. 192:1-479. Bannerot, S. P., and C. B. Austin. 1983. Using frequency distributions of catch per unit effort to measure fish-stock abundance Trans. Am. Fish. Soc. 112:608-617. Barans, C. A., and D. V. Holliday. 1983. A practical technique for assessing some snapper/group- er stocks. Bull. Mar. Sci. 33:176-181. Barkley, R. A. 1972. Johnston Atoll's wake J. Mar. Res. 30:201-216. Bray, R. N. 1981. Influence of water currents and zooplankton densities 9Polovina, J. J. A variable catchability version of the Leslie model with application to an intensive fishing experiment on a multispecies stock. Manuscr. submitted to U.S. Natl. Mar. Fish. Serv., Fish. Bull. 153 FISHERY BULLETIN: VOL. 84, NO. 1 on daily foraging movements of blacksmith, Chromis puncti- pinnis, a planktivorous reef fish. Fish. Bull., U.S. 78:829-84 1. Brock, R. E. 1973. A new distributional record for Panulirus marginatus (Quoy & Gaimard, 1825). Crustaceana 25:111-112. 1982. A critique of the visual census method for assessing coral reef fish populations. Bull. Mar. Sci. 32:269-276. Brock, V. E., and T. C. Chamberlain. 1968. A geological and ecological reconnaissance off western Oahu, Hawaii, principally by means of the research submarine "Asherah." Pac Sci. 22:373-394. Carlson, H. R., and R. R. Straty. 1981. Habitat and nursery grounds of Pacific rockfish, Sebastes spp., in rocky coastal areas of southeastern Alaska. Mar. Fish. Rev. 43(7):13-19. Carpenter, K. E., R. I. Miclat, V. D. Albaladejo, and V. T. Corpuz. 1981. The influence of substrate structure on the local abun- dance and diversity of Philippine reef fishes. In E. D. Gomez, C. E. Birkeland, R. W. Buddemeier, R. E. Johannes, J. A. Marsh, Jr., and R. T. Tsuda (editors). The reef and man, p. 497-502. Proceedings of the Fourth International Coral Reef Symposium, Vol. 2. Mar. Sci. Cent., Univ. Philippines, Quezon City. Clarke, T. A. 1972. Collections and submarine observations of deep benthic fishes and decapod Crustacea in Hawaii. Pac Sci. 26:310- 317. Colin, P. L. 1974. Observation and collection of deep-reef fishes off the coasts of Jamaica and British Honduras (Belize). Mar. Biol. (Berl.) 24:29-38. COLTON, D. E., AND W. S. ALEVIZON. 1981. Diurnal variability in a fish assemblage of a Bahamian coral reef. Environ. Biol. Fishes 6:341-345. Forster, G. R. 1984. The distribution of fishes on the outer reef slopes of Aldabra (Indian Ocean) in relation to bottom temperatures. Environ. Biol. Fishes 10:129-136. Gladfelter, W. B., J. C. Ogden, and E. H. Gladfelter. 1980. Similarity and diversity among coral reef fish com- munities: A comparison between tropical western Atlantic (Virgin Islands) and tropical central Pacific (Marshall Islands) patch reefs. Ecology 61:1156-1168. Gosline, W. A. 1955. The inshore fish fauna of Johnston Island, a central Pacific atoll. Pac Sci. 9:442-480. 1965. Vertical zonation of inshore fishes in the upper water layers of the Hawaiian Islands. Ecology 46:823-831. Grimes, C. B., K. W. Able, and S. C. Turner. 1982. Direct observation from a submersible vessel of com- mercial longlines for tilefish. Trans. Am. Fish. Soe 111:94- 98. Gulland, J. A. 1974. The management of marine fisheries. Univ. Washing- ton Press, Seattle, 198 p. High, W. L. 1 980. Bait loss from halibut longline gear observed from a sub- mersible. Mar. Fish. Rev. 42(2):26-29. Ivo, C. T C, and A. J. Hanson. 1982. Aspectos da biologia e dinamica populacional do pargo, Lutjanus purpureus Poey, no norte e nordeste do Brasil. Arq. Cienc. Mar 22(1/2):1-41. Kami, H. T. 1973. The Pristipomoid.es (Pisces: Lutjanidae) of Guam with notes on their biology. Micronesica 9:97-118. LUCKHURST, B. E., AND K. LUCKHURST. 1978. Analysis of the influence of substrate variables on coral reef fish communities. Mar. Biol. (Berl.) 49:317-323. MacCall, A. D. 1976. Density dependence of catchability coefficient in the California Pacific sardine, Sardinops sagax caerula, purse seine fishery. Calif. Coop. Oceanic Fish. Invest. Rep. 18:136-148. Moffitt, R. B. 1980. A preliminary report on bottomfishing in the North- western Hawaiian Islands. In R. W. Grigg and R. T. Pfund (editors), Proceedings of the Symposium on Status of Re- source Investigations in the Northwestern Hawaiian Islands, April 24-25, 1980, University of Hawaii, Honolulu, Hawaii, p. 216-225. Sea Grant Misc Rep. UNIHI-SEAGRANT-MR- 80-04. Munro, J. L. (editor). 1983. Caribbean coral reef fishery resources. ICLARM Studies and Reviews 7, Manila, 276 p. Pitcher, T J., and P. J. B. Hart. 1982. Fisheries ecology. Avi Publ. Co. Inc., Westport, Conn., 414 p. POWLES, H, AND C. A. BARANS. 1980. Groundfish monitoring in sponge-coral areas off the southeastern United States. Mar. Fish. Rev. 42(5):21-35. Ralston, S. 1980. An analysis of the Hawaiian offshore handline fishery: A progress report. In R. W. Grigg and R. T Pfund (editors), Proceedings of the Symposium on Status of Resource Inves- tigations in the Northwestern Hawaiian Islands, April 24-25, 1980, University of Hawaii, Honolulu, Hawaii, p. 216-225. Sea Grant Misc Rep. UNIHI-SEAGRANT-MR-80-04. 1984. Biological constraints on production and related man- agement issues in the Hawaiian deepsea handline fishery. In R. W. Grigg and K. Y. Tanoue (editors), Proceedings of the Symposium on Resource Investigations in the Northwest- ern Hawaiian Islands, May 25-27, 1983, University of Hawaii, Honolulu, Hawaii, p. 248-264. Sea Grant Misc Rep. UNIHI- SEAGRANT-MR-84-01. Ralston, S., and G. T. Miyamoto. 1983. Analyzing the width of daily otolith increments to age the Hawaiian snapper, Pristipomoides filamentosus. Fish. Bull., U.S. 81:523-535. Ralston, S., and J. J. Polovina. 1982. A multispecies analysis of the commercial deep-sea handline fishery in Hawaii. Fish. Bull., U.S. 80:435-448. Randall, J. E., P. S. Lobel, and E. H. Chave. In press. Annotated checklist of the fishes of Johnston Island. Pac Sci. Ricker, W. E. 1973. Linear regressions in fishery research. J. Fish. Res. Board Can. 30:409-434. 1975. Computation and interpretation of biological statistics of fish populations. Fish. Res. Board Can. Bull. 191, 382 p. Rothschild, B. J. 1977. Fishing effort. In J. A. Gulland (editor), Fish popula- tion dynamics, p. 96-115. John Wiley and Sons, N.Y. Sale, P. F., and W. A Douglas. 1981. Precision and accuracy of visual census technique for fish assemblages on coral patch reefs. Environ. Biol. Fishes 6:333-339. Shipp, R. L., and T S. Hopkins. 1978. Physical and biological observations of the northern rim of the De Soto Canyon made from a research submersible Northeast Gulf Sci. 2(2): 1 13-121. 154 RALSTON ET AL.: BOTTOM FISH RESOURCE AT JOHNSTON ATOLL Strasburg, D. W.. E. C. Jones, and R. T. B. Iversen. S. S. Lampton (editors), Fisheries techniques, p. 239-259. 1968. Use of a small submarine for biological and oceano- Am. Fish. Soc., Bethesda, MD. graphic research. J. Cons. Cons. Perm. Int. Explor. Mer Uzmann, J. R., R. A. Cooper, R. B. Theroux, and R. L. Wigley. 31:410-426. 1977. Synoptic comparison of three sampling techniques for Talbot, F H. estimating abundance and distribution of selected mega- 1960. Notes on the biology of the Lutjanidae (Pisces) of the fauna: Submersible vs camera sled vs otter trawl. Mar. Fish. East African coast, with special reference to L. bohar (For- Rev. 39(12):11-19. skal). Ann. S. Afr. Mus. 45:549-573. Thorne, R. E. 1983. Hydroacoustics. In L. A. Nielsen, D. L. Johnson, and 155 PATCHINESS AND NUTRITIONAL CONDITION OF ZOOPLANKTON IN THE CALIFORNIA CURRENT Stewart W. Willason, John Favuzzi, and James L. Cox1 ABSTRACT Zooplankton and water samples were collected from 81 stations off the California coast in April 1981 during CalCOFI cruise 8104 aboard the RV David Starr Jordan. Abundance, weight (wet and dry), digestive enzyme activity (laminarinase), and biochemical composition of three zooplankton species were determined. The indices measured provided estimates of zooplankton nutritional history on time scales of 1 day to 3 weeks. Upwelling was taking place along the California coast, from Point Conception to San Francisco dur- ing the study period. The resulting low surface temperatures were most evident south of San Francisco and just north of Point Conception. Just south of these areas patches of high phytoplankton standing crop (up to 14.7 mg chlorophyll a/m3) were found. The two herbivorous species, Euphausia pacifica and Calanus pacificus, showed highest laminarinase activity in areas with the highest density of phytoplank- ton: enzyne activity was particularly high in the waters off Point Conception. Zooplankters in the southern and offshore regions of the sampling grid showed very low digestive enzyme activity. The larger size (weight) and higher lipid content of C. pacificus near Point Conception and south of San Francisco in comparison to animals in other parts of the California Current suggest that animals in these areas experience pro- longed periods of better nutrition. Nematoscelis difficilis, which is not a herbivore, did not show these patterns. This study illustrates the importance of upwelling regions, such as Point Conception, and shows the large spatial variability of trophic interactions within the California Current System. The nearshore, pelagic marine environment is ex- tremely variable and heterogeneous. Spatial hetero- geneity of physical conditions elicit behavioral or physiological responses from marine organisms which contribute to biological patchiness (Haurey et al. 1978; Steele 1978). Patchiness of pelagic marine organisms occurs on all temporal and spatial scales (Haury et al. 1978); one of the most important of these is the mesoscale (a few kilometers to 100's of kilometers, and a few weeks to months). Mesoscale processes, such as coastal upwelling, play a major role in structuring the physical and biological en- vironment at all scales (Haury 1982). Although up- welling regions are very productive (eg., Ryther 1969), trophic interactions within these important areas are poorly understood. Along the California coast episodic upwelling takes place during the spring and summer months (Reid et al. 1958; Bernstein et al. 1977; Owen 1980; Lasker et al. 1981; Parrish et al. 1981). Upwelling results in mesoscale phytoplankton patchiness along the coast and in the southward flowing California Cur- rent (Owen 1974; Cox et al. 1982; Smith and Baker 1982; Pelaez and Guan 1982). It is thought that phy- toplankton patchiness in this area influences the sur- ^arine Science Institute, University of California, Santa Bar- bara, CA 93106. vival and physiological condition of larval fish popula- tions (Lasker 1975; Lasker and Smith 1977; Lasker and Zweifel 1978; O'Connell 1980). In addition, nutri- tion of herbivorous zooplankton (estimated by diges- tive enzyme activity) is influenced by phytoplankton patchiness (Cox et al. 1982; Cox et al. 1983; Willa- son and Cox in press). This study investigates the impact that mesoscale and larger scale phytoplankton patchiness have on zooplankton populations within the California Cur- rent along the central and southern California coast. Results of measurements of temperature, phyto- plankton biomass, zooplankton abundance, and zoo- plankton nutrition are presented. Nutritional status was evaluated using intrinsic properties which reflect previous feeding conditions. Short-term feeding his- tory was estimated from measurements of the acti- vity of the digestive enzyme, laminarinase Although digestive enzyme levels of zooplankton do not always provide a good measure of instantaneous digestive or feeding rates (Hassett and Landry 1983; Head et al. 1984; Willason and Cox in press), the level of activity in field captured animals does give an indica- tion of relative feeding history on the order of 1 to 5 d (Cox 1981; Cox and Willason 1981; Cox et al. 1983; Willason 1983). Longer term nutritional con- dition was assessed from biochemical composition and animal size (wet and dry weight) measurements. Manuscript accepted April 1985. FTSWRRV RTTT.T.F.TTN- VDT . RA NO 1 1 QRfi /r* -/?4 157 FISHERIES BULLETIN: VOL. 84, NO. 1 Lipid content, size, and water content of a zooplank- ton species reflect feeding history on the order of 1 to 3 wk (Omori 1970; Lee et al. 1970, 1971; Bam- stedt 1975; Childress 1977; Boyd et al. 1978; Vidal 1980; Hakanson 1984). Spatial patterns derived from these data are used to estimate relative differences in feeding and nutritional condition of zooplankton from different areas within the California Current. An understanding of the interrelationships of these variables in different areas may provide insights in- to mechanisms which generate and maintain physical and biological mesoscale features. METHODS Species Studied Two euphausiid species, Euphausia pacifica Han- sen and Nematoscelis difficilis Hansen, and the copepod, Calanus pacificus Brodsky were chosen for the present study because 1) all are common in the California Current region (Fleminger 1964; Brinton 1967b), 2) all have been used in previous digestive enzyme studies (Cox 1981; Cox and Willason 1981; Hassett and Landry 1982, 1983; Cox et al. 1983; Willason 1983; Willason and Cox in press), and 3) a large base of information exists on the sizes, feeding rates, and energetics of these zooplankters (Brinton 1967a; Mullin and Brooks 1976; Vidal 1980; Ross 1982; Cox et al. 1983; Torres and Childress 1983; Willason 1983; Hakanson 1984; Willason and Cox in press). Euphausia pacifica, the most abun- dant euphausiid in the California Current (Brinton 1967b; Brinton and Wyllie 1976; Youngbluth 1976), and C. pacificus, the most abundant copepod along the California coast (Fleminger 1964; Star and Mullin 1981), are considered primarily herbivorous (Mullin and Brooks 1976; Ross 1982; Willason and Cox in press). By contrast, N. difficilis does not ap- pear to be a herbivore (Nemoto 1967; Mauchline and Fisher 1969; Willason and Cox in press). Sample Collection The sampling program was conducted off the Cali- fornia coast from 7 to 27 April 1981 in conjunction with the California Cooperative Fisheries Investiga- tion (CalCOFI) survey. Zooplankton and water sam- ples were collected from 81 stations during CalCOFI cruise 8104 aboard RV David Starr Jordon. Figure 1 shows the stations sampled and the sampling se- quence during the cruise. The grid covered an area of about 270,000 km2; nearshore stations were sometimes within 1 km of the coast and offshore stations were located up to 300 km from the coast. Although the mean flow of the California Current is south through the sampling grid at this time of the year (Lynn et al. 1982), smaller regions within the grid are often subjected to different hydro- graphic influences. For example, the waters of the offshore regions intergrade with the waters of the Central Pacific Gyre (Bernstein et al. 1977); the nearshore region south of Point Conception (the Southern California Bight) is characterized by a semipermanent, counterclockwise eddy and is hydro- graphically distinct from the other areas of the grid (Owen 1980); and the nearshore area adjacent to and north of Point Conception is characterized by periods of intense coastal upwelling during the spring and summer months (Parrish et al. 1981). To compare the biological and nutritional properties of zooplank- ton in the different hydrographic regions, the sam- pling grid was divided into four sections: southern nearshore (I), northern nearshore (II), southern off- shore (III), and northern offshore (IV) (Fig. 1). Surface chlorophyll a concentration (depth of 2 m) was used as an indicator of phytoplankton standing crop. Previous studies have shown that there are positive correlations between surface chlorophyll a, integrated chlorophyll a, and primary production in the waters of the California Current (Lorenzen 1970; Hayward and Venrick 1982). Measurements of sur- face chlorophyll a, therefore, give a relative approx- imation of phytoplankton biomass within the sam- pling grid. Two replicate water samples (0.25 to 2.0 L) for chlorophyll a analysis were taken at each of the 81 stations from a depth of about 2 m using the ship's seawater pumping system. Each sample was filtered through a 4.5 cm Whatmann GF/C filter; two drops of a seawater-saturated MgC03 solution were add- ed during filtrations. The filters were folded in half and stored frozen in aluminum foil at -20°C. An additional 15 water samples were taken for chloro- phyll a analysis along the cruise track adjacent to and immediately south of the Point Conception region while the ship was under way. Measurements of surface water temperature (±0.1°C) were also taken at each station using a glass mercury thermo- meter. Paired bongo nets (designated net 1 and net 2) with mouth openings of 0.396m2 and mesh openings of 505 ptm were used for the collection of zooplankton samples. A General Oceanics2 flowmeter was mount- 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 158 WILLASON ET AL.: ZOOPLANKTON IN CALIFORNIA CURRENT I Southern Nearshore II Northern Nearshore III Southern Offshore IV Northern Offshore San Francisco ~&Sjii: Monterey Bay II J \ > v oj £///. Point Conception San Diego 38c 36( 34° 32° 124< 122< 120° 118' Figure 1.— Sampling grid. Open circles are day stations and closed circles are night stations. Arrows show the sampling sequence The first station, adjacent to San Diego, was occupied on 7 April 1981. The last station, just north of San Francisco, was occupied on 27 April 1981. ed inside the mouth of each net to measure the volume of water filtered. An oblique net tow was made to a depth of about 210 m at each station (bot- tom depth permitting); each net filtered about 400 m3 of water. Ship speed during the net tows was 1.5 to 2.0 kn. Thirty-six stations were occupied at night (after sunset and before sunrise) and 45 were oc- cupied during the day. The euphausiids, Euphausia pacifica and Nema- toscelis difficilis, and the copepod, Calanus pacificus, were separated from the catch of net 1 immediately after collection. Adult euphausiids were sorted for males and females and copepods sorted for females and stage V copepodites. Specimens of E. pacifica and N. difficilis were considered adults if they were larger than 11 mm and 15 mm, respectively (Brin- ton and Townsend 1981). Fifty undamaged animals of each species and sex (or stage) were saved from each net tow if adequate numbers were captured. For C. pacificus, which were very abundant, 50 females and stage V's were saved from 72 and 75 of the 81 stations, respectively. Two replicate groups of 50 females of C. pacificus were taken from 7 stations and two replicate groups of 50 stage V copepodites from 9 stations. After sorting, animals from each net tow were wrapped in parafilm in groups (5 to 50 animals of each sex or stage) and frozen at -20°C for biochemical analyses in the laboratory. Catches from net 1 that could not be sorted on the ship (10 of the 81 stations sampled) were frozen whole at -20°C and sorted in the laboratory after the cruise. The entire catch of net 2 was preserved in Formalin immediately after collection. The abundances (numbers per 1,000 m3) of adult euphausiids at each station were estimated by count- ing all adults captured in net 1 and dividing by the volume of water filtered. Copepod abundances (num- bers per 1 m3) were estimated by counting all 159 FISHERIES BULLETIN: VOL. 84, NO. 1 females and stage V copepodites in triplicate aliquots taken from the preserved catches of net 2. Sample Analyses All frozen samples were analyzed in the laboratory within 6 wk of the time of collection. Plant pigments were extracted from the filters in 90% acetone in darkness at 4°C for 48 h. Chlorophyll a concentra- tion was determined by the method of Strickland and Parsons (1972) using a model 10-005 Turner Designs fluorometer. The two chlorophyll a measurements from each station were averaged. Groups of frozen animals (separate species and sexes) were thawed in the laboratory, blotted lightly to remove excess water, and weighed (±0.01 mg). Animals were then freeze-dried for 24 h at -50°C and reweighed. Groups were then immediately ground in cold (4°C) succinic acid buffer (pH 5.0) using a Polytron grinder (for euphausiids) or a hand glass tissue grinder (for copepods). Homogenates were analyzed for total proteins by the Lowry method using Sigma protein standard (Merchant et al. 1964). Laminarinase activity (LA) of the homo- genates was determined by the methods described by Cox (1981) and Willason (1983). LA was expressed as a function of the animal's wet weight: yg glucose produced per gram wet weight per minute of incu- bation. Copepod homogenates were also analyzed for total lipids using stearic acid as the standard (Bligh and Dyer 1959; Marsh and Weinstein 1966). Data Analysis Willason and Cox (in press) found that E. pacifica exhibits a diel rhythm in enzyme activity associated with feeding activity at night. Thus, to compare LA of E. pacifica collected at different times of the day from different localities, enzyme levels had to be standardized with respect to the time of capture. Calibration factors, which convert the LA of E. pacifica collected at different times to a standardized maximum value (between 0200 and 0800 h), were derived from the results of the 24-h time-series col- lections in Willason and Cox (in press). These fac- tors are based on the average relative increases and decreases of enzyme activity over a 24-h period (Table 1). LA of AT. difficilis and C. pacificus do not show diel changes (Cox et al. 1983; Willason and Cox in press) and, therefore, were not standardized. The data set for each station consists of surface temperature, surface chlorophyll a, zooplankton abundance, LA, individual wet and dry weights, pro- tein content, and lipid content (copepods only). To permit parametric statistical comparisons between the various biological and physical properties and between regions, chlorophyll a, zooplankton abun- dance, and zooplankton LA were normalized by log transformation. The log transformed values were used for all parametric statistical tests. Zooplankton wet weight, dry weight, protein content, and lipid content were found to be normally distributed by probit analysis and were not log transformed. Non- transformed values from all data sets were used to construct contour maps. The contour maps are in- tended to show general trends and patchiness within the sampling grid. Table 1.— Correction factors for standard- izing laminarinase activity (LA) of Euphausia pacifica. These factors account for diel changes in LA and are based on the time of capture. They were derived from the 24-h time-series collections of Willason and Cox'. LA was standardized to the 0200-0800 time period. LA of euphausiids captured during other time periods was multiplied by the corresponding factor. Correction factor Time period Females Males 2000-0200 0200-0800 0800-1400 1400-2000 1.042 1.000 1.253 1.486 1.132 1.000 1.281 1.453 'Willason, S. W. and J. L. Cox. In press. Diel feeding, laminarinase activity and phytoplankton consumption by euphausiids. Biol. Oceanogr. RESULTS Surface Water Temperature and Surface Chlorophyll a Surface water temperatures along the California coast during April 1981 ranged from 9.6° to 16.0°C. The coldest water was located in the northern near- shore region and the warmest was found in the southern offshore region (Table 2, Fig. 2). Two small areas showed very low surface water temperatures: close to the shore along the central coast of Califor- nia and just off San Francisco Bay (Fig. 2). A cold water plume extended from Point Conception south into the Southern California Bight. Chlorophyll a concentrations showed greater than 100-fold variation between stations and were inverse- ly correlated with surface water temperatures (r = 0.83, P < 0.001). Lowest values, 0.09 to 0.16 mg chlorophyll a/m3, were found in the southern off- shore region. Highest concentrations occurred in the northern nearshore region (Table 2, Fig. 3). Within 160 WILLASON ET AL.: ZOOPLANKTON IN CALIFORNIA CURRENT Table 2— Mean surface water temperature and mean surface chlorophyll a. Chlorophyll a expressed as mg/m3. The numbers in parentheses are one standard deviation. Southern Nearshore (1) Southern Offshore (III) Northern Nearshore (II) Northern Offshore (IV) Temperature (°C) 14.85 (0.72) 15.03 (0.49) 11.56 (0.82)* 13.68 (0.78) Chlorophyll a Log Chlorophyll a 0.659 (0.88) -0.378 (0.38) 0.141 (0.04) -0.883 (0.12) 5.110 (4.42) 0.555 (0.39)* 0.485 (0.29) -0.404 (0.31) No. of stations 27 12 25 17 indicates value(s) significantly different from those of other regions (P < 0.05, t-test). Surface Temperature (°C) 9.0- 9.9 10.0-10.9 11.0-11.9 12.0-12.9 13.0-13.9 14.0-14.9 15.0-16.0 Figure 2 — Surface water temperatures (°C) along the California coast. L 124c 122' 120c 118c this region two areas of very high chlorophyll a (up to 14.7 mg/m3) were found: near Point Conception and just south of San Francisco Bay. These areas were located just south of the areas of coldest sur- face waters. Euphausiid Distribution and Abundance Euphausia pacifica adults were captured at 43 of the 81 stations sampled and Nematoscelis difficilis adults were captured at 38 stations. As there was no significant difference between numbers of males and females captured of either species (P > 0.3, Wilcoxon test), the abundances shown in Figures 4 and 5 represent the sum of both sexes. Both the number of specimens ofN. difficilis captured at each station (P < 0.01, t-test) and the proportion of sta- tions where individuals were caught (P < 0.01, x2 test) were greater at night. For E. pacifica, there were no significant day-night differences in the num- bers of animals captured (P > 0.2, £-test), however, like N. difficilis, the proportion of stations where individuals were captured was greater at night (P < 0.05, x2 test). The day-night differences may represent net avoidance by euphausiids or under- sampling during the day because of vertical migra- tion. Thus, the data presented in Figures 4 and 5 represent general trends and are intended to show relative differences between areas. Because euphau- siids were captured at only about one half of the sta- tions, statistical comparisons were made only between the north and south (i.e, nearshore and 161 FISHERIES BULLETIN: VOL. 84, NO. 1 Figure 3— Surface chlorophyll a. Ex- pressed as mg chlorophyll a per m3. Surface Chlorophyll a mg/m3 > 7.0 3.5-7.0 1.3-3.4 0.4-1.2 < 0.4 San £/;;.Diego 38c 36< 34< 32( 124c 122c 120c 118c Euphausia pacifica Abundance San Adults/1000 m3 Francisco > 1500 400-1500 40-399 < 40 None Captured ;;j. Point Conception San £:•;... Diego J L J L 38c 36° 34< 32< Figure A— Euphausia pacifica abundance Expressed as number of adults per 1,000 m3. 124< 122° 120° 118c 162 WILLASON ET AL.: ZOOPLANKTON IN CALIFORNIA CURRENT offshore regions for the north and south were combined). Specimens of E. pacifica were captured in signifi- cantly greater numbers north of Point Conception (Table 3) and were rare or absent at most offshore stations (regions III and IV). This species was espe- cially abundant off Point Conception and just south of Monterey Bay along the central coast (Fig. 4). These two areas were located close to the areas of highest chlorophyll a concentration. The abundance of E. pacifica was significantly correlated with chlorophyll a over the entire grid (Table 4). The distribution of N. difficilis (Fig. 5) was quite different from that of E. pacifica. This species was captured at only 30% of the stations where E. pacifica was found and was distributed farther off- Table 3.— Mean abundance and laminarinase activity (LA) of Euphausia pacifica and Nematoscelis difficilis in the north and south regions. Numbers in parentheses are one standard deviation. Log values were used for statistical comparisons. South Regions (I & North (Regions (II & IV) Males Females Males Females Euphausia pacifica Abundance (No./1,000 m3) Log abundance LA Log LA No. of stations Nematoscelis difficilis Abundance (No./1,000 m3) Log abundance LA Log LA No. of stations 96.07 (100.4) 1.604 (0.623)* 122.5 (47.8) 2.058 (0.167) 16 13.71 (12.78) 1.001 (0.327)* 167.3 (87.5) 2.172 (0.237)* 16 96.41 (102.5) 1.647 (0.666)* 165.1 (59.9) 2.186 (0.183) 15 18.06 (16.07) 1.061 (0.461)* 208.6 (102.9) 2.270 (0.207)* 18 200.6 (234.4) 2.035 (0.551) 109.7 (68.9) 1 .965 (0.263) 27 55.11 (70.31) 1.530 (0.441) 104.4 (41.9) 1.992 (0.174) 20 270.2 (337.3) 2.119 (0.579) 153.2 (111.9) 2.099 (0.269) 27 75.17 (83.84) 1 .657 (0.453) 130.4 (55.1) 2.041 (0.191) 19 indicates value(s) significantly different between north and south (P < 0.05, f-test). Figure 5.— Nematoscelis difficilis abun- dance Expressed as number of adults per 1,000 m3. Nematoscelis difficilis Abundance Adults/1000 m3" >250 50-250 10-49 <10 None Captured '///..Point Conception San V^.Diego 38< 36c 34< 32c 124° 122c 120c 118c 163 FISHERIES BULLETIN: VOL. 84, NO. 1 shore As with E. pacifica, both sexes of AT. difficilis were found in significantly greater numbers in the north (Table 3). The abundance of N. difficilis was not correlated with surface chlorophyll a (Table 4). Euphausiid Laminarinase Activity Similar to the results of Willason (1983) and Willa- son and Cox (in press), males of both euphausiid species showed significantly less LA than females (P < 0.01, both cases, Wilcoxon test). Males in this study had about 70% (Euphausia pacifica) or 80% (Nematoscelis difficilis) of the LA of females (Table 3). To simplify the presentation of the data on the contour maps, LA values of males and females at each station were averaged. The values of LA for Euphausia pacifica within the sampling grid ranged from 50 to 430. Euphau- siids with the lowest LA values were found in off- shore areas and in the nearshore area along the cen- tral coast. Euphausia pacifica with the highest levels of LA were found just south of San Francisco Bay and adjacent to the south of Point Conception (Fig. 6). These areas overlapped with and extended just south of the regions of highest surface chlorophyll a. There was a positive correlation between LA of E. pacifica and chlorophyll a over the entire grid (Table 4). Table 4— Correlations between chlorophyll a, zooplankton abundance, and laminarinase activity (LA) for Euphausia pacifica, Nematocelis difficilis, and Calanus pacificus. For euphausiids, abundance and LA values used in the analyses are the averages of males and females. Numbers in parentheses refer to the number of samples used in regression analyses. Correlation coefficients Correlation E. pacifica (43) N. difficilis (38) C. pacificus 9(81) C. pacificus V(81) Chlorophyll a vs. abundance Chlorophyll a vs. LA LA vs. abundance 0.61 0.57 0.40 10.27 10.03 10.14 0.24 0.53 0.38 0.31 0.62 0.48 'Correlation coefficients which were not significant at the 95% level. J L Euphausia pacifica Laminarinase Activity >300 200-300 100-199 < 100 None Captured ///.Point Conception San rv/;;;. Diego J L j i 38c 36° 34c 32e Figure 6— Euphausia pacifica laminari- nase activity (LA). Expressed as fig glucose per gram wet weight per minute 124< 122c 120c 118c 164 WILLASON ET. AL.: ZOOPLANKTON IN CALIFORNIA CURRENT The values of LA for Nematoscelis difficilis were in the same range as those of Euphausia pacifica (50 to 400), but showed a different distributional pat- tern (Fig. 7). Regions of highest activity were located in three small areas: adjacent to San Diego, in the Santa Barbara Channel (just south of Point Concep- tion), and in an area about 150 km off Monterey Bay. Both males and females of N. difficilis had signifi- cantly higher levels of LA in the southern portion of the grid (Table 3). LA of AT. difficilis was not corre- lated with chlorophyll a (Table 4). Nematoscelis dif- ficilis with high LA were often found in areas with very low phytoplankton biomass and vice versa. Euphausiid Size and Chemical Composition Mean wet and dry weights, water content, and pro- tein content (expressed as percent dry weight and percent wet weight) of Euphausia pacifica and Nematoscelis difficilis are presented in Table 5. Female E. pacifica and both sexes oiN. difficilis had significantly higher wet and dry weights in the north. The water content of both euphausiid species ranged from 76.5 to 81.7% and was very similar between species, sexes, and regions (Table 5). Protein content was also very similar between species, sexes, and regions. The protein values reported here (51 to 56% of dry weight) are within the range of previously reported values (Childress and Nygaard 1974). Copepod Distribution and Abundance Female and stage V copepodites of Calanus paci- ficus were captured at all 81 stations sampled. There were no significant differences between day and night catches for either C. pacificus (P < 0.01, £-test). For comparisons between regions, mean abundances were calculated using both the log transformed and nontransformed values (Table 6). The log transform- ed values were used for statistical comparisons. The overall abundances of females and stage V copepo- dites were similar to one another in all regions (P > 0.1, t-test, all cases). Both C. pacificus stages were significantly more abundant in the two nearshore regions (I and II) than in the two offshore regions (III and IV) (Table 6). Figures 8 and 9 show that the distributions of females and stage V C. pacificus were patchy within regions. Copepods were particu- larly abundant in the area close to and just south of Point Conception. An extremely dense aggrega- Figure 1 —Nematoscelis difficilis lami- arinase activity (LA). Expressed as fig glucose per gram wet weight per minute. Nematoscelis difficilis Laminarinase Activity. >300 200-300 100-199 <100 None Captured Point Conception San Diego 38c 36c 34c 32( 124< 122< 120' 118c 165 FISHERY BULLETIN: VOL. 84, NO. 1 Table 5.— Mean individual wet weight, dry weight, % water, and protein content of Euphausia pacifica and Nematoscelis difficilis from the north and south. Numbers in parentheses are one standard deviation. South (Regions 1 & III) North (Regions II & IV) Males Females Males Females Euphausia pacifica Wet weight (mg) 31.01 (11.55) 32.57 (10.81)* 37.73 (11.24) 42.38 (12.11)* Dry weight (mg) 6.48 (2.38) 6.75 (2.53)* 7.91 (2.57) 8.98 (2.41)* % water 79.10 79.28 79.04 78.81 Protein (% dry wt) 54.57 56.16 52.62 52.26 Protein (% wet wt) 11.40 11.62 11.05 11.02 No. of stations 16 15 27 27 Nematoscelis difficilis Wet weight (mg) 27.63 (7.07)* 34.73 (11.28)* 35.23 (7.26)* 43.59 (8.72)* Dry weight (mg) 5.96 (2.21) 7.22 (2.49)* 7.43 (2.22) 9.19 (2.78)* % water 78.43 79.22 78.82 78.94 Protein (% dry wt) 56.59 51.23 52.92 54.96 Protein (% wet wt) 12.22 10.65 11.17 11.58 No. of stations 16 18 20 19 indicates value(s) significantly different between north and south (P < 0.05, Mest). Calanus pacificus Abundance copepods/m3 San :.. Diego 38c 36c 34c 32( 124° 122° 120c 118° Figure 8.— Calanus pacificus females, abundance Expressed as number of copepods per m3. tion of stage V C. pacificus (474 copepods/m3) was found at the station adjacent to Point Conception. The areas where C. pacificus showed the highest abundances were located near regions of high chloro- phyll a concentration. However, the abundances of both C. pacificus stages were poorly correlated (al- though significant at the 95% level) with chlorophyll a over the entire grid (Table 4). Copepod Laminarinase Activity LA of female and stage V copepodites was much higher than the levels of both euphausiid species when expressed on a per weight basis. Like the euphausiid results, there was large variability in the LA of C. pacificus among stations. For example, LA of stage V copepodites ranged from < 150 at offshore 166 WILLASON ET. AL.: ZOOPLANKTON IN CALIFORNIA CURRENT Table 6. — Calanus pacificus. Mean abundance and laminarinase activity (LA) of stage V copepodites and females from each region. Numbers in parentheses are one standard deviation. Log values were used for statistical comparisons. Southern Nearshore (1) Southern Offshore (III) Northern Nearshore (I!) Northern Offshore (IV) Stage V copepodites Abundance (No./m3) Log abundance 22.89 (32.11) 0.924 (0.681)* 1.70 (0.86) 0.175 (0.233) 26.07 (93.90) 0.571 (0.733)* 1.82 0.139 (1.44) (0.299) LA Log LA 825.4 (455) 2.845 (0.272) 538.6 (254) 2.688 (0.201) 1,527.5 (792.1) 3.129 (0.231)* 933.2 2.891 (659.4) (0.258) Females Abundance (No./m3) Log abundance 14.21 (14.72) 0.807 (0.692)* 2.54 (2.21) 0.253 (0.411) 6.67 (10.79) 0.621 (0.400)* 3.05 0.343 (2.03) (0.351) LA Log LA 927.6 (466.2) 2.913 (0.281) 635.2 (413.5) 2.734 (0.222) 1,272.9 (610.3) 3.072 (0.204)* 1,041.5 (547.5) 2.856 (0.261) No. of stations 27 12 25 17 indicates value(s) significantly greater than those of other regions (P < 0.05, r-test). 41/ Figure 9— Calanus pacificus Stage V copepodites, abundance Expressed as number of copepods per m3. Calanus pacificus V Abundance - copepods/m3 San Diego 38c 36' 34c 32c 124c 122c 120c 118c stations to 3,855 at the station adjacent to Point Con- ception. LA of replicate groups of 50 copepods from the same station were very similar indicating that the variability was due to differences between sta- tions (P < 0.05, ANOVA). Calanus pacificus LA also showed large differ- ences among the four hydrographic regions. Both females and stage V copepodites from the northern nearshore region (II) had significantly higher levels of LA than copepods from the other regions (Table 6). Copepods in the southern offshore region had the lowest levels. The contour maps of C. pacificus LA show patches of copepods with high LA located ad- jacent to and just south of Point Conception and off Monterey Bay (Figs. 10, 11). These areas were located near the regions of highest E. pacifica LA (Fig. 6) and close to the regions of highest chloro- phyll a (Fig. 3). There were significant positive cor- 167 FISHERY BULLETIN: VOL. 84. NO. 1 Figure 10— Calanus pacificus females, laminarinase activity (LA). Expressed as /jg glucose per gram wet weight per minute Calanus pacificus q Laminarinase Activity >2000 1100-2000 500-1099 < 500 7/. Point Conception San .Diego 38' 36° 34c 32c 124< 122< 120c 118c Calanus pacificus V 124c Figure 11.— Calanus pacificus Stage V copepodites, laminarinase activity (LA) Expressed as ng glucose per gram wet weight per minute 122c 120c 118' 168 WILLASON ET AL.: ZOOPLANKTON IN CALIFORNIA CURRENT relations between the LA of both C. pacificus stages and the concentration of chlorophyll a (Table 4). Copepod Wet and Dry Weights The largest female and stage V C. pacificus in terms of weight were located in the northern near- shore region and the smallest copepods were found in the southern regions (Table 7). The average water content of both C. pacificus stages from the four regions was inversely related to the average dry weights. Specimens of C. pacificus with the lowest water content were found in the northern nearshore region and those with highest water content were located in the southern offshore region (Table 7). Figures 12 and 13 show the distribution of wet weights of C. pacificus females and stage V copepo- dites, respectively. Since wet and dry weights were highly correlated (r = 0.81 and 0.83, P < 0.001) only wet weights are shown. Both figures show a band of large copepods in the nearshore region along the central coast. The figures also show the variation in size of each stage between areas. Copepods (both stages) in the "heavy band" along the central coast were almost twice the weight of copepods at some of the offshore and southern stations. Copepod Protein and Lipid Content Total protein content 0*g per copepod) of both C. pacificus stages was highest in the northern near- shore region and lowest in the two southern regions (Table 7). This appears to reflect differences in cope- pod size between regions as there were highly sig- nificant correlations between the protein content and the wet weight for both female (r = 0.82, P < 0.001) and stage V C. pacificus (r = 0.69, P < 0.001). Pro- tein content was not mapped since the patterns were very similar to those of wet weight. Protein content of C. pacificus, expressed as per- cent of wet weight, was quite similar between re- gions: 8.9 to 10.5% for stage V copepodites and 9.3 and 10.8% for females (Table 7). However, both stages from the southern offshore region did show slightly higher protein content when expressed as percent dry weight. This probably reflects the high water content of copepods from the southern off- shore region. The distributions of lipid content of female and stage V C. pacificus were very patchy and showed greater than fourfold variation between areas (Figs. 14, 15). Copepods with highest lipid values were found in the area surrounding Point Conception and off San Francisco Bay. Although copepod size (wet weight) probably influenced the total lipid content of C. pacificus to some extent, the variability of lipid content cannot be attributed solely to weight. Lipid content, unlike protein content, was poorly cor- related with wet weight (r = 0.26 for females and r = 0.38 for stage V copepodites). Table 7. — Calanus pacificus. Mean wet weight, dry weight, percent water, protein content, and lipid content for stage V copepodites and females from each region. Numbers in parentheses are one standard deviation. Southern Nearshore (1) Southern Offshore (III) Northern Nearshore (II) Northern Offshore (IV) Stage V copepodites Wet weight (^g) 471 (81) 447 (83) 555 (92)* 465 (95) Dry weight fag) 98 (21) 88 (23) 125 (26)* 98 (25) % water 79.20 80.31 77.54 78.89 Protein (^g/copepod) Protein (% dry wt) Protein (% wet wt) 41.88 (12.24) 44.12 8.89 44.82 (9.31) 49.25 10.03 52.15 (11.26) 42.75 9.40 48.58 (8.02) 48.10 10.45 Lipid (^g/copepod) Lipid (°/o dry wt) Lipid (% wet wt) 19.74 (7.82) 20.78 4.19 13.94 (4.96) 15.32 3.12 29.33 (7.72)* 24.04 5.28 15.74 (5.71) 15.58 3.38 Females Wet weight (^g) 1,023 (170) 1,083 (160) 1,278 (180)* 1,125 (190) Dry weight fag) 191 (47) 185 (38) 263 (40)* 225 (29) % water 81.34 82.83 79.40 80.20 Protein (^g/copepod) Protein (% dry wt) Protein (% wet wt) 94.92 (22.71) 49.70 9.28 100.84 (24.84) 54.51 9.31 137.81 (26.40)* 52.74 10.78 115.62 (23.33) 51.38 10.28 Lipid (fjg/copepod) Lipid (% dry wt) Lipid (% wet wt) 26.71 (13.31) 13.81 2.61 21.96 (9.00) 11.69 2.03 35.27(11.47) 13.41 2.76 30.19 (10.21) 13.47 2.68 No. of stations 27 12 25 17 indicates value(s) significantly greater than those of other regions (P < 0.05, f-test). 169 FISHERY BULLETIN: VOL. 84, NO. 1 Figure 12— Calanus pacificus females. Average individual wet weight in mg. Calanus pacificus Individual Wet Weight (mg) 1.34-1.54 1.14-1.33 0.94-1.13 0.74-0.93 //.Point Conception San Diego 38° 36° 34c 32' 124° 122' 120° 118° Calanus pacificus V San Francisco ■i-t Monterey ": Bay Individual Wet Weight (mg) 0.57-0.68 0.48-0.56 0.39-0.47 0.27-0.38 ///.Point Conception San Diego 38c 36c 34c 32< 124' 122' Figure 13.— Calanus pacificus Stages V copepodites. Average individual wet weight in mg. 120< 118c 170 WILLASON ET AL.: ZOOPLANKTON IN CALIFORNIA CURRENT Calanus pacificus °. San Francisco ::'. Monterey HI Bay Lipid yiLg/copepod > 40 30-40 20-29 10-19 4 Point Conception San £/.■;. Diego 38c 36c 34c - 32c Figure 14— Calanus pacificus females. Average lipid content per copepod in Hg- 124< 122' 120< 118c Figure 15.— Calanus pacificus Stage V copepodites. Average lipid content per copepod in ytg. Calanus pacificus V * /] UJiHn|l PJf W$r Lipid 'ml: Francisco ^.g/COpepod V^m//.': Monterey \).\};:i; Bay III >30 20-30 10-19 <10 P « y r T T ffilTmlt- v/- Point Conception ~ 1 ■ VI lnw::y ' • J LLnulnllfT IfllliinilSi-'-. •• < i,j yjj pi pjm * ^i m n ;:::. San )&::. Diego 3 l i R f < in/ 4 « 4 LI | » J-r-M-U s 1 • > /T | s o ■ hh:;:;-' 1 1 1 1 o 1 1 1 1 1 1 38< - 36c 34< /••^ 32° 124c 122< 120° 118e 171 FISHERY BULLETIN: VOL. 84, NO. 1 Lipid content of female C. pacificus, expressed as percent dry weight or percent wet weight, was lowest in the southern offshore region, but was quite similar between the other three regions (Table 7). Lipid con- tent (percent dry or wet weight) of stage V copepo- dites from the northern nearshore region was higher than the other regions. This stage showed the lowest lipid content in the southern offshore region (Table 7). DISCUSSION Upwelling was taking place along the California coast during April 1981. The resulting coastal low surface water temperatures were most evident in the northern part of the sampling grid, especially just north of Point Conception. An upwelling index calcu- lated for this region during mid-April was higher than the 20-yr mean (Howe et al. 1981). The cold- water plume extending into the Southern Califor- nia Bight (Fig. 2) is a common phenomenon that occurs when cold, upwelled water from the Point Conception region becomes entrained into the south- ward flowing California Current (Reid et al. 1958; Bernstein et al. 1977; Lasker et al. 1981). The distri- bution of phy toplankton biomass (estimated by sur- face chlorophyll a) was the most obvious biological feature associated with coastal upwelling. Phyto- plankton patchiness in turn influenced zooplankton biomass and nutritional parameters. The following discusses 1) the relationships between various biol- ogical properties influenced by upwelling and 2) the persistence and consequences of biological meso- scale patchiness within the California Current System. The distributions and abundances of both euphau- siid species were similar to previous reports (Brin- ton 1962, 1967b, 1976, 1981; Brinton and Wyllie 1976; Youngbluth 1976). Euphausia pacifica is gen- erally more abundant than Nematoscelis difficilis, and the center of its distribution is located closer to the coast. The abundance of E. pacifica within the sampling grid was positively correlated with phy- toplankton biomass, as has been noted by Young- bluth (1976). Other herbivorous euphausiids (eg., Thysanoessa raschii and T. inermis) also show this same relationship (Sameoto 1976). The distribution and abundance of Calanus paci- ficus stages were also similar to previous reports (Fleminger 1964; Longhurst 1967). Both females and stage V copepodites were most abundant close to the coast near upwelling regions. In contrast to E. paci- fica, abundances of the two C. pacificus stages show- ed rather poor (but significant at 95% level) correla- tions with phytoplankton biomass (r values of 0.24 and 0.31). This result was surprising since both species are considered herbivores. The weak corre- lations between C. pacificus abundance and phyto- plankton standing crop probably resulted from small-scale heterogeneity and poor mobility of the C. pacificus population. Populations of C. pacificus along the California coast show a great deal of small- scale patchiness on the order of 10's to 100's of meters (Mullin and Brooks 1976; Star and Mullin 1981; Cox et al. 1982). Grazing by copepods within these patches can greatly reduce the local phyto- plankton standing crop. When samples are taken on scales of 1 km or less, a poor or inverse correlation between phytoplankton and zooplankton biomass results (Mackas and Boyd 1979; Star and Mullin 1981). Zooplankton samples in this study were col- lected from net tows that covered distances of about 1 km or less. Thus, the poor correlations in the present study confirm results of previous studies and can be explained on the basis of the sampling procedure Laminarinase activity (LA) of C. pacificus and E. pacifica was positively related to phytoplankton standing crop. However, a strong relationship be- tween these variables did not exist for either species (correlation coefficients between 0.53 and 0.62). These results were expected because, although most studies agree that zooplankton digestive enzyme ac- tivity and feeding rates are closely linked, enzyme levels do not always represent instantaneous inges- tion rates nor are they always related to the food en- vironment at the time of collection (Head and Con- over 1983; Hassett and Landry 1983; Head et al. 1984; Willason and Cox in press). We propose three, non-exclusive explanations for the observed weak correlations between LA and phy- toplankton biomass. First, time lags of 1 to 7 d in the response of zooplankton digestive enzymes to changing food concentrations (Mayzaud and Poulet 1978; Cox and Willason 1981; Willason 1983) can in- fluence the association between enzyme levels and the food environment. Because the standing stock of phytoplankton is often very patchy and can change rapidly, especially in upwelling regions, zooplankters are probably continually acclimating to new condi- tions and an equilibrium may seldom be reached between enzyme activity, feeding rates, and food concentration. Second, phytoplankton concentration may occa- sionally be high in terms of chlorophyll a, but poor in quality resulting in low consumption rates and low digestive enzyme activity. Herbivorous zooplankton feeding rates have been shown to be greatly de- 172 WILLASON ET AL.: ZOOPLANKTON IN CALIFORNIA CURRENT pressed by the presence of unpalatable or toxic phytoplankton (Fielder 1982). Third, recent evidence indicates that zooplankton digestive enzymes do not show a substrate-specific response. Head and Conover (1983) found that LA in C. hyperboreus was induced in animals which were fed an algae that did not contain laminarin. Willa- son (1983) found that levels of laminarinase in E. pacifica increased when animals consumed small, nonreactive charcoal particles. This increase in ac- tivity, however, was less than that of animals given phytoplankton as a food source Hence, some types of nonphytoplankton food, such as detrital particles or fecal pellets, may also elicit a positive digestive enzyme response. However, since E. pacifica and C. pacificus are primarily herbivorous and are found close to the coast where phytoplankton is abundant, LA of these zooplankters is probably, for the most part, controlled by phytoplankton consumption. Because of large-scale patchiness within the sam- pling grid, relationships between the various biol- ogical properties are much clearer when stations were grouped and regions or mesoscale features compared. Mesoscale patches (100 to 200 km) of C. pacificus and E. pacifica with high LA values were clearly associated with areas of highest phytoplank- ton standing crop: south of San Francisco Bay and particularly in the area adjacent to and just south of Point Conception. Although laminarinase levels may not always accurately represent the feeding con- ditions at a single station (because of the reasons stated above), large-scale comparisons indicate that digestive enzyme levels of herbivorous zooplankton are stongly influenced by overall food concentration within an area. This suggests that animals near the coastal upwelling regions were feeding at higher rates than animals from other areas of the sampling grid. In contrast to E. pacifica, neither the abundance nor the LA of N. difficilis were correlated with phy- toplankton standing crop. These differences between the two euphausiid species are due most likely to dif- ferent feeding modes or different food preferences. Nematoscelis difficilis, unlike E. pacifica and C. paci- ficus, is probably not a herbivore Nemoto (1967) con- cluded that its mouthparts were very different from those of most herbivorous euphausiids, and Willason and Cox (in press) found that phytoplankton was only a small part of the diet of N. difficilis. What is puzzling, however, are the high levels of LA we some- times found in N. difficilis, a range of values similar to those of E. pacifica. Laminarinase levels in N. dif- ficilis are apparently controlled by consumption of a food source other than phytoplankton. Since we did not examine the gut contents of AT. difficilis nor quantify potential food other than phytoplankton, the type of food eaten by N. difficilis could not be determined. Based on the weight and biochemical composition of C. pacificus, the areas of high feeding activity along the California coast appear to have been per- sistent for periods of at least 1 to 2 wk. Calanus pacificus from the northern nearshore region and from the area near Point Conception were heavier, had a lower water content, and a higher lipid con- tent than copepods from other areas. This indicates that these copepods have had prolonged exposure to better feeding conditions. The use of zooplankton biochemical composition and weight as indices of relative "physiological" or "nutritional" state has been documented in laboratory experiments. Vidal (1980) showed a direct relationship between food con- centration and weight of adult and stage V C. paci- ficus. Since C. pacificus completes a life cycle in about 30 d (Vidal 1980; Huntley and Brooks 1982) and has a fixed number of molts to maturity, 1 or 2 wk at higher food concentrations can have a large impact on adult size The lipid content of a zooplank- ton species represents an energy reserve and is an excellent indicator of nutritional state Lipid content increases in well-fed animals and decreases in starved animals (Lee et al. 1970, 1971; Mayzaud 1976; Hakanson 1984). During periods of starvation, crustaceans in the laboratory also show an increase in water content (Hiller-Adams and Childress 1983). Two field studies have shown that changes in food quality and quantity can cause physiological or nutri- tional changes in zooplankton populations (Omori 1970; Boyd et al. 1978). In both of these cases, zoo- plankters were displaced from their optimal habitat to areas of lower food concentration by currents or eddies. The displaced zooplankters showed a lower lipid content and a higher water content presumably due to suboptimal nutrition. This may be what hap- pened to individuals of C. pacificus in the offshore areas of the California Current. These copepods weighed less and were in poorer physiological con- dition (high water content and low lipid content) than C. pacificus located close to the upwelling regions. Although the origins of these copepods are not known, physical processes within the California Cur- rent System such as eddy extensions (Bernstein et al. 1977; Pelaez and Guan 1982; Haury 1984) or off- shore surface transport mechanisms (Parrish et al. 1981) could displace zooplankters such as C. paci- ficus to the food-poor offshore waters. Because euphausiids were captured at only about 173 FISHERY BULLETIN: VOL. 84, NO. 1 one-half of the stations, comparisons of weight and water content between specific regions were diffi- cult. Although the average weight of adults of both euphausiid species were greater in the northern area (nearshore and offshore combined), water content of both species was similar in all areas. The weight and biochemical composition of adult euphausiids may be less susceptible to short-term changes in food concentration than copepods because of their larger size and longer life cycle (>1 yr, Ross 1982). Thus far, it is apparent that processes which oc- cur in relatively small areas along the California coast, in particular the area near Point Conception, have a considerable influence on the nutritional state of two common herbivorous zooplankters, E. paci- fied, and C. pacificus. What are the long-term im- plications of this mesoscale patchiness? The regions of high phytoplankton standing crop found in April 1981 appear to be relatively predict- able from year to year. Although upwelling events in these areas are episodic and seasonal, previous studies have shown similar patterns. CalCOFI sur- veys (Owen 1974) and recent satellite imagery (Smith and Baker 1982; Pelaez and Guan 1982) indicate that in past years Point Conception and the area off Monterey Bay have consistently been regions of high phytoplankton production during the spring and summer months. This enhanced production has un- doubtedly influenced zooplankton populations in pre- ceding years in much the same way that was found during the present study. Previous investigations concerning zooplankton distributions and grazing ac- tivity along the California coast support this conclu- sion (Fleminger 1964; Brinton 1976, 1981; Cox et al. 1982, 1983). Although reproduction was not estimated, it is like- ly that well-fed zooplankters in the California Cur- rent produce more eggs than poorly fed animals. This has clearly been demonstrated in the laboratory for copepods (Marshall and Orr 1955; Checkley 1980) and has been suggested for euphausiids (Brinton 1976). Larger individuals of a species also produce more eggs (Brinton 1976; Nemoto et al. 1972; Ross et al. 1982). Thus, the larger, better fed copepods and euphausiids near Point Conception and off Mon- terey Bay probably have a higher reproductive out- put than animals from other areas. There is some evidence which suggests that enhanced reproduction of zooplankton takes place near Point Conception. Arthur (1977) noted that the highest densities of copepod nauplii in the Southern California Bight were located in a cold-water upwelling plume extend- ing south from Point Conception. In addition, eggs and larvae of E. pacifica are more abundant in the Southern California region following periods of up- welling (Brinton 1976). In summary, our results show that upwelling and phytoplankton variability have a significant impact on the herbivorous zooplankton in the California Current. Not only did we find patchiness of zooplank- ton abundances, but more importantly, zooplankton nutritional states were also highly variable (i.e, meso- scale and larger scale patchiness of trophic inter- actions). Zooplankton in upwelling regions appear to experience better feeding conditions for periods of up to several weeks. Prolonged periods of better feeding conditions in specific areas should influence secondary production as well. This implies that the relatively small, productive regions along the Cali- fornia coast, south of San Francisco Bay and par- ticularly the area near Point Conception, have a disproportionally large impact on the biology of marine organisms within the California Current System. ACKNOWLEDGMENTS We thank M. Page, T Bailey, L. Haury, D. Morse, and R. Trench for critical review of the manuscript. We also thank P. Smith and the research staff at the Southwest Fisheries Center in La Jolla for support during CalCOFI cruise 8104 and for providing ac- cess to preserved samples. This work was supported by NSF grants OEC 79-9317 and OEC 81-09934 and by the Marine Science Institute at the University of California, Santa Barbara. LITERATURE CITED Authur, D. K. 1977. Distribution, size, and abundance of microcopepods in the California Current system and their possible influence on the survival of marine teleost larvae Fish. Bull., U.S. 75:601-611. Bamstedt, U. 1975. Ecological aspects of individual variations in weight and protein and lipid content of Euchaeta norvegica (copepoda). Sarsia 59:31-46. Bernstein, R. L., L. Breaker, and R. Whritner. 1977. California Current eddy formation: ship, air, and satel- lite results. Science 195:353-359. Bligh, E. G., and W. J. Dyer. 1959. A rapid method of total lipid extraction. Can. J. Bio- chem. Physiol. 37:911-917. Boyd, S. H., P. H. Wiebe, and J. L. Cox. 1978. Limits of Nematoscelis megalops in the northwestern Atlantic in relation to Gulf Stream cold core rings. II. Physio- logical and biochemical effects of expatriation. J. Mar. Res. 36:143-159. Brinton, E. 1962. The distribution of Pacific euphausiids. Bull. Scripps Inst. Oceanogr., Univ. Calif. 8:51-270. 174 WILLASON ET AL.: ZOOPLANKTON IN CALIFORNIA CURRENT 1967a. Vertical migration and avoidance capability of euphau- siids in the California Current. Limnol. Oceanogr. 12:451- 483. 1967b. Distributional atlas of Euphausiacea (Crustacea) in the California Current region. Part I. Calif. Coop. Oceanic Fish. Invest. Atlas, Vol. 5, 275 p. 1976. Population biology of Euphausia pacifica off southern California. Fish. Bull., U.S. 74:733-762. 1981. Euphausiid distributions in the California Current dur- ing the warm winter-spring of 1977-78, in the context of a 1949-1966 time series. Calif. Coop. Oceanic Fish. Invest. Rep. 22:135-154. Brinton, E., and A. W. Townsend. 1981. A comparison of euphausiid abundances from bongo and 1-m nets. Calif. Coop. Oceanic Fish. Invest. Rep. 22:111-125. Brinton, E., and J. G. Wyllie. 1976. Distributional atlas of euphausiid growth stages off southern California, 1953-1956. Calif. Coop. Oceanic Fish. Invest. Atlas Vol. 24, 289 p. Checkley, D. M., Jr. 1980. The egg production of a marine planktonic copepod in relation to its food supply: Laboratory studies. Limnol. Oceanogr. 25:430-446. Childress, J. J. 1977. Physiological approaches to the biology of midwater organisms. In N. R. Andersen and B. J. Zahuranec (editors), Oceanic sound scattering prediction, Vol. 5, p. 301-324. Plenum Press, N.Y. Childress, J. J., and M. Nygaard 1974. Chemical composition and buoyancy of midwater crus- taceans as a function of depth of occurrence off Southern California. Mar. Biol. (Berl.) 27:225-238. Cox, J. L. 1981. Laminarinase induction in marine zooplankton and its variability in zooplankton samples. J. Plankton Res. 3:345- 356. Cox, J. L., L. R. Haury, and J. J. Simpson. 1982. Spatial patterns of grazing-related parameters in Cali- fornia coastal surface waters, July 1979. J. Mar. Res. 40:1127-1153. Cox, J. L., and S. W. Willason. 1981. Laminarinase induction in Calanus pacificus. Mar. Biol. Lett. 2:307-311. Cox, J. L., S. W. Willason, and L. Harding. 1983. Consequences of distributional heterogeneity of Cala- nus pacificus grazing. Bull. Mar. Sci. 33:213-226. Fiedler, P. C. 1982. Zooplankton avoidance and reduced grazing responses to Gymnodinium splendens (Dinophyceae). Limnol. Oceanogr. 27:961-965. Fleminger, A. 1964. Distributional atlas of calanoid copepods in the Califor- nia Current region, Part I. Calif. Coop. Oceanic Fish. In- vest. Atlas, Vol. 2, 313 p. Hakanson, J. L. 1984. The long and short term feeding condition in field- caught Calanus pacificus, as determined by the lipid con- tent. Limnol. Oceanogr. 29:794-804. Hassett, R. P., and M. R. Landry. 1982. Digestive carbohydrase activities in individual marine copepods. Mar. Biol. Lett. 3:211-221. 1983. Effects of food-level acclimation on digestive enzyme activities and feeding behavior of Calanus pacificus. Mar. Biol. (Berl.) 75:47-55. Haury, L. R. 1982. Mesoscale processes: The biological and physical connec- tions. [Abstr.] EOS 63-68. Trans. Am. Geophys. Union. 1984. An offshore eddy in the California Current System Part IV: Plankton distributions. Prog. Oceanogr. 13:95-111. Haury, L. R., J. A. McGowan, and P. H. Wiebe. 1978. Pattern and processes in the time-space scales of plankton distributions. In J. H. Steele (editor), Spatial pat- terns in plankton communities, p. 277-327. Plenum Press, N.Y. Hayward, T L., and E. L. Venrick. 1982. Relation between surface chlorophyll, integrated chloro- phyll and integrated primary production. Mar. Biol. (Berl.) 69:247-252. Head, E. J. H., and R. J. Conover. 1983. Induction of digestive enzymes in Calanus hyperboreus. Mar. Biol. Lett. 4:219-231. Head, E. J. H., R. Wang, and R. J. Conover. 1 984. Comparison of diurnal feeding rhythms in Temora longi- cornis and Centre-pages hamatus with digestive enzyme ac- tivity. J. Plankton Res. 6:543-551. Hiller-Adams, P., and J. J. Childress. 1983. Effects of prolonged starvation on O, consumption, NH/ excretion, and chemical composition of the bathy- pelagic mysid Gnathophausia ingens. Mar. Biol. (Berl.) 77: 119-127. Howe, J. T, D. B. Gibson, T O. Evans, L. Breaker, R. C. Wrigley, and W. W Broenkow. 1981. Biological and physical oceanographic observations per- taining to the trawl fishery in a region of persistent coastal upwelling. NASA Tech. Mem. 81332. Huntley, M., and E. R. Brooks. 1982. Effects of age and food availability on diel vertical migration of Calanus pacificus. Mar. Biol. (Berl.) 71:23-31. Lasker, R. 1975. Field criteria for survival of anchovy larvae: The rela- tion between inshore chlorophyll maximum layers and suc- cessful first feeding. Fish. Bull, U.S. 73:453-462. Lasker, R., and P. E. Smith. 1977. Estimation of the effects of environmental variations on the eggs and larvae of the northern anchovy, Engraulis mordax. Calif. Coop. Oceanic Fish. Invest. Rep. 19:128-137. Lasker, R., and J. R. Zweifel. 1978. Growth and survival of first-feeding northern anchovy larvae (Engraulis mordax) in patches containing different proportions of large and small prey. In J. H. Steele (editor), Spatial pattern in plankton communities, p. 329-354. Ple- num Press, NY. Lasker, R., J. PelAez, and R. M. Laurs. 1981. The use of satellite infrared imagery for describing ocean processes in relation to spawning of the northern an- chovy [Engraulis mordax). Rem. Sens. Environ. 11:439-453. Lee, R. F, J. C. Nevenzel, G. A. Paffenhofer, and A. A. Benson. 1970. The metabolism of wax esters and other lipids by the marine copepod, Calanus helgolandicus. J. Lipid Res. 11: 237-240. Lee, R. F, J Hirota, and A. M. Barrett. 1971. Distribution and importance of wax esters in marine copepods and other zooplankton. Deep-Sea Res. 18:1147- 1165. Longhurst, A. R. 1967. Diversity and trophic structure of zooplankton commu- nities in the California Current. Deep-Sea Res. 14:393-408. Lorenzen, C. J. 1970. Surface chlorophyll as an index of the depth, chlorophyll content, and primary productivity of the euphotic layer. Limnol. Oceanogr. 15:479-480. 175 FISHERY BULLETIN: VOL. 84, NO. 1 Lynn, R. J., K. A. Bliss, and L. E. Eber. 1982. Vertical and horizontal distributions of seasonal mean temperature, salinity, sigma-t, dynamic height, oxygen, and oxygen saturation in the California Current, 1950-1978. Calif. Coop. Oceanic Fish. Invest. Atlas, Vol. 30, 513 p. Mackas, D. L., and C. Boyd. 1979. Spectral analysis of zooplankton spatial heterogeneity. Science 204:62-64. Marsh, J. B., and D. B. Weinstein. 1966. Simple charring method for determination of lipids. J. Lipid Res. 7:574-576. Marshall, S. M., and A. R Orr. 1955. The biology of a marine copepod, Calanusfinmarchicus. Oliver and Boyd., Edinb., 188 p. Mauchline, J., and L. R. Fisher. 1969. The biology of euphausiids. Adv. Mar. Biol. 7:1-454. Mayzaud, P. 1976. Respiration and nitrogen excretion of zooplankton. IV The influence of starvation on the metabolism and the bio- chemical composition of some species. Mar. Biol. (Berl.) 37:47-58. Mayzaud, P., and S. A. Poulet. 1978. The importance of the time factor in the response of zooplankton to varying concentrations of naturally occurring particulate matter. Limnol. Oceanogr. 23:1144-1154. Merchant, D. J., R. H. Kahn, and W. H. Murphy, Jr. 1964. Handbook of cell and organ culture Burgess Publ., Minneap., 269 p. Mullin, M. M., and E. R. Brooks. 1976. Some consequences of distributional heterogeneity of phytoplankton and zooplankton. Limnol. Oceanogr. 21:784- 796. Nemoto, T. 1967. Feeding pattern of euphausiids and differentiations in their body characteristics. Inf. Bull Planktol. Jpn, Commem. No. Y. Matusue, p. 157-171. Nemoto, T., K. Kamada, and K. Hara. 1972. Fecundity of a euphausiid crustacean, Nematoscelis dif- ficilis, in the North Pacific Ocean. Mar. Biol. (Berl.) 14:41-47. O'CONNELL, C. P. 1980. Percentage of starving northern anchovy, Engraulis mordax, larvae in the sea as estimated by histological methods. Fish. Bull., U.S. 78:475-489. Omori, M. 1970. Variations of length, weight, respiration rate, and chem- ical composition of Calanus cristatus in relation to its food and feeding. In J. H. Steele (editor), Marine food chains, p. 113-126. Oliver and Boyd, Edinb. Owen, R. W., Jr. 1974. Distribution of primary production, plant pigments, and secchi depth in California Current 1969. Calif. Coop. Oceanic Fish. Invest. Atlas 20:98-117. 1980. Eddies of the California Current system: physical and ecological characteristics. In D. M. Power (editor), The Cali- fornia Islands: Proceedings of the Multidisciplinary Sympo- sium, p. 237-263. Santa Barbara Mus. Nat. Hist. Parrish, R. H., C. S. Nelson, and A. Bakun. 1981. Transport mechanisms and reproductive success of fishes in the California Current. Biol. Oceanogr. 1:175-203. PELAEZ, J., AND F. GUAN. 1982. California Current chlorophyll measurements from satellite data. Calif. Coop. Oceanic Fish. Invest. Rep. 23: 211-225. Reid, J. L., G. I. RODEN, AND J. G. Wyllie. 1958. Studies on the California Current system. Calif. Coop. Oceanic Fish. Invest. Rep. 6:27-57. Ross, R. M. 1982. Energetics of Euphausia pacifica I. Effects of body car- bon and nitrogen and temperature on measured and pre- dicted production. Mar. Biol. (Berl.) 68:1-13. Ross, R. M., K. L. Daly, and T. S. English. 1982. Reproductive cycle and fecundity of Euphausia pacifica in Puget Sound, Washington. Limnol. Oceanogr. 27:304- 314. Ryther, J. H. 1969. Photosynthesis and fish production in the sea. Science 166:72-76. Sameoto, D. D. 1976. Distribution of sound scattering layers caused by eu- phausiids and their relationships to chlorophyll a concentra- tions in the Gulf of St. Lawrence Estuary. J. Fish. Board Can. 33:681-687. Smith, R. C, and K. S. Baker. 1982. Oceanic chlorophyll concentrations as determined by satellite (Nimbus-7 coastal zone color scanner). Mar. Biol. (Berl.) 66:269-279. Star, J. L., and M. M. Mullin. 1981. Zooplankton assemblages in three areas of the North Pacific as revealed by continuous horizontal transects. Deep- Sea Res. 28A:1303-1322. Steele, J. H. 1978. Some comments on plankton patches. In J. H. Steele (editor), Spatial patterns in plankton communities, p. 1-20. Plenum Press, N.Y. Strickland, J. D. H., and T. R. Parsons. 1972. A practical handbook of seawater analysis. Fish. Res. Board Can., Bull. 167, 310 p. Torres, J. J., and J. J. Childress. 1983. The relationship of oxygen consumption to swimming speed in Euphausia pacifica I. Effects of temperature and pressure Mar. Biol. (Berl.) 74:79-86. Vidal, J. 1980. Physioecology of zooplankton. I. Effects of phytoplank- ton concentration, temperature, and body size on the growth rate of Calanus pacificus and Pseudocalanus sp. Mar. Biol. (Berl.) 56:111-134. Willason, S. W. 1983. Spatiotemporal patterns of grazing and laminarinase activity of zooplankton off the California coast. Ph.D. Thesis, Univ. of California, Santa Barbara, 176 p. Willason, S. W, and J. L. Cox. In press. Diel feeding, laminarinase activity, and phytoplank- ton consumption by euphausiids. Biol. Oceanogr. YOUNGBLUTH, M. J. 1976. Vertical distribution and diel migration of euphausiids in the central region of the California Current. Fish. Bull., U.S. 74:925-936. 176 RHIZOCEPHALAN INFECTION IN BLUE KING CRABS, PARALITHODES PLATYPUS, FROM OLGA BAY, KODIAK ISLAND, ALASKA P. T. Johnson,1 R. A. Macintosh,2 and D. A. Somerton3 ABSTRACT An isolated population of blue king crabs, Paralithod.es platypus, in Olga Bay, Kodiak Island, was sampled quarterly during 1980-81. It was found to contain abnormal mature females with degenerate ovaries and/or no sign of having extruded ova following molt. Histological studies of these females and of males and females collected subsequently in April 1982 showed that rhizocephalan internas (roots) were present in up to 50% of the population. Both males and females were infected, but male gonads and secondary sexual characteristics were apparently unaffected. Presence of the rhizocephalan was strongly related to ovarian abnormalities. Evidence suggests that infected females can molt, but do not extrude or retain embryos. The Olga Bay rhizocephalan is not related to Briarosaccus callosus, which parasitizes several species of Alaskan king crabs, including the blue king crab. Externas of the Olga Bay parasite were not found. The possible relationship of this rhizocephalan to the genus Thompsonia, which has minute multi- ple externa that might be missed during gross examination, and the possibility that the blue king crab is an abnormal host that does not allow development of externas are discussed. Molting, mating, and extrusion of ova occur annually in red king crabs, Paralithodes camtschatica, and biennially in blue king crabs, P. platypus. Because embryos of both species hatch within about 1 yr, empty embryo cases are carried on blue king crabs in the second year (Powell and Nickerson 1965; Sasa- kawa 1973, 1975; Somerton and Macintosh in press). Somerton and Macintosh (1982)4 studied an isolated population of blue king crabs in Olga Bay (Kodiak Island, AK) and found abnormal females that were of mature size but lacked external evidence of having extruded eggs or that had apparently degenerate ovaries. This paper reports results of gross and histological examination of blue king crabs from the aberrant Olga Bay population and from three ap- parently normal eastern Bering Sea populations. A rhizocephalan, which was found only in the Olga Bay crabs, appears to be responsible for the abnormal reproductive pattern. Northeast Fisheries Center Oxford Laboratory, National Marine Fisheries Service, NOAA, Oxford, MD 21654. 2Northwest and Alaska Fisheries Center Kodiak Laboratory, Na- tional Marine Fisheries Service, NOAA, P.O. Box 1638, Kodiak, AK 99615. 3Northwest and Alaska Fisheries Center, National Marine Fisheries Service, NOAA, 7600 Sand Point Way, N.E., Seattle, WA 98115. 4Somerton, D. A., and R. A. Macintosh. 1982. Aspects of the life history of the blue king crab (Paralithodes platypus) in Alaska. Document submitted to the annual meeting of the International North Pacific Fisheries Commission, Tokyo, Japan, October 1982. MATERIALS AND METHODS Blue king crabs in Olga Bay were sampled quarter- ly: spring (March-April 1980), summer (June 1980), autumn (October 1980), and winter (January 1981). Seasonal sample sizes ranged from 155 to 229 crabs, and a total of 422 males and 337 females was ex- amined. Both sexes were measured to the nearest millimeter in carapace length (see Wallace et al. 1949, for measurement). Carapace lengths ranged from 12 to 162 mm for males and 16 to 143 mm for females. Data were taken on external egg clutches of females by relative volume, color of embryos, and presence or absence of eyespots on embryos. Pres- ence or absence of empty embryo cases on non- ovigerous females was also noted. For the purposes of this paper, "oogonia" are stem cells; "oocytes" are developing cells before full maturity; and "ova" are cells that have completed vitellogenesis, have a thick chorion, and are ready for fertilization. "Embryo" refers to an external, fer- tilized, and developing egg or ovum. The entire ovary and a pleopod with attached em- bryos or empty embryo cases (if present) were re- moved from each female considered to be mature or in the prepubertal stadium (>68 mm carapace length (CL)). These were preserved in 10% freshwater (river water) Formalin5 solution buffered with sodium 5Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Manuscript accepted April 1985. FTSHF.RY RTTT.T.KTTN- VOT. 84 NO 1 1986. 177 FISHERY BULLETIN: VOL. 84. NO. 1 borate (10 g/L solution). The wet weight of preserved ovaries was recorded to the nearest g and diameters of a sample of oocytes/ova were recorded to the near- est 0.1 mm using a stereomicroscope. Because many of the ovaries appeared abnormal and could not be classified easily by oogenetic stage, histological examination was undertaken of ovaries and pleopods from the largest sample, collected in January 1981 (Table 1). To provide material for a more detailed examination, the Olga Bay population was sampled again in April 1982, and three ap- parently normal Bering Sea populations of blue king crabs were also sampled (Table 1). Except as in- dicated, tissues taken in these collections included portions of the central nervous system, gut, hepato- pancreas, gills, eyestalks, epidermis, heart, anten- nal gland, bladder, ovary, female pleopods, anterior vas deferens, and, in some cases, testis and hemo- poietic tissue Except for the January samples from Olga Bay (fixed in borate Formalin), all tissues were fixed in Kelly's solution (containing zinc chloride rather than mercuric chloride) for 3-4 d, washed 1-2 h in 50% ethyl alcohol, and stored in 70% ethyl alcohol until being processed by standard histological methods. To provide a basis for comparison, ovaries and pleo- pods of 1 1 female red king crabs collected at Olga Bay, January 1981, and fixed in borate Formalin, and tissues from two blue king crabs collected at Glacier Bay, AK, infected with the rhizocephalan Brian- saccus callosus, and fixed in Helly's solution, were also prepared for histological examination. RESULTS Prevalence of the Rhizocephalan The roots (internas) of a rhizocephalan were asso- ciated with either or both the ovary and the pleopod in 52% of the 104 blue king crab females taken from Olga Bay in January 1981, and with various tissues in 40% of the 15 females and 33% of the 15 males taken from Olga Bay in April 1982 (Table 2). The rhizocephalan was also found in 1 of the 11 red king Table 1.— Origins of blue king crabs examined histologically. Carapace length Location Date Number of specimens (mm) Olga Bay 8-14 Jan. 1981 104 females (ovaries and pleopods) 69-136 Olga Bay 5-9 Apr. 1982 15 males 88-151 15 females 90-128 Pribilof Is. 25 June-3 July 1982 10 males 10 females (plus ovaries and pleopods from an additional 83-155 10 females) 96-145 Pribilof Is. 21 Feb. 1983 10 females 113-137 St. Matthew I. 10-13 July 1983 17 males 68-158 9 females 61-129 St. Lawrence I. 5-11 Sept. 1982 5 males 85-106 5 females 79-104 Table 2.— Rhizocephalans in individual male and female blue king crabs, Olga Bay, Kodiak Island, AK, April 1982. Intensity of infection Degenerate roots Major areas parasitized (in tissue sections) Sex Nerve cord, assoc. bladder Bladder in other areas Gut Gonad Antennal gland Hepato- pancreas Female ±1 +2 + + + + + + + + + 3 + + + + + + + + + + + + + + + + + + + + + + + + + Male ± + + + + + + + + + + + + + + + + + + + + + + + + + + + + '± = light infection; + to + + + 2+ = parasite present. 3+ = present medium to very heavy infection. 178 JOHNSON ET AL.: RHIZOCEPHALAN INFECTION IN BLUE KING CRABS crab females taken from Olga Bay in January 1981. Rhizocephalan externas were never detected. Rhizo- cephalan tissue was not found in any of the 76 blue king crabs collected from the Bering Sea and ex- amined by us. Data on females collected from Olga Bay in January 1981 and April 1982 were combined and then separated into various categories of reproduc- tive condition, based on both histological condition and reproductive features of the ovary and on ex- ternal reproductive features. Females in all cate- gories were further classified by the presence or absence of rhizocephalan infection, as determined histologically (Table 3). The effect of the rhizocephalan on female repro- duction was examined by testing the independence of probable future reproductive success and rhizo- cephalan presence Based on ovarian categories (Table 3), probable future reproductive success was judged as either successful (no degenerating gonadal cells) or unsuccessful (ovary empty or ovary with degenerate gonadal cells). Independence of probable future success and rhizocephalan presence was re- jected for both measures, implying that rhizo- cephalan infestation significantly reduces the prob- ability of future reproductive success (x2 = 16.81, df = 1, P < 0.001 for empty ovary; x2 = 20.41, df = 1, P < 0.001 for ovary with degenerate gonadal cells). Three of the external categories of females (Table 3) represent crabs at different times after extrusion of ova. Embryos begin to develop eyes about 4 mo after extrusion. Hatching occurs slightly more than 12 mo after extrusion. Following hatching, empty embryo cases persist on the pleopod setae until the crab molts again, usually slightly <12 mo later (Somerton and Macintosh in press). Therefore, the Table 3.— Prevalence of rhizocephalan infection in female blue king crabs (>68 mm CL) collected in Olga Bay, Kodiak Island, AK, January 1981 and April 1982. Parasitized Not n % parasitized Ovarian categories Ovary empty 15 71 6 Ovary with gonadal cells1 With some degenerate cells 38 64 21 No degenerate cells 7 18 32 External categories Clean pleopod setae 19 51 18 Ovigerous Uneyed embryos 1 10 9 Eyed embryos 12 48 13 Previously ovigerous 27 59 19 (embryo cases) 'Oocytes and/or ova. generalized time since extrusion for the uneyed, eyed, and empty-embryo-case categories is 0-4 mo, 4-14 mo, and 14-24 mo, respectively. If parasitic at- tacks are random and prevent successful extrusion and embryo attachment, then prevalence of the parasite should be low for females with uneyed em- bryos and should increase with time Independence between prevalence and time since extrusion (using uneyed and empty-embryo-case categories) was re- jected (x2 = 7.79, df = 1, P < 0.01). Females are grasped by males and held in a "pre- copulatory embrace" before molting and mating. Of the 10 grasped females collected January 1981, 5 showed no evidence of previous reproductive activity, and 5 had empty embryo cases. None were infected with the rhizocephalan, although three of the females with empty embryo cases had some degen- erate gonadal cells. Based on the April 1982 sample, which includes males, independence between sex and rhizocephalan presence was not rejected (x2 = 0.14, df = 1, P = 0.75). The rhizocephalan, therefore, does not appear to discriminate by host sex. Presence of the rhizocephalan apparently did not affect the gonads of males. Both infected and non- infected males had numerous spermatophores in the anterior vas deferens. Spermatocytes, some of them dividing, and developing and mature sperm were present in the four crabs whose testes were sampled (one parasitized and three nonparasitized). In the field, we saw no males exhibiting female secondary sexual characteristics. Histological Observations Rhizocephalan roots occupied the hemal spaces of the pleopods, were associated with the exterior of the ovary, and occasionally lay within internal hemal spaces of the ovary of infected females collected in January 1981. Roots were associated with various tissues of males and females collected from Olga Bay in April 1982 (Table 2). Hemal sinuses of the ovary and those abutting the gut, the bladder, and the thoracic ganglia were the most frequently invaded sites. Roots lay within the glia of the thoracic ganglia of one crab, but otherwise were confined to hemal spaces and did not invade tissues. Roots were cylindrical and surrounded by a PAS- positive cuticle of variable thickness (Figs. 1, 3). Cells within the roots usually had large vesicular nuclei, and refractile spherules were sometimes present in the cytoplasm. Usually the roots were tubular, with a defined lumen, and those with large, empty lumens often had a flattened epithelium. Loosely anasto- 179 1 J ■« ■-,:,/'.., :g:'":"-' I . • < Figure 1.— Olga Bay rhizocephalan: Cross sections of roots with occluded lumens. PAS. C, cuticle; S, refractile cytoplasmic spherules. Bar =10 ^m. mosing cells filled the lumen of some tubules, and a defined epithelium was not present in these (Fig. 2). Roots with narrow or occluded lumens often had smaller, denser nuclei in the epithelium, or an addi- tional interior layer or group of cells with small, dense, or condensed nuclei (Fig. 2). The occluded roots may represent the distal, growing portions of the organism. Intensity of infection varied (Table 3). In all of the heavier infections and most of the medium ones, por- tions of the roots were degenerate or necrotic (Fig. 3). Host hemocytes had aggregated in such areas and often had encapsulated the degenerate roots. In heavy infections with many degenerating and necrotic roots, blackened areas, probably due to melanin deposition in the roots, were visible with the naked eye in the tissues. Sometimes hemocytes had invaded the lumens of degenerate and necrotic roots, and other roots had been reduced to amorphous material surrounded by hemocytes (Fig. 3). In all cases, roots of normal appearance were also pres- ent in the same areas. In only one instance were nor- mal roots surrounded by hemocytes (Fig. 2). Prob- st FISHERY BULLETIN: VOL. 84, NO. 1 n ! 0 ilp *>" i 1 I v ** . # i A Figure 2— Olga Bay rhizocephalan: Normal roots, lying in an area invaded by hemocytes. Note variable size of the lumen and one tubule with a group of small, central nuclei and another with anastomosing cells in the lumen (arrows). PAS. H, hemo- cytes; T, tubular roots. Bar = 20 (jm. ably the section had been cut just peripherally to a large area of degenerating roots. Ovaries of 88% (53/60) of parasitized females as opposed to 46% (27/59) of normal females either con- tained no oocytes or had some or all degenerate oocytes (Fig. 4). Figure 5 shows a normal ovary with previtellogenic oocytes. Grasped females all had nor- mal oocytes that were in late vitellogenesis and en- closed by a thick chorion. Of the 10 grasped females, 9 were in the premolt condition, and the 10th, a precocious juvenile 77 mm CL, was in the intermolt. None of the parasitized crabs were in advanced premolt, although some were judged to be in early premolt because the pleopod epidermis was thick- ened, and occasionally a developing epicuticle was present. Excepting the ovary, tissues and organs appeared normal in the parasitized crabs. Whether or not there was reduced lipid storage in the hepatopan- creas was not evident by histological examination of the present series. 180 JOHNSON ET AL.: RHIZOCEPHALAN INFECTION IN BLUE KING CRABS 1 4 V / * % — N,# ^ ^J||L<« - ** -I . u ; *. * .0 * t „• • « Figure 3— Olga Bay rhizocephalan: Degenerating and normal roots. PAS. N, normal tubule; C, cuticle; D, tubules with sloughing epithelium; M, completely necrotic tubule; H. hemocytes. Bar = 0.05 mm. 4 ,>'£#*, 4flWMHBfe*i/r'w . V * 1 1 DISCUSSION The presence of the rhizocephalan in female blue king crabs appears to impair reproductive function. Most parasitized crabs have empty ovaries or ovaries that contain degenerate gonadal cells. We assume that these traits are linked to reproductive failure, although there are also unparasitized crabs within each category. It is not unusual to find a few retained ova— destined to be resorbed— in a normal post- extrusion ovary. Therefore, these crabs are also a source of degenerate gonadal cells. The 2-yr reproductive cycle of the blue king crab might also lead to presence of degenerate gonadal cells that had been produced early in the cycle and had become senescent. This speculation remains to be investi- gated. The increase in the incidence of infection over time in postextrusion crabs also suggests reproductive im- pairment. Not only is the prevalence very low (10%) among females that had recently extruded (with uneyed embryos), it is zero among grasped premolt females that were presumably about to molt, mate, Figure 4— Olga Bay rhizocephalan: Empty ovary of an infected crab. Arrows point to roots of the parasite PAS. Bar = 0.2 mm. 181 FISHERY BULLETIN: VOL. 84, NO. 1 Figure 5.— Normal ovary with oogonia and previtellogenic oocytes. PAS. Same scale as Fig. 4. and extrude. These facts suggest that the rhizo- cephalan might preclude mating and subsequent ex- trusion and attachment of fertilized ova. The external category of reproductive condition we term "clean pleopod setae" would normally be associated with immature crabs. In this study, it con- tained both small females and females of mature size (total size range 69-133 mm CL). The average size at maturity of females in Alaskan populations lack- ing the rhizocephalan ranges from 80 to 96 mm (Somerton and Macintosh 1983). Crabs larger than 114 mm could reasonably be expected to be carry- ing embryos or empty embryo cases, but 10 crabs in the combined January-April sample (9 of which had the rhizocephalan) were not. Two of the para- sitized females were soft-shelled, suggesting that molting can occur in parasitized females. Presence of the rhizocephalan in male crabs from Olga Bay apparently did not interfere with normal gonadal function. Species of Sacculina and many other rhizocephalans cause a varying degree of ex- ternal feminization and gonadal dysfunction of their male hosts (Reinhard 1956). For example, Thomp- sonia mediterranea causes external appendages of males of Callianassa truncata to approach the female condition (Caroli 1931), but a species of Thompsonia parasitizing Portunus pelagicus does not affect males (Phang 1975). Briarosaccus callosus parasitizes the blue, red, golden (Lithodes aequis- pina), and deep-sea {Lithodes couesi) king crabs in the Gulf of Alaska (McMullen and Yoshihara 1970; Somerton 1981; Hawkes et al. 1985). Meyers6 found testicular regression and broadening of the abdomen in Briarosaccus-'mfected male blue king crabs from Glacier Bay. High prevalences of infection with rhizocephalans have been reported previously in other decapod species, so the high prevalence in blue king crabs of Olga Bay is not surprising. McMullen and Yoshihara (1970) found 14 of 21 golden king crabs, captured near Kodiak Island, infected with B. callosus, and Hawkes et al. (1985) reported 76% prevalence of the same species in blue king crabs from Glacier Bay; Phang (1975) reported prevalences between 24% and 68% of Thompsonia sp. in groups of Portunus pela- gicus captured near Singapore; and Perry (1984) said that sometimes over 50% of blue crabs sampled from a single population in the Gulf of Mexico were in- fected with Loxothylacus texanus. Although nearly 800 blue king crabs were sampled from Olga Bay at quarterly intervals, no rhizoceph- alan externas were observed, and the one red king crab female found infected with what appeared to be the same rhizocephalan also lacked an externa. Due to the absence of externas, the Olga Bay rhizo- cephalan cannot be indentified with certainty. Its roots are similar histologically to those of other rhi- zocephalans [Thompsonia (Potts 1915); Sacculina (Fischer 1927; Dornesco and Fischer-Piette 1931); and Peltogaster and Gemmosaccus (Nielsen 1970)], corresponding best with the roots of Thompsonia, which have a thinner cuticle than the others (Potts 1915). Roots of the Olga Bay parasite differ histologically in several ways from those of Briarosaccus callosus. They are of lesser diameter, have a thinner cuticle, lack large peripheral nuclei, often have a large lumen and flattened epithelium, and seldom have the cytoplasmic vacuoles (probably representing lipid storage) that are common in the B. callosus roots. (Compare Figures 1, 2, and 3 with Figure 6.) The Olga Bay parasite and B. callosus also differ in that the roots of B. callosus are a bright green when fresh (Hawkes et al. 1985) and blue- green when fixed in Helly's solution, whereas the roots of the Olga Bay parasite are colorless. 6T. Meyers, Assistant Professor of Fisheries, School of Fisheries and Science, University of Alaska, 11120 Glacier Highway, Juneau, AK 99801, pers. commun. October 1984. 182 JOHNSON ET AL.: RHIZOCEPHALAN INFECTION IN BLUE KING CRABS The lack of obvious externas on the parasitized crabs is puzzling. One possibility is that externas are produced but are inconspicuous and/or evanescent. Most rhizocephalans produce easily detected exter- nas that emerge from the venter of the abdomen. Species of Thompsonia, however, produce multiple small externas 1-4.5 mm long and no more than 1.1 mm in diameter. These externas occur on the ap- pendages and venters of the thorax and abdomen, depending on the species, and those of at least one of the species are easily dislodged (Hafele 1911; Potts 1915; Phang 1975). If few and scattered externas of the Thompsonia type were present, they could have escaped notice on animals as large as the blue king crabs investigated. The second possibility is that externas are not developed in the blue king crab. Host ranges of rhizocephalans are often broad, but some of the host/parasite associations may be acci- dental or not fully evolved. Sacculina carcini is known to react differently in different species of crabs. In Carcinus maenas multiple broods of lar- vae are produced by S. carcini, but if the host is Por- tunus holsatus, it breeds but once and then is shed, which suggests that C. maenas is a natural host but P. holsatus is an adventitious and not entirely com- petent one (Baer 1951). Perhaps the Olga Bay para- site is not a usual parasite of the blue king crab, and although the interna develops extensively and causes severe damage to female gonads, externas cannot be produced in this species. The fact that some roots of the parasite were degenerating or necrotic in most infected crabs suggests that parasites do die within the blue king crab, and that infections might be lost before externas are formed. ACKNOWLEDGMENTS We are grateful to E. Munk, J. Bowerman, and R. Otto of the Kodiak Laboratory for assistance with fieldwork; to S. Meyers, also of the Kodiak staff, for laboratory assistance; to G. Roe and C. Smith of the Oxford Laboratory for preparing tissues for histo- logical examination; to R. Otto for reviewing the manuscript; to T R. Meyers, University of Alaska, Juneau, for providing tissues of blue king crabs in- fected with Briarosaccus callosus; and finally, to Bill Pinnell and Morris Talifson of Olga Bay, without whose logistic support and hospitality the fieldwork would have been twice as difficult and infinitely less enjoyabla # # A • c ■||: c • ♦ i •< • « 1| » FIGURE 6.— Briarosaccus callosus: Roots. Note lack of a central lumen and the very large, peripheral nuclei (arrows). Feulgen. C, cuticle Bar =10 ^m. 183 FISHERY BULLETIN: VOL. 84, NO. 1 LITERATURE CITED Baer, J. G. 1951. Ecology of animal parasites. Univ. Illinois Press, Ur- bana, IL, 224 p. Caroli, E. 1931. Azione modificatrice dei Bopiridi e dei Rizocefali sui caratteri sessuali secondarii delle Callianasse Arch. Zool. Ital. 16:316-322. DORNESCO, G. T., AND E. FlSCHER-PlETTE. 1931. Donnees cytologiques sur les "racines" de la Sacculine, Crustace parasite Bull. Histol. Appl. 8:213-221. Fischer, E. 1927. Sur le tissu constituant les "racines" endoparasitaires de la Sacculine C. R. Soc. Biol. 96:329-330. Ha'fele, F. 1911. Anatomie und Entwicklung eines neuen Rhizocephalen: Thompsonia japonica. Beitrage zur Naturgeschichte Osta- siens. Abh. bayer. Akad. Wiss. Math.-phys. Kl., Suppl.-Bd. 2, Abh. 7, p. 1-25. Hawkes, C. R., T. R. Meyers, and T. C. Shirley. 1985. Parasitism of the blue king crab, Paralithodes platypus, by the rhizocephalan, Briarosaccus callosus. J. Invertebr. Pathol. 45:252-253. MCMULLEN, J. C, AND H. T. YOSHIHARA. 1970. An incidence of parasitism of deepwater king crab, Lithodes aequispina, by the barnacle Briarosaccus callosus. J. Fish. Res. Board Can. 27:818-821. Nielsen, S.-O. 1970. The effects of the rhizocephalan parasites Peltogaster paguri Rathke and Gemmosaccus sulcatus (Lilljeborg) on five species of paguridan hosts (Crustacea Decapoda). Sarsia 42:17-32. Perry, H. M. 1984. A profile of the blue crab fishery of the Gulf of Mexico. Gulf States Mar. Fish. Comm., Spec Publ. 9, 80 p. Phang, V. P. E. 1975. Studies on Thompsonia sp. a parasite of the edible swim- ming crab Portunus pelagicus. Malay. Nat. J. 29:90-98. Potts, F. A. 1915. On the rhizocephalan genus Thompsonia and its rela- tion to the evolution of the group. Pap. Dep. Mar. Biol. Carnegie Inst. Wash. 8:1-32. Powell, G. C, and R. B. Nickerson. 1965. Reproduction of king crabs, Paralithodes camtschatica (Tilesius). J. Fish. Res. Board Can. 22:101-111. Reinhard, E. G. 1956. Parasitic castration of Crustacea. Exp. Parasitol. 5: 79-107. Sasakawa, Y. 1973. Studies on blue king crab resources in the western Ber- ing Sea. I. Spawning cycle [In Jpn.] Bull. Jpn. Soc. Sci. Fish. 39:1031-1037. (Engl, transl. NOAA Lang. Serv. Branch.) 1975. Studies on blue king crab resources in the Western Ber- ing Sea. II. Verification of spawning cycle and growth by tag- ging experiments. [In Jpn.] Bull. Jpn. Soc. Sci. Fish. 41: 937-940. (Engl, transl. NOAA Lang. Serv. Branch.) Somerton, D. A. 1981. Contribution to the life history of the deep-sea king crab Lithodes couesi, in the Gulf of Alaska. Fish. Bull., U.S. 79: 259-269. Somerton, D. A., and R. A. Macintosh. 1983. The size at sexual maturity of blue king crab, Para- lithodes platypus, in Alaska. Fish. Bull., U.S. 81:621- 628. Somerton, D. A., and R. A. Macintosh. In press. Reproductive biology of the blue king crab, Para- lithodes platypus, in the eastern Bering Sea. J. Crustacean Biol. Wallace, M. M., C. J. Pertuit, and A. H. Hvatum. 1949. Contributions to the biology of the king crab Para- lithodes camtschatica (Tilesius). U.S. Fish Wild. Serv., Fish. Leafl. 340, 49 p. 184 NOTES THE SEX RATIO AND GONAD INDICES OF SWORDFISH, XIPHIAS GLADIUS, CAUGHT OFF THE COAST OF SOUTHERN CALIFORNIA IN 1978 In the tropical and subtropical Pacific, swordfish, Xiphias gladius, about to spawn are found through- out the year but are most abundant from March to July (Palko et al. 1981). There is, however, little in- formation on the reproductive potential of swordfish during their summer and autumn migrations into the Southern California Bight, a temperate region encompasing the principal U.S. west coast swordfish fishing grounds. In 1978 scientists from the South- west Fisheries Center collected the gonads of sword- fish harpooned in the Bight (from Point Conception to the United States-Mexico border) in order to determine sex ratios, gonad indices, and the repro- ductive condition of these fish. Methods Ninety swordfish were sampled from 25 August through 20 November 1978. After capture their gonads were preserved in 10% Formalin1 and, in the laboratory, were weighed to the nearest gram and their sex determined visually. Ovarian sections used in the histological analysis were obtained from seg- ments removed from the centers of the ovaries. Seg- ments were imbedded in Paraplast and 8 ^m sections were cut, stained in iron hematoxylin, and counter- stained in eosin. Two gonad indices were calculated for each pair of ovaries to permit comparisons with two existing studies on the sexual maturity of Pacific swordfish. The first (from Uchiyama and Shomura 1974) is simply the percentage of the fresh weight of the ovaries to the total weight of the fish: GI = (W/L3) x 104 (2) r,r WT-0 ^nn GI = - x 100 WT-F (1) where GI = gonad index, WT-0 = fresh weight of both ovaries, and WT-F = fresh weight of whole fish. The second index (from Kume and Joseph 1969) is 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. FISHERY BULLETIN: VOL. 84, NO. 1, 1986. where GI = gonad index, W = fresh weight of both ovaries in grams, and L = post-orbital fork length in centime- ters. Because the gonads used in this study were pre- served, and thus subject to shrinkage and loss of weight, it was necessary to estimate their fresh weight using the relationship (from Uchiyama and Shomura 1974): Y = e In X-0.155 0.969 (3) where Y = estimated fresh weight of ovaries, and X = weight of preserved ovaries. The estimated weight loss due to preservation was as high as 7%. Results and Discussion All 90 swordfish collected were mature with fork lengths ranging from 133 to 218 cm. Of these, 23 (26%) were males and 67 (74%) were females for a sex ratio of 0.34:1 (M:F). Although the proportion of females varied among months, our sample sizes were too small to demonstrate such variation. Female swordfish in our sample all had gonad in- dices that were considerably lower than those of com- parable studies. Uchiyama and Shomura (1974) col- lected 16 pairs of ovaries from swordfish caught near Hawaii and found three pairs to be ripe These had gonad indices (from Equation (1)) of 6.4, 8.4, and 9.8 whereas our highest value (from Equations (1) and (3)) was 1.0. Kume and Joseph (1969) examined 362 pairs of ovaries from swordfish captured in the east- ern Pacific (east of long. 130°W) and found two ripe specimens whose gonad indices (from Equation (2)) were 10.8 and 11.1. By comparison, the highest from our study (from Equations (2) and (3)) was 1.8. These results indicate swordfish in the Southern Califor- nia Bight during our sampling period were not spawning. A histological analysis was performed on a subset of 16 pairs of ovaries from our sample Histological analyses can be used to determine not only if a fish 185 is in spawning condition but, also, if it has recently spawned (Hunter and Macewicz 1985). Ovaries from our sample contained no mature oocytes and, in addition, did not contain abundant atretic oocytes indicative of the resorption process. Instead the ovaries were in the regressed stage and contained primary oocytes lining connective tissue septa. These results indicate that the swordfish were reproduc- tively inactive during the sampling period and for at least a month or two before capture Although this conclusion does not preclude the possibility of spawn- ing early in the year, swordfish then are scarce Also water temperatures favorable for spawning (Palko et al. 1981) are not widespread in the summer and autumn, and are virtually nonexistant the remainder of the year. Acknowledgments The authors are indebted to the cooperating com- mercial swordfish fishermen and the scientific observers, particularly Dimitry Abramenkoff and Lynn Shipley, who conducted field sampling. The comments of Gary Sakagawa, Norm Bartoo, and Pierre Kleiber were greatly appreciated. Literature Cited Hunter, J. R., and B. J. Macewicz. 1985. Rates of atresia in the ovary of captive and wild north- ern anchovy, Engraulis mordax. Fish. Bull., U.S. 83:119- 1. •',<',. Kume, S., and J. Joseph. 1969. Size composition and sexual maturity of billfish caught by the Japanese longline fishery in the Pacific Ocean east of 130 W. Bull. Far Seas Fish. Res. Lab. (Shimizu) 2:115- 162. Palko, B. J., G. L. Beardsley, and W. J. Richards. 1981. Synopsis of the biology of the swordfish, Xiphias gladius Linnaeus. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ. 441, 21 p. UCHIYAMA, J. H., AND R. S. SHOMURA. 1974. Maturation and fecundity of swordfish, Xiphias gladius, from Hawaiian waters. In R. S. Shomura and F. Williams (editors), Proceedings of the International Billfish Sympo- sium Kailua-Kona, Hawaii, 9-12 August, 1972. Part 2. Review and contributed papers, p. 142-148. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF 675. Earl C. Weber Southwest Fisheries Center La Jolla Laboratory National Marine Fisheries Service, NOAA 8604 La Jolla Shores Drive La Jolla, CA 92038 Stephen R. Goldberg Department of Biology Whittier College Whittier, CA 90608 GROWTH OF DOLPHINS, CORYPHAENA HIPPURUS AND C. EQUISELIS, IN HAWAIIAN WATERS AS DETERMINED BY DAILY INCREMENTS ON OTOLITHS The dolphin, Coryphaena hippurus, and pompano dolphin, C. equiselis, are widely distributed pelagic fishes in tropical and subtropical oceans (Beardsley 1967; Rose and Hassler 1968; Shcherbachev 1973). In Hawaiian waters C. hippurus is caught through- out the year, but its abundance fluctuates. Small fish (<2.3 kg) are plentiful in summer and large fish (13.6-18.1 kg) are more abundant from February to April (Squire and Smith 1977). Coryphaena hippurus is important to the commercial and recreational fish- eries; C. equiselis, a smaller fish with a maximum length of 74 cm (Herald 1961), is occasionally caught by recreational fishermen. Although much is known about the life history of C. hippurus in the Atlantic (Palko et al. 1982), the biology of the Hawaiian population has been only sketchily investigated. Lit- tle is known about C. equiselis. At least three age and growth studies on C. hippu- rus have been reported. Annual marks on scales have been used to age C. hippurus off Florida (Beards- ley 1967) and North Carolina (Rose and Hassler 1968) in the western North Atlantic Ocean. Wang (1979) used monthly modal progression of length- frequency distributions to estimate the growth rate of C. hippurus off eastern Taiwan in the western Pacific Ocean. The estimated growth rates of C. hip- purus off Florida and North Carolina differed slight- ly, but the growth rate of C. hippurus in the western Pacific Ocean was reported to be about twice as great as those in the western North Atlantic Ocean. The purpose of this study was to validate estimates of age and growth of larval and juvenile C. hippurus and C. equiselis based on microstructure of otoliths (sagittae) from fish of known age reared in captivity. Otoliths from wild specimens captured in Hawaiian waters were also used as a source of age and growth information and these data were fitted to the von Bertalanffy growth model. Ages of cultured and cap- tured wild specimens were estimated by enumer- ating presumed daily increments on the sagitta following Pannella (1971). The daily nature of the increments was validated by counts from sagittae of fish reared in captivity and whose age was known. Knowledge of growth rates of both species of dolphins are useful to mariculturists who would like to compare the growth rates of wild and cultured individuals. Information on the growth rate of C. hip- purus can also be of use to managers of Hawaiian fishery resources. 186 FISHERY BULLETIN: VOL. 84, NO. 1 Materials and Methods Validation Fertilized eggs of C. hippurus and C. equiselis were obtained between January 1982 and February 1983 from captive broodstock held at the University of Hawaii's Waikiki Aquarium (WA); the Kewalo Re- search Facility (KRF) of the Southwest Fisheries Center Honolulu Laboratory, National Marine Fish- eries Service; and The Oceanic Institute, Waimanalo, HI. Larvae of both species were reared at the WA in 4,000 L circular fiber glass tanks with flow- through water exchange and under shaded natural light condition. Water temperature ranged between 23° and 27°C. Both species were fed an unlimited supply (a density of 1-5/mL) of cultured copepod, Euterpina acutifrons, and Artemia sp. until they were large enough to accept chopped fish and squid (about 30 d after hatching), which were then pro- vided several times during the day. These fish were fed to satiation. One 167-d-old and three 191-d-old C. hippurus were reared at the KRF under similar environmental conditions and feeding regime as at the WA.1 These juvenile C. hippurus were trans- ferred to 8 m diameter tanks when they were about 25 cm long. One to three larvae of C. equiselis were sampled on the day of hatching (D-0), and each day thereafter (D-l, D-2, D-4, etc.). However, after the fourth day, there were few survivors, so only a single specimen was taken at intervals of 4 d from D-l 9. Three lar- vae of C. hippurus were sampled on D-4 and single specimens were sampled at various intervals or ob- tained after accidental deaths for validating the growth increments. Other larvae were sampled from other batches on D-0, D-l, and D-2 for measure- ments. Specimens were sampled around noon. Total length of the larvae was measured under a micro- scope with an ocular micrometer while the specimen was alive or within an hour after death. To facilitate measurement, each larva was put on a glass slide, extended to its full length, and measured. For the examination of otoliths, the larva on the slide was immersed in 70% ethanol and allowed to fix for an hour. The larva was then removed from the ethanol bath, blotted, and mounted in Euparal,2 a water solu- ble mounting medium, and covered with a cover slip. Otoliths could be examined in the squashed whole mount without extracting them. After measuring the fork length of juvenile and adult dolphins with a caliper to the nearest milli- meter, otoliths were extracted, cleaned, and mounted whole To extract the otoliths, the head was removed from the body, and the flesh removed from the head to expose the skull. With a saw or knife, most of the supraoccipital and roof of the skull were removed. After careful removal of the brain, the sagittae (largest of the three otoliths) could be found in the sacculi located anteriorly on the right and left sides of the first vertebra at the caudal end of the brain cavity. Under a dissecting microscope, the sagitta was teased out of the sacculus, and extraneous tissues were brushed away. The pair of sagittae was then placed on a clean glass slide, permitted to dry, and mounted in Euparal. Segments of monofilament line slightly thicker than the sagittae were placed on both sides of the sagittae to prevent the cover slip from crushing it. After clearing for a month, presumed daily incre- ments on a sagitta were enumerated using a com- pound binocular microscope with transmitted light at 600 x magnification. Increments were counted starting from the core out to the edge of the post- rostrum, or from the core to the tip of the rostrum. Usually, counts could not be made in a direct line from the core to the edge of the rostrum or post- rostrum of the sagitta; rather, a somewhat circuitous route was taken from one area of the sagitta to another by following a prominent growth increment. Increments were also counted inward from the edge to the core. Two independent age estimations were made separately on the rostrum and postrostrum on a sagitta to verify the age of fish. In some samples, it was possible only to make a single age estimate since the sagitta was incomplete, having just a rostrum or postrostrum. The reader had no infor- mation such as specimen size or previous counts to prevent bias in the counting. The arithmetic mean of 3-14 counts was used to estimate a fish's age The number of counts from the rostrum and postrostrum varied from as few as 3 for a larva to 14 for a sagitta of a juvenile The rela- tionship between counts of otolith increments and days was assessed for both species by regression analysis. 'Thomas K. Kazama, Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, Honolulu, HI 96812, pers. commun. October 1984. 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Growth of Wild Specimens Juveniles of both species were dip netted from Kaneohe Bay, HI. Large juveniles and adults of both species were obtained from private and chartered 187 sport fishing boats in Honolulu, and C. hippurus specimens were also obtained from cruises of the NOAA ship Townsend Cromwell to the Northwestern Hawaiian Islands from October 1976 to September 1981. Fork lengths were measured to the nearest millimeter with calipers. The extraction and slide preparation of sagittae, and counting method were the same as described for the validation experiments above But before reading the sagittae of fish caught in the wild, the sagitta of a known-age fish was re- examined to review the difference between known daily increments and subdaily increments. Concen- tric daily increments, which consist of an inner light band and an outer dark band, were distinguished from subdaily increments by carefully focusing to the plane of maximum clarity. The dark band of the subdaily increment appeared less defined than the dark band of daily increments. Misinterpretation and counting subdaily increments as daily increments could result in an overestimation of aga The mean of 10-20 counts was used as the age estimate of older fish. Age estimates of wild fish were fitted to the von Bertalanffy growth model using NLIN Procedure, a nonlinear regression routine (SAS Institute 1982). The three juvenile C. hippurus whose sex was un- determined were added to both the male and female groups when fitting the curves. Results Validation Fertilized eggs of C. equiselis and C. hippurus began to hatch after 48-50 h at 24°-25°C and all hatched within 2 h. The larvae of both species were 4.0-4.6 mm TL and had two pairs of otoliths, the sagitta and lapillus, at time of hatching. Otoliths of C. equiselis and C. hippurus on D-0 ranged from 16 to 20 nm in diameter and consisted of the core and primordium. An hour after hatching, the larvae were from 5.2 to 5.4 mm TL but did not grow during the next 3 d and even shrank from 0.1 to 0.2 mm. Oto- liths of both species on D-l had a dark ring near the edge which the otoliths of D-0 larvae did not have and were 22-24 ^m in diameter. The sagittae of both species on D-4 had four increments (Fig. 1) and were now slightly larger than the lapillus. Sagittal diameters were 29-36 fim for C. equiselis and 34-41 ^m for C. hippurus. Mean counts of growth increments on the sagit- tae of 10 C. hippurus (Table 1) and 13 C. equiselis Figure 1— Sagitta of a day-4 Coryphaena hippurus larva. Diameter of sagitta is 17 ^m. 188 Table 1. — Mean of counts on known age sagittae of Coryphaena hippurus. Table 2.— Mean counts on known age sagittae of Coryphaena equiselis. Mean Total Fork Mean Total Fork Known increment No. of length length Known increment No. of length length age counts SD counts (mm) (mm) age counts SD counts (mm) (mm) 0 0 5.3 0 0 4.0 1 1 — 1 1 0.00 3 — 4 4 0.00 3 6.7 1 1 0.00 3 — 4 4 0.00 3 6.8 4 3 0.00 3 4.6 4 4 0.00 3 6.8 4 4 0.00 3 5.2 20 20.0 + 1.26 5 — 19 19 0.00 3 14.2 35 33.6 + 2.06 9 — — 23 23.6 + 2.23 8 23.0 47 45.2 + 3.16 10 — 95.0 27 21.5 + 1.59 14 25.2 167 166.8 + 7.14 11 — 383.0 31 31.7 + 1.38 7 — 29.5 191 190.3 + 6.92 6 — 510.0 36 35.3 + 2.45 10 — 48.0 191 191.0 + 0.71 4 — 554.0 51 51.7 + 2.13 13 — 82.0 191 192.8 + 7.44 5 — 491.0 52 53.6 + 4.81 14 — 72.0 63 63.3 + 3.19 14 — 89.0 63 63.4 ±8.59 13 — 112.0 (Table 2) were plotted against corresponding known ages (Figs. 2, 3). The relationships of mean incre- ment counts (7) to known age (X) were Y = -0.5295 + 1.0035X(r = 0.999, P < 0.01) for 10 C. hippurus and Y = -0.6986 + 1.0164X(r = 0.997, P < 0.01) for 13 C. equiselis. These results demonstrated that growth increments are formed daily, and validated their use for aging wild fish up to 191 d for C. hip- purus and 63 d for C. equiselis. Growth of Wild Specimens Because of sexual dimorphism, separate von Ber- talanffy growth parameters were calculated for male and female C. hippurus (Table 3). The male and female von Bertalanffy growth curves and 18 age- length relationships of C. hippurus are shown in Table 3.— Von Bertalanffy growth parameters calculated from cap- tured wild specimens of Coryphaena hippurus. Sex Number Parameter Estimate SE Male Female 10 fn 0.0790 yr 0.0305 K 1.1871 0.5218 i-oo 189.9301 cm FL 48.9702 'n 0.0731 yr 0.0126 K 1.4110 0.2454 L„ 153.2676 cm FL 14.2168 200 DAYS Figure 2— Validation of daily increments on sagittae of Cory- phaena hippurus by relationship of known age (X) to mean incre- ment count (Y) up to 191 d (r = 0.999). FIGURE 3— Validation of daily increments on sagittae of Cory- phaena equiselis by relationship of known age (X) to mean incre- ment count (Y) up to 63 d (r = 0.997). 189 Figure 4. A single set of growth parameters (Table 4) was calculated for C. equiselis since the largest specimen in the sample had just reached sexual maturity, and the calculation of separate growth curves by sex was not warranted. The von Berta- lanffy growth curve and 13 age-length relationships of C. equiselis are shown in Figure 5. Discussion Validation A pair of otoliths was present at the time of hatch- ing for both dolphins, and the first increment was formed on the otoliths on D-l, identical to Kat- suwonus pelamis, another tropical pelagic species (Radtke 1983). The strong correlation of mean incre- ment counts of sagittae to known age of fish validated the use of growth increments in the aging of C. equiselis up to 63 d and C. hippurus up to 191 d. Since regular incremental formation began on D-l, no adjustment is required to the incremental counts 140 120 100 - 80 I I- o z Id cc O 60 40 20 / ^ MALES / / / t\ ° / / FEMALES // / / Q IMMATURE (N =3) o FEMALES (N =8) A MALES (N =7) ^^= VALIDATED — == UNVALIDATED 0 I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 ESTIMATED AGE (MONTHS) Figure 4— Von Bertalanffy growth curves of male and female Cory- phaena hippurus in Hawaiian waters. Table 4.— Von Bertalanffy growth parameters calculated from captured wild specimens of Coryphaena equiselis. Number Parameter Estimate SE 13 K 0.0648 yr 2.1734 61.3914 cm FL 0.0131 0.9750 17.8000 of wild fish sagittae to estimate age Ideally, valida- tion of daily increments should cover 1) the time when the first daily increment is formed, 2) the regularity in the formation of increments in all stages of life, and 3) events such as spawning, migra- tion, and periods of starvation which may affect the regularity of increment formation. Having achieved only part of these requirements, validation of daily increments on otoliths should continue as older known-age specimens become available, and the ef- fects of spawning and starvation on increment for- mation should also be examined. Growth of Wild Specimens The plot of age-length relationships of male C. hip- purus showed that there was at least one extreme variant. This 111.0 cm FL male greatly affected the growth curve, resulting in a lower estimated L^ and causing most of the male age estimates to fall below the growth curve (Fig. 4). Thus, age-length relations of wild C. hippurus should be examined further to shed light on the extent of variation in size at given ages. Additional age determinations might also im- prove the confidence intervals of the von Bertalanffy growth parameters. Growth rates of C. hippurus to age 1 around Hawaii appeared to be greater than those reported from the western North Atlantic Ocean. Beardsley (1967) reported a mean length of 72.5 cm in age group 1 for C. hippurus off Florida. Rose and Hassler (1968) reported a mean length of 65.3 cm at the end of 1 yr for fish off North Carolina. Around Hawaii male C. hippurus were estimated to attain 40 30 £ 20 cc O 10 "l r VALIDATED 2 3 4 5 ESTIMATED AGE (MONTHS) Figure 5.— Von Bertalanffy growth curve of Coryphaena equiselis in Hawaiian waters derived from 13 age estimates. 190 a length of about 126 cm at 1 yr and about 112 cm for females. The slower growth rate of C. hippurus in the western North Atlantic Ocean may be the result of a decrease in feeding rate when water temperature goes below 23.0°C and a cessation of feeding at 18.0°C (Hassler and Hogarth 1977). Cory- phaena hippurus feed throughout the year in Hawaii and can be expected to grow continuously. Wang (1979) used the monthly progression of modes in length-frequency distributions to estimate growth rates of about 10 cm/mo from February through June for C. hippurus between 50 and 100 cm FL. This growth rate is similar to that found for C. hippurus in Hawaiian waters. Growth rates of captive C. hippurus were similar to those of wild fish in Hawaiian waters. Beardsley (1967) reported rapid growth rates of three captive C. hippurus. These fish grew from about 35 to 125 cm in 7 to 8 mo.3 Soichi (1978) reported that 11 C. hippurus 35-50 cm TL grew to a mean 123 cm TL in 7-8 mo. Their observations also support our estimates of rapid growth for C. hippurus around Hawaii. Coryphaena equiselis appeared to grow as rapid- ly as C. hippurus during the first 4 mo, then grew at a slower rate (Fig. 5). At about 4 mo, C. equiselis reached sexual maturity. Coryphaena hippurus also reached sexual maturity at 4-5 mo, but have been observed to mature as early as 3 mo in captivity. The daily regularity of increment formation has been demonstrated from D-l to D-191 for C. hip- purus and from D-l to D-63 for C. equiselis. So the use of daily increment counts on the sagitta of wild fish for estimating age has only been partially validated for these dolphins. The age-length relation- ships are valid for the first 6 mo for wild C. hippurus and the first 2 mo for wild C. equiselis. Thus, the von Bertalanffy growth curves calculated for wild C. hippurus in Hawaiian waters should be viewed with caution despite good agreement with several other growth observations in the literature Acknowledgments Our thanks to Richard W. Brill, Richard E. Brock, Leighton R. Taylor, and Jerry A. Wetherall for their critical reviews of this manuscript. Carol Hopper greatly assisted our sampling efforts for wild-caught specimens and Thomas K. Kazama provided the oldest known-age C. hippurus. 3 A length-weight relationship (Gibbs and Collette 1959) was used to estimate lengths in centimeters from weights, given in pounds, by Beardsley (1967). Literature Cited Beardsley, G. L., Jr. 1967. Age, growth, and reproduction of the dolphin, Cory- phaena hippurus, in the Straits of Florida. Copeia 1967: 441-451. Gibbs, R. H., Jr., and B. B. Collette. 1959. On the identification, distribution, and biology of the dolphins, Coryphaena hippurus and C. equiselis. Bull. Mar. Sci. Gulf Caribb. 9:117-152. Hassler, W. W., and W. T. Hogarth. 1977. The growth and culture of dolphin, Coryphaena hip- purus, in North Carolina. Aquaculture 12:115-122. Herald, E. S. 1961. Living fishes of the world. Doubleday and Co., Inc., Garden City, NY, 304 p. Palko, B. J., G. L. Beardsley, and W. J. Richards. 1982. Synopsis of the biological data on dolphin-fishes, Cory- phaena hippurus Linnaeus and Coryphaena equiselis Lin- neaus. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ 443, 28 p. [Also FAO Fish. Synop. 130.] Pannella, G. 1971. Fish otoliths: Daily growth layers and periodical pat- terns. Science (Wash., D.C.) 173:1124-1127. Radtke, R. L. 1983. Otolith formation and increment deposition in labora- tory-reared skipjack tuna, Euthynnus pelamis, larvae. In E. D. Prince and L. M. Pulos (editors), Proceedings of the Inter- national Workshop on Age Determination of Oceanic Pelagic Fishes: Tunas, Billfish.es, and Sharks, p. 99-103. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 8. Rose, C. D., and W. W. Hassler. 1968. Age and growth of the dolphin, Coryphaena hippurus (Linnaeus), in North Carolina waters. Trans. Am. Fish. Soc 97:271-276. SAS Institute. 1982. SAS user's guide: Statistics. SAS Institute Inc., Cary, NC, 584 p. Shcherbachev, Yu. N. 1973. The biology and distribution of the dolphins (Pisces, Coryphaenidae). [In Russ.] Vopr. Ikhtiol. 13:219-230. (Engl, transl. in J. Ichthyol. 13:182-191.) Soichi, M. 1978. Spawning behavior of the dolphin, Coryphaena hip- purus, in the aquarium and its eggs and larvae [In Jpn., Engl, summ.] Jpn. J. Ichthyol. 24:290-294. Squire, J. L., Jr., and S. E. Smith. 1977. Anglers' guide to the United States Pacific coast: marine fish, fishing grounds & facilities. U.S. Dep. Commer., NOAA, NMFS, 139 p. Wang, C H. 1979. A study of population dynamics of dolphin fish (Cory- phaena hippurus) in waters adjacent to eastern Taiwan. [In Chin., Engl, abstr.] Acta Oceanogr. Taiwan. 10:233-251. James H. Uchiyama Southwest Fisheries Center Honolulu Laboratory National Marine Fisheries Service, NOAA P.O. Box, 3830, Honolulu, HI 96812 Raymond K. Burch Syd A. Kraul, Jr. Waikiki Aquarium, University of Hawaii 2777 Kalakaua Avenue Honolulu, HI 96815 191 SIZES OF WALLEYE POLLOCK, THERAGRA CHALCOGRAMMA, CONSUMED BY MARINE MAMMALS IN THE BERING SEA In the Bering Sea at least 11 species of marine mam- mals, 13 seabirds, and 10 fishes are known to feed on walleye pollock, Theragra chalcogramma (Frost and Lowry 1981a). Walleye pollock are a major food of most pinnipeds, particularly in the southern Ber- ing Sea (Lowry and Frost 1981), and are sometimes eaten by several species of baleen and toothed whales (Frost and Lowry 1981b). In recent years, walleye pollock have been the prin- cipal target species in the Bering Sea commercial groundfish fishery. Annual catches have been as high as 1,840,000 t in 1972 (Bakkala et al. 1981). While there can be little doubt that both the fishery and marine mammal predation affect pollock stocks and perhaps also one another, the interactions are poorly understood at present (Lowry et al.1; Swartzman and Harr 1983). An important aspect of marine mammal-fishery interactions is the size composition of fishes eaten in relation to that of the commercial catch. For ex- ample, if a marine mammal consumes fishes smaller than those taken by the fishery, the fishery would be unlikely to influence availability of food to the predator unless it affected recruitment. If marine mammals and the fishery remove fishes of similar sizes, competition would be expected (IUCN2). Stomach contents of marine mammals seldom con- tain intact fishes in a condition suitable for mea- suring. However, the sagittal otoliths of species such as walleye pollock are easily identified (Frost 1981), and equations are available that estimate the length and weight of fishes from otolith lengths (Frost and Lowry 1981a). We present here information on the sizes of walleye pollock consumed by marine mam- mals in the Bering Sea, based on otoliths from gastrointestinal tracts. Methods Specimens were collected during the months of March to October 1975-81, at the locations shown in Table 1. With the exception of a minke whale, Balaenoptera acutorostrata, which was stranded on shore, all specimens were from animals collected for scientific purposes. Stomachs were removed and opened, and the contents gently washed on a 1 mm mesh sieve. Otoliths were sorted from other ingesta and identified using the descriptions of Morrow (1979) and Frost (1981). Since fresh walleye pollock otoliths have fine lobulations around their perimeter (Frost 1981) which disappear during digestion, degraded otoliths were easily detected by compari- 'Lowry, L. F., K. J. Frost, D. G. Calkins, G. L. Swartzman, and S. Hills. 1982. Feeding habits, food requirements, and status of Bering Sea marine mammals. North Pac Fish. Manage Counc. Doc. 19 and 19A, Anchorage, Alaska, Contract 81-4, 574 p. 2IUCN. 1981. Report of IUCN workshop on marine mammal- fishery interactions, La Jolla, Calif., 30 March-2 April. IUCN, Gland, Switzerland, 68 p. Table 1. — Location and dates of capture of marine mammals from which otoliths of walleye pollock were obtained. No. of No. of otoliths Species Dates Location specimens measured Harbor seal, 13 Apr. 1979 Otter Island 4 23 Phoca vitulina richardsi 9 Oct. 1981 Port Heiden 1 12 Spotted seal, 6 May 1978 61°42.3N, 175°36.0W 1 11 Phoca largha 23 May 1978 63°25.8N, 173°05.6W 1 10 Ribbon seal, 19-20 Apr. 1976 57°20.1N-57°28.0N 5 256 Phoca fasciata 172°30.9W-173°07.5W 21-22 Mar. 1977 58°51.0N-58°56.0N 172°40.0W-173°08.0W 4 67 5-31 May 1978 61°23.0N-64°39.4N 169°07.0W-176°08.8W 10 145 Steller sea lion, 20 Mar. 1976 56°04.8N, 168°32.9W 1 274 Eumetopias jubatus 13 Apr. 1979 Otter Island 1 6 24 Mar., 59°30.0N-60°11.5N 32 497 10-11 Apr. 1981 176°43.5W-179°55.0W 30 Mar.-4 Apr. 59°08.0N-60°13.0N 56 638 1981 165°45.0E-170°46.0E Minke whale, 5 Aug. 1975 Unalaska Island 1 121 Balaenoptera acutorostrata 192 FISHERY BULLETIN: VOL. 84, NO. 1, 1986. son with those taken from trawl-caught fishes. The maximum length of nondegraded otoliths was measured to the nearest 0.1 mm using vernier calipers. When more than 20 otoliths occurred in a single stomach, a subsample of 20 was measured. Very few otoliths were found in the stomachs of ribbon, Phocafasciata, and spotted, P. largha, seals. For those species, additional otoliths were obtained from small intestines which were split along their entire length and examined for parasitological studies. There was no significant difference between sizes of otoliths obtained from stomachs and intes- tines of ribbon seals (Frost and Lowry 1980). Too few otoliths were retrieved from spotted seal stomachs to test their sizes relative to otoliths from intestines. However, otoliths from intestines were of the same general size range and condition as those from stomachs. We therefore pooled the measurements of otoliths from stomachs and intestines. The fork lengths and weights of walleye pollock consumed were estimated from equations in Frost and Lowry (1981a). Results We measured a total of 2,060 otoliths from 117 in- dividual marine mammals belong to 5 species (Table 1). Most of the otoliths were from the stomachs and small intestines of 19 ribbon seals and 90 Steller sea lions, Eumetopias jubatus. Ribbon seals, spotted seals, and a minke whale fed primarily on walleye pollock <20 cm long (Table 2, Fig. 1). Harbor seals, Phoca vitulina richardsi, fed on a wide size range of pollock, including equal numbers of fishes 8-15 cm and 20-35 cm long and a few individuals 45-56 cm in length. Most pollock eaten by sea lions (76%) were 20 cm or longer. Young sea lions (<4 yr) collected in 1981 (all were males) ate significantly smaller fish (x = 22.4 cm, n = 37) than did older animals (x = 26.9 cm, n = 51; P < 0.005). There were some differences in sizes of pollock consumed at different localities and in different years. The sizes of pollock eaten by harbor seals col- lected at Otter Island in 1979 ranged from 10.3 to 56.3 cm (i = 31.8 cm), while those eaten by a seal collected at Port Heiden in 1981 were all <12.6 cm long (x = 10.6 cm). Two sea lions collected in 1976 and 1979 near the Pribilof Islands had eaten pollock averaging 46.9 cm in length (range 18.4-61.4 cm), while those collected in 1981 to the west had eaten substantially smaller pollock averaging 25.2 cm in length (range 8.3-64.2 cm). In Figure 1, the smaller size mode corresponds to 1981 collections and the larger mode to those from 1976 and 1979. In 1981 sea lions collected in the central Bering had eaten larger pollock than those off the Kamchatka Penin- sula (x = 26.8 cm vs. 23.5 cm; P < 0.001). This was not attributable to different age or size composition of the samples, since the difference was apparent for older sea lions (>5 yr; x = 21.8 cm vs. 25.6 cm; P < 0.01) as well as the samples as a whole, and the mean age and standard length of all sea lions >5 yr in the Kamchatka sample (x age = 9.1 yr, x SL = 297 cm, n = 27) was greater than that of the cen- tral Bering sample (x age = 8.2 yr, x SL = 282 cm, n = 25). Discussion Of the marine mammal species we examined, rib- bon seals, spotted seals, and a minke whale ate almost exclusively small pollock, whereas Steller sea lions and harbor seals ate pollock of a wide range of sizes. There are few other data available on the sizes of pollock consumed by marine mammals in the Bering Sea. Nemoto (1959) indicated that the length of pollock eaten by fin whales, Balaenoptera physa- lus, never exceeded 30 cm, while larger pollock were sometimes eaten by humpback whales, Megaptera navaeangliae. Fiscus et al. (1964) reported that in 1962 northern fur seals, Callorhinus ursinus, ate mostly whole pollock <30-35 cm long. McAlister et al.3 found intact pollock in fur seal stomachs collected in the eastern Bering Sea, July- September 1974, to range from 10 to 35 cm, with a mean length of 19.3 cm. Most specimens were between 16 and 21 cm long. In 1981, Loughlin4 collected fur seals north of Unalaska Island and found the average size of pollock consumed to be 30.4 cm. Antonelis5 found that bearded seals, Erignathus barbatus, collected near St. Matthew Island in the central Bering Sea had eaten only small pollock (x length = 8.2 cm). It is unknown whether the consumption patterns described above are a result of actual size selection of prey or if they result from coincidental distribu- tion of predators and prey size classes. The overall density of pollock and distribution by age classes are far from uniform in the southern Bering Sea (Smith 1981; Bakkala and Alton6). The sizes of fishes con- 3McAlister, W. B., G. A. Sanger, and M. A. Perez. 1976. Pre- liminary estimates of pinniped-finfish relationships in the Bering Sea. Unpubl. background paper, 19th meeting North Pac. Fur Seal Comm., Moscow, 1976. 4T. R. Loughlin, National Marine Mammal Laboratory, 7600 Sand Point Way N.E., Seattle, WA 98115, pers. commun. November 1983. 5G. Antonelis, National Marine Mammal Laboratory, 7600 Sand Point Way N.E., Seattle, WA 98115, pers. commun. December 1983. 6Bakkala, R., and M. Alton. 1983. Evaluation of demersal trawl survey data for assessing the condition of eastern Bering Sea 193 Table 2.— Summary of sizes of walleye pollock consumed by marine mammals in the Bering Sea. Size of walleye pollock consumed Marine mammal Fork length height of mean 1Mean weight of species Mean (cm) Range (cm) length fish (g) 8.6 fishes consumed (g) Ribbon seal 11.2 6.5-34.4 11.2 Spotted seal 10.9 8.0-15.0 7.9 8.4 Harbor seal 24.5 8.2-56.3 83.8 174.3 Steller sea lion 29.3 8.2-64.2 140.5 204.3 Minke whale 14.5 11.8-17.5 18.3 18.7 'The weight of the mean length fish does not correspond to the mean weight of fishes consumed due to the exponential nature of the length-weight relationship for fishes and the distribution of lengths of fishes consumed. sumed generally agree with the basic distribution pattern for pollock in that sea lions collected near the continental slope ate many large pollock, while ribbon and spotted seals collected north of St. Mat- thew Island ate almost entirely small pollock. However, concurrent sampling of prey in stomachs and those available in the environment suggest that some selection does occur. Fur seals were found to eat smaller pollock than those caught in otter trawls taken nearby (x length = 30.4 cm in seals, 38.3 cm in trawls), while sea lions appeared to select larger fishes (x length = 29.9 cm in sea lions, 25.5 cm in trawls) (Loughlin fn. 4). Such comparisons must be interpreted with caution since demersal trawl samples underestimate the abundance of young pollock, most of which occur several meters off the bottom (Traynor7). Other information also indicates that marine mam- mals sometimes select fishes of certain size classes. The sizes of arctic cod, Boreogadus saida, caught in otter trawls in the northern Bering Sea were com- pared with the estimated lengths of fishes eaten by spotted and ribbon seals collected in the same area and time period (Frost and Lowry 1980; Bukhtiyarov et al. 1984). While the distribution of trawl-caught fishes was distinctly bimodal, seals ate predominant- ly fishes of the larger size classes. Saffron cod, Eleginus gracilis, eaten by adult white whales, Del- phinapterus leucas, in the Kotzebue Sound region of the southern Chukchi Sea were larger than those eaten by younger animals collected at the same loca- tion on the same dates (Seaman et al. 1982). We ob- tained similar results in this study for young versus old sea lions. Pitcher (1981) found that pollock eaten by sea lions were significantly longer (x = 29.8 cm) pollock. Unpubl. Rep., 43 p. Northwest and Alaska Fisheries Center, NMFS, NOAA, Seattle, WA. 7Traynor, J. J. 1983. Midwater pollock (Theragra chalcogram- ma) abundance estimation in the eastern Bering Sea. Unpubl. Rep., 7 p. Northwest and Alaska Fisheries Center, NMFS NOAA Seattle, WA. 194 than those eaten by harbor seals (x = 19.2 cm; P < 0.001) collected in the same general locations in the Gulf of Alaska. The factors involved in the apparent size selection of prey are poorly known for marine mammals. A strict relationship between the size of predators and the size of their prey is not to be expected in such behaviorally complex and morphologically diverse animals. For example, the prey of ringed seals, Phoca hispida, range in length from 1 cm (euphausiids) to at least 121 cm (wolffish, Anarhichas sp.) (Frost and Lowry 1981c). The largest animal we examined in this study, a minke whale 7.3 m long, ate uniformly small pollock. Age-related differences in sizes of fishes eaten by sea lions and belukha whales are more likely due to morphological and behavioral development than to size relationships per se. Although size may affect a sea lion's ability to catch large pollock, and old sea lions are larger than young ones (i SL = 212 cm for sea lions age 1-4 yr, n = 33 vs. x SL = 289 cm for those >5 yr, n = 52), the size range of pollock eaten by both young and old sea lions was similar. The largest pollock (64 cm) represented in our samples was eaten by a 215 cm long, 3-yr-old sea lion which indicates that physical differences due strictly to predator size are not the sole factor influencing preference for a particular prey siza Aspects of feeding strategy, including size selectivity, are the result of a complex and inter- acting suite of morphological, physiological, and behavioral adaptations which allow an organism to ' gather food in the most efficient manner (Schoener 1971). Size-specific feeding may have important conse- quences for predators. For example, the length of 1-yr-old pollock fluctuates markedly among years, as ! does the numerical abundance of the first year class. In 1976 abundance was low (729 million individuals in the NMFS Bering Sea survey area) and fishes were small (x = 11.6 cm), while in 1974 abundance was high (2,840 million individuals) and fishes were SPOTTED SEALS FISH LENGTH Com) HARBOR SEALS <" = 5) FISH LENGTH Com) MINKE WHALES Cn = l) ga ea in I ^- -J C D U. O 70 ea 5a K 48 3a I 3 z za IB za 30 4a » FISH LENGTH (cm) ea ?a Figure 1— Size distributions of walleye pollock eaten by five species of marine mammals collected in the Bering Sea, 1975-81. ia 2e oe 7e FISH LENGTH Com) 195 considerably larger (x = 15.9 cm) (Smith 1981). The corresponding average individual weights can be estimated as 9.5 and 23.7 g, giving an estimated biomass of age 1 pollock about 10 times greater in 1974 than in 1976. Therefore, the total food available to predators that specialize on small pollock can vary markedly, as can the energy obtained from each fish consumed. Lengths and population sizes of older pollock also vary somewhat among years (Smith 1981); however, predators feeding on large pollock will undoubtedly be exploiting several age classes. Three species of marine mammals— harbor seals, sea lions, and fur seals— consume age classes of pollock that are also exploited by the commercial fishery (Table 3). A major effect of the pollock fishery has been a reduction in the abundance of older, larger individuals (Pereyra et al.8). Major declines in abundance of sea lions and fur seals in the eastern Bering Sea have been reported since the 1950's (Braham et al. 1980; Fowler 1982). Although the evidence is equivocal, especially for the fur seal (see Swartzman and Haar 1983), reduced food availability due to expansion of the pollock fishery has been sug- gested as a possible cause of the decline in popula- tions. The present population status of other pollock- eating marine mammals in the Bering Sea is not known. The sizes of fishes consumed by marine mammals are obviously very important for determining the nature and magnitude of marine mammal-fishery interactions. It is particularly important to recognize that because of different feeding strategies, changes 8Pereyra, W. T., J. E. Reeves, and R. G. Bakkala. 1976. Demer- sal fish and shellfish resources of the eastern Bering Sea in the baseline year 1976. Processed Rep., 619 p. Northwest and Alaska Fisheries Center, NMFS, NOAA, Seattle, WA. Table 3.— Age-class distribution of walleye pollock con- sumed by marine mammals in the Bering Sea, and caught in the commercial fishery in 1978, based on length-at-age data from Smith (1981). Percent of fishes in age class Predator species 1 23456789 >10 Harbor seal 43 20 23 3 0 3 3 6 Spotted seal 100 — Ribbon seal 98 1 1 — Steller sea lion 21 40 14 3 5 6 4 2 2 3 Fur seal1 49 44 7 — Minke whale 100 — — — — Commercial fishery2 2 20 40 18 20 (>5 yr old) 1from McAlister et al. 1976. 2from Smith 1981. in fish stock characteristics caused by fishing may benefit some marine mammal species while having no effect or being detrimental to others. Acknowledgments Support for this study was provided by the U.S. Bureau of Land Management Outer Continental Shelf Environmental Assessment Program and the Federal Aid in Wildlife Restoration Program. Num- erous colleagues, particularly John J. Burns and Larry M. Shults, assisted in the collection and processing of specimens. We are particularly grateful to Donald G. Calkins, Thomas R. Loughlin, and George Antonelis for providing us unpublished in- formation. Graphics and statistical analyses were done by Jesse Venable Clifford H. Fiscus made help- ful comments on an earlier draft of the manuscript. We also thank two anonymous reviewers whose com- ments substantially improved the manuscript. Literature Cited Bakkala, R., K. King, and W. Hirschberger. 1981. Commercial use and management of demersal fish. In D. W. Hood and J. A. Calder (editors), The eastern Bering Sea shelf: oceanography and resources, Vol. 2, p. 1015-1036. U.S. Dep. Commer., Off. Mar. Pollut. Assessment, NOAA, Rockville, MD. Braham, H. W., R. D. Everitt, and D. J. Rugh. 1980. Northern sea lion population decline in the eastern Aleutian Islands. J. Wild]. Manage 44:25-33. Bukhtiyarov, Y. A., K. J. Frost, and L. F. Lowry. 1984. New information on foods of the spotted seal, Phoca largha, in the Bering Sea in spring. In F. H. Fay and G. A. Fedoseev (editors), Soviet-American cooperative research on marine mammals, Vol. 1 - Pinnipeds, p. 55-59. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 12. Fiscus, C. H., G. A. Baines, and F. Wilke. 1964. Pelagic fur seal investigations, Alaska waters, 1962. U.S. Fish Wildl. Serv., Spec Sci. Rep. Fish. 475, 59 p. Fowler, C. W. 1982. Interactions of northern fur seals and commercial fish- eries. Trans. N. Am. Wildl. Nat. Resour. Conf. 47:278-292. Frost, K. J. 1981. Descriptive key to the otoliths of gadid fishes of the Ber- ing, Chukchi, and Beaufort Seas. Arctic 34:55-59. Frost, K. J., and L. F. Lowry. 1980. Feeding of ribbon seals (Phoca fasciata) in the Bering Sea in spring. Can. J. Zool. 58:1601-1607. 1981a. Trophic importance of some marine gadids in north- ern Alaska and their body-otolith size relationships. Fish. Bull., U.S. 79:187-192. 1981b. Foods and trophic relationships of cetaceans in the Ber- ing Sea. In D. W. Hood and J. A. Calder (editors), The eastern Bering Sea shelf: oceanography and resources, Vol. 2, p. 825-836. U.S. Dep. Commer., Off. Mar. Pollut. Assess- ment, NOAA, Rockville, MD. 1981c Ringed, Baikal, and Caspian Seals. In S. H. Ridgway and R. J. Harrison (editors), Handbook of marine mammals, Vol. 2, Seals, p. 29-53. Acad. Press, N.Y. 196 Lowry, L. R, and K. J. Frost. 1981. Feeding and trophic relationships of phocid seals and walruses in the eastern Bering Sea. In D. W. Hood and J. A. Calder (editors), The eastern Bering Sea shelf: oceanog- raphy and resources, Vol. 2, p. 813-824. U.S. Dep. Commer., Off. Mar. Pollut. Assessment, NOAA, Rockville, MD. Morrow, J. E. 1979. Preliminary keys to otoliths of some adult fishes of the Gulf of Alaska, Bering Sea, and Beaufort Sea. U.S. Dep. Commer., NOAA Tech. Rep., NMFS Circ. 420, 32 p. Nemoto, T. 1959. Food of baleen whales with reference to whale move- ments. Sci. Rep. Whales Res. Inst. 14:149-291. Pitcher. K. W. 1981. Prey of the Steller sea lion, Eumetopias jubatus, in the Gulf of Alaska. Fish. Bull., U.S. 79:467-472. SCHOENER, T W. 1971. Theory of feeding strategies. Annu. Rev. Ecol. Syst. 2:369-404. Seaman, G A., L. F Lowry, and K. J. Frost. 1982. Foods of belukha whales (Delphinapterus leucas) in western Alaska. Cetology 44:1-19. Smith, G. B. 1981. The biology of walleye pollock. In D. W. Hood and J. A. Calder (editors), The eastern Bering Sea shelf: oceanog- raphy and resources, Vol. 1, p. 527-551. U.S. Dep. Commer., Off. Mar. Pollut. Assessment, NOAA, Rockville, MD. SWARTZMAN, G. L., AND R. T HAAR. 1983. Interactions between fur seal populations and fisheries in the Bering Sea. Fish. Bull., U.S. 81:121-132. Kathryn J. Frost Lloyd F Lowry Alaska Department of Fish and Game 1300 College Road Fairbanks, AK 99701 OCCURRENCE OF SOME PARASITES AND A COMMENSAL IN THE AMERICAN LOBSTER, HOMARUS AMERICANUS, FROM THE MID-ATLANTIC BIGHT1 Larvae of the nematode Ascarophis sp. were reported by Uzmann (1967b) from American lobsters collected from Hudson, Block, Veatch, and Corsair Canyons on the edge of the continental shelf east and south of southern New England (Fig. 1). Follow- ing parasitological examinations of over 3,000 coastal and offshore lobsters, Uzmann (1970) reported that the nematode larvae were restricted almost ex- clusively to offshore lobsters. Adult Ascarophis sp. are intestinal parasites of fishes (Uspenskaya 1953). Although coastal and offshore lobsters occur off Contribution No. 1277, Virginia Institute of Marine Science, Gloucester Point, VA 23062. northern and central New Jersey, coastal lobsters are scarce or absent south of Cape May NJ. There is an active offshore commercial lobster fishery along the edge of the continental shelf south to Norfolk Canyon (Fig. 1). Materials and Methods To determine whether offshore lobsters in the Mid- Atlantic Bight have larval Ascarophis sp., we ex- amined the guts of 218 American lobsters, Homarus americanus, collected from August 1975 through March 1977. Lobsters from this region had not been examined previously for parasites. One hundred and ninety-seven of the lobsters ex- amined were caught in lobster traps or trawl nets by commercial and research vessels in Norfolk and Washington Canyons and from the shelf and slope between and adjacent to those canyons (areas III-V, Fig. 1) at depths of 73-402 m. The remaining 21 lobsters were caught by trawl nets from research vessels off the coasts of Delaware and New Jersey at depths of 57-95 m (area VIII, Fig. 1). The intestines and rectum were excised from live lobsters on shipboard (70% of the samples) or in the laboratory at the Virginia Institute of Marine Science, split longitudinally, and fixed in 10% Formalin2 or in Davidson's fixative No free parasites were found in the gut contents. In the laboratory, the gut was transferred to 35% glycerine in 70% ethanol, and part of the ethanol evaporated in a 55° C oven. Pieces of the gut were then laid open, pressed between two 35 x 50 mm slides, and examined for the presence of cysts. This procedure followed the recommendation of J. R. Uzmann3. Results Thirty-nine American lobsters were infected with larval Ascarophis sp., encapsulated in the anterior wall of the rectum (Table 1). The proportion of infec- tion in 218 lobsters (17.9%) from the Mid-Atlantic Bight was similar to that reported by Uzmann (1967b), when examined in a 2 x 2 contingency table and using Yates' correction for continuity (Elliott 1971). Uzmann (1967b) reported 77 infections in 314 lobsters (24.5%) collected east and south of southern New England. However, Boghen (1978) reported in- fection in the gills of 82 out of 233 lobsters (35.2%) 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 3J. R. Uzmann, Northeast Fisheries Center Woods Hole Labora- tory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543, pers. commun. June 1974. FISHERY BULLETIN: VOL. 84, NO. 1, 1986. 197 Figure 1— Canyons and lobster sampling sites along the edge of the continental shelf, between Cape Hatteras and the eastern edge of Georges Bank. 198 Table 1. — Prevalence of American lobsters infected with nematodes, Ascarophis sp., in the Mid-Atlantic Bight, August 1975- March 1977. No. lobsters sampled Prevalence of (No. infected) infection (%) Sexes Sexes Date Area' M F combined M F combined Aug., Sept. III 26(1) 236(6) 63(7) 3.8 16.2 11.1 1975 Dec. 1975 III 18(3) 18(2) 36(5) 16.7 11.1 13.9 Jan. 1975 III 3(1) 16(5) 19(6) 33.3 31.3 31.6 Jan. 1976 IV 11(1) 13(1) 24(2) 9.1 7.7 8.3 Apr. 1976 III 6(3) 9(2) 15(5) 50.0 22.2 33.3 Apr. 1976 V 4(2) 16(4) 20(6) 50.0 25.0 30.0 July 1976 V 7(1) 5(2) 12(3) 14.3 40.0 25.0 Oct. 1976 V 3(0) 5(2) 8(2) 0.0 40.0 25.0 Nov. 1976 VIII 11(1) 6(2) 17(3) 9.1 33.3 17.6 Mar. 1977 VIII 2(0) 2(0) 4(0) 0.0 0.0 0.0 Total 91(13) 127(26) 218(39) 14.3 20.5 17.9 1 III. Norfolk Canyon and adjacent slope IV. Between Norfolk and Washington Canyons V. Washington Canyon VIM. Between Wilmington and Hudson Canyons. 2One 86 mm female contained 33 acanthocephalan cysts, Corynosoma sp. from Northumberland Strait, southern Gulf of St. Lawrence That higher proportion of infection was highly significantly different from that reported off southern New England and in the Mid-Atlantic Bight. Mid-Atlantic Bight lobsters examined for parasites ranged from 49 to 179 mm carapace length (CL) (Table 2). Larval Ascarophis sp. were found in 13 (14.3%) of 91 male lobsters and in 26 (20.5%) of 127 female lobsters. No significant difference in preva- lence of infection between males and females, when size was ignored, could be demonstrated with a 2 x 2 contingency table analysis. This agrees with the absence of sex specificity in the canyon lobsters Table 2.— Numbers of American lobsters examined and prevalence of infection by the larvae of the nematode Ascarophis sp. in the Mid-Atlantic Bight. Size range, No. examined No. infected Percent of group Percent of total No. larvae, CL mm M F Sum M F Sum infected infected range 40-49 0 2 2 0 2 2 100.0 0.9 1-12 50-59 5 7 12 2 1 3 25.0 1.4 1-9 60-69 9 21 30 1 8 9 30.0 4.1 1-13 70-79 27 29 56 4 7 11 19.6 5.0 1-4 80-89 20 37 57 2 4 6 10.5 2.8 1-5 90-99 16 19 35 3 3 6 17.1 2.8 1-8 100-109 7 8 15 1 1 2 13.3 0.9 2-3 110-119 2 2 4 0 0 0 120-129 2 1 3 0 0 0 130-139 0 0 0 0 0 0 140-149 1 1 2 0 0 0 150-159 0 0 0 0 0 0 160-169 1 0 1 0 0 0 170-179 1 0 1 0 0 0 Total 91 127 218 13 26 39 17.9 110-149 5 4 9 0 0 0 150-179 2 0 2 0 0 0 reported by Uzmann (1967b) and also reported from Northumberland Strait by Boghen (1978). Almost one-half (46.3%) of all infections occurred in the 60-79 mm size classes; intensity of infection ranged from 1 to 13 (mean 3.0) (Table 2). None of the 11 lobsters >110 mm CL contained parasites. Boghen (1978) reported 51.3% infection in the 60-69.9 mm range When the occurrences of para- sites in males and females are arranged in three size groups, 40-59, 60-79 and 80-109 mm, and statistically examined with a 2 x 3 contingency table, no depar- ture from the expected 1:1 ratio was observed. A single specimen of the commensal polychaete, Histriobdella homari, was obtained from the gills of a female lobster, 82 mm CL, caught in Norfolk Canyon in June 1974. Gills of four other lobsters were excised, placed in dilute seawater in specimen bowls, and refrigerated overnight. The polychaete was found in the sediment collected from one gill. Because of the small number of lobster gills ex- amined, an estimate of prevalence is inappropriate Previously, Histriobdella was reported by Uzmann (1967a) in the gills and by Simon (1968) in the gills and bodies of New England lobsters, and by Boghen (1978) in the branchial chamber and gills of lobsters from Northumberland Straits. One female lobster, 86 mm CL, caught in Norfolk Canyon in August 1975, was infected with cysts of an acanthocephalan, Corynosoma sp. Thirty-three cysts were found in the intestinal wall and in the mesenteries along the outside of the intestine Adult Corynosoma sp. are parasites of mammals and aquatic birds; crustaceans are first intermediate hosts and fishes are second intermediate hosts (Yamaguti 1963). According to Uzmann (1970), Corynosoma sp. is a discriminator of coastal lobster stocks. Therefore its presence in a lobster taken in Norfolk Canyon indicates that migration from inshore to offshore waters occurs. Montreuil (1954) reported that the acanthocephalan infections in lobsters from the Magdalen Islands, Gulf of St. Lawrence, varied with the sex of the lobster and by season: 20% of females and 20% of males had cysts seemingly acquired towards the end of summer and early fall. Boghen (1978) attributed the absence of cysts in his North- umberland Strait samples to the fact that the lob- sters were collected before the end of summer. Discussion The variety of animal parasites and their inten- sity of infection are small in the Mid-Atlantic Bight lobsters. Differences in the occurrence and rates of 199 infection of Ascarophis and Corynosoma and of the commensal Histriobdella reported from American lobsters of the Mid-Atlantic Bight, southern New England waters, and the Gulf of St. Lawrence, are not large and could be attributed to differences in sample sizes or season of sampling. Peculiarly, cysts of the sporozoan Porospora sp. were not seen in Mid- Atlantic Bight lobsters, but occurred in most lobsters in the Gulf of St. Lawrence (Montreuil 1954; Boghen 1978) and were reported by Uzmann (1970) from southern New England waters. Cysts of the trema- tode Stichocotyle sp. were reported by Nickerson (1895) from Penobscot Bay, ME, and from lobster dealers in Boston, MA; by Linton (1940) from an un- stated region, probably Woods Hole, MA; by Uzmann (1970) from southern New England waters; and by Montreuil (1954) from southern Nova Scotia or southeastern New Brunswick. Nickerson (1895) found the cysts only in the intestinal tract at the union of the intestine and rectum. Literature Cited Boghen, A. D. 1978. A parasitological survey of the American lobster Homarus americanus from the Northumberland Strait, southern Gulf of St. Lawrence Can. J. Zool. 56:2460-2462. Elliott, J. M. 1971. Some methods for the statistical analysis of samples of benthic invertebrates. Freshw. Biol. Assoc, Sci. Pub. 25, 148 P- Linton, E. 1940. Trematodes from fishes mainly from the Woods Hole region, Massachusetts. Proa U.S. Natl. Mus. 88:1-172. Montreuil, P. 1954. Parasitological investigations. Rapp. Ann. Stn. Biol. Mar. Dep. Peches Quebec, Contrib. 50:69-73. Nickerson, W. S. 1895. On Stichocotyle nephropsis Cunningham, a parasite of the American lobster. Zool. Jahrb., Abt. Anat. Ontog. Tiere 8:447-480. Simon, J. L. 1968. Incidence and behavior of Histriobdella homari (An- nelida: Polychaeta), a commensal of the American lobster. Bioscience 18:35-36. Uspenskaya, A. B. 1953. The life cycle of nematodes of the genus Ascarophis van Beneden (Nematodes - Spirurata). [In Russ.] Zool. Zh. 32: 828-832. (Translated by J. M. Moulton, Bowdoin College, Brunswick, ME, 1966). Uzmann, J. R. 1967a. Histriobdella homari (Annelida:Polychaeta) in the American lobster, Homarus americanus. J. Parasitol. 53: 210-211. 1967b. Juvenile Ascarophis (Nematoda:Spiruroidea), in the American lobster, Homarus americanus. J. Parasitol. 53: 218. 1970. Use of parasites in indentifying lobster stocks. (Abstr.) In Section II, Proceedings of the Second International Con- gress of Parasitology, p. 349. J. Parasitol. 56(4). Yamaguti, S. 1963. Classification of the Acanthocephala. Systema Helmin- thum, Vol. V, Acanthocephala. Interscience Publ., 423 p. W. A. Van Engel R. E. Harris, Jr. D. E. Zwerner Virginia Institute of Marine Science School of Marine Science College of William and Mary Gloucester Point, VA 23062 RESILIENCE OF THE FISH ASSEMBLAGE IN NEW ENGLAND TIDEPOOLS1 Factors regulating density and species composition of tidepool fishes have been little studied, partic- ularly in comparison to other elements of the inter- tidal community (Gibson 1982). Twenty-two collec- tions of fishes were made in two tidepools at the Marine Science and Maritime Studies Center of Northeastern University at Nahant, MA, during summers from 1967 to 1985. Initially, the purpose was simply to demonstrate to my summer class in ichthyology the technique of collecting fishes with rotenone. After several years, it became apparent that there would be interest in examing long-term effects of repeated poisoning of the same pools. The purpose of this paper is to report the data from this series of samples and to compare the resilience of this New England tidepool fish fauna with studies done in the Gulf of California (Thomson and Lehner 1976), the central California coast (Grossman 1982), and South Africa (Beckley 1985). Unfortunately, there are no other similar tidepools in the area, so it was not possible to make control collections from unsampled pools. Methods The same two tidepools were sampled each sum- mer from 1967 to 1985. The tidepools are located on the ocean side of East Point, in Broad Sound. The higher pool is at about 2 m elevation and is about 1 m deep at high tide; the lower pool is slightly below 1 m elevation, contains extensive red and brown algal growth, and is shallower. Average tidal amplitude is slightly over 3 m. One collection was made each year except for 1969, 1982, and 1983 when two collections were made, spaced about 2 wk apart. Collections 'Contribution No. 134 from the Marine Science Institute, North- eastern University, Nahant, MA 01908. 200 FISHERY BULLETIN: VOL. 84, NO. 1, 1986. were made with rotenone (about 1 qt Noxfish2) at low tide in August, except in 1983 and 1985, when they were made in July and in 1984 when they were made in September. Specimens were taken by dip net from the pools by my students and me An at- tempt was made to collect and then count and measure (mm SL) all fishes. Sometimes I used a face mask to find fishes at the bottom of the pool which was closer to the ocean. Many invertebrates also were killed, but no attempt was made to record num- bers. The most abundant invertebrates in the 1984 collection were the green crab, Carcinus maenas (Linnaeus), and the sea urchin, Strongylocentrotus droebachiensis (Miiller). Also collected were amphi- pods, Gammarellus angulosus (Rathke), Calliopius laeviusculus (Kroyer), and Gammarus oceanicus Segerstrale; isopods, Idotea baltica (Pallas); and scale worms, Harmothoe imbricata (Linnaeus). Results Thirteen species of fishes were collected (Table 1). The number of species per collection varied from 3 to 8 (x 5.3). One species, the rock gunnel, Pholis gun- nellus (Linnaeus), was collected in all 22 samples. Young cunner, Tautogolabrus adspersus (Walbaum), were found in all but two collections. The grubby, Myoxocephalus aenaeus (Mitchill), and the threespine stickleback, Gasterosteus aculeatus Linnaeus, were present in 17 and 15 collections, respectively. The radiated shanny, Ulvaria subbifurcata (Storer), was taken 12 times. The seasnail, Liparis atlanticus (Jor- dan and Evermann), was taken in 10 collections, the mummichog, Fundulus heteroclitus Linnaeus, in 8. The American eel, Anguilla rostrata (LeSueur), was taken four times; young lumpfish, Cyclpterus lum- pus Linnaeus, three times. Four of the 13 species were taken only once or twice: the Atlantic tomcod, Microgadus tomcod (Walbaum); Atlantic silverside, Menidia menidia (Linnaeus); ninespine stickleback, Pungitius pungitius (Linnaeus); and northern pipefish, Syngnathus fuscus Storer. I can detect no long-term change in species composition or number of individuals over the 19-yr period. The number of specimens per sample varied from 17 to 1,850 (x 197.5), but the mean is distorted by the 1,842 young (9-28 mm SL) Tautogolabrus adsper- sus taken in sample 16. Deleting this number, the figures are 17-343 (x 119.2). Thus, a "typical" sam- ple would consist of 41 Pholis gunnellus, 49 young Tautogolabrus adspersus, 12 Myoxocephalus aenaeus, 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 7 Gasterosteus aculeatus, and 2 Fundulus heteroclitus. One other species might be present, 1 or 2 specimens of any of the other 8 species, most likely Ulvaria subbifurcata or Liparis atlanticus. There is great variation from collection to collec- tion in numbers of specimens of the most abundant 4 species: 2-232 Pholis gunnellus; 2-1,842 Tauto- golabrus adspersus; 1-127 Myoxocephalus aenaeus; and 1-44 Gasterosteus aculeatus. Ulvaria subbifur- cata, Liparus atlanticus, Cyclopterus lumpus, and Fundulus heteroclitus showed much less variation, 1-12 per collection. The other 5 species were uncom- mon, numbering 1-4 specimens. Discussion To evaluate short-term effects, comparisons can be made between pairs of collections made in 1969, 1982, and 1983 at 2-3 wk intervals. The number of species decreased from 8 to 6 in the 1969 pair and from 7 to 5 in 1982, but the number increased from 3 to 5 in 1983. Four of the 8 species in the first sam- ple in 1969, and 3 of the 7 species in the first sam- ple in 1982, numbered only 1 or 2 specimens, as did one of the species in the second sample of 1983. Numbers of individuals were about the same in the 1969 pair of collections (over 50) and the 1983 pair (74 and 86), but decreased (54 to 17) in the second collection of the 1982 pair. Rapid recolonization of the tidepools clearly takes place. Differences in thor- oughness of collecting, plus apparent random varia- tion in the 7 least commonly taken species, can explain the few differences between the paired collections. Thomson and Lehner (1976) sampled a large tide- pool in the Gulf of California 11 times over the period 1966-73. The period of time between sampling ranged from 13 to 78 wk. Number of species ranged from 16 to 26, total 50; number of individuals 435- 2,627, total 11,701. No decrease in number of species or individuals over time is apparent from their data (Thomson and Lehner 1970:table 1). Grossman (1982) sampled a series of rocky tide- pools with quinaldine at Dillon Beach in northern California 15 times from January 1979 to May 1981. The period of time between sampling ranged from 4 to 21 wk. Number of species per sample varied from 9 to 18 (excluding the first sample, 12-18), total 29 species; number of individuals was 71-517 per sample [not 520 as in Grossman's (1982) table 3], total 2,853 individuals. The structure of this rocky intertidal fish taxocene was persistent over 29 mo through 15 defaunations (Grossman 1982:table 3). Beckley (1985) sampled three South African pools 201 in CO CO CD E o 03 co Z JO O o a. a> 73 o g ■a o .2> o o CO Q) CO o 0) Q. 0) CO CO E 0) CD c CO 1_ CD N CO "O c (0 CD o c CO "D C 3 < m ,< oo CO 0) o a> Q. CO n "* SK cm t- t- CD co 4 6 m t- CO CM cd CD CD CM CO t- 00 00 * CM d) CD CD to co ;r 4 CO in o t cm \- o Si S CO CO ■- CD CO s CM t- i- U cm 6 <3- i- CMS cd CM CO CM CO R m 00 o cm 00 t- ^" CM 4 CM ■* 7- 00 ^. CO CO O |^ t^ CO cxj CO *~ o •* 1- co "55 c c o it <0 3 CD Q. « CD 10 -Q -2 O CD s l<0 cd CO CO CO -c & o I cm 6o CM CO CM CO m CM 4 o CM co CD CM in cm in 4 CO t- OO 1^ 00 CM t- CO CO CD in CO CM i— 00 m 2. 4 CO CO 3 *— CO 0) "3 o CO CO 3 & CO 8 s> CO CO C3 ■2 CO -3 •Q 3 CO to 3 y C CO co CO CO CM CM CM T3 CO CM O 00 6 CO CO m CO I CM in co 6 CM CM CD CO CM O CM O0 CD CO t- lO i- N CD --» O) r-(0i- "3- CO CD 00 t* in CM ■* ^. |^ CD ;s CD ,- CO ,- C) CM 00 N in 7t m cd r- r-~ 1- CO ,- in t ■* n ko OO Cfli-N i- m -s cd r- 00 r- N. CO CD CO CM CM ^" "^" CM P** y- m ^ OO t- eo y- N CO CO 't K i- CM CO U Is- T- CO ,- in oo cm i- co o ini-nSN in oo oo T r«- ,- a> CO i- ("- CD CM CO in co ^ cd ^J" CO CD r^ CD in m ~; co CO 00 00 h~ in in o oo co cm t- in cm co CM ^ CD y- 00 t- m cd in t- co N CO CO r- CD CM ~; CD i- 00 ^ CO 3 o o CD -c co 3 c 5 to o 3 c 3 o CO CO a o .3 3 E 6 C CO O CO 5 CO CO 3 Q. CO .3 CD CO 2 ■2 CO E CO 3 CO c D) Q. O CD s S) !6 £ O o C C O § 3 CL CO . o . O CD CL CD = Q. Q. O CO CO O ^ CD O O 6 .CD 08 ZZZEQ? 202 to CM CM 73 CD .C c o o CD ,< CO CD O CD D- CM C\J O CM CM t- t- 5*8 CO t- CO CO? co oo '-to i"» to *- S? ev ft. 2. en - S 6 o t^- LT) CM ,- CD ^ o co cx> °°o CO i- CO ^-^ ^-^ cm ^r CO r- CO cd 3 co ■* ID 00 m cm co CM t; cm CM f-^ *~ CD CM CO ^ CM (C •>- CM T- CO CD O t- CM CM CO i- in CD O CO m ;= CD CD CD r- m cm t- op CO t^- CO CO t- CM CM CO t-- CM CM r- 00 CO CM id CD r- CD CM o> in in LO co CO 3 "53 c c CO o co 52 CD f CD CO 2 -Q CD O O) O .CD CO 3 CO CD 3 CD C CD CD CO .3 4w CD ■2 iS "3 o CD CD ,2 CO 3 .O "cB CO 3 5 c -C -Q .CD a 3 CD CO CO CD o p 8 ^ £ .TO CO o CO 5 CD Sn CD ^ A CJ3 CO 3 O o CD -C CO .3 "3 ■Q c 3 so 03 p 3 O) c co 3 Q. E co 3 Q. o u 73 O O co 3 fc o en c 3 !6 CO Q. CD 3 E "D CO CD 13 CD CO D O) !6 c 3 c CD § 0. § CO 3 o CO ,3 CO 3 CD C D) C ^ CO U Q 1) n Q CO en ro in ^ ^r m CD CO N. CM co m CO O " CO CM CO Is" CD •* CO CM i- •<* if •* O ^- CO CO CO CO CO CD CD CO O h- CO CO CO CM CO ■»— r^ ■*— C) ^ O N mW r-- t- r~- ~- co h- N CD tn l~~ i- en i- CM i- Is- CO °o •^ ^ CO i— 00 T- CM i-~ ^ r~- •<* r^- •■ in r>- ,- oo ^ ^ co x- 00 r- "2 CO ^ 00 CO tv. ^ CD " T- CO ,- 1D CM CD O CO o CO CO S co ^ r--. co co CO CO f CM CO CO •3- in ,- r~- co I*- ^ co ,- 00 -r- CO Is- CO CO in 00 CM CD ^ h- t- CD -2- CD T- CO -r- CMSSoS CD ID c\J f~- ^- CO S CO i- 00 ,- . o • O CD Q. CD ^ Q. Q. O CO CO O O O O 9? CD CD Z Z Z L. Q > 203 with rotenone over a 2-yr period at intervals of 1 mo, 3 mo, and 6 mo. She found rapid recolonization but with lower densities of recolonizers in winter than in summer. During 26 monthly samples, only one of the original species did not recolonize the pool, while 13 additional species were found. In Pool 2, which was sampled in 3-mo intervals, 14 species were taken in the initial sample, 7-12 in subsequent samples. Three of the original 14 species failed to recolonize, but 8 additional species were taken. During four repeat visits to Pool 3, the number of species varied between 9 and 14, all but 1 species recolonized the pool, and 5 additional species were recorded. My study and those of Thomson and Lehner (1976), Grossman (1982), and Beckley (1985) indicate great resilience of species of tidepool fishes in tropical and temperate waters. Recolonization is quite rapid, within a matter of weeks. Acknowledgments I thank the students and teaching assistants in my ichthyology course for helping to collect the material. James Dooley made the second 1982 collection. N. W. Riser of the Marine Science and Maritime Studies Center identified the invertebrates. Comments on drafts of this note were provided by G. D. Grossman, J. Randall, N. W. Riser, V. G. Springer, and A. B. Williams. Literature Cited Beckley, L. E. 1985. Tidepool fishes: Recolonization after experimental elim- ination. J. Exper. Mar. Biol. Ecol. 85:287-295. Gibson, R. N. 1982. Recent studies on the biology of intertidal fishes. Oceanogr. Mar. Biol. Ann. Rev. 20:363-414. Grossman, G. D. 1982. Dynamics and organization of a rocky intertidal fish assemblage: the persistence and resilience of taxocene struc- ture Am. Nat. 119:611-637. Thomson, D. A., and C. E. Lehner. 1976. Resilience of a rocky intertidal fish community in a physically unstable environment. J. Exper. Mar. Biol. Ecol. 22:1-29. Bruce B. Collette Marine Science and Maritime Studies Center, Northeastern University, Nahant, MA 01908 and Systematics Laboratory, National Marine Fisheries Service, National Museum of Natural History, Washington, DC 20560 PARASITES OF BENTHIC AMPHIPODS: CILIATES Benthic gammaridean amphipods were sampled dur- ing a 2V2-yr period as a part of the Northeast Moni- toring Program (NEMP) of the Northeast Fisheries Center, National Marine Fisheries Service The am- phipod survey was designed to determine the kinds of parasites and pathological conditions occurring in amphipod populations that live in and on the sediments of the continental shelf from Maine to North Carolina. Microsporidans of the sampled am- phipods have been discussed by Johnson (1985), and this paper presents and discusses data on host distribution, prevalence, effects on the host, and probable relationships, of ciliates parasitizing am- phipods from the same samples. Materials and Methods Benthic amphipods were collected from 35 sta- tions, mainly on the Georges Bank and Mid-Atlantic Bight (Fig. 1). Amphipods were sampled during 11 cruises, July 1980-November 1982 (Table 1). Each station was sampled from 1 to 10 times during the survey. The 11 stations indicated by solid circles on Figure 1 had the most consistent and numerous populations of amphipods, were sampled at least five times each, and yielded the majority of data presented here A Smith-Mclntyre grab and occa- sionally an epibenthic sled or scallop dredge were used to obtain the samples. Up to 30 individuals of each species present in a sample, and sometimes more depending on numbers present, were prepared for histological study. Details of collecting pro- cedures and preparation of the amphipods for study are given by Johnson (1985). Results Host and geographic distribution of ciliate infec- tion is given in Table 1. Ciliate-infected amphipods were taken in samples from at least one station on every cruise There was no indication that prevalence was influenced by the season of the year or location of the positive stations. The majority of infected specimens were Ampelisca agassizi (Judd), but prevalence of ciliate infection was lower in A. agassizi than in the other species found infected (Pontogeneia inermis Kr0yer, Phoxocephalus holbolli Kr0yer, Har- pinia propinqua Sars, and unidentified haustoriids) (Table 2). In three instances, at station 33, cruise G; station 48, cruise I; and station 57, cruise E, in- dividuals ofH. propinqua or P. holbolli were infected 204 FISHERY BULLETIN: VOL. 84, NO. 1 50 100 150 200 KILOMETERS 68° 40- 36- Figure 1.— Number designations and positions of Northeast Monitoring Program (NEMP) benthic stations where gammaridean am- phipods were sampled during the survey. but A. agassizi collected at the same times were not. Except for A. agassizi, all the species with ciliate infections were rare (Table 2). The most numerous species collected, after A. agassizi, were Leptochei- rus pinguis Stimpson, which made up 11% of the total collected (2,655/24,244), and Unciola species (probably all U. irrorata Say and U. inermis Shoe- maker), which made up 10% of the total (2,356/ 24,244). Despite their abundance, these species were never found infected with ciliates. Considering all amphipods sectioned and examined, overall prevalence of ciliate infection was 0.6% (41/7,363). Light infections consisted mainly of large ciliates. Heavier infections had medium to small ciliates, but 205 Table 1.— Stations with ciliate-infected amphipods, by cruise and host species. Cruise1 Station A B C D E F G H 1 J K 20 P|2 3 23 AA* AA AA — — AA — — 25 — PH2 33 — AA AA AA AA AA HP2 — — 35 — AA — AA — — — AA 37 AA 38 — — AA — 48 — AA — — HP 49 AA 50 AA 51 AA AA — 57 — — — — AA PH — AA AA 62 HAU2 — — — — 78 — — HAU 'Dates of cruises: A, July 1980; B, Sept. 1980; C, Dec. 1980; D, Apr 1981; E, July 1981; F, Aug. 1981; G, Nov. 1981; H, Jan. 1982; I, Mar 1982; J, Aug. 1982; K, Nov. 1982 2lnfected amphipods present at the station PI = Pontogeneia inermis, AA = Ampelisca agassizi, PH = Phoxocephalus holbolli, HP = Harpinia propin- qua, HAU = unidentified haustoriids. 3— = station sampled, no ciliate Infections found. sometimes large forms were also present. The largest ciliates were in the gill of a specimen of Pon- togeneia inermis (Fig. 2). Measured in paraffin sec- tions, they were about 17 ^m x 80 ^m. Large forms in other infected amphipods were 16-20 jjm x 40-50 jim. The majority of small- and medium-sized ciliates were 17-30 ^m in the greater dimension; none were less than 14 fiin (Fig. 3). Ciliates were elongate- spindle-shaped, with pointed or sharply rounded ends in P. inermis, and oval to subspherical in the other amphipods. The macronucleus of the large ciliates in P. inermis was sometimes ribbonlike (Fig. 2), and macronuclei of the smaller ciliates in P. in- ermis and those from the other amphipod species were elongate cylinders or elongate ovals in section (Fig. 3). None of the infections showed recent evidence of host reaction against the ciliates. The melanized nodules and small hemocytic encapsulations occa- sionally seen in infected amphipods did not contain recognizable ciliates, and may have been due to other causes. Few pathological effects were visible in ciliate- infected tissues. Two infected subadult males of A. agassizi had karyorrhexis and probable lysis in the transverse abdominal muscles, and one heavily in- fected female of A. agassizi, which had a few early embryos in the brood pouch, also had retained necrotic, mature ova in the ovaries. All infected am- phipods had material in the gut, indicating that feeding was continuing. Hemocytes were present in Figure 2 — Pontogeneia inermis: large and small ciliates in the gill. L, large form; S, small form. Bar = 10 /im. > .1 206 r % # < Figure 3.— Ampelisca agassizi: medium-sized ciliates. The small, pale micronucleus is visible close to the macronucleus in one of the ciliates (arrow). Bar = 10^m. light to medium infections, but essentially missing in heavy infections. None of the ciliates were posi- tioned in such a way that they appeared to have been phagocytizing hemocytes or other cells at the time of fixation. The granular inclusions commonly pres- ent in the cytoplasm of the ciliates bore no resem- blance to food vacuoles or phagocytized material. Discussion Two groups of ciliates contain species that para- sitize crustaceans. Paranophrys, in crabs, lobsters, and possibly isopods, and Parauronema, in penaeid shrimp, belong in the class Oligohymenophora, order Scuticociliatida (Corliss 1979). They are apparently opportunistic parasites (Bang 1970; Sindermann 1977; Couch 1978; Armstrong et al. 1981; Hibbits and Sparks 1983). The remaining parasites are members of the class Kinetofragminophora, order Apostomatida (Corliss 1979). Typical apostomes are obligate commensals of aquatic crustaceans and have a life cycle geared to their hosts' molting cycles (Bradbury 1966, 1973), but some apostomes have become internal parasites of various invertebrates, including polychaetes, cephalopods, ophiurans, coelenterates, ctenophores, and isopod, amphipod, and decapod crustaceans (Corliss 1979). Because specialized fixation and staining of whole Table 2.— Species of amphipods infected by ciliates: proportion of the amphipod population and prevalence of ciliate infection. Species of amphipod Percent prevalence at positive stations Proportion of the total amphipods collected Ampelisca agassizi Harpinia propinqua Haustoriidae spp. Phoxocephalus holbolli Pontogeneia inermis 3.8% (31/812) 18.2% (2/11) 5.4% (3/56) 9.5% (2/21) 10.3% (3/29) 54.3% (13,165/24,244) 0.6% (146/24,244) 0.9% (225/24,244) 0.5% (125/24,244) 0.7% (164/24,244) 207 ciliates is necessary for firm identification (Corliss 1979), the amphipod ciliates can be only provision- ally assigned to a ciliate group, as is true in other studies based on fixed and embedded material (Sparks et al. 1982; Hibbits and Sparks 1983). On the basis of similarities in hosts and morphology, the amphipod ciliates discussed here are like the apos- tome genus Collinia, whose members parasitize am- phipods (Summers and Kidder 1936; de Puytorac and Grain 1975). Like Collinia, size of individual ciliates from the benthic amphipods varied greatly and there was no indication that the ciliates were phagocytic Paranophrys and Parauronema, on the other hand, belong to a group that ingests particu- late material. Paranophrys is known to ingest hemocytes and other cells of its hosts, and does not exhibit major size differences (Bang 1970; Sparks et al. 1982; Hibbits and Sparks 1983). Provisionally, the ciliates of benthic amphipods are being con- sidered apostomes. Whether more than one species of ciliate was in- volved in the infections is uncertain, but probably the ciliate of Pontogeneia inermis represented a species apart from the others. Its very large forms with the ribbonlike macronucleus were not dupli- cated in other infections. Although more A. agassizi were found infected with ciliates than any other species of amphipod, this was apparently because it was the most abundant and widespread of the susceptible species sampled. A. agassizi had the lowest overall prevalence of ciliate infection and sometimes was not parasitized when other species in the same samples were parasitized. There are at least two possible explanations for the odd host distribution of the amphipod ciliates. First, the ciliates might be highly host specific, each am- phipod species having its own species of ciliate Sec- ond, the ciliates might be either primary parasites of some other member(s) of the benthic community, or incompletely adapted to a parasitic existence, and thus only occasionally parasitizing the least resis- tant species of the sampled amphipods. Unciola species and Leptocheirus pinguis were often the most abundant amphipods at certain stations, but ciliates were never found in individuals of these species, suggesting that they are resistant to ciliate attack. Conversely, the relatively high prevalence of ciliates in the rare species of amphipods could in- dicate less resistance than is exhibited by most of the species of amphipods sampled. Presumably, infected amphipods would eventually die of their ciliate infections because of the massive loss of hemocytes. The infrequency of ciliate infec- tion, except in certain rare species, indicates that these parasites are not important in regulating the general amphipod populations they infect. Acknowledgments Thanks are due the following: Frank Steimle and Robert Reid of the Northeast Fisheries Center, Sandy Hook Laboratory, and Linda Dorigatti, Gret- chen Roe, and Sharon MacLean of the Oxford Laboratory collected the amphipods; Ann Frame, Sandy Hook Laboratory, provided advice and train- ing in identification of amphipods; Linda Dorigatti identified material from cruises A to C, and she, Gretchen Roe, Dorothy Howard, and Cecelia Smith, Histology Section, Oxford Laboratory, prepared the specimens for histological examination. Literature Cited Armstrong, D. A., E. M. Burreson, and A. K. Sparks. 1981. A ciliate infection (Paranophrys sp.) in laboratory-held Dungeness crabs, Cancer magister. J. Invertebr. Pathol. 37:201-209. Bang, F. B. 1970. Disease mechanisms in crustacean and marine arthro- pods. In S. F. Snieszko (editor), a Symposium on Diseases of Fishes and Shellfishes, p. 383-404. Am. Fish. Soc, Wash., D.C., Spec Publ. No. 5. Bradbury, P. C. 1966. The life cycle and morphology of the apostomatous ciliate, Hyalophysa chattoni n.g., n. sp. J. Protozool. 13: 209-225. 1973. The fine structure of the cytostome of the apostomatous ciliate Hyalophysa chattoni. J. Protozool. 20:405-414. Corliss, J. O. 1979. The ciliated Protozoa. Characterization, classification and guide to the literature 2d ed. Pergamon Press, Ox- ford, 455 p. Couch, J. A. 1978. Diseases, parasites, and toxic responses of commercial penaeid shrimps of the Gulf of Mexico and South Atlantic coasts of North America. Fish. Bull., U.S. 76:1-44. de Puytorac, P., and J. Grain. 1975. Etude de la tomitogenese et de l'ultrastructure de Col- linia orchestiae, cilie apostome sangnicole, endoparasite du crustace Orchestia gammarella Pallas. Protistologica 11: 61-74. Hibbits, J., and A. K. Sparks. 1983. Observations on the histopathology caused by a parasitic ciliate (Paranophrys sp.?) in the isopod Gnorimosphaeroma oregonensis. J. Invertebr. Pathol. 41:51-56. Johnson, P. T. 1985. Parasites of benthic amphipods: microsporidans of Ampelisca agassizi (Judd) and some other gammari- deans. Fish. Bull., U.S. 83:497-505. Puytorac, P. de, and J. Grain, See de Puytorac, P., and J. Grain. Sindermann, C. J. 1977. Ciliate disease of lobsters. In C. J. Sindermann (editor), Disease diagnosis and control in North American marine aquaculture, p. 181-183. Elsevier Sci. Publ. Co, Amsterdam. 208 Sparks, A. K., J. Hibbits, and J. C. Fegley. 1982. Observations on the histopathology of a systemic ciliate (Paranophrys sp.?) disease in the Dungeness crab, Cancer magister. J. Invertebr. Pathol. 39:219-228. Summers, F. M., and G. W. Kidder. 1936. Tkxonomic and cytological studies on the ciliates asso- ciated with the amphipod family Orchestiidae from the Woods Hole district. II. The coelozoic astomatous parasites. Arch. Protistenkd. 86:379-403. Phyllis T. Johnson Northeast Fisheries Center Oxford Laboratory National Marine Fisheries Service, NOAA Oxford, MD 21654 FECUNDITY OF THE PACIFIC HAKE, MERLUCCIUS PRODUCTUS, SPAWNING IN CANADIAN WATERS Previous studies on the fecundity of Pacific hake, Merluccius productus, have been concentrated on the coastal stock in Baja California (MacGregor 1966, 1971; Ermakov et al. 1974), although large-scale spawning events have been recorded as far north as lat. 38°N, near San Francisco, CA (Stepanenko 1980). The present work was undertaken in conjunc- tion with ichthyoplankton surveys, aimed at esti- mating the released egg production and spawning biomass of the Pacific hake stock resident in the Strait of Georgia, a semi-closed marine basin in British Columbia (Thomson 1981). The spawning season extends from February through June, peaks in early April, and is 90% complete by mid-May (Mason et al. 1984). In comparison with the coastal stock of some 1-2 million metric tons (t) (Bailey et al. 1982), this in- shore stock, of about 140,000 t, is subject to modest annual exploitation (1-500 t) and resides in a semi- estuarine environment on the known northernmost edge of the reproductive range. The coastal stock undertakes a northward feeding migration after the spring spawning and reaches the southwest coast of Vancouver Island by late summer (Bailey et al. 1982). There is no evidence of intermingling between these two stocks, based on their distributional patterns. The inshore stocks in the Strait of Georgia and Puget Sound may undergo some exchange, possibly due to surface transport of larvae produced in the central Strait of Georgia (Mason et al. 1984). The Puget Sound and coastal stocks have been identified as genetically distinct by Utter and Hodgins (1971), but the two inshore stocks in Puget Sound and the Strait of Georgia have not been similarly com- pared. Histological analysis has indicated that only one mode of oocytes developes in Georgia Strait hake. However, like the Baja, California form and hake species elsewhere, some Strait of Georgia hake show evidence of ovarian resorption following spawning (Foucher and Beamish 1980). The quantitative sig- nificance of resorption relative to individual and stock fecundities, or to their potential physiological and environmental correlates have not yet been ex- amined. This report considers the "apparent fecun- dity" as an annual expression of reproductive poten- tial applicable to the stock in the Strait of Georgia, determines that fecundity, and concludes that ovarian resorption is of minor consequence in the stock. Materials and Methods The ovaries of 97 Pacific hake females 39-82 cm FL were collected during late February and early March of 1980 and 1981, 71 of which were collected in 1981 (McFarlane et al. 1983). Unspawned females were selected in maturity stages R2 and R (Foucher and Beamish 1977) when the ovary is yellow and opaque, has prominent blood vessels, and fills one- third to one-half of the coelomic cavity. No ovaries contained translucent oocytes which signify immi- nent spawning. Fresh ovaries were preserved in 10% formaldehyde solution. In the laboratory, the pre- served ovaries were transferred to modified (Simp- son 1951) Gilson's fluid for several months to allow breakdown of connective tissue Ovaries were then washed thoroughly in cold water over a series of stainless steel screens of 40 jum and larger aperture, and gently broken up by hand when necessary to separate the hardened eggs from the ovarian tissue The mesh size of the finest screen was determined by the difficulty encountered in separ- ating oocytes <40 ^m diameter from ovarian tissue The cleaned eggs were then stored in 5% formal- dehyde solution in preparation for analysis. Eggs from a single ovary were transferred to a 20 L glass reservoir filled to either 10 or 15 L. While the reservoir was being stirred vigorously with a wooden paddle in a rotating figure-eight pattern, a second worker extracted 50 1-2 mL volumetric sub- samples using Stempel pipettes and transferred them to petri dishes. Under the dissecting micro- scope at 50 x magnification, all eggs in five subsam- ples were sized and counted in 20 /urn intervals of oocyte diameter. These results were then combined to construct oocyte size-frequency histograms and FISHERY BULLETIN: VOL. 84, NO. 1, 1986. 209 to allot proportions of the combined egg count to the various size intervals. All eggs were counted in the remaining 45 subsamples to provide with the previous 5 subsamples, 50 counts of eggs per unit volume The total number of eggs in the ovary was calculated from the product of mean subsample count per milliliter and the reservoir volume prior to subsampling. The number of eggs in various size categories was obtained by applying the appropri- ate proportional value to the estimated total number of eggs in the ovary. Subsample egg counts averaged between 50 and 150 eggs, with the majority falling within 75 and 100. Size-frequency histograms were based on 250-750 sized eggs with the majority bas- ed on 375-500 sized eggs. Initial procedural evalua- tion indicated that 200 sized eggs was sufficient to obtain a replicable size-frequency distribution. Eighteen ovaries from postspawned females were collected on 3 July 1981 and were similarly processed. Prespawning females collected in 1981 were ag- ed by the otolith break and burn method (Chilton and Beamish 1982). Results and Discussion Frequency Distributions of Oocyte Diameter for Prespawners Most of the 97 ovaries of prespawners examined contained a pronounced bimodal distribution of oocyte diameters with peaks at about 100 p*m and between 500 and 600 ^m (Figs. 1-3). Oocytes <150 jjm in diameter contained no yolk materials and are taken to constitute a reserve fund for subsequent years (Foucher and Beamish 1980). Oocytes >150 ^m diameter were undergoing vitellogenesis, and a few ovaries contained nonhydrated oocytes reaching 700-750 nm diameter. Hydrated eggs were not seen in these ovaries collected in early March and hydra- tion probably does not occur in oocytes <700 /urn, although hydrated oocytes from 350 to 950 /^m diameter were found by Foucher and Beamish (1980). This apparent discrepancy may reflect their underestimation of oocyte diameters in histological preparations of translucent oocytes due to the plane of sectioning. The unimodal distribution of yolked oocytes, also reported for M. m. hubbsi in the Argentine Sea (Christiansen and Cousseau 1971) does not comple- ment the findings of MacGregor (1966, 1971). He found that ovaries of prespawning coastal hake taken off Baja California contained distinct groups of "small" and "large" yolked oocytes, of which only the latter were destined for release Furthermore, Er- makov et al. (1974) reported 21% of the 93 female Pacific hake taken off Baja California in 1972 had unimodal, 55% bimodal, 18% trimodal, and 6% quadrimodal oocyte distributions. Similarly, their subsequent sample of 45 ovaries collected in the Oregon-Washington region in late November contain- ed 22% unimodal, 65% bimodal, and 6% trimodal distributions, with major peaks at 200 and 600 /um diameter. Nearly half of the ovaries collected and ex- amined by Ermakov et al. (1974) did not contain a bimodal distribution of yolked oocytes, although these authors concluded that asynchronous develop- ment of yolked oocytes indicated the probability of multiple spawnings, most likely two batches within the spawning season. Estimates of Total Fecundity Standard errors of mean egg counts for total fecundity estimates of total fecundity (oocytes ^40 (jm diameter) ranged between 0.4 and 4.4% of the means and were <3% in nearly 70% of the 97 ovaries processed. The variability of the enumeration tech- nique compares favorably with that reported by Mason et al. (1983) in an analysis of the fecundity of the sablefish, Anoplopoma fimbria, and with that reported by Pitt (1963) on the fecundity of the American plaice Hippoglossoides platessoides, using Wiborg's whorling vessel (Wiborg 1951). The estimates of total fecundity (oocytes ^40 ^m diameter) increased with fork length according to the equation F = 0.3081FL3-7605, [where FL = fork length in centimeters]. The correlation coefficient (r) for the regression was 0.93. An insignificant F ratio from analysis of variance of slope and inter- cept values allowed pooling of the 1980 and 1981 data. The smallest and largest Pacific hake females in the sample (39 and 82 cm FL) contained estimated total oocyte complements of 202,100 and 3,009,900 oocytes >40 ^m, respectively. All 97 fecundity esti- mates fell within the range of 165,700 and 3,108,000 oocytes ^40 \xm. Estimates of Fecundity Within Size Classes of Oocytes The estimated number of oocytes with 20 \xm in- tervals of diameter were summed within five inter- vals and regressed against fork length to examine the correlation coefficients (Table 1). Coefficients declined progressively with increased oocyte diameter, reflecting increasing variability among 210 18 16 14 - 12 10- 8 - 6- 4 2 0 18 16 14- 12 10- 8 6 4- 2- 0 18 - Z UJ 14- o UJ IT U. I- 8- 2 O 6 . 18 16 14 12- 10 8- 6- 54 cm wJjLP 0 2 4 6 8 10 0 2 4 6 8 10 54cm 0 2 4 6 e 10 OOCYTE DIAMETER (Jim « I0Z) Figure 2— Representative frequency distributions of oocyte diameter from ovaries of Pacific hake 47-54 cm FL. 212 12 10- >- o o cc o LL 56 cm Uiliiii 59 cm 62 cm 69 cm 57 cm 60 cm 63 cm Mkiliitu 70 cm 57 cm illllll 61 cm B 16 14- 12- 10 63 cm 70 cm JJiUilil 58 cm 62 cm llll llllllllll 65 cm 73 cm llllllllllllllll, 18- 16 14- 12- 10 8 6- 74 cm Mil I, 2 4 6 74 cm i- 16- 14- 12 10 8- 6- 4 2 80 cm 111 llll 82 cm llll 2 4 6 2 4 6 8 10 2 4 6 8 10 OOCYTE DIAMETER (jjm x I02 ) Figure 3.— Representative frequency distributions of oocyte diameter from ovaries of Pacific hake 56-82 cm FL. 213 Table 1.— Regression equations for oocytes of several size classes, and some combinations of same, found in prespawned ovaries of Pacific hake from the Strait of Georgia, B.C. Oocyte Regression Correlation diameter Oocyte equation coefficient G*m) description (F = aFLb) (r) 40-780 all oocytes F = 0.3081 FL37605 0.93 40-180 unyolked reserve F = 0.0692FL3 9766 0.88 200-380 small, yolked F = 0.0446FL37097 0.86 400-580 medium, yolked F = 0.2078FL34174 0.71 600-780 large, yolked F = 0.0008FL4 6370 0.65 400-780 medium plus large yolked F = 0.1872FL35640 0.75 200-780 all yolked F = 0.5501 FL33896 0.81 females in the number of maturing oocytes as their maturity stage advanced towards hydration. This may be both a reflection of the range in stage of maturity among individual females at a common time of collection, and variation among females in the proportion of yolked oocytes destined for hydra- tion and release Apparent fecundity taken as the number of yolked oocytes >200 /urn was best expressed by the equa- tion Fa = 0.5501FL3-3896. The averaged female hake in the Strait of Georgia stock (43.3 cm FL) contain- ed an estimated 193,868 yolked oocytes >200 /urn and had a relative apparent fecundity of 382.3 eggs/g. In comparison, an uncommonly large female (80 cm FL) could contain more than 1.5 million yolked oocytes for a specific fecundity of 477 oocytes/g (Table 2). Pacific hake in the Strait of Georgia grow rapidly to age 4, showing almost linear growth in length (McFarlane et al. 1983). Thereafter, growth de- creases rapidly and is accompanied by considerable individual variation in annual growth. The largest female in the sample (82 cm FL) was age 18 whereas another female age 15 was only 49 cm FL. Not surprisingly, age was weakly related to apparent fecundity and wide individual differences in ap- Table 2.— Total and relative (oocytes/g body weight) fecundity estimates at fork length for unyolked (40-180 ^m diameter) and yolked (200-780 /^m diameter) oocytes found in prespawned ovaries of Pacific hake from the Strait of Georgia, B.C. Fork length Unyolked oocytes Yolked oocytes % yolked of (cm) Total Relative Total Relative unyolked 40 162,502 406 148,178 370 91.1 45 259,580 455 220,887 388 85.3 50 394,666 507 315,679 403 79.5 55 576,544 551 436,089 417 75.7 60 814,896 598 585,684 430 71.9 65 1,120,308 645 768,233 443 68.7 70 1,504,260 693 987,611 455 65.7 75 1,979,132 739 1,247,812 466 63.1 80 2,558,196 786 1 ,552,943 477 60.7 parent fecundity are evident within age classes (Fig. 4). Frequency Distributions of Oocyte Diameter in Postspawners Gonads of 276 adult Pacific hake, trawl-caught on 3 July 1981, were staged superficially for maturity after Foucher and Beamish 1977. All gonads were in postreproductive state The ovaries of 18 of 111 females retained for microscopic analysis were dis- tributed within the various maturity states with these results: spent (1), recovering (7), and resting (10). Yolked oocytes (200-500 /mi) were found in 7 ovaries: spent (1), recovering (4), and resting (2). Number of oocytes ^200 /mi, expressed as a percent- age of the oocytes <200 pm (40-180 /mi) was <3% in 6 of these fish, and 11% in the seventh, compared with 85-90% in prespawned ovaries collected in March (Table 2). These results support previous conclusions that not all yolked oocytes larger than 200 /mi diameter are released, as suggested by Foucher and Beamish (1980) and MacGregor (1966). They also suggest that resorbtion in postspawned females probably does not exceed about 5% of the yolked oocytes destined for release The female Pacific hake in the Strait of Georgia appears to use progressively less of the reserve fund of unyolked oocytes present during gonadal matura- tion in subsequent spawnings (Table 2), although relative and apparent fecundity increases with in- creased fork length. This can be illustrated by com- paring females <55 cm FL (Figs. 1, 2) with larger females (Fig. 3). The number of reserve fund oocytes in the size fraction 40-180 /mi increases at a faster rate, almost doubling the relative fecundity for reserve fund oocytes in this size fraction by 80 cm FL than does production of larger oocytes. The reserve fund may have several origins, and cyto- logical evidence was presented by Foucher and Beamish (1980) that the fund may be supplemented by cells of follicular origin in the postspawned ovary. Such a mechanism to increase potential fecundity would appear to be rather redundant if significant resorbtion of yolked oocytes commonly occurs. Stock Differences in Fecundity and Estimates of Spawning Stock Methodological differences or lack of disclosure, and lack of substantiated assessment of stock- specific resorption following spawning, render it im- possible to draw very useful comparisons of fecun- 214 LU lu 180- 170 160- 150- 140 130- S 120- E =«, O o CVJ co CD O LU LL o CO Q Z < co D O I LL o co Z LU 1 10 - 100- 90 80- 70 60- 50 40 30- 20 10- •'° 9 0/< • 10 »9 • 10 »I3 • 14 • 12 • • 8 '-» -I?9 • 13 • • •/•o 4»* §5^6 6 O •^ 3s*„ o o o o+^v- 0 40 44 48 52 56 60 64 68 FORK LENGTH (cm) 72 76 80 84 Figure 4— Estimated number of yolked oocytes >200 ^m diameter in 97 hake ovaries from the Strait of Georgia, B.C., plotted against fork length of female hake Numbers adjacent to individual plots indicate estimated age of female; open circles - 1980 females, closed circles - 1981 females. 215 dity between coastal and inshore stocks of Pacific hake at this time Ermakov et al. (1974) excluded oocytes <100 pirn diameter, thus excluding a large fraction of unyolked oocytes constituting the reserve fund. Their estimates of total fecundity (compar- able fork length) are one-half to one-third of those reported here for hake in the Strait of Georgia (>40 mm) and are also lower than the present estimates for apparent fecundity (oocytes ^200 ^m diameter). MacGregor (1966, 1971) counted advanced, yolked oocytes (>600 /urn) only, premised on his assumption that only these cells were destined for release On the basis of relative fecundity (eggs per gram), for yolked oocytes >580 fim diameter of comparable size to MacGregor's "large, yolked" or "advanced" oocytes, the female Pacific hake in the Strait of Georgia are considerably less fecund (54-164 eggs/g) over the fork length range of 40-80 cm than are Baja California hake which averaged 216 eggs/g (MacGre- gor 1971). However, the lack of distributional bi- modality in the Canadian ovaries renders such a com- parison unrealistic, for a common size threshold for resorption, even if appropriate, cannot be applied conveniently to individual ovaries. We can state with reasonable certainty that re- sorption of yolked oocytes is a common occurrence in both coastal and inshore stocks of Pacific hake, as has been found in other forms of Merluccius (Hickling 1930; Christiansen 1971). The influence of ovarian resorption on annual fecundity of stock and on the magnitude of released egg production from individual females remains unknown. It follows that the application of existing fecundity information to problems of assessing magnitude of Pacific hake spawning stock from released egg production, as determined through ichthyoplankton surveys, should reflect these reservations. For the Pacific hake stock in the Strait of Georgia, British Columbia, resorption may not involve more than 5-10% of the apparent fecundity. Hence spawn- ing biomass estimates based on released egg pro- duction and the apparent fecundity could be rendered conservative by the observed extent of resorption in this stock. Acknowledgments Staff of the Groundfish Program at Nanaimo are thanked for collecting biological materials and related statistics at sea, and for aging the female Pacific hake used in the study (Aging Unit). Par- ticular appreciation is extended to Susan Johnston for her able assistance in the laboratory and to R. Foucher for helpful discussion and comments on the original draft. Literature Cited Bailey, K. M., R. C. Francis, and P. R. Stevens. 1982. The life history and fishery of Pacific whiting, Merluc- cius productus. NOAA, NMFS, Proc Rep. 82-03, NWAFC, Seattle, 81 p. Chilton, D. E., and R. J. Beamish. 1982. Age determination for fishes studied by the Groundfish Program at the Pacific Biological Station. Can. Spec Publ. Fish. Aquat. Sci. 60, 102 p. Christiansen, H. E. 1971. La reproduccion de la merluza en el Mar Argentino (Merlucciidea, Merluccius merluccius hubbsi). 1. Descripci6n histologica del ceclo del ovario de merluza (Reproduction of the hake in the Argentine Sea. 1. Histological description of the spawning cycle of the hake). Bol. Inst. Biol. Mar., Mar del Plata 20:1-41. (Engl, transl., Can. Fish. Mar. Serv., Transl. Ser. 4003.) Christiansen, H. E., and M. B. Cousseau. 1971. La reproduccion de la merluza en el Mar Argentino (Merlucciidea, Merluccius merluccius hubbsi). 2. La reproduccion de la merluza y su relaci6n con otros aspectos biologicos de la especie (Reproduction of the hake in the Argentine Sea. 2. Hake reproduction and its relationship with other biological aspects of the species). Bol. Inst. Biol. Mar, Mar del Plata 20:41-73. (Engl, transl., Can. Fish. Mar. Serv., Transl. Ser. 4003.) Ermakov, J. K., V. A. Snytko, L. S. Kodolov, 1. 1. Serobaba, L. A. Borets, and N. S. Fadeev. 1974. Biological characteristics and conditions of the stocks of Pacific hake, sea perches, sablefishes, and walleye pollock in 1972. Engl, transl., Can. Fish. Mar. Serv. Transl. Ser. 3066, 37 p. Foucher, R. P., and R. J. Beamish. 1977. A review of oocyte development in fishes with special reference to Pacific hake (Merluccius productus). Can. Fish. Mar. Serv. Tech. Rep. 755, 16 p. 1980. Production of nonviable oocytes by Pacific hake (Merluc- cius productus). Can. J. Fish. Aquat. Sci. 37:41-48. Hickling, C. F. 1930. The natural history of the hake Part III. Seasonal changes in the condition of the hake. Fish. Invest. Minist. Agric. Fish. Food (G.B.), Ser. II, XII(l):l-78. McFarlane, G. A., W. Shaw, and R. J. Beamish. 1983. Observations on the biology of Pacific hake, walleye pollock, and spiny dogfish in the Strait of Georgia, February 20-May 2, and July 3, 1981. Can. MS Rep. Fish. Aquat. Sci. 1722, 109 p. MacGregor, J. S. 1966. Fecundity of the Pacific hake, Merluccius productus (Ayres). Calif. Fish Game 52:111-116. 1971. Additional data on the spawning of the haka Fish. Bull., U.S. 69:581-585. Mason, J. C, R. J. Beamish, and G. A. McFarlane. 1983. Sexual maturity, fecundity, spawning, and early life history of sablefish (Anoplopoma fimbria) off the Pacific coast of Canada. Can. J. Fish. Aquat. Sci. 40:2126-2134. Mason, J. C, A. C. Phillips, and O. D. Kennedy. 1984. Estimating the spawning stocks of Pacific hake (Merluc- cius productus) and walleye pollock (Theragra chalcogram- ma) in the Strait of Georgia, B.C. from their released egg pro- duction. Can. Tech. Rep. Fish. Aquat. Sci. 1289, 51 p. 216 Pitt, T. K. 1964. Fecundity of the American plaice, Hippoglossoides platessoides (Fabr.) from Grand Bank and Newfoundland areas. J. Fish. Res. Board Can. 21:597-612. Simpson, A. C. 1951. The fecundity of the plaice Fish. Invest. Ministr. Agric Fish. Food (GB), Ser. II, 17(5):l-27. Stepanenko, M. A. 1980. Reproductive conditions and the assessment of the spawning part of the Pacific hake, California anchovy, horse- mackerel, and some other fish species in the California Cur- rent Zone in 1979. Pac. Inst. Fish. Oceangr. (TINRO), Manuscr. Rep., 29 p. Thomson, R. E. 1981. Oceanography of the British Columbia coast. Can. Spec Publ. Fish. Aquat. Sci. 56, 219 p. Utter, F M., and H. 0. Hodgins. 1971. Biochemical polymorphisms in the Pacific hake (Merluc- cius productus). Rapp. P.-v. Reun. Cons. int. Explor. Mer 161:87-89. WlBORG, K. F 1951. The whirling vessel: An apparatus for the fractioning of plankton samples. Rep. Norw. Fish. Mar. Invest. 9(13): 1-16. J. C. Mason Department of Fisheries and Oceans Fisheries Research Branch Pacific Biological Station Nanaimo, British Columbia V9R 5K6, Canada STRANDED ANIMALS AS INDICATORS OF PREY UTILIZATION BY HARBOR SEALS, PHOCA VITULINA CONCOLOR, IN SOUTHERN NEW ENGLAND Since Federal protection began in 1972, the New England population of harbor seals, Phoca vitulina concolor, has more than doubled (Gilbert and Stein 19811; Payne and Schneider 1984), increasing at a site in southeastern Massachusetts at an average rate of 11.9% per year (Payne and Schneider 1984). One of the primary management concerns regarding the New England seal population is the increasing potential for conflict between commercial fisheries and harbor seals (Prescott et al. 19802). Seals have been shown to be significant consumers Gilbert, J. R., and J. L. Stein. 1981. Harbor seal populations and marine mammal fisheries interactions. National Marine Fish- eries Service, NOAA, Northeast Fisheries Center, Contract No. NA-80-FA-C-00029, Woods Hole, MA 02345, 55 p. 2Prescott, J. H., S. D. Kraus, and J. R. Gilbert. 1980. East Coast/Gulf Coast Cetacean and Pinniped Workshop. Marine Mam- mal Commission (MMC), Final Report, Contract 79/02. (Available National Technical Information Service, Springfield, VA 22151 as PB80-160104, 142 p.) of marine production (Brodie and Pasche 1982) and have been implicated as competitors for commer- cially valuable fish stocks, impacting fisheries through direct predation, gear damage, and en- tanglement (Boulva and McLaren 1979; Everitt and Beach 1982; Brown and Mate 1983). Despite the significant increase in harbor seal abundance, only anecdotal information exists on the diet of harbor seals along the eastern United States, lb assess the impact of this common predator on fish and squid, information is required on the food species exploited. In the past, seals were killed to facilitate quanti- tative analysis of their stomach contents (Imler and Sarber 1947; Spalding 1964; Boulva and McLaren 1979; Pitcher 1980a), although this procedure is im- practical in New England. Two alternatives to this method are the analysis of the stomachs of strand- ed animals, and the examination of seal feces col- lected on accessible haul-out sites (Pitcher 1980b; Treacy and Crawford 1981; Brown and Mate 1983). The first alternative for determining the food habits of the southern New England seal population was provided by the more than 500 harbor seals that have been found stranded south of Maine since 1977. The stranded seals were collected by the New England Aquarium (NEA), Boston, MA. The major- ity (59%) of the seals were collected between January and March (Table 1) along the perimeter of Cape Cod Bay, MA, primarily on the eastern side. This corre- sponds to the time when the peak number of seals occur south of Maine (Schneider and Payne 1983). Most of the stranded seals (65%) came from one year, 1980 (Table 1), when over 445 seals died of acute pneumonia associated with influenza virus (Geraci et al. 1982). Upon necropsy at the NEA, most of the stomachs and intestinal tracts of the stranded seals were found to be empty. Only 63 stomachs contained food mat- ter, and the contents from those were frozen for later Table 1.— Monthly distribution of stranded P. v. concolor contain- ing prey items examined 1977-83. Month 1977 1978 1979 1980 1981 1982 1983 Total Jan. 1 15 1 17 Feb. 7 2 1 10 Mar. 10 10 Apr. 1 1 May 1 1 1 2 1 6 June 1 2 1 4 July 1 1 Aug. 2 1 1 4 Sept. 3 2 5 Oct. 1 1 Nov. 1 1 Dec. 2 1 3 Totals 1 1 3 41 3 9 5 63 FISHERY BULLETIN: VOL. 84, NO. 1, 1986. 217 examination. In the fall of 1983, we pilot-tested the analysis of stomach contents from stranded seals using those 63 stomach samples as an indicator of prey utilization. The objectives of this study were 1) to identify prey items selected by seals in southern New England and 2) to determine whether stomach contents from stranded animals can provide accurate information on the utilization of most kinds of prey. Methods The stomachs were thawed and the contents wash- ed with water through a series of nested sieves (1.80, 1.00, and 0.50 mm2). Identifiable materials were rough-sorted into fish and fish components, inverte- brates and invertebrate components. Intact speci- mens and cephalopod beaks were preserved in a 70% ethanol-30% glycerin solution. Persistent prey hard parts (primarily otoliths) were removed and stored dry in glass vials. Otoliths from the stomach samples were identified against a reference collection at the National Marine Fisheries Service, Northeast Fisheries Center (NMFS/NEFC), Woods Hole, MA. Cephalopod beaks were identified against a reference key (Clarke 1962). To estimate the size of fish taken by harbor seals, otoliths removed from the stomach samples were measured under a dissecting microscope using ver- nier calipers. Regression equations relating otolith length to fish length (Frost and Lowry 1980; Brown and Mate 1983) were calculated using measurements obtained from the reference collection of fishes col- lected in the Gulf of Maine, located at the NMFS/ NEFC. Fork lengths were estimated for four prey species. Results Fifty-three stomachs (84%) held identifiable food items (Table 2). Cephalopod beaks were recovered from 35 stomachs, representing at least 168 in- dividuals and 2 species. Thirty-three stomachs con- tained beaks from the short-finned squid, Illex il- lecebrossus, with a range of 1-22 beaks per stomach. Beaks of the long-finned squid, Loligo pealei, were found in two stomachs, ranging from 4 to 5 beaks per stomach, and accounted for only 5% of the squid recovered. The two species were not found together in any of the stomachs. Twenty-nine stomachs con- tained squid remains and no other type of prey. Six stomachs contained both squid and fish remains. Seventeen stomachs contained some fish remains, including intact specimens, copious semidigested flesh, and 121 free otoliths. In total, seven species and five families were represented. Fourteen stomachs held otoliths from only one species of fish, while seven stomachs contained otoliths from more than one fish species. Four species of Gadidae comprised the majority of all fish species found in the stomachs of the stranded seals. A total of 86 otoliths in six stomachs were recovered. Haddock, Melanogrammus aegle- finus, was the most frequently found gadid (45 otoliths in four stomachs) with a maximum of 24 otoliths recovered from a single stomach. Silver hake, Merluccius bilinearis, remains were found only slightly less frequently (34 otoliths from three stomachs). Pollock, Pollachius virens, otoliths were found in one stomach (five otoliths), and two red hake, Urophycis chuss, otoliths of equal length were recovered from one stomach, presumably from a single fish. Fifteen free otoliths and three intact specimens of American sand lance, Ammodytes americanus, were recovered from two stomachs, and three stomachs contained otoliths from members of the flatfish family Pleuronectidae Two stomachs contained shells: the Atlantic mussel, Mytilus edulis, and the common slipper shell, Crepidula fornicata. The estimated mean fork length for the four gadid prey species ranged from 170 to 340 mm (Table 3). Regressions were not available to estimate the lengths of the sand lance found in the stomachs; however, studies on sand lance in Cape Cod Bay found a mean size of 93 mm SL (Richards 1982). Table 2. — Analysis of stomach contents from stranded harbor seals, P. v. concolor in Southern New England, 1977-83. Stomach {N = 63) Frequency Min. no. Species N % animals Cephalopoda: Illex illecebrossus 33 58.4 159 Loligo pealei 2 3.7 9 Mytilidae: Mytilus edulis 2 3.7 12 Calyptraeidae: Crepidula fornicata 2 3.7 10 Clupeidae: Clupea harengus 1 1.8 1 Gadidae: Melanogrammus aeglefinus 4 5.6 23 Pollachius virens 1 1.8 3 Urophycis chuss 1 1.8 1 Merlucciidae: Merluccius bilinearis 3 5.6 17 Ammodytidae: Ammodytes americanus 2 3.7 11 Pleuronectidae: Pseudopleuronectes americanus 3 5.6 10 Unidentified pisces 11 20.8 218 Table 3. — Estimated sizes of four fish prey species of harbor seals in Southern New England, based on regression equations relating otolith length (OL) to fish fork length (FL). Species Regression equation Estimated prey size (FL, mm) r2 Range Mean Melanogrammus aeglefinus Merluccius bilinearis Pollachius virens Urophycis chuss FL = 3.4(OL) - 9.32 0.97 45 110-310 230 FL = 22.4(OL) - 1.44 0.98 34 30-460 170 FL = 4.9(OL) - 22.58 0.95 5 160-310 280 FL = 25.0(OL) + 0.63 0.96 2 340 Discussion Analyzing stomach contents from stranded ani- mals to determine prey preference or selection does yield a partial list of prey species exploited; however, several apparent biases prohibit the realization of ac- curate quantitative results. Therefore, the utility of this method is questionable The limited number of stomachs containing food was likely due to the weakened condition of seals prior to stranding and their inability to obtain food. The stomachs that did contain food all came from stranded animals, and therefore may not reflect on what a healthy seal was feeding. The stranded seals were generally animals with debilitating conditions like lungworm and heartworm, and may not have been able to feed in usual feeding areas, or secure usual prey, and thus were probably less selective about prey items. For example, the shells found in the two stomachs may represent prey items desirable only to a disease- weakened seal. The size and number of these shells suggest that they were not ingested incidentally. Comparing the stomach contents to a "condition index", such as length vs. girth or blubber thick- ness, might indicate whether the stranded animals are less selective about prey species than healthy ones. The abundance of squid beaks found in the stomachs suggests that squid are an important part of the diet of harbor seals along coastal New England; however, our own finding of squid beaks in 56% of 63 stomachs may be inflated. Boulva and McLaren (1979) found squid remains in 20.6% of 279 stomachs examined from eastern Canada, and Pit- cher (1980b) similarly found cephalopod beaks in 21.1% of 351 harbar seals collected in the Gulf of Alaska. Seals have been shown to retain, then re- gurgitate, cephalopod beaks rather than pass them through their digestive tract (Miller 19783; Pitcher 1980b). Retention of squid beaks will tend to over- represent the utilization of squid as a prey species (Pitcher 1980a). The retention of beaks during a period of fasting prior to death may also account for the large percentage (41%) of stomachs containing squid beaks and no other type of prey remains. Large fish may be underrepresented if the heads (i.e, otoliths) are not eaten (Boulva and McLaren 1979; Brown and Mate 1983). Pitcher (1980b) sug- gested that seals often fragment large fish while eating them, usually discarding the head. Finally, the relationship between the time when prey was eaten and when the stomach was collected may determine what types of prey remains will be recovered (Frost and Lowry 1980; Pitcher 1980a; Brown and Mate 1983). For example, the low num- ber of sand lance otoliths found in the stomachs may not accurately represent the importance of sand lance as a prey species of harbor seals in southern New England because otoliths of the size of the ones recovered are very small and delicate and may not remain for long in the seal stomachs once freed from the skull (Smith and Gaskin 1974). Thus, using only frequency of occurrence as a measure of prey preference or selection may be mis- leading by overemphasizing the importance of some species. For example, based on number, cephalopods were the major prey item; however fewer otoliths representing fish of greater weight may show that fish indeed are more improtant. The full importance of fish or squid in the diet of seals can be accurately described only if quantitative assessments such as weight or volume of food items in the stomachs can be determined (Rae 1973; Frost and Lowry 1980). In summary, given a large sample of animals the analysis of stomach contents from stranded seals does provide information on the types of prey selected. However, the analysis of stomach contents from stranded seals greatly overemphasizes cephal- opod remains while likely underrepresenting most 3Miller, L. K. 1978. Energetics of the northern fur seal in rela- tion to climate and food resources of the Bering Sea. Marine Mam- mal Commission, Final Report, Contract MM5AC025. (Available National Technical Information Service, Springfield, VA 22151 as PB-275 296, 32 p.) 219 species of fish prey due to an extended period of fasting prior to stranding. We consider comparative frequencies of selected prey to be too biased to be useful in any ranking of prey items. Therefore, this technique of analyzing prey utilization should be con- sidered only if the examination of feces or the stomach contents from seals that were healthy when collected are not possible options. Acknowledgments We wish to thank all those from the New England Aquarium, Marine Mammal Rescue and Release Pro- gram, who helped collect the stranded animals. Paul J. Boyle and Kevin D. Powers commented on previous drafts of this manuscript. Research was conducted with the New England Aquarium's Edgerton Re- search Laboratory. This study was funded by Na- tional Marine Fisheries Service/Northeast Fisheries Center, Contract No. NA-82-FA-00007. Literature Cited Boulva, J., and I. A. McLaren. 1979. Biology of the harbor seal, Phoca vitulina, in Eastern Canada. Fish. Res. Board Can., Bull. 200:1-24. Brodie, P. F., and A. J. Pasche. 1982. Density-dependent condition and energetics of marine mammal populations in multispecies fisheries management. In M. C. Mercer (editor), Multispecies approaches to fisheries management advice, p. 35-38. Can. Spec Publ. Fish. Aquat. Sci. 59. Brown, R. F, and B. R. Mate. 1983. Abundance, movements, and feeding habits of harbor seals, Phoca vitulina, at Netarts and Tillamook Bays, Ore- gon. Fish. Bull., U.S. 81:291-301. Clarke, M. R. 1962. The identification of cephalopod "beaks" and the rela- tionship between beak size and total body weight. Bull. Br. Mus. (Nat. Hist), Zool. 8:421-480. Everitt, R. D., and R. J. Beach. 1982. Marine mammal-fisheries interactions in Oregon and Washington: An overview. In K. Sabol (editor), Transactions of the North American Wildlife and Natural Research Con- ference, p. 265-277. Wildl. Manage. Inst, Wash., DC. Frost, K. J., and L. F. Lowry. 1980. Feeding of ribbon seals (Phoca fasciata) in the Bering Sea in Spring. Can. J. Zool. 58:1601-1607. Geraci, J. R., D. J. St. Aubin, I. K. Barker, R. G. Webster, V. S. Hinshaw, W. J. Bean, H. L. Ruhnke, J. H. Prescott, G. Early, A. S. Baker, S. Madoff, and R. T. Schooley. 1982. Mass mortality of harbor seals: pneumonia associated with influenza A virus. Science 215:1129-1131. Imler, R. H., and H. R. Sarber. 1947. Harbor seals and sea lions in Alaska. U.S. Fish. Wildl. Serv., Spec. Sci. Rep. 28, 23 p. Payne, P. M., and D. C. Schneider. 1 984. Yearly changes in abundance of harbor seals at a winter haul-out site in Massachusetts. Fish. Bull., U.S. 82:440-442. Pitcher, K. W. 1980a. Food of the harbor seal, Phoca vitulina richardsi, in the Gulf of Alaska. Fish. Bull., U.S. 78:544-549. 1980b. Stomach contents and feces as indicators of harbor seal, Phoca vitulina, foods in the Gulf of Alaska. Fish. Bull., U.S. 78:797-798. Rae, B. B. 1973. Further observations on the food of seals. J. Zool. (Lond.) 169:287-297. Richards, S. W. 1982. Aspects of the biology of Ammodytes americanus from the St. Lawrence River to Chesapeake Bay, 1972-75, in- cluding a comparison of the Long Island Sound postlarvae with Ammodytes dubius. J. Northwest Atl. Fish. Sci. 3:93- 104. Schneider, D. C, and P. M. Payne. 1983. Factors affecting haul-out of harbor seals at a site in southeastern Massachusetts. J. Mammal. 64:518-520. Smith, G. J. D, and D. E. Gaskin. 1974. The diet of harbor porpoises (Phocoena phocoena (L.)) in coastal waters of eastern Canada, with special reference to the Bay of Fundy Can. J. Zool. 52:777-782. Spalding, D J. 1964. Comparative feeding habits of the fur seal, sea lion, and harbour seal on the British Columbia coast. Fish. Res. Board Can., Bull. 146, 52 p. Treacy, S. D, and T W. Crawford. 1981. Retrieval of otoliths and statoliths from gastrointestinal contents and scats of marine mammals. J. Wildl. Manage. 45:990-993. Lawrence A. Selzer Manomet Bird Observatory, Manomet, MA 0231*5 Greg Early Patricia M. Fiorelli New England Aquarium, Boston, MA 02110 P. Michael Payne Manomet Bird Observatory, Manomet, MA 0231*5 Present address: Boston University Marine Program, Woods Hole, MA 0231*5. Robert Prescott Massachusetts Audubon Society, South Wellfleet, MA 02663 SCAVENGER FEEDING BY SUBADULT STRIPED BASS, MORONE SAXATILIS, BELOW A LOW-HEAD HYDROELECTRIC DAM1 A spawning run of striped bass, Morone saxatilis, has not been found in the Connecticut River, but subadults from other rivers were reported in the lower 100 km of the river in the 1930's (Merriman Contribution No. 84 of the Massachusetts Cooperative Fishery Research Unit, which is supported by the U.S. Fish and Wildlife Service, Massachusetts Division of Fisheries and Wildlife, Massa- chusetts Division of Marine Fisheries, and the University of Massachusetts. 220 FISHERY BULLETIN: VOL. 84, NO. 1, 1986. 1941). Subadults enter the river in the spring and summer, often in enough abundance to support a sport fishery in Connecticut (Moss 1960). No striped bass were passed upstream in the two Holyoke Dam fish lifts located at river km 140 from the initial operation in 1955 until 1979, when 103 were lifted. Each year from 1980 to 1984, 110-510 striped bass have used the fish lifts (O'Leary 1985). In 1982, 83.5% of the fish were age II; 16.5% were age III; and none were sexually mature (Warner 1983). Because the striped bass did not migrate into the river to spawn, they probably entered to feed. The food of striped bass has been extensively studied, but there is no published report about the food of young fish that gather below a hydroelectric dam. We studied the food of the striped bass that were lifted at Holyoke Dam in 1982. Methods The stomachs of fish were removed and frozen, and the contents were examined in the laboratory with a dissecting microscope Stomach contents were classified as small forage fish, body parts of large fish (i.e., fish larger than the striped bass could eat whole), insects, plant material, and empty. Body parts were the scales, bones, flesh, and ovaries of adult alosids (i.e, American shad, Alosa sapidissima, and blueback herring, A. aestivalis), and pieces of adult sea lamprey, Petromyzon marinus. The body parts originated from the following sources: fish that were injured or killed while attempting to pass the dam or to use the fish lifts, American shad that were discarded below the dam by sport fishermen, or turbine-induced injuries or mortalities of fish that passed through the hydropower turbine at the dam (Bell and Kynard 1985). When possible, small forage fish were identified to species and measured for total length. Insects were identified to order. We compared the frequency of occurrence of the four foods eaten by striped bass that were lifted early (25 May-14 June), when average daily passage of adult alosids in the lifts was about 28,000, with the foods eaten by striped bass that were lifted late (after 21 June), when the average daily lift of alosids was about 3,000. Results and Discussion We examined 78 stomachs of striped bass— 65 (83%) contained food. Sixty-nine percent of the stomachs with food contained the body parts of large fish (Fig. la). Of the stomachs with the body parts of large fish, 93% contained the scales of adult alosids, with many containing over 20 scales; 16% contained the body parts of adult sea lampreys. Small forage fish were second in the frequency of occurrence at 61%, and insects were third at 21% (Fig. la). Elvers of the American eel, Anguilla rostrata, (96 mm mean total length, range: 70-125 mm, N = 24) dominated the small forage fish category, occurring in 58% of the stomachs that con- tained forage fish. Elvers, migrating upstream from the ocean, may be delayed and concentrated by Holyoke Dam; perhaps striped bass follow the elvers upriver— both species occur in the fish lifts at the same time Cyprinids were identified in six of the stomachs with forage fish. All had a 2,4-4,2 tooth formula and were probably spottail shiners, Notropis hudsonius, a commonly observed minnow. Insects in stomachs were mayfly nymphs, order Ephemerop- tera, but only one or two mayflys were found in any stomach. There was a significant difference in the frequency of the four food groups in fish collected early and late (x2 = 12.6, P < 0.01). Fish parts dominated the stomach contents of early-lifted fish, whereas in late- lifted fish 54% contained parts of large fish, but 77% contained small forage fish (Fig. lb). Fifteen per- cent of the stomachs of early-lifted fish were empty, UJ o a: 100 UJ a. uu (a ) 80 60 40 20 n m FISH FORAGE PARTS FISH INSECTS PLANTS Figure 1— Percent occurrence of the four major foods in the stomachs of striped bass passed by the Holyoke fish lifts a) in all of 1982 and b) in fish sampled early (25 May-14 June, N = 39) and late (after 21 June, N = 26) 1982. 221 and 19% of the stomachs of late-lifted fish were empty. Food of the striped bass at Holyoke Dam was dominated by the body parts of adult American shad, blueback herring, and sea lamprey when many in- dividuals of these species were being lifted, and dominated by forage fish and insects, when the alosids and sea lampreys were scarce (Fig. lb). The reduced incidence of feeding on the body parts of large fish by striped bass lifted after 21 June was probably the result of a dramatic reduction in the availability of this food that occurred when the run of anadromous alosids diminished. Hollis (1952) found alosid scales in the stomachs of adult striped bass captured below Conowingo Dam on the Susquehanna River in Maryland, but he dis- missed these as accidental. In our study, alosid body parts occurred in stomachs too frequently to be ac- cidental. Many authors consider the food that is selected by striped bass to be directly related to the availability (Hollis 1952; Thomas 1967; Schaefer 1970). During the run of anadromous fish at Holyoke Dam, the most abundant food that is available for striped bass is likely the body parts of dead or in- jured American shad, blueback herring, and sea lam- prey although we were not able to confirm this by sampling below the dam. About 900,000 adult alosids were passed upstream in the fish lifts in 1982, and injuries and mortalities were commonly observed at the dam and fish lifts. Subadult striped bass may typically concentrate below hydroelectric dams and feed on the parts of fish (anadromous or freshwater species) that die or sustain injury while attempting to move upstream or downstream of the dam. Acknowledgments This research was supported by Federal Aid Pro- ject AFS-4-R-21 and Dingell- Johnson Project 5-29328 to the Massachusetts Division of Fisheries and Wildlife and the Massachusetts Cooperative Fishery Research Unit. Literature Cited Bell, C. E., and B. Kynard. 1985. Mortality of adult American shad passing through a 17-megawatt Kaplan turbine at a low-head hydroelectric dam. North Am. J. Fish. Manage 5:33-38. Hollis, E. H. 1952. Variations in the feeding habits of the striped bass, Roc- cus saxatilis (Walbaum), in Chesapeake Bay. Bull. Bingham Oceanogr. Collect, Yale Univ. 14:111-131. Merriman, D. 1941. Studies on the striped bass (Roccus saxatilis) of the Atlantic coast. U.S. Fish Wildl. Serv., Fish. Bull. 50:1-77. Moss, D. D. 1960. A history of the Connecticut River and its fisheries. Conn. Board Fish. Game, Hartford, 12 p. O'Leary, J. O. 1985. Connecticut River anadromous fish investigations. Mass. Coop. Fish. Res. Unit, Univ. Mass., D-J Proj. F-45-R-2 Rep., 19 p. Schaefer, R. H. 1970. Feeding habits of striped bass from the surf waters of Long Island. N.Y. Fish Game J. 17:1-17. Thomas, J. L. 1967. The diet of juvenile and adult striped bass, Roccus sax- atilis, in the Sacramento-San Joaquin River system. Calif. Fish Game 53:49-62. Warner, J. R 1983. Demography, food habits, and movements of striped bass, Morone saxatilis Walbaum, in the Connecticut River, Massachusetts. M.S. Thesis, Univ. Massachusetts, Amherst, 94 p. John Warner Boyd Kynard Massachusetts Cooperative Fishery Research Unit 204 Holdsworth Hall University of Massachusetts Amherst, MA 01003 GENETIC CONFIRMATION OF SPECIFIC DISTINCTION OF ARROWTOOTH FLOUNDER, ATHERESTHES STOMIAS, AND KAMCHATKA FLOUNDER., A. EVERMANNI The uncertain taxonomic status of two morphologi- cal types of Atheresthes (family Pleuronectidae) has led to some problems in fisheries surveys and stock assessments. Although data collection would be simplified if these types were conspecific morphs, a single classification would mask differences of distribution and abundance if each type actually represented a distinct species. Each type is described as a separate species: arrowtooth flounder, A. stomias, and Kamchatka flounder, A. evermanni, based on morphological differences in gill raker count, dorsal and anal fin rays, caudal vertebrae number, eye-dorsal fin distance, and relative position of the upper eye (Norman 1934; Wilimovsky et al. 1967). However, the differences are subtle, and both types have generally been considered A. stomias in fisheries surveys (e.g., Smith and Bakkala 1982). Atheresthes stomias occurs in the eastern Bering Sea and eastern North Pacific Ocean from about St. Matthew Island, southward through the eastern Ber- ing Sea and Gulf of Alaska, and along the North American coast to central California (Hart 1973). Atheresthes evermanni is distributed in the western 222 FISHERY BULLETIN: VOL. 84, NO. 1, 1986. Bering Sea and western North Pacific Ocean from the Anadyr Gulf, south along the Kamchatka Pen- insula, through the Sea of Okhotsk, and to northern Japan (Andriyashev 1939; Wilimovsky et al. 1967). The geographic ranges of the two types overlap in some areas of the Aleutian Islands and eastern Ber- ing Sea. Biochemical data have recently provided valuable insights towards clarifying genetic relationships among fishes. Findings have ranged from identify- ing previously unknown species (eg., Shaklee et al. 1982) to grouping taxa previously considered distinct (eg, Wishard et al. 1984). Biochemical data were therefore used to clarify the taxonomic status of A. stomias and A. evermanni through an electro- phoretic examination of known individuals of both types. The level of genetic difference observed in this study is compared with that found between two other groups of marine fishes of the Bering Sea and the North Pacific Ocean. Materials and Methods Collections were made in the Bering Sea near Unalaska Island by National Marine Fisheries Ser- vice research vessels Oregon (lat. 53°45'N, long. 166°56'W, August 1980) and Miller Freeman (lat. 54°44'N, long. 166°29'W, February 1981). The 12 Kamchatka flounder (4 taken in 1980 and 8 in 1981) included males and females with fork lengths ranging from 24 to 43 cm. The 13 arrowtooth flounder, taken only in 1981, also included both sexes and ranged in fork lengths from 33 to 43 cm. Mor- phological types were distinguished by the gill raker counts and position of the upper eye In specimens identified as Kamchatka flounder, the upper eye did not reach the edge of the head and the mean total gill raker count was 12.4. The upper eye of specimens identified as arrowtooth flounder reached the edge of the head, breaking the dorsal profile and the mean total gill raker count was 15.3. Fish were frozen in- tact at -20°C following collection and remained frozen up to 30 mo until tissues were removed for electrophoretic analysis. Sample preparation and electrophoresis followed methods given by Utter et al. (1974). Buffer systems included 1) a discontinuous tris-citric acid (gel pH 8.2), lithium hydroxide-boric acid (tray pH 8.0) buf- fer, described by Ridgway et al. (1970); 2) a tris-boric acid - 0.004 M EDTA (pH 8.5) buffer, described by Markert and Faulhaber (1965); and 3) an aminopro- pylmorpholine-citric acid - 0.01 M EDTA (pH 6.5) buf- fer, described by Clayton and Tretiak (1972). Procedures of visualizing enzyme activities follow- ing electrophoresis were those outlined by May et al. (1979). We followed the criteria of Allendorf and Utter (1979) for the inference of Mendelian inheri- tance in the absence of breeding data. Genetic data were collected from 22 protein systems (Table 1). A system of nomenclature suggested by Allendorf and Utter (1979) was used where the most common allelic form of a locus was designated as 100, and other allelic forms were assigned values based on their mobility relative to the common form. Alleles migrating cathodally were given negative values. Phenotypic frequencies of the overall sample (all specimens of both presumed species pooled together) at each polymorphic locus were tested for expected binomial (i.e, Hardy-Weinberg) distributions using a G statistic for goodness of fit (Sokal and Rohlf 1969; Goodenough 1978). Multiple allelic cases were col- lapsed to two allelic classes to allow for small sam- ple sizes. A contingency table analysis of allelic fre- quencies testing the null hypothesis of no difference between the two groups also used the G statistic, Table 1.— Protein systems used in this study including tissues and appropriate buffer systems for detection of suitable activity. Enzyme commission Protein system number Tissues1 Buffed Acid phosphatase (ACP) 3.1.3.2 M,L,H 1,2,3 Adenosine deaminase (ADA) 3.5.4.4 M,E 1 Alcohol dehydrogenase (ADH) 1.1.1.1 L 3 Aldolase (ALD) 4.1.2.13 M 1,3 Aspartate aminotransferase (AAT) 2.6.1.1 M 1,2 Creatine kinase (CK) 2.7.3.2 M 1,3 Esterase (EST) 3.1.1.1 L,H,E 3 General protein (GP) M,E 2,3 Glucosephosphate isomerase (GPI) 5.3.1.9 M,E 1 Glyceraldehydephosphate dehydrogenase (GAP) 1.2.1.12 E,M 1,3 Glycerol-3-phosphate dehydrogenase (G3P) 1.1.1.8 M 3 Glycylleucine peptidase (GL) 3.4.11 E,M 1,2 Isocitrate dehydrogenase (IDH) 1.1.1.42 M,H,E 3 Lactate dehydrogenase (LDH) 1.1.1.27 M,E 3 Leucylglycylglycine peptidase (LGG) 3.4.11 M 1 Malate dehydrogenase (MDH) 1.1.1.37 H.L.E.M 3 Malate dehydrogenase (ME) (decarboxylating - NADP+) 1.1.1.40 M 2 Mannosephosphate isomerase (MPI) 5.3.1.8 M 2 Phosphoglucomutase (PGM) 2.7.5.1 M 1 6-phosphogluconate dehydrogenase (PGD) 1.1.1.44 M,E 3 Phosphoglycerate kinase (PGK) 2.7.2.3 M 3 Superoxide dismutase (SOD) 1.15.1.1 M,H 1,3 1M = muscle, L = liver, H = heart, E = eye. 21 = discontinuous tris citrate, lithium borate; 2 EDTA; 3 = continuous amine citrate, EDTA. continuous tris, borate, 223 with Yates correction for small sample sizes (Sokal and Rohlf 1969). Nei's (1978) measure of genetic distance for small sample sizes was calculated be- tween the two groups. Results and Discussions Data were collected from 22 enzyme systems en- coding the following 32 presumed loci (polymorphic loci having one or more variant alleles are indicated by *): AAT*, ACP-1, ACP-2*, ADA*, ADH*, ALD, G3P-1*, G3P-2, CK-1, CK-2, EST, GAP-1*, GAP-2, GL-1, GL-2, IDH*, LDH-1*, LDH-2, LDH-3, LGG*, MDH-1, MDH-2, MDH-3, ME, PGD*, PGM-1, PGM-2, GPI-1, GPI-2*, PGK*, MPI, SOD. Allelic distributions for the 13 polymorphic loci indicate considerable similarity for most of the systems but some distinct differences as well, based on the contingency analysis (Table 2). The nonsig- nificant differences observed at nine of the loci are not highly informative given the limited number of individuals that were sampled. However, the differences that were statistically sig- nificant provide considerable information. The most striking difference is at the ADH locus, where no alleles were shared between the 12 arrowtooth and the 10 Kamchatka flounders. These data alone con- firm the genetic distinctness of the two types. The allelic distribution between the two forms is almost as distinct at the GAP-1 locus; a lesser, but signifi- cant difference also exists at the ACP-2 locus. Gel banding patterns observed for these three loci are shown in Figure 1. Not surprisingly, the genotypic frequencies at the ADH and GAP-1 loci also deviated significantly (P < 0.001) from the ratios of a binomial expansion of allelic frequencies (Hardy-Weinberg equilibrium ex- pected in a single, randomly mating population). This difference resulted from excesses of homozygous and deficits of heterozygous classes, a situation expected in population mixtures (i.e., the Wahlund effect, see Futuyma 1979). The distinct genotypic distribution of the two forms at the ADH and GAP-1 loci, coupled with their sympatric occurrence and subtle but consistent mor- phological identities, support their present tax- onomic status as distinct congeneric species. How- ever, the value of genetic distance observed, 0.052, is rather low for distinct species suggesting recent speciation (Avise 1976). Recent genetic studies of two other pleuronectid species sampled from the same geographic region indicate only conspecific variation. The Alaska Pen- insula separates two population groups of yellowfin sole, Limanda aspera, at a mean genetic distance of 0.005 (Grant et al. 1983). No significant differences of allelic frequencies were detected in Pacific halibut, Hippoglossus stenolepis, sampled in the Bering Sea and the North Pacific Ocean (Grant et al. 1984). These various outcomes among confamilial group- ings undoubtedly reflect the past and present actions of numerous variables; two major factors are differ- ing capabilities for gene flow based on distinct life history patterns, and differing times and degrees of isolation imposed by glaciation events within the past 2 million years (discussed by Grant and Utter 1984). Finally, the possibility of hybridization and intro- gression between the two species of Atheresthes should be examined through more extensive sam- pling of these two forms over a broader geographic range The distinct distribution of ADH alleles ex- cluded a hybrid origin of any individuals in this study. Table 2. — Observed number and (in parentheses) within group fre- quency of alleles of 13 polymorphic loci in samples of arrowtooth and Kamchatka flounder. Allele Observed alleles (frequencies) P1 Subunit structure2 Locus Arrowtooth Kamchatka AAT 92 100 106 2(0.100) 10(0.500) 8(0.400) no data d ACP-2 100 109 20(0.769) 6(0.231) 22(1 .000) 0(0.000) <0.05 m ADA-1 100 108 24(0.923) 2(0.077) 19(0.792) 5(0.208) ns m ADH -100 -75 -13 24(1.000) 0(0.000) 0(0.000) 0(0.000) 1(0.050) 19(0.950) <0.001 d G3P-1 100 150 24(1.000) 0(0.000) 19(0.950) 1(0.050) ns d GAP-1 13 70 100 0(0.000) 0(0.000) 26(1.000) 9(0.375) 12(0.500) 3(0.125) <0.001 t GPI-2 100 107 25(0.962) 1(0.038) 24(1.000) 0(0.000) ns d IDH 70 100 0(0.000) 26(1.000) 3(0.125) 21(0.875) ns d LDH-3 100 117 26(1 .000) 0(0.000) 22(0.917) 2(0.083) ns t LGG 86 100 1(0.038) 25(0.962) 0(0.000) 22(1.000) ns d PGD 75 100 4(0.154) 22(0.846) 0(0.000) 22(1.000) ns d PGM-1 84 100 105 113 0(0.000) 23(0.885) 3(0.115) 0(0.000) 1(0.042) 22(0.916) 0(0.000) 1(0.042) ns m PGK 100 133 26(1.000) 0(0.000) 19(0.950) 1(0.050) ns m 'Contingency tests of allelic frequencies using the G-statistic with Yates cor- rection for small sample sizes, assuming all samples drawn from the same population; ns = not significant. 2Protein subunit structure based on observed banding patterns of variants; m = monomer, d = dimer, t = tetramer. 224 Literature Cited 'en ACP-2 qo oo o o o o c 'en I Q < 1 1 c 'en GAP-1 a; a < C a! Q s~ -C CO n -M T r~ o «4H -a 0) > in 00 a> I T O w -*-> C ea o CJ o o cS I I > o cfl c ■-. 0) -t-» c3 a bO C -o c ea _Q cB" 73 E 73 CD ci-O *— to CO CD Q. O O O .c ■d a 03 —1 0 > CO 73 73 "5 ■a &o co CD a. ■0 0 c O to O O) ffl > to — 73 73 73 O Q. CD Q. O O 73 Oo CD a. 0 0 O) P_ '1— 3 73 > to 73 CD 73 3 O 0 CO CD Q. __ O CO O c w. CD *- X 03 ■_ 73 — CD 73 O 73 CD 0 CD a. to CD Q. O O C O O CO 'C CO __ a. — £ 73 0 73 O O l a. 1 CD CO Q. O O m £ 73 73 0 a. CD a. 0 O 73 S P CD E a> CD o £ 1- CO 73 p CD E a> CD CO O) CO 73 0) P CD E O) CD ep 4 73 03 p CD E O CD to O .E 73 © g I £<■> E, CO C33 O *° CO .s 4 05 73 S -~ 31 £ o Q. 3> P c ® 5, t 10 01 0,i « U® «) .E 4 73 CD P CD E O) CD CO 00 to 73 CD 73 03 P CD E O) CD CO 4 73 CD p c CD CD E E O) C3> CD CD to to to in 4 73 CD 73 CD c p CD CD E E o> OS CD CD CO CO l*» in 4 73 73 CD £ c P CD CD E E u> TO CD CD CO cp co in 4 CD O 1 1 "o S CD O || Q. 73 03 P CD E CD to in in >>«9 |i 0 g> 6 <" z "5 8 CD O II CD O .p V 9, CD O I I CD O 1 1 CD O p E 6 <" Q. •o CD 73 P I >, 03 .E = P Q. 73 <* O .E 5 CO c c 03 c co co c C CD II £ "55 co a 4 $ n g CD O 03 c CD 03 co 0 CO c E C33 O) c 75 c E CD O E 73 03 P CD E s 11 * C 3 O C CD CD O i to to T3 .b s> 2»s e o "CO c S Q3 E CO ii CVJ ■5 "P CD 03 C C 03 03 E E C33 C33 03 03 tp to CO CO "2 S 03 O a. S •a 73 •a CO CO 03 03 £ co c 0 C -0 03 C c c CD — 03 03 CO 03 CD E E a) c E c i E E 03 0)7; to $ &co O) CT S 03 E 03 to 03 CO 00 m >* to CO c C 03 II S Id ffl o "ffl CD CD J3 C CO iii C3) CD = to CO CO 2 W CO CO 73 73 « 52 C C 03 03 E E a> 03 CD 03 to to CO CO to CO c C 03 II £ "5 ffl o "CO c co £ c6 73 03 £ "o S 03 03 CO E c CD C3)= 10 CD 2 io Q-Cvj 73 03 c C CD 03 E E a C33 03 CD CO CO c C CD II 5J © CO CO CD 573f CM « .-E 03 CM 5 ^ » ai 01 01 S E CM CM s to CO CO c 0 c CD — E E 03 CO c .2 a> £ co S c 03 E CD S "CD CO c I ES _- CO O) « ^ s p 3 (/) CO ^E S - C 03 3 CO 73 73 CD 03 C c CD 03 E E en O) CD 03 CM CM co m CO E to 73 73 03 £ C C CD 03 E E C33 C33 03 CD to CM OJ CD 73 J5 n £ a. E CD CO ^* © , m CO 03 « 4 5 03 C — CD to CO ■-- C 73 C CD 03 E E ■5. ~ 73 S? c a> 03 03 >- 03 £ "55 CD 03 CO CO 73 C o o 03 CO E 3 1— CO _03 n 73 c CO X CO E 73 c o o 03 C/3 73 03 Q. CD 03 £ £ ^_ ^_ 0 0 I I CO CO CC a: c CD E o) a> u> S 03 03 03 03 to to co to CO CO CM 1- 73 73 73 £ £ £ t- P P P O — 03 03 03 CO E E E S f a, as 03 O) -2 p cd 03 03 03 03 03 i- 10 CO CO tO 03 CO CO CO CM CM 73 73 73 £ £ £ P P P 03 03 03 E E E C3) a> a) -S CD 03 03 03 CO CO to to co CO CO CM 1- 73 73 73 03 03 £ C P P CD 03 03 E E E ffl C33 03 a> 03 CD 03 03 to T CO to CO CM CM T_ 73 73 73 a> CD 03 c P P 03 03 03 E E E ffl C33 O) u> 03 CD 03 03 to CO to to C\J CM CM T~ 73 73 73 £ £ £ C p E CD CD CD E E E s DJ 03 C33 CD 03 CD CD n to to to CM CM CM T~ 73 73 03 £ C P 0) CD E C33 a> CO E C3) CD CO co 03 to c CD to CM T- CM CO 73 03 C CD E 03 CO c c CD 03 to 03 to CD to a CD to v- CM CO co >. c 0 03 4— ^- ^, co c c c 03 to CD CD CD to to to JD X3 JD CM ro CO CO CO ■* a> £ a> 0 E CO CC 0 I CO DC m CD 03 03 _1 C33 03 _J 243 FISHERY BULLETIN: VOL. 84, NO. 2 tained in Gooding's thesis (1963) have not been re- examined, it appears likely that the presence of a very reduced maxilliped in the adult female is a generic character in Leptinogaster. Gooding (1963:218-220) discussed the generic status ofSaphirella T. Scott, 1894, pointing out that species of Saphirella may represent Copepodid I stages of clausidiids. In his thorough description of Copepodid I of Leptinogaster a significant difference seems to be in the body length, which Gooding gave as 0.45 mm, while in this study the length is 0.57 mm (0.45-0.60 mm). Although the genus Leptinogaster has been assign- ed to various families (Table 1), its presently agreed location appears to be in the Clausidiidae Embleton, 1901, along with Clausidium Kossmann, 1874, Con- chyliurus Bocquet and Stock, 1957a, Giardella Canu, 1888, Hemieyclops Boeck, 1873, Hersiliodes Canu, 1888, and Hippomolgus G.O. Sars, 1917. [According to the phylogenetic analysis of Ho (1984), the genus Myzomolgus Bocquet and Stock, 1957b, should be removed from the Clausidiidae and placed close to the Catiniidae Bocquet and Stock, 1957b.] The family Clausidiidae, containing seven genera of certain status, shows several features: first anten- na 6- or 7-segmented; second antenna 4-segmented with third segment having in some cases prehensile elements and fourth segment without a strong claw; mandible with spine (or spinelike process) and 2 or 3 accessory elements (setae, spines); labrum with rounded margin, mostly entire without median in- dentation, except triangular in Leptinogaster; first maxilla often with 2 lobes, but with 1 lobe having 2 groups of setae in Leptinogaster and 1 lobe with a few setae in Clausidium; maxilliped in female mostly 2-, 3-, or 4-segmented, but in Leptinogaster reduced to 2 setae; maxilliped in male 2- or 3-seg- mented plus claw (in Hippomolgus male unknown); legs 1-4 biramous and 3-segmented (endopod of leg 1 bearing suckers in Clausidium); leg 5 2-segmented (though in some first segment not clearly separated from body). Leptinogaster falls within this concept of the family Clausidiidae Neighboring families have fun- damentally different features, e.g., the Clausiidae (first antenna 3-6 segmented; legs 1-4 showing various degrees of reduction (as characterized by Wilson and Illg (1955)), the Myicolidae (3-segmented second antenna with strong terminal claw, max- illiped in female a small unarmed lobe), and the Ergasilidae (second antenna with a strong terminal claw, maxilliped often absent in female, legs 1-4 with some reduction). More information on the develop- mental stages of the members of these families would contribute greatly to understanding their interrelationships. ACKNOWLEDGMENTS I thank Roger F. Cressey who aided in the collec- tion of the copepods from Mya at Cotuit in 1957 and who provided M. S. Wilson's notes and correspon- dence concerning Leptinogaster (= Myocheres) ma- jor which are in the custody of the National Museum of Natural History, Smithsonian Institution. I thank also Geoffrey A. Boxshall, British Museum (Natural History), and Paul L. Illg, University of Washington, for helpful suggestions. LITERATURE CITED Allen, J. A. 1956. Myocheres inflata a new species of parasitic copepod from the Bahamas. J. Parasitol. 42:60-67. B&CESCU, M., AND F. POR. 1959. Cyclopoide comensale (Clausidiide si Clausiide) din Marea Neagra si descrierea unui gen nou, Pontoclausia gen. nov. In Omagiu lui Traian Savalescu cu prilejul implinirii a 70 e ani, p. 11-30. Acad. Rep. Pop. Rom. Bocquet, C, and J. H. Stock. 1957a. Copepodes parasites d'invertebres des cotes de France I. Sur deux genres de la famille des Clausidiidae, commen- saux de mollusques: Hersiliodes Canu et Conchyliurus nov. gen. Proa K. Ned. Akad. Wet, Ser. C Biol. Med. Sci. 60: 212-222. 1957b. Copepodes parasites d'invertebres des cotes de France IVa. Le double parasitisme de Sipunculus nudus L. par Myzomolgus stupendus. nov. gen., nov. sp., et Catinia plana nov. gen., nov. sp., copepodes cyclopoi'des tres remarquables. Proa K. Ned. Akad. Wet, Ser. C Biol. Med. Sci. 60:410- 431. 1958. Copepodes parasites d'invertebres des cotes de la Man- che, IV. Sur les trois genres synonymes de copepodes cyclopoi'des, Leptinogaster Pelseneer, Strongylopleura Pelseneer et Myocheres Wilson (Clausidiidae). Arch. Zool. Exp. Gen. 96:71-89. Boeck, A. 1873. Nye Slaegter og Arter af Saltvands-Copepoder. Forh. Vidensk.-Selsk. Christiania (1872), p. 35-60. Canu, E. 1888. Les copepodes marins du Boulonnais. III. Les Her- siliidae, famille nouvelle de copepodes commensaux. Bull. Sci. Fr. Belg. 19:402-432. Causey, D. 1953. Parasitic Copepoda from Grand Isle, Louisiana Occas. Pap. Mar. Lab., La. State Univ. No. 7, 18 p. Deevey, G. B. 1948. The zooplankton of Tisbury Great Pond. Bull. Bingham Oceanogr. Collect, Yale Univ. 12:1-44. 1960. The zooplankton of the surface waters of the Delaware Bay region. Bull. Bingham Oceanogr. Collect, Yale Univ. 17:5-53. Embleton, A. L. 1901. Goidelia japonica - a new entozoic copepod from Japan, associated with an infusorian (Trichodina). Trans. Linn. Soa Lond. 2d Ser., Zool. 28:211-228. 244 HUMES: COPEPODIDS AND ADULTS OF LEPTINOGASTER MAJOR Gooding, R. U. 1963. External morphology and classification of marine poecilostome copepods belonging to the families Clausidiidae, Clausiidae, Nereicolidae, Eunicicolidae, Synaptiphilidae, Catiniidae, Anomopsyllidae, and Echiurophilidae. Ph.D. Thesis, Univ. Washington, Seattle, 247 p. Ho, J.-S. 1984. New family of poecilostomatoid copepods (Spiophani- colidae) parasitic on polychaetes from southern California, with a phylogenetic analysis of nereicoliform families. J. Crustacean Biol. 4:134-146. Humes, A. G., and R. F. Cressey. 1958. Copepod parasites of mollusks in West Africa. Bull. Inst. Ft. Afr. Noire 20(A):92 1-942. 1960. Seasonal population changes and host relationships of Myocheres major (Williams), a cyclopoid copepod from pelecy- pods. Crustaceana 1:307-325. Humes, A. G, and R. U. Gooding. 1964. A method for studying the external anatomy of cope- pods. Crustaceana 6:238-240. Kossmann, R. 1874. Ueber Clausidium testudo, einen neuen Copepoden, nebst Bemerkungen uber das System der halbparasitischen Copepoden. Verh. Phys.-Med. Ges. Wurzburg, n.f. 7:280-293. Monod, T., and R.-Ph. Dollfus. 1934. Des copepodes parasites de mollusques (deuxieme sup- plement). Ann. Parasitol. Hum. Comp. 12:309-321. Pearse, A. S. 1947. Parasitic copepods from Beaufort, North Carolina. J. Elisha Mitchell Sci. Soc. 63:1-16. Pelseneer, P. 1929. Copepodes parasites de mollusques. Ann. Soc R. Zool. Belg. (1928) 59:33-49. Sars, G. O. 1917. An account of the Crustacea of Norway with short descriptions and figures of all the species. In Vol. VI Cope- poda Cyclopoida Parts XI and XII Clausidiidae, Lichomolgidae (part), p. 141-172. Bergen Museum, Bergen. Scott, T. 1894. Report on Entomostraca from the Gulf of Guinea, col- lected by John Rattray, B.Sc Trans. Linn. Soc Lond. 2d Ser., Zool. 6:1-161. Sewell, R. B. S. 1949. The littoral and semi-parasitic Cyclopoida, the Monstril- loida and Notodelphyoida. Sci. Rep. John Murray Exped. 9(2):17-199. Sharpe, R. W. 1910. Notes on the marine Copepoda and Cladocera of Woods Hole and adjacent regions, including a synopsis of the genera of the Harpacticoida. Proc. U.S. Nat. Mus. 38:405-436. Williams, L. W 1907. A list of the Rhode Island Copepoda, Phyllopoda, and Ostracoda, with new species of Copepoda In Thirty-seventh Annual Report of the Commissioners of Inland Fisheries of Rhode Island, p. 69-79. Spec. Pap. 30. Wilson, C. B. 1932. The copepods of the Woods Hole region Massachusetts. U.S. Nat. Mus. Bull. 158, 635 p. Wilson, M. S. 1950. A new genus proposed for Lichomolgus major Williams (Copepoda, Cyclopoida). J. Wash. Acad. Sci. 40:298-299. Wilson, M. S., and P. L. Illg. 1955. The family Clausiidae (Copepoda, Cyclopoida). Proa Biol. Soc. Wash. 68:129-141. 245 REPRODUCTIVE BIOLOGY OF FEMALE SPOTTED DOLPHINS, STENELLA ATTENUATA, FROM THE EASTERN TROPICAL PACIFIC A. C. Myrick, Jr., A. A. Hohn, J. Barlow, and P. A. Sloan1 ABSTRACT Reproductive parameters were estimated from about 4,700 female spotted dolphins collected in the eastern tropical Pacific from 1973 to 1981. From this sample, specimens for which ages were estimated were divided into two subsets and were used to estimate age-specific rates for the northern offshore stock of this species. The youngest sexually mature individual was 10 years old; the oldest immature was 17 years; the youngest and oldest pregnant individuals were 10 and 35 years, respectively. There was high individual variability in the accumulation of corpora with age; the ovulation rate appears to slow abruptly after the eighth ovulation. Average age at attainment of sexual maturity (ASM) for all years ranged from 10.7 to 12.2 years (x = 11.4 years) for two sets of age estimates; no significant temporal change in ASM was detected. Correlation between color phase and state of sexual maturity suggests that color phase may be a good indicator of maturity for this stock. The average annual pregnancy rate was about 0.33; this rate did not change significantly with age. The calving interval was 3.03 years (SE = 0.205). The lactation period was 1.66 years, but there was a significant increase noted in the percent lactating from 1973 to 1981. A low percentage of postreproductive females was found in the sample (0.4%) in- dicating that reproductive senescence is of little importance in reproductive rates of this stock. Purse seine operations of the yellowfin tuna fishery in the eastern tropical Pacific Ocean (ETP) have caused high mortality of the spotted dolphin, Stenella attenuata (Perrin 1969a, 1970). Estimated incidental kills for the northern offshore stock of spotted dolphins were between 100,000 and 400,000 annually throughout the 1960's and early 1970's (Smith 1983). Since 1968, research efforts by the Na- tional Marine Fisheries Service (NMFS) have focused on assessing the biological consequences of the large incidental kill of this and other affected dolphins using specimens and data collected by NMFS observers aboard U.S. tuna seiners. Perrin et al. (1976) presented the first comprehensive description of spotted dolphin life history and reproduction for specimens from the ETP. The ac- cumulation of thousands of additional specimens, the sharp decline in dolphin mortality (Smith 1983; Ham- mond and Tsai 1983), and the improvements made in estimating age since that study (Myrick et al. 1983) have made a new analysis necessary. The purpose of this paper is to estimate the reproductive parameters of the female spotted dolphin, based on analyses which include more data and a better age estimating method than previous 1 Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. studies. Reproductive features of the male spotted dolphin (Hohn et al. 1985) and temporal trends in reproduction in the northern offshore stock (Barlow 1985) are discussed in separate papers. MATERIALS AND METHODS Samples The specimens were analyzed as three samples. The "overall" sample contained about 4,700 speci- mens that had been collected from 1973 through 1981. A second sample for which ages were esti- mated contained 580 specimens selected randomly from more than 3,500 specimens collected in 1973 through most of 1978 (the 1973-78 aged sample). The randomly chosen 1973-78 aged sample did not in- clude any of the specimens studied by Perrin et al. (1976). The third sample (the 1981 aged sample) was composed of 226 specimens which had been collected in 1981 and for which ages were estimated. It in- cluded almost all specimens for which ovaries and teeth were collected in that year. The two aged sam- ples, referred to collectively as the aged sample, are subsets of the overall sample In several analyses the 1973-78 aged sample was divided into 1973-74 and 1975-78 subsamples in an effort to detect possible temporal changes in reproductive rates. Only the Manuscript accepted June 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 247 FISHERY BULLETIN: VOL. 84, NO. 2 northern offshore stock of spotted dolphins (as de- fined by Smith 1983) is treated in this analysis. The geographic boundary used to divide it from a south- ern stock is lat. 1°S (Henderson et al. 1980). Life History Data Data and specimens were collected by biological technicians aboard tuna purse seine vessels in the ETP. Biological data used in this analysis are body length, color phase, reproductive condition (preg- nant, lactating, or resting), and corpora counts for each specimen (see Perrin et al. [1976] for a descrip- tion of collection and examination procedures). Although there is no certainty that all ovarian cor- pora persist for life in all delphinids (Perrin and Re illy 1984), corpora counts were used with age to estimate ovulation rates. Counts included corpora albicantia (CAs), corpora lutea (CLs), and in some cases corpora atretica (atretic follicles). Only specimens that had both ovaries examined were in- cluded in the ovulation rate analyses. Age Estimates Ages were estimated for about 800 specimens (from 1973 to 1978 and 1981 samples) by counting growth layer groups (GLGs, Perrin and Myrick 1980) in the dentine and cementum of decalcified and hematoxylin-stained thin sections (Myrick et al. 1983). Tooth readings were made independently by two readers (A. C. Myrick and A. A. Hohn), without referring to field or laboratory data on size or repro- ductive condition. For the 1973-78 sample, a tooth of each specimen was read at least three times by each reader. Age estimates by each reader were sig- nificantly different (Reilly et al. 1983). To minimize the differences, the mean of the multiple age esti- mates by each reader was calculated and the average of the two means was used as the estimate of a specimen's aga For the 1981 sample a tooth from each specimen was read once by each reader after calibration tests showed that differences in estimates between readers were acceptably small (Reilly et al. 1983). An average of these two readings was used for specimen age. We consider the method we used to estimate ages improved over that used by Perrin et al. (1976) because 1) the preparation technique we used provides superior resolution of GLGs (Myrick et al. 1983); 2) the new method of reading utilizes GLGs in the cementum as well as in dentine and allows a more accurate estimate of maximum age for adults (Myrick et al. 1983; see also Kasuya 1976); 3) calibration of GLGs in tetracycline-labeled teeth of Hawaiian spinner dolphins, Stenella longi- rostris (Myrick et al. 1984), has provided a basis for interpreting dental layering within an absolute-time framework (Myrick et al. 1983; Myrick et al. 1984). Perrin et al. (1976) used the term tooth layers in lieu of known time units. RESULTS AND DISCUSSION Composition of Samples Chi-square (contingency) tests were used to evalu- ate whether fractions of mature, pregnant, and lac- tating females in the 1973-78 aged sample were a representative subset of the overall sample for those years. For all three tests, differences were not sig- nificant (P > 0.05). Reproductive statistics showed some differences between years (Table 1). Chi-square tests were carried out for homogeneity between 1973-74 and Table 1.— Number of sexually mature, pregnant only, lactating only, simultaneously pregnant and lactating, and "resting" female spotted dolphins, and the proportion of the sample pregnant or lactating in the aged and overall samples. The proportion pregnant and proportion lactating in- clude the simultaneously pregnant and lactating specimens. Number Proportion Years Sexually mature Pregnant only Lactating only Pregnant and lacting Resting Pregnant Lactating Aged 1973-74 188 57 87 7 38 0.34 0.50 1975-78 205 48 100 13 44 0.30 0.55 1981 149 34 86 9 17 0.29 0.64 Total 542 139 273 29 99 0.31 0.56 Aged and unaged 1973-81 2,979 780 1,480 151 568 0.31 0.55 248 MYRICK ET AL.: REPRODUCTIVE BIOLOGY OF SPOTTED DOLPHINS 1981 aged samples and between 1975-78 and 1981 aged samples for numbers of specimens pregnant, lactating, and resting. These tests revealed signifi- cant differences (1973-74 vs. 1981: xl = 7.46, P = 0.024; 75-78 vs. 1981: xl = 6.16, P = 0.046.). These differences are the result of an increase in the relative frequency of lactating females (see section on Lactation Period). There were no differences in percent pregnant during this time (see also Barlow 1985). Ovulation Rate Individual Variability Perrin et al. (1976) found high variability in the number of corpora (corpora atretica included) for a given age (in tooth layers). Nevertheless, by fitting a power curve to the average number of corpora as a function of average reproductive age, they deter- mined that the average ovulation rate slowed abruptly from about "four during the first layer, [to] two during the second, and about one per layer there- after" (Perrin et al. 1976, p. 261). The sexually mature specimens in the combined aged samples were used in our study to plot average frequency of corpora (corpora atretica excluded) on estimated age (Fig. 1). Regressions for the 1973-74 sample and for the 1981 sample are not significant- ly different; when the samples are pooled, the resulting slope is 0.61 corpora/yr. A plot of number of corpora on age for all individuals (n = 542) in mature age classes (10 through 38 yr old) for all aged specimens (Fig. 2) showed a significant slope (P < 0.0001) but a low correlation (r2 = 0.397), indicating high individual variability. For example, the sample included 12- and 13-yr-olds with 7 or 8 corpora, and 21-yr-olds with 4 or fewer corpora. A 38-yr-old had only 1 1 corpora (Table 2). These results support those of Sergeant (1962), Brodie (1971), Kasuya et al. (1974), and Perrin et al. (1976), that great individual variation occurs in ovulation rates among odonto- cetes. Table 2. — Summary of age-related reproductive statistics for female spotted dolphins taken in 1973-78 and 1981. Estimated age Variable (years) Range of ages with no corpora 0-17 Oldest with one corpus 23 Youngest with one corpus 10 Youngest pregnant 10 Oldest pregnant 35 Average age pregnant 18 Oldest simultaneously pregnant and lactating 29 Oldest lactating 36 Youngest lactating 10 Oldest 38 Changes in Rate Ovulation rate apparently decreases with repro- ductive age If ovulation and mortality rates were Figure 1— Linear regression of number of cor- pora on estimated age as gross estimates of ovulation rates in female spotted dolphins. Points represent averages for 1-yr age classes (1973-78 samples = closed circles; 1981 sam- ple = open circles). < a. O Q. cc O O u. O < 26 28 30 AGE ESTIMATES (years) 249 FISHERY BULLETIN: VOL. 84, NO. 2 • ! . I • : i : • • • * . I • I I I— L J—J L_l I I L_l ' ■ ' ( - t I I I I I _ J I I I I I I I I o 00 CO CO CO CO CNJ CO o CO 00 C\J CD CNJ CNJ CNJ CNJ o CNJ 00 CO CNJ 00 CO CNJ <0 (0 0) UJ o < CNJ CNJ O CNJ CNJ 00 CO CNI 00 CO CNJ 00 iH s 00 t> CO t~ 05 -a v o o in C IS a, "o -o -a o Q. w a E c 3 s a o © o S o •2J- ea I-. o o 13 CL, w & 5 daawnN vdoddoo 250 MYRICK ET AL.: REPRODUCTIVE BIOLOGY OF SPOTTED DOLPHINS constant, a semilog plot of the frequency distribu- tion of corpora counts would be linear. The slope of this line would be negative, and its value would be determined by mortality and ovulation rates. The observed shape of the log-frequency distribution of corpora counts for spotted dolphins (Fig. 3) suggests that ovulation and/or mortality rates are not con- stant. After about eight ovulations, log-frequencies decrease monotonically and nearly linearly. For up to the first eight ovulations, the rate is apparently much higher (presuming, again, that mortality rates do not change with the number of ovulations and that all CAs persist for life [Perrin and Reilly 1984]). This supports the findings of Perrin et al. (1976) that ovulation rates decrease with reproductive age in spotted dolphins. mates using a variation of the method described by DeMaster (1978). Age-specific maturation rates were used to calculate mean ASM as ASM = J. (x - 0.5) Px where x is age class, Px is the probability of first ovulating in age class x, and w is the maximum age in the sample. The term (x - 0.5) was substituted for DeMaster's (x) so that the mean age in an age- class interval would be represented by the midpoint of that interval. The terms Px were estimated as Px =f(x + 1) -f{x), Sexual Maturity The age at which a female first ovulates is con- sidered the age at attainment of sexual maturity (DeMaster 1978, 1984). Using the aged samples, we estimated average age at sexual maturity (ASM) using two methods. For these estimates, ages were grouped by 1-yr intervals: age-class 1 included specimens 0-1.0 yr, age-class 2 from 1.1 to 2.0 yr, etc The mean age of sexually mature females was 18.7 yr. Method-One ASM was estimated from both readers' age esti- where/(:r) is the probability of being mature at age x. The function f(x) was estimated as the best least- squares fit of a curve (York 1983) to the observed values of percent mature by age class. A 3-parameter sigmoid curve based on a modification of the logistic equation was found to give an adequate fit of the data (Fig. 4). ASMs were calculated separately for the aged samples, 1973-74, 1975-78, and 1981. There were no significant differences among these samples (P > 0.05). The ASM for all samples combined was 10.7 (var. = 0.03) to 12.2 (var. = 0.05) yr for the two readers. The average of these two ASM estimates was 11.4 yr. The precision between readers in age estimates of the 1981 specimens was greater than (O 400 300 200 2 100 z yo < 80 LL 70 o 60 0.50). This result differs from that of Perrin et al. (1976) which indicated a significant reduction in pregnancy rate with age. Calving Interval Calving interval is an estimate of the mean period between births for mature females. Typically, it is estimated as the inverse of the annual pregnancy rate (Perrin and Reilly 1984). The principal require- ments for calculating the calving interval are un- biased estimates of gestation time and of the frac- tion of mature females that are pregnant. The standard error in an estimate of calving interval (CI) by these methods is approximated by SE (CI) = (APR-4) var (APR) (Perrin and Reilly 1984). Given our calculated APR estimate of 0.330 for the overall sample, the calving interval is 3.03 yr. The standard error of this estimate is about 0.205. Al- though it is difficult to prove that our estimates of the percent of pregnant females are unbiased, sup- port for such a position is given by Barlow's finding that the percent of pregnant females varies little with sampling conditions (including sampling season, geographic area, dolphin school size, and dolphin kill- per-set) (Barlow 1985). However, if annual variability in the percent of pregnant females is important, binomial sampling theory is likely to underestimate our certainty in estimating the percent of pregnant females, APR, and calving interval. Because no significant trends were detected in the percentage 0.70 r 0.60 0.50 2 0.40 o § 0.30 E O. 0.20 0.10 - Lactating Pregnant <15 16-19 20-23 ^24 ESTIMATED AGE (years) Figure 6— Proportion lactating and proportion pregnant as a function of age for sexually mature female spotted dolphins, in 1973-78 and 1981. Bars represent one standard error from the mean (n = 542). 254 MYRICK ET AL.: REPRODUCTIVE BIOLOGY OF SPOTTED DOLPHINS of pregnant females from 1974 to 1983 (Barlow 1985) and because no significant changes were found in pregnancy rates with age, estimates of calving in- terval were not calculated for any of these possible stratifications. Previous estimates of calving interval for S. at- tenuata include 2.5 yr for the southern offshore ETP stock, 2.7-3.4 yr for the northern offshore ETP stock, and 3.5-3.9 yr for a western North Pacific popula- tion (all values taken from Perrin and Reilly 1984, table 6). Our estimate, 3.06 yr, is thus close to previous estimates for the ETP northern stock and falls between the estimates for two other popula- tions. Lactation Period The calving cycle in mammals can be thought of as a gestation period, a lactation period, and (in some cases) a resting period. Since gestation and lac- tation can overlap, the calving interval can be less than the sum of the gestation and lactation peri- ods. In this study, the duration of the lactation period was estimated as the fraction of mature females that are lactating multiplied by the calving interval in years. Again, the assumption is that all reproduc- tive stages of mature females are sampled without bias. The estimated lactation period for the overall sample is 1.66 yr. Unlike the percent pregnant, the percentage of lac- tating females has apparently increased over the years between 1973-74 and 1981 (Table 1). Collabor- ative evidence is provided by Barlow (1985). Barlow's weighted regression of the percent of lactating females regressed against year predicts values of 46% lactating for 1971 and 69% for 1983. These cor- respond to a change in mean lactation period from 1.4 to 2.1 yr. There were no significant differences in propor- tion of lactating females in different age-classes for all aged samples combined (x| = 2.58, P > 0.25) (Fig. 6). Evidence exists for considerable individual vari- ability in calving interval and lactation period. The sum of the estimated gestation time (0.958 yr) plus the mean lactation period (1.66 yr) is about 2.6 yr; the mean calving interval, estimated as the inverse of APR, is roughly 3 yr. We might pre- dict from this that individuals would never be simultaneously pregnant and lactating. In fact, 16% of the sampled pregnant females were lactat- ing. This is implicit evidence of individual variabil- ity. Postreproductive Females Several criteria have been used to identify post- reproductive female odontocetes. Perrin et al. (1976) described postreproductive spotted dolphins and Per- rin et al. (1977) described postreproductive spinner dolphins, S. longirostris. Both studies were based on the presence of atrophic ("regressed" or "withered") ovaries. In both cases, the incidence of postreproduc- tive females was 1% or less of the sample In pilot whales, Globicephala macrorhynchus, Marsh and Kasuya (1984) found changes in the histology of the ovary, such as a decrease in the volume of the cor- tex and sclerosis of the arterial walls that are age related and associated with senescence Senescent females were characterized on the basis of follicle abundance and the incidence of follicular atresia. Postreproductive females also occurred in our sam- ple Nine of the mature females collected from 1973 to 1982 had atrophic ovaries and thus are considered to have been reproductively senescent. Their mean ovary weights and maximum follicle diameters were significantly different from the means of the other mature females collected during these years (£-test, P < 0.005) (Table 3, Fig. 7). None was lactating. Evidence of decreased fertility was found in some females without atrophic ovaries. Two groups were extracted from the aged sample: 1) those specimens that had 20 or more corpora (all but one was 20 yr old or older), and 2) those specimens that were 20 yr old or older and had only four or fewer total cor- pora (including atretica). Of the first group (n = 12), the mean maximum follicle diameter was larger than that of the atrophic-ovary sample (i-test, P < 0.005), but the mean weights for both ovaries combined were not significantly different (Table 4). Atretic corpora constituted 24% of the total corpora, less than the frequency of atresia found in the atrophic ovaries (39%). The two specimens in this sample with the highest proportion of corpora atretica also had ovaries with maximum follicle diameter and ovary weights within the range of the atrophic ovaries; in addition, they had no CLs (corpora lutea) or Type 1 corpora. We consider these two females to have been postreproductive Of the second group (n = 14), the mean maximum follicle diameter and ovary weight were not different from those in the sample with more total corpora, but were markedly different from those of the atrophic ovaries (£-test, P < 0.025). None of these ovaries contained corpora atretica. Comparison of females in the two groups provides evidence that when the complement of follicles has nearly been expended (through ovulation or atresia), fertility diminishes. Of the first group, 5 of the 12 255 Table 3. — Combined ovary weights, maximum follicle diameter, and corpora counts in "non-atrophic" (normal) ovaries with no corpus lutem (n = 3,455) and atrophic ovaries (n = 9) of sexually mature female spotted dolphins collected in 1973-82. Non-atrophic Atrophic ovaries ovaries Variable Mean SE Mean SE Combined ovary weight 4.9 0.05 3.0 0.30 Maximum follicle diameter 2.8 0.06 0.4 0.07 Total corpora excluding atretica 6.8 0.09 12.4 1.36 Total corpora including atretica 7.5 0.11 20.9 1.13 Corpora atretica 0.7 0.04 8.4 1.67 Percent of corpora atretic 6.4 0.30 40.0 7.6 FISHERY BULLETIN: VOL. 84, NO. 2 lacked macroscopic follicles. Such specimens have in common: 1) the absence of CLs and Type 1 cor- pora, 2) a large number of total corpora, 3) a high frequency of atresia (a relatively large proportion of the total corpora), and 4) a maximum follicle diameter of 0.5 mm or less. The incidence of obvious senescence in the sample of spotted dolphins (0.4%) is much less than that in pilot whale samples studied (5% in Globicephala melaena from the northern Atlantic Ocean [Sergeant 1962] and 25% in G. macrorhynchus from the western Pacific [Marsh and Kasuya 1984]). This may be indicative of inherent dif- ferences in the social structure or longevity between pilot whales and spotted dolphins. Table 4.— Mean age, maximum follicle diameter, ovary weight, corpora counts, and reproductive states for female spotted dolphins. Type 1 and Type 2 corpora defined by Perrin et al. (1976). Maximum Combined follicle ovary Pregnant/ Age diameter weights Total Corpora Percent Type 1 Type 2 lactating years (mm) (g) corpora1 atretica atretic corpora corpora (0/0) A. Females with 20 or more corpora (n = 12) Mean 20.2 2.7 4.2 21.3 4.7 21.9 0.5 1.4 40 SE 1.0 0.5 0.5 0.3 0.8 3.6 0.2 0.4 B. Females 20 yr or older with four or fewer corpora (n = 14) Mean 22.6 2.5 5.1 2.7 0 0 0.9 0.9 100 SE 0.5 0.3 0.6 0.3 0 0 0.2 0.2 includes atretica. specimens were pregnant or lactating. All 14 of the second group were pregnant or lactating. Thus, the first group shows reduced fertility when compared with the second group. Marsh and Kasuya (1984) described an age-related decline in follicle abundance in pilot whales, stating that when follicles are "depleted" the animals become senescent. The reduc- tion in fertility indicated in our sample of spotted dolphins is not strictly age-related; it is more depen- dent on the number of corpora (including corpora atretica) already present in the ovaries. This has been shown to be true in western Pacific spotted dolphins (Kasuya et al. 1974) and in sperm whales (Best 1967). In addition to the postreproductive females with atrophic ovaries, four mature females with normal- appearing ovaries had no macroscopic follicles (one of the atrophic ovaries contained no macroscopic follicles). This is similar to the condition described by Marsh and Kasuya (1984). The ovaries of these specimens weighed from 2.2 to 5.9 g, had no CLs or Type 1 or Type 2 corpora, and contained 8-22 total corpora, 12% of which were atretic None were lac- tating. They are considered to have been postrepro- ductive also. Spotted dolphin specimens were judged to have been senescent when they had atrophic ovaries or CONCLUSIONS Several of our analyses have yielded results similar to those reported previously for spotted dolphins by others, notably Perrin et al. (1976) and Kasuya et al. (1974). We found ovulation rates to have high in- dividual variability with a markedly higher rate of corpus formation in the earlier reproductive years that decreases after a fixed number of ovulations has occurred. The conclusions reached by Perrin (1969b) and par- ticularly by Kasuya et al. (1974) with regard to the close correlation between color pattern and sexual maturity in spotted dolphins are also supported by our study. Ninety-six percent of the fused, 50% of the mottled, and only 4% of the speckled specimens were sexually mature Fused specimens had more corpora and appeared to have been sexually mature longer than mottled specimens of the same age or length. Our estimated length of the calving interval (3.03 yr) is within the range of earlier estimates calculated for this stock by Perrin and Reilly (1984). It is also within the range of estimates for two other spotted dolphin stocks. Some of our analyses, however, produced results 256 MYRICK ET AL.: REPRODUCTIVE BIOLOGY OF SPOTTED DOLPHINS - CN . O • t • • • • t • * * • • • • ft* ft* M • ■ ■ • t t t tt t • ft** • 0*4 • • • **» t IIMf • «* • » * •» t * t * t* • • • t t f • • < • •«•••• ••• • »• * I f • •» • • w t t*t*t *»•• mm « iiwii ■ * •••••*!•» •##•«• #* » *•* ♦• » • • •■* •♦ •» • * • —m • •*• •*•• •*• • •* **♦ *** • • — IM — III «M ■ I ■ MMH t # • 1 »«M* W— MM H ft t « t Mtta* • »•••» ttvt t**t t ft • t * t ft* — «»»« 4 Wt> »•*• » • t M*tt *•• • m * • t tmtmm • • • * t t I « • ft t ■ t f \ — ■ — I — ' — 1 — I CO g HI 00 £ < > O ■ CD ... c\J o co co ^r CM t- r- t- T ' I ' 1— ■ 1 ' I ■ I ■ 1 ' I ■ 1— ■ 1 ■ I ■ 1- ■" CNJ O 00 CD Tf CVJ O H3±3wvia 3iomoj wnwixvw a. ~o a 03 3 SO u .o ho JS .ho '33 -a c o >- 3) s 3 a) Oh I w OS D o 257 FISHERY BULLETIN: VOL. 84, NO. 2 that contradict earlier findings. Based on the more reliable method of estimating age in spotted dol- phins, we believe that our findings present a clearer picture of the reproductive information than has been reported previously. Our aged samples showed that the youngest sexually mature female was 10 yr old— the same age as the youngest pregnant and the youngest lactating specimens. This suggests that some females must become sexually mature before the age of 10, even though mature specimens younger than 10 were not found in our sample The average age of a pregnant female in our sample was about 18 yr, and some females of about 35 yr old were pregnant or nursing. These values are substan- tially higher than estimated previously for this stock (Perrin et al. 1976), but they are similar to, though still somewhat higher than, estimates for the western Pacific stock (Kasuya 1976). The ASM estimate in this study (about 11.4 yr) is higher than that estimated by Perrin et al. (1976). Our calculations showed no significant difference between the ASM calculated for the 1973-74 sam- ple (taken during years of heavy fishing mortality) and the ASM for the 1981 sample (taken after at least 5 yr of reduced fishing mortality). An ASM of 11.4 yr means that the youngest average age of first parturition would be 12.3 yr (11 mo later). Since not all females would conceive at first ovulation, the actual average age would be greater than this. The implication of this protracted period before reproduction and a long (3.03 yr) calving interval is that spotted dolphin survival rates must be very high in order to maintain a stable population level. There is a significant depression in the age struc- ture of the 1973-78 and 1981 aged sample in the 6-12 yr age classes (Hohn and Myrick in prep.2). Similar age-structure patterns, interpreted as reflecting some sort of schooling segregation, have been en- countered in studies of other delphinids (see review by Perrin and Reilly 1984). If animals at or near the age of sexual maturity have been regularly under- sampled because their schools were not targets of purse seines (Hohn and Scott 1983), the ASMs calculated for the aged samples could be upwardly biased. However, there is no evidence that the depression in the age structure represents missing animals that were sexually mature The annual pregnancy rate averaged 0.33 from 1973 through 1981. There were no sustained upward or downward changes in age-specific pregnancy rates with increased age A similar result was shown by Kasuya (1976) for the western stock, although his values were somewhat lower than the rates we have estimated for the northern offshore stock. Our esti- mates are different from those of Perrin et al. (1976) who reported high pregnancy rates among younger specimens and a decreasing rate with increased age The implications of an apparent progressive in- crease in the lactation period are enigmatic It is probable that the increase in lactation period reflects the decrease in per capita mortality of calves due to the more efficient releasing procedures employed by the purse seine fleet from the mid-1970's onwards. Decreased mortality of nursing calves would be reflected by an apparent increase in the number of lactating females because fewer nursing periods were ended prematurely. Our study of postreproductive specimens suggests that fertility diminishes as the complement of follicles for a female becomes expended through ovulation or atresia. Female spotted dolphins with atrophic ovaries or with no macroscopic follicles are reproductively senescent. Although the expenditure of follicles progresses with age, reduction in fertility is not strictly age related. The occurrence of repro- ductive senescence in spotted dolphins in this study was negligible and the number of specimens in this state probably is of limited importance to estimates of reproductive parameters. ACKNOWLEDGMENTS We thank D. DeMaster, W. F. Perrin, and S. Reilly for their helpful comments and recommendations on early drafts of the manuscript. We are grateful to J. Bengtson, D. Chapman, F Hester, J. Mead, A. York, and R. Wells for their very thorough reviews. J. Walker and S. Chivers assisted in organizing and accessing the life history data and S. Chivers helped with the analyses. D. Stanley and M. Kimura prepared the tooth sections for the aged subsamples. Special thanks go to H. Orr who prepared the figures and to H. Becker and S. Richardson and the SWFC Technical Support Staff who typed parts of the manuscript. D. DeMaster, N. Lo, and S. Reilly assisted in statistical testing of some of the samples. J. Michalski edited the final draft. 2Hohn, A. A., and A. C. Myrick, Jr. The age structure of north- ern offshore dolphins, Stenella attenuata, from the eastern tropical Pacific Manuscr. in prep. Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. LITERATURE CITED Barlow, J. 1985. Variability, trends, and biases in reproductive rates of 258 MYRICK ET AL.: REPRODUCTIVE BIOLOGY OF SPOTTED DOLPHINS spotted dolphins, Stenella attenuata. Fish. Bull., U.S. 83:657- 669. Best, P. B. 1967. The sperm whale (Physeter catodon) off the west coast of South Africa. 1. Ovarian changes and their significance S. Afr. Div. Sea Fish. Invest. Rep. 61:1-27. Brodie, P. F 1971. A reconsideration of aspects of growth, reproduction, and behavior of the white whale (Delphinapterus leucas) with reference to the Cumberland Sound, Baffin Island, popula- tion. J. Fish. Res. Board Can. 28:1309-1318. DeMaster, D. P. 1978. Calculation of the average age of sexual maturity in marine mammals. J. Fish. Res. Board Can. 35:912- 915. 1984. Review of techniques used to estimate the average age at attainment of sexual maturity in marine mammals. In W. F. Perrin, R. L. Brownell, Jr., and D. P. DeMaster (editors), Reproduction in whales, dolphins and porpoises, p. 175-179. Rep. Int. Whaling Comm. (Spec Issue No. 6). Hammond, P. S., and K. T. Tsai. 1983. Dolphin mortality incidental to purse-seining for tunas in the eastern Pacific Ocean, 1979-1981. Rep. Int. Whaling Comm. 33:589-597. Henderson, J. R., W. F Perrin, and R. B. Miller. 1980. Rate of gross annual production in dolphin populations (Stenella spp. and Delphinus delphis) in the eastern tropical Pacific 1973-78. SWFC, NMFS, NOAA, Admin. Rep. LJ-80-02, 51 p. Hohn, A. A., J. Barlow, and S. J. Chivers. 1985. Reproductive maturity and seasonality in male spotted dolphins, Stenella attenuata, in the eastern tropical Pacific Mar. Mammal Sci. l(4):273-293. Hohn, A. A., and M. D. Scott. 1983. Segregation by age in schools of spotted dolphin in the eastern tropical Pacific [Abstr.] Fifth Bienn. Biol. Mar. Mammals, p. 47. Kasuya, T. 1976. Reconsideration of life history parameters of the spotted and striped dolphins based on cemental layers. Sci. Rep. Whales Res. Inst. (Tokyo) 28:73-106. Kasuya, T, N. Miyazaki, and W. H. Dawbin. 1974. Growth and reproduction of Stenella attenuata in the Pacific coast of Japan. Sci. Rep. Whales Res. Inst. (Tokyo) 26:157-226. Marsh, H., and T. Kasuya. 1984. Changes in the ovaries of the short-finned pilot whale, Globicephala macrorhynchus, with age and reproductive ac- tivity. In W F. Perrin, R. L. Brownell, Jr., and D. P. DeMaster (editors), Reproduction in whales, dolphins and porpoises, p. 311-335. Rep. Int. Whaling Comm. (Spec Issue No. 6). Myrick, A. C, Jr., A. A. Hohn, P. A. Sloan, M. Kimura, and D. Stanley. 1983. Estimating age of spotted and spinner dolphins (Sten- ella attenuata and Stenella longirostris) from teeth. U.S. Dep. Commer., NOAA Tech. Memo. NMFS SWFC-30, 17 p. Myrick, A. C, Jr., E. W Shallenberger, I. Kang, and D. B. MacKay. 1984. Calibration of dental layers in seven captive Hawaiian spinner dolphins, Stenella longirostris, based on tetracycline labeling. Fish. Bull, U.S. 82:207-225. Perrin, W F. 1969. Using porpoise to catch tuna. World Fishing 18(6): 42-45. 1970a. Color pattern of the eastern Pacific spotted porpoise Stenella graffmani; Lonnberg (Cetacea, Delphinidae). Zool- ogica (NY) 54:135-149. 1970b. The problem of porpoise mortality in the U.S. tropical tuna fishery In Proceeding of the 6th Annual Conference on Biological Sonar and Diving Mammals, p. 45-48. Stan- ford Res. Inst., Menlo Park, CA. Perrin, W F, J. M. Coe, and J. R. Zweifel. 1976. Growth and reproduction of the spotted porpoise, Stenella attenuata, in the offshore eastern tropical Pacific Fish. Bull, U.S. 74:229-269. Perrin, W F, D. B. Holts, and R. B. Miller. 1977. Growth and reproduction of the eastern spinner dolphin, a geographical form of Stenella longirostris in the eastern tropical Pacific Fish. Bull., U.S. 75:725-750. Perrin, W F, and A. C. Myrick, Jr. (editors) 1980[1981]. Age determination of toothed whales and siren- ians. Rep. Int. Whaling Comm. (Spec Issue No. 3), 229 p. Perrin, W F, and S. B. Reilly. 1984. Reproductive parameters of dolphins and small whales of the family delphinidae In W F. Perrin, R. L. Brownell, Jr., and D. P. DeMaster (editors), Reproduction in whales, dolphins and porpoises, p. 97-133. Rep. Int. Whaling Comm. (Spec Issue No. 6). Reilly, S. B, A. A. Hohn, and A. C. Myrick, Jr. 1983. Precision of age determination of northern offshore spotted dolphins. U.S. Dep. Commer., NOAA Tech. Memo. NMFS SWFC-35, 27 p. Sergeant, D. E. 1962. The biology of the pilot or pothead whale Globicephala melaena (Traill) in Newfoundland waters. Fish. Res. Board Can. Bull. 132:1-84. Smith, T. D. 1983. Changes in size of three dolphin (Stenella spp.) popula- tions in the eastern tropical Pacific Fish. Bull, U.S. 81:1-13. York, A. E. 1983. Average age at first reproduction of the northern fur seal (Callorhinus ursinus). Can. J. Fish. Aquat. Sci. 40: 121-127. 259 CHINOOK SALMON, ONCORHYNCHUS TSHAWYTSCHA, SPAWNING ESCAPEMENT BASED ON MULTIPLE MARK-RECAPTURE OF CARCASSES Stephen D. Sykes and Louis W. Botsford1 ABSTRACT Mark-recapture data from a population of chinook salmon, Oncorhynchus tshawytscha, carcasses were collected for escapement estimates in a northern California stream. Escapement was taken to be im- migration into the population of carcasses. Results from three methods of estimating total immigration into this population— Jolly-Seber, Manly and Parr, and Jolly-Seber with a modified data set— were com- pared to a weir count. Sources of violations of modeling assumptions, age-dependent catchability, and survival were identified, but the estimates appeared to be relatively insensitive to these. The effect of lower sampling intensity, which exacerbates effects of age-dependent catchability, was evaluated through simulation. The third method appears to be the best of the three because 1) it requires the least sampling effort, 2) it is the most robust with respect to violations of the assumption of equal catchability, and 3) it enables reanalysis of previously collected data. Standard errors and 95% confidence intervals of estimates obtained by the third method were computed by simulation. Since the distribution of estimates is asymmetrical, these confidence limits are preferred over standard expressions. Pacific salmon fisheries are currently managed by attempting to allow a specified number of fish to escape the fishery, migrate upstream and spawn. Proper management therefore requires accurate estimates of this escapement. Since Pacific salmon die immediately after spawning, escapement can be estimated from the number of carcasses that accu- mulate during a spawning season. The California Department of Fish and Game (CDF&G) estimates escapement of chinook salmon, Oncorhynchus tshawytscha, each year using the methods of Schaef- fer (Schaeffer 1951; Darroch 1961) and Peterson (Seber 1982) to analyze mark-recapture data from surveys of accumulated carcasses. Since the fish enter the stream to spawn during the sampling periods, the assumption of a closed population re- quired by the Peterson estimate does not hold. The Schaeffer method is designed to estimate numbers from a stratified two sample experiment in which fish are tagged at different locations (or different times at one location as fish migrate upstream) and are sampled at the same locations (or an upstream point) at a later time. CDF&G carcass surveys, on the other hand, involve sampling the same unstrati- fied stretch of spawning stream several times. The results described here are part of an attempt to develop an accurate, efficient, and robust procedure for estimating escapement from carcass data. A department of Wildlife and Fisheries Biology, University of California, Davis, CA 95616. Manuscript accepted July 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. technique that allows not just estimates for current and future years, but also could be used to analyze mark-recapture data taken by CDF&G in past years was desired. Parker (1968) and Stauffer (1970) used standard Jolly-Seber methods to estimate spawning run sizes from mark-recapture data obtained from carcass counts. However, they did not examine departures from modeling assumptions by collecting appropri- ate data in the field or statistically testing assump- tions. Also, an independent count of the population size was unavailable, hence actual errors in their estimates could not be computed. In addition, car- casses were carefully replaced where they had been found after sampling and tagging, hence captured carcasses would have a high probability of being recaptured. Thus, their results were probably biased because of heterogeneous capture probabilities. To develop the estimation technique a mark-recap- ture experiment was performed in the Bogus Creek spawning area of the Upper Klamath River drainage during the 1981 chinook salmon spawning run. As a check on the estimates, a counting weir was placed at the mouth of Bogus Creek. Salmon were counted while they were in the weir trap, and were sub- sequently released upstream. This mark-recapture study differed from the usual mark-recapture studies of fish and wildlife populations in that the population was composed of carcasses (i.e., in- dividuals enter the population by dying and leave by predation and decay). Thus, the age of a carcass, 261 FISHERY BULLETIN: VOL. 84, NO. 2 as used here, refers to time since death rather than time since recruitment. The procedures followed here differed from previous CDF&G surveys in that more data were taken than were actually needed for the estimate so that departures from model assumptions could be examined. The additional data enabled simula- tion of the sampling procedure to estimate bias and variances, and allowed us to determine the sources of failure of assumptions. We were also able to develop estimates from which some sources of bias had been removed. METHODS The study was conducted on a chinook salmon spawning area of a small northern California stream, Bogus Creek (Fig. 1). The stream was sampled over a 6.5-mi reach from a counting weir upstream to Bogus School road. Sampling was begun on 15 September 1981, at the very beginning of the spawning run, and discontinued on 12 Novem- ber 1981, by which time very little spawning activ- ity was apparent. The stream was sampled weekly during that period; sampling took 2 d during the peak of the run, with one half of the stream being sampled per day. The stream was sampled by two people walking upstream and capturing with a gaff any carcasses seen. Data on each capture were described as follows: Place of capture: Edge top, edge bottom, middle top, middle bottom, snagged, dry or buried. Size: Small (<65 cm), medium minus (65-69 cm), Klamath River Figure 1— Study area in north- ern California. 262 SYKES and BOTSFORD: CHINOOK SALMON SPAWNING ESCAPEMENT medium (70-80 cm), medium plus (81-85 cm), or large (>85 cm). Sex: Male or female. Condition: Alive, fresh (eyes clear), decayed minus (eyes cloudy, flesh firm), decayed (flesh soft), decayed plus (flesh very soft), or skeleton (flesh falling off). Carcasses were individually tagged with fingerling fish tags which were attached around the maxillary bone. Data on place of release for each released car- cass were recorded as follows: Pool, pool/riffle, or riffle. The presence or absence of obstructions which would trap and remove a carcass. The speed of water flow. Thus movements of individual carcasses and their condition, both of which might affect catchability and survival, could be examined on an individual basis. During the sampling process about one-third of the unmarked, captured carcasses was random- ly removed from the population by cutting the fish in two. This was done because of limited time available for recording data. These individuals were considered "trap mortalities" (i.e., they are counted in the sample size but not in the total releases for that time period). Because the mark-recapture methods used allow for capture loss, removal of these fish has no effect on errors other than lower- ing sample sizes. Two existing methods, those of Jolly and Seber (Seber 1982) and Manly and Parr (1968), and a third, a modified Jolly-Seber method, were used to es- timate population sizes, recruitment, survival, and their standard errors (when expressions were available). The corrected estimates of Seber (1982) were used for the Jolly-Seber method. When sur- vival was estimated as greater than unity, or immi- gration as <0.0, those values were replaced with 1.0 and 0.0 respectively in subsequent calculations. In the third method, standard Jolly-Seber estimates were calculated after modifying the mark-recapture data so that all decayed (decayed minus or worse) carcasses (marked and unmarked) were assumed to have been destroyed upon capture. This method simulates the way CDF&G has traditionally col- lected data. After these estimates had been calculated, the estimated escapement, E, was calculated as the number present at the first sample period, plus the number of individuals immigrating during each subsequent period. E = nx + 02 - R, * *,)/(*!) b) + D2 + D3 + D4 (1) where nx = the number sampled at the first sam- ple time, Rx = the number tagged and released at the first sample time, N2 = the estimated population size at sam- ple time two, A = Mn5, i = the survival rate from i to i + 1, and B{ = the estimated number of carcasses still present at the sample time i + 1 which immigrated between i and i + 1. In this expression the initial number present at time period 1 is conservatively taken to be the sample size at time period 1 {n{). The number immigrating during the subsequent period is taken to be the estimated population at time period 2, minus the number of tagged fish which had been accounted for in the first sample. Immigration during the next two periods are standard estimates. Each immigration rate is divided by the square root of the survival rate (the survival rate for half the sample period), to ac- count for fish that enter the population and leave it between sampling periods, and thus are never sampled (Stauffer 1970). Estimates of immigration during the last time period are not computed in standard multiple mark- recapture experiments; however, this immigration (.B4 here) can be estimated from the standard Jolly- Seber expression (Seber 1982), if the final numbers (Nb) and survival rate (4) can be estimated. If sur- vival varies little from sample to sample, 4 can be estimated by assuming that mortality is equal to the value estimated over an earlier period in this study. Since survival varied little between sampling periods and the x2 test of Seber (1982) failed to reject the null hypothesis of constant survival (x2 = 0.4648, df = 2), we estimated survival from period 3 to period 4 as the average of 4>1; 4>2> ana" ^3- To esti- mate iV5, we estimated the capture probability at sample period 5 (P5) as the ratio of the number of carcasses released at sample 4 and recaptured at sample 5 (r4) to the number released at sample 4 (R4) times survival to sample 5 (4>4), P5 = rJ(R, * 4>4). (2) We then estimated the population size at sample 5 263 FISHERY BULLETIN: VOL. 84, NO. 2 (N5) as the sample size (n5) divided by the capture probability (P5). Standard errors and 95% confidence limits for the third method were obtained by simulation. The sam- pling process was simulated by generating a popula- tion of carcasses based on population size estimates from the third method. We then sampled the popula- tion by comparing a uniformly distributed random number with the appropriate probability of capture [see Sykes (1982) for a more detailed description of the simulation process, and a Fortran program]. From each simulation we calculated Jolly-Seber estimates of survival, immigration, population sizes, and their standard errors. An estimate of E was then calculated as above. This simulation process was repeated 1,000 times. In addition to calculating the average and standard error of each of these estimates, 95% confidence limits were calculated by Buckland's (1980) method 1. To obtain 95% con- fidence limits by this method, one adds the dif- ference between the average of the 25th and 26th lowest estimates (out of 1,000) and the average value to the field estimate to obtain the upper bound and subtracts the difference between the average of the 25th and 26th highest estimates and the average value to obtain the lower bound. All three methods assume that all individuals are equally catchable. The methods based on the Jolly- Seber model also assume that all individuals have equal probabilities of survival. Since violation of these assumptions could result in biased estimates, we determined whether catchability and survival varied and the effects of these on the estimates. Several statistical tests can be used to check for differential catchability and mortality, but only among animals that are already marked. Two x2 tests, which compare expected frequencies of cap- ture histories with actual frequencies (Seber 1982; Jolly 1982) were calculated from the unmodified field data. The test of Leslie and Carothers (Carothers 1971) was not performed because of the small number of sampling periods. Since both tests yielded expected values less than unity, pooled x2 values were also calculated, using a conservative df value of df = (number of pools - 1). For Seber's test, all values less than unity were pooled; for Jolly's, each value less than unity was pooled with the next highest value. Following Leslie et al. (1953, cited by Seber 1982) we tested for homogeneity of catchability and sur- vival by comparing estimates of population param- eters obtained by different methods. These methods differ in sensitivity to survival and capture heter- ogeneity, hence the presence of heterogeneity should cause differences in estimates of the same parameter by the different methods. We tested the unmodified field data by calculating the following parameter estimates as per Leslie et al. (1953): v{: the estimated number of new marks re- leased at time i 4> . {: the estimated survival for the subpopulation of marked carcasses, and N.z: the number of marked carcasses. and compared them with, respectively, v{: the actual number of new marks released at time i fy: the Jolly-Seber estimate of survival, and M^ the Jolly-Seber estimate of the number of marked carcasses. If differential catchability or survival, when present, results in significant bias, these estimates will be different. Since only marked (and thus decayed) carcasses are considered in the statistical tests discussed thus far, these tests do not address the potential for age- dependent catchability. To evaluate possible effects of age-dependent catchabilities we "corrected" the sample size n{ by reducing it to account for the fact that fewer fresh (shiny, silver colored) carcasses would have been captured if they had not been more visible than decayed (dull brown colored) carcasses. We then recalculated the escapement estimates using the corrected sample size. We used two ratios of average fresh to decayed catchability: 2.0 and 1.4. Since visibility only differed among carcasses on the stream bed, and only 30% of the captures were on the stream bed, these values represented actual ratios for carcasses on the stream bed of approx- imately 6.7 and 4.7, respectively. To evaluate the potential advantage of increasing the efficiency of the third method by lowering the sampling effort we examined the effect of lowered sampling intensity on behavior of the three estimators. Lower effort would most likely result in less searching on the bottom of the stream for carcasses. We therefore simulated lowered sampling by generating new capture histories for each in- dividual according to the following set of rules: 1) If an individual was buried at a capture event, that and all subsequent captures were ignored, 2) cap- tures of decayed carcasses on the stream bed and surface were ignored according to comparison of a uniform random number with the appropriate decrease in capture probability, and 3) the next cap- 264 SYKES and BOTSFORD: CHINOOK SALMON SPAWNING ESCAPEMENT ture of an individual whose previous bottom capture was ignored was considered to be a bottom capture, as movement was probably the result of the previous capture event. RESULTS Total escapement estimates for the three methods and the weir count of fish moving into the spawn- ing area are presented in Table 1. All three methods result in escapement estimates that are close to the weir count. The third method is the most efficient Table 1.— Estimates of total escapement and the estimates used to compute them for each of the three methods. Jolly-Seber Manly and Parr Method 3 N2 999 1,076 1,063 SE A/2 95 128 139 N3 2,302 2,312 1,886 SEA/3 166 184 161 W4 . 1,845 1,853 1,452 SEA/4 67 72 93 B2 1,801 1,740 1,459 SES2 174 (1) 183 S3 150 136 371 SE03 128 (1) 179 02 0.7617 0.7789 0.7297 SE 4>2 0.353 (1) 0.439 *>3 0.7878 0.7940 0.8578 SE*3 0.0305 (1) 0.0548 "1 87 87 87 [A/2 - flt ,]/<*>Ub 1,042 1,139 1,142 D2 2,062 1,970 1,708 03 169 151 401 D4 84 91 170 E 3,445 3,438 3,508 Weir count: 3,642 'Estimate of these standard errors are not available. in the sense that it requires the least sampling effort. For the third method, Jolly-Seber estimates and associated estimated standard errors, computed from the survey data along with the average value, standard errors, and 95% confidence limits obtained from simulation, are presented in Table 2. Esti- mated standard errors and simulated standard errors are in close agreement, except that the distri- bution of estimates around the mean value is clear- ly asymmetrical. Since they are based on simulation of the actual process rather than approximate analytical expressions, confidence limits obtained from simulation are presumably more realistic than those estimated by the methods of Jolly and Seber. The sum of the estimated escapement by time i + 2 is plotted with the sum of the weir count at time i in Figure 2. Since the numbers of fish which migrated through the weir correlates well with the estimated number of fish that died 2 wk later, most salmon probably spawned and died within 2 wk of having entered the stream. Since the estimate of immigration during the last sampling interval seems to fit the known number of fish immigrating, the assumption of constant survival seems to be a good one. It is clear that our criteria for stopping sam- pling when most spawning activity had ceased resulted in an estimate of the complete run. Sam- pling for another week would have removed the need to make any assumptions in estimating B4, but since this value will always be small in relation to the total escapement, the increase in accuracy does not seem worth the additional effort. Data regarding the condition of carcasses at the time of capture reflect a declining trend in catch- Table 2.— Estimates of escapement (E), population size (W), immigration (S), survival (O), and associated standard errors obtained from a Jolly-Seber analysis of data for method three. Also shown are the computed mean, standard error, and 95% con- fidence intervals obtained by simulation. Simulation value Upper Lower Field estimate Mean SE 95% C.I. 95% C.I. A?2 1,063 1,041 145 + 222 -344 SE N2 139 138 43 + 66 -100 N3 . 1,886 1,889 166 + 289 -360 SEA/3 161 165 28 + 46 -62 N4 1,452 1,458 94 + 167 -199 SEA/4 93 94 19 + 33 -43 *2 0.7297 0.7327 0.0459 + 0.0892 -0.0929 SE M; N 1 84 — — 0.7995 — 67 — 2 311 — — 0.7617 — 288 319 3 724 680 44 0.7878 0.7969 797 796 4 741 756 214 — — 1,201 1,234 5 — — — — — — — far too low, indicates that this bias was probably not present in our sampling process. Thus biases en- countered here are insignificant, both in relation to possible imprecision in estimating the percent run and area covererd, and the estimated standard errors. Estimates computed to evaluate the effects of lowering sampling intensity are shown in Table 7. Simulations are listed according to the percent of top and the percent of bottom captures ignored for that simulation. The estimates obtained by the third 267 FISHERY BULLETIN: VOL. 84, NO. 2 Table 6.— Escapement estimates obtained by correcting for differential catchability of fresh and decayed carcasses for three methods of estimating escapement. For each correction, the ratio of the average fresh to decayed catchabilities that was assumed to obtain the corrected estimate is given. Assumed fresh/decayed Catchabilities Corrected escapement Jolly-Seber Manly and Parr Method 3 Original estimate 3,445 3,438 3,508 1.4/1.0 3,446 3,471 3,274 2.0/1.0 3,321 3,319 3,262 Table 7.— Escapement estimates obtained by simulation of reduced sampling effort for three methods of estimating escapement. For each simulation the fraction of decayed top carcass captures and the fraction of decayed bottom carcass captures ignored is given. Fraction of decayed Carcass captures ignored Escapement estimate Top Bottom Jolly-Seber Manly and Parr Method 3 Original estimate 3,445 3,438 3,508 0.0 0.4 3,740 3,765 3,676 0.0 1.0 3,944 4,058 3,777 0.2 0.4 3,890 3,917 3,977 0.4 0.6 4,844 4,934 4,364 method are less biased than those obtained by the other two methods. DISCUSSION The estimates of total immigration are all remark- ably close to the weir count. This accuracy is even more remarkable in light of the fact that CDF&G has traditionally used a correction factor of 0.95 to account for an estimated 5% of the spawning grounds that is not sampled on Bogus Creek. Inclu- sion of this factor brings all of the estimates to within 1.4% of the weir count. Since the third method provides a high degree of precision (Table 2) at much less sampling cost, it is preferable over the other two methods. We can compare the preci- sion of the third method with the Jolly-Seber and Manly and Parr methods by comparing the standard error estimates that are available for those two methods (Table 1). The Jolly-Seber method is more precise in estimates of AT, B, and . This is expected, since both the Manly and Parr method and the third method use fewer individuals in estimates than the Jolly-Seber method does. However, the precision of the third method is more than adequate: 95% con- fidence intervals are +5.3% and -5.5% of the escapement estimate. The detected violations of assumptions, age- dependent catchability and heterogeneity of capture probabilities and survival, are those that would be expected on the basis of physical considerations. Survival of carcasses is a function of two processes: fresh carcasses being removed by carnivores, and old carcasses decaying and becoming buried in the stream bed. Rates of disappearance could thus be affected by condition, and therefore age and size, of carcasses. Older carcasses and smaller carcasses, which decay more quickly and are buried more easily than larger carcasses, would be expected to have lower survival rates. Catchability is a function of both visibility and loca- tion, both of which would be expected to vary with condition and size of carcasses. This causes two dif- ferent types of problems: age-dependent catchability and size-dependent catchability. Shiny, fresh car- casses were much more visible on the bottom of the stream than the brown, decayed carcasses. Car- casses on the stream surface were in general visi- ble regardless of their condition. Since carcasses lost their high visibility in about a week, no marked car- casses will be in this high visibility category, and un- marked carcasses will on the average be more catch- able than marked carcasses. This can be thought of as age-dependent catchability. Size-dependent catch- ability stems from the fact that decayed individuals that were large were more visible than those that were small. This can be viewed as capture heter- ogeneity. Since fresh fish were high visible regard- less of their size, this heterogeneity existed only among decayed individuals. Based on these con- siderations we would expect catchability to vary with age and size according to Figure 4. 268 SYKES and BOTSFORD: CHINOOK SALMON SPAWNING ESCAPEMENT While both Jolly's (1982) and Seber's (1982) tests indicate differential catchability and/or mortality are present, the issue of real importance is the amount of any resulting bias. Manly (1970) concluded that if age-specific mortality is present in a sampled population, Manly and Parr (1968) estimates should fare better than those of Jolly and Seber (Seber 1982). Both methods, however, are biased for the case in which mortality increases with age; in fact, Manly's (1970) estimates of bias for additions (B) are greater for Manly and Parr estimates than for Jolly- Seber estimates for those simulations with param- eters closest to our population. Survival, population size, and catchability estimates were negatively bi- ased by only 1 or 2%. Seber (1982) pointed out that Jolly-Seber estimates should be relatively unbiased even with differential mortality if mark status and mortality were not correlated. Both estimators, then, should have relatively unbiased estimates of survival and catchability for "marked" animals. A positive bias in estimates of immigration, B, (and consequently in E) would arise primarily from apply- ing mortality of marked animals to the entire population, when marked animals are in general older, and thus have lower survival than unmarked animals. The age-dependent catchability detected in this study would be expected to result in a positive bias in the estimate of total escapement, E. Because each capture sample includes fresh, recently immigrated individuals, and recapture samples include older, decayed individuals, we expect N to be overesti- mated (i.e., nIN > m/M in Jolly-Seber and pN < n in Manly and Parr), which results in estimates of B and E being positively biased also. Since bias from age-dependent catchability in N decreases as M ap- frash decayed Age Figure 4.— Expected changes in capture prob- abilities with age at different sizes. proaches N, and removing carcasses after capture in the third method decreases the ratio of marked to total carcasses, we would expect the third esti- mator to be more biased by age-dependent catch- ability problems than the first two methods. However, the simulations of lower sampling in- tensity, which would exacerbate the effects of age- dependent catchability, show that the estimate obtained by the third method is more robust with regard to lowered sampling intensity. This unex- pected result is probably due to compensating effects which decrease bias in E. The two most im- portant components of E are the second ((N2 - Ri Oi)/$i'5) and third (D2). In the standard estimates these values both increase with increases in the number of captures ignored. In the third method, however, the second component increases, but the third decreases. This is because as catchability declines, fewer marks are captured and "removed", hence more carcasses are available for later capture. This is not the case in the first two methods because marked carcasses are not removed at capture. Since in the third method the composition of M and N is relatively unchanged at the second sample period, but at the third sample period, M increases relative to N (because of the increase in the number of decayed marks present), the estimate of population size at the third sample period will be less biased than the estimate for the second sample period. This results in a negative bias in the estimated immigra- tion from time period two to three. This compensa- tion makes the third method more robust with respect to age-dependent catchability problems than the other two methods. Bias in the estimates is not severe until large numbers of capture events are ig- nored (Table 7). While all three methods produce accurate estimates, even when lowered sampling exacerbates differential catchability problems, the magnitude of the bias relative to standard errors can be substantial. For this reason, samples must be carefully taken if estimates from different streams or different years (which will have different biases because of different conditions) are to be compared statistically. Heterogeneity of capture probabilities affects Jolly-Seber and Manly and Parr estimates in the same manner. Since in the Jolly-Seber method the individuals marked and released at sample i, Rit are on the average younger than the individuals marked and released prior to sample i, M{ is a low estimate (i.e., rlR > zl(M - m), or M > (Rzlr) + m). This decreases the positive bias in N which is caused by age-dependent catchability. Since bias in M in- creases as more individuals are marked, we expect 269 estimates of M from the third method to be less bi- ased than those from the first two. Usually, capture heterogeneity leads to the more catchable animals joining the marked population, and we expect marked animals to be more catchable than unmarked animals. Capture heterogeneity, however, is only prevalent among decayed in- dividuals who are all less catchable than fresh, un- marked individuals. Thus, capture heterogeneity, by placing the more catchable decayed individuals in the marked population, results in the capture prob- ability of marked animals being closer to the cap- ture probability of unmarked animals. This reduces the negative bias in population size (N), immigra- tion (B), and escapement (E) estimates, which was caused by age-dependent catchability. Again, the third method, by removing decayed individuals and decreasing the fraction of the population which is decayed, will not be affected by capture heteroge- neity as strongly as the other two methods. Manly and Parr estimators will have the same ameliorating affects because of capture heteroge- neity as their Jolly- Seber counterparts. Since the estimate of catchability, p, should be accurate for the more catchable animals, estimated survival should be accurate for that group. Bias would result from correlations between catchability and survival. Also, since p is estimated for marked (and thus decayed) individuals, using the more catchable decayed individuals to estimate p brings the estimated catchability closer to the actual catch- ability of the unmarked individuals. Again, this reduces the bias in N, B, and E which is caused by age-dependent catchability. There are other approaches to estimating param- eters from populations with age-dependent survival and capture rates. By placing carcasses in two readi- ly identifiable age classes, fresh (and thus <1 wk old) or decayed (and thus older than 1 wk), Pollock's (1981) modified Jolly-Seber analysis of the data could have been made. Since this method requires recap- tures of decayed individuals, it could not be used to analyze data from previous surveys, and it would require more sampling effort in future surveys than the method 3 estimate. If different age classes have sufficiently different capture or survival rates, then this method will provide more accurate estimates. If not, then it will yield the same estimate as the third method, but would have higher variances, as more parameters are estimated. FISHERY BULLETIN: VOL. 84, NO. 2 ACKNOWLEDGMENTS We would like to thank L. B. Boydston of the California Department of Fish and Game for making us aware of this problem, for many helpful discus- sions, and for assisting in data collection. S. Sykes was supported by the California Department of Fish and Game during the sampling. We are grateful for the comments by K. Pollock, T. Schoener, and N. Matloff on an earlier version of this manuscript. We would also like to thank Ivan Paulsen for assistance in data collection. LITERATURE CITED BUCKLAND, S. T. 1980. A modified analysis of the Jolly-Seber capture-recapture model. Biometrics 36:419-435. Carothers, A. D. 1971. An examination and extension of Leslie's test of equal catchability. Biometrics 27:615-630. Darroch, J. N. 196 1 . The two-sample capture-recapture census when tagging and sampling are stratified. Biometrika 48:241-260. Jolly, G. M. 1982. Mark-recapture models with parameters constant in time Biometrics 38:301-321. Manly, B. F. J. 1970. A simulation study of animal population estimation using the capture-recapture method. J. Appl. Ecol. 7:13-39. Manly, B. F. J., and M. J. Parr. 1968. A new method of estimating population size, survivor- ship, and birth rate from capture-recapture data Trans. Soc Br. Entomol. 18:81-89. Parker, R. R. 1968. Marine mortality schedules of pink salmon of the Bella Coola River, central British Columbia. Can. J. Fish. Res. Board 25:757-794. Pollock, K. H. 1981. Capture-recapture models allowing for age-dependent survival and capture rates. Biometrics 37:521-529. Schaefer, M. B. 1951. Estimation of the size of animal populations by mark- ing experiments. U.S. Fish Wildl. Serv., Fish. Bull. 52: 191-203. Seber, G. A. F. 1982. The estimation of animal abundance and related param- eters. MacMillan Publishing Co., Inc, N.Y. 654 p. Stauffer, G. 1970. Estimates of population parameters of the 1965 and 1966 adult Chinook salmon runs in the Green-Duwamish River. M.S. Thesis, Univ. Washington, Seattle, 155 p. Sykes, S. D. 1982. Multiple mark-recapture estimators of salmon spawn- ing runs sizes. M.S. Thesis, Univ. California, Davis, 64 p. 270 THE DISTRIBUTION OF THE HUMPBACK WHALE, MEGAPTERA NOVAEANGLIAE, ON GEORGES BANK AND IN THE GULF OF MAINE IN RELATION TO DENSITIES OF THE SAND EEL, AMMODYTES AMERICANUS P. Michael Payne,1 John R. Nicolas,2 Loretta O'Brien,2 and Kevin D. Powers1 ABSTRACT The distribution of the humpback whale, Megaptera novaeangliae, (based on shipboard sighting data) is significantly correlated (r = 0.81, df = 13) with the number of sand eel, Ammodytes americanus, per standardized tow (based on NMFS/NEFC groundfish surveys) by strata within the Gulf of Maine A demonstrated increase in the number of humpback whale sightings in the southwest Gulf of Maine since 1978 concurrent with an increase in the number of sand eel in the same area supports the hypothesis that within the Gulf of Maine the present distribution of humpback whales is due to the distribution of their apparent principal prey, the sand eel. A similar correlation between humpback whale sightings and sand eel abundance on Georges Bank was not significant (r = 0.24, df = 18) despite dense patches of sand eel in that region. Therefore, within the combined Gulf of Maine-Georges Bank regions, factors other than simply prey availability must influence the feeding distribution of the humpback whale We argue that the bottom topography of the Gulf of Maine and the foraging behavior of the whales are critical factors influencing their present feeding distribution. In the northwest Atlantic, the major summer con- centrations of humpback whales, Megaptera novae- angliae, occur off the coasts of Newfoundland- Labrador and off the coast of New England in the Gulf of Maine which includes Georges Bank (Katona et al. 1980; Whitehead et al. 1982). During this period feeding is their principal activity. The major winter concentrations in the western North Atlan- tic occur along the Antillean Chain in the West Indies, principally on Silver and Navidad Banks which lie north of the Dominican Republic (Winn et al. 1975; Balcomb and Nichols 1978; Whitehead and Moore 1982). During this season conception and calving are their primary activities; food does not seem to be an important determinant of the hump- backs in these areas (Whitehead and Moore 1982). Humpbacks have been generally considered coastal animals (Mackintosh 1965). However, their migratory routes between regions of winter breed- ing and summer feeding in the northwest Atlantic (based on sighting data) occur in deeper, slope waters off the continental shelf (Hain et al. 1981; Kenney et al. 1981; Payne et al. 1984). Several possi- ble offshore routes between winter and summer grounds suggest reasonably distinct stocks (Katona ^anomet Bird Observatory, Marine Mammal and Seabird Studies, Box 936, Manomet, MA 02345. 2Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. et al. 1980). Kenney et al. (1981) suggested that for the Gulf of Maine stock, the Great South Channel (Fig. 1) is the major exit-entry between the Gulf of Maine feeding area and the deeper, offshore migra- tion route. Humpback whales have been described as general- ists in their feeding habits (Mitchell 1974). The reported prey of humpbacks in the Gulf of Maine are Atlantic herring, Clupea harengus; Atlantic mackerel, Scomber scombrus; pollock, Pollachius virens; and the American sand eel, Ammodytes americanus (Gaskin 1976; Katona et al. 1977; Watkins and Schevill 1979; Kraus and Prescott 1981). In recent years, observations of feeding humpbacks indicate that sand eels have become an increasingly important prey item in the Gulf of Maine (Overholtz and Nicolas 1979; Hain et al. 1982; Mayo 1982). Kenney et al. (1981) hypothesized that the ob- served distribution of the Gulf of Maine humpback stock was due to the distribution of sand eel, their apparent principal prey species. However, the pres- ent distribution of the humpback whale in the Gulf of Maine and throughout the remaining shelf waters of the northeastern United States is not so clearly related to the distribution of sand eel as was sug- gested. Although we recognize an important predator-prey interaction between humpbacks and sand eel, we hypothesize that behavior and bottom Manuscript accepted July 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 271 FISHERY BULLETIN: VOL. 84, NO. 2 GOM = Gulf of Maine GB=Georges Bank Figure 1— The geographical areas and NMFS/NEFC bottom-trawl survey strata in the study area (upper) and the combined strata into regions (lower) referred to throughout the text. 272 PAYNE ET AL.: DISTRIBUTION OF HUMPBACK WHALE topography are also critical factors in the foraging strategy of humpbacks, hence the present distribu- tion of these whales. We base this hypothesis on observed sightings of humpbacks throughout the shelf waters of the northeastern United States in relation to sand eel abundance, and on an apparent shift in the center of feeding areas used by hump- backs in the Gulf of Maine since the mid-1970's. METHODS The collection of fisheries data used in these analyses was carried out by National Marine Fish- eries Service/Northeast Fisheries Center (NMFS/ NEFC) scientists and technicians on domestic research vessels during standardized spring bottom- trawl surveys. These surveys measure trends in fin- fish population abundance and have been used to monitor changes in the size and composition of fin- fish biomass (Clark and Brown 1977; Grosslein et al. 1980). Meyer et al. (1979) found that spring (March-May) bottom-trawl surveys accurately reflect trends in sand eel abundance. Therefore, the fisheries data we examined were from these surveys, 1978-82. The stratified mean catch per tow of sand eel was calculated for each region and considered propor- tional to the population size within each region. We transformed the mean catch into logarithmic values; then, using a two-way analysis of variance (F- statistic), we compared sand eel population size by region and year. The survey area includes shelf waters from Cape Hatteras north to Nova Scotia and has been spatially stratified by the NMFS/NEFC, based principally on depth and latitude (Grosslein 1969). Sampling sta- tions are randomly assigned within a stratum and the number of stations allocated to strata approx- imately in proportion to the area of each stratum (Grosslein 1969). In this study, individual stratum have been combined into regions (Fig. 1), in a man- ner consistent with NMFS/NEFC management units. The two important regions emphasized are the Gulf of Maine and Georges Bank. Sightings of humpback whales were recorded by observers from the Manomet Bird Observatory (MBO) on NMFS/NEFC research vessels conducting standardized surveys. Observations were recorded continuously along the predetermined cruise path between the sampling stations (following Payne et al. (1984)) in 15-min periods where each period represents a transect. Thus, the duration of each observation period was constant, but the linear km surveyed within each 15-min period depended upon vessel speed. The location (latitude-longitude) of each 15-min observation and the location and num- ber of humpback whales observed were recorded and assigned to appropriate regions to facilitate direct comparisons between the observed number of humpbacks per linear km (humpbacks/effort) and potential prey densities. Humpback whales are generally present in the study area from spring through fall (March-Novem- ber) and absent during the winter (CETAP 1982). Therefore, sighting data and effort for winter months were excluded from the analyses. We also examined sighting data collected only during op- timum sea conditions less than Beaufort (Kenney et al. 1981) (<16 nmi/h). Difference between the number of humpbacks/effort sighted by region and year were also compared by a two-way analysis of variance (F-statistic). A coefficient of correlation (r) from the linear regression between the stratified mean catch of sand eel (log) and the number of humpbacks/effort was used to determine whether concentrations of hump- back whales co-occurred with patches of sand eel within regions of the Gulf of Maine and Georges Bank. A P < 0.05 was considered statistically significant. RESULTS Distribution of Sand Eel The stratified mean number of sand eel varied sig- nificantly between regions on Georges Bank (F = 14.14, df = 3, 12) and in the Gulf of Maine (F = 16.90, df = 2, 8). On Georges Bank, sand eel were very abundant on the shoals with catches ranging from 1.117 sand eel/tow (log value) in 1978 to 2.846 (log value) in 1982 (Table 1). Sand eel were absent from most tows along the northern and shelf edges. Sand eel were also abundant in the southwest Gulf of Maine ranging from 0.670 sand eel/tow (log value) in 1978 to 2.422 in 1981 (Table 1). Sand eel were not abundant in the deeper, central Gulf of Maine This patchy distribution reflects a known preference of the sand eel for sand-bottom substrates (Bigelow and Schroeder 1953) characteristic of submarine banks and shoals. No significant differences were found between the stratified mean catch per tow (log value) by year. Distribution of Humpback Whales Since 1978, the observed number of humpbacks/ effort in the Gulf of Maine has steadily increased 273 FISHERY BULLETIN: VOL. 84, NO. 2 Table 1.— Stratified mean number of sand eel per tow + SE (in parentheses) and the number of sampling tows (lower number) by region and year. Region 1978 1979 1980 1981 1982 Georges Bank shoals northern edge shelf edge central bank Gulf of Maine central gulf southern southwest 1.117 (0.233) 15 0.000 9 0.100 (0.707) 15 0.941 (0.182) 21 0.000 64 0.000 9 0.670 (0.371) 20 1.200 (0.305) 30 0.256 (0.211) 16 0.000 2.752 (0.590) 15 0.000 8 0.000 1.850 (0.499) 15 0.747 (0.464) 8 0.000 2.846 (0.691) 15 0.000 8 0.000 14 14 0.410 0.236 (0.202) (0.132) 38 18 10 14 0.654 0034 (0.396) (0.341) 19 19 0.012 (0.012) 61 0.141 (0.101) 47 0.055 (0.545) 45 0.625 0.116 (0.422) (0.115) 12 6 1.286 1.240 (0.289) (0.384) 34 16 0.000 47 1.077 0.116 (0.617) (0.115) 6 6 2.422 0.860 (0.756) (0.318) 18 21 (Table 2). Over 90% of the humpbacks/effort ob- served each year in the combined Georges Bank-Gulf of Maine waters were seen in the Gulf of Maine. The increased number of humpbacks/effort observed was significantly different between regions in the Gulf of Maine (F = 7.098, df = 2, 8). The greatest con- centrations of humpbacks in the Gulf of Maine are located in the southwest region (Table 2). Between 1978 and 1982, 82% of the total humpbacks/effort in the Gulf of Maine were observed in the southwest region. The importance of this region for feeding humpbacks has been previously reported (Kenney et al. 1981; Hain et al. 1982). Although there were no significant differences between the number of humpbacks/effort seen by year (F = 0.824, df = 4, 12) or region (F = 0.609, df = 3, 12) on Georges Bank, the number of hump- backs/effort observed on the bank has steadily de- clined since 1978. Sixty percent of the humpbacks/ effort observed on Georges Bank between 1978 and 1982 occurred during 1978 (Table 2). We examined the apparent increase in the south- west Gulf of Maine more thoroughly by dividing it into two smaller components (Table 3), a southern which extends from the Great South Channel north along the outside of Cape Cod (NMFS/NEFC strata 23, 25, from Figure 1) and a northern which centers on Stellwagen Bank (NMFS/NEFC strata 26, 27, from Figure 1). The number of humpbacks/effort observed within the southwest Gulf of Maine-north- ern segment steadily increased by an order of mag- nitude from 1.86 x 10 ~2 whales/effort in 1978 to 29.01 x 10"2 whales/effort in 1982. Therefore, the observed increase in the number of humpbacks/ effort in the southwest Gulf of Maine since 1978 has occurred primarily in the northern half of this region (NMFS/NEFC strata 26, 27). Table 2.— The number of humpback whales per linear km x 10 "2 (humpbacks/effort) seen during shipboard observations and the total number of linear km surveyed (in parentheses) by region and year. Region 1978 1979 1980 1981 1982 Georges Bank shoals northern edge shelf edge central bank Gulf of Maine central gulf southern southwest — 0.189 — — — (480.9) (529.0) (190.0) (342.6) (744.5) 1 .500 — — (200.0) (176.8) (66.5) (230.0) (213.6) (115.6) 0.168 0.285 0.299 (593.6) (701.9) (334.4) 0.750 (933.1) 2.449 (489.8) 1.174 (681 .2) 0.119 (841.7) 0.828 (482.8) 2.817 (745.4) (966.0) (267.6) 7.679 (547.0) (89.8) (207.0) (895.9) 0.855 (467.6) 0.393 (254.2) 11.172 (454.9) (222.7) 0.225 (198.6) 0.116 (863.5) (1,172.8) 1.662 (223.5) 6.814 (692.5) -2 Table 3.— The number of humpback whales per linear km x 10 (humpbacks/effort) seen during shipboard observations and the total number of linear km surveyed (in parentheses) within the par- titioned southwest Gulf of Maine. Region 1978 1979 1980 1981 1982 Northern 1.864 2.655 10.794 22.469 29.014 (strata 26, 27) (34.9) (263.6) (333.5) (252.6) (299.6) Southern 0.556 3.113 2.811 1.987 3.308 (strata 23, 25) (359.3) (481.8) (213.5) (202.3) (392.9) Correlation Between Humpback Whale Distribution and Sand Eel Abundance A significant correlation (r = 0.81, df = 13) ex- ists between the observed number of humpbacks/ effort and the log-mean number of sand eel/tow by region within the Gulf of Maine (Fig. 2). This in- dicates that within the Gulf of Maine the distribu- tion of humpback whales do co-occur with dense patches of sand eel in that region. The greatest den- sities of sand eel in the Gulf of Maine and the greatest observed numbers of humpbacks/effort have both occurred in the southwest Gulf of Maine since 1978. This supports the hypothesis by Kenney et al. (1981) that within the Gulf of Maine, the 274 PAYNE ET AL.: DISTRIBUTION OF HUMPBACK WHALE 10.0 8.0- 6.0 O Ul m 4.0 HI _l < I o < en Q. D I 2.0- Georges Bank y=0.20-0.08x n=20 r---0.24 —i- 1.0 2.0 STRATIFIED MEAN SAND EEL PER TOW (LOG) 3.0 Figure 2— The regression and correlation coefficient (r) between the stratified mean number of sand eel/tow (log value) and the number of humpback whales/effort x 10 ~2 by region and year on Georges Bank (closed circles) and in the Gulf of Maine (open cir- cles). observed distribution of the humpback whale was due to the distribution of sand eel. However, the correlation between the observed number of humpbacks/effort and the log mean num- ber of sand eel/tow by region on Georges Bank (Fig. 2) was not significant (r = 0.24, df = 18). The mean number of sand eel/tow (log value) on Georges Bank was greatest on the shallow shoals. Only one hump- back whale was observed on the shoals between 1978 and 1982. Our data does not support any co-occurrance between humpback whale distribution and sand eel abundance on Georges Bank despite dense patches of sand eel in that region. DISCUSSION Our data suggest that the distribution of hump- back whales in the Gulf of Maine-Georges Bank region is presently centered in the southwest Gulf of Maine. This distribution is correlated with dense concentrations of sand eel, a principal prey item, which has dramatically increased throughout shelf waters of the eastern United States including the southwest Gulf of Maine since the mid-1970's (Meyer et al. 1979; Sherman et al. 1981). This increase in sand eel followed a decline of Atlantic herring stocks from the mid-1960's to the mid-1970's (Anthony and Waring 1980; Grosslein et al. 1980), and possible replacement by sand eel of depleted fish stocks in the northwest Atlantic (Sherman et al. 1981). The correlations between the humpback distribution in the Gulf of Maine and sand eel abundance supports the theory by Kenney et al. (1981) that the present distribution of the whales in that region is due to the distribution of sand eel. A demonstrated shift in the humpback distribution since the mid-1970's from the upper Gulf of Maine-lower Bay of Fundy southward into the southwest Gulf of Maine also supports this theory. A 10-yr summary of observations from Mt. Desert- Rock, ME (MDR, Fig. 1) in the northern Gulf of Maine shows a dramatic decrease in the number of humpback sightings/observer hour since 1977 (Mul- lane and Rivers 1982). The maximum number of humpbacks observed in that summary occurred in 1975 (98 whale sightings, 0.123 humpbacks/observer hour). Only 10 humpbacks were seen from 1978 to 1982, and the maximum number of humpbacks/ef- fort since 1975 has been 0.005/observer hour in 1982. This decline in the number of humpbacks at MDR coincides with the increased numbers of hump- backs observed in the southwest Gulf of Maine. Twelve of the 17 humpbacks photo-identified from 1975 to 1977 at MDR have subsequently been seen in the southwest Gulf of Maine, principally on Stell- wagen Bank. At least three of these whales have been observed during three different years on Stell- wagen Bank since they were first identified at MDR (Mullane and Rivers 1982). In comparison, only one whale identified at MDR has consistently returned to the coastal waters of eastern Maine and New Brunswick. Katona et al. (1977) also listed the Grand Manan Banks, Briers Island-St. Mary's Bay, Nova Scotia, and the lower Bay of Fundy as areas of humpback congregations. However, humpbacks were not common in the Bay of Fundy during 1981 and 1982 (Kraus and Prescott 1981, 1982). Shifts in the distribution of humpbacks caused by changes in the distribution and density of prey species have been shown elsewhere (Lien and Merd- soy 1979; Whitehead et al. 1980). We believe that the correlations between humpbacks/effort and mean sand eel catches in the southwest Gulf of Maine, and the demonstrated decline of humpbacks throughout the upper Gulf of Maine-lower Bay of Fundy concurrent with an increase in the numbers 275 FISHERY BULLETIN: VOL. 84, NO. 2 of humpbacks in the southwest Gulf of Maine reasonably explains the present distribution of humpbacks within the Gulf of Maine. However, it does not adequately explain the paucity of hump- backs on Georges Bank (Table 2) and throughout the remaining shelf waters of the northeastern United States (Hain et al. 1981; Kenney et al. 1981; Payne et al. 1984), areas where sand eel have also increased since 1975. The nonsignificant correlation between humpbacks/effort and the log-mean catches of sand eel/tow on Georges Bank suggests that factors other than simply food concentrations, perhaps behavioral or environmental, may influence the humpback's feeding strategy and location. Sutcliffe and Brodie (1977) reported that hump- backs are led into ecological or oceanographic bound- aries (i.e., isopleths or shelf-edges) and feed in patchy areas of dense prey aggregations along these boundaries. A change in depth on the shelf is often accompanied by a concentration of near-surface zoo- plankton; in general, the more abrupt the change, the greater the concentration (Sutcliffe and Brodie 1977). Concentrations are especially noticeable along the edge of banks where the availability of prey is most affected (Jaansgard 1974). Reay (1970) found that sand eel concentrations are greatest on the edges of sandy banks where currents and prey (zooplankton) are optimum; thus the whales, in seek- ing the highest concentrations of prey, feed most frequently along the edges of the banks (Sutcliffe and Brodie 1977; Brodie et al. 1978). Observations of feeding humpbacks in the Gulf of Maine have oc- curred primarily along the edge of submarine banks or canyons (Hain et al. 1982; CETAP 1982). If bottom topography influences feeding behavior of humpbacks (by concentrating prey), then the paucity of humpbacks on Georges Banks and throughout the mid-Atlantic Bight regions becomes more understandable. The floor of the broad mid- Atlantic Bight is gently sloping continental shelf with no relief until it steepens sharply at the shelf break, at about 200 m depth, to form the continen- tal slope. Since the feeding behaviors for humpbacks described by Hain et al. (1982) occur principally over a shelf-floor with rugged relief, the strategies used by humpbacks seem most efficient in these waters. This also explains the present lack of sightings in the mid-Atlantic shelf waters and the offshore migration route between calving and feeding areas. It seems energetically advantageous for the hump- back, a relatively slow-moving whale, to migrate over deep water with little apparent feeding, then feed on the densely concentrated prey along the bot- tom profiles of the Gulf of Maine. We maintain that humpbacks are merely utilizing the first concentrations of prey available to them in spring, after they reach shelf-waters from their offshore migration route between winter-calving and summer-feeding grounds. The humpbacks seem to use the Great South Channel as an entry-exit in- to the Gulf of Maine (as hypothesized by Kenney et al. (1981)), and follow the bottom profile northward, using this profile to their feeding advantage until they reach the dense concentrations of sand eel available within the southwest Gulf of Maine. The quantities of sand eel available to humpbacks at this location have allowed the whales to remain through- out the feeding season; therefore, the recent paucity of sightings in the northern Gulf of Maine. ACKNOWLEDGMENTS The authors wish to thank T. R. Azarovitz, S. K. Katona, P. Major, M. P. Pennington, M. P. Sissen- wine, G. Waring, H. Whitehead, and anonymous reviewers for criticizing previous drafts of this manuscript. The study was funded by the National Marine Fisheries Service, Northeast Fisheries Center, Woods Hole, MA. LITERATURE CITED Anthony, V. C, and G. Waring. 1980. The assessment and management of the Georges Bank herring fishery. Int. Counc. Exp. Sea, Rapp. P.-v. Reun. 177:72-111. Balcomb, K. C., and G. Nichols. 1978. Western north Atlantic humpback whales. Rep. Int. Whaling Comm. 28:159-164. BlGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53:1-577. Brodie, P. F., D. D. Sameoto, and R. W. Sheldon. 1978. Population densities of euphausiids off Nova Scotia as indicated by net samples, whale stomach contents, and sonar. Limnol. Oceanogr. 23:1264-1267. CETAP. 1982. A characterization of marine mammals and turtles in the mid- and north Atlantic areas of the U.S. outer continen- tal shelf. Final Rep. Cetacean and Turtle Assessment Pro- gram, Univ. Rhode Island, Bur. Land Manage. Contract No. AA551-CT8-48, 450 p. U.S. Dep. Inter., Wash., D.C. Clark, S. H., and B. E. Brown. 1977. Changes in biomass of finfishes and squid from the Gulf of Maine to Cape Hatteras, 1963-74, as determined from research vessel survey data. Fish. Bull., U.S. 75:1- 21. Gaskin, D. E. 1976. The evolution, zoogeography, and ecology of Cetacea. Oceanogr. Mar. Biol., Annu. Rev. 14:247-346. Grosslein, M. D. 1969. Groundfish survey program of BCF Woods Hole. Commer. Fish. Rev. 31(8-9):22-35. 276 PAYNE ET AL.: DISTRIBUTION OF HUMPBACK WHALE Grosslein, M. D., R. W. Langton, and M. P. Sissenwine. 1980. Recent fluctuations in pelagic fish stocks of the North- west Atlantic, Georges Bank region, in relation to species interactions. Rapp. R-v. Reun. Cons. Perm. int. Explor. Mer 177:374-404. Hain, J. H., G. R. Carter, S. D. Kraus, C. A. Mayo, and H. E. Winn. 1982. Feeding behavior of the humpback whale, Megaptera rwvaeangliae, in the western north Atlantic Fish. Bull., U.S. 80:259-268. Hain, J. H. W., R. K. Edel, H. E. Hays, S. K. Katona, and J. D. Roanowicz. 1981. General distribution of cetaceans on the continental shelf waters of the northeastern United States. In A charac- terization of marine mammals and turtles in the mid- and north Atlantic areas of the U.S. outer continental shelf, p. II-1-257. Ann. Rep. 1979, Cetacean and Turtle Assessment Program, Univ. Rhode Island, Bur. Land Manage. Contract No. AA551-CT8-48, U.S. Dep. Inter, Wash, D.C. Jangaard, P. M. 1974. The capelin (Mallotus villosus): biology, distribution, exploitation, utilization and composition. Bull. Fish. Res. Board Can. 186, 70 p. Katona, S, W Steiner, and H. E. Winn. 1977. Marine mammals. In Center for natural areas, a sum- mary and analysis of environmental information on the Con- tinental Shelf from the Bay of Fundy to Cape Hatteras, Vol. 1, No. 2, p. 1-167. Final Rep. Bur. Land Manage, U.S. Dep. Inter, Wash, D.C, Contract No. AA550-CT6-45. Katona, S. K, P. Harcourt, J. S. Perkins, and S. D. Kraus. 1980. Humpback whales: A catalog of individuals identified by fluke photographs. College of the Atlantic, Bar Harbor, ME. Kenney, R. D, D. R. Goodale, G. P. Scott, and H. E. Winn. 1981. Spatial and temporal distribution of humpback whales in the CETAP study area. In A characterization of marine mammals and turtles in the mid- and north Atlantic areas of the U.S. outer continental shelf, p. V-l-9. Ann. Rep. 1979, Cetacean and Turtle Assessment Program, Univ. Rhode Island, Bur. Land Manage Contract No. AA551-CT8-48, U.S. Dep. Inter, Wash, D.C. Kraus, S. D, and J. H. Prescott. 1981. Distribution, abundance and notes of the large ceta- ceans of the Bay of Fundy, summer and fall 1980. Natl. Mar. Fish. Serv, Northeast Fish. Cent. Contract No. NA-80-FA- 2-00048, 65 p. 1982. The North Atlantic right whale (Eubalaena glacialis) in the Bay of Fundy, 1981, with notes on distribution, abun- dance, biology and behavior. Natl. Mar. Fish. Serv, North- east Fish. Cent. Contract No. NA-81-FA-C-00030, 105 p. LEIN, J, AND B. MERDSOY. 1979. The humpback is not over the hump. Nat. Hist. 88(6): 46-49. Mackintosh, N. A. 1965. The stocks of whales. Fishing News (Books) Ltd, Lond, 232 p. Mayo, C. A. 1982. Observations of cetaceans: Cape Cod Bay and Southern Stellwagen Bank, Massachusetts, 1975 - 1979. Natl. Tech. Inf. Serv. Rep. No. PB82-186263, 68 p. Meyer, T. L, R. A. Cooper, and R. W Langton. 1979. Relative abundance, behavior, and food habits of the American sand lance, Ammodytes americanus, from the Gulf of Maine Fish. Bull, U.S. 77:243-253. Mitchell, E. 1974. Trophic relationships and competition for food in North- west Atlantic whales. In M. D. B. Burt (editor), Proceedings of the Canadian Society of Zoologists annual meeting, June 2-5, p. 123-133. MULLANE, S. J, AND A. RIVERS. 1982. Mt. Desert Rock, Maine - Annual report. Allied Whale, College of the Atlantic, Bar Harbor, Maine, 27 p. OVERHOLTZ, W J, AND J. R. NICOLAS. 1979. Apparent feeding by the fin whale, Balaenoptera physalus, and humpback whale Megaptera novaeangliae, on the American sand lance, Ammodytes americanus, in the northwest Atlantic Fish. Bull, U.S. 77:285-287. Payne, P. M, L. A. Selzer, and A. R. Knowlton. 1984. Distribution and density of cetaceans, marine turtles, and seabirds in the shelf water of the northeastern United States, June 1980 - December 1983, based on shipboard observation. Natl. Mar. Fish. Sen, Northeast Fish. Cent. Contract No. NA-81-FA-C-00023, 246 p. Reay, P. J. 1970. Synopsis of biological data on North Atlantic sand eels of the genus Ammodytes: A. tobianus, A. dubius, A. ameri- canus, and A. marinus. FAO Fish. Synop. No. 82, 28 P- Sherman, K. C, C. Jones, L. Sullivan, W Smith, P. Berrein, and L. Ejstmont. 1981. Congruent shifts in sand eel abundance in western and eastern North Atlantic ecosystems. Nature 291:486- 489. Sutcliffe, W H, AND P. F. Brodie. 1977. Whale distribution in Nova Scotia Waters. Fish. Mar. Serv. Tech. Rep. 722, 83 p. Fisheries and Marine Service, Bedford Institute of Oceanography, Dartmouth, Nova Scotia Watkins, W A, and W. E. Schevill. 1979. Aerial observation of feeding behavior in four baleen whales: Eubalaena glacialis, Balaenoptera borealis, Megap- tera novaeangliae, and Balaenoptera physalus. J. Mammal. 60:155-163. Whitehead, H, P. Harcourt, K. Ingham, and H. Clark. 1980. The migration of humpback whales, past the Bay de Verde Peninsula, Newfoundland, during June and July, 1978. Can. J. Zool. 58:687-692. Whitehead, H, and M. J. Moore. 1982. Distribution and movements of West Indian humpback whales in winter. Can. J. Zool. 60:2203-2211. Whitehead, H, R. Silver, and P. Harcourt. 1982. The migration of humpback whales along the northeast coast of Newfoundland. Can. J. Zool. 60:2173-2179. Winn, H. E, R. K. Edel, and A. G. Taruski. 1975. Population estimate of the humpback whale (Megaptera novaeangliae) in the West Indies by visual and acoustic tech- niques. J. Fish. Res. Board Can. 32:499-506. 277 SEABIRDS NEAR AN OREGON ESTUARINE SALMON HATCHERY IN 1982 AND DURING THE 1983 EL NINO Range D. Bayer1 ABSTRACT In the summer of 1982, 14.4 million salmon, Oncorhynchus sp., smolts were released at the Yaquina Estuary, OR; and in the summer of 1983, 12.8 million salmon smolts were released. Within hours after release, fish-eating seabirds aggregated at the estuary mouth. In 1982, the number of no seabirds was significantly correlated with the number of days since a release. In 1983, however, numbers of common murres, Uria aalge; gulls, Larus sp.; and brown pelicans, Pelecanus occidentalis, were significantly in- versely correlated with the date of a release, and the number of cormorants, Phalacrocorax sp., was significantly more abundant the second day after a release. In contrast, numbers of Caspian terns, Sterna caspia, and pigeon guillemots, Cepphus columba, showed no relationship with releases in 1983. There were significantly more cormorants and marbled murrelets, Brachyramphus marmoratus, in 1983 than in 1982. There were also significantly more murres in 1983 than in 1982 before 1 August, but fewer afterwards. Gull and brown pelican numbers were about the same between years, but significant- ly fewer pigeon guillemots were present in 1983 than in 1982. Seabirds have been estimated to consume 29% of the pelagic fish production within 45 km of a British seabird colony (Furness 1984b), and several simula- tion models for various geographical areas indicate that 20-30% of the annual pelagic fish production may be preyed upon by seabirds (Furness 1984a). Since the mortality of salmon, Oncorhynchus sp., smolts as a result of predation and environmental factors is greater shortly after they first enter the ocean than after they move offshore (Parker 1962, 1968), the impact of seabird predation on salmon smolts just released along a coast could also be significant. El Nino is the intrusion of anomalously warm water off the coast of Peru and Ecuador (Barber and Chavez 1983); an El Nino of varying intensity oc- curs on the average of every 3-5 yr (Quinn et al. 1978; Duffy 1983a). Along the Oregon coast, warm- water conditions concurrent with an El Nino appear much more rarely, and in the last century have oc- curred only in 1983, 1957-1958, and perhaps in 1941 (Huyer 1983; Reed 1983). The impact of seabirds on hatchery-released salmon smolts would be expected to be greater in years of anomalously warm water associated with El Nino, when natural prey for sea- birds become rare and seabirds starve or have low nesting success (Boersma 1978; Duffy 1983a, b; Ainley 1983; Schreiber and Schreiber 1984). 'Oregon Aqua-Foods, Inc., 2000 Marine Science Drive, New- port, OR 97365; present address: P.O. Box 1467, Newport, OR 97365. Here, I correlate bird numbers with salmon smolt releases at Yaquina Estuary, OR, and examine variation in bird numbers between the summer of 1982 and the summer of 1983, when warm water associated with an El Nino was present. STUDY AREA AND METHODS Yaquina Estuary (Fig. 1) is located on the mid- Oregon coast and is a drowned river valley. It has an intertidal and submerged area of 15.8 km2 (Oregon State Land Board 1973). During this study, all releases were from the site designated as OAF in Figure 1. The most abundant seabird nesting nearby was the common murre, Uria aalge, but western gulls, Larus occidentalis; Brandt's cormorants, Phalacro- corax penicillatus; pelagic cormorants, P. pelagicus; and pigeon guillemots, Cepphus columba, also nested there (Table 1; Pitman et al. in press). Within Ya- Table 1.— Distance of nesting birds from the mouth of Yaquina Estuary in 1979 (calculated from Pitman et al. in press). Cumulative number of nesting birds <7 km <20 km <25 km <45 km <50 km common murres1 22,800 26,800 26,800 western gulls 398 528 536 cormorants 418 653 1,581 pigeon guillemots 45 191 201 26,800 322,000 541 1,231 1,727 3,041 206 220 'Includes all breeding and nonbreeding adults at colony. ^Estimated for 1983 (USFWS, aerial survey; pers. obs.). includes 1983 as well as 1979 estimates. Manuscript accepted July 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 279 aS'T-f- 2-*^ FISHERY BULLETIN: VOL. 84, NO. 2 x =JETTY OBSERVAT ION POINT REGIONS 0A SC SB (D D o 33 KM TO HEAD OF TIDE I 2 4"^ 0 2 Figure 1.— Yaquina Estuary study regions. OAF indicates site of smolt releases. quina Estuary, <30 pairs of western gulls (Bayer 1983) and an undetermined number of pigeon guillemots also nested in association with manmade structures. The typical nesting phenology of these birds at Yaquina Head (which is about 6.5 km north of Yaquina Estuary) has been examined by Scott (1973) and Bayer (1983) with murres beginning to fledge young in early July, gulls and pelagic cor- morants in late July, and Brandt's cormorants and pigeon guillemots in early to mid- August. However, it would be invalid to assume that nesting in 1983 followed the chronologies of typical years because nesting success for cormorants and murres was ab- normally low in 1983 with eggs and young being abandoned (Bayer2). Although the nesting success of gulls was not unusually low in 1983 (Bayer fn. 2), the chronology of their nesting might have been different than in 1982. Thus, comparing 1982 and 1983 bird numbers at Yaquina Estuary for the same stage of the nesting cycle would be tenuous. Brown pelicans, Pelecanus occidentalis, and Caspian terns, Sterna caspia, do not nest in this area. I divided the estuary and the area around its mouth into four censusing regions (Fig. 1) with region A having an area of about 1.8 km2; region B, 0.5 km2; region C, 3.0 km2; and region D, 3.2 2Bayer, R. D. In prep. Breeding success of seabirds along the mid-Oregon coast concurrent with the 1983 El Nino. Unpubl. manuscr. P.O. Box 1467, Newport, OR 97365. km2. I censused birds from observation points where I could overlook the estuary or estuary mouth with a 20 x telescope when glare, heat waves, and water conditions did not obscure birds. I censused the areas around the mouth of the jetties from an observation point about halfway out on the south jetty (Fig. 1). The boundaries of region A were estimated -by using the distance to the first naviga- tion buoys to the west of the jetties as a radius that was about 1.5 km from the observation point and 1.0 km from the end of the jetties to estimate the outer boundary. All taxa except pigeon guillemots were censused during a single continuous sweep of nonoverlapping portions of a region; pigeon guille- mots were enumerated during two sweeps per por- tion with the maximum number of the two sweeps recorded. I censused "active" (see below) gulls and cor- morants, nonflying common murres and pigeon guillemots (including guillemots standing on station- ary objects), roosting Caspian terns, all brown pelicans, and all marbled murrelets, Brachyramphus marmoratus. "Active" gulls were those that flew over or sat in the water (gulls sitting on stationary roosts were not included). Gull species included western, glaucous-winged, L. glaucescens, and western x glaucous-winged gull hybrids (Hoffman et al. 1978). Cormorants present were usually either Brandt's or pelagic cormorants, but some double- crested cormorants, P. auritus, were also included. "Active" cormorants were those on the water sur- face or those making short flights in association with a feeding flock; cormorants on transit flights through a region or roosting on stationary objects were not counted. Only nonflying murres and guille- mots were included because others flew through regions A and B without landing (and feeding). Al- though roosting Caspian terns were obviously not feeding, they were recorded because their numbers were an index of the total numbers present and because it was not possible to count foraging (i.e., flying) Caspian terns accurately. There were 166 censuses during 37 d from 1 June to 16 September 1982 at regions A-D during variable tidal conditions, and 39 censuses within 2 h of low tides before 1500 Pacific Daylight Time (PDT) during 39 d from 1 June to 30 August 1983 at regions A-C. Each census took 45-75 min, depending upon the number of birds present. Comparisons of bird numbers between 1982 and 1983 were only made for censuses within 2 h of low tides before 1500 PDT. Comparisons were made during the 1 June to 30 August period for brown pelicans, "active" cormorants, "active" gulls, and 280 BAYER: SEABIRDS NEAR OREGON ESTUARINE SALMON HATCHERY pigeon guillemots because the numbers of these birds during this period did not show any signs of seasonal variation. But for common murres, the periods of comparison were 1 June-31 July and 1-30 August, and the periods for Caspian terns were 1 June-10 July, 11 July-5 August, and 6-30 August. The periods for common murres and Caspian terns were chosen because in one or both years there were marked seasonal changes in bird numbers between or among these periods. The number of days postrelease refers to the num- ber of daylight periods after a smolt release (Myers 1980). For example, if smolts were released on Mon- day night or early Tuesday morning, then Tuesday after dawn would be considered as 1-d postrelease (i.e., the first day, or first daylight period, after a release). If variances were not significantly different, then the student's £-test for two means or the analysis of variance (ANOVA) for three or more means were calculated to determine statistical differences be- tween or among means. If variances were signifi- cantly different, the Mann-Whitney U test or nor- malized Mann-Whitney z test (Zar 1974, p. 109-113) for two samples or the Kruskal-Wallis rank H or Hc (if ranks were tied) test (Zar 1974, p. 139-142) for three or more samples was used. All tests were two-tailed. RESULTS AND DISCUSSION Smolt Releases Oregon Aqua-Foods, Inc. (OAF) has released 2 million or more salmon smolts (almost all coho salmon, Oncorhynchus kisutch) each year since 1977 into Yaquina Estuary between June and August. In 1982 and 1983, the proportion that were coho salmon was 98% and 94%, respectively; the re- mainder were chinook salmon, 0. tshawytscha. Un- til 1983, these releases were under variable tidal con- ditions in the evening just after dark to minimize bird predation of smolts as they were released. In 1983, salmon smolts were released either in the evening or early morning on the ebbing tide while it was still dark. Salmon smolts do not immediately swim to the ocean after they are released. Myers (1980) found that the number of OAF smolts in the Yaquina Estuary declined exponentially after a release. Dur- ing June- August releases in 1978, half the smolts left the estuary within an average of 3.9 d (SE = 0.7 d, range 1.7-9.0 d, N = 9 releases) with a few smolts remaining in the estuary several months (calculated from Myers 1980). There are no data to determine if the smolt residency time in the estuary differed between 1982 and 1983. In 1982 and 1983 from June through August, the interval between releases averaged <2.5 d, and an average of 0.2-0.3 million fish were released each time (Table 2). Although the average release inter- val was longer and the number of fish per release usually greater in 1983, these differences were not significant (Table 2). But the biomass of fish per release was significantly greater in 1983 than in 1982 in the June- July and June- August periods (Table 2). Overall, 1.6 million fewer fish were re- leased in 1983 than 1982, but the total biomass released was almost 38 metric tons (t) greater (Table 2); this resulted from smolts weighing more on the average in 1983 (32.9 g/smolt) than in 1982 (26.7 g/smolt) (calculated from Table 2). Bird Predation of Salmon Smolts Although all birds in this study except marbled murrelets were observed with salmon smolts in their bills, the importance of smolts in these birds' diets was not documented in this study. However, Mat- thews (1983) found that coho salmon smolts com- Table 2.— Releases of salmon smolts in 1982 and 1983 at Yaquina Estuary. Total = total number or biomass of fish released during a period. Differences between years tested with student's f, Mann-Whitney U, or normalized Mann-Whitney z test. NS = not significant. Re- Release interval (d) No. fish/release (millions) Fish biomass/release (t) Period Year N X SD P x SD P Total x SD P Total June-July 1982 30 2.0 1.2 0.3 0.1 9.6 7.5 2.4 225.3 1983 25 2.4 1.7 NS 0.4 0.1 NS 9.0 11.3 4.6 <0.01 268.4 August 1982 20 1.6 0.7 0.2 0.1 4.7 7.9 3.4 158.6 1983 12 2.4 1.6 NS 0.3 0.2 NS 3.7 12.8 9.0 NS 153.0 June- 1982 50 1.8 1.0 0.3 0.1 14.4 7.7 2.8 383.9 August 1983 37 2.4 1.6 NS 0.3 0.2 NS 12.8 11.8 6.3 <0.01 421.4 281 FISHERY BULLETIN: VOL. 84, NO. 2 posed 13% of 287 prey items of common murres col- lected within 2 km of the Yaquina's jetties during the summer of 1982. Salmon smolts appeared to be most vulnerable to predation soon after a release. When they first entered the estuary after exiting a pond through a large tube, smolts seemed disoriented and milled around the surface where they could easily be caught by birds. Night releases allowed smolts several hours to become adjusted before becoming vulnerable to predators at daylight. (The only somewhat signifi- cant nocturnal bird predator were heerman's gulls, L. heermanni, but they usually numbered <50 birds, were not present for every release, and were pres- ent mainly in late July and August.) Within about 4 h after daylight after a release, some smolts were observed jumping at the mouth of the jetties in regions A and B, where birds also concentrated. For censuses of regions A-D within 2 h of low tides and within 2 d of a release in 1982, an average of 97.9% (SD - 6.3, N - 17 d) of the common murres, 91.5% (SD = 16.2, N = 17 d) of the "active" gulls, and 90.5% (SD = 26.8, N = S d) of the "active" cormorants censused were at regions A and B. But regions A and B accounted for only about 27% of the area of regions A-D. Evidently, the turbulent action of the estuarine water entering the ocean and/or the funneling ef- fect of the jetties (Fig. 1) caused the smolts to be particularly vulnerable to predators there. During the first 12 h of daylight after a release, some smolts within 0.5 km of the release site were still vulnerable to bird predation as many smolts were near the water surface. Many jumped out of the water, and some rolled on their sides exposing their silver undersides, which were highly conspic- uous against the dark water background. Gulls often sat on the water and grasped a fish as it jumped into the air. Schools of smolts also milled near the sur- face where they were clearly visible to humans (and presumably birds). Within-Day Variation in Bird Numbers Bird abundance was clearly not constant within a day, and taxa did not reach maxima synchronously (Fig. 2). Censuses within 2 h of early low tides (i.e., low tides before 1500 PDT) averaged closer to the maximum number censused daily for all taxa, and censuses near high tide were usually closer to the daily maximum than counts within 2 h of evening low tides (i.e., after 1800 PDT) for all taxa except brown pelicans (Table 3). But differences in censuses among tidal conditions within a day were only sig- IOO COMMON MURRES ( MAX=442 I) LO A 0.8 0600 0800 1000 1200 1400 1600 1800 2000 PACIFIC DAY L IGHT TIME Figure 2.— Percentage of daily maximum number of common murres, "active" gulls, "active" cormorants, and pigeon guillemots (PCs) observed on 5 August 1982 (which was two days postrelease) at regions A-D. Times and heights of measured low (LO) and high (HI) tides are indicated by open and closed triangles, respective- ly. MAX - maximum number of birds seen on 5 August. nificant for common murres and "active" gulls (Table 3). A single census at any time of day is unlikely to estimate accurately the maximum number of birds of any taxon present that day (Table 3). The average census only ranged from 10.8% to 63.7% of the daily maximum (Table 3). The best censuses to use for estimates would be those within 2 h of a morning or afternoon low tide because their averages (44-64% of daily maxima) were greater than for high and evening low tides, and their CV's (41-82%) were generally lower than for other tides (Table 3). Daily Variation in Bird Numbers On a day to day basis, bird numbers could often be seen to increase in the first day postrelease and then to decline (Fig. 3). However, the degree of in- crease was variable. Overall, murres, "active" gulls in 1983, and brown pelicans exhibited the same pat- 282 BAYER: SEABIRDS NEAR OREGON ESTUARINE SALMON HATCHERY Table 3.— Percentage of daily maximum number of birds observed within 2 h of ac- tual high tides, early low tides (i.e., time of low tide before 1500 PDT), and late low tides (i.e., time of low tide after 1800 PDT). Censuses between 6 July and 17 September 1982 at regions A-D with 9-11 censuses/d (i.e., 13-14 h period). N = total censuses; CV = coefficient of variation. Days Percent of daily maximum birds within 2 h of High tide Early low tide CV N x (%) Late low tide N X CV (0/0) CV N x (%) common murres "active" gulls brown pelicans "active" cormorants pigeon guillemots 9 10 6 4 3 31 35 21 14 10 130.8 237.9 339.1 "32.6 552.9 92.5 81.5 79.0 112.6 58.8 24 163.7 55.3 28 249.5 71.9 14 344.0 81.8 7 "61.3 65.3 8 558.O 41.0 11 120.3 100.0 11 218.1 100.0 8 349.1 58.9 6 "10.8 142.6 3 533.9 93.8 'Heterogeneity, Kruskal-Wallis Hc = 16.36, P< 0.01. heterogeneity, Kruskal-Wallis Hc = 7.62, P< 0.10. heterogeneity, Kruskal-Wallis Hc = 0.80, P > 0.10. "Heterogeneity, Kruskal-Wallis Hc = 5.62, P > 0.10. 5Heterogeneity, Kruskal-Wallis Hc = 1.87, P > 0.10. tern of more birds present the first day after a release than later; this pattern, however, was sig- nificant only in 1983 (Tables 4, 5). In contrast, only "active" cormorants were more numerous the second day after a release than on the first day; however, the differences in cormorant numbers among days were only significant in 1983 (Table 5). Numbers of pigeon guillemots and Caspian terns did not show any indication of dependence on the number of days postrelease. The differences in pigeon guillemot numbers in the 1 June-30 August period among 1,2, and 3-6 d postrelease was insig- nificant (1982: F = 0.23, df = 2, 34; P > 0.10; 1983: Kruskal-Wallis Hc = 0.61, P > 0.10). Sample sizes were too small to test differences for Caspian terns in 1982, but in 1983 variation with 1, 2, and 3-6 d postrelease was insignificant in either the 1 June- 14 July period (when there were few Caspian terns (Kruskal-Wallis Hc = 2.74, P > 0.10)) or the 15 July-30 August period (when they were abundant (Kruskal-Wallis Hc = 2.74, P > 0.10)). T FIRST DAY POSTRELEASE 80 0 \ in 0 O CORMORANTS 0 *40 CD \\ \\ 0 — 1 , x / 0N / l/\ W \ / % \ ° 9/ / 0 T 1 1 . t r t , ? T 4000 cr 3 2000 - 800 o c 400 1 J I I T , T , ■ T - 0 13 14 15 16 17 18 19 20 21 22 JULY 1983 Figure 3.— Number of brown pelicans, "active" cormorants, "active" gulls, and common murres with relation to dates of salmon smolt releases during 14-22 July 1983 censuses that were within 2 h of low tides before 1500 PDT. Table 4.— Numbers of common murres at regions A-C in 1982 and 1983 during the 1 June-31 July period when murres were abundant and the 1-30 August period when murres were infrequent in 1983. N = number of cen- suses (1 census/d within 2 h of low tides before 1500 PDT); MAX = maximum number of birds counted. 1 June-31 July 1-30 August 1-d postrelease 2-d postrelease 3-6 d postrelease N x SD MAX 1-3 d postrelease Year N x SD MAX N x SD MAX N x SD MAX 1982 1983 8 i23,053 967 4,310 13 2.63J10 2,746 9,638 6 8 131,823 2,114 5,988 362,462 2,063 6,206 2 1 "1,276 824 1,858 6 46561 711 1,972 4 51,860 2,091 4,419 10 S106 280 901 1 Heterogeneity among days: Kruskal-Wallis H = 4.88, P > 0.10. 21982 vs. 1983: Mann-Whitney U = 52, P > 0.10. 31982 vs. 1983: student's f = 2.12, df = 12, P< 0.10. "1982 vs. 1983: not tested because of small sample sizes in 1982. M982 vs. 1983: Mann-Whitney U = 38, P < 0.02. 6Heterogeneity among days: Kruskal-Wallis H = 8.91, P < 0.05. 283 FISHERY BULLETIN: VOL. 84, NO. 2 Table 5.— Comparison of bird numbers at regions A-C during 1 June-30 August period in 1982 with 1983. Day(s) = days post- release of salmon smolts, N = number of censuses (1 census/d within 2 h of low tides before 1500 PDT), and MAX = maximum number of birds counted. "active" brown "active" gulls pelicans 1 2 3-6 cor 1 morants Days(s): 1 2 3-6 2 3-6 1982 N 10 7 2 10 7 3 9 7 3 Birds (x) 1391 13811445 231 211 27 318 328 347 SD 272 294 36 36 10 10 13 41 38 MAX 919 729 470 106 30 19 38 110 88 1983 N 20 9 9 20 9 9 20 9 9 Birds (x) MOO 1332 !26 225 217 27 346 381 321 SD 349 450 25 19 22 8 33 90 15 MAX 1,311 1,200 77 84 69 20 128 286 52 11 d vs. 2 d vs. 3 Kruskal-Wallis Hc = 0.07, df = 28, P > 21 d vs. 2 d vs. 3 Kruskal-Wallis Hc = = 107,P>0.10;2d, U = 14, P> 0.10. 31 d vs. 2 d vs. 3- Kruskal-Wallis Hc = = 142.5, P< 0.02; 2 U = 20, P>0.10. 6 d: 1982, Kruskal-Wallis Hc = 0.44, P > 0.10; 1983, 14.62, P < 0.01. 1982 vs. 1983; 1 d, student's f = 0.10; 2 d, student's f = 0.25, df = 14, P > 0.10. 6 d: 1982, Kruskal-Wallis Hc = 2.44, P > 0.10; 1983, 8.71, P< 0.02. 1982 vs. 1983: 1 d, Mann-Whitney U Mann-WhitneyU = 32.5, P>0.10; 3-6 d, Mann-Whitney 6 d: 1982, Kruskal-Wallis Hc = 1.84, P > 0.10; 1983, 6.14, P < 0.05. 1982 vs. 1983: 1 d, Mann-Whitney U d, Mann-Whitney U = 49, P< 0.10; 3-6 d, Mann-Whitney Yearly Variation in Bird Numbers Cormorants were significantly more abundant for 1 and 2 d postrelease in 1983 than in 1982 but not for 3-6 d postrelease (Table 5). Brown pelicans were about as numerous in 1983 as in 1982 in the 1 June-30 August period (Table 5). Gulls were not significantly more abundant in 1983 than in 1982 in the 1 June-30 August period (Table 5), and their nesting success was also not lower in 1983 than in other years (Bayer fn. 2). But Caspian terns were significantly more abundant dur- ing the 11 July-5 August period (when many emigrated) in 1983 than in 1982 (Bayer 1984). There were an average of about 650 more com- mon murres per census in 1983 than in 1982 dur- ing the 1 June-31 July period for either 1 or 2 d post- release, but the differences were only significant for 2 d postrelease (Table 4). In contrast, there were more murres in 1982 than in 1983 during this period for 3-6 d postrelease, but there were only two samples in 1982 (Table 4). In the 1-30 August period, there were significantly fewer murres in 1983 than in 1982 (Table 4). The low numbers in 1983 resulted from the mass exodus of murres after 31 July, whereas in 1982 murre num- bers did not decline as dramatically until after 12 August. In fact, there were still more murres pres- ent within 2 h of low tides on 3 and 16 September 1982 (186 and 318 murres, respectively) than in 10 censuses on different days between 1 and 18 August 1983 (i.e., <56 murres). The early exodus of murres in 1983 probably resulted from them migrating north early because they were unusually numerous in inland marine waters of Washington during the summer of 1983 (Mattocks et al. 1983). During the June through August period at regions A-C, pigeon guillemot numbers were about 29% greater during 1982 (x = 23.9, SD = 11.0, N = 13 d) than in 1983 (x = 17.1, SD = 7.8, N = 35 d), a significant difference (t = 2.39, df = 46, P < 0.05). This decrease could have resulted from the large number of mortalities in the spring of 1983 (Hodder3). Marbled murrelets were not observed in any of 120 censuses of regions A-D in the June through 20 August period of 1982. In 1983 at regions A-C, they were observed in only 1 of 21 censuses in June and August, but an average of 3.9 murrelets/census (SD = 8.7, range 0-32, N = 17 censuses) were counted in July. The difference in the number of murrelets per census in July was significantly greater in 1983 than in 1982 (normalized Mann-Whitney z = 2.18, P < 0.05). They were only observed at region A. CONCLUSIONS It is not possible to relate the number of birds nesting near the Yaquina Estuary with the number feeding there for several reasons. First, the num- ber of nesting and nonbreeding birds is unknown, so it is not possible to determine what proportion of the birds censused were nonbreeders. Second, censuses of feeding birds represent the number of birds feeding at only one point in time, but nesting birds probably fed serially at the Yaquina Estuary (i.e., birds came and went as individuals or small flocks not as massive synchronous flocks). With serial use, the number of nesting birds using the Ya- quina Estuary could be much larger than indicated by censuses. Unfortunately, birds would have to be individually recognizable to determine the degree of serial use, and this was beyond the scope of this study. It also was not possible to tell from how far nest- ing birds came to feed at the Yaquina Estuary in either year because birds were not individually marked. Murres, however, may have come from long distances. In both years, the average number of murres one day after a salmon release (Table 4) was greater than the number of murres at a colony <7 km away (Table 1), and the maximum number 3J. Hodder, Institute of Marine Biology, Charleston, OR 97420, pers. commun., 1984. 284 BAYER: SEABIRDS NEAR OREGON ESTUARINE SALMON HATCHERY of murres simultaneously seen at the Yaquina (Table 4) was greater than the number of murres at colonies within 45 km of the Yaquina (Table 1). It was somewhat surprising that more cormorants and common murres were not at the Yaquina Estuary in 1983, because they then had a poor nesting season, probably as a result of a food short- age (Bayer fn. 2). There are several possible reasons why there were not more cormorants and murres counted in 1983. First, the number of salmon smolts available at the Yaquina Estuary might have been insufficient or the distance between the Yaquina and their nesting site too great for these birds to be dependent solely on salmon smolt releases. If the salmon smolt releases had been oftener and nearer to bird nesting colonies, the numbers of birds pres- ent could have been much greater. Second, there may have actually been many more birds in 1983 than in 1982, but a single census per day regime was inadequate to measure this (Table 3). Censuses throughout the day in 1983 or measurements of the serial use of the Yaquina Estuary in 1982 and 1983 might have indicated that there were dramatically more birds using the Yaquina in 1983 than in 1982. Finally, the lack of there not being a greater influx of birds in 1983 might be because many of the murres and cormorants that normally remained near the Yaquina dispersed to avoid the generally poor feeding conditions between releases. Many Oregon pelagic and Brandt's cormorants had aban- doned nesting by mid-July 1983 (see Bayer fn. 2; Hodder fn. 3), and many murres may have left the Oregon coast before it became apparent at the Ya- quina Estuary at the end of July. Early dispersal or migration is known for southern seabirds during an El Nino (Duffy 1983a; Schreiber and Schreiber 1984). ACKNOWLEDGMENTS I am grateful to Bill McNeil, Vern Jackson, Rob Lawrence, Mike Bauman, and Andy Rivinus of Oregon Aqua-Foods for facilitating the logistics and funding of this project; to Dan Varoujean for advice about censusing murres prior to the 1982 field season; and to Jan Hodder, Dan Matthews, Daniel W. Anderson, Peter Stettenheim, and two anony- mous reviewers for constructive comments on an earlier draft of this manuscript. LITERATURE CITED AlNLEY, D. 1983. El Nino in California? Point Reyes Bird Observ. Bull. 62:1-4. Barber, R. T., and F. P. Chavez. 1983. Biological consequences of El Nino. Science 222:1203- 1210. Bayer, R. D. 1983. Nesting success of western gulls at Yaquina Head and on man-made structures in Yaquina Estuary, Oregon. Mur- relet 64:87-91. 1984. Oversummering of whimbrels, Bonaparte's gulls, and Caspian terns at Yaquina Estuary, Oregon. Murrelet 65:87-90. Boersma, P. D. 1979. Breeding patterns of Galapagos penguins as an in- dicator of oceanographic conditions. Science 200:1481- 1483. Duffy, D. C. 1983a. Environmental uncertainty and commercial fishing: effects on Peruvian guano birds. Biol. Conserv. 26:227-238. 1983b. The foraging ecology of Peruvian seabirds. Auk 100: 800-810. Furness, R. W. 1984a. Modelling relationships among fisheries, seabirds, and marine mammals. In D. N. Nettleship, G. A. Sanger, and P. F. Springer (editors), Marine birds: their feeding ecology and commercial fisheries relationships, p. 117-126. Proc. Pacific Seabird Group, 6-8 January 1982, Can. Wildl. Serv., Can. Minist. Supply Cat. No. CW66-65/1984. 1984b. Seabird-fisheries relationships in the northeast Atlan- tic and North Sea. In D. N. Nettleship, G. A. Sanger, and P. F. Springer (editors), Marine birds: their feeding ecology and commercial fisheries relationships, p. 162-169. Proc. Pacific Seabird Group, 6-8 January 1982, Can. Wildl. Serv., Can. Minist. Supply Cat. No. CW66-65/1984. Hoffman, W., J. A. Wiens, and J. M. Scott. 1978. Hybridization between gulls (Larus glaucescens and L. occidentalis) in the Pacific Northwest. Auk 95:441-458. Huyer, A. 1983. Anomalously warm water off Newport, Oregon, April 1983. Trop. Ocean-Atmos. Newsl. 21:24-25. Mattocks, P., Jr., B. Harrington-Tweit, and E. Hunn. 1983. Northern Pacific Coast region. Am. Birds 37:1019- 1022. Matthews, D. R. 1983. Feeding ecology of the common murre, Uria aalge, off the Oregon coast. M.S. Thesis, Univ. Oregon, Eugene, 108 P- Myers, K. W. 1980. An investigation of the utilization of four study areas in Yaquina Bay, Oregon, by hatchery and wild juvenile sal- monids. M.S. Thesis, Oregon State Univ., Corvallis, 234 p. Oregon State Land Board. 1973. Oregon estuaries. State of Oregon, Div. State Lands. Parker, R. R. 1962. Estimations of ocean mortality rates for Pacific salmon (Onc&rhynchus). J. Fish. Res. Board Can. 19:561-589. 1968. Marine mortality schedules of pink salmon of the Bella Coola River, central British Columbia. J. Fish. Res. Board Can. 25:757-794. Pitman, R. L., M. R. Graybill, J. Hodder, and D. H. Varoujean. In press. The catalog of Oregon seabird colonies. U.S. Dep. Fish Wildlife, USFWS FWS/OBS. Quinn, W. H., D. O. Zopf, K. S. Short, and R. T. W. Kuo Yang. 1978. Historical trends and statistics of the Southern Oscilla- tion, El Nino, and Indonesian droughts. Fish. Bull., U.S. 76:663-678. 285 FISHERY BULLETIN: VOL. 84, NO. 2 Reed, R. K. Scott, J. M. 1983. Oceanic warming off the U.S. West Coast following the 1973. Resource allocation in four syntopic species of marine 1982 El Nino. Trop. Ocean-Atmos. Newsl. 22:10-12. diving birds. Ph.D. Thesis, Oregon State Univ., Corvallis, SCHREIBER, R. W., AND E. A. SCHREIBER. 107 p. 1984. Central Pacific seabirds and the El Nino Southern Zar, J. H. Oscillation: 1982 to 1983 perspectives. Science 225:713- 1974. Biostatistical analysis. Prentice-Hall, Englewood 716. Cliffs, N.J., 620 p. 286 DEVELOPMENT AND EVALUATION OF METHODOLOGIES FOR ASSESSING AND MONITORING THE ABUNDANCE OF WIDOW ROCKFISH, SEBASTES ENTOMELAS Mark E. Wilkins1 ABSTRACT Rapid expansion of a new fishery for widow rockfish, Sebastes entomelas, stocks off the Pacific coast of the United States began in 1979. Within 3 years, landings rose from <1,000 t to almost 30,000 t of a species for which little information on abundance or life history was available. It was known that widow rockfish occurred in irregularly distributed, dense, midwater, and semidemersal schools primarily during the night, which posed problems in directly assessing this resource Therefore, a project was designed to further investigate the habits and distribution of the species and develop an adequate assessment methodology. Line transect survey methods, using sector scanning sonar to estimate the number of schools per unit area and standard hydroacoustic echo integration techniques to estimate school biomass, were used in study areas off Washington and Oregon. The applicability of this methodology will depend on our abil- ity to resolve technical problems and minimize the effects of distributional variability by refining survey design. The need for more sophisticated sonar equipment to improve data collection and processing, the extreme temporal and spatial variability of widow rockfish school size and location, and the difficulty of identifying the species composition of observed schools are matters of special concern. The rockfish (genus Sebastes) of the Pacific Ocean are comprised of over 65 species exhibiting a wide array of colors, sizes, body forms, behavior, and life history characteristics. Members of this family are generally demersal or semidemersal and school over hard substrate on the continental shelf and slope. The widow rockfish, Sebastes entomelas, is atypical. As an adult it aggregates in dense midwater schools during the night.2 These schools tend to disappear from established fishing grounds at dawn or shortly thereafter, becoming less vulnerable to the fishery. The role of this species in the Pacific coast ground- fish fishery changed from an undesirable incidental catch in 1978 to a major target species by 1980. Ad- vances in fishing technology and product handling and marketing, as well as new vessels seeking alter- native fisheries, promoted an increase in landings from 1,107 t in 1978 to 28,419 t in 1981 (Table 1). By 1981, schools were becoming more difficult to locate and there was concern that the resource was being overharvested. The fishery began expanding into new areas to maintain profitable catch rates. During late 1981 and early 1982, most of the widow Northwest and Alaska Fisheries Center Seattle Laboratory, Na- tional Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Building 4, BIN C15700, Seattle, WA 98115. 2Groundfish Management Team. 1981. Status of the widow rockfish fishery. Unpubl. manuscr., 41 p. Pacific Fishery Manage- ment Council, 526 S.W Mill Street, Portland, OR 97201. rockfish were being taken from the vicinities of Bodega Bay and Monterey, CA, though fishing was taking place as far north as Cape Flattery, WA. The rapid growth of this new fishery resulted in large catches from a resource about which little was known. Research on this species prior to 1979 was limited to general descriptions of distribution, habitat, and biological characteristics (Hitz 1962; Phillips 1964; Pereyra et al. 1969). Scientists began gathering data in 1978 to determine the impact of the fishery on the condition of the stock, to define the distribution and size of the stock, and to establish a baseline of biological characteristics of the species. Commercial landings have been sampled by State Table 1.— Landings of widow rockfish by state for years 1973-83 in metric tons. Year Washington Oregon California Total 1973 81 15 29 125 1974 18 7 47 72 1975 13 11 57 81 1976 51 55 147 253 1977 277 34 267 578 1978 428 472 207 1,107 1979 1,697 1,960 636 4,293 1980 6,632 8,718 4,808 120,677 1981 7,211 14,689 6,519 28,419 1982 6,030 9,262 10,270 25,562 1983 3,293 3,151 3,455 9,899 This also included 519 1 of joint venture and foreign catch. Manuscript accepted July 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 287 FISHERY BULLETIN: VOL. 84, NO. 2 and Federal agencies in Washington, Oregon, and California for information on size and age composi- tion, sex ratio, maturity, feeding habits, morpho- metries, meristics, and fecundity. Widow rockfish abundance was estimated by the Groundfish Management Team (fn. 2, 19823) of the Pacific Fisheries Management Council, using cohort and stock reduction analyses (SRA) (Kimura and Tagart 1982). These stocks were found to have been fished down from their virgin level and were thought to be approaching a biomass level which would, under prudent management, produce a maximum sustainable yield of about 12,000 t in the INPFC (In- ternational North Pacific Fisheries Commission) Col- umbia and Eureka areas. Research surveys were needed to complement these analyses by providing independent estimates of abundance, describing the distribution, and col- lecting biological information not available from fishery data (for example, data on prerecruits and fish in areas which will not support a profitable fishery). Widow rockfish present special problems to those seeking to estimate their abundance through research surveys. The species is not usual- ly available to bottom trawls, precluding traditional "area-swept" trawl surveys, and its tightly clustered distribution and inconsistent schooling behavior reduce the effectiveness of traditional hydroacous- tic surveys. In 1980, the Northwest and Alaska Fisheries Center (NWAFC) began developing a practicable method to survey widow rockfish stocks. Scientists needed to understand the distribution and behavior of widow rockfish to determine which survey methods might be most appropriate to measure the size of the resource The first objective of the project, therefore, was to study aspects of the behavior, distribution, and biology of the species. The distri- bution of its characteristic nighttime aggregations relative to features of submarine topography was of particular interest. The distribution of this species is highly variable both on a diel basis and over longer periods, and the reasons for this variability were also of interest. Another question concerned what pro- portion of the total resource is present in detectable schools and how that proportion changes in space and time Clark and Mangel (1979) described a theoretical situation in yellowfin tuna stock dynamics wherein detectable, fishable schools are constantly being replenished from an undetectable portion of 3Groundfish Management Team. 1982. Status of the widow rockfish fishery. Unpubl. manuscr., 22 p. Pacific Fishery Manage- ment Council, 526 S.W. Mill Street, Portland, OR 97201. the population. They discussed the implications of this behavior in a fishery. If such a phenomenon could be confirmed in widow rockfish, determining the detectable proportion of the population might enable us to estimate the absolute size of the resource The second objective of the project was to inves- tigate methodologies with potential for estimating widow rockfish stock size, considering the species' behavior and distribution patterns. The final objec- tive was to evaluate the effectiveness of the chosen technique when actually implemented. The project was conducted in three phases: 1) an examination of the biology and behavior of widow rockfish on commercial fishing grounds, 2) the development of a practical survey method for assess- ing distribution and abundance, and 3) an evaluation of the feasibility and effectiveness of applying such assessment methodology to widow rockfish on a routine coastwide monitoring basis. Field studies were initiated in March 1980 and concluded in April 1982. Behavior studies were conducted during August 1980 and April 1981. Field work focusing on methodology development took place during late March 1980 and mid-March 1981, and the trial assessment survey took place during mid-March to early April 1982. All field work was conducted off Oregon and southern Washington (Fig. 1). The purpose of this report is to document the work done to date on the development of widow rockfish assessment methodologies, to evaluate the utility of those methods for routine assessment and monitor- ing of widow rockfish stocks and other species ex- hibiting a similar behavior, and to recommend means of enhancing future assessment efforts. BEHAVIOR STUDIES (1980-81) The nature of the fishery made it apparent that the behavior of widow rockfish differed from that of other commercially important species of the genus Sebastes. Extremely large widow rockfish catches were taken by midwater trawlers operating almost exclusively at night and fishing on very dense mid- water schools in only a few small areas along the coast. The first phase of the project studied the behavior and habits of widow rockfish to determine their distribution patterns, using demersal and midwater trawls and hydroacoustic observations. This included determining where the fish go when the dense, mid- water schools disperse; whether there are compo- nents of the stock other than the typical midwater aggregations; and at what period in their daily cycle 288 WILKINS: ABUNDANCE OF WIDOW ROCKFISH Nelson Islanc Halibut Hill The Fingers- Heceta Bank Cape Blanco - 45° 00' 47° 00' N 46° 00' 44° 00' - 43° 00' 126° 00'W 125° 00' 124° 00' 123° 00' Figure 1.— Widow rockfish survey areas off the coasts of Washington and Oregon occupied during field work conducted between 1980 and 1982. 289 FISHERY BULLETIN: VOL. 84, NO. 2 their availability is most stable. Other objectives were to investigate the possible causes of their diel aggregation habits and to develop an ability to distinguish widow rockfish schools from those of other species on the basis of echosign4 characteris- tics and test fishing. Methods The behavior study was initiated 11-13 August 1980 aboard the chartered trawlers Pat San Marie and Mary Lou. Concurrently, scientists aboard the NOAA RV Miller Freeman conducted a conventional echo integration survey in the study area and made four midwater tows to identify the species composi- tion of the schools sighted. The survey was repeated during 10-26 April 1981 aboard the NOAA RV Chap- man and included 7 d of hydroacoustic and sonar observations.6 Descriptions of the vessels, trawls, and hydroacoustic equipment employed appear in Tables 2, 3, and 4, respectively. Demersal trawl stations were located around a seabed rise known as Nelson Island off Newport, OR, to determine if significant quantities of widow rock- fish occurred on or near the bottom in an area where they were known to form dense midwater aggrega- tions. A 4 x 4 station grid with interstation distances of 4.6 km (Fig. 2) was established between the depths of 110 and 360 m with the rise at the center. Two trawl hauls were attempted at each sta- tion: one during daylight and one during darkness. When significant midwater fish schools were ob- served, they were sampled with midwater trawl gear for species composition. The contents of each trawl haul were sorted by species, weighed, counted, and recorded. Otoliths were removed from samples selected for age deter- mination and stage of maturity was recorded for some individuals. Stomach sample collections, stratified by fish length, were also taken and pre- served for feeding studies.6 No meaningful descrip- tion of age and length composition was possible because of the small catches. Consultations with fishermen, observation trips aboard commercial trawlers, and observations dur- ing research operations provided further informa- tion about school characteristics and diel behavior patterns of widow rockfish and other species on and around widow rockfish fishing grounds. Results Twenty-seven demersal tows were completed dur- ing the August 1980 widow rockfish behavior study, including 12 at night and 15 during the day. The trawl was damaged during two night hauls. The wi- dow rockfish catch was small, with 1 or 2 specimens in six hauls and 20 specimens in one of the night hauls during which the trawl was damaged (Fig. 3, 1980). Therefore, no conclusions about diel move- ment patterns were possible from the 1980 study. The Miller Freeman transected the Nelson Island area during the same study period and found one 4"Echosign" can be defined as the echo return output (paper echo- grams, video chromoscope displays, etc) of an echo sounder aimed at targets in the water column. 6Thomas, G. L., C. Rose, and D. R. Gunderson. 1981. Rockfish investigations off the Oregon coast, annual report. Unpubl. manuscr., 20 p. Univ. Wash., Fish. Res. Inst, FRI-UW-8119. 6Adams, P. B. 1984. The diet of widow rockfish (Sebastes en- tomelas) in northern California. Unpubl. manuscr. Southwest Fisheries Center Tiburon Laboratory, National Marine Fisheries Service, NOAA, 3150 Paradise Drive, Tiburon, CA 94920. Table 2.- -Vessels used during the widow rockfish assessment project. Main Length engine Survey Vessel (m) (hp) type Agency1 Dates Muir Milach 26 800 Hydroacoustic sonar FRI 19 Mar.-2 Apr. 1980 Pat San Marie 31 765 Behavior NWAFC 11-13 Aug. 1980 Mary Lou 26 700 Behavior NWAFC 11-13 Aug. 1980 Miller Freeman 66 2,200 Behavior and hydroacoustic NWAFC 11-13 Aug. 1980 Alaska 30 600 Hydroacoustic sonar FRI 12-23 Mar. 1981 Chapman 39 1,165 Behavior and hydroacoustic sonar NWAFC 7-26 Apr. 1981 Ocean Leader 36.5 1,125 Hydroacoustic sonar NWAFC 14 Mar.-7 Apr. 1982 1FRI = Fishery Research Institute; NWAFC = Northwest and Alaska Fisheries Center. 290 WILKINS: ABUNDANCE OF WIDOW ROCKFISH Table 3.— Fishing gear used during the widow rockfish assessment project. Trawl type Vessels Doors and accessory gear Approximate fishing dimensions Bottom trawl Nor'eastern Midwater trawl Alaska Diamond Norsenet No. 7 Gourock rope wing No. 8 Gourock rope wing Pat San Marie 1.5 x 2.1 m steel V-doors, 55 m triple and Mary Lou dandylines, 32 mm mesh cod end liner, roller gear Muir Milach Same as above but with 1.8 x 2.7 m and Chapman steel V-doors 2,500 lb Alaska Same as above but with 1.6 x 2.9 m aluminum V-doors Chapman 1.8 x 2.7 m steel V-doors, 55 m dou- ble dandylines with 4 sets of 5.5 m bridles, 125 kg weights attached to the bottom of each wingtip, 32 mm mesh cod end liner Alaska Same as above but with 1.6 x 2.9 m aluminum V-doors Miller Freeman 6 m2 Waco doors, 75 m double dandy- lines, 46 mm mesh cod end covered with a double braided 144 mm mesh bag Muir Milach 4.6 m2 Suberkrub doors, 73.2 m dou- ble dandylines, 114 mm mesh cod end (no liner) Ocean Leader 4.5 m2 Suberkrub doors, 100 m dandy- lines 200 kg weights attached to the bottom of each wing, 32 mm mesh cod end liner 9.1 m headrope height, 13.4 m wingspread 6.10 m headrope height, 16.7 m wingspread (Wathne1) (not measured) 11.0-14.6 m vertical opening 15.2 m wingspread Same as above 18-20 m vertical opening 18.3 m vertical opening, wingspread not measured 21.3 m vertical opening, wingspread not measured 'Wathne, R, Northwest and Alaska Fisheries Center, 2725 Montlake Blvd. E., Seattle, WA 98115, pers. commun. June 1981. Table 4.— Hydroacoustic equipment used during widow rockfish behavior and assessment surveys, 1980-82. Institute; NWAFC = Northwest and Alaska Fisheries Center. FRI = Fisheries Research Vessel: Muir Milach Miller Freeman Alaska Chapman Ocean Leader (FRI) (NWAFC) (FRI) (NWAFC) (NWAFC) Dates used 19 March- 2 April 1980 11-13 August 1980 12-23 March 1981 21-26 April 1981 14 March- 7 April 1982 Echo sounder and Simrad1 EK-38 Simrad EK-38 Simrad EK-38 Simrad EK-38 Biosonics 101 transducer 11° beam at -3dB 12° beam at -3dB 11° beam at -3dB 11° beam at -3dB 7° beam at -3dB Towed V-fin 2-ft Braincon 2-ft Braincon 2-ft Braincon 2-ft Braincon 2-ft Braincon transducer housing Tape recorder TEAC 3440A cassette TEAC 3440A TEAC 3440A cassette reel-to-reel reel-to-reel reel-to-reel Chart recorder Simrad wet paper Simrad dry paper Simrad wet paper Simrad wet paper EPC 1600 dry paper Portable echo Biosonics 120 NWAFC acoustic Biosonics 120 Biosonics 120 Biosonics 120 integrator research container system Computer Not used NWAFC acoustic research container system Not used Not used Radio Shack TRS-80 Sonar system C-Tech LSS-68 Not used C-Tech LSS-68 Simrad SQ Furuno FSS-75 68 kHz sector 68 kHz sector searchlight beam 75 kHz sector scanning scanning scanning Video camera and RCA C004 camera Not used RCA C004 camera RCA C004 camera RCA C004 camera recorder Panasonic recorder Panasonic recorder Panasonic recorder Panasonic recorder 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 291 FISHERY BULLETIN: VOL. 84, NO. 2 44° 50' N - 44° 40' 44° 30' 124° 50' W 124° 40' Figure 2.— The demersal trawl station grid occupied during 1980 and 1981 widow rockfish behavior studies on the Nelson Island ground off Newport, OR. The 16 trawl stations are marked with a (+). school of widow rockfish, which was sampled with midwater trawl gear (Fig. 4). It was not possible to stay in contact with the school long enough to observe diel changes in behavior. When the study was repeated in April 1981, only 4 of 20 demersal tows contained widow rockfish. Two of these tows contained only a single specimen each, while the others contained 20 and 28 specimens. Results again indicated that widow rockfish were relatively unavailable to demersal trawl gear and that their distribution was somewhat more closely associated with Nelson Island during the night than during the day (Fig. 3, 1981). It is important to be able to distinguish widow rockfish from other species on the basis of echosign in order to draw conclusions about their behavior, distribution, and abundance Commercial fishermen targeting on this species have shown that this can be done. We characterized the echosign produced by widow rockfish and other species occurring on widow rockfish grounds using echograms obtained aboard research and commercial vessels and through discus- sions with commercial fishermen on echograms and corresponding catches. Widow rockfish schools most frequently appeared on echograms as tall, slender columns suspended over an irregular bottom (Fig. 5). These were often accompanied by less dense layers probably composed of salps and other zoo- plankton. Widow rockfish were sometimes present during evening and morning in smaller schools high in the water column (Fig. 6). Shortbelly rockfish, Sebastes jordani, and redstripe rockfish, S. proriger, have similar echosign characteristics and are most likely to be confused with widow rockfish off the Oregon coast (Figs. 7, 8). Other midwater targets in the area were identified as layered schools of 292 WILKINS: ABUNDANCE OF WIDOW ROCKFISH I 44° 50' N Figure 3.— Locations of demersal tows which contained widow rockfish during day (D) and night (N) sampling during the 1980 and 1981 behavior studies. 44[> 50' N 44° 40' 44° 30' 124° 50' W 124° 40' 293 FISHERY BULLETIN: VOL. 84, NO. 2 44° 50' N - 44° 40' 44° 30' 124° 50' W 124° 40' Figure 4.— Hydroacoustic transects (dashed lines) and midwater trawl hauls (solid ar- rows) conducted by the RV Miller Freeman during the 1980 behavior study. Only haul 43 contained widow rockfish (1,247 kg). Pacific whiting, Merluccius productus, (Fig. 8) or less dense layers of zooplankton. The formation and dispersal of widow rockfish ag- gregations was observed during the research cruises. During a typical night, small schools would appear in late evening (from 2000 to 2400) either near bot- tom or high in the water column. As the night pro- gressed, these schools tended to grow and those high in the water would settle toward the bottom. Peak school size and density usually occurred between 0200 and dawn. Shortly after daybreak, most schools would separate into smaller schools and rise off the bottom. The schools would sometimes move over deeper water while maintaining their nighttime configuration. Departures from the typical behavior patterns have been reported. For example, while observing widow rockfish schools over the continental shelf (not aggregating around a seamount), Gunderson et al.7 noted a progressive offshore shift in the location of the schools during one night. By sunrise most of the schools were located near the edge of the shelf. Most of these schools dispersed after dawn, but some re- mained on the bottom in the area (in one case as late as 1037 when observations were terminated). This apparent shift may have been related to diurnal ver- tical migration behavior (Pereyra et al. 1969). METHODOLOGY DEVELOPMENT (1980-81) The methodology development was conducted by 7Gunderson, D. R., G. L. Thomas, P. Cullenberg, and R. E. Thome 1981. Rockfish investigations off the coast of Washing- ton and Oregon. Final report. Unpubl. manuscr., 45 p. Univ. Wash., Fish. Res. Inst, FRI-UW-8125. 294 WILKINS: ABUNDANCE OF WIDOW ROCKFISH . > ■ •*'■ , 14 Figure 5— Echogram showing the typical configuration of widow rockfish schools at night (arrows). Figure 6 -Echogram showing the configuration of "evening and morning" widow rockfish schools (arrows). 295 FISHERY BULLETIN: VOL. 84, NO. 2 Figure 7— Echogram showing the typical configuration of shortbelly rockfish schools (arrows). Figure 8— Echogram showing configuration of Pacific whiting (W), redstripe rockfjsh (R), and shortbelly rockfish (S) schools. 296 WILKINS: ABUNDANCE OF WIDOW ROCKFISH the University of Washington's Fishery Research In- stitute (FRI) under contract with the NWAFC (Gunderson et al. fn. 7, 8). The objectives of the work were to evaluate the applicability of several resource assessment techniques and refine the most prom- ising approaches. In particular, it involved a compar- ison of three methods of quantifying widow rockfish abundance in small areas off southern Washington and northern Oregon: conventional echo integration, line transect survey theory (Burnham et al. 1980; Seber 1980), and line intercept survey theory (Seber 1973, 1980). Methods This study involved three research cruises off southern Washington and northern Oregon. Tables 2-4 present the dates of these cruises and specifica- 8Gunderson, D. R., G. L. Thomas, P. Cullenberg, D. M. Eggers, and R. E. Thorna 1980. Rockfish investigations off the coast of Washington. Annual report. Unpubl. manuscr., 68 p. Univ. Wash., Fish. Res. Inst., FRI-UW-8021. tions of the vessels, fishing gear, and hydroacoustic equipment employed. The field work entailed system- atically transecting the survey areas, simultaneously recording data from quantitative echo integration equipment and sector scanning sonar. Data were col- lected on the number of fish schools, their perpen- dicular distance from the transect, their depth below sea surface, the size and density of selected schools, and the distribution of schools in relation to various features of submarine topography. The echo integra- tion system was used in a conventional manner to obtain a measure of the density of fish within a relatively narrow acoustic beam of 10°-11° directly below the vessel (Fig. 9). Sector scanning sonar can- not measure fish density, but by employing an ar- ray of transducers radiating an acoustic signal over a 200° x 9° semicircular wedge perpendicular to the path of the vessel (Fig. 9), it can be used to count schools within about 100-200 m to each side of the vessel, measure their dimensions, and determine their perpendicular distance from the transect. The sonar's transducer array was aimed straight down- ward for these studies. The entire wedge was ECHOSOUNDER SONAR s . 1 . - Editing of data to include only schools likely to be widow rockfish. Jl"""'\—>-ll * Within-school fish density Schools that were also detected on echo integration system Mean school biomass estimate (metric tons/school) » School height * School width * School length ' ' Biomass estimate (t) Area of survey area (km2 ) i Number of schools sighted with sonar School density estimate (schools/km2 ) Distance between school center and transect plane Figure 9— Schematic diagram depicting the analysis of echo sounder and sonar data collected during hydroacoustic line transect surveys. 297 FISHERY BULLETIN: VOL. 84, NO. 2 displayed simultaneously on a 10-in diameter cathode ray tube (CRT) screen which provided information on the location and size of fish schools within its 200-400 m wide path (Fig. 10). Data was collected electronically during the echo integration and sonar surveys. Echo sounder return signals were processed by an echo integrator capable of measuring voltages in variable-sized depth inter- vals. The echo integrator produced periodic printouts of summed integrated voltage values which corre- sponded to relative fish densities along the transect in various depth intervals. Analog data (receiver out- put voltages) were recorded onto magnetic tape as a back-up procedure and for further processing. The sonar CRT display screen was video-taped for play- back and data reduction in the NWAFC laboratory. Survey design of the 1980 and 1981 FRI studies off southern Washington was generally similar, though some aspects differed. In 1980, preselected tracklines were run and were between lat. 46°20'N and 46°48'N and between 55 and 183 m isobaths at intervals of of 3.7 km. When a significant aggrega- tion of fish was encountered, its bounds were deter- mined by making several mapping runs perpen- dicular to the main trackline. Trawling followed to determine the species composition of the aggrega- tion and to collect biological samples. Most of the 1980 work was done during daylight with the intent of mapping and measuring yellowtail rockfish, Sebastes flavidus, schools. After encountering numerous widow rockfish schools at night, it became apparent that this species' schooling behavior was better suited for evaluating this methodology. Thereafter, three nights were spent transecting a smaller "widow rockfish subarea". Diel behavior and distribution were examined by making several repeti- tions of three selected tracklines. Near the end of this cruise an area occupied by a dense aggregation of widow rockfish schools was encountered. A short nonrandom transect was run to obtain comparable line intercept and line transect results. In 1981, tracklines spaced every 3.7 km were transected between the depths of 128 and 220 m off northern Oregon between lat. 45°50'N and 46°18'N. Figure 10.— Measurements and calculated dimensions of fish schools from videotaped sonar records. 298 WILKINS: ABUNDANCE OF WIDOW ROCKFISH The same procedures were used as in 1980 except that nearly all operations were conducted after dark and no mapping runs were made to define the bounds of school groups. A diel variability study was conducted on 20 March 1981 between the hours of 0153 and 1037, consisting of 13 replicates of track- line 21. The conventional analysis of echo sounder data (in- tegration) is based on the principle that the acoustic intensity of a signal reflected from fish targets is pro- portional to the density of fish in the region irradi- ated by the echo sounder. Detailed descriptions of the technique can be found in Moose and Ehrenberg (1971), Forbes and Naaken (1972), and Thorne (1977). During the 1980 and 1981 surveys, density estimates from this method were obtained by aver- aging returning acoustic signals over a series of transmissions (25 transmissions over 12.5 s during the Muir Milach cruise and 40 transmissions over 50 s during the Alaska and Chapman cruises). These averages were then converted from relative to ab- solute densities (kg/m2) for various depth intervals using calibration data and a scaling factor based on an average target strength of -35 dB/kg.9 Absolute abundance (biomass) was estimated by extrapolating absolute density estimates to the survey area. Each survey area was systematically transected using the echo sounder and sonar to search for fish schools and, thereby, to derive line intercept and line transect estimates of school abundance (schools/ km2). Data on school dimensions and density were collected from those schools sighted. With the line intercept method, only the presence of a school (as detected by the echo sounder) and its width were used to estimate school abundance. This technique is based on the theory that, for systematically located transects, the probability of intersecting school i equals wJW, where w^ is the width of school i and W is the distance between adjacent transects. The number of schools per unit area {D) can then be estimated by 0-1 y-i wjL (Seber 1980) where n Wi number of schools measured on a transect of length L width of jth school. 9The target strength value used in these analyses (-35 db/kg) was not derived during work on widow rockfish. Since accurate target strength estimation was not necessary for evaluating the utility of the methodology, we used a value which had been esti- mated for Pacific whiting (Dark et al. 1980) which has a similar scattering cross section. The line intercept method was applied only to data collected from the nonrandom run made on the night of 27-28 March 1980. The data from this line were subdivided into two artificial transects of unequal length and the jackknife method (Seber 1980) was used to estimate D and its variance. This technique is described fully by Gunderson et al. (fn. 8). Line transect theory is based on the premise that the probability of sighting a given object (or school) is a function of its perpendicular distance from the transect. A "detection function" is derived from school sighting data which relates the probability of a school being sighted to its distance from the transect. This function is used to expand the number of schools actually sighted to obtain an estimate of school abundance. The advantage of this method is that not all schools within sighting range need to be detected in order to estimate the number of schools in the area. Using line transect estimation, the school abun- dance (schools per unit area) was estimated by D = nf(0) 2L n L /(0) where D = estimated number of schools per unit area number of schools sighted length of transect "detection function'— a parameter estimated from probability function for the perpendicular distances off transect of schools sighted. The assumptions necessary for the use of this method are 1) Schools directly on the transect plane will always be sighted. 2) Schools are sighted in the position they occupied prior to the approach of the vessel, i.e. there is no avoidance of or attraction to the vessel. 3) Perpendicular distances off transect are mea- sured precisely, particularly near the transect plane 4) The detection function remains constant. The computer program TRANSECT (Laake et al. 1979) was used to estimate the probability density function of the perpendicular distance of schools from the transect. The estimator model used is based on a nonparametric Fourier series expansion fit to data sets of observed perpendicular distances of 299 FISHERY BULLETIN: VOL. 84, NO. 2 schools off the transect plane Quinn (1979) and Burnham et al. (1980) showed that this model is robust and flexible and provides the best fit to the detection function in most applications. This esti- mator, at zero distance, is 1(0) = ^ + £ w k=\ % where w* = truncation width, or the effective limit of the range of detection, beyond which all observations are discarded and <*fc nw Z. COS 1=1 knX; w (Burnham et al. 1980) where n = number of schools observed x{ = perpendicular distance off transect for the ith school k = term number = 1,2,3,. . .m [The num- ber of terms (m) is determined by a stopping rule in the computer pro- gram TRANSECT]. TRANSECT also computes the school abundance estimate D and its variance which is estimated by the equation: var 0) = (Df var (n) var [/"(0)] nc Lf(0)]2 Dimensions measured directly included depth of school from the surface, distance off bottom, width, thickness, radial distance from vessel, and bearing to the right or left of a vertical line below the vessel (Fig. 10). The perpendicular distance of the school from the vertical plane of the vessel's path ("distance off transect") was calculated from the radial distance and bearing. All distances were measured or cal- culated to the apparent geometric center of each school (Burnham et al. 1980). The length of each school was calculated from the product of vessel speed and the duration that the school was being detected by the sonar, and was corrected to account for the variable sonar beam width parallel to the vessel's path due to depth. The biomass of individual schools was estimated by the formula h% = ti l{ w% di where bt = estimated biomass of school i t{ = average thickness of school i, top to bottom (echo sounder data) l{ = length of school i, parallel to transect (sonar data) w{ = average width of school i perpendicular to transect plane (sonar data) di = mean integration density for school i (g/m3) assuming a target strength of -35 dB/kg (see footnote 9) (echo sounder data). The mean school biomass (MSB) was estimated from the individual school biomass estimates; its variance was determined from Mean school biomass estimates were derived from density information (from echo sounder data), school dimension information (from sonar data), and an assumed target strength of -35 dB/kg. These esti- mates were used in the line transect and line inter- cept analyses. All information on schools detected by the hydroacoustic systems was edited to discrim- inate widow rockfish from other species using judg- ments based on school form, density, location, and test trawl records. Data on each school identified as widow rockfish were then integrated to obtain mean within-school density. The CRT display of the sec- tor scanning sonar provided representations of the size, shape, and position of fish schools within its range of detection. The dimensions of all schools identified as widow rockfish were measured on the screen of a video monitor using the slow motion and freezeframe features of the video recorder-player. " (b, - MSB)2 var (MSB) = 2. — v ; i-i N(N-1) where N = number of schools averaged for MSB. Total biomass estimates from the line transect and line intercept methods were calculated for each survey area using the formula B = AD (MSB) where B = estimated total biomass for the survey area, and A = total area (km2) of the survey area. The variance of these estimates was determined from 300 WILKINS: ABUNDANCE OF WIDOW ROCKFISH var B = A2 [(D)2 var(MSB) + (MSB)2 var0) - var (MSB) var 0)] (Goodman) 1960) Results Twenty one trawl hauls were made during the 1980 FRI survey aboard the Muir Milach; 6 with bottom gear and 15 with midwater gear. Widow rockfish were caught only in midwater hauls and comprised 99% of those catches. The most abundant species in the bottom tows were spiny dogfish, Squalus acan- thias, and black rockfish, Sebastes melanops. The acoustic survey consisted of 22 systematic transects covering about 550 km and employed sonar and echo integration equipment. Twenty six schools were sighted and measured to provide data for a line transect estimate of school abundance. During the nonrandom transect run on the night of 27-28 March 1980, 73 schools were sighted and measured for use in developing line transect and line intercept esti- mates of school abundance in a small subarea. Only four trawl hauls were attempted during the 1981 FRI survey due to severe gear damage. Red- stripe rockfish, Sebastes proriger, comprised 90% or more of the two catches which contained fish (one midwater haul and one bottom haul). The midwater haul was made quite close to bottom near midnight and contained small quantities of sharpchin rockfish, Sebastes zacentrus; widow rockfish; and greenstriped rockfish, S. elongatus, suggesting an association of these species in nearbottom schools at night. Fifteen systematic transects were covered during this survey (about 400 km) during which 49 schools were sighted and measured. One of the transects was replicated 13 times during one night to observe the behavior of a group of schools over the continental shelf just south of the Columbia River. These schools were not gathered around a prominent bottom feature As the night progressed they moved deeper and further off- shore, reaching the shelf break about sunrise After sunrise most of the schools dispersed, though some remained on bottom at least until observations ceased at 1037 (Gunderson et al. fn. 7). During the 1981 NMFS cruise, quantitative hydro- acoustic data were collected from 21 transects on the Nelson Island, The Fingers, Heceta Bank, and Cape Blanco grounds (Fig. 1, Table 5) using echo integration (Thomas et al. fn. 5). The searchlight- beam sonar available on the Chapman was inade- quate to identify school types or provide estimates of school density. This is because it employed only a single transducer programmed to sweep back and forth and did not provide continuous coverage of the area within its range Therefore, all density and biomass figures for this survey refer to total nekton rather than widow rockfish. Table 5.— The mean fish and nekton density (g/m2) and biomass (metric tons) by location, date, and transect estimated by a conventional echo integration survey performed aboard the NOAA RV Chap- man, 21-26 April 1981 .1 Transect Mean Trans- length density Var Area Biomass Var Location Date sect (km) Density D D (km2) B B Crater 4/21 1 18.56 1.78 4/22 2 17.11 5.86 2.82 2.45 228.87 646 1.28 x 10s 4/22 3 15.02 0.66 Cape Blanco 4/23 4/23 4/23 4 5 6 4.19 4.35 4.67 16.17 26.22 4.50 4/23 7 5.32 1.12 6.47 11.31 200.81 1,301 4.56 x 10s 4/23 8 6.11 1.79 4/24 9 2.44 1.63 4/24 10 3.87 0.34 4/24 11 3.87 0.06 Heceta Bank 4/24 12 17.59 4.67 4/25 13 18.37 0.12 4.89 9.44 87.15 427 7.17 x 104 4/25 14 15.30 10.88 The Fingers 4/25 15 14.74 1.78 4/25 16 14.74 1.01 1.73 0.19 75.11 130 1.07 x 103 4/25 17 12.22 2.54 Crater 4/26 18 5.57 0.00 4/26 19 6.96 0.00 1.79 2.83 34.78 62 3.43 x 103 4/26 20 6.28 6.75 4/26 21 5.63 0.26 1Thomas, G. L, C. Rose, and D. R. Gunderson. 1981. Rockfish investigations off the Oregon coast, annual report. Unpubl. manuscr, 20 p. Univ. Wash., Fish. Res. Inst. FRI-UW-8119. 301 FISHERY BULLETIN: VOL. 84, NO. 2 The results of echo integration, line intercept, and line transect analyses were compared using data col- lected during the 1980 and 1981 FRI cruises (Gunderson et al. fn. 7, 8). Large differences were seen between echo integration and line transect estimates in a situation where schools were relatively small and scarce (1980 transect data, Table 6). The principal reason for this is that the threshold echo voltage required to trigger the sonar CRT display was higher than that needed to detect a school on the echo integration system, so many of the sparser schools detected by the echo sounder were not detected with the sonar. In situations where schools were larger and more plentiful (1980 nonrandom runs and 1981 transects) all three methods produced similar estimates. The precision of abundance estimates generated by line transect and line inter- cept methods is usually comparable to that of con- ventional echo integration methods and can exceed it in some cases (Gunderson et al. fn. 7). The major factors which led us to concentrate our efforts on line transect surveys were the ability to cover large areas rapidly and the ability to expand the number of schools sighted by a detection function, yielding more accurate estimates of school abundance. APPLICATION OF ASSESSMENT METHODOLOGY By 1982 the aforementioned studies had provided a foundation of information on which to expand developmental research. The behavioral observations suggested that widow rockfish aggregations were most stable and susceptible to assessment during the night. Line transect estimation of school abundance through the use of sonar and echo integration equip- ment was found to be the most effective of the tech- niques compared, especially when school abundance was likely to be low. The next step in the project was to evaluate the feasibility of applying the line transect survey method in a comprehensive survey to assess and monitor widow rockfish stocks. Methods The trawler Ocean Leader was chartered to survey five areas off Oregon (Fig. 1) where widow rockfish had been caught consistently between 1980 and 1982. Specifications of the vessel, fishing gear, and hydroacoustic equipment used appear in Tables 2-4. The proximity of alternative grounds was important for the success of the survey, should widow rockfish not be found in one or more of the areas. At each of the grounds the survey procedure was as follows: 1) The ground was systematically surveyed with hydroacoustic equipment during the night to determine whether fish schools were in the area. The locations of schools suspected to be composed of widow rockfish, or species likely to be confused with widow rockfish, were noted. The final bound- aries of the study area were then delineated. 2) The study area was surveyed at night along parallel tracklines about 1 km apart using the line transect survey technique The tracklines were replicated as many times as practical throughout Table 6.— Summary of estimates of school abundance (D), mean school biomass (MSB), and total biomass (6) for widow rockfish. Coefficients of variation (CV) are given for each estimate.1 D (schools/ MSB No. Of B nm2) CV (t) schools CV (t) CV 1980 Transect data 26 schools, 2 transects Line transect estimate 69.5 0.84 0.12 15 0.20 204 0.84 Echo integration estimate 778 0.16 Nonrandom run data 73 schools, 1 transect Line transect estimate 242.2 0.19 0.85 16 0.50 5,003 0.53 Line intercept estimate 248.9 0.10 0.85 16 0.50 5,139 0.51 Echo integration estimate 6,453 — 1981 Transect data 29 schools, 3 transects Line transect estimate 12.1 0.24 0.62 27 0.33 342 0.40 Echo integration estimate 342 0.77 'Gunderson, D. R., G. L. Thomas, P. Cullenberg, and R. E. Thorne. 1981. Rockfish investigations off the coast of Washington and Oregon. Final report. Unpubl. manuscr, 45 p. Univ. Wash., Fish. Res. Inst. FRI-UW-8125. 302 WILKINS: ABUNDANCE OF WIDOW ROCKFISH the night to provide information on variability of abundance and distribution within a given night. Selected study areas were again surveyed after an interlude of several days to study variability over longer periods. 3) Fish aggregations noted during transecting were sampled with midwater trawls for species iden- tification. This was done on alternate nights so as not to impede the progress of the acoustic assessment portion of the survey. Biological data (eg, size composition, maturity, stomach con- tents) were collected from widow rockfish in the catches. Results About 725 km of transects were covered in the five study areas during the 12 nights of hydroacoustic data collection. Ten midwater trawl hauls were made to identify species present in various schools. Widow rockfish schools were seen in all areas, but were sparse on the Cape Blanco, Heceta Bank, and The Fingers grounds. The Halibut Hill ground, only recently exploited, contained the highest density of widow rockfish schools and also the largest average school size After editing videotaped sonar records, 127 schools were identified as widow rockfish; data from 37 of these were integrated on the echo sounder system and used to calculate school biomass esti- mates. Ideally, a mean school biomass would have been derived for each ground, but because few schools were observed there, school biomass estimates were pooled and averaged for the Nelson Island, The Fingers, and Heceta Bank grounds. No measurable widow rockfish schools were seen dur- ing surveys of the Cape Blanco ground. School abun- dance was estimated for each area by treating each pass through the area as a replicate and pooling data from all replicates within the area. School abundance (excepting Cape Blanco) ranged from 0.6035 schools/ km2 on The Fingers ground to 1.4810 schools/km2 on the Halibut Hill ground. Area biomass estimates are summarized in Table 7. The total estimated biomass for the five survey areas was about 830 t; 50% at Halibut Hill ground, 30% at Heceta Bank, 11% at The Fingers, and 9% at Nelson Island. Sampling was concentrated on the Halibut Hill ground, where widow rockfish schools were largest and most plentiful, in order to investigate the diel and night-to-night variability in school abundance The survey of this ground was repeated seven times; three times each night on 26-27 March and 31 March-1 April and once on 30 March. Separate sighting functions for each night were estimated by pooling observations. Corresponding school abun- dance and mean school biomass estimates were then calculated for each night. School abundance ranged from 0.39 schools/km2 on 26-27 March to 4.50 schools/km2 on 30 March. Mean school biomass tended to decline as school abundance increased, however, so biomass estimates for each of the sam- pling periods changed less than either school abun- dance or mean school biomass (Table 8). It was not possible to analyze the Halibut Hill data on a replicate-by-replicate basis because few schools were sighted during any single replicate The number of sightings per replicate ranged from 4 to 34. Burn- ham et al. (1980) cautioned that such stratification procedures for line transect surveys should be "severely limited to those few surveys where the number of objects seen on replicate lines is fairly large (perhaps at least in the 20 to 30 range)". Table 7— Summary of estimates of school abundance (D), mean school biomass (MSB), and biomass (fl) in each of four study areas covered during the 1982 widow rockfish assessment feasibility survey. Study area , D schools km2 j Var0) MSB Area B school / Var(M$8 (km2) (t) Var(£) CV(6) [Var(g)]1/2 6 Heceta Bank 0.9490 0.0305 1.968 0.656 68.60 245.76 6,758.37 0.335 (6 schools) The Fingers 0.6035 0.0711 6.409 5.486 40.47 92.20 2,212.03 0.510 (4 schools) Nelson Island 0.7587 0.3010 3.924 9.514 27.44 78.59 3,467.75 0.749 (2 schools) 3 above areas pooled1 3.775 1.151 Halibut Hill 1.4810 0.2028 9.639 21.668 28.81411.27 51,758.47 0.553 (25 schools) 'School biomass data from Heceta, The Fingers, and Nelson Island were pooled to provide a mean school biomass which was used to calculate total biomass in each area. 303 FISHERY BULLETIN: VOL. 84, NO. 2 Table 8.— Summary of estimates of school abundance (D), mean school biomass (MSB), and biomass (8) of widow rockfish on the Halibut Hill ground during replicates on 26-27 March, 30 March, and 31 March-1 April 1982. Sampling period No. of repli- cates No. of schools sighted/ replicate schools | km2 / Var(D) MSB t Area Var(MSB) (km2) B (t) Var(8) CV(B) [Var(B)]1'2 school B 26-27 March 3 9 14 13 0.394 0.1264 21.674 153.878 28.81 MSB based on 8. schools) 245.96 52,852.69 0.935 30 March 1 34 4.499 0.8908 0.729 0.160 28.81 MSB based on 7 schools) 94.49 2,962.75 0.576 31 March- 3 9 1.325 0.1687 0.935 0.088 28.81 35.69 238.52 0.433 1 April 11 11 MSB based on 9 schools) Variations in the pattern of school abundance over the course of a night were common. Echograms recorded during the seven replicates of one transect on the Halibut Hill ground (Fig. 11) illustrate one case when abundance was high early in the night and decreased toward dawn (26-27 March). The op- posite trend of low abundance increasing toward dawn is illustrated (31 March-1 April) in the same figure. DISCUSSION The objectives of this 3-yr project were to study the schooling behavior of widow rockfish to provide the background needed to design effective abun- dance estimating surveys; then to develop an appro- priate survey methodology for the species; and, finally, to test the feasibility of implementing such a survey. Substantial progress was made toward satisfying these objectives. The studies of widow rockfish habits and distribution have provided a base for designing surveys which cover its range and pro- duce the best likelihood of encountering the ex- ploitable population at a time when it will be most available Understanding the schooling and dispersal be- havior of widow rockfish was important to develop an appropriate survey approach for estimating abun- dance The nighttime aggregations which are the targets of the commercial fishery tend to disperse about daybreak, perhaps scattering throughout the water column or seeking shelter near the bottom. If the latter had been the case, more conventional survey methods (i.e, bottom trawl or conventional echo integration surveys) might have been more appropriate Although daytime concentrations of widow rock- fish were observed, bottom trawl catches during the 1980 and 1981 surveys showed that this species is relatively unavailable to bottom trawls in an area where widow rockfish are known to aggregate at night.10 This is substantiated by low incidences of widow rockfish in catches of other bottom trawl surveys during periods when midwater trawlers were making large landings. Consequently, when mid- water schools disappear during the day, it is unlike- ly that they disperse along the bottom. In recent years, skippers of midwater trawlers have com- mented that widow rockfish are becoming more evasive and dive below their nets to avoid capture Some skippers have taken advantage of this behavior by purposely driving the schools toward bottom with engine noise where they capture them with bottom trawls equipped with roller gear. Although these are classified as bottom trawl landings, the fishermen are, in a sense, capturing midwater schools. Fisher- men have also reported encountering daytime aggre- gations of this species over the continental slope in waters deeper than they are usually found at night (>500 m) and some have been able to catch them on or near the bottom during the day. Thus the distribu- tion of widow rockfish relative to the sea bottom is quite unpredictable during the daytime These schools are also not as large as those that occur at night. The appropriate time to survey this resource thus appeared to be at night. The line transect survey method, adapted for use with sector scanning sonar and echo integration equipment, was chosen over conventional echo integration and the line intercept method because of its ability to survey areas more quickly and thoroughly. Application of the method exposed several problems affecting the precision and accuracy of the abundance estimates. The estimation of school abun- dance was hampered primarily by limitations of the sonar equipment and by small samples. We were not '"Observations of midwater echosign and landing information from commercial vessels fishing in the area confirmed that the usual dense midwater widow rockfish aggregations were present in the area at night during the 1981 bottom trawl survey. 304 WILKINS: ABUNDANCE OF WIDOW ROCKFISH K (J «i S ~ e) r- T^ If) o CM O CM O CM O E CO in CM CM CM CM CM a "el (V V > i- 0) « c - en T3 cd 0) TO Qt 0) O. C3 is o c CD CD be 2 bb® 2 g 3 « £ § m /. — -— fl « « & 8 '5b 3 c7M c c o c eg — c 3 1 g CO CM CM CM in CM "5 CM CM <--* Z.2 " 92 M3JB1AI oe iPje|AI L IMdv - L£ M3JB|A| 305 FISHERY BULLETIN: VOL. 84, NO. 2 able to calibrate the sonar systems so that the sensi- tivity of all transducers in the array were equal. Hence, the probability of detecting a given school in one sector of the sonar display was not necessarily the same as detecting it at an equal distance in another sector. The inability to calibrate the trans- ducers may have compromised our ability to detect all schools directly below the transect. This is the most important assumption of line transect surveys; school abundance estimates will be biased downward if it is violated. Intercalibration of the transducers would also help establish a more accurate detection function which would apply throughout the sonar's range The limited lateral resolution of sector scanning sonar hampers the accurate measurement of school width, an important value for determining mean school biomass. Each transducer in the fan-shaped array acts as an independent echo sounder and if any portion of a school enters the radiation pattern, the entire width of the 9° -10° sector sampled by that beam will be displayed as a reflective target (Fig. 12). This results in an overestimation of school width and a distortion of the school's size and location, yield- ing overestimates of biomass and inaccurate measures of distance from the transect plane The detection function will be altered by these inac- curacies and may modify estimates of school abun- dance depending on the magnitude and the direc- tion of the errors. The distortion may be aggravated by interference of side lobes in the directivity pat- tern of individual transducer beams (Fig. 13). Even these lower power lobes can produce echo signals if very dense targets are encountered and may inter- fere with the acoustic signals from adjacent trans- ducers. Another weakness of sector scanning sonar in this application is insufficient detection sensitivity. This weakness became apparent during calculations of the lengths of individual schools. Lengths were cal- culated twice for each school, once based on echo sounder data and again based on sonar data. The theoretically correct method would employ the sonar data because schools could be detected further to each side of the vessel. The echo sounder could only detect the portion of the school within the 10° -11° beam directly below the vessel. Consequently, if a large part of the school was outside the beam, its length was underestimated. In practice however, the length estimates based on sonar detections were usually shorter than those based on echo sounder data (Table 9) due to the lower sensitivity of the sonar system. The sonar-based lengths were chosen, how- ever, because they measured the dimensions of the part of the school having densities above the thres- hold required to trigger the sonar. This is probably Segment covered by one element in sonar transducer array True school center Apparent school center Apparent location of school as seen on the sonar display True location of school Figure 12— A facsimile of the sector scanning sonar output display exemplifying biases in apparent school loca- tions resulting from the limited resolution of the instrument. 306 WILKINS: ABUNDANCE OF WIDOW ROCKFISH •?0° 3dB ?°° Half-power point roc YOo Figure 13— The theoretical directivity pattern of one transducer element of the sector scanning sonar show- ing side lobes which may interfere with the signals received by adjacent transducers. a more proportional measurement of school length than the echo sounder-based lengths. The accuracy of school dimension measurements could be im- proved by using more sensitive and specialized sonar equipment. These problems with the limitations of sector scan- ning sonar should not be difficult to overcome More sensitive quantitative sonar equipment is now avail- able or relatively easy to develop. Lateral resolution may remain a problem because of the difficulty and expense of producing narrow-beam transducers, but the errors it causes are relatively unimportant. The accuracy of mean school biomass estimates would be improved by target strength studies specific to widow rockfish. Calculation of average fish den- sity within each school was relatively straight- forward but involved assuming a target strength of -35 dB/kg. Ideally, the target strength should be calculated specifically for widow rockfish but such specialized work was beyond the scope of this study. The ability to distinguish widow rockfish schools from those of other species using hydroacoustic equipment is an important element of this tech- nique Through these studies, our ability to correct- ly identify widow rockfish echo sign has been im- proved. The accuracy of species identification varies depending on the nature of the species complex in the survey area. Where shortbelly and redstripe rockfish are present, the potential for misidentifica- tion increases. Technological improvements in sonar equipment may help to reduce this problem. The density of a school is an important criterion for distinguishing widow rockfish from other species and newer sonar equipment includes density-graded color video displays. Other techniques, such as underwater photography or remote video camera vehicles, might also improve our ability to identify species. I believe, however, that test fishing will always be a necessary component of hydroacoustic resource assessment surveys. Surveys of widow rockfish resources must be designed with the behavior and distributional vari- ability of the species in mind. The diel behavior of the species indicates that the most effective sampling period is at night, but even then unpredictable be- havior places special demands on survey design. Observations from hydroacoustic transects which were replicated on several nights (Fig. 11) show that long-term variability in abundance (eg, night-to- night or week-to-week) is even more marked than that over a shorter time These results are substan- tiated by other surveys (see footnotes 5, 7, and 8) and illustrate the difficulty of estimating widow rockfish abundance Long-term variability is also a factor in area-swept bottom trawl surveys. The 307 FISHERY BULLETIN: VOL. 84, NO. 2 Table 9.— Comparison of school length measurements (m) derived from echo sounder versus sector-scanning sonar data collected aboard the FV Ocean Leader, 14 March-7 April 1982. Density (kg/m2) Length from Length from School echo sounder sonar 1 0.1292 48.5 45.0 2 0.0265 85.8 2.1 3 1 .4860 66.2 5.8 4 1.6996 51.8 19.2 5 0.6374 37.4 74.0 6 0.3559 37.4 79.6 7 1.2415 59.9 52.3 8 0.7513 93.2 66.5 9 0.2119 117.5 118.1 10 0.7567 81.9 236.2 11 0.0686 526.9 497.5 12 0.0453 45.0 31.9 13 0.0490 96.9 353.1 14 0.0068 138.0 18.7 15 1.1400 511.9 393.2 16 0.6089 189.7 64.2 17 0.3055 76.6 18.4 18 0.4564 84.6 70.6 19 0.2154 27.8 93.0 20 0.0080 62.5 116.6 21 0.0487 103.0 16.9 22 0.1057 130.3 49.3 23 0.0936 147.5 28.5 24 0.1730 27.4 9.9 25 0.0103 46.8 39.6 26 0.1262 53.6 40.9 27 0.0182 67.7 43.9 28 0.1013 30.1 124.9 29 0.0830 67.7 84.2 30 0.0430 67.7 189.4 31 0.0208 48.9 33.5 32 0.0117 23.8 41.4 33 0.0128 22.4 14.8 34 0.1808 106.4 50.2 35 0.1809 292.2 216.7 36 0.0424 81.0 66.4 37 3.3097 158.3 101.6 x 0.3990 105.8 94.8 s 0.6587 113.5 112.2 assumption is that the variability has a strong ran- dom component and catch per unit effort values are consequently unbiased. The same situation may well be true here, in which case an important component of the survey design would be multiple replication to obtain good estimates of both long- and short-term variance Burnham et al. (1980) reported that good results from line transect surveys require observation of a minimum of about 40 objects per replicate Fitting the observed perpendicular sighting distances to a detection function becomes less reliable with a smaller number of objects. Widow rockfish abun- dance is now low on all major grounds and the recommended minimum number of schools was not observed during any single replicate in the 1982 survey, but by pooling replicates a sufficient data- base was constructed. Sample sizes could be in- creased through more intense sampling. A time- stratified analysis of the data would be desirable to define within-night variability, but this would place even further demands on a sampling pro- gram. Surveys of the type used for widow rockfish must cover the geographic range of the species of interest more thoroughly than most other survey methods. The dynamic behavior of widow rockfish suggests that the survey method should cover large areas in a relatively short time in order to survey a given fishing ground at least once during the night. Be- cause of day-to-day variability, surveys should include sampling each area during several nights over a 1- or 2-wk period. Most areas containing fishable widow rockfish concentrations have probably been iden- tified and there are a limited number of these grounds (probably 12-20); nearly all are character- ized by ridges or rises on the outer continental shelf or upper slope and are relatively small in area. In- tensive sampling of widow rockfish, therefore, is more feasible than for most other groundfish species inhabiting less well-defined areas. Because widow rockfish schools are continually forming and breaking up, there may be a significant portion of the population which is not schooling at any given time and is therefore not susceptible to these survey techniques. This project did not answer whether this is so, but nothing was found to suggest that widow rockfish are significantly detectable by trawl or hydroacoustic surveys in any form other than midwater schools. Until more is learned about the proportion of the stock occurring as schools, surveys must be considered as yielding minimum biomass estimates. Clark and Mangel (1979) pro- posed a study of rates of school formation and disper- sal to explain and evaluate a similar relationship be- tween overall stock size and the proportion of a yellowfin tuna stock occurring as schools. Such a technique should receive further consideration in this situation, but present low widow rockfish school abundance (schools/km2) and lack of a consistent pattern of school formation and dispersal would probably make its application in widow rockfish assessment difficult. This question is analogous to that of defining catchability coefficients (i.a, what proportion of those fish in the path of a net are ac- tually captured) for quantitative trawl surveys. Changes in relative abundance can be monitored by such surveys without knowing the catchability if one assumes that the available proportion of the popula- tion is constant. Results of other analyses of widow rockfish be- 308 WILKINS: ABUNDANCE OF WIDOW ROCKFISH havior and stock size should be used to evaluate survey methodology. The groundfish management team of the Pacific Fisheries Management Council (see footnotes 2 and 3) used stock reduction and cohort analyses to estimate the abundance of this species. In an area comparable to our 1982 survey area, the widow rockfish biomass was estimated to be 21,664 t at the begining of 1982. This estimate is based in part on commercial landing information and, consequently, the definition of the grounds to which it applies is somewhat vague. The fishery- based estimates are much higher than those derived from the 1982 survey data (about 830 1). The relative- ly low sensitivity of the sonar systems used would result in underestimating biomass and is undoubted- ly responsible for much of this difference The discrepancy is also partly due to the fact that our survey methods only estimate the portion of the stock present as detectable schools and are therefore a measure of relative, rather then absolute, abun- dance. This is true to some extent for most types of surveys. Innovations are also needed to resolve the techni- cal problems related to data collection, identification of school species composition, and survey design. Some suggestions include 1) a two-vessel survey to improve the efficiency of data collection— such a technique would separate the chore of delineating areas of widow rockfish ag- gregations, estimating school abundance, and test fishing from that of estimating mean school biomass (Gunderson et al. fn. 7); 2) a means of recording a time base on both the audio and video tape records of the echo sounder and sonar to simplify finding the same school on each system for school dimension measurements; and 3) a method of estimating all school dimensions and the density within the school from a single data collection system— this would entail development of a sophisticated, quantitative sonar-integration sys- tem with the capability of recording the output onto videotape (Ehrenberg 1979). Such refinements could probably be implemented with relative ease. The methodology should be re- evaluated when these technological and sampling improvements have been made. Widow rockfish management could have been significantly improved with the knowledge of stock size from an effective resource assessment survey. There are also other species which exhibit similar behavior and which, although presently unexploited, need to be assessed (eg, shortbelly, redstripe, and black rockfish). This methodology could probably be easily adapted for surveying these resources. CONCLUSIONS Based on the results of research conducted dur- ing this project, the line transect survey method using a sector scanning sonar and a quantitative echo sounder appears to be the best means of assess- ing widow rockfish abundance with research surveys. A weakness of this method is that it only measures the portion of the population existent as distinguish- able schools and that portion may be quite variable It also relies heavily on subjective experience for identifying the species composition of schools. Its strengths are that large areas can be covered quickly and it is not necessary that all schools within sighting range be detected in order to estimate school abun- dance It appears that this could be a useful assess- ment method for widow rockfish and for several other Pacific coast groundfish species which are not yet being seriously exploited. The effectiveness of the technique could be enhanced by employing or developing more sensitive and specialized quan- titative sonars and by improving the methods of data collection. The technological and survey design prob- lems encountered should be relatively easy, though somewhat costly, to resolve The method should then be reevaluated to determine its utility. As the tech- nique is used, scientists will gain a better under- standing of the behavior and habits of the target species. ACKNOWLEDGMENTS The work described in the developmental section was ably conducted by Donald R. Gunderson and Gary L. Thomas and their associates at the Fisheries Research Institute, University of Washington Col- lege of Fisheries, Seattle, WA, under contract to the National Marine Fisheries Service (contract no. 79- ABC-00203). Much of the information presented in that section is extracted from their contract reports. The hydroacoustic expertise during the 1982 survey work was provided by the Pelagic Resources Assess- ment Task of the Resource Assessment and Conser- vation Engineering (RACE) Division, Northwest and Alaska Fisheries Center, NMFS; in particular, Ed- mund Nunnallee, Jimmie J. Traynor, and John Gar- rison. I am expecially grateful to Nunnallee for advice and guidance during the analysis of echo sounder and sonar data. I also wish to thank Thomas A. Dark, RACE Division, and Nunnallee and Traynor 309 for their thoughtful and constructive review of this manuscript. LITERATURE CITED Burnham, K. P., D. R. Anderson, and J. L. Laake. 1980. Estimation of density from line transect sampling of biological populations. Wildl. Monogr. 72, 202 p. Clark, C. W., and M. Mangel. 1979. Aggregation and fishery dynamics: a theoretical study of schooling and the purse seine tuna fisheries. Fish. Bull., U.S. 77:317-337. Dark, T. A., M. 0. Nelson, J. J. Traynor, and E. P. Nunnallee. 1980. The distribution, abundance and biological character- istics of Pacific whiting, Merluccius productus, in the Cali- fornia-British Columbia region during July-September 1977. Mar. Fish. Rev. 42(3-4): 17-33. Ehrenberg, J. E. 1979. The potential of the sector-scanning sonar for in situ measurements of fish target strengths. In Proceedings of the 1979 Institute of Acoustics Meeting on sector scanning sonars. Lowestoft, England. Forbes, S. T., and 0. Nakken (editors). 1972. Manual of methods for fisheries resource survey and appraisal. Part 2. The use of acoustic instruments for fish detection and abundance estimation. FAO Man. Fish. Sci. 5, 138 p. Goodman, L. A. 1960. On the exact variance of products. J. Am. Stat. Assoc 55:708-713. Hitz, C. R. 1962. Seasons of birth of rockfish (Sebastes spp.) in Oregon FISHERY BULLETIN: VOL. 84, NO. 2 coastal waters. Trans. Am. Fish. Soc. 91:231-233. Kimura, D. K., and J. V. Tagart. 1982. Stock reduction analysis, another solution to the catch equation. Can. J. Fish. Aquat. Sci. 39:1467-1472. Laake, J. L., K. P. Burnham, and D. R. Anderson. 1979. User's manual for program TRANSECT. Utah State Univ. Press, Logan, 26 p. Moose, P. H., and J. E. Ehrenberg. 1971. An expression for the variance of abundance estimates using a fish echo integrator. J. Fish. Res. Board Can. 28: 1293-1301. Pereyra, W. T, W. G. Pearcy, and F. E. Carve y, Jr. 1969. Sabastodes Jlavidus, a shelf rockfish feeding on meso- pelagic fauna, with consideration of the ecological implica- tions. J. Fish. Res. Board Can. 26:2211-2215. Phillips, J. B. 1964. Life history studies on ten species of rockfish (genus Sabastodes). Calif. Dep. Fish Game, Fish Bull. 126, 70 p. Quinn, T. J., II. 1979. The effects of school structure on line transect estima- tors of abundance In G. P. Patil and M. L. Rosenzweig (editors), Contemporary quantitative ecology and related ecometrics, p. 473-491. Int. Coop. Publ. House, Fairland, MD. Seber, G. A. F. 1973. The estimation of animal abundance and related param- eters. Hafner Press, N.Y., 506 p. 1980. Some recent advances in the estimation of animal abun- dance Univ. Wash., Wash. Sea Grant Tech. Rep. 80-1, 101 p. Thorne, R. E. 1977. A new digital hydroacoustic data processor and some observations on herring in Alaska. J. Fish. Res. Board Can. 34:2288-2294. 310 POPULATION AND FISHERY CHARACTERISTICS OF GULF MENHADEN, BREVOORTIA PATRONUS Walter R. Nelson1 and Dean W. Ahrenholz2 ABSTRACT Landing data from 1964 to 1978 for the purse seine fishery in the north-central Gulf of Mexico for gulf menhaden, Brevoortia patronus, were analyzed to determine growth rate, yield-per-recruit and spawner- recruit relationships, and maximum sustainable yield (MSY). Estimates of stock size, year-class size, and rates of fishing were obtained from cohort analysis. The fishery is characterized by high rates of both fishing and natural mortality. During the period studied, an average of 40% of the population of age-1 and older fish were taken by the fishery and 47% was lost to other causes annually. Although there was substantial scatter about the fitted curve, a Ricker-type spawner-recruit relationship was found to be suitable The number of age-1 recruits fluctuated annually between 7.5 and 25.4 billion during the period studied. Maximum biomass of a year class is reached at an age of about 1.5 years. Yield-per-recruit estimates were obtained for an array of fishing mortalities and ages of entry. A deterministic simulation model incorporating growth, the spawner-recruit relationship, and age-specific rates of fishing provided an estimate of MSY at 585,118 t with 127% of the current mean rate of fishing. Implications for the current and future status of this fishery are discussed. Gulf menhaden, Brevoortia patronus, are filter-feed- ing, surface-schooling clupeids that are subjected to an intensive purse seine fishery in the northern Gulf of Mexico. Although annual landings have fluc- tuated, there has been a general increase since the inception of the modern fishery in 1946 to a high of 820,000 metric tons (t) in 1978. The fishery con- sists of about 80 refrigerated vessels serving 11 reduction plants at 6 ports in Mississippi and Loui- siana. The fishing season is currently set by State law from mid-April to mid-October. Although a majority of the catch is taken off Louisiana and Mississippi, vessels range west into eastern Texas coastal waters and east to the coastal waters of the Florida panhandle Vessels, aided by spotter aircraft, land from 6,000 to 10,000 t/6-mo fishing season. Ex- cellent background information and descriptions of the fishery have been published by Christmas and Etzold (1977) and Nicholson (1978). Considerable literature exists on the general biology of gulf menhaden (Reintjes et al. 1960; Rein- tjes 1964; Reintjes and Keney 1975; Christmas and Etzold 1977); however, information is scarce on the population dynamics of gulf menhaden and on the dynamics and impact of the fishery. Chapoton (1972) 'Southeast Fisheries Center Beaufort Laboratory, National Marine Fisheries Service, NOAA, Beaufort, NC; present address: Southeast Fisheries Center Miami Laboratory, National Marine Fisheries Service, NOAA, 75 Virginia Beach Drive, Miami, FL 33149. Southeast Fisheries Center Beaufort Laboratory, National Marine Fisheries Service, NOAA, Beaufort, NC 28516-9722. and Schaaf (1975a) estimated maximum sustainable yield (MSY). Ahrenholz (1981) described recruitment patterns and estimated natural and fishing mortality rates from returns of tagged juvenile and adult menhaden. Gulf menhaden have a life history similar to many other estuarine-dependent coastal species. Spawn- ing takes place in coastal and offshore waters in the winter (Christmas and Waller 19753; Lewis and Roithmayr 1981). Larvae move onshore into Gulf estuaries in winter and early spring, transform to juveniles, and remain in the nursery areas until the following fall. Juveniles move offshore during the winter and back into coastal waters the following summer. Spawning occurs for the first time at the end of their second year. A joint State-Federal-Industry plan developed for gulf menhaden identified the lack of a reliable mea- sure of effective effort and questionable MSY estimates as major concerns in evaluating the gulf menhaden stock and fishery (Christmas and Etzold 1977). Problems encountered in determining the status of gulf menhaden stocks and estimating a long-term yield from catch-effort data on schooling species subjected to a purse seine fishery are com- pounded by the "dynamic aggregation process" described by Clark and Mangel (1979). Basically, they 3Christmas, J. Y, and R. S. Waller. 1975. Location and time of menhaden spawning in the Gulf of Mexico. Unpubl. manuscr., 20 p. Gulf Coast Research Laboratory, Ocean Springs, MS 39564. Manuscript accepted July 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 311 FISHERY BULLETIN: VOL. 84, NO. 2 hypothesized that surface schooling species are more susceptible to fishing effort than nonschooling species, and indicators of abundance such as catch and catch per unit effort (CPUE) are not reliable when the "intrinsic schooling rate is greater than the intrinsic (population) growth rate". Thus, severe stock depletion could occur in the gulf menhaden fishery before indications of such a situation were evident from catch and CPUE data. The dynamic aggregation process may be further aggravated when the vessels are assisted by spotter aircraft which greatly reduce search time. We have attempted to estimate characteristics of the gulf menhaden stock, such as population size, biomass, growth rate, spawner-recruit relationship, and to determine characteristics of the fishery, such as fishing mortality, catchability coefficient, yield- per-recruit, equilibrium yield levels, and MSY. These characteristics were determined through application of cohort analysis, yield-per-recruit and spawner- recruit models, and a deterministic simulation model of the Gulf of Mexico population and fishery. Our overall objectives are to evaluate the status of the gulf menhaden stock, determine the impact of the fishery, and provide an outlook for the stock and fishery for resource managers and the purse seine fishing industry in the Gulf of Mexico. GULF MENHADEN DATA BASE The National Marine Fisheries Service (formerly Bureau of Commercial Fisheries) has maintained a sampling program for gulf menhaden since 1964. Details of the sampling methodology are given by Nicholson (1978) and Huntsman and Chapoton (1973), and a description of the aging technique is provided by Nicholson and Schaaf (1978). Vessel landings by trip have been recorded, along with per- tinent data on vessel size and characteristics. Overall summaries of landings by year and nominal effort (measured in vessel-ton-weeks) are available back to 1945, but the basis for the bulk of this analysis is the catch and effort data (1964-79) and estimated number of fish landed at age for these years (Table 1). WEIGHT-LENGTH RELATIONSHIP AND GROWTH Estimates of growth rate are needed for yield analyses and estimates of size at age are needed to determine the spawner-recruit relationship. Al- though some calculations use length and others weight, all growth estimates were calculated for length, and when required, weight was estimated from the weight-length relationship. For each age group, there was no major systematic variation in the mean length over the 15 yr period (Fig. 1). In addition, no density-dependent correla- tions were detectable for mean length at age on stock size or on year-class size, estimated from the subsequent cohort analysis. Hence there appeared to be little, if any, potential gain in estimate accuracy by computing and using year-class specific growth rates when reconstructing the historical population biomass and average size at age for the subsequent spawner-recruit analysis, or to incorporate a density- dependent growth function in the subsequent popula- tion simulations for total yield. Estimates of overall mean length at age for each quarter for the year classes that had passed com- Table 1. — Catch, effort, and estimated number of gulf menhaden landed at age for the 1964-79 fishing seasons (1964-78 for number at age). Catch (metric No. of Effort (vessel-ton- Number at age x 106 Year tons x 103) vessels' weeks x 103) 0 1 2 3 4 Total 1964 409.4 76 272.9 6.3 3,135.6 1 ,365.2 108.1 3.9 4,619.1 1965 463.1 82 335.6 46.6 4,888.1 966.3 69.9 1.5 5,972.4 1966 359.1 80 381.3 46.8 3,126.8 850.2 30.5 0.5 4,054.8 1967 317.3 76 404.7 18.7 4,129.2 309.9 10.5 — 4,468.3 1968 373.5 69 382.3 35.4 3,311.5 850.0 27.0 0.2 4,224.1 1969 523.7 72 411.0 10.8 5,766.8 1,011.1 30.4 — 6,819.1 1970 548.1 73 400.0 49.2 3,256.4 2,197.2 34.4 — 5,537.2 1971 728.2 82 472.9 25.3 5,763.3 1,838.1 166.2 3.7 7,796.6 1972 501.7 75 447.5 17.6 3,146.3 1,615.6 68.7 4.4 4,852.6 1973 486.1 65 426.2 57.2 3,012.4 1,082.7 108.2 1.3 4,261.8 1974 578.6 71 485.5 20.0 3,747.3 1,399.0 60.2 — 5,226.5 1975 542.6 78 536.9 96.4 2,512.3 1,453.1 428.2 0.8 4,490.8 1976 561.2 81 575.9 1.8 4,517.7 1,273.1 190.2 — 5,982.8 1977 447.1 80 532.7 1.6 4,800.2 1,209.6 104.3 7.3 6,123.0 1978 820.0 80 574.3 0.0 6,784.7 2,578.8 48.3 3.6 9,415.4 1979 777.9 77 533.9 — — — — — — 'Includes only vessels that fished 9 or more weeks. 312 NELSON and AHRENHOLZ: CHARACTERISTICS OF GULF MENHADEN Figure 1— Mean length of gulf menhaden at ages 1-3 taken from commercial landing sam- ples (April- June), 1964-78. E E I I- O z LU —I rr O Li. < LU 250r 200 150 100 of J I J I I L J L 1964 65 66 67 68 69 70 71 72 73 74 75 76 77 78 YEAR pletely through the fishery were used in the growth computations (Table 2). The von Bertalanffy growth- in length equation lt = Loo(l - e ■W-L ') (1) where lt = fork length (mm) at time t (years), Leo = theoretical length at t = infinity (asymptote), K = growth coefficient, t0 = theoretical age when length = 0 was fitted to the data by the computer program BGC3 (Abramson 1971). Although the data points appear stepped between whole age units, they are reasonably well described by the fitted curve (Fig. 2). Table 2.— Mean length and number of fish sampled at age from 1963 to 1974 year classes of gulf menhaden. Mean fork length Age (mm) Number 1.125 121.7 59 1.375 148.3 43,284 1.625 160.5 57,286 1.875 161.3 1,538 2.125 — — 2.375 182.9 16,687 2.625 190.6 16,452 2.875 194.4 260 3.125 — — 3.375 210.5 1,063 3.625 216.4 1,368 3.875 220.2 14 4.125 — — 4.375 227.8 16 4.625 227.5 32 4.875 — — E E X t- a z LU _) cr O LL 24U 200 160 120 S* 80 / lt=252.893(l-e-°-4748(t+0-3585)) 40 o ■ i i . i . i . i . i i ' " i D 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 AGE IN YEARS Figure 2— Von Bertalanffy growth curve fitted to average length at age data for gulf menhaden sampled from commercial landings, 1963-74 year classes. 313 FISHERY BULLETIN: VOL. 84, NO. 2 Weight-length regression coefficients were calcu- lated for each of three 2-mo intervals for the major portion of the fishing season for each year class, 1960-77. No systematic variation in the parameter estimates was apparent within years by 2-mo inter- vals, so the data were pooled between seasons and years. An overall weight-length relationship was ob- tained from a GM-functional regression (Ricker 1973) on the pooled data. The results are log, w = 3.2669 log, I - 12.1851 (2) where w = weight in g, and I = fork length in mm. The correlation coefficient (r) was 0.976 and the sam- ple size was 168,397. COHORT ANALYSIS Estimates of mortality rates and population sizes were obtained by using the Cohort Analysis tech- nique developed by Murphy (1965) and later modified by Tomlinson (1970). Calculations were made with the computer program MURPHY (Tomlinson 1970). This technique does not involve estimates of CPUE. The backward estimation procedure was used. Since the catch equations and general method of applica- tion are given in Tomlinson's paper, discussion here will be limited to the source and nature of input data and parameters. The calendar year was divided into four periods (quarters) of approximately equal length: Quarter 1 = 1 January to 3 April, Quarter 2 = 4 April to 3 July, Quarter 3 = 4 July to 3 October, Quarter 4 = 4 October to 31 December. Numbers of fish at each age landed quarterly were sums of weekly estimates obtained by sampling methods outlined by Nicholson (1978). Annual sum- maries of these data were given earlier (Table 1). An estimate of the annual rate of instantaneous natural mortality (M) was obtained from an analysis of mark-recapture data (Ahrenholz 1981). M, equal to 1.1 (0.275 per quarter), was assumed to be con- stant for all ages and seasons. Because backward sequential computations, using a range of trial estimates of input F (instantaneous fishing mortality) for the oldest age, tend to con- verge on the correct value of F for the youngest and forward calculations tend to diverge (unless true starting values are used), it is desirable to begin with the oldest age for which reliable landing data are available (Ricker 1975). Because of aging difficulties (Nicholson and Schaaf 1978), we assumed that catch estimates of older fish, mainly age 4, were not reliable, hence for most year classes estimates of the annual rate of instantaneous fishing mortality (F) for age-2 fish were derived from catches of age 2 and age 3 from Fn = (log, Cn - log, Cn+1) - M (3) where C = annual catch in numbers at age (n) from a given cohort. Initial starting values of F for the oldest age group landed in a year class were adjusted by trial and er- ror until the sum of the quarterly Fs for age-2 fish were virtually equal to the estimate of annual F2 derived from Equation (3). This technique was ap- plicable for all year classes except 1960 and 1961, where no 2-yr-old fish were available in the landing data, the 1976 year class where no 3-yr-old fish were available in the landing data, and the 1972 year class, where the 2-yr-old fish apparently were not fully recruited. For the 1960 and 1961 year classes, trial and error adjustments were made to the starting F value until the annual Fz estimate for the 1961 year class and the annual F4 estimate of the 1960 year class were virtually equal to the unweighted mean F3 estimate derived from the sequential computa- tions of the 1963-71 and 1973-75 year classes. Similarly, the mean F2 estimate was used for the 1976 year class and the mean F3 estimate for the 1972 year class. Estimates of number- at-age by quarter by year class obtained from cohort analysis permitted the reconstruction of population structure for the ex- ploited gulf menhaden stock from 1964 to 1977 (Table 3). Numbers of newly recruited age-1 fish varied as much as threefold between years. Because age-1 fish were numerically the most abundant age group each year, the population size fluctuated in close concert with their numbers (Fig. 3). Resultant age-specific annual Fs by fishing season demonstrate that 1-yr-olds are incompletely re- cruited to the fishery and that age 2's are fully re- cruited (Table 4). These results are in accord with those of Ahrenholz (1981), who concluded that fish from more distant eastern and western, areas of the Gulf of Mexico (Gulf) shifted toward the more heavily fished central Gulf areas as they aged. The slightly higher values for both the weighted and unweighted mean F's for 3- and 4-yr-olds could be due to either small numbers of fish from the most distant eastern 314 NELSON and AHRENHOLZ: CHARACTERISTICS OF GULF MENHADEN Table 3.— Population size (in millions) of gulf menhaden on 4 April estimated by cohort analysis, 1964-77. Age Year 1 2 3 4 1964 8,189.2 2,048.0 156.8 5.5 1965 9,796.0 1,329.2 105.0 7.4 1966 5,703.8 1,111.9 41.9 5.2 1967 9,215.6 548.5 14.4 0.0 1968 9,256.7 1,249.0 47.6 0.3 1969 19,311.9 1,539.6 44.6 0.0 1970 12,454.5 3,817.6 59.4 0.0 1971 15,860.1 2,635.0 289.3 4.7 1972 9,580.3 2,704.4 97.0 25.6 1973 15,793.9 1,796.2 181.5 2.6 1974 15,107.1 3,849.3 99.6 0.0 1975 10,220.9 3,324.1 668.4 5.5 1976 11,467.8 2,216.2 435.6 0.0 1977 18,584.2 1,739.4 181.4 57.9 and western areas reaching the more intensively fished waters, or simply a sampling variance. The F estimates from cohort analysis for age-3 and age-4 fish are somewhat suspect, especially for age 4, since the cohort analysis technique used (iterating to a preset F2) actually makes the F3 and F4 estimates of a forward computational nature, rather than backward as for age 1. The divergent nature of the estimates is clearly evident in the values for age 4, although the mean value is realistic for subsequent yield computations, and numbers of fish at this age are of very low magnitude as well. Because a year class is well represented in the fishery for only 3 yr, a short time span is available for the convergence of estimates of numbers and fishing mortality. This short time span was ap- 25r Figure 3— Population number of gulf menha- den as of 4 April 1964-77, Estimated from cohort analysis on 1960-76 year classes. CO c o LU N CO O I- < -J a. O 0. ■ ■ ' I I 1 1 1 1964 65 66 67 68 69 70 71 72 73 74 75 76 77 YEAR Table 4.— Annual instantaneous fishing mortality rate (F) for gulf menhaden for ages 1-3, by year, 1964-77, and fishing mortality rate applied at age 4 (age 3 for year classes without age-4 landings) to initiate the cohort analysis. F Year Age 1 Age 2 Age 3 Age 4 1964 0.7182 1 .8706 1 .9547 1 .9504 1965 1 .0757 2.3576 1.9112 0.3546 1966 1.2431 3.2468 2.5992 0.1399 1967 0.8991 1 .3447 2.7786 0.0000 1968 0.6938 2.2323 1 .4032 110.6065 1969 0.5211 2.1553 1 .6225 0.0000 1970 0.4521 1 .4760 1 .4392 0.0000 1971 0.6681 2.2033 1 .3287 110.4097 1972 0.5740 1.6024 2.5195 0.2688 1973 0.3120 1.7933 2.0281 1.2313 1974 0.4140 0.6507 1 .7885 0.0000 1975 0.4287 0.9322 1 .9484 0.1950 1976 0.7860 1 .4030 0.9184 0.0000 1977 0.4375 2.1293 1 .4229 0.1923 Mean F 0.6588 1.8141 1 .8331 1.8106 (unweighted) parently adequate, however, as cohort runs on the year classes with 4-yr-olds in the landings, using starting estimates of F for age 4 obtained from catch curves, converged to very similar estimates to those obtained by the analysis used here. Ulltang (1977) emphasized that when F is high, convergence is rapid. The short-term impact of the fishery on the stock can be assessed by comparing the estimated number- at-age in the population for any given year with the number-at-age landed by the fishery, or simply by using the estimated rate of fishing and calculating the exploitation rate (u) by un = (Fn (1 - e-^W)Wn + M). (4) initial F set equal to 10.0. From 1964 to 1977 the fishery took an average of 31% of the 1-yr-olds in the population and about 61% of the older fish each year. At these exploitation rates 315 FISHERY BULLETIN: VOL. 84, NO. 2 the population loses 52% of the age-1 fish and 35% of the older fish to natural mortality. The short-term impact on the entire population was determined by 1) calculating a mean F weighted by the number of individuals taken by age for ages 1-3 and then estimating u by Equation (4) and 2) directly comparing numbers landed with the recon- structed population sizes. The average annual loss of individuals from the population to the fishery was about 40% by both methods. However, recruitment is only partial at age 1, and u is much higher at older ages. Natural mortality losses averaged about 47%/year for the overall population. In the absence of fishing, annual losses to natural mortality would be about 67% for all ages. A measure of how a unit of fishing effort affects the population is commonly quantified through its effect on F. Traditionally this effect, the catchabil- ity coefficient (q), is assumed to be a constant. The total fishing effort times this constant should equal F for the year: F=qf (5) where / = a unit of fishing effort (here, a vessel- ton-week). if the catchability coefficients were independent of this variable (Fig. 4). An inverse relationship was noted, a situation which also exists for the Atlantic menhaden (Schaaf 1975b). The data were fitted to the power function to demonstrate the curvilinear inverse relationship. SPAWNER-RECRUIT RELATIONSHIP The cohort analysis provides estimates of popula- tion size at ages 1-4 from 1964 to 1977. All fish mature by the end of their second year, and spawn- ing apparently reaches a peak in December and January (Lewis and Roithmayr 1981). Therefore, estimates of number-at-age in the population as of 1 January were used to provide estimates of spawn- ing stock size and subsequent recruitment (Table 6). Spawning stock was identified as all fish that had reached at least their second birthday by 1 January. Lewis and Roithmayr also showed that length ac- counted for a greater porportion of the variance in fecundity than either age or weight. Our fecundity estimates, assuming a 1:1 sex ratio, were based on Lewis and Roithmayr's relationship: log, E = -9.8719 + 3.8775 (log, I) (6) Estimates of q for the 1964-77 fishing years were obtained by solving for q in the above equation for the population F for ages 1-3 (i^.3) weighted by number taken at age, and also for the population total F (Table 5). The resultant g's are quite variable (in excess of fourfold). Estimates of q were plotted against corresponding population size to determine where E = fecundity in number of eggs and I = fork length in millimeters. Because there was little variation in size at age by year class, and the differences noted were not related to population size, estimates of mean length- at-age were obtained from the overall von Bertal- Table 5. — Estimated gulf menhaden population size as of April 4, number caught by year, population exploitation rate (u), estimated population fishing mortality rate (F), population catchability coefficient (q) x 10"3, weighted annual mean fishing mortal- ity rate from cohort analysis (^.3), and the corresponding Fv3 catchability coefficient (q) x 10 "3 calculated from vessel-ton-weeks (Table 1), 1964-77. Population Number size (millions) caught (millions) Population Year age 1-4 age 1-4 u F Qx10-3 K3 Q1.3XIO"3 1964 10,399.5 4,612.8 0.444 1.10 4.03 1 .0886 3.99 1965 11,237.6 5,925.8 0.527 1.46 4.35 1 .2946 3.86 1966 6,862.8 4,008.0 0.584 1.78 4.67 1 .6785 4.40 1967 9,778.5 4,449.6 0.455 1.14 2.82 0.9346 2.31 1968 10,553.6 4,188.7 0.397 0.93 2.46 1.0176 2.64 1969 20,896.1 6,808.3 0.326 0.70 1.70 0.7687 1.87 1970 16,331.5 5,488.0 0.336 0.73 1.83 0.8682 2.17 1971 18,789.1 7,771.3 0.414 0.99 2.09 1 .0455 2.21 1972 12,407.3 4,835.0 0.390 0.90 2.01 0.9456 2.11 1973 17,774.2 4,204.6 0.237 0.47 1.10 0.7377 1.73 1974 19,056.0 5,206.5 0.273 0.56 1.15 0.4935 1.02 1975 14,218.9 4,394.4 0.309 0.66 1.23 0.7433 1.38 1976 14,119.6 5,981.0 0.424 1.02 1.77 0.9215 1.60 1977 20,562.9 6,121.4 0.298 0.63 1.18 0.7890 1.48 316 NELSON and AHRENHOLZ: CHARACTERISTICS OF GULF MENHADEN 6r n i O X 4- Z UJ o LLI o o > b m < i o i- < °0k 5 6 _i_ J_ _L 9 10 11 12 13 14 15 16 17 18 19 20 21 POPULATION SIZE (billions) Figure 4— Catchability coefficients calculated from population fishing mor- talities (open circles, dashed line), and from cohort annual weighted mean fish- ing mortalities (dots, solid line) plotted on population number estimated as of 4 April, for the 1964-77 fishing seasons (see Table 5). Table 6.— 1 January estimates of number of spawners, number of eggs produced by the spawn- ing stock, biomass of the spawning stock, and number and biomass of recruits at age 1 for gulf menhaden. Preliminary estimates in parentheses. Total No. Spawning Resultant Recruitment No. at age ( spawners of eggs biomass recruitment biomass Year 2 3 4 (millions) (trillions) (t) (millions) (t) 1964 2,696.3 206.4 7.2 2,909.9 36.1 305,468 12,896.7 410,630 1965 1,749.9 138.2 9.7 1,897.8 23.7 200,150 7,519.5 239,421 1966 1,463.9 55.1 6.8 1,525.8 18.4 156,705 12,138.2 386,480 1967 722.2 19.0 — 741.2 8.8 75,118 12,186.7 388,025 1968 1,644.3 62.6 0.4 1 ,707.3 20.5 174,454 25,424.7 809,522 1969 2,026.9 58.7 — 2,085.6 24.8 211,752 16,396.8 522,074 1970 5,026.0 78.2 — 5,104.2 60.0 513,461 20,889.9 665,134 1971 3,472.8 382.4 6.2 3,861.4 49.0 412,808 12,618.5 401,773 1972 3,565.3 127.7 33.7 3,726.7 45.2 384,521 20,796.4 662,157 1973 2,365.8 239.0 3.4 2,608.2 32.8 277,323 19,889.0 633,266 1974 5,067.7 131.1 — 5,198.8 61.7 526,725 13,456.1 428,442 1975 4,376.3 879.9 7.3 5,263.5 70.5 588,668 (15,097.7) (480,711) 1976 2,917.7 573.5 — 3,491.2 46.6 389,073 (24,466.7) (779,020) 1977 (2,290.0) 238.8 76.2 (2,605.0) (34.3) (286,686) 1978 (5,258.5) ( 90.6) 19.2 (5,368.3) (63.6) (543,194) anffy growth function presented earlier. Thus, length-at-age estimates are taken as constants, and differences in among year estimates of egg produc- tion are due to differences in both total numbers and age composition of the spawning stock (Table 6). Similarly, the weight-length relationship was used in conjunction with the mean length-at-age esti- mates to obtain weight-at-age estimates for com- putation of spawning and recruitment biomass (Table 6). Least-square regressions of second and third degree polynomials were run with numbers of re- cruits on number of spawners to determine the general shape of the spawner-recruit relationship. Dome-shaped functions provided the least residual sum of squares, indicating that a Ricker-type curve (Ricker 1975) is appropriate A Ricker-type function has been applied to Atlantic menhaden data (Nelson et al. 1977), and the same criteria appear to apply to gulf menhaden data, i.&, that there is a size- dependent fecundity relationship and that adult menhaden are filter feeders which are known to in- gest their own eggs. Additionally, the calculation of an index of density dependence, as detailed by Cushing (1971) (loge recruitment regressed on loge spawning stock), provides a slope of 0.159. This slope, 317 FISHERY BULLETIN: VOL. 84, NO. 2 plus the average fecundity of gulf menhaden (about 25,000 eggs/female) places the gulf menhaden among Cushing's clupeoid groups which have a slightly domed spawner-recruit curve Accordingly, a spawner-recruite relationship was applied of the form: R = Se^ ~ 5>/s« (7) where R S e Sr = recruitment at age 1 = spawning stock size = base of natural logarithm = maximum equilibrium stock = spawning stock size yielding maxi- mum absolute recruitment. The model, fitted by a nonlinear least squares technique (Marquardt 1963), predicts an average maximum recruitment of 18.4 billion individuals at a spawning stock of 3.22 billion (Table 7). The curve is a reasonably good fit (Fig. 5), considering the variability inherent in clupeoid recruitment. Data were available over a wide range of spawning stock sizes and recruitment levels. Although recruitment tended to fluctuate widely at lower spawning stock sizes, estimates appear to converge at higher spawn- ing stock levels, indicating the possibility of a strong density-dependent response as spawning stock size increases. The Ricker function appears to describe the data, thus an estimate of spawning stock size premits a general estimate of anticipated re- cruitment at moderate to high numbers of spawners. Because fecundity increases with age and because age structure of spawners varies from year to year, estimates of the number of eggs produced should provide a more accurate estimate of spawning stock size than estimates of the numbers of spawners. When the Ricker equation was fitted to number of eggs and number of recruits, the estimate of op- timum spawning stock size was similar to the estimate based on the number of spawners and recruits (Table 7) (SM of 39.66 trillion eggs = 2.3 billion spawners). The unrealistic replacement level (Sr) of 283.32 trillion eggs was generated by scaling factors involved in the comparison of unequal spawner and recruit units (Ricker 1975). Applying the function to spawning and recruitment biomass also provided similar estimates of maximum recruit- ment and optimum spawning stock size (Table 7, Table 7. — Ricker spawner-recruit estimates of maximum equilibrium stock (Sr), stock size for maximum recruitment (Sm), and recruit- ment at Sm, for models incorporating number of spawners on number of recruits, number of eggs on number of recruits, and spawning biomass on recruitment biomass, 1964-76 year classes of gulf menhaden. Stock for Maximum maximum equilibrium recruitment Recruitment stock (Sf) (Sm) atSm No. of spawners on no. of recruits 8.83 billion 3.22 billion 18.42 billion No. of eggs on no. of recruits 283.32 trillion 39.66 trillion 18.48 billion Spawning biomass on recruit biomass 524,172 t 336,011 t 588,236 t NUMBER OF SPAWNERS (billions) Figure 5.— Ricker spawner-recruit relationship for number of spawners and recruits at age 1, estimated as of 1 January, for the 1964-76 gulf menhaden year classes. 318 NELSON and AHRENHOLZ: CHARACTERISTICS OF GULF MENHADEN Fig. 6). The maximum recruitment level of 588,236 t is equal to about 18.5 billion recruits. The optimum spawning stock biomass of 336,011 t is equal to about 3.2 billion spawners, assuming the age distri- bution for the spawning stock is average. The func- tion for biomass accounts for changes in age struc- ture of the spawning stock and since age-2 fish consistently represent over 90% of the spawners, differences between plots of numbers and biomass (Figs. 5, 6) are minor. Spawning stock size has generally remained with- in a range of potentially good recruitment and has not undergone years of extreme highs or lows (Figs. 5, 6). Trends indicating a steady decrease or increase in stock size and recruitment are not apparent, al- though the general increased level of recruitment in recent years may be part of a cyclic recruitment fluctuation that is found in many stocks. YIELD-PER-RECRUIT We applied what is essentially a Ricker type yield- per-recruit model that was initially developed to evaluate a multiple-gear fishery (M-GE AR) and later modified to accommodate a multiple-area fishery (M- AREA) (Lenarz et al. 1974; Epperly et al. 19794). Yield is summed by time intervals, and individual weights and estimates of natural and fishing mor- tality can be inserted for each interval (Ricker 1975). An option developed by Epperly et al. (fn. 4) allows 4Epperly, S. P., W. H. Lenarz, L. T. Massey, and W. R. Nelson. 1979. A generalized computer program for yield per recruit analysis of a migrating population with area specific growth and mortality rates. Unpubl. manuscr., 14 p. Southeast Fisheries Center Beaufort Laboratory, National Marine Fisheries Service, NOAA, Beaufort, NC 28516. for calculation of biomass within intervals by either exponential or arithmetic means. We applied the model in its simpliest form: one set of growth data because the stock was not divided into subareas, con- stant natural mortality rate, and the exponential growth mode for biomass calculation. The year was divided into quarters to simulate the seasonal nature of the fishery (Table 8). Quarterly fishing mortality rates were developed from the cohort analysis. Estimates were obtained for periods of low popula- tion size and high fishing mortality (1964-68), high population size and low fishing mortality (1974-77), and "average" population size and mortality (1964-77) (Table 8). Age of entry into the fishery was Table 8.— Input array of quarterly length (mm), weight (g), and fishing mortality rates (F) used in the calculation of yield-per-recruit of gulf menhaden under average fishing conditions (1964-77), years of low stock size (1964-68), and years of high stock size (1974- 77). Age 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 Months W F (64-77) F (64-68) July-Sept. Oct.-Dec. Jan. -Mar. Apr.-June July-Sept. Oct.-Dec. Jan. -Mar. Apr.-June July-Sept. Oct.-Dec. Jan.-Mar. Apr.June July-Sept. Oct.-Dec. Jan.-Mar. Apr.-June July-Sept. Oct.-Dec. 84.7 103.5 120.2 135.1 148.2 160.0 170.4 179.6 187.8 195.1 201.6 207.3 212.4 216.9 221.0 224.5 227.7 230.5 10.1 19.5 31.8 46.6 63.2 81.0 99.5 118.2 136.8 154.9 172.3 188.9 204.5 219.1 232.7 245.2 256.7 267.2 0.0013 0.0003 0.0002 0.1850 0.4437 0.0299 0.0002 0.4478 1.2213 0.1448 0.0003 0.4500 1 .2722 0.1106 0.0000 0.1605 1.6501 0.0018 0.0001 0.0004 0.2677 0.6315 0.0264 0.0000 0.5652 1 .5858 0.0594 0.0000 0.5407 1.5752 0.0133 0.0000 0.3085 2.3018 F (74-77) 0.0008 0.0005 0.0000 0.1244 0.3593 0.0329 0.0000 0.3683 0.8253 0.0852 0.0000 0.2966 1 .0670 0.1557 0.0000 0.0488 0.0480 CO CO < r 800r 600 1- z LU o 40C > H o ) ft CD O E 200 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 AGE AT ENTRY Figure 7.— Yield-per-recruit of gulf menhaden under average con- ditions of growth and with multiples of average fishing mortality by 3-mo interval (F-multiple = 1.0) for the 1964-77 fishing seasons (average conditions indicated by □). 320 NELSON and AHRENHOLZ: CHARACTERISTICS OF GULF MENHADEN SUSTAINABLE YIELD AND POPULATION SIMULATION Production functions were developed from the 1946-79 catch and effort data to provide an estimate of maximum sustainable yield (MSY) for gulf men- haden. Application of a standard parabolic surplus production model (Schaefer 1954, 1957) yields an MSY estimate of 553,000 t at 555,000 vessel-ton- weeks. Past updates of MSY for the Gulf fishery have shown continual increases as additional years are added. Chapoton (1972) estimated an MSY of 430,000 t for the 1946-70 period, and Schaafs (1975a) estimate of 478,000 t included the 1971 and 1972 catch and effort. For the years in which estimates of catchability coefficient (q) were calculated (1964-77) nominal ef- fort was adjusted to the mean population q of that period. For that time period, mean catchability coef- ficient was divided by the estimate of population F each year, to provide an estimate of effort adjusted for "average" conditions from 1964 to 1977. A parabolic surplus production function was ap- plied to the 1946-79 data set, with adjusted effort used instead of nominal effort for 1964-77. The results were similar to model results using nominal effort with an estimated MSY of 541,904 t at an ef- fort of 505,483 vessel-ton-weeks (Fig. 8). A general- ized stock production model (PRODFIT) which allows the shape of the curve to vary based on a least squares fit to the data (Fox 1975) was also applied, yielding an estimate of MSY of 636,886 t at an ef- fort of 531,201 vessel-ton-weeks (Fig. 8). The two curves provide estimates that vary by about 95,000 1 with the PRODFIT model indicating a sharp drop in yield after MSY is exceeded. An estimate of MSY based on biological charac- teristics should be more reliable than one based on yield and nominal effort, particularly when there is not a clear nominal effort-effective effort relation- ship. Accordingly, we applied a population simula- tion model (Walters 1969) for the 1964-77 period which incorporated our estimates of growth, spawner-recruit relationship, fishing mortality, and natural mortality. This estimated the impact of fish- ing mortality on stock and yield at an array of fish- ing mortality rates. The model can also iterate to MSY. Underlying assumptions of the Walters' model are that the 1) spawner-recruit relationship incorporated is realistic, 2) array of F's accurately reflect the distribution of fishing effort and avail- ability at age, and 3) time increment estimates of weight are sufficiently brief to realistically estimate both population and catch biomass during the fishing periods. The model calculates population biomass, yield, residual spawners of age 2 and greater, and incoming recruitment. We used weight- at-age data described in the section on average size and growth (Equations (1) and (2)), and used the spawner-recruit relationship developed for the number of spawners and recruits (Equation (7)). The instantaneous natural mortality rate was 1.1 as discussed earlier. Fishing mortality could not precisely mimic that for the fishery, because the program requires either zero fishing mortality or a constant fishing mortality for any within-year incre- ment. However, it does allow for an array of multi- pliers at a given fishing mortality, providing dif- ferent F's for each age, if desired. Therefore, we were able to vary fishing mortality by age, but used either zero or a constant fishing mortality for quarterly increments within each year. Since fishing mortality was essentially zero on age-0 fish and was inconsistent between years, a fishing mortality rate of zero was applied to that age group. For age groups 1-4, all of the fishing mortality was by defini- tion imposed equally in quarters 2 and 3 (April- June, July-September), and no fishing mortality was ap- plied in quarters 1 and 4, even though we knew that fishing mortality during the July- September period was consistently higher than that observed for the 900r 800- 700- m O ^ 600 X CO c O 500 o »- ♦- « E 400 I O 1- 300 < o 200 100- 0 100 200 300 400 500 600 700 800 EFFORT (Thousands of Vessel Ton Weeks) Figure 8.— Parabolic (dashed line) and prodfit (solid line) surplus production function models fitted to catch and effort data for the gulf menhaden fishery from 1946 to 1979, with 1964-77 data being estimates of effective effort, based on adjustments from calculated catchability coefficients for those years. 321 FISHERY BULLETIN: VOL. 84, NO. 2 previous quarter (i.e., the same nominal effort was applied to a smaller population). The result was that yield was overestimated for April-June and under- estimated for July-September, but estimated reason- ably accurately for the season. We attempted to simulate reality by using multi- ples of the fishing mortality distribution that we observed in the 1964-77 data base. Fishing mortal- ity imposed to mimic current conditions was ob- tained by taking the mean fishing mortality at age by quarter from the cohort analysis conducted on the 1960-76 year classes (1964-77 fishing years). The mean mortality on ages 2-4 fish was used, along with a mortality obtained from a scaling factor of 0.362 for age 1 (Table 10). Input population size to start the simulation runs was the mean population num- ber-at-age as of 1 January, the arbitrarily assigned birth date of gulf menhaden. Those numbers were 16,030 million, 2,813 million, 227.9 million, and 10.78 million for ages 1-4. The model was run over a range from 0 to 2.75 times the average fishing mortality and was also used to iterate to MSY under the current distribution of fishing mortality by age (Table 10). The overall catch-effort curve from multiple runs indicates that the fishery is operating slightly before the MSY level (Fig. 9, Table 10). At the currents- multiple of 1.0, the fishery should sustain an average yield of about 565,581 t, assuming no variance in recruitment from the hypothetical spawner-recruit curve. The model predicts a MSY of about 585,118 t at 127% of the average fishing mortality for the 1964-77 fishing seasons. We feel that this model, which incorporates a spawner-recruit relationship and recruitment pattern plus growth and natural mortality rates, provides a better estimate of long- term MSY than does a model based on a simple catch-effort production function. Considerable fluc- tuation in yield will result from fluctuations in recruitment, but the long-term MSY estimate appears to be realistic, provided that the esti- mated spawner-recruit relationship is valid and that the basic pattern of recruitment remains unchanged. The Walters' model also identifies the level of fishing mortality at which the population is no longer sustainable, i.e., a biological break-even point. The extinction point occurs at an F- multiple of 2.50 (150% greater than current fishing mortality), al- though the model indicates that such extinction would involve a gradual decline over a period of many years, again assuming that "average" condi- tions prevailed (Fig. 9). Increasing the fishing mor- tality beyond an F-multiple of 2.50 results in a more rapid rate of extinction (Table 10). Results of low and high F-multiple levels show steep slopes on the ascending and descending limbs of the production function curve (Fig. 9). The ascending limb behaves similarly to the curves in the yield-per-recruit model as fishing mortality rates go from low to current levels (Fig. 7). At mortality rates higher than current levels, however, the yield- per-recruit model cannot be used to evaluate poten- tial yield because of the impact of heavy fishing mortality on the spawning stock and the subsequent reduction in recruitment. For example, under the average recruitment level of 16.03 billion fish at age Table 10. — Annual age-specific fishing mortality rates for gulf menhaden, ex- pressed as multiples of the average fishing mortality rate at age, 1964-77, (F- multiple = 1.00), actual fishing mortality rates at age used in the population simulation model, sustainable yield, population biomass, and years to stabilization. f. Actual F at age Sustainable yield level (t) Population biomass (t) Years to stabi- lization multiple 0 1 2-4 0 0 0 0 0 1 ,268,348 97 0.25 0 0.1647 0.4550 266,878 1,151,345 53 0.50 0 0.3294 09100 419,813 1,072,885 35 0.75 0 0.4941 1 .3650 512,568 1,009,288 27 1.00 0 0.6588 1 .8200 565,581 945,740 8 1.25 0 0.8236 2.2750 585,010 871,695 20 1 .27 (MSY) 0 0.8367 2.3114 585,118 865,300 22 1.50 0 0.9883 2.7300 569,823 778,012 32 1.75 0 1.1530 3.1850 514,388 655,278 42 2.00 0 1.3177 3.6400 409,304 492,702 78 2.25 0 1 .4824 4.0950 241,462 277,288 210 2.50 0 1 .6471 4.5500 0 0 1>300 2.75 0 1.8118 5.0050 0 0 '250 1To extinction. 322 NELSON and AHRENHOLZ: CHARACTERISTICS OF GULF MENHADEN 1 (28.34 billion at age 0.5) and an F-multiple of 2.00, the yield-per-recruit model predicts a total yield of 546,395 t; the population simulation model predicts a gradual decline from current levels and stabiliza- tion at about 409,304 t. Thus, when using average recruitment levels and yield-per-recruit results, estimates of yield at F-multiple levels higher than about 1.75 times the average fishing mortality for the 1964-77 period will be unrealistic. The impact of increasing levels of fishing mortality on the stock is also reflected in estimates of popula- tion biomass under an array of F-multiples (Table 10). Biomass estimates were based on predicted population size as of 1 January (i.e., after recruit- ment and before application of fishing mortality). These estimates show a pre-exploitation population biomass exceeding 1.268 million t, followed by an accelerating decline as increased fishing mortality takes progressively larger fractions of the popula- tion and disproportionately larger fractions of older and heavier fish. STATUS AND OUTLOOK FOR THE GULF MENHADEN FISHERY The gulf menhaden population appears to be healthy, highly productive, and capable of supporting yearly harvests exceeding 500,000 t, although con- siderable variation can be expected. It has shown a general increase in abundance through the period covered in this report, although this increase may be a portion of a general cycle of this clupeid stock. The high natural mortality rate indicates that fish- ing mortality has to be applied at a fairly high rate and on young fish to avoid loss of surplus biomass. Peak cohort biomass is reached at an age of 1.5 yr. It is not all available to the fishery, because age-1 fish are only partially recruited. Partial recruitment appears to have some benefit in that it affords some protection for the spawning stock. Recruitment fluctuation appears to be greater at low spawning stock sizes. Initial spawning before full recruitment would assure moderate to high levels of recruitment and reduce chances for large recruit- ment fluctuation. Therefore, if recruitment failure were to occur, it would likely arise from biotic or en- vironmental factors rather than from excessive fishing mortality. Significant increases in fishing mortality are unlikely to occur, given the present distribution and operating procedure of the fishery, unless there is a series of recruitment failures. The current fleet of about 80 purse seine vessels appears to be more than adequate to harvest the recruited gulf menhaden stock during years of low to moderate stock size, and 700- 600- n o 500 X (0 c o •- 400 o — a F 300 •— ' D -1 UJ > 200 100- 025 O50 "0775 1^0 T725 175( F-MULTIPLE F00 05 T50 Figure 9— Sustainable yield predicted by a deterministic population simulation model of the gulf menhaden fishery at multiples of the average fishing mortality (F-multiple = 1.00) for the 1964-77 fishing season (see Table 10 for scaling values). 323 FISHERY BULLETIN: VOL. 84, NO. 2 capable of taking advantage of those years when a large harvestable stock is available (1971, 1978, and 1979). Total mortality rates (averaging 83% for age-1 fish and 95% for ages 2-4 fish) are extremely high. Major expansions of the fleet and processing facil- ities necessary to substantially increase the fishery's share of population biomass would require enormous capital investment. Based on results of the simula- tion model, large increases in fishing effort would also result in an overall average decline in landings that would likely be followed by an economically forced reduction in effort. Under present circum- stances, we do not envision the sustained intensifica- tion of effort necessary to drive the gulf menhaden stock to biological extinction. The simulation model estimates that the effort cur- rently applied in the fishery is probably very close to that which is necessary to produce MSY (Fig. 9), while it exceeds the necessary level in the catch- effort production function (Fig. 8). Assuming that the simulation model reasonably approximates average conditions, some increase in overall yield could be obtained through a modest increase in effort, which has in fact occurred in more recent years. Based on recruitment levels for 1964-77, it is evi- dent that considerable variation will occur around a long-term sustainable yield level, regardless of the level of fishing mortality. We varied recruitment level in the population simulation model through periods of high (25 billion) and low (10 billion) levels of recruitment to provide estimates of the yield from the fishery under good and poor recruitment regimes, and to observe the rate of response to recruitment changes. The results range from an ap- proximate high of 757,000 t to a low of 303,000 t at the high and low recruitment levels (Fig. 10). Since only age-1 and age-2 fish predominate in the fishery, only 2 years were required for the full impact of a change in recruitment to be shown, with a majority of the impact occurring in the first year. We then allowed the average spawner-recruit relationship to operate, stabilizing yield at 565,580 t. Actual low yield predictions are probably underestimated, in that fishing mortality increases in years of low stock size, and the fishery would produce higher yield than through the fishing mortality imposed under average conditions. Nevertheless, these extremes are near the actual ranges in yield observed in the fishery dur- ing the study period (316,100-820,000 t) and should provide reasonable estimates of mean yield and range expected in future years. Since considerable variation does exist around the spawner-recruit curve and simulations were all con- ducted in deterministic fashion, the model was run with recruitment varying randomly between the recruitment extremes calculated from our data set (7.5 billion-25.0 billion). The results of that simula- tion (Fig. 10) provide a long-term (50 yr) average of 467,459 t, but it varies from 718,000 to 263,000 t. We anticipate that the fishery will continue to operate somewhat in this fashion, unless there is a cyclic environmental or biological influence on recruitment. i,uoo- *■■-» n O 800- •*■ X ■ (O c 600- o . t^ +■* F 400- *■ — o _i LU > 200- \/ b. ' ' V/fl. -0- O -6- -O- O -D- O^-O^-oVo 10 15 20 25 30 35 40 45 50 YEARS Figure 10.— Annual yield of the gulf menhaden fishery projected by the population simulation model when upper and lower values of recruitment from the 1964-77 year classes are inserted (dashed line) and when recruitment varies randomly within limits of observed recruitment for the same data set (solid line). 324 NELSON and AHRENHOLZ: CHARACTERISTICS OF GULF MENHADEN SUMMARY The fishery for gulf menhaden appears to be at parity with the stock. There is ample capacity to harvest available biomass and segments of the stock are not available to the fishery until after spawning has occurred. The fishery appears to be near the level of estimated maximum sustainable yield, but will be subject to wide ranges in annual yield. Substantial- ly increased effort will likely reduce long-term average yield, but should not drive the stock to biological extinction. Maintenance of current catch and stock conditions is dependent on the biology of gulf menhaden, the pattern of recruitment, and on maintaining the current fishing strategy. Major changes in the operation of the fishery, such as an expansion of effort east and west of the present range, or offshore on winter spawning concentra- tions, will change the basis on which these analyses were formulated, and would have consequences which are not predictable at this time LITERATURE CITED Abramson, N. J. 1971. Computer programs for fish stock assessment. FAO Fish. Tech. Pap. 101, 154 p. Ahrenholz, D. W. 1981. Recruitment and exploitation of Gulf menhaden, Brevoortia patronus. Fish. Bull., U.S. 79:325-335. Chapoton, R. B. 1972. The future of the Gulf menhaden, the United States' largest fishery. Proc Gulf Caribb. Fish. Inst. 24:134-143. Christmas, J. Y., Jr., and D. J. Etzold. 1977. The menhaden fishery of the Gulf of Mexico United States: a regional management plan. Gulf Coast Res. Lab. Tech. Rep. Ser. 1, 53 p. Clark, C. W., and M. Mangel. 1979. Aggregation and fishery dynamics: a theoretical study of schooling and the purse seine tuna fisheries. Fish. Bull., U.S. 77:317-337. Cushing, D. H. 1971. The dependence of recruitment on parent stock in dif- ferent groups of fishes. J. Cons. Int. Explor. Mer 33:340-362. Fox, W. W., JR. 1975. Fitting the generalized stock production model by least- squares and equilibrium approximation. Fish. Bull, U.S. 73:23-37. Huntsman, G. R., and R. B. Chapoton. 1973. Biostatistical data acquisition in the menhaden fisheries. Trans. Am. Fish. Soc 102:452-456. Lenarz, W. H., W. W. Fox, Jr., G. T Sakagawa, and B. J. Roth- child. 1974. An examination of the yield per recruit basis for a minimum size regulation for Atlantic yellowfin tuna, Thun- nus albacares. Fish. Bull., U.S. 72:37-61. Lewis, R. M., and C. M. Roithmayr. 1981. Spawning and sexual maturity of Gulf menhaden Brevoortia patronus. Fish. Bull., U.S. 78:947-951. Marquardt, D. W. 1963. An algorithm for least-squares estimation of nonlinear parameters. SIAM J. App. Math. 11:431-441. Murphy, G. I. 1965. A solution of the catch equation. J. Fish. Res. Board Can. 22:191-202. Nelson, W. R., M. C. Ingham, and W. E. Schaaf. 1977. Larval transport and year-class strength of Atlantic menhaden, Brevoortia tyr -annus. Fish. Bull., U.S. 75:23-41. Nicholson, W. R. 1978. Gulf menhaden, Brevoortia patronus, purse seine fishery: catch, fishing activity, and age and size composition, 1964-73. US. Dep. Commer., NOAA Tech. Rep. NMFS SSRF 722, 8 p. Nicholson, W. R., and W. E. Schaaf. 1978. Aging of Gulf menhaden, Brevoortia patronus. Fish. Bull., U.S. 315-322. Reintjes, J. W. 1964. Annotated bibliography on biology of menhadens and menhadenlike fishes of the world. US. Fish. Wildl. Serv., Fish. Bull. 63:531-549. Reintjes, J. W., and P. M. Keney. 1975. Annotated bibliography on the biology of the menhadens, genus Brevoortia, 1963-1973. US. Dep. Com- mer., NOAA Tech. Rep. NMFS SSRF 687, 92 p. Reintjes, J W., J. Y. Christmas, Jr., and R. A. Collins. 1960. Annotated bibliography on biology of American men- haden. U.S. Fish Wildl. Serv., Fish. Bull. 60:297-322. Ricker, W. E. 1973. Linear regressions in fishery research. J. Fish. Res. Board Can. 30:409-434. 1975. Computation and interpretation of biological statistics offish populations. Bull. Fish. Res. Board Can. 191, 382 p. Schaaf, W. E. 1975a Status of the Gulf and Atlantic menhaden fisheries and implications for resource management. Mar. Fish. Rev. 37(9):l-9. 1975b. Fish population models: potential and actual links to ecological models. In C. S. Russell (editor), Ecological modeling in a resource management framework, p. 211-239. Resources for the Future, Washington, D.C. Schaefer, M. B. 1954. Some aspects of the dynamics of populations important to the management of the commercial marine fisheries. Inter-Am. Trop. Tuna Comm. Bull. 1:27-56. 1957. A study of the dynamics of the fishery for yellowfin tuna in the eastern tropical Pacific Ocean. Inter-Am. Trop. Tuna Comm. Bull. 2:247-268. Tomlinson, P. K. 1970. A generalization of the Murphy catch equation. J. Fish. Res. Board Can. 27:821-825. Ulltang, 0. 1977. Sources of errors in and limitations of virtual popula- tion analysis (cohort analysis). J. Cons. Int. Explor. Mer 37:249-260. Walters, C. J. 1969. A generalized computer simulation model for fish population studies. Trans. Am. Fish. Soc 98:505-512. 325 LENGTH-WEIGHT RELATIONSHIPS OF BLUE, PARALITHODES PLATYPUS, AND GOLDEN, LITHODES AEQUISPINA, KING CRABS PARASITIZED BY THE RHIZOCEPHALAN, BRIAROSACCUS CALLOSUS BOSCHMA Clayton R. Hawkes, Theodore R. Meyers, and Thomas C. Shirley1 ABSTRACT Length-weight relationships and condition factors of nonparasitized blue king crabs, Paralithodes platypus, and golden king crabs, Lithodes aequispina, in southeastern Alaska were compared with crabs parasi- tized by the rhizocephalan, Briarosaccus callosus. Species, sex, and shell condition were considered in all analyses. Parasitized male blue king crabs and parasitized male golden king crabs weighed significant- ly less than nonparasitized individuals. Golden king crabs may be more resistant to infection and the ef- fects of B. callosus parasitism than blue king crabs. They had a lower prevalence of infection, and the percent difference between the body mass of parasitized and nonparasitized crabs was considerably less. In both crab hosts the prevalence of infection was greater in samples where sublegal or smaller size classes of adults were included in analyses, suggesting that crab growth was reduced by the parasite A parasite of lithodid crab species in Alaska is the rhizocephalan barnacle, Briarosaccus callosus Boschma (Boschma and Haynes 1969; Boschma 1970; McMullen and Yoshihara 1970; Somerton 1981; Hawkes et al. 1985). The parasite's distribution in Alaskan waters, its life history, and its effects on king crab hosts are almost unknown except that parasitized crabs become castrated (Boschma and Haynes 1969; McMullen and Yoshihara 1970). The prevalence of this barnacle parasite varies between areas and species and is especially high in south- eastern Alaska. Parasitism by B. callosus might decrease the productivity of king crab stocks through sterilization and may also reduce crab growth rates. Therefore, parasitized crabs of the same size as nonparasitized crabs may weigh less. In this study we examined the influence of B. callosus on the length-weight relationships and condition fac- tors of parasitized and nonparasitized blue king crab, Paralithodes platypus, and golden king crab, Li- thodes aequispina. MATERIALS AND METHODS Two methods were used to compare the growth of parasitized and nonparasitized crabs. A Fulton's con- dition factor (w/ls x 10 ~4, where w = weight in grams and I = carapace length in mm) was used for 'School of Fisheries and Science, University of Alaska, Juneau, 11120 Glacier Highway, Juneau, AK 99801. comparing different individuals of the same species (Ricker 1975). This method assumes that all body parts grow isometrically. The second method used for comparison assumes allometric growth, where different body parts grow at different rates. Con- stants were determined empirically by linear regres- sion using the model, w = ALB, and logarithms of the carapace lengths and body weights (Everhart et al. 1976, p. 70-71). The length-weight relationships of parasitized and nonparasitized crabs were com- pared with analysis of covariance (ANCOVA). All mean values (X) are given ± 1 standard deviation. Probabilities <0.05 are considered significant and those <0.01 are considered highly significant. The analysis of length-weight relationships was based on wet weights taken in the field (nearest 25 g) and in the laboratory (nearest gram). Crabs with missing or partially regenerated appendages were not weighed. Carapace lengths were measured to the nearest 1 mm (Wallace et al. 1949). Shell condition was classified according to a four point scale (Somer- ton and Macintosh 1983). A new shell condition is found in crabs that have recently molted, and skip- molt crabs are those that have not molted within the last year. Skipmolts or old shell crabs were identified by worn spines and dactyl tips and accumulations of shell epifauna. Infections were diagnosed gross- ly by the presence of externae or scars, indicative of lost externae. A scar is a short chitinous brown pedicel from which an externa was attached and pro- trudes from underneath the abdomen. Manuscript accepted July 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 327 riontni Dui_,j_in,in\: vul. 04, inu. z Blue King Crab Male and female blue king crab of various sizes from Muir and Adams Inlets in Glacier Bay (Fig. 1) were measured, weighed, and examined for B. callosus by the authors in March 1984. Commercial gear was used but with modified escape ports to pre- vent loss of juvenile crabs. Data on large male blue king crabs from Lynn Canal and Glacier Bay were also gathered at dockside areas before sale to processors or the public Since state regulations for southeastern Alaska restrict the commercial harvest of blue king crabs to males M65 mm in carapace width, all commercial samples, therefore, excluded females and smaller adult males. Golden King Crabs Male and female golden king crabs of various sizes were collected by the authors from Lynn Canal near Haines, AK (Fig. 1), using standard pot gear in May 1984. Commercial catches in November 1983 pro- vided legal sized (M78 mm carapace width) males. RESULTS The prevalences of B. callosus in the commercial catches of male blue king crabs were 6.3% and 11.6% for Lynn Canal and Glacier Bay, respectively. Sam- ples from Glacier Bay, which contained males and females of all sizes, had a prevalence of 76%. The prevalence in varisized male and female L. aequi- spina collected from the Haines area was 20%. Linear length-weight relationships of log trans- formed data best defined our data, since no trends were present in the residuals (differences between predicted lines and actual data) of parasitized or non- parasitized crabs. Figure 1.— Sampling sites of blue, Paralithodes platypus, and golden, Lithodes aequispina, king crabs in south- eastern Alaska. 328 HAWKES ET AL.: RHIZOCEPHALAN PARASITISM OF ALASKA KING CRAB Blue King Crab Glacier Bay and Lynn Canal blue king crab data were pooled. The populations were considered to be identical because the two groups were regarded as having the same linear relationship (ANCOVA). Smaller crabs (<134 mm in carapace length) not common to data sets from both areas and skipmolts were eliminated from this analysis. Significantly (chi-square test) more skipmolts were found among the nonparasitized crabs (45/237) than the parasitized crabs (9/131). Because skipmolts tend to be heavier than new shell crabs (Somerton and Macintosh 1983), skipmolting was analyzed as a possible source of bias. In male blue king crabs the new shell crabs had a higher mean weight than the skipmolts at greater carapace lengths, while the skipmolts had a higher mean weight at the smaller lengths (Fig. 2). Although individual linear relation- ships did not describe the data as well as a common line, the skipmolts were eliminated from further analyses of both blue and golden king crab data. Subsequently, in the length-weight relationships of male blue king crabs pooled from both areas, with small crabs represented in each group, the nonpara- sitized crabs were heavier at a highly significant level than the parasitized crabs (ANCOVA) (Fig. 3). Non- parasitized males were 8.7% heavier than parasitized crabs. Nonparasitized male blue king crabs also had a significantly (£-test) higher condition factor (8.5 ± 0.8) than parasitized crabs (7.2 + 0.6), indicating that nonparasitized crabs were heavier for a given length. Condition factor did not vary with size in non- parasitized blue king crabs but the slope was sig- nificant and negative for the parasitized crabs. This indicates that the condition factor of parasitized blue king crabs decreased with increased size Only five nonparasitized female blue king crabs were available for length-weight relationships and condition factor comparisons. More samples are needed for further analysis of female blue king crabs. Golden King Crabs Males with carapace lengths common to both para- sitized and nonparasitized crabs, 117 to 159 mm, pro- vided linear relationships that were parallel and significantly different (Fig. 4). Briarosaccus callosus was not present in any of the large commercial-size crabs sampled in 1983; therefore, these samples were excluded from analysis. The percent weight dif- ference between parasitized and nonparasitized male golden king crabs was about 2.6%. Weight conver- sion in parasitized male P. platypus of similar sizes 4000- 3500 O) 3000 O) 2500 2000J Old Shell lognY = -5.04 + 2.60loghX r*=0.70 New Shell lognY=-6.54 +2.89lognX H =0.68 n i 181 130 135 140 ^45 150 15JF Length (mm) 160 Figure 2— Length-weight linear relationships of new shell and skipmolt nonparasitized male Paralithodes platypus. 329 FISHERY BULLETIN: VOL. 84, NO. 2 O) '3 5 4000 3000 2000 1000 Nonparasitized I09.Y-- lognX 100 110 120 130 140 ISO 160 170 180 Length (mm) Figure 3— Length-weight linear relationships of parasitized and nonparasitized male Paralithodes platypus with skipmolts eliminated. 2500 ^ 2000 O) O) 1500 1000 log, U 150 160 120 130 140 Length (mm) Figure 4— Length-weight linear relationships of parasitized and nonparasitized male Lithodes aequispina after elimination of 1983 data. 330 HAWKES ET AL.: RHIZOCEPHALAN PARASITISM OF ALASKA KING CRAB was inhibited considerably more than in parasitized male L. aequispina. The condition factor for non- parasitized male L. aequispina (6.5 ± 0.5) was also greater at a highly significant level than for male parasitized crabs (6.1 ± 0.4). The condition factor in parasitized and nonparasitized male golden king crabs did not vary significantly with size Nonparasitized female L. aequispina (n = 77) were heavier than parasitized females (n = 43) over most of the length range The linear relationships were significantly different but not parallel, preventing a comparison of the intercepts. Condition factors were not significantly different between the parasitized (5.9 + 0.5) and nonparasitized (5.7 + 0.4) females. Condition factors varied significantly with size and in the nonparasitized crabs but not in the parasitized crabs. DISCUSSION Weights and, consequently, condition factors were significantly lower in male blue and golden king crabs parasitized by B. callosus. A difference in mean weight was also present in female blue king crabs that were parasitized, although an adequate com- parable sample size of nonparasitized females was not available The prevalence of the parasite was con- siderably greater in king crab populations where sub- legal or smaller size classes of adult crabs were in- cluded in the sample number. In blue king crabs from Glacier Bay, the inclusion of females in the sample also raised prevalence figures since females had a significantly higher prevalence of barnacle para- sitism than male crabs. A potential reason for in- creased barnacle prevalence in smaller crabs could include differential mortality such that fewer parasitized crabs survive to larger size classes. Other explanations include reduced molting frequencies, reduced number of instars and/or reduced growth represented by a reduction in relative molt increment (Hawkes et al. in press). However, reduced weights in parasitized crabs within the same size classes as nonparasitized individuals suggest that growth of the host crab is decreased by B. callosus. The higher parasite prevalence in smaller crabs also supports this conclusion. Parasitized crabs may develop significantly less body tissue after molting, which is likely to be a cumulative effect occurring over more than one season. Although the complete life history of B. callosus is unknown, other species of Rhizocephala are known to require at least 9 to 12 mo to reach reproductive maturity and develop an externa in host crabs (Ritchie and H0eg 1981). In males that be- come castrated and weight loss of testes is insignif- icant in total body mass (0.2%) and does not account for the weight difference observed. Also testes weigh less than the interna and externa of the parasite In female king crabs a considerable amount of the wet body weight can be attributed to the egg clutch and ovaries. Consequently, gonadal atrophy, nonovigerous conditions and reduced somatic growth rates all may account for the lesser weights observed in parasi- tized female king crabs. The percentages of weight difference between parasitized and nonparasitized males was con- siderably different between the two species of king crabs. Golden king crab was less affected by the parasite, sustaining less growth inhibition due to bar- nacle parasitism than parasitized blue king crabs. Parasitized golden king crabs have significantly higher hemolymph protein concentrations in com- parison to either their nonparasitized conspecifics or parasitized blue king crabs. The additional pro- tein may be attributed to the presence of lectins, specific carbohydrate-binding proteins suspected of playing a role in crustacean immunity (Shirley et al. 1985). If we are correct, reduced crab growth as an ef- fect of B. callosus parasitism would conflict with data from other peltogastrid rhizocephalans (O'Brien and Van Wyk 1985). Other rhizocephalan species tend to be more prevalent on larger crab hosts, making enhanced growth or enhanced survivorship a plaus- ible effect of parasitism. Another explanation is that parasitized crabs have less somatic growth and, as a result, have fewer molts. Molting is a time of greatest mortality for most decapods, and those with lower molting frequencies would have greater sur- vival. The probability of infection may also be greater in certain size classes. Behavioral differences or sampling bias could affect the parasite's relative fre- quency within the host population. Sacculinidae ap- pear to be distributed differently within host popula- tions (O'Brien and Van Wyk 1985). Pugettia producta is a majid crab from California and does not molt after reaching maturity. When parasitized by the rhizocephalan Heterosaccus californicus, there is no significant effect on molt increments of juveniles and the pubertal molt increment is not affected in adults. However, P. producta that are parasitized pass through fewer instars before reaching maturity, and the mean size of these individuals is significantly less than in nonparasitized crabs (O'Brien 1984). Blue crabs, Callinectes sapidus, also have retarded growth when parasitized by Loxothylacus texanus, with most adults appearing as miniature adult females (Over- street 1978). 331 FISHERY BULLETIN: VOL. 84, NO. 2 Prevalence of the parasite as a function of host size and field length-weight comparisons are still only in- direct measurements of host growth. Consequently, further laboratory studies measuring growth directly in parasitized king crabs are needed to positively prove our hypothesis. ACKNOWLEDGMENTS We thank R. Hakala of the FV Fortune and J. Donahue of the FV Stormfront for assistance with field sampling and use of their vessels. Steve Ignell of the National Marine Fisheries Service's Auke Bay Laboratory was also very helpful as a consultant on statistical procedures. Funding was received from the Research Council of the University of Alaska, Juneau, and the Alaska Fisheries Research Center of the University of Alaska (Project number RC/ 84-04). A research fellowship was provided by Alaska Sea Grant for the senior author (CRH). LITERATURE CITED BOSCHMA, H. 1970. Notes on Rhizocephala of the genus Briarosaceus, with a description of a new species. Proc, K. Ned. Akad. Wet. C73:233-242. BOSCHMA, H., AND E. HAYNES. 1969. Occurrence of the rhizocephalan Briarosaceus callosus Boschma in the king crab Paralithodes camtschatica (Tilesius) in the Northeast Pacific Ocean. Crustaceana 16: 97-98. EVERHART, W. H., A. W. ElPPER, AND W. D. YOUNGS. 1976. Principles of fishery science Comstock Publ. Assoc., Ithaca, N.Y., 288 p. Hawkes, C. R., T. R. Meyers, and T. C. Shirley. 1985. Parasitism of the blue king crab, Paralithodes platypus, by the rhizocephalan Briarosaceus callosus Boschma. J. In- vertebr. Pathol. 45:252-253. In press. Growth of Alaskan blue king crabs, Paralithodes platypus (Brandt), parasitized by the rhizocephalan Briaro- saceus callosus Boschma. Crustaceana. MCMULLEN, J. C, AND H. T. YOSHIHARA. 1970. An incidence of parasitism of deepwater king crab, Lithodes aequispina, by the barnacle Briarosaceus callosus. J. Fish. Res. Board Can. 27:818-821. O'Brien, J. 1984. Precocious maturity of the majid crab, Pugettia produc- ta, parasitized by the rhizocephalan barnacle, Heterosaccus californicus. Biol. Bull. 166:384-395. O'Brien, J., and P. Van Wyk. 1985. Effects of crustacean parasitic castrators (epicaridean isopods and rhizocephalan barnacles) on growth of crusta- cean hosts. In A. Wenner (editor), Crustacean issues 3, Fac- tors in adult growth, p. 191-218. A.A. Balkema Pubis., Rot- terdam, Neth. Overstreet, R. M. 1978. Marine maladies? Worms, germs, and other symbionts from the Northern Gulf of Mexico. Miss. -Ala. Sea Grant Consortium, MASGP -78-021, 140 p. RlCKER, W. E. 1975. Computation, and interpretation of biological statistics of fish populations. J. Fish. Res. Board Can. Bull. 191, 382 p. Ritchie, L. E., and J. T. H6eg. 1981. The life history of Lemaeodiscus porcellanae (Cirri- pedia: Rhizocephala) and co-evolution with its procellanid host. J. Crustacean Biol. 1:334-347. Shirley, S. M., T. C. Shirley, and T. R. Meyers. 1985. Hemolymph studies of blue (Paralithodes platypus) and golden (Lithodes aequispina) king crab parasitized by the rhizocephalan barnacle, Briarosaceus callosus. In Pro- ceedings of the International King Crab Symposium, p. 341-352. Alaska Sea Grant Rep. 85-12. SOMERTON, D. A. 1981. Contribution to the life history of the deep-sea king crab, Lithodes couesi, in the Gulf of Alaska. Fish. Bull., U.S. 79:259-269. SOMERTON, D. A., AND R. A. MACINTOSH. 1983. Weight-size relationships for three populations in Alaska of the blue king crab, Paralithodes platypus (Brandt, 1850) (Decapoda, Lithodidae). Crustaceana 45:169-175. Wallace, M., C. J. Pertuit, and A. R. Hvatum. 1949. Contribution to the biology of the king crab, Para- lithodes camtschatica Tilesius. U.S. Fish. Wildl. Serv. Leafl. 340, 50 p. 332 DISTRIBUTION AND ABUNDANCE OF COMMON DOLPHIN, DELPHINUS DELPHIS, IN THE SOUTHERN CALIFORNIA BIGHT: A QUANTITATIVE ASSESSMENT BASED UPON AERIAL TRANSECT DATA Thomas P. Dohl, Michael L. Bonnell, and R. Glenn Ford1 ABSTRACT On 35 aerial transect surveys of the Southern California Bight, 157 sightings of common dolphin, Delphinus delphis, schools were observed and mapped for distributional analysis. Sightings were pooled into 30' of latitude by 30' of longitude sampling quadrats, and density estimates were obtained by fitting a Fourier series to a frequency distribution of perpendicular sighting distances. Two distinct seasonal distributions are represented by density contour maps: a winter-spring distribution when schools were confined to the easternmost and warmest waters of the area, and a summer-autumn distribution when schools were widespread. Mean seasonal population estimates were 15,448 for winter-spring and 57,270 for summer-autumn (cv of 0.36 and 0.17, respectively). During the warmer water months, the common dolphin population expands its use of the Southern California Bight. They enter from the south, apparently following the major undersea ridges and escarpments, and flow through the Southern California Bight in a generalized counterclockwise fashion. Observational evidence suggests that there is mixing of both the nearshore and pelagic forms of this species in the offshore waters over the Santa Rosa-Cortes Ridge and Patton Escarpment. The common dolphin, Delphinus delphis, is the most abundant cetacean in the waters of the Southern California Bight (SCB). On an annual basis the num- bers of common dolphins exceed, on average, the combined total of all other cetaceans in this area by 2.75 times (Dohl et al. 1980). Common dolphins inhabit subtropical waters of Mexico and the SCB throughout the year (Norris and Prescott 1961). Density estimates for this species and other dolphins (Stenella sp.) in waters offshore of Mexico and Central America were calculated by the National Marine Fisheries Service in 1974 (Smith 1981). The distribution of common dolphins in the eastern parts of the Southern Califor- nia Bight was described by Evans (1975). In order to understand the role of the common dolphin in the ecology of the SCB and to understand when and where this population is mostly vulnerable to human activities, we have constructed a spatial- seasonal distributional model with two aims: 1) to generate population estimates for the entire area and 2) to describe the general features of seasonal distribution patterns. This is the first study to ex- amine the spatial heterogeneity of common dolphin distribution in the SCB and to generate confidence limits for density and seasonal mean population size estimates. 'Institute for Marine Sciences, University of California, Santa Cruz, CA 95064. Manuscript accepted July 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. From April 1975 through March 1978, nearly 110,000 nmi (200,000 km) of combined aerial and ship surveys were conducted within the SCB for the Department of the Interior, Bureau of Land Man- agement (now the Minerals Management Service). During this marine mammal and seabird study, a total of 505 schools of 134,675 Delphinus delphis were recorded. This paper is primarily concerned with one subset of the 3 yr, common dolphin sighting data base. To avoid the statistical pitfalls of pooling data obtained from a variety of platforms performing their mis- sions at different speeds, at different altitudes, and over varying portions of the study area, we re- stricted these analyses to 35 monthly flights flown at 1,000 ft above sea level (ASL). Each of these surveys required about 15 overwater flight hours and covered about 1,350 nmi (2,500 km) of track- line. All species of cetaceans encountered were recorded as to location, number, behavior, direction of movement, and number of juveniles. Common dolphins were encountered 157 times in this flight series, for a total of 46,153 animals or 69% of all cetaceans observed. The results of the distributional study and accom- panying figures were derived from the 1,000 ft ASL aerial survey data defined above. However, material in the Discussion section draws upon observations made from all survey platforms used during this study. 333 FISHERY BULLETIN: VOL. 84, NO. 2 METHODS Aerial surveys were flown at an altitude of 1,000 ft ASL (328 m) at about 90 kn (167 km/h) in a high- wing, twin-engine Cessna2 337. The crew consisted of a pilot and three experienced marine mammal observers, one acting as recorder. Surveys were flown along 15 parallel, predetermined tracklines, separated by 15 nmi and extending from the shore to a maximum distance of 100 nmi (185 km; Fig. 1). Tracklines were oriented from northeast to south- west and were roughly perpendicular to the shore- line, as well as to most major features of submarine topography in the study area. Whenever possible, all transect lines were surveyed on each 3-d flight. Transect lines were not replicated on a single survey, nor were they flown in a predetermined order or direction. The first line flown on a given day was oc- casionally dictated by weather or military activity in the area; subsequent lines were chosen to optimize coverage and simplify logistics. Observers searched unbounded corridors on each side of the aircraft trackline Sightings were recorded 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. and coded for computer entry at the time of occur- rence The aircraft was diverted to circle those schools located off the trackline for positive iden- tification, animal count, and photographs. The total animal count recorded for each school was a consen- sus of the observers on board, derived from multi- ple orbits of the school. Any additional sightings ob- tained while "off transect" were not included in later density calculations due to the possibility that the secondary sighting was prompted by the first. All transect segments where observer effectiveness might have been hampered by fog and/or sea state were deleted from the data base; only transect seg- ments where visibility exceeded 1 nmi and the sea state was Beaufort 3 (few, scattered whitecaps) or less were retained. Aerial photographs were used to validate observ- er estimates of school size. The aerial photographs were taken on 9" x 9" film from a vertically mounted camera and on 4" x 5" and 35 mm films in hand- held cameras for oblique views. The large, 9" x 9" vertical photographs soon proved to be the most useful and were used almost exclusively for count verification. Observer counts and film counts on average-sized schools (up to 100 animals) varied only slightly, but not in a consistent manner. The 3-5% 121 120" 119" 118" 117" SOUTHERN CALIFORNIA BIGHT I n 120 119" 118" Figure 1.— Map of the Southern California Bight study area showing aerial survey tracklines. 334 DOHL ET AL.: COMMON DOLPHIN DISTRIBUTION AND ABUNDANCE variations in counts occurred randomly, with no pat- tern to indicate in which method the higher counts would occur. Small schools of <100 animals repre- sented most of the sightings (53%). In medium-sized schools, up to 300 animals, the variation was higher (about 11%), and the photographs indicated probable observer underestimation in 62% of the counts. The largest underestimates occurred in large schools, >300 animals, and were found in 76% of the observ- er counts. These underestimates ranged up to 30% in some circumstances. Within the large-school category, two subcategories became evident: 1) Dis- persed schools with multiple discrete subgroups of animals gave the observers less of a problem than 2) the tightly grouped, rapidly moving, compact schools. The dispersed large schools yielded under- estimate values in the range of 14-16%, while the compact, large groups were usually 21-23%. Ex- tremely large schools of over 1,000 animals were responsible for the highest error values of up to 30%; these schools accounted for only 6.6% of total sightings. Generally, we found that aerial estimates were lower than numbers based on photographs and that the larger the school, the higher the difference. We attribute some of the difference to the time lag be- tween when the count was made while circling the school and the photo run over the center of the school. Results of photo runs made either before or after the counting effort did not vary significantly, but occasionally, continued circling scattered larger schools into several smaller subgroups. Sea surface glare affected observation efficiency to some degree on about 10% of all survey days. Due to the orientation of transect lines, glare conditions could impair the search ability of only the left-side observer on southwest-bound legs (up to 26% of total search effort per survey day). Holt (19843) found density estimates of dolphin schools to be 39% lower under poor sun conditions than during good sun con- ditions. Using his figure, we calculate that our over- all seasonal density estimates might be low by about 1%. Because of the lack of any systematic bias resulting from glare affecting density estimates in one particular region or season more than another, we made no corrections to adjust for this slight underestimate. The perpendicular distance from the trackline to the sighting was calculated from the declination angle obtained using a hand-held inclinometer. Per- pendicular distances were recorded for 112 sightings of common dolphin schools, representing 74.2% of all sightings used in density calculations. Distributional Model Inspection of the first year's common dolphin sighting numbers and plots of monthly distribution indicated seasonal fluctuations of residency within the Southern California Bight. Examination of the 3-yr database showed two distinct seasons of occupancy for the species in the SCB (Fig. 2). A comparison of the two sets of data on a monthly basis show a significant statistical dif- ference (^(1,34) = 7.66, P < 0.01). In view of these observations, two seasons were defined for the development of the distributional model: a summer- autumn season (July through December) when com- mon dolphin sightings were widespread in the SCB, and a winter-spring season (January through June) when most schools were confined to the southeast- ern portion of the surveyed area. Common dolphin sightings were assigned by their latitude and longi- tude to 30' x 30' grid-cells (sampling quadrats) centered on degree and half-degree lines of latitude and longitude. Data were pooled to provide seasonal estimates of common dolphin abundance for each 30' x 30' grid-cell. The estimate of density of groups in cell i, Dh was calculated from the relationship: Di = nlf{Q>)IZLl (Burnham et al. 1980) (1) 3Holt, R. S. 1984. Testing the validity of line transect theory to estimate density of dolphin schools. U.S. Dep. Commer., NOAA Admin. Rep., NMFS-SWFC LJ-84:31, 56 p. where n{ is the number of groups encountered, /(0) is the probability density function of perpendicular distances evaluated at the ^-intercept, and h% is the sum of all transect lengths in cell i contributing to the seasonal estimate. The value of the f(0) term was calculated using the nonparametric Fourier- series estimator of Crain et al. 1978 (see Burnham et al. 1980 for a complete discussion of this esti- mator). Computations were made employing the program TRANSECT (Laake et al. 1979). For calcu- lation of the/(0) term, the perpendicular distance of each sighting was reduced by one-half the width of the exclusion area under the aircraft, where visibility was obstructed by the fuselage (total ex- clusion area = 530 ft at 1,000 ft ASL). This ap- proach, in effect, moves the transect centerline out- board to the point of nearest possible sighting distance— a point where it is assumed that all animals present will be seen and counted. The ques- tion of how to deal with the problem of restricted downward visibility and line transect theory has been considered by others; however, the best treat- 335 FISHERY BULLETIN: VOL. 84, NO. 2 7000 6000 5000 4000 3000 2000 1000 1 = Winter (Jan., Feb., March) 2 = Spring (April, May, June) 3 - Summer (July, Aug., Sept.) 4 = Autumn (Oct., Nov., Dec.) Figure 2.— Comparison of total counts of common dolphins on aerial surveys of the Southern California Bight by season, 1975-78. ment of the subject, in print, is found in two papers by Leatherwood et al. (1982, 1983). Because sample size was small in each grid-cell and in each season, data were combined to calculate a single value of f(0). The pooling of data was based on the assumption that the sightability of common dolphin groups did not vary between seasons or be- tween regions of the surveyed area. Violation of this assumption would lead to biases in the estimates of relative densities between seasons or regions, al- though it would not necessarily effect mean popula- tion size estimates. The assumption of seasonal homogeneity was tested using a single classification ANOVA (two groups, unequal samples; Sokal and Rohlf 1969, p. 208). No significant difference be- tween the distribution of perpendicular sighting distances collected in summer-autumn and winter- spring seasons was found (F1(111 = 2.01, P = 0.18). The same test was used to compare frequency distributions with distance of sightings collected in calmer inshore waters, with sightings collected in rougher offshore waters, since this seemed to be the most likely source of bias in sightability. No signifi- cant difference was found between the distribution of perpendicular sighting distances in the two sub- regions (F1>108 = 1.78, P = 0.20). The rescaled frequency distribution of perpendi- cular sighting distance is shown in Figure 3. The probability density function, f(x), is from a three- term, Fourier-series model, which provides the best fit to these data (x2 = 6.026, df = 3, P = 0.11). Data were truncated at 6,600 ft in order to remove two extreme values. Intervals were specified, by in- spection of the data, in order to smooth the func- tion and minimize the effects of "heaping" in per- pendicular distance measurements (Burnham et al. 1980, p. 47). For estimation of common dolphin density (ani- mals/km2) in a given grid-cell for a given season, we multiplied the density of groups in a given cell 336 DOHL ET AL.: COMMON DOLPHIN DISTRIBUTION AND ABUNDANCE by the mean group size throughout the SCB ob- tained for that season. The small sample size in any cell and the very large variability in the size of groups necessitated pooling of all sightings within a season to calculate mean group size. The mean group size in summer and autumn was 338 ± 38 SE (n = 115), while that of winter and spring was 231 ± 73 SE (n = 36). While not significantly different (^1,149 = 1-42, P > 0.25), we used separate mean group size in calculations of seasonal abundance. We tested the assumption that mean group size in each season was constant throughout the SCB, using a bootstrap procedure (Efron 1982). For a given season, cell i contained n{ observations of groups of mean size s{. For each cell i, we randomly drew 10,000 sets of values of size n{ from the group size distribution based on all observations recorded in that season, computed the mean of this subsample, and formed a frequency distribution of these mean values. If the percentile ranking of the observed mean group size in cell i was >97.5% or <2.5%, s, was assumed to be a nonrandom sample. For the summer-autumn season, only 1 cell of the 26 cells containing observations of common dolphins had means which differed significantly from the rest of the surveyed area. Similarly, for the winter-spring season, only 1 cell in 10 showed a significant dif- N = 112 .13 .27 .40 .67 .94 DISTANCE IN KM 1.22 1.48 Figure 3.— Probability density function f(X) fit to histogram of sighting frequency and perpendicular distance (rescaled; see text). ference from the overall group size distribution. Therefore, group size homogeneity was assumed for these data, and a single seasonal value of mean group size (s) was used in all calculations of cell density for each season. If f(0) and s may be assumed to be homogeneous, the remaining source of between-cell variability is the density of groups. We tested the hypothesis that the density of groups is homogeneous through the SCB as follows: taking the mean number of sight- ings of common dolphin schools per kilometer of transect for the entire surveyed area, A*. We com- puted the expected number of cells containing a specified number of sightings of groups, using the formula: [Expected number of cells with k sightings] = 1 e-rL. (A*L^ (2) i=i where m is the total number of cells sampled, k is the specified number of sightings of groups, and L% is the length of trackline surveyed in cell i. The ex- pected number of cells containing k sightings were compared with the observed number for all k using a chi-square test. No significant spatial heteroge- neity was evident for data collected in summer and autumn (x2 = 5.06, df = 5, P > 0.5). However, the winter and spring distribution showed clear heter- ogeneity in the density of groups by cell (x2 = 12.85, df = 3, P < 0.005). We used the method of Chernoff and Moses (1959) to place confidence limits on the estimate of the number of groups per km of transect in cell i, A; (see also Clopper and Pearson 1934). We used a com- puter program which finds a density value, r1( such that the probability of observing n% or more groups in a transect segment of length L{ is 0.025; this is the lower confidence bound on A,. Similarly, we find a density value, r2, such that the probability of ob- serving n{ or fewer groups is 0.025; this forms the upper bound on A^. Tx and V2 are defined as satis- fying the equations: and k = n. k = n k\ = 0.025 (3) 2. \J_±L = o>025. *=o k\ (4) 337 FISHERY BULLETIN: VOL. 84, NO. 2 Such confidence limits are asymmetric about Xx and decrease in size with increasing transect coverage. They have the important properties that r2, the up- per limit, tends to be large when the transect length L; is small, even when the number of groups ob- served is zero, and the lower limit r1 is bounded by zero. Population size estimates were made for each cell i in each season from the relationship N{ = D{ • s • A{, where N{ is the cell population, Dt is the estimated density of groups based on Equation (1) (groups/km2), s is the seasonal mean group size, and A{ is the open-water area of cell i. Total population size in each season, (N), was estimated as from the sum of populations in each cell, and from the theoretical formula: -T w/(0) _ . N = J -s-A 2L (5) where n is the total number of groups observed, L is the total transect length, s is the seasonal mean group size, and A is the areal extent of the study area. The variance of N was estimated from the rela- tionship (K. Burnham4). var (N) = A2 ■ var 0,) (6) where var 0t) = 0,)2 var(s) (E(i)f var(n) var(/"(0)) (E(n)f (£(A0)))2 + . The variance of n was calculated assum- ing that n had a Poisson distribution; if this assump- tion holds, var(n) = n (Burnham et al. 1980). The variance of /(0) was calculated by program TRANSECT, using the method of Burnham et al. (1980). Variance of s was estimated as the standard error of the mean group size. The formula for variance requires that/(0) and s be independent, an assumption that may be violated due to the diff eren- 4K. Burnham, Department of Statistics, School of Physical and Mathematical Sciences, North Carolina State University, Raleigh, NC 27650-5457, pers. coramun. Figure 4.— Common dolphin distribution in the Southern California Bight, winter and spring, 1975-78. Density contours show animals/km2. 338 DOHL ET AL.: COMMON DOLPHIN DISTRIBUTION AND ABUNDANCE tial sightability of large and small groups (discussed below). Because we could not be sure that the assumptions of the theoretical formula were met, we also calculated the variance of population size for the summer-autumn season, using a jackknife estimator (Miller 1974; Burnham et al. 1980). Pseudovalues of the area-wide population were generated by sequentially deleting pairs of surveys from the database. All sources of variance were con- sidered in estimation of total variance: /(0), mean group size, and spatial variability of sightings. Because of the small number of perpendicular sight- ing distances for winter-spring season (31), we were unable to obtain a stable value of /(0), thus pre- cluding the estimation of jackknife variance of that season. Distribution maps were prepared using Surface Display Library software (Dynamic Graphics, Inc., Berkeley, CA). Contour lines, generated by linear interpolation between density values assigned to grid-cell centerpoints, were smoothed using a cubic spline function. RESULTS Two distinct seasonal distributions were found for common dolphins in the Southern California Bight (SCB). In winter and spring months (January through June), common dolphin sightings were almost completely confined to the eastern part of the SCB (Fig. 4). Within the area occupied, three cells in the southernmost rank and one shore- bounded cell north of San Diego showed significantly higher density than the overall seasonal mean (P > 0.95 in all cases). In summer and autumn months (July through December), common dolphin sightings were widespread from Rodriguez Seamount and the Patton Escarpment in the west to the mainland shore in the east (Fig. 5). Cell density estimates in this season were relatively homogeneous through- out the area. Only a single cell in the San Diego Basin could be shown to be significantly higher than the seasonal mean at the P > 0.95 level. Neverthe- less, we believe that the clustering of moderately high-density cells east of Santa Catalina and San Figure 5.-Common dolphin distribution in the Southern California Bight, summer and fall, 1975-78. Density contours show animals/km2. 339 FISHERY BULLETIN: VOL. 84, NO. 2 Clemente Islands and west of San Nicolas Island represents a real distributional pattern. Cell-density estimates and 95% confidence limits are provided in Tables 1 and 2. Confidence limits were calculated considering only sampling error due to number of groups sighted (Equations (3) and (4)) and not uncertainty in/(0) or mean group size. Sam- pling error associated with the number of groups sighted was the dominant source of variation in cell- by-cell estimates of density, typically exceeding variance of the/(0) term by three times and variance associated with mean group size by five times. It should be remembered that the density estimates are mean values computed from pooled data col- lected over a several month period in 3 successive years. From these density estimates, we computed seasonal mean population size estimates. By cal- culating population size as the sum of the numbers in each 30' x 30' cell, we estimate a winter-spring population of 15,448 animals. This figure is a mean population occurring in the months of January through June and includes months of higher and lower numbers. Using Equation (5), we calculate a theoretical winter-spring population size of 18,933 animals. This second estimate for the SCB, based on pooled data, may be high because survey effort was 6.7% greater in the higher density parts of the study area in winter and spring. Based on Equation (6), the coefficient of variation of the winter-spring population was 36%. The coefficients of variation for number of groups, f(0), and mean group size were 16%, 8%, and 31%, respectively. The relatively large variability in mean group size was due to a Table 1.— Relative abundance of common dolphins in the winter and spring. Mean density (animals/ km2) is provided for each 30' x 30' cell; latitude and longitude indicate center point of cell. Upper and lower values are 95% confidence limits derived from the spatial variability of sightings along aerial transect lines. 121 °00' 120°30' 120°00' 119°30' 119°00' 118°30' 118°00' 117°30' 7.41 1.71 1.33 2.42 34°30' 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3.42 1.05 0.81 0.62 0.86 1.28 34°00' 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 5.32 1.19 0.90 1.05 0.48 1.09 1.81 33°30' 0.00 0.00 0.00 0.19 0.10 0.43 0.00 0.00 0.00 0.00 0.05 0.00 0.19 0.00 4.89 1.71 0.76 1.14 1.38 2.00 33°00' 0.00 0.48 0.00 0.33 0.48 0.95 0.00 0.14 0.00 0.10 0.19 0.48 1.24 0.71 1.95 2.57 2.80 32°30' 0.00 0.00 0.76 1.09 1.38 0.00 0.00 0.33 0.48 0.67 Table 2.— Relative abundance of common dolphins in the summer and fall. Mean density (animals/km2) is provided for each 30' x 30' cell; latitude and longitude indicate the center point of cell. Upper and lower values are 95% confidence limits derived from the spatial variability of sightings along aerial transect lines. 121°00' 120°30' 120°00' 119°30' 119°00' 118°30' 118°00' 117°30' 34°30' 34°00' 33°30' 33°00' 32°30' 3.53 1.45 1.32 3.12 0.00 0.00 0.00 0.05 0.00 0.00 0.00 0.14 1.80 1.04 1.25 1.52 2.15 2.08 0.35 0.21 0.35 0.62 0.83 0.35 0.07 0.07 0.07 0.28 0.35 0.07 5.82 2.70 1.73 2.70 2.15 1.25 2.56 1.04 1.04 0.62 1.25 1.25 1.42 0.48 0.28 0.42 0.21 0.55 0.76 0.14 0.14 4.92 2.29 2.91 2.56 4.09 2.98 1.66 0.76 1.32 1.25 2.56 1.66 0.62 0.28 0.62 0.62 1.59 0.90 2.49 2.08 1.94 3.39 3.12 0.69 0.83 0.69 1.45 1.66 0.21 0.35 0.21 0.62 0.90 340 DOHL ET AL.: COMMON DOLPHIN DISTRIBUTION AND ABUNDANCE single sighting of 2,450 animals; we choose not to treat this observation as an outlier because the oc- casional occurrence of very large groups is typical of this species. For the summer-autumn season of greatest abun- dance, the stock size estimate based on summing in- dividual cell populations and the estimate derived from Equation (5) were 57,270 and 46,675, respec- tively. The theoretical estimate based on pooled data may be low because survey effort was 7.8% greater in the lower density parts of the study area in the summer-autumn season (i.e., the offshore waters in the west). The coefficient of variation computed from the theoretical variance formula (Equation (6)) was 17%. Coefficients of variation for number of groups, /(0), and mean group size were 9%, 8%, and 11%, respectively. The jackknife estimator gave a higher coefficient of variation for population size of 27%. Components of this estimate for number of groups, /(0), and mean group size were 15%, 18%, and 14%, respectively. Differences between the two types of estimators may be due, in part, to the in- herently conservative nature of the jackknife (Efron 1982), but probably result primarily from within- survey correlation of variables. In addition, the jack- knife estimate of /(0) relied on a smaller subset of sighting distances measured only during summer- autumn surveys (n = 81). DISCUSSION Even in an area as heavily utilized as the South- ern California Bight, sightings of common dolphin schools are not common events. For this reason it was necessary to pool aerial survey data collected over several months in each of three years to describe their distribution in statistical terms. The two seasonal views of common dolphin distribution in the SCB are shown for contrast in Figures 4 and 5. It is apparent that the population makes season- ally greater use of the SCB in summer and autumn months. The months of greatest numbers, based on sightings per km of trackline, were September through November. During these months, the popu- lation far exceeds the mean value of 57,000 and probably approaches 100,000 animals. A potential source of bias in our mean population size estimates was the differential sightability of groups of various sizes. The detection function for common dolphin sightings declined sharply beyond about 1,650 ft (500 m), suggesting that mostly large or conspicuous groups were seen at relatively great distances. The Fourier estimator is robust to varia- tion in sighting efficiency (Burnham et al. 1980). For comparison, the/(0) term of 2.29 for common dol- phins was quite close to the/(0) estimate of 2.16 more recently obtained for 136 sightings of Pacific white-sided dolphin schools on aerial surveys off- shore of central and northern California (Dohl et al. 1983). However, variable sighting effectiveness may also bias the estimation of mean group size. Holt and Powers (1982) found that smaller groups of dolphins were more likely to be missed on aerial surveys than larger groups, resulting in a 25% overestimation of mean group size. For our data on common dolphins, we did not find a significant difference in mean group size between sightings within the first 1,650 ft and beyond due to high variability in sightings size (213 ± 46 SE, n = 65, compared with 308 ± 49 SE, n = 50; Fin3 = 1.94, P = 0.18). Nevertheless, our calculations show that stratification of mean group size by distance from the trackline (< 1,650 ft and >1,650 ft) would result in an 18% decrease in mean density values. The distribution shown for summer and autumn can be viewed as a composite of monthly distribu- tions. Common dolphin distribution expands from the southeast into the central and western parts of the SCB in late spring and early summer and recedes toward the east and south in late autumn and early winter. Common dolphin movement into and out of the SCB appears to be temperature related. As sea surface temperatures (SST) rise in late spring-early summer, animals begin to be sighted more often along the Coronado Escarpment. Peak numbers of common dolphins were found in open water regions of the SCB 3-5 wk after intru- sion of the warmer waters. During cool-water months, when SSTs down to 10.0°C were recorded and the SCB-wide mean was 14.6°C, no animals were observed in waters cooler than 14.0° C. Distributional patterns of the common dolphin within the SCB may be changing. Hui (1979) ana- lyzed data collected on Naval Ocean Systems Center (NOSC) surveys from 1968 through 1976 and showed no common dolphin sightings north of Point Vincente (lat. 33°45'N) or west of approximately San Nicolas Island. Our surveys in summer and autumn months found 29.9% of all sightings and 30.8% of all animals occurred in the northern and western portion of the SCB— an area largely un- sampled by the NOSC surveys. Hui's results agreed with those of Evan's (1975), who found only a small fraction of the total sightings recorded on aerial and shipboard surveys to occur in this northern and western portion of the SCB; however, aerial sam- pling effort in Evan's earlier study also favored the inshore and southern portions of the SCB. 341 FISHERY BULLETIN: VOL. 84, NO. 2 Based upon the distribution of sightings on our bimonthly aerial surveys, movement of common dolphins into the SCB appeared to follow the net- work of escarpments and seamounts noted by Evans (1971). The major corridor was along the Coronado Escarpment to Thirty-Mile Bank, up to the Cata- lina Escarpment, around both sides of Santa Cata- lina Island, along the western margins of the San Pedro and Santa Monica basins to Santa Cruz and Santa Rosa Islands (Fig. 1). The population front then advanced westward along the southern margin of these islands until reaching the Santa Rosa-Cortes Ridge where it shifted south, spreading out along the western slope of this prominant underwater feature. Some elements of this influx stopped and along the way, increasing summer-autumn popula- tions significantly in the San Pedro Channel, Gulf of Santa Catalina, and, to a lesser extent, in near- shore waters from Dana Point to La Jolla. A sec- ondary pathway was from Forty-Mile Bank in the south, up the San Clemente Escarpment west of San Clemente Island to reach the Santa Rosa-Cortes Ridge area. During periods of peak occupancy common dolphin sightings west of long. 119°W were dis- tributed along the western slope of the Santa Rose- Cortes Ridge centered at lat. 33°00'N, long. 120°00'W. As waters cooled, the distributional center shifted eastward to locate over the eastern slope of the Santa Rosa-Cortes Ridge at 33°00'N, 119°20'W, while a smaller element moved north- westerly to a new location around 33°30'N, 120°30'W. With continued cooling of the western waters, the majority of the animals along the east- ern edge of the Ridge appeared to move southeast- erly to merge with existent populations south and east of San Clemente Island. The remaining small number of animals wintering-over moved westward, centering near 33°00'N, 119°30'W, south of San Nicolas Island. The destination of common dolphins that moved northwesterly from the summering grounds over the western edge of the Santa Rosa-Cortes Ridge is unknown. However, several pieces of incomplete evidence lead us to believe that they are part of a "pelagic" population that returns in late autumn or early winter to offshore waters over the Rodriguez Seamount or Patton Escarpment. During several midsummer ship surveys and three aerial surveys of offshore waters over the Patton Escarpment and San Juan Seamount, we recorded sightings of large schools of robust-bodied, brilliantly marked, "pelagic" common dolphins. On two occasions, our crew on the catch boat head-netted, brought on board, photographed, measured, tagged, freeze- branded, and released, examples of these "pelagic" animals from within schools containing predom- inantly the paler, smaller, nearshore variety of Delphinus. Ships' logs indicate that the presence of these "pelagic" animals increased with distance from shore, and percentages as high as 50% were found in mixed schools of common dolphins at the western boundary of catch trips, usually south of lat. 33°45'N and west of long. 120°00'W. West of the Patton Escarpment, mixed schools were not noted, and the few schools encountered contained only "pelagic" animals (Dohl unpubl. data). In summary, this study establishes an extended distributional range of the common dolphin within the SCB, identifies areas of significantly greater seasonal use, and provides seasonal mean popula- tion estimates. Our results confirm the findings of earlier studies that common dolphins move into the SCB following major features of underwater topog- raphy in response to increasing seasonal water temperatures. Observations on surveys also seem to indicate that most of the population moves through the SCB in a generalized counterclockwise direction, and that the western summer-autumn population is augmented by an influx of "pelagic" animals from far offshore. ACKNOWLEDGMENTS The original data for this paper were collected under contract to the University of California, Santa Cruz, from the Minerals Management Service (formerly a part of Bureau of Land Management), U.S. Department of the Interior. The analysis of these data and the development of the distributional model described here were sup- ported by Woodward-Clyde Consultants (WCC), Walnut Creek, CA, by a contract from the Minerals Management Service, Department of the Interior. We are grateful to the many individuals involved in the collection of these data: J. D. Bryant, R. C. Guess, J. D. Hall, L. J. Hobbs, M. W. Honig, K. S. Norris, and P. N. Sund. We also thank T. P. Win- field and R. K. Christiansen at WCC for technical assistance, and particularly K. P. Burnham, T. D. Smith, and R. S. Holt for valuable assistance and comments on the manuscript. LITERATURE CITED Burnham, K. P., D. R. Anderson, and J. L. Laake. 1980. Estimation of density from line transect sampling of biological populations. J. Wildl. Manage. Manage. Monogr. 342 72, 202 p. Chernoff, H., and L. E. Moses. 1959. Elementary decision theory. John Wiley and Sons, Inc., New York, NY. Clopper, C, and E. S. Pearson. 1934. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 26:404-413. Crain, B. R., K. P. Burnham, D. R. Anderson, and J. L. Laake. 1978. A Fourier-series estimator of population density for line transect sampling. Utah State Univ. Press, Logan, 25 p. Dohl, T. P., K. S. Norris, R. C. Guess, J. D. Bryant, and M. W. Honig. 1980. Cetacea of the Southern California Bight. Part II of summary of marine mammal and seabird surveys of the Southern Calfiornia Bight Area, 1975-1978. Final Report to the Bureau of Land Management, 414 p. [Available at U.S. Dep. Commer., Natl. Tech. Inf. Serv., Springfield, VA as NTIS Rep. #PB81248189.] Dohl, T. P., R. C. Guess, M. L. Duman, and R. C. Helm. 1983. Cetaceans of central and northern California, 1980-1983: Status, abundance, and distribution. Final report to the Minerals Management Service, Contract #14- 12-0001-29090, 284 p. Efron, B. 1982. The jackknife, the bootstrap, and other resampling plans. The Society for Industrial and Applied Mathematics, Philadelphia, PA. Evans, W. E. 1971. Orientation behavior of delphinids: radio-telemetric studies. Ann. New York Acad. Sci. 188:142-160. 1975. Distribution, differentiation of populations, and other aspects of the natural history of Delphimcs delphis Linneaus in the northeastern Pacific. Ph.D. Thesis, Univ. California, Los Angeles, 164 p. Holt, R. S., and J. E. Powers. 1982. Abundance estimation of dolphin stocks involved in the eastern tropical Pacific yellowfin tuna fishery determined from aerial and ship surveys to 1979. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-SWFC-23, 95 p. Hui, C. A. 1979. Undersea topography and distribution of dolphins of the genus Delphinus in the Southern California Bight. J. Mammal. 60:521-527. Laake, J. L., K. P. Burnham, and D. R. Anderson. 1979. User's manual for program TRANSECT. Utah State Univ. Press, Logan, 26 p. Leatherwood, S., A. E. Bowles, and R. R. Reeves. 1983. Endangered whales of the eastern Bering Sea and Shelikof Strait, Alaska: Results of aerial surveys, April 1982 through April 1983, with notes on other marine mammals seen. Final report to NOAA/OCSEAP, Juneau, AK. Leatherwood, S., I. T. Show, Jr., R. R. Reeves, and M. B. Wright. 1982. Proposed modification of transect models to estimate population size from aircraft with obstructed downward visibility. Int. Whaling Comm. 32:577-580. Miller, R. G. 1974. The jackknife— a review. Biometrika 61:1-15. Norris, K. S., and J. H. Prescott. 1961. Observations on Pacific cetaceans of California and Mexican waters. Univ. Calif. Publ. Zool. 63:291-402. Smith, T. D. 1981 . Line-transect techniques for estimating density of por- poise schools. J. Wildl. Manage. 45:650-657. SOKAL, R. R., AND F. J. ROHLF. 1969. Biometry. W. H. Freeman and Co., San Franc, 776 P- 343 CETACEAN HIGH-USE HABITATS OF THE NORTHEAST UNITED STATES CONTINENTAL SHELF1 Robert D. Kenney and Howard E. Winn2 ABSTRACT Results of the Cetacean and Turtle Assessment Program previously demonstrated at a qualitative level that specific areas of the continental shelf waters off the northeastern U.S. coast consistently showed high-density utilization by several cetacean species. We have quantified, on a multispecies basis and with adjustment for level of survey effort, the intensity of habitat use by whales and dolphins, and defined areas of expecially high-intensity utilization. The results demonstrate that the area off the northeast United States, which is used most intensively as cetacean habitat, is the western margin of the Gulf of Maine, from the Great South Channel to Stellwagen Bank and Jeffreys Ledge Secondary high-use areas include the continental shelf edge and the region around the eastern end of Georges Bank. High-use areas for piscivorous cetaceans are concentrated mainly in the western Gulf of Maine and secondarily at mid-shelf east of the Chesapeake region, for planktivores in the western Gulf of Maine and the southwestern and eastern portions of Georges Bank, and for teuthivores along the edge of the shelf. In general, habitat use by cetaceans is highest in spring and summer, and lowest in fall and winter. From October 1978 through January 1982, the Ceta- cean and Turtle Assessment Program (CETAP) at the University of Rhode Island conducted surveys of the waters of the U.S. continental shelf from Cape Hatteras, NC, to the northern Gulf of Maine. The purpose of these surveys was to provide data on the distribution and abundance of whales, dolphins, and sea turtles inhabiting the northeast shelf for input to decision-making relative to offshore oil and gas resource development. Twenty-six species of ceta- ceans were observed during the study, and their distributions have been described in some detail (CETAP 1982). Each species exhibited a distinctive pattern of distribution in space and time, inhabit- ing some small portion(s) of the study area at higher relative densities. When comparing distributions of individual species, there appear to be specific geographic areas which consistently contained higher abundances of several cetacean species. This phenomenon had been noted during the CETAP study (CETAP 1982), but had not been analyzed quantitatively. An individual species approach to the analysis of such multispecies phenomena has certain limitations. One cannot simply combine the sighting distributions of several species; the different cetacean species vary widely 'This report has been reviewed by the Minerals Management Ser- vice and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Ser- vice, nor does mention of trade names or commercial products con- stitute endorsement or recommendation for use 2Graduate School of Oceanography, University of Rhode Island, Narragansett, RI 02882-1197. Manuscript accepted July 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. in size and may have quite different ecological requirements. An additional complication in a study of habitat use, based on sighting data, is introduced by the uneven allocation of sighting effort. One can- not be certain whether a lack of sightings is due to absence of whales or absence of observers, or, con- versely, whether a concentration of sightings repre- sents a real concentration of whales or simply a con- centration of effort. Thus it is difficult to simply or directly combine single-species sighting distributions in any sort of multispecies habitat use analysis. In this paper, we have attempted to synthesize, from the CETAP individual species sighting data, a mea- sure of the intensity of habitat use by the total ceta- cean fauna in the study area which accounts for both interspecific differences and differences in allocation of effort. These results then serve to delineate those specific habitat areas which are used at particularly high levels by whales and dolphins off the north- eastern United States. An underlying assumption in this paper is that a habitat which is occupied by whales or dolphins is necessarily utilized by them. Previous results from CETAP data have shown that the distribution of sightings of a particular species where definite feeding behavior was observed tended to closely mir- ror the overall sighting distribution for that species. Only feeding activity at or very near the surface can be seen by observers on ships or airplanes, but much feeding behavior likely occurs below the surface For some species, observations of surface feeding are very rare. In addition, cetaceans are large mammals 345 jyr-^?- FISHERY BULLETIN: VOL. 84, NO. 2 with high metabolic rates and accordingly high feed- ing rates. They are estimated to consume prey equi- valent to 1.5-4% of their body weight daily (Sergeant 1969; Lockyer 1981), with some estimates for smaller species as much as 10% of body weight per day (eg. Smith and Gaskin 1974). The CETAP study concluded that cetaceans "would be expected to feed virtually every day while in the study area" and that "each species of cetacean was likely feeding, either at the surface or below, in any area in which it was seen regularly" (CETAP 1982, p. 417). For the pur- poses of the current study, we have also followed this reasoning and assumed that a habitat which is be- ing occupied by one or more cetacean species is therefore being utilized by those species as a feed- ing area. METHODS The CETAP study area was defined as the waters of the U.S. continental shelf north of Cape Hatteras, from the shoreline to 5 nmi (9.3 km) seaward of the 1,000 fathom (1,829 m) isobath. Surveys were con- ducted from October 1978 through January 1982. Data collected from two types of surveys have been used in this analysis: 1) Dedicated aerial surveys: Random transect aerial surveys were conducted in defined blocks within the study area, including both regular surveys throughout the year and special surveys targeted at endangered species, particularly right whales. The primary objective of these surveys was to estimate the absolute abundance, e.g., the total number of in- dividuals in the population, of each species in the study area, using line transect census methods (Burnham et al. 1980; Scott and Gilbert 1982). This methodology requires consistent use of rigorously standardized sampling, e.g., use of the same plat- form, even allocation of sampling across the different blocks, and random selection of transects within a block. The two aircraft used for these surveys were a Beechcraft3 AT-11 and a Cessna 337-G Skymaster, both twin-engine planes. The ATI 1 crew consisted of a pilot, a navigator, and four observers; two observers at a time were stationed in a clear acrylic observation bubble in the nose of the plane The Sky- master carried a pilot, a navigator, and two observ- ers, who sat in the rear seats and watched out the side windows. All surveys were conducted at an 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. altitude of 750 ft (229 m) and a groundspeed of 120 kn (222 km/h). For any particular survey, a series of parallel track lines was flown. For the regular surveys, the lines sampled were randomly chosen from a pool of lines running northwest-southeast (roughly perpendicular to the bathymetry) and spaced at 2 nmi intervals throughout the block to be sampled. For the en- dangered species surveys, the lines were systema- tically spaced at a predetermined interval, with the first line placed at a randomly determined distance from the edge of the block. 2) Platforms of opportunity (POP) surveys: Trained observers were placed aboard various ships and air- craft operating within the study area in order to col- lect distributional data to supplement the dedicated surveys. The platforms most often used included Coast Guard cutters, U.S. and foreign oceanographic and fisheries research vessels, and Coast Guard fish- eries patrol and thermography aircraft. The track of the ship or aircraft was wholly determined by its primary mission. These data could not be used in abundance estimation because effort was not al- located randomly or evenly, and the platforms used were not exactly comparable Observers on both types of surveys recorded a variety of information. The data collected included date, time, latitude and longitude, platform heading, beginning and end of periods when the observer(s) were actively on watch, and environmental informa- tion (air temperature, water temperature, depth, weather, visibility, sea state, wind direction, and cloud cover). The data were recorded at each sight- ing, as well as at periodic intervals (typically 5 min for aerial and 30 min for shipboard surveys) during all on-watch periods. This allowed for subsequent reconstruction of flight-cruise tracks. Additional data recorded at sightings included species, reliabil- ity of identification, number of animals, distance from the platform, animal heading, and behaviors. The data were transcribed from the field forms to coding forms, keypunched, and input to a com- puter data base A number of quality control steps were included in the process, and all discovered er- rors were corrected. In addition to the two types of survey data described above, historical sighting data collected prior to CETAP and opportunistic sight- ing data provided by fisherman, mariners, whale- watchers, fish-spotters, pilots, etc are included in the CETAP data. None of these data have associated track-line information, and are therefore not in- cluded in this paper. After completion of the CETAP 346 KENNEY and WINN: CETACEAN HIGH-USE HABITATS study, the entire data base was archived on magnetic tape at the University of Rhode Island Academic Computer Center. The data base is very large, com- prising nearly 70,000 entries and 112 variables; it includes almost 25,000 sightings of cetaceans, sea turtles, or other large marine animals (eg., sharks, ocean sunfish, swordfish, rays, etc.). For this paper, the study area was partitioned in- to blocks measuring 10 minutes of latitude by 10 minutes of longitude The area of the blocks ranges from about 243 km2 at the northern extreme of the study area to about 281 km2 at the southern end, due to the curvature of the earth's surface and resulting convergence of the meridians toward the north pole The data were further grouped by calen- dar seasons across all the years of sampling. All dedicated aerial and POP data which met defined criteria were included in the analysis. These criteria included observer(s) formally on watch, clear visibility of at least 2 miles, and sea states of Beau- fort 3 or lower. Although the dedicated aerial and POP data were not directly compatible for the pur- pose of absolute abundance estimation, we are justified in combining them for this analysis. An ex- amination of sighting effort in the 1979 CETAP data (Hain et al. 1981) demonstrated a significant corre- lation between numbers of sightings and length of line surveyed for both aerial and POP surveys. Re- analysis of these same data shows that the average number of sightings per mile of track line surveyed was somewhat higher for the POP surveys, but that the difference is not statistically significant at the 5% level (paired Student's £-test). Since we are in ef- fect using the number of sightings per unit length of track line as a measure of relative abundance in this analysis, the two data types can be combined. To remove any bias due to uneven allocation of sighting effort among the blocks, the effort was first quantified. A computer program was developed which calculated the length of track line surveyed each season within each of the 10-minute blocks, in- cluding only line segments surveyed within the criteria defined above Each line surveyed is recorded in the data base as a sequence of latitude-longitude positions. For any pair of successive positions, the length of track line between the points (D, in km) can be calculated by: D = 111.12 arccos [sin (X:) sin (X2) + cos (X:) cos (X2) cos (Y2 - Yx)], where Xj and X2 are the latitudes of the two posi- tions, and Yj and Y2 are the corresponding longi- tudes. This calculates great circle distance Flight or cruise tracks would actually be rhumb lines rather than great circles, but the algorithm required to calculate rhumb line distance is much more complex. Furthermore, for two points around 10 km apart, typical of track line segments in the data, great cir- cle and rhumb line distance differ by * 359 MSHfciKY BULLETIN: VOL. 84, NO. 2 Figure l.-Track of Saluda in the Gulf of California (June 1969) with numbered cetacean contacts. 360 CUMMINGS ET AL.: SOUND FROM BRYDE AND FINBACK WHALES preamplifier (Wilcoxon,5 Type M-H90-A) suspended at depths of 6 to 53 m below the surface Up to 800 m of floating cable carried the signals to the ship, allowing the hydrophone to be stationary until the ship drifted out to this distance The hydrophone was suspended from an inflatable 8 m spar buoy which provided effective acoustic isolation from low-fre- quency acceleration caused by surface waves. The hydrophone's response was attenuated at low fre- quencies (beginning with 3 dB down at 12 Hz) to fur- ther reduce low-frequency noise and to prevent most of the preamplifier blockage from any drag motion that remained. Without these or similar measures, we have found that hydrophone and sea noise below 100 Hz, even in relatively smooth seas, usually prevents satisfactory recordings of low-frequency mysticete sounds with suspended systems. One track of a magnetic tape recorder (Magnecord 1020), powered by a DC-AC converter, carried a run- ning commentary and airborne whale sounds from a radio microphone (Vega Telemike). The other track recorded signals from the hydrophone. Continuous visible records were made on station with a level recorder (Briiel & Kjaer, Type 2301), also powered by the converter which was acoustically isolated. A sound analyzer (General Radio, Type 1558) was used to monitor incoming signals and their absolute levels and to provide power to the hydrophone-preamplifier. Calibration was by means of a 1,000-Hz tone and pink or white noise which were inserted through the system and recorded at frequent intervals. Overall response of the recording system was +5 dB from 25 Hz to 18 kHz. Without a hydrophone array we could not precisely localize sound sources. However, correlations be- tween whale movements and changes in received sound level provided evidence that those sounds came from the whales observed. At sea we find it difficult to distinguish the Bryde BReference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. whale from other balaenopterids, especially the sei whale, B. borealis. An exception was the circum- stance noted here, involving long contacts and good visibility above and below water, so that identifying features of behavior and form were revealed. Most useful of these field characteristics were 1) the pres- ence of ridges on top of the head of Bryde whales, 2) the asymmetrical coloration of finbacks, usually a yellowish white on the lower right jaw and baleen that is contrasted with the darker appearance of the left area, and 3) the peculiar surfacing of sei whales whereby head and fin appear nearly simultaneous- ly, without arching. Received overall sound levels are reported in dB re 1 jiPa, and source levels are referenced to 1 m. Analysis was accomplished using graphic level recorders, oscilloscopes, a sound spectrographic recorder, and a RTA (real time analyzer). RESULTS Sightings and Recordings The locations of whale sightings associated with recordings of whale sounds are listed (Table 1). Un- identified balaenopterid whales were sighted off La Paz, where two low-level whale sounds were recorded during Contact 1. We spotted two Bryde whales, about 11 m long, southeast of Loreto (Contact 2). The sea was calm and the surface water temperature was 24 °C. The two animals separated as the ship approached. One swam away and remained mostly out of sight. The other began passing back and forth under the ship's keel. It dove about 10 m and surfaced every 1 to 6 min. W C. Cummings dove on the whale and photo- graphed it underwater for identification. We recorded 288 low-frequency moans in 50 min from the Bryde whales during Contact 2. Some of these sounds were of very low signal-noise ratio (down into the ambient level of the sea noise) and presumably originated from the more distant of the Table 1.— Contacts with sound producing whales in the Gulf of California. Contact Date Time Location Subjects (No.) 1 6-11 1000 24°43. 2 6-13 0700 25°57. 3 1930 26°50' 4 6-17 1815 28°18' 5 6-18 0530 28°25' 6 1330 28°58' 7 6-19 0900 29°35' 8 1430 29°41. 9 6-20 1430 29°14' 10 1530 29°15. 5'N, 110°36'W, 2 km S of Isla San Francisco 5'N, 110°19'W, 8 km SSE of Loreto N, 111°42'W, 14.8 km SE of Pta. Concepcion N,111°46'W, midway, Guaymas to Isla San Esteban N, 112°9.5'W, 18.5 km ENE of San Pedro Martir Island N, 112°53.5'W, 24.1 km ESE of Isla Angel de la Guarda N, 113°31'W, 3.7 km ENE of Puerto Refugio 5'N, 113°27'W, 17.6 km NE of Puerto Refugio N, 113°33'W, N. end of Ballenas Channel 5'N, 113°30'W, N. end of Ballenas Channel Balaenoptera sp. (1) B. edeni (2) B. edeni (1) Large whale (1) Balaenoptera sp. (1) B. physalus (3) B. physalus (about 35) B. physalus (2) Balaenoptera sp. (1) Balaenoptera sp. (1) 361 nonn.n.1 DUL.ijt.iiiM; »ul. 01, i\\j. i. two whales. Coincidently we were recording low- level, high-pitched whistles and squeals from a dis- tant group of saddleback porpoises, Delphinus delphis. It was obvious that changes in the loudness of other low-frequency signals, as aurally monitored, and in the level on the graphic recorder were cor- related with the nearby Bryde whale's proximity to the ship. Later the same day, another whale was sighted off Pta. Conception in the Mulege area and tentatively identified as either a sei or Bryde whale of about 12 m (Contact 3). We recorded 407 sounds from this whale The sounds were essentially the same as those recorded earlier from the Bryde whales of Con- tact 2. After analyzing the sounds, Contact 3 was identified as a Bryde whale Sounds of these characteristics were not encountered again during the cruise, nor were any other Bryde whales seen. About 100 km northwest of Guaymas, a large whale was sighted at a range of about 450 m (Con- tact 4). A brisk wind and choppy seas prevented iden- tification, but one distinctly whalelike moaning sound appeared in the accompanying noisy record- ing. East of San Pedro Martir Island, we recorded 42 sounds from another whale (Contact 5) identified as a finback, about 15 m in length. Three large finback whales were sighted off the southern tip of Isla Angel de la Guarda (Contact 6). All of the 376 moans recorded from these whales occurred when the animals were below the surface On 19 June, we sighted about 35 finback whales outside the entrance of Puerto Refugio (Contacts 7, 8). They surfaced in series of 2 to 7 times, usually in pairs or in trios. Their blows were accompanied by smooth resonant sounds similar to that expected from air rushing through a confined space Climax- ing the final appearance in a series of surfacings, the whales strongly arched their backs and appeared to dive at a steep angle Some of the finbacks' dor- sal fins were distorted. Large concentrations of whales, porpoises, and sea lions occurred over an area of at least 6 km around the ship where they 0 1 TIME IN SECONDS Figure 2.-Spectrograms of typical Bryde whale moans. The effective analyzing filter bandwidth was 3 Hz. 362 CUMMINGS ET AL.: SOUND FROM BRYDE AND FINBACK WHALES were feeding on red crabs, Pleuroncodes planipes, that swarmed at the surface during the early morn- ing and evening. We distantly accompanied two of the whales which were swimming at 18 km/h and surfacing every 1 to 1.5 min. They rose high enough above the surface for us to clearly identify them as finbacks. Extensive sound recordings were made among the large concentration of whales near shore (Contact 7) and also much farther offshore (Contact 8), away from the main group. Recordings of whale sounds from Contacts 9 and 10 were made in Ballenas Channel near finned whales on the west side of Isla Angel de la Guar- da. Analysis of Whales Sounds Most sounds attributed to Bryde and finback whales, other than those from blows, were in a class we called "moan'— emissions longer than 0.2 s and <250 Hz in frequency. Many other sounds of biological origin, including clicks, knocks, etc, were recorded in the presence of the whales, but only when other possible sources were present, such as porpoises and sea lions. Bryde Whales As seen in Figures 2 and 3, upper, Bryde whale IV RMS LU LU Q < j 0 -«Ol CO 5 IV R MS IPP^r 0 LU 70 LU 03 C/) SEC 4 1 -jjA -?rW 0 HZ 100 Figure 3— Waveform and spectrum (/Hz) for Bryde whale (upper) and finback whale (lower). Effective analyzing filter bandwidth was 0.75 Hz (Bryde whale sound), 0.375 Hz (finback whale). 363 FISHERY BULLETIN: VOL. 84, NO. 2 moans varied widely in duration and frequency (Hz). Of the 93 miscellaneous moans analyzed (Table 2), the principal sound energy occurred at a mean fre- quency of 124 Hz; that of individual moans varied from 70 to 245 Hz. Seventy-three percent of these sounds exhibited frequency shifts (mean of 15.2 Hz) that were downward or upward, or a combination thereof. The mean duration of the moans was 0.42 s (range, 0.2 to 1.5). These sounds occurred at inter- vals of 0.2 to 9 min. The Bryde whale that apparently was attracted to the ship (Contact 2) did not emit moans when very closeby. The received overall sound level for a typical moan, when this whale was estimated to be 300 m away, was 102 dB. Assuming a spherical spreading loss of 20 log10 1.094(R), R being distance in Table 2.— Analysis of whale sounds. 68/34-Hz moans Miscellaneous moans Received Source Mean Received Source Range level level frequency I Duration Range level level Contact Identification No.1 (m) (dB)* (dB)3 No.1 (Hz) (s) (m) (dB)* (dB)3 1 Balaenoptera sp. 2(2) — 83 — 0 — — — — — 2 B. edeni 0 — — — 288(93) 123.9 0.42 300 102 152 3 B. edeni 0 — — — 407(35) 132.0 0.40 600 250 116 126 168 174 4 Large whale 0 — — — 1(1) 75.0 0.60 — 90 — 5 Balaenoptera sp. 30(10) 250 121 169 12(8) 49.6 0.55 200 121 166 No visual contact 44(16) — 90 — 21(21) 50.7 0.63 — 92 — 6 Balaenoptera sp. 203(6) 2,000 115 183 173(14) 53.7 1.23 2,000 115 183 7 B. physalus 164(20) — 108 — 468(131) 59.8 0.59 100 125 165 8 B. physalus 201(30) — 99 — 550(42) 65.5 0.73 — 117 — 9 Balaenoptera sp. 3(3) 2,000 95 — 102(17) 63.3 0.70 2,000 118 181 10 Balaenoptera sp. 90(5) 800 108 166 12(8) 77.5 0.68 800 101 159 1Number of sounds encountered with number analyzed in parenthesis. 2dB re 1 iPa 3dB re 1 jiPa at 1 m. AIR-BORNE 4- 4- 3- 3- X > o o 2 1 3- 2- 1 * * v - — • 0.5 1.0 K WATER-BORNE 4- 3- 2- Kfi - » ^ . - -- - a. •'*" . *r~ o 0.5 L0 1.5 0 0.5 1.0 1.5 TIME IN SECONDS Figure 4— Spectrograms of two blows from a Bryde whale recorded in air (upper) and in water (lower). The effective analyzing filter bandwidth was 20 Hz. 364 CUMMINGS ET AL.: SOUND FROM BRYDE AND FINBACK WHALES meters, this received level would indicate an overall source level of 152 dB in the effective bandwidth. The whales were close enough and the frequencies low enough that attenuation was probably minimal. However, these particular moans could possibly have been emitted by the other whale that was about 500 m away at the time In this case, the estimated overall source level would have been 157 dB in the effective bandwidth. Weak exhalation sounds were recorded simultan- eously from underwater and in the air from the near- by surfaced Bryde whale. The exhalation sounds received underwater were nearly obscured by splash- ing sounds as the animal broke the surface (Fig. 4). For the 35 moans analyzed from the Bryde whale of Contact 3, the mean frequency of the strongest component was 132 Hz and the mean duration was 0.40 s, both values close to those from Contact 2 (Table 2). However, the overall source level estimates of 168 and 174 dB (in the effective band widths) were greater. The nearby Bryde whale (Contact 2) was totally submerged as it produced all of its moaning sounds, but no other apparent behavior was associated with the moans. Finback Whales In addition to miscellaneous moans (Fig. 5) that were similar to, but lower in frequency than those recorded from Bryde whales, the sounds of identified finback whales (Contacts 7, 8) included unique moans characterized by a long 68-Hz component that was usually followed by another component at 34 Hz (Figs. 6, 3 (lower)). Of the miscellaneous moans analyzed from recordings of the finback whales of Contacts 7 and 8, the mean frequency of the strong- est component was 59.8 and 65.5 Hz, and the mean duration was 0.59 and 0.73 s, respectively (Table 2). Typically, these moans showed some frequency shift with <10% of the signals changing more than 20 Hz, generally downward. Overall source level 200 100 - M I > u z LU o Sf 200 100- 0 1 TIME IN SECONDS Figure 5— Spectrograms of miscellaneous whale moans from finback whales off Isla Angel de la Guarda. The effective analyzing filter bandwidth was 3 Hz. 365 FISHERY BULLETIN: VOL. 84, NO. 2 200 - 100 - M U z o Sf 200 - TIME IN SECONDS Figure 6— Spectrograms of typical 68/34-Hz moans from finback whales recorded off Isla Angel de la Guarda. The first component of the moans began at 65 Hz and increased to 68 Hz in the first sec It was accompanied by weaker modulation products at about 23-Hz intervals, mostly above the main frequency. The 34-Hz component followed and sometimes overlapped the first component (lower spec- trogram). The effective analyzing filter bandwidth was 3 Hz. of the sounds was 165 dB in the effective band- widths. Of the 50 long moans from finbacks that were analyzed, the mean frequency of the main, or first, component was 68.2 Hz; the mean frequency at onset being 66.1 Hz. The mean duration was 1.5 s. Thirty moans exhibited additional lower frequency compo- nents with a mean frequency of 33.5 Hz and a mean duration of 1.3 s. The overall mean duration of these two-part moans was 3.1 s. The 365 moans of this type encountered in Contacts 7 and 8 occurred on the average of 1.6 and 2.2 times/min, respectively. In the case of unidentified balaenopterid whales, the mean frequency of the strongest component of the 68 miscellaneous moans analyzed was 58.5 Hz (range from 15 to 95 Hz), and the mean duration was 0.8 s. Of these sounds <10% had any frequency shift >10 Hz. Thirty-seven of the analyzed moans were the same as the long two-part moans recorded in the presence of finbacks. Their mean frequencies were 68.1 and 34 Hz, the mean component duration was 1.9 and 2.6 s, respectively, and the mean total dura- tion was 3.4 s. The mean starting frequency of the 68-Hz component was 63.9 Hz. These two-part moans occurred at a rate of 1.5 to 3.2/min. Overall source levels ranged from 159 to 183 dB in the ef- fective frequency bandwidth. The blows of finback whales were as high as about 7 m above the water's surface, and often they were clearly audible in air at distances out to 200 m. The last blow in a series was followed by an inhalation that sometimes involved a low-frequency whistlelike sound just before a long dive (Fig. 7). The physical characteristics of blow sounds varied slightly from one whale to another, providing a certain degree of uniqueness for an individual whale (Fig. 7). Wheez- ing, shriek, and hornlike sounds produced by hump- back whales in association with their blows have been described by Watkins (1967) and Thompson et al. (1977). 366 CUMMINGS ET AL.: SOUND FROM BRYDE AND FINBACK WHALES WHALE I q gggag o a t^Br gSjStSg^f*- 9fe iriTttrnrnWirra 14 21 «St»Tr ngKiiarr-iTinmi .awe 32 5 2 ui as Li. H ■-£,--•- v -5-5 ■J < i - V M* 3tfc --gr-Tis jjjgge '- 1- i ■ t - 4froa^ai^iBa5t*»ttia»agw4^ muk. 'jfrsasmiAe •-Sfe jte 41 51 64 71 84 TIME SCALE I 1 1 SECOND Figure 7— Spectrograms of whale blow series recorded in air. Running time in seconds relative to the first blow is indicated on the abscissa. Whales II and III (second series) can be distinguished throughout the first 105 min by the unique physical characteristics of their alter- nating blows. Just before a long dive, the whales produced a low-frequency whistlelike sound at inhalation Oast spectrogram, first row; last spectrogram, last row) which was not apparent during earlier blows of a series. In the second series, two low-level blow sounds at 110 and 132 min are not shown. The effective analyzing filter bandwidth was 20 Hz. DISCUSSION The moans recorded on this cruise from visually unidentified or unseen whales were very similar to those found to be from finbacks, except for Contact 3 involving Bryde whale sounds. Thus we believe the former also were from finback whales. Some of the moans recorded in this study only slightly resembled short "20-Hz signals" described by several investigators (Walker 1963; Patterson and Hamilton 1964; Schevill et al. 1964; Weston and Black 1965; Cummings and Thompson 1966 [fn. 4]; Northrop et al. 1968; Watkins 1981). However, none of the presently described signals could be categ- orized as short "20-Hz signals" noted in other studies, because of differences in frequency (Hz) of major sound energy, signal repetition, and inter- vals between repetitions. Typical short "20-Hz signals" are narrowband pulses with principal sound energy near 20 Hz. They are repeated at remark- ably constant intervals. Only about 3% of the sounds reported here had components as low as 20 Hz. The miscellaneous moans that were recorded from finbacks mainly resemble the category that Watkins (1981) called "higher frequency sounds". However, most of his recordings of these sounds were down- ward-sweeping pulses, eg., 75-40 Hz, with emphasis around 40 Hz. We did not record sounds similar to Watkins' low-frequency rumble or ragged pulse 367 FISHERY BULLETIN: VOL. 84, NO. 2 categories, nor did we record his nonvocal, sharp im- pulsive category. Our experience with finback and Bryde whales in the Gulf of California showed that underwater- generated sounds were not produced when visible animals were at or very close to the surface. Excep- tions were those sounds which, although principally airborne (eg., blow and snort sounds), established a physical coupling with the water medium allow- ing detection by hydrophone The typical short "20-Hz signals" noted from finback whales in other locations (Northrop et al. 1968) appear in trains that are interrupted after 3 to 22 min of pulsing (equi- valent to expected dive times, Fig. 8). We believe that these interruptions that last from 1 to 6 min represented surface time. Blue whale sounds in southeast Pacific waters had silent interruptions that were associated with surfacing and ventilation (Cum- mings and Thompson 1971). Winn et al. (1970) cor- related certain "cries" and "ratchet" sounds with sur- facing behavior of humpback whales. Data from the present cruise, our recordings of typical short "20-Hz signals", our recordings from blue whales, and from work on humpback whales, apparently reveal sur- face and dive times as learned through monitoring underwater whale sounds. Possible explanations for our lack of 20-Hz short pulses in the presently described recordings and for the absence of other classes of sounds that Watkins (1981) has commonly recorded from finbacks are seasonality and insufficient sampling. We now know that seasonality is involved. Watkins (1981) recorded the pulses in the North Atlantic only from late October to early May. Cum- mings and Thompson (fn. 4) recorded them in the North Pacific from September to April, and Thomp- 20 40 60 TIME (sec) 100 00 (/> LU > r- < -J 111 CC 20 40 60 80 TIME (min) 100 120 140 Figure 8— (a) Spectrogram of short "20-Hz signals" from finback whales; the effective analyzing filter bandwidth was 0.4 Hz. (b) Strip chart showing 11 trains of short "20-Hz signals" with interruptions between; the filter passband was 12.5-25 Hz. 368 CUMMINGS ET AL.: SOUND FROM BRYDE AND FINBACK WHALES son and Friedl (1982), working off Hawaii, recorded them only from the end of August to late April. Nor- throp et al. (1968), in the North Pacific, noted them from October to March. Finally, in recordings from finbacks in March 1985 (Gulf of California) typical 20-Hz short pulses were the predominant sound (Thompson et al.6). Like the well-known songs of humpback whales, these sounds are probably a manifestation of social or other behavior which oc- curs seasonally. According to Watkins (1981) they "perhaps were a courtship or reproductive display". Watkins and others apparently have not noted our frequently recorded 68-34 Hz long moans. There have been many technical advances in bio- acoustic signal acquisition and processing. Long- term recordings can be used for obtaining informa- tion about certain behaviors, presence or absence of animals, or perhaps distribution of a given species, without the presence of an observer (Cummings et al. 1983). Great gains are being made in the field of sighal processing wherein computer- and optically aided automatic acoustic pattern recognition is possi- ble for a number of sounds with recognizable physical criteria. However, regardless of technical ad- vances, the use of such tools is severely limited without first knowing the behavioral significance of the animal sound production. In reality, the two are mutually dependent. An analogous situation would be the use of the most refined instrumentation available for listening in on a conversation carried out in a foreign language that is unfamiliar to the observer. Although extremely difficult to fulfill, the need for related behavioral information on finback whales is paramount. For these and other reasons, descriptions of sounds from identified sources should be given in detail along with adequate description of the recording in- struments. Recording procedures and analyses can greatly affect the apparent variability of sounds. Moreover, one must be careful to consider the large variety of sounds that is apparent in any species of marine mammal (including the finback whale, as shown in this report) and the relatively limited number of recorded sounds of any species. ACKNOWLEDGMENT The authors thank R. S. Gales for assisting in the field observations; D. R. Nelson for assisting in the "Thompson, P. 0., L. T. Findley, and 0. Vidal. Finback whale underwater sounds recorded near Guaymas, Mexico. Manuscr. in prep. Paul 0. Thompson, 4256 Sierra Vista, San Diego, CA 92103. diving and other field work; R. Ludwig and crew for operating the ship; W. A. Watkins and W. E. Schevill for comments on the research; and M. Richardson, R. Hawley, and T Rydlinski for producing the finished manuscript. This work was supported by funds for Independent Research (Naval Ocean Systems Center) and by the Office of Naval Research (R. C. Tipper and B. Zahuranec) by means of con- tracts with the San Diego Society of Natural History and Scripps Institution of Oceanography. LITERATURE CITED Beamish, P. 1978. Evidence that a captive humpback whale (Megaptera novaeangliae) does not use sonar. Deep-Sea Res. 25:469- 472. Beamish, P., and E. Mitchell. 1971. Ultrasonic sounds recorded in the presence of a blue whale, Balaenoptera musculus. Deep-Sea Res. 18:803-809. 1973. Short pulse length audio frequency sounds recorded in the presence of a minke whale (Balaenoptera acutorostrata). Deep-Sea Res. 20:375-386. Cummings, W. C, D. V. Holliday, W. T Ellison, and B. J. Graham. 1983. Technical feasibility of passive acoustic location of bow- head whales in population studies off Pt. Barrow, Alaska. Tracor, Inc., San Diego, Doc No. T-83-06-002, 169 p. Cummings, W. C, and P. O. Thompson. 1971. Underwater sounds from the blue whale, Balaenoptera musculus. J. Acoust. Soc Am. 50:1193-1198. Edds, P. L. 1981. Variations in vocalizations of fin whales, Balaenoptera physalus, in the St. Lawrence River. M.S. Thesis, Univ. Maryland, College Park, 126 p. Northrop, J., W. C. Cummings, and M. F. Morrison. 1971. Underwater 20-Hz signals recorded near Midway Island. J. Acoust. Soc Am. 49:1909-1910. Northrop, J., W. C. Cummings, and P. O. Thompson. 1968. 20-Hz signals observed in the central Pacific J. Acoust. Soc. Am. 43:383-384. Patterson, B., and G. R. Hamilton. 1964. Repetitive 20 cycle per second biological hydroacoustic signals at Bermuda. In W. N. Tavolga (editor), Marine bio- acoustics, p. 125-145. Pergamon Press, N.Y. Schevill, W. E., and W. A. Watkins. 1962. Whale and porpoise voices. Woods Hole Oceanogr. Inst, (with a phonograph record), 24 p. Schevill, W. E., W A. Watkins, and R. H. Backus. 1964. The 20-cycle signals and Balaenoptera (fin whales). In W N. Tavolga (editor), Marine bio-acoustics, p. 147-152. Per- gamon Press, N.Y. Thompson, P. O., and W C Cummings. 1969. Sound production of the finback whale, Balaenoptera physalus, and Eden's whale, B. edeni, in the Gulf of Califor- nia (A). In Proceedings of the Sixth Annual Conference on Biological Sonar and Diving Mammals, p. 109. Stanford Res. Inst., Menlo Park, CA. Thompson, P. O., W C. Cummings, and S. J. Kennison. 1977. Sound production of humpback whales, Megaptera novaeangliae, in Alaskan waters. [Abstr.] J. Acoust. Soc Am. 62(Suppl. 1):S89. 369 FISHERY BULLETIN: VOL. 84, NO. 2 Thompson, P. 0., and W. A. Friedl. 1982. A long term study of low-frequency sounds from several species of whales off Oahu, Hawaii. Cetology 45, 19 p. Walker, R. A. 1963. Some intense, low-frequency underwater sounds of wide geographical distribution, apparently of biological origin. J. Acoust. Soc Am. 35:1816-1824. Watkins, W. A. 1967. Air-borne sounds of the humpback whale, Megaptera nwaeangliae. J. Mammal. 48:573-578. 1981. Activities and underwater sounds of fin whales. Rep. Whales Res. Inst. 33:83-117. Weston, D. E., and R. I. Black. 1965. Some unusual low- frequency biological noises under- water. Deep Sea Res. 12:295-298. Winn, H. E., P. J. Perkins, and T. C. Poulter. 1970. Sounds of the humpback whale In Proceedings of the Seventh Conference on Biological Sonar and Diving Mam- mals, p. 39-52. Stanford Res. Inst., Menlo Park, CA. 370 INCREASED FOOD AND ENERGY CONSUMPTION OF LACTATING NORTHERN FUR SEALS, CALLORHINUS URSINUS Michael A. Perez and Elizabeth E. Mooney1 ABSTRACT Data from pelagic northern fur seals, Callorhinus ursinus, taken during 1958-74 by the United States and Canada in the eastern Bering Sea were analyzed to determine relative feeding rates of lactating and nonlactating females. Estimates of the quantity of food and energy consumed by these seals during July-September were evaluated. The average daily feeding rate (adjusted for percentage of time feeding at sea) for lactating seals is 1.6 times that for nonlactating seals. During July-September, the total popula- tion of lactating and nonlactating females were estimated to consume 146.5 x 103 1 (204.5 x 109 kcal) and 43.1 x 103 1 (60.2 x 109 kcal) of food respectively. Fish accounted for 66.4% of food biomass (69.4% of total energy consumption); squid, the remainder. The energetics of reproduction, especially during lactation, are poorly documented for free-ranging animals. The various reproductive states of domestic mammals, e.g., cattle, sheep, etc., have been exten- sively studied; and there has also been considerable research on rodents, e.g., mice, voles, etc., under both laboratory and field conditions. As a result of these studies it is widely accepted that most nursing females require considerably more energy than do nonnursing females of the same species, age, and size. Brody (1945) also noted that the maintenance requirements of lactating animals are elevated relative to nonlactating animals. In some mammalian species, food intake during lactation may be up to 5 times greater than that observed in nonpregnant, nonlactating adult females, and lactating animals often convert con- siderable body substance to support the lactation process (Baldwin 1978). Previous studies on ter- restrial mammals have specifically shown increased energy consumption by lactating females relative to nonlactating females. For example, captive deer mice, Peromyscus maniculatus, have a 96% to a 194% increase (Stebbins 1977); and ewes have a 116% increase (Engels and Malan 1979). The bat, Myotis thysanodes, which undergoes thermoregula- tory physiological changes during reproductive stages, also has higher energy requirements for lac- tating females (Studier et al. 1973). Lactating humans are recommended to increase food con- sumption by at least 25% (Eagles and Randall 1980); however, some lactating humans in Guatemala meet their additional lactation energy costs by fat loss (Schutz et al. 1980). There are few studies on the energetics and con- sumption of food during lactation by marine mam- mals. Lactation appears to drain the energy reserves of large baleen whales; the blubber of lactating females (e.g., blue, Balaenoptera musculus, and fin, Balaenoptera physalus, whales) is lean and emaciated compared with nonlactating females (Lockyer 1978, 1981a). Lockyer (1981b) estimated that adult female sperm whales, Physeter macro- cephalus, need to increase their food intake by about 32-63% during lactation, meaning that they would need to feed 4 or 5 times daily to meet higher energy requirements. Lockyer (1981b) also estimated that minke, Balaenoptera acutorostrata, and fin whales increase their food intake by 75 and 86%, respec- tively. Spotte and Babus (1980) did not find a significantly increased mean feeding rate for one captive, pregnant bottlenosed dolphin, Tursiops truncatus, but standard deviations were consistently greater. In addition, during the first 3V2 mo of lactation, a captive mother bottlenosed dolphin con- sumed 170% more food than she did while not lac- tating the following year (Mooney2). Costa and Gen- try (in press) derived metabolic rates for lactating female northern fur seals from measurements of water flux and discussed the components of the 'Northwest and Alaska Fisheries Center National Marine Mam- mal Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way, N.E., Seattle, WA 98115. 2Mooney, E.E. 1981. Unpubl. data. Northwest and Alaska Fish. Cent. Natl. Mar. Mammal Lab., Natl. Mar. Fish. Serv., NOAA, 7600 Sand Point Way NE, Seattle, WA 98115. Manuscript accepted August 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 371 FISHERY BULLETIN: VOL. 84, NO. 2 energy budget for females and pups during the first two months of the reproductive cycle. Although most mammals ingest more food while they are lactating than they would in a nonlactating state, many species of phocid seals fast during the lactation period (Harrison 1969). These seals (e.g., gray seal, Halichoerus grypus, and northern ele- phant seal, Mirounga angustirostris) do not feed from parturition to weaning of the young, and all of their energy needs during lactation must be met by metabolism of in situ energy such as fat reserves. This behavior has been well documented for the gray seal (e.g., Amoroso and Matthews 1951, 1952; Fedak and Anderson 1982) and also for the harp seal, Phoca groenlandica, (Lavigne et al. 1982). However, metabolism of fat reserves does not reduce the energetic costs of producing offspring; it merely shifts the time that energy must be acquired, at some energy cost for storage (Millar 1978). The objectives of this study were to show, using both stomach content and body mass data, that lac- tating female fur seals ingest more food than nonlac- tating females in order to meet their increased energy requirements for maintenance and milk pro- duction, and to make estimates of the magnitude of this difference in food ingestion. For this study, we utilized data from postpartum and nonpregnant adult, female northern fur seals, Callorhinus ur- sinus, taken pelagically during 1958-74. METHODS Data on the contents of stomachs from the female northern fur seals taken pelagically (Fig. 1) in the eastern Bering Sea during 1958-74 by the United States and Canada were analyzed to determine the relation between lactation and food consumption during the summer breeding season. Only data from female fur seals (age >4 yr) which had information on both body mass and stomach content mass were included. Age was determined from longitudinal half-sections of the upper canine teeth by counting the annual growth layers in the dentine, a method widely accepted by researchers during recent decades to determine the age of many species of mammals (Klevezal' and Kleinenberg 1967). Methods used during 1958-74 to determine age, reproductive status, and the different items in the stomachs were discussed by Lander (1980). The data used in this study represented stomach contents under different stages of digestion; how- ever, it was not possible to make comparisons be- tween stages because no data on stages of diges- tion were recorded. Rates of digestion of all prey were assumed to be similar for all females during the same time interval. In our study, all postpartum females were considered lactating, and all non- pregnant (not postpartum) females were considered nonlactating. Statistical Methods The cumulative frequency distributions of data on mass of total stomach contents for both lactating and nonlactating females were compared using the one-tail Kolmogorov-Smirnov two-sample test (Siegel 1956). Data from seals with empty stomachs or stomachs with only a trace of contents (i.e., <10 cc) were con- sidered as zero mass and pooled with data from seals with food in their stomachs. Data for different ages and months were pooled to provide sufficient sam- ple size for analysis because the normal approxima- tion to compute confidence limits is only valid if sam- ple sizes are adequate (Cochran 1977). In order to use parametric statistics, and yet not seriously violate basic assumptions of normality, data were transformed by the modified arcsine transformation discussed by Zar (1974): X = \/ M + 0.5 arcsin \/ (S + 0.375)/(M + 0.75) where M is the net body mass (excluding mass of stomach contents, S) and X is the transformed value. This equation was used because of its utility where a large number of the data were from stomachs con- taining only a trace or less. The transformed values on the mass of total stomach contents (expressed as a percentage of net body mass) obtained from the above equation were transformed back to percentages to obtain means. We calculated an index of the relative intake of food by lactating females compared with that of non- lactating females by multiplying the ratio of their respective mean mass of stomach contents by 100. The £-test for two independent samples, with the assumption of unequal variance (Snedecor and Cochran 1980), was used on the transformed data to determine if any significant difference in total food consumption and body mass existed between females of different reproductive status. The relative importance of individual prey in the total diet was assessed using the modified volume percentage method (Bigg and Perez 1985). Only foods with fleshy remains were used as evidence of diet in this method, and the procedure combined the traditional methods of volume and frequency of oc- currence. The proportion of total fish and total squid 372 PEREZ and MOONEY: LACTATING NORTHERN FUR SEALS + + + & -*&=- Unimak Pass 61° N 59° 57° 55° 53° 51' 177°W 167° 157c Figure 1.— Locations where 3,494 adult female northern fur seals (ages >4 yr), whose data were analyzed in this study, were taken by the United States and Canada in the eastern Bering Sea during July-September 1958-74. in the diet was determined by frequency, while the ration of each species within only fish and squid was determined by volume. Statistical comparison of the diets of lactating and nonlactating females included 1) the Spearman rank correlation coefficient (Siegel 1956; Fritz 1974), 2) percentage similarity (Goodall 1973), and 3) 2 x 2 contingency table analysis (Zar 1974) on the number of stomachs with fish or squid. Feeding Time at Sea The largest breeding population of northern fur seals (currently estimated at 8.7 x 105 for a declin- ing population; North Pacific Fur Seal Commission 1984) resides on the Pribilof Islands during the sum- mer months. Pups first appear in late June (Bar- tholomew and Hoel 1953) and the mean date of pup birth based on recent data is 5 July (Gentry and Holt in press); a date median between values cited by Bar- tholomew and Hoel (1953) and Peterson (1968). After this time, adult females spend a number of days on shore in several visits to the islands during June-November, and the intervening days between these visits at sea foraging for food (Bartholomew and Hoel 1953; Peterson 1968). They do not feed daily. 373 FISHERY BULLETIN: VOL. 84, NO. 2 Once arriving at the rookery, a parturient female gives birth to one pup, initiates lactation, comes into estrus, and copulates with a male, but does not feed. Gentry and Holt (in press) provided data showing that the average adult female is on shore about 1 d before and 7.4 d after parturition. Each subse- quent shore visit lasts about 2.2 d (Peterson 1968; Gentry and Holt in press). The duration of the first sea trip is the shortest (4.8 d), with the duration of the subsequent sea trips increasing at a rate of an additional 1.2 d/30 d postpartum (Gentry and Holt in press). Recent data collected on the Pribilof Islands by Gentry and Holt (in press) suggests that nonpreg- nant (= nonlactating) adult females arrive later (about 8 d) on the rookeries and that they may show a somewhat different behavioral pattern than preg- nant females. Their first foraging trip at sea is longer (8.9 d), but each of their subsequent shore visits is of constant duration (2.5 d). From these data we derived values for total percent of time spent at sea during July- September (92 d) for the average adult female. Assuming birth of pups on 5 July, this was 69.3 and 75.9% for lactating and nonlactating females, respectively. However, it should be noted that individual females vary from these averages because the period during which adult females first arrive on the rookeries extends over 30 d (Bartholo- mew and Hoel 1953; Peterson 1968; Gentry and Holt in press). Feeding Rate Calculations Bigg et al. (1978) provided data on feeding rates for three captive adult female northern fur seals. Their data for these seals were 5,977 kcal/d (3.0 kg; 6.7% of body mass), 6,118 kcal/d (3.1 kg; 7.6% of body mass), and 5,055 kcal/d (2.5 kg; 8.5% of body mass). These captive northern fur seals were main- tained with a diet of Pacific herring (2.01 kcal/g dur- Table 1 .—Body mass (minus stomach contents mass) of lactating (postpartum) and nonlactating female northern fur seals (ages >4 yr pooled) taken pelagically in the eastern Bering Sea and western Alaska, 1958-74. Lactating Nonlactating x and 95% C.I. x and 95% C.I. Month n (kg) n (kg) June M99 41.10 + 0.54 128 29.77 + 1.41 July 743 34.04 + 0.42 376 31.49 ± 0.70 Aug. 1,481 35.62 + 0.30 551 31.05 + 0.57 Sept. 305 36.46 ± 0.34 118 30.19 ± 1.36 July-Sept. 2,529 35.26 ± 0.23 1,045 31.11 ± 0.42 ing winter), and it was necessary to consider the energetic concentration of the seal's diet in the wild with respect to the data in Bigg et al. (1978). We derived the following relationship from these data: Daily energy consumption (kcal/d) = 375.47 M° 75 by averaging the results given for the three captive seals. We calculated average daily feeding rates using this relationship and data on seal body mass. RESULTS Body Mass Table 1 gives the mean values of body mass of adult female northern fur seals (age >A yr) taken during June-September in the eastern Bering Sea and western Alaska. During July-September, the average lactating female (mean 35.3 kg, median age 10 yr) had a body mass 1.13 times that of the average adult nonlactating female (mean 31.1 kg, median age 5 yr; seals age ^4 yr only). However, as Figure 2 shows, lactating and nonlactating females of the same age were similar in body mass. The differences shown in Table 1 are primarily due to the higher proportions of lactating females at older ages (Lander 1981). Lactating females exhibited a significant (P < 0.001) loss of 7.1 kg of body mass between June and July following parturition (Table 1). This is based 60 r- 50 - _ 40 - en ■o, CD E 30 20 10 10 Age 15 > 18 'Pregnant (prepartum) females. Body mass does not include fetal mass. Figure 2.— Mean body mass (minus stomach contents mass) of lac- tating and nonlactating female northern fur seals by age taken pelagically in the eastern Bering Sea and western Alaska during July-September 1958-74. 374 PEREZ and MOONEY: LACTATING NORTHERN FUR SEALS on data for pregnant females, after excluding fetal mass, which we used to represent body mass of lac- tating females prior to parturition. Figure 3 shows that this loss in body mass occurred for all ages. 60 50 - 40 E 30 20 10 ■ Pregnant ▲ Lactating 10 Age 15 > 18 Figure 3.— Mean body mass (minus stomach contents and fetal mass) of pregnant (prepartum) and lactating (postpartum) female northern fur seals by age taken pelagically in the eastern Bering Sea and western Alaska during June and July respectively (1958-74). Relative Food Intake We found a significant difference (P < 0.05) in the relative magnitude of food consumption between lac- tating and nonlactating female northern fur seals during July-September, but not October, and Figure 4 shows the relative percentage frequency of the number of lactating and nonlactating adult females showing different masses of stomach contents dur- ing July-September (pooled data). It is apparent that a greater proportion of lactating females contained food in their stomachs. Lactating females signifi- cantly (P < 0.001) ingested more food because they had lower cumulative percentages of empty stom- achs and stomachs with smaller quantities of food than did nonlactating females. Table 2 presents the results of analyses between lactating and nonlactating females for the July- September period by time of collection during the day. Our calculated values of the index of relative food intake after sunrise were 162% during 0-3 h, 166% during 4-7 h, 537% during 8-11 h, and 585% during 12-15 h (P < 0.05). The calculated index values during 8-15 h after sunrise are excessive, presumably an artifact of food digestion in the stomach. c 01 u c 01 a. >■ u c 01 3 or 01 > 45 40 - 35 30 - 25 £ 20 - 75 15- 10 ■ Lactating (N=981) "J Nonlactating (N=2513) rirl J rl •* J LM hi [■ tH- ■" "■» i" ' " i " — ' '" " Empty Trace <0.2 CM CM CN CN T o CM I O m I o o T— CM CO Tf o A Total mass of stomach contents (kg) Figure 4.— The relative percentage frequency of lactating and nonlactating female northern fur seals (age >4 yr) by total mass of stomach contents during July-September. 375 FISHERY BULLETIN: VOL. 84, NO. 2 Table 2.— Body mass and arcsine transformed mass of stomach contents (expressed as a percent- age of body mass) for lactating (LACT) and nonlactating (NON) female northern fur seals (age >4 yr) by hour of collection after sunrise during July-September 1958-74. The relative consumption index (percentage expression of the ratio of the proportion of body mass which was stomach contents for lactating females relative to that for nonlactating females) is also given. Hours after sunrise Repro- ductive condition n Body mass (kg) x and 95% C.I. Mass of stomach percentage of t Arcsine units x SE contents as >ody mass Percentage units Relative consumption index (0/0) X Pi 0-3 LACT NON 312 108 35.7 + 0.7 31.4 ± 1.5 1,355.81 997.74 57.24 83.33 1.558 0.964 161.6 <0.05 4-7 LACT NON 1,070 381 35.5 ± 0.4 31.1 ± 0.7 623.22 453.29 25.86 35.42 0.333 0.201 165.7 <0.05 8-11 LACT NON 906 365 35.0 ± 0.4 31.4 + 0.7 408.80 167.27 24.46 20.22 0.145 0.027 537.0 <0.05 12-15 LACT NON 225 127 34.6 + 0.7 30.5 ± 1.2 415.60 161.66 47.64 36.38 0.152 0.026 584.6 <0.05 'Significance levels for comparisons between the mean proportions of body mass which was stomach contents (arcsine units) for lactating and nonlactating females were determined by t tests. To derive a single index value for relative food consumption between lactating and nonlactating females, we performed alternative calculations. In this case we did not simply pool data because that would not adequately account for digestion trends. Northern fur seals feed primarily at night or in the early morning hours (Fiscus et al. 1964; Gentry et al. in press); therefore, we considered the value at 0-3 h after sunrise (0.96; Table 2) as the relative daily quantity of stomach contents for nonlactating seals. Feeding more than once a day to satisfy only energy needs of maintenance and routine activity should be done by all fur seals, and would already be included in these results (Table 2) when the inherent relative rate of digestion is examined. However, lactating females require additional food intake for milk pro- duction, and we added an increment (0.12) to the value observed at 0-3 h after sunrise (1.56; Table 2) to calculate an adjusted index of 1.68% of body mass. This incremental value was derived first by calculating the rate of decrease between data values for partially digested stomach contents at the dif- ferent hourly time intervals. We assumed the rate of digestion throughout the day was the same for lactating females as that observed for nonlactating females. Next, keeping the value for lactating females at 0-3 h after sunrise (1.56) as constant, we summed the absolute value of the differences be- tween the expected values for remaining stomach contents and the observed values in Table 2 to ob- tain a value of 0.12. We then calculated a value of 174% as our index of relative food intake (i.e., the ratio of 1.68 for lactating females relative to 0.96 for nonlactating females) for a typical foraging day. However, because females do not feed every day during the breeding season (Bartholomew and Hoel 1953; Peterson 1968; Gentry and Holt in press), the average daily feeding rate (adjusted for percent- age of time feeding at sea) for lactating seals is 1.6 times that for nonlactating seals during July- September, i.e, the increased cost of lactation is + 59.8%. Estimated Energy and Food Requirements Lactating and nonlactating female northern fur seals consumed the same species of prey in relatively similar proportions within their diet, when feeding in the same general area at the same time during 1958-74. Ranks of importance of prey to the diet were significantly correlated (P < 0.05); the percent similarity of relative prey importance by percent modified volume was 80%; and there was no signifi- cant difference in the frequency of food stomachs containing fish or squid. Being culled from the same region and for the same season, data for all adult females were pooled. We derived a gross energy estimate of 1.40 kcal/g as the average energetic density of northern fur seal prey during July-September based on their relative dietary importance and information in the literature on their energy content (Table 3). Using the data on seal body mass (Table 1) and increased cost of lactation ( + 59.8%), we calculated average daily feeding rates of 18.2% (6.42 kg) and 11.4% (3.55 kg) of total body mass, respectively, for the average lac- tating and nonlactating adult female. This repre- sents daily energy consumption requirements of 376 PEREZ and MOONEY: LACTATING NORTHERN FUR SEALS Table 3.— Relative dietary importance, energy value and average daily consumption of prey by individual lactating and nonlactating female northern fur seals (age >4 yr) in the eastern Bering Sea during July-September. Relative Relative dietary importance Energy energy value in diet Estimated average consum ption Biomass (kg/d) Energy (kcal/d) Prey (o/o)1 (kcal/g)2 (%)3 Lactating Nonlactating Lactating Nonlactating Pacific herring 7.67 "2.17 11.95 0.49 0.27 1,070 590 Salmonids 1.87 52.01 2.69 0.12 0.06 240 130 Capelin 14.85 61.31 14.00 0.95 0.53 1,250 690 Deepsea smelt 3.30 70.76 1.81 0.21 0.12 160 90 Walleye pollock 36.11 61.41 36.51 2.32 1.28 3,270 1,800 Atka mackerel 1.05 81.58 1.19 0.07 0.04 110 60 Pacific sand lance 0.43 51.22 0.38 0.03 0.02 40 20 Flounders 1.10 51.20 0.94 0.07 0.04 80 50 Subtotal (fish) 66.38 91.46 69.47 4.26 2.36 6,220 3,430 Gonatid squid 33.62 101.27 30.53 2.16 1.19 2,740 1,510 Total 100.0 91.40 100.00 6.42 3.55 8,960 4,940 'Percent modified volume of stomach contents data collected during 1958-74. 2For some species, data were derived from results of proximate analyses on muscle tissue composition using energetic density factors of 9.50, 5.65 and 4.00 kcal/g (gross energy), respectively for fat, protein, and carbohydrate (Watt and Merrill 1963). Data for other species were based on bomb calorimetry analyses of whole specimens. 3Derived by multiplying columns 1 and 2, and summing to 100%. 4Based on proximate analysis data for Pacific herring, Clupea harengus pallasi, in Bigg et al. (1978). 5Based on proximate analysis data for salmonids (Salmonidae); Pacific sand lance, Ammodytes hexapterus; and flounders (Pleuronectidae) in Sidwell (1981). 6Based on data from heat of combustion in analyses of whole fish specimens of capelin, Mallotus villosus, and walleye pollock, Theragra chalcogramma [Miller, L K. 1978. Energetics of the northern fur seal in relation to climate and food resources of the Bering Sea. U.S. Mar. Mammal Comm. Rep. MMC-75/08, 27 p.] 'Based on proximate analysis data for deepsea smelt (Bathylagidae) in Childress and Nygaard (1973). 8Based on proximate analysis data for Atka mackerel, Pleurogrammus monopterygius, in Kizevetter (1971). 9Average value of prey species in diet adjusted by their relative dietary importance. 10Perez, M. A. 1984. Unpubl. data. Northwest and Alaska Fish. Cent. Natl. Mar. Mammal Lab., Natl. Mar. Fish. Serv., NOAA, 7600 Sand Point Way NE., Seattle, WA 98115. 8,960 kcal and 4,940 kcal for the average, individual lactating and nonlactating adult female northern fur seal during July-September (Table 3). Average postpartum females not in a lactating state would have a daily energy consumption requirement of 5,430 kcal or feeding rate of 11.0% (3.89 kg) of total body mass. Table 3 also provides estimates for each food item of the total energy and biomass consumed daily by the average individual adult female. Lactating females each consume about 6,220 kcal/d gross energy (4.3 kg/d) of fish and 2,740 kcal/d gross energy (2.2 kg/d) of squid, and each nonlactating in- dividual consumes about 3,430 kcal/d gross energy (2.4 kg/d) offish and 1,510 kcal/d gross energy (1.2 kg/d) of squid. Female northern fur seals are not able to feed every day, and thus estimated consump- tion for the average foraging day is 8,980 kcal/d gross energy (6.1 kg/d) of fish and 3,950 kcal/d gross energy (3.1 kg/d) of squid by lactating seals, and 4,530 kcal/d gross energy (3.1 kg/d) of fish and 1,990 kcal/d gross energy (1.6 kg/d) of squid by nonlac- tating females. We also calculated estimates of the total energy in biomass consumed by all adult females during July-September in the eastern Bering Sea (Table 4). Because the northern fur seal population has been declining in recent years (North Pacific Fur Seal Commission 1984) we used 80% of the estimated population values given by Lander (1981): 2.61 x 105 pregnant/postpartum and 1.19 x 105 nonpreg- nant adult females (age ^4 yr). We assumed all of these seals are present in the eastern Bering Sea during this period. Because 5% of the pups born on St. Paul Island, Pribilof Islands, between 1975 and 1982 died on the rookeries during July and August (Kozloff 1985), we modified our calculations to reflect the number of postpartum females which are nonlactating. We thus estimated a total of 2.48 x 105 lactating and 1.32 x 105 nonlactating adult females (age >A yr). Multiplying individual estimates by these population totals, lactating females con- sume an estimated collective total of 204.5 x 109 kcal gross energy (146.5 x 103 1) and nonlactating females consume an estimated collective total of 60.2 x 109 kcal gross energy (43.1 x 103 1) of food. Therefore, all adult female northern fur seals con- sume an estimated collective biomass of 189.6 x 103 1 with a gross energy value of 264.7 x 109 kcal during July-September, of which 69.4% of this energy (183.7 x 109 kcal; 125.9 x 103 t) are fish and 30.6% (81.0 x 109 kcal; 63.7 x 103 1) are squid. 377 FISHERY BULLETIN: VOL. 84, NO. 2 Table 4.— Estimated energy value and consumption of fish and squid by the total population of lactating and nonlactating female northern fur seals (age >4 yr) during July-September (92 days). Prey Lactating females Nonlactating females Individual average Energy consumption (kcal/g)1 (kg/d)1 Total seasonal consumption by population (2.48 x 105) Biomass Energy (x 103 t) (x 109 kcal) Individual average consumption (kg/d)' Total seasonal consumption by population (1.32 x 105)2 Biomass Energy (x 103t) (x 109 kcal) Fish 1.46 4.26 97.2 141.9 2.36 28.7 41.8 Squid 1.27 2.16 49.3 62.6 1.19 14.4 18.4 Total 1.40 6.42 146.5 204.5 3.55 43.1 60.2 'From Table 3. includes postpartum females that fail to lactate. DISCUSSION The food consumption data presented in Table 2 were based on partially digested stomach contents, and thus underestimate the actual feeding rates of adult female northern fur seals. It is apparent from these data that lactating seals obtain most of their energy needs by filling their stomachs slightly more than the nonlactating seals early in the day and by eating additional food later in the day. Any female, whether lactating or not, may eat more than once during the day, as captive northern fur seals often do (Spotte 1980). Females must feed more than once during the 24-h period (on those days when they are able to feed) to meet their daily food requirements because the maximum observed stomach contents by percentage of body mass during July-September 1958-74 were 13.8 and 8.2%, respectively, for lac- tating and nonlactating females (Perez3), which are less than their predicted feeding rates. In addition, digestion does vary among individual seals and with the type and amount of prey eaten (Bigg and Faw- cett 1985). However, the data in Table 2 should be typical of the relative relationship between lactating and nonlactating females if actual feeding rates could be measured for free-ranging seals. Lactating northern fur seals were estimated to consume 8,960 kcal/d (gross energy), of which 3,520 kcal/d (gross energy) represent the additional intake of food related to lactation. Energy expenditures for maintenance and routine activity not directly at- tributable to lactation were estimated to be 5,440 kcal/d (gross energy). This estimate is about 5.4 times the amount predicted (1,010 kcal/d metaboliz- able energy or 49.0 WO for basal metabolism by the relationship between metabolic rate (MR) in watts 3Perez, M. A. 1981. Unpubl. data. Northwest and Alaska Fish. Cent. Natl. Mar. Mammal Lab., Natl. Mar. Fish. Serv., NOAA, 7600 Sand Point Way NE, Seattle, WA 98115. (W) and body mass (M) shown by Kleiber (1961) (MR (W) = 3.39 M° 75). These estimates are not typical of energy expen- diture during the first week (7.4 d average) post- partum, a period during which the parturient female does not feed. Lactating seals must metabolize their energy from fat reserves during this period (in- cluding the day before parturition when they usually do not feed, although we considered only the post- parturition period). The loss in body mass (Table 1) in postpartum females following parturition ac- counts for some of this metabolism of energy from fat reserves. This loss includes about 0.6 kg (12% of pup mass as in harp seals, Lavigne and Stewart 1979) of placental matter and 3.3 kg (7% prepar- turient female mass) of amniotic and other fluids during parturition (Costa and Gentry in press). There is a calculated net mass loss of 3.2 kg. Loss of body water, as has been reported for some mam- mals, e.g., cattle (Degen and Young 1980) is also probably part of this loss. In addition, this loss in- cludes the utilization of fat reserves to satisfy energy requirements for lactation (Sadleir 1969) during the first few days of the pup's life, a period when par- turient females remain on shore and do not feed (Bartholomew and Hoel 1953; Peterson 1968; Gen- try and Holt in press). Our estimate of net mass loss, presumably through fat metabolism, is an underestimate because it was derived from mean body mass data from seals taken at sea, and, therefore, includes lactating animals which probably regained some body mass after their first foraging trip at sea. Costa and Gentry (in press) measured an average of 8.75 kg of mass loss, presumably by tissue metabolism and water loss, prior to the female's initial departure to sea, after which they gained additional body mass. This situa- tion is analogous to that in gray seals. The gray seal does not feed during its entire 18-d lactation period from parturition to weaning (Amoroso and Mat- 378 PEREZ and MOONEY: LACTATING NORTHERN FUR SEALS thews 1951, 1952) and over 80% of the female's stored energy reserves are used to feed their pup (Fedak and Anderson 1982). We conducted similar analyses of data comparing pregnant and nonpregnant adult females (age >4 yr) during June-July, but we did not find any sig- nificant difference in relative feeding rates. We, therefore, conclude that the onset of the lactation process, and not pregnancy, initiates increased feeding behavior in parturient fur seals. Pregnant northern fur seals presumably consume more food than required by nonpregnant females (i.e., more than that simply required as a function of body mass). This would be necessary for growth of the fetus, especially during winter and spring months when they are in the North Pacific. This conclusion was based on a preliminary examination of the pelagic fur seal data, although the results were not statistically conclusive. Female northern fur seals probably also store energy in fat reserves for the stresses of birth and the first week of lactation. Nevertheless, any additional food intake required by pregnant females is substantially less than that of lactating seals. We believe lactating females may reduce their need for additional food intake during the last month prior to weaning of pups because we did not find a significant difference in food consumption between lactating and nonlactating females during October; however, data were few. Weaning does not occur until late October or early November when females abandon their pups; the mean date of weaning is about 2 November (Peterson 1968). It should be noted that births occur over at least a 30-d period (Peterson 1968), and weaning of individual pups will likewise occur over a similar time frame. It is thus possible that pups born earlier will quit nursing earlier than those born later in the season. The total lactation period is about 3-5 mo. Therefore, the feeding rate relationships and energy estimates presented in this paper should typify those during the first three months of lactation only, and not necessarily during July-September. We assumed that all postpartum females taken during 1958-74 were lactating. We believe that this assumption does not significantly affect our results because only a small percentage of the postpartum females fail to lactate or terminate lactation for one reason or another (such as still birth or death of the pup). Therefore, our estimate of the difference in consumption between lactating and nonlactating females is a conservative indicator of the magnitude of this ratio. This is because inclusion of postpartum females that did not lactate would have decreased the mean value of stomach contents for the lactating group. Individual northern fur seals show variations in their feeding locations. Differences may occur over location and time. For example, lactating females may travel great distances, e.g., at least 160 km from the Pribilof Islands (Perez4), during their sea trips in search of food, and they may dive up to 200 m (Gentry et al. in press) to catch prey. There are, of course, differences in availability (e.g., walleye pollock, Theragra chalcogramma, Smith and Bak- kala 1982) and energetic density (e.g., Pacific herring, Clupea harengus pallasi, Bigg et al. 1978; deepsea smelt, Bathylagidae, Childress and Nygaard 1973) of prey by season, region, and depth. The 95% C.I. for the importance of fish biomass in the fur seal diet in the Bering Sea is 64.0-68.6% (Perez and Bigg5). Therefore, the estimated quantity of fish and squid consumed, and their relative energy contribu- tion, may vary ±5%. It should also be stressed that the estimates pre- sented in this paper also depend heavily on metabolic rate information for adult females which we ob- tained from the literature. Individual variations among seals will cause differences in results ob- tained from several experiments, and future research may provide somewhat different metabolic rates. Should feeding rates be revised substantial- ly, then the magnitude of energetic estimates from these data will be affected in a corresponding direc- tion. However, the relative ratio of food consump- tion between lactating and nonlactating females dur- ing the breeding season will be unaffected, and remain about 1.6. We suggest the need for further studies on feeding behavior and energetics of lac- tating females and pups prior to weaning. ACKNOWLEDGMENTS We thank Michael Bigg and Peter Olesiuk of the Pacific Biological Station, Nanaimo, B.C.; Daniel Costa of the University of California at Santa Cruz, and Roger Gentry and the late Mark Keyes of the National Marine Mammal Laboratory, Seattle, WA, for valuable information on the biology and behavior of fur seals. We also thank Charles Fowler and "Perez, M. A., Northwest and Alaska Fisheries Center National Marine Mammal Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way NE, Seattle WA 98115, pers. observ., 1984. 6Perez, M. A., and M. A. Bigg. 1984. Food habits of northern fur seals (Callorhinus ursinus) off western North America. Un- publ. rep., 67 p. Northwest and Alaska Fish. Cent. Natl. Mar. Mammal Lab., Natl. Mar. Fish. Serv., NOAA, 7600 Sand Point Way NE, Seattle, WA 98115. 379 FISHERY BULLETIN: VOL. 84, NO. 2 Thomas Loughlin for constructive criticism of the draft manuscript. We are grateful to Teresa Clocksin, Carol Hastings, Sherry Pearson, James Kenagy, and George Antonelis, Jr., for comments, suggestions, and technical assistance. LITERATURE CITED Amoroso, E. C, and L. H. Matthews. 1951. The growth of the grey seal (Halichoerus grypus) from birth to weaning. J. Anat. 85:427-428. 1952. Reproduction and lactation in the seal. Proc. Int. Congr. Anim. Reprod. Artif. Insemin. 1:193-203. Baldwin, R. L., Jr. 1978. The initiation and maintenance of lactation. In A. Yokoyama, H. Mizuno, and H. Nagasawa (editors), Physiology of mammary glands, p. 5-19. Jpn. Sci. Soc. Press, Tokyo. Bartholomew, G. A., and P. G. Hoel. 1953. Reproductive behavior of the Alaska fur seal, Callorhinus ursinus. J. Mammal. 34:417-436. Bigg, M. A., and I. Fawcett. 1985. Two biases in diet determination of northern fur seals (Callorhinus ursinus). In J. R. Beddington, R. J. H. Bever- ton, and D. M. Lavigne (editors), Marine mammals and fish- eries, p. 284-291. George Allen and Unwin, Publ., Lond. Bigg, M. A., I. B. MacAskie, and G. Ellis. 1978. Studies on captive fur seals. Progress rep. 2. Can. Fish. Mar. Serv. Manuscr. Rep. 1471, 21 p. Bigg, M. A., and M. A. Perez. 1985. Modified volume: a frequency-volume method to assess marine mammal food habits. In J. R. Beddington, R. J. H. Beverton, and D. M. Lavigne (editors), Marine mammals and fisheries, p. 277-283. George Allen and Unwin, Publ., Lond. Brody, S. 1945. Bioenergetics and growth. Reinhold, N.Y., 1023 p. Childress, J. J., and M. H. Nygaard. 1973. The chemical composition of midwater fishes as a func- tion of depth of occurrence off Southern California. Deep- Sea Res. 20:1093-1109. Cochran, W. G. 1977. Sampling techniques. 3d ed. John Wiley and Sons, N.Y., 428 p. Costa, D. P., and R. L. Gentry. In press. Free-ranging energetics of the northern fur seals. In R. L. Gentry and G. L. Kooyman (editors), Fur seals: Maternal strategies on land and at sea, p. 79-101. Princeton Univ. Press, Princeton, NJ. Degen, A. A., and B. A. Young. 1980. Live-weight, total body-water and maternal body-solid changes in pregnant and lactating beef cows. J. Agric. Sci. 95:1-5. Eagles, J. A., and M. N. Randall. 1980. Handbook of normal and therapeutic nutrition. Raven Press, N.Y., 323 p. Engels, E. A. N., and A. Malan. 1979. Feed intake of non-pregnant, pregnant and lactating ewes on native pasture and lucerne hay. S. Afr. J. Anim. Sci. 9:27-32. Fedak, M. A., and S. S. Anderson. 1982. The energetics of lactation: accurate measurements from a large wild mammal, the grey seal (Halichoerus grypus). J. Zool. (Lond.) 198:473-479. Fiscus, C. H., G. A. Baines, and F. Wilke. 1964. Pelagic fur seal investigations, Alaska waters, 1962. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 475, 59 P- Fritz, E. S. 1974. Total diet comparison in fishes by Spearman rank cor- relation coefficients. Copeia 1974:210-214. Gentry, R. L., and J. R. Holt. In press. Attendance behavior of female northern fur seals. In R. L. Gentry and G. L. Kooyman (editors), Fur seals: Maternal strategies on land and at sea, p. 41-60. Prince- ton Univ. Press, Princeton, NJ. Gentry, R. L., G. L. Kooyman, and M. E. Goebel. In press. Feeding and diving behavior of northern fur seals. In R. L. Gentry and G. L. Kooyman (editors), Fur seals: Maternal strategies on land and at sea, p. 61-78. Prince- ton Univ. Press, Princeton, NJ. Goodall, D. W. 1973. Sample similarity and species correlation. In R. Tux- en (editor), Handbook of vegetation science, Part 5: Ordina- tion and classification of communities, p. 105-106. Dr. W. Junk B. V. Publ., The Hague. Harrison, R. J. 1969. Reproduction and reproductive organs. In H. T. Andersen (editor), The biology of marine mammals, p. 253- 348. Acad. Press, N.Y. KlZEVETTER, I. V. 1971. Chemistry and technology of Pacific fish. [In Russ.] Dal'izdat, Vladivostok. [Translated by Isr. Program Sci. Transl., 1973, available U.S. Dep. Commer., Natl. Tech. Inf. Serv., Springfield, VA, as TT72-50019, 304 p.] Kleiber, M. 1961. The fire of life, an introduction to animal energetics. John Wiley and Sons, N.Y., 454 p. Klevezal', G. A., and S. E. Kleinenberg. 1967. Age determination of mammals from annual layers in teeth and bones. [In Russ.] Izdatel'stvo "Nauka", Moscow. [Translated by Isr. Program Sci. Transl. 1969, available U.S. Dep. Commer., Natl. Tech. Inf. Serv., Spring- field, VA, as TT69-55033, 128 p.] Kozloff, P. (editor). 1985. Fur seal investigations, 1982. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-71, 127 p. Lander, R. H. (editor). 1980. Summary of northern fur seal data and collection pro- cedures. Vol. 2: Eastern Pacific pelagic data of the United States and Canada (excluding fur seals sighted). U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-4, 541 p. Lander, R. H. 1981 . A life table and biomass estimate for Alaskan fur seals. Fish. Res. 1:55-70. Lavigne, D. M., and R. E. A. Stewart. 1979. Energy content of harp seal placentas. J. Mammal. 60:854-856. Lavigne, D. M., R. E. A. Stewart, and F. Fletcher. 1982. Changes in composition and energy content of harp seal milk during lactation. Physiol. Zool. 55:1-9. Lockyer, C. 1978. A theoretical approach to the balance between growth and food consumption in fin and sei whales, with special reference to the female reproductive cycle. Int. Whaling Comm. Rep. Comm. 28:243-249. 1981a. Growth and energy budgets of large baleen whales from the Southern Hemisphere. In Mammals in the seas; 3, General papers and large cetaceans, p. 379-487. FAO 380 PEREZ and MOONEY: LACTATING NORTHERN FUR SEALS Fish. Ser. 5, Vol. 3. 1981b. Estimates of growth and energy budget for the sperm whale, Physeter catodon. In Mammals in the seas; 3, Gen- eral papers and large cetaceans, p. 489-504. FAO Fish. Ser. 5, Vol. 3. 1981c. Estimation of the energy costs of growth, mainten- ance, and reproduction in the female minke whale, (Balaenoptera acutorostrata), from the Southern Hemis- phere. Int. Whaling Comm. Rep. Comm. 31:337-343. Millar, J. S. 1978. Energetics of reproduction in Peromyscus leucopus: the cost of lactation. Ecology 59:1055-1061. North Pacific Fur Seal Commission. 1984. Proceedings of the 27th annual meeting. North Pac. Fur Seal Comm., Wash., D.C., 50 p. Peterson, R. S. 1968. Social behavior of pinnipeds with particular reference to the northern fur seal. In R. J. Harrison, R. C. Hubbard, R. S. Peterson, C. E. Rice, and R. J. Schusterman (editors), The behavior and physiology of pinnipeds, p. 3-53. Apple- ton-Century-Crofts, N.Y. Sadleir, R. M. F. S. 1969. The ecology of reproduction in wild and domestic mam- mals. Methuen and Co., Lond., 321 p. Schutz, Y., A. Lechtig, and R. B. Bradfield. 1980. Energy expenditures and food intakes of lactating women in Guatemala. Am. J. Clin. Nutr. 33:892-902. Sidwell, V. D. 1981. Chemical and nutritional composition of finfishes, whales, crustaceans, mollusks, and their products. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/SEC-11, 432 p. SlEGEL, S. 1956. Nonparametric statistics for the behavioral sciences. McGraw-Hill, N.Y., 312 p. Smith, G. B., and R. G. Bakkala. 1982. Demersal fish resources of the eastern Bering Sea: spring 1976. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-754, 129 p. Snedecor, G. W., and W. G. Cochran. 1980. Statistical methods. 7th ed. Iowa State Univ. Press, 507 p. Spotte, S. 1980. Acclimation of adult northern fur seals (Callorhinus ur- sinus) to captivity. Cetology 36:1-8. Spotte, S., and B. Babus. 1980. Does a pregnant dolphin (Tursiops truncatics) eat more? Cetology 39:1-7. Stebbins, L. L. 1977. Energy requirements during reproduction of Peromys- cus maniculatus. Can. J. Zool. 55:1701-1704. Studier, E. H., V. L. Lysengen, and M. J. O'Farrell. 1973. Biology of Myotis thysanodes and M. lucifugus (Chiroptera: Vespertilionidae)— II: Bioenergetics of preg- nancy and lactation. Comp. Biochem. Physiol. 44A:467-471. Watt, B. K., and A. L. Merrill. 1963. Composition of foods... raw, processed, prepared. U.S. Dep. Agric, Agric. Handb. 8, 190 p. Zar, J. H. 1974. Biostatistical analysis. Prentice Hall, Englewood Cliffs, NJ, 620 p. 381 ARRIVAL OF NORTHERN FUR SEALS, CALLORHINUS URSINUS, ON ST. PAUL ISLAND, ALASKA Michael A. Bigg1 ABSTRACT The age-specific arrival times and relative numbers of northern fur seals, Callorhinus ursinus, on St. Paul Island, Alaska, were determined from an analysis of kill data collected during 1956-82, and a review of the fur seal literature. Arrival times differed by sex, age, and reproductive state. Arrival took place progressively earlier with age in young males and females. Most males age >6 arrived by late June, while most males age 5 arrived by late June to early July, those age 4 by mid-July, those age 3 by late July, those age 2 by mid- to late August, and those age 1 by late September to early October. Females tended to arrive later than males of the same age. Nonpregnant females age >3 arrived by mid-August, while those age 2 arrived by mid- to late September, and females age 1 by October to early November. Pregnant females age >4 arrived mainly by mid-July, about 1 month before nonpregnant females of the same age. For both sexes, the number of seals returning increased between age 1 and age 3. Both sexes appeared to stop arriving earlier and in larger numbers at about the age of sexual maturity. The process of gradual maturation may play a role in inducing a cohort to undertake the return migration at earlier times with age, and to cause a greater proportion to return. The northern fur seal, Callorhinus ursinus, inhabits the North Pacific Ocean mainly between lat. 32 °N and 60°N (Fiscus 1978; King 1983). The species is migratory, being pelagic and widely dispersed in winter, and gathering on rookeries to give birth, mate, nurse, and rest in summer. Rookeries occur along the Asian coast on Robben, Kurile, and Com- mander Islands, and along the North American coast mainly on the Pribilof Islands and on San Miguel Island. The presence of large numbers of animals on Robben Island, Commander Islands, and the Pribilof Islands has allowed an annual commer- cial kill for pelts over many years. The Pribilof Islands, in particular St. Paul Island and St. George Island has the largest stock of seals, numbering currently about 0.9 million (North Pacific Fur Seal Commission 1984a). The species has been harvested there almost every year since discovery in 1786 (Roppel and Davey 1965; Roppel 1984). Over the years, fishery managers learned to adjust the kill quite specifically for seals of a particular age and sex by making use of the arrival sequence of migrants and their preferences for haul-out sites. For example, Russians in the early 1800's took juvenile males on hauling grounds, and left the breeding adults and pups undisturbed on nearby rookeries. Americans in the late 1800's knew that the largest, and thus oldest, juvenile males arrived before small males (Jordan and Clark 1898). Follow- ^epartment of Fisheries and Oceans, Pacific Biological Station, Nanaimo, British Columbia, Canada V9R 5K6. ing the discovery in 1950 that teeth could be used for aging, the kill was refined further to focus on 3- and 4-yr-old males. Although the kill has been directed primarily at young males since the early 1900's, females were taken during a herd reduction program from 1956 to 1968. Behavioral studies on the Pribilof Islands have documented the arrival times for broad population categories, such as adult and juvenile males, and pregnant females (Jordan and Clark 1898; Barthol- omew and Hoel 1953; Peterson 1965, 1968; Gentry 1981). However, these studies could not determine the age-specific arrival times because no method was available to distinguish the age of the live animals being observed. The widely accepted arrival se- quence was for bulls to arrive on land first, followed by progressively younger males, progressively younger pregnant females, and later by mostly young nonpregnant cows (Kenyon and Wilke 1953; Fiscus 1978). This arrival sequence was deduced from preliminary examinations of the age and sex composition of commercial kills and from the arrival times of tagged individuals and to some extent from differences in body size, at least for the 1- and 2-yr- olds. There are no published analyses that describe age-specific arrival times, although some unpub- lished reports give information on arrival times. In this study, I determine the arrival times for seals of each age, sex, and reproductive condition on hauling grounds and rookeries of St. Paul Island, the largest of the Pribilof Islands. The study is based Manuscript accepted July 1985. FISHERY BULLETIN: VOL. 84, No. 2, 1986. 383 FISHERY BULLETIN: VOL. 84, NO. 2 mainly on an analysis of seasonal changes in the number of animals killed of each age during harvests. I examine the evidence for arrival times by order of decreasing age within each sex, and com- pare the relative numbers returning for each age of young seals. The published and unpublished literature on northern fur seals is reviewed for in- formation on arrival times and abundance. The rela- tionship between arrival schedules, relative number returning, and onset of sexual maturity is discussed. METHODS The kill data from St. Paul Island used in this study were collected during 1956-82 by the National Marine Mammal Laboratory, National Marine Fish- eries Service, Seattle. Most data up to 1979 were listed by Lander (1980), who noted the method of data collection and the number killed by age, sex, date, and location. Kozloff (19812, 1982, 1983) listed the data collected during 1980-82. Abegglen et al. (1956, 1957, 1958, 1959) determined the age-specific pregnancy rates of females killed during 1956-59. These authors considered a female to be pregnant when parous (carrying a term fetus), or recently postpartum (lactating or uterus involuting). They did not separate females into these two categories, or determine whether postpartum females were carry- ing a new conceptus. Almost all males and females were killed on haul- ing grounds rather than on rookeries. No commer- cial kills for males took place on rookeries, and only a few took place for females. Typically, the kill of both sexes on hauling grounds was made between late June and mid- August. It consisted of a series of consecutive 5-d circuits, or rounds, of all hauling ground sites. During each round, a crew undertook one killing operation at each site, and killed all seals present of a particular sex and length. The body length limits for harvesting were set in inches (in) from nose to tip of tail, or from nose to base of tail. I converted all lengths to cm and standard length, using 1 in for tail length. Lander (1980) and the North Pacific Fur Seal Commission (1984b) noted the annual changes in management practices on St. Paul Island. The changes included variations in body length limits, kill dates, quotas, kill locations, and special kills for sex and age. I used only data that were collected under comparable management restrictions. 2P. Kozloff (editor). 1981. Fur seal investigations, 1980. NWAFC Processed Rep., 96 p. National Marine Mammal Lab- oratory, National Marine Fisheries Service, NOAA, Seattle. Probit plots of age-specific cumulative length fre- quencies were used to determine the percentage of males and females of each age present in the kill for each set of length limits. Sufficient age-length data were not available for the plots from kills made on St. Paul Island, but were available from samples col- lected pelagically for research purposes by the United States and Canada under the terms of the North Pacific Fur Seal Commission. These data are on file at the Pacific Biological Station, Nanaimo, and at the National Marine Mammal Laboratory, Seattle. The age-length data used were from seals collected near St. Paul Island during June-August 1958-74. The lengths of females used were those of postpartum and nonpregnant seals, the main cat- egories of females killed on land. I assumed that seals were arriving on St. Paul Island when the number killed increased in suc- cessive rounds and that arrival was completed when the number killed reached an asymptote. These assumptions were valid only under certain circum- stances. One was that all seals encountered of a designated sex and length were killed, which was the case. Another was that the number of seals hauled out, and thus available for killing, was a con- stant proportion of the number alive through the harvest season. The assumption seems reasonable in that Gentry (1981) estimated an average of about 19% of marked juvenile males were ashore at any one time on St. George Island. Finally, the propor- tion of a particular age and sex killed during each year must have been sufficiently small so as not to have substantially reduced cohort size, and thus altered the trend in numbers killed by round. This qualification was probably true for all ages, except perhaps for 4-yr-old males. Lander (1981) estimated the harvest utilization rate of males on St. Paul Island to be only 2.8% for age 2, 40.3% for age 3, 14.7% for age 5, but 57.3% for age 4. Escapement rates of females from the commercial harvest were not calculated, but were probably high. The females killed were mainly of ages 3 and 4 with the largest annual take for age 3 in the years studied being 9,700, and for age 4 being 6,300. These figures com- pared with about 55,000 and 48,000, respectively, for females present in the whole population, based on Lander's (1981) life table for the species. The number killed of each age up to the last day of each round for each year was plotted to describe the seasonal change in numbers killed. For males, the most common last-day dates for each round were in the series of 5-d rounds ending between 1 July and 5 August. For years in which the dates for last- day rounds differed from this series, the number of 384 BIGG: ARRIVAL OF NORTHERN FUR SEALS males killed was interpolated from the annual plots, so as to standardize the number killed by date. The mean number of males killed, and standard error of the mean, were determined for each date of the last-day of rounds. During 1965-72, a kill of males sometimes took place twice at a haul-out site in one round, but was missed at this site in the preceding or following round. In these cases, one of the two kills was selected randomly and transposed to the other round. Occasionally, sites were visited extra times without being missed in the adjacent round. These data were omitted. The kill data used for females on hauling grounds were from years in which the kills on rookeries and hauling grounds were recorded separately, and in which the pregnancy rates were noted. Such kills took place only during 1956-59. These kills were made during the 5-d rounds with the last-day dates between 1 July and 20 August. The only kills on rookeries for which pregnancy data could be used were in 1956 and 1957. On 1-6 July 1956, a kill was made on Polivina rookery. All kills made in the region of this rookery on 1-21 July 1957 were in fact made only on the rookery (A. Roppel3). The number of females killed on rookeries was set by quota, rather than by all available animals being taken, as on hauling grounds. No body length limits were im- posed on the kill of females on rookeries in 1956 and 1957. To determine the relative number of each age that returned to St. Paul Island, I reviewed the largely subjective comments on abundance given in the literature, and also compared the number killed when arrival was believed to have been completed. For the latter, the only data used were from years when body length limits included at least 50% of the individuals of the relevant age and when the total number of living animals of a particular age did not change substantially between years. The main change in herd size was between 1956 and 1959, when pup production on St. Paul Island decreased by about 27% due to the killing of adult females dur- ing the herd reduction program (York and Hartley 1981; Fowler 1982). Pup production changed little between 1960 and 1980, although declined slightly in 1981-82. The cumulative effect of harvesting a cohort over several years was considered when com- paring the relative number of each age killed. The relative numbers of females of each age killed be- tween 1956 and 1959 were biased slightly downward with time by the herd reduction program during the 3A. Roppel, Northwest and Alaska Fisheries Center, National Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA 98115, pers. commun. July 1983. intervening years. The bias was only slight because of the years and ages selected for analysis, the lack of time for the herd reduction program to have potentially changed age distribution, and the fact that most seals ages 1 and 2 remained at sea. RESULTS Effect of Body Length Limits The lower length limit of 107 cm for males in- cluded essentially no individuals age 1, few age 2, but most of those >2> yr (Table 1). The upper length limit varied by year, with the smallest upper limit including a few ^4 yr, and the largest, a few >6 yr. I used kill data collected from the years 1969-82 to describe arrival times and relative numbers of males ages 1 and 2. Data from the years 1962-82 were used to describe the arrival time and relative number of 3-yr-olds. For males >4 yr, the relative numbers returning by age could not be compared with one another, or with younger males, because of the cumulative reduction in the size of a cohort by the harvest, and the exclusion of seals by upper length limits. I used data from the years 1963-72 and 1980-82 to describe the arrival schedule for age 4, and 1964-71 for ages 5, 6, and >7. The lower length limit of 104 cm for females in- cluded most individuals >4 yr, while the upper length limit of 116-117 cm included mostly <5 yr. Data collected in 1956 were used to describe the ar- rival schedules for females >A yr, and 1958-59 for those <5 yr. The number of females killed at age 3 during 1959 was not used due to an unusually low pup survival in 1956 (Abegglen et al. 1959; Lander 1979). Arrival of Males on Hauling Grounds 1 -Year-Olds No yearling males were taken in the kill by 5 August, and thus none were likely to have been on hauling grounds up to this time. However, few year- ling males apparently go to hauling grounds. Osgood et al. (1915) and Roppel et al. (1965a) indicated that yearlings of both sexes preferred rookery edges, near cows and pups, and only occasionally went to hauling grounds (see section on Arrival of Males on Rookeries). 2-Year-Olds Very few 2-yr-old males arrived by 1 July (Fig. 385 FISHERY BULLETIN: VOL. 84, NO. 2 Table 1 .—Percent of each age included in standard length restrictions for kills of male and female northern fur seals on hauling grounds. Percentages determined from Probit plots of age-length cumulative length frequencies of seals collected at sea near St. Paul Island by the United States and Canada. Sample sizes are in parentheses. Length limit (cm) Age (yr) Years 1 2 3 4 5 6 >7 Males (24) (166) (251) (117) (48) (20) (43) 107-119 1956-58, 1960 1.6 27.7 71.0 44.2 5.0 3.0 0.0 107-121 1959 1.6 27.9 76.5 55.7 8.3 4.0 0.0 107-124 1961-63 1.6 28.0 79.5 71.2 15.5 6.5 0.0 107-1135 1964-68 1.6 28.0 82.0 96.8 64.0 28.0 <1.2 1<135 1969-71 100.0 100.0 100.0 98.6 64.0 28.0 <1.2 <124 1972, 1980-82 100.0 100.0 97.5 73.0 15.5 6.5 0.0 <117 1973-75 100.0 99.2 82.0 35.0 3.0 1.5 0.0 <119 1976-79 100.0 99.5 86.0 40.0 -emales 4.0 2.0 0.0 (18) (69) (297) (465) (301) (136) (530) >104 1956 0.0 16.0 48.0 89.0 97.0 99.4 3>99.9 <116 1958 100.0 99.8 98.4 80.0 54.0 31.5 <12.0 <117 1959 100.0 99.9 99.0 84.0 63.0 40.0 <16.0 1 Upper body size was the presence of a mane. A. Roppel (National Marine Mammal Laboratory, National Marine Fisheries Service, NOAA, Seattle, WA 981 15, pers. comm. July 1983) felt that the mane developed at a body length of about 135 cm. 1). Numbers began to increase in early July and con- tinued to increase up to 5 August. This age group began to arrive earlier than the yearlings. Osgood et al. (1915) observed the first branded 2-yr-old in- dividuals on 12 June, about IV2 mo before the first branded yearling males on rookeries. As found in the current study, Kenyon and Wilke (1953) noted that 2-yr-olds were quite common by the end of July, and after 1 August became increasingly abundant. 5000- 4000 3000 2000- 1000 21 14 ^4 a 2yr • 3yr O 4yr 1 ,9,o/a '* ? n Figure 1.— Mean number, and standard error, of northern fur seal males killed of age 2-4 on hauling grounds of St. Paul Island, by date. Data from Lander (1980) and annual reports of the National Marine Mammal Laboratory, Seattle. Number of years of data for each date indicated above means. The date of peak numbers, and thus the date when most arrived, was probably after early August. The date when most would have arrived may be deter- mined by assuming that the interval between the time when seals clearly began to increase in number and the time when essentially all seals had arrived was the same for 2-yr-olds as for bulls and cows. Observations by Peterson (1968) suggested that this interval was about 1-1 V2 mo for bulls and pregnant females. Because the number of 2-yr-old males began to increase in early July, the arrival time for most was probably mid- to late August. A similar arrival time was also indicated by subtracting I-IV2 mo, the interval separating the first sightings of tagged yearlings and 2-yr-olds, from the arrival time of late September to early October for yearling males on rookeries. The number of 2-yr-olds returning appeared to be greater than that for yearlings, but less than that for 3-yr-olds. Roppel (fn. 3) felt that more 2-yr-old males returned than yearling males, and Kenyon et al. (1954) noted that many 2-yr-olds remained at sea. 3-Year-Olds The 3-yr-olds were already quite abundant by 1 July and reached a peak in numbers by late July (Fig. 1), suggesting that arrival was completed by late July. Kenyon and Wilke (1953) similarly noted the maximum number of 3-yr-olds on hauling grounds was after mid-July. This age group 386 BIGG: ARRIVAL OF NORTHERN FUR SEALS appeared to have the largest number of males returning. 4-Year-Olds The number of males killed of age 4 remained essentially constant during July, except for a decrease in late July (Fig. 1). Although no distinc- tive peak in numbers was evident, several factors suggest the main arrival was probably completed by mid-July. First, the number killed in the first round (i.e., up to 1 July) was likely to have been too large relative to later rounds because of an accum- ulation of males that arrived before the kill began. This situation was most obvious for kills of males ages 5 and 6 (Fig. 2), but also could have existed to some extent for the kill of males ages 2 and 3. For ages 2 and 3, the accumulation would not have been as obvious because the main arrival time was after kills began. Secondly, the true peak in number killed of 4-yr-olds was probably flattened by the high harvest utilization rate of this age. Finally, an ex- amination of the trend in numbers killed by round for individual years indicated the seasonal pattern was quite variable, ranging between that noted for males age 3, and that for males age 5. For exam- ple, the arrival time for 4-yr-olds in 1971 was similar to that seen for the typical 3-yr-olds; it was similar for the typical 5-yr-olds in 1968; and in 1980 it was intermediate, with a distinctive peak in mid-July. Such variations tended to dampen the peak. Kenyon and Wilke (1953) remarked that the maximum number of males older than 3 yr arrived before mid- July. Fewer age-4 males returned than age-3 males, probably due to the large kill at age 3. 5-Year-Olds Most 5-yr-olds appeared to have already arrived 400- S- a 5yr 300- 200- e s B 7 7 4- 4 ♦ 6yr 100- 6 e a 8 7 6 4 0- -£=*= =f M= =4= =+= =*= =t r ■ i ■ ■ i 6 II 16 21 26 31 JULY 5 10 15 20 AUGUST Figure 2.— Mean number, and standard error, of northern fur seal males killed of age 5-6 on hauling grounds of St. Paul Island, by date. Data from Lander (1980) and annual reports of the National Marine Mammal Laboratory, Seattle. Number of years of data for each date indicated above means. by early July (Fig. 2). However, as noted for 4-yr- olds, the kill by 1 July was probably large relative to the number killed in later rounds. Most males probably arrived by late June to early July, assum- ing the time in peak numbers of 5-yr-olds was earlier than mid-July, but not earlier than for territorial bulls (>7 yr) on rookeries. Fewer 5-yr-olds returned than 4-yr-olds because of the large kill of males at age 4. 6-Year-Olds As with 4- and 5-yr-olds, the first kill was likely too large. Most 6-yr-olds probably arrived by late June. Gentry (1981) tagged juvenile males on haul- ing grounds of St. George Island in 1977 and count- ed them during late May to mid- August 1980. Al- though the ages were not known with certainty, the most common age in 1977 was likely 3 yr, with a range of 2-5 yr (R. Gentry4), and thus most males in 1980 were probably 6 yr of age. His counts indicated numbers began to increase in late May, reached a peak on 19-28 June 1980, and declined thereafter. ^7-Year-Olds No males older than 6 yr of age were taken in the annual kills on hauling grounds. This was because the upper length limits excluded these ages from kills, and because many males of these ages go to rookeries for breeding rather than to hauling grounds. Arrival of Males on Rookeries 1 -Year-Olds Behavioral studies suggest most yearling males probably arrived on rookeries by late September to early October, and the number returning was the smallest of any age group of males. Osgood et al. (1915) reported that branded male yearlings were rarely seen between late July and mid- August but became more numerous later, although they always remained small in number. Kenyon and Wilke (1953) mentioned yearlings of unspecified sex returned principally in September to November, and that only a few individuals were involved. Using counts of tagged yearlings seen on rookeries between 17 September and 17 October, Roppel et al. (1965a) 4R. Gentry, Northwest and Alaska Fisheries Center, National Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA 98115, pers. commun. February 1984. 387 FISHERY BULLETIN: VOL. 84, NO. 2 suggested that the largest number of yearlings of unspecified sex was present on 27 September to 11 October. These animals were predominantly males, as indicated by the recorded sex ratio of 84% males in a sample of 356 yearlings seen during 1961-65 (Roppel et al. 1965a, 1965b, 1966). Osgood et al. (1915) noted all yearlings examined during his study were males. Surveys by Abegglen et al. (1961) in- dicated very few yearlings of either sex were pres- ent on rookeries after early November. ^7-Year-Olds Essentially all males present on rookeries during the pupping season were bulls (Jordan and Clark 1898). According to Johnson (1968), the age of such bulls would have been >1 yr. Peterson (1965, 1968) noted that bulls began to arrive on rookeries in mid- May, reached peak numbers by late June, and declined in numbers after mid-July. No data exist on whether old bulls arrived before young bulls. (Lander 1981). Pregnant females age ^4 were rarely taken on hauling grounds during July, but were in- creasingly common during 1-15 August (Fig. 3). Using the trend in the number of 4- and 5-yr-olds killed after 15 August, most pregnant females prob- ably arrived by mid-August. Because essentially all pregnant females gave birth in July, the pregnant females killed on hauling grounds during August would have been postpartum. An examination of the median dates for collection of pregnant females sug- gested that arrival times on hauling grounds of age ^4 did not differ among ages (Table 2). Nonpregnant 1-YEAR-OLDS.— As with yearling males, year- ling females apparently preferred rookeries to haul- ing grounds (Jordan and Clark 1898; Roppel et al. 1965a). No yearling females were taken on hauling grounds during the commercial kill for females up to 20 August. Arrival of Females on Hauling Grounds Pregnant, >4 Years Very few females younger than 4 yr give birth 2000 1000 II 16 21 26 31 JULY 5 10 15 20 AUGUST Figure 3.— Mean number, and range, of pregnant females of northern fur seal killed of age >4 on hauling grounds of St. Paul Island, by date. Data from Lander (1980) and annual reports of the National Marine Mammal Laboratory, Seattle. 2-YEAR-OLDS.-Jordon and Clark (1898) and Osgood et al. (1915) suggested 2-yr-old females also preferred rookeries to hauling grounds. However, a few were taken on the hauling grounds during the harvest for females. Numbers began to increase in mid- August (Fig. 4), and thus increases began about 1 mo later than males of the same age. Assuming a 1-1 V2 mo interval for essentially all animals to ar- rive, as assumed for 2-yr-old males, then 2-yr-old females probably arrived by mid- to late Septem- ber. ^3-YEAR-0LDS.-Very few nonpregnant fe- males >3 yr were taken on hauling grounds in July, but many were present by 15 August (Figs. 4, 5). Based on the trend in the number of females killed at 3-5 yr, the arrival of ages >S yr was essentially completed by mid-August. Support for this conclu- sion comes from Peterson (1965, 1968), who counted Table 2.— Median dates of collection of pregnant and nonpregnant females of north- ern fur seals taken during 1956, 1958, and 1959 on hauling grounds of St. Paul Island. All dates are in August. Data from annual reports of the National Marine Mammal Laborabory, National Marine Fisheries Service, NOAA, Seattle. State Age (yr) Year 3 4 5 6 7 8 9 >10 1956 Pregnant 11 11 9 11 10 10 9 Nonpregnant 12 11 11 10 11 10 10 9 1958 Pregnant — 9 9 9 8 10 15 10 Nonpregnant 13 11 10 10 8 10 10 8 1959 Pregnant — 13 12 12 12 12 12 13 Nonpregnant 14 13 13 12 12 13 11 12 388 BIGG: ARRIVAL OF NORTHERN FUR SEALS "nonbreeders" on hauling grounds and the inland edges of rookeries. "Nonbreeders" were thought to consist of idle females and young males. He ob- served a sharp increase in numbers in early August and that most arrived by mid- August. The current study indicated the female component of Peterson's "nonbreeders" were mainly nonpregnant females, plus a few postpartum females. Abegglen et al. (1956) noted an increase in the number of seals on hauling grounds and rookery edges between 15 August and 4 September. While this increase may have resulted from a continued influx of nonpreg- nant females at >S yr, it may also have been due, at least in part, to the arrival of some 2-yr-old males and females. The increase in number of nonpregnant females during August consisted primarily of 3- and 4-yr- olds. A comparison of the median dates for collec- tion of nonpregnant females at ^3 yr on hauling grounds suggests that arrival times were similar for each age (Table 2). 3000- 2000- * 2yr • 3yr O 4yr / < N 1000- / / > 0- -i 1 9- -9- ^ -r— + — t==t — i ' Arrival of Females on Rookeries Pregnant, >4 Years Females gave birth on St. Paul Island during 15 June to 10 August, with about 90% of all births com- pleted by 20 July (Bartholomew and Hoel 1953; Peterson 1965, 1968). The general belief that preg- nant females arrived by order of decreasing age ap- parently originated from Wilke (1953). He collected 571 females on rookeries from 15 June to 4 Septem- ber and showed the median date of collection for each age became progressively earlier with age. For example, the median collection date for females at MO yr was 7 July, while that for females at 3 yr was 23 August. However, Wilke did not separate pregnant and nonpregnant females in his calcula- tions. The large shift in median dates probably resulted mainly from an influx of young nonpreg- nant females on rookeries during August, as took place on hauling grounds. An analysis of arrival times for pregnant females of each age should not include seals that are non- pregnant. Such an analysis can be made using data collected by Wilke between 15 July and 22 July 1953 (Table 3). Although Wilke did not record pregnancy 2000 1000 16 21 JULY 26 31 5 10 15 AUGUST 20 II 16 21 26 31 JULY 5 10 15 20 AUGUST Figure 4.— Mean number, and range, of nonpregnant females of northern fur seal killed of ages 2-4 on hauling grounds of St. Paul Island, by date. Data from Lander (1980) and annual reports of the National Marine Mammal Laboratory, Seattle. Figure 5.— Mean number, and range, of nonpregnant females of northern fur seal killed at age >5 on hauling grounds of St. Paul Island, by date. Data from Lander (1980) and annual reports of the National Marine Mammal Laboratory, Seattle. Table 3. — Median dates of collection of northern fur seal females on rookeries of St. Paul Island during 17 June to 22 July 1953. Data from Wilke (1953) and the current study. Age (yrs) Number collected by age Date 4 5 6 7 8 9 >10 n 17 June 0 2 2 1 0 0 20 25 22 June 1 0 7 3 2 2 22 37 27 June 0 2 3 1 1 3 26 36 2 July 0 4 5 7 5 5 23 49 7 July 0 2 6 3 3 6 20 40 12 July 1 3 5 1 1 0 2 13 17 July 2 2 5 3 5 5 21 43 22 July 3 9 8 7 6 1 16 50 Median date 16 Jul 10 Jul 6 Jul 4 Jul 9 Jul 3 Jul 29 Jun 389 FISHERY BULLETIN: VOL. 84, NO. 2 rates for this sampling period, the rates were prob- ably 90-100%, as will be shown later on rookeries for the period 1-21 July. A comparison of median collection dates suggests arrival may have taken place slightly earlier with increasing age, but no clear shift in arrival times was evident, as previously believed. Unfortunately, the true age-specific arrival times of parous females cannot be determined readily from these data. The main difficulty is that the pregnant females used in the analysis included not just parous seals, but postpartum seals as well. Postpartum seals usually remain on land for 2 d, then go to sea to forage for 8 or 9 d, and repeat this pattern about 10 times throughout the nursing period (Peterson 1958; Gentry and Holt in press). The potentially complex effect that returning post- partum females could have on the trend in the num- ber of parous females arriving of a particular age must be considered. Other difficulties were the small sample sizes, and the fact that the sample sizes taken on each date did not reflect the increase in numbers on rookeries. At this time, while a slight shift in ar- rival times of parous females may exist with age, more research is needed for confirmation. Nonpregnant l-YEAR-OLD.-Jordan and Clark (1898) felt year- ling females did not arrive on rookeries before September. As noted earlier for yearling males, Kenyon and Wilke (1953) felt yearlings returned to the Pribilof Islands mainly during September to November, and only a few individuals were involved. The date of arrival for most yearling females is unclear, although it is probably after yearling males, during October to early November. Only a small number of yearling females had arrived by late September to early October compared to males. However, they arrived presumably no later than early November, because few yearlings were pres- ent on the rookeries after that time. 2-YEAR-OLDS. -The arrival of 2-yr-old females on rookeries began in August, a similar time to that seen on hauling grounds. Branding studies by Osgood et al. (1915) suggested a few individuals began to arrive about one month after males. The first branded 2-yr-old female was seen on 19 July compared to 12 June for males age 2. Thus, arrival was probably completed also a month later than males, by mid- to late September. Jordan and Clark (1898) reported that 2-yr-old females began to in- crease in numbers by about 1 August, while Ken- yon and Wilke (1953) noted they did not begin until late August, and the current study suggested arrival on hauling grounds began in mid- August. Kenyon and Wilke (1953) believed the largest number were present in October, slightly later than suggested by the current study. Based on the comments by Ken- yon and Wilke (1953) and Kenyon et al. (1954), fewer 2-yr-olds returned than 3-yr-olds, but more 2-yr-olds returned than yearlings. >4-YEAR-OLDS.-A total of 1,533 females were collected on rookeries during 1-6 July 1956 and 1-21 July 1957, a period covering the main pupping season. All females were >4 yr of age. Of these, only 2% were nonpregnant, a low rate compared to 31% nonpregnancy for the population as a whole, based on the life table derived by Lander (1981). The low rate likely resulted from the small number of non- pregnant females on the rookeries, as was found on Table 4.— Summary of the times of arrival and relative numbers for males and females of northern fur seal rookeries and hauling grounds of St. Paul Island, based on the current study and a review of the literature. Age Sex Site1 State2 (yr) Arrival time3 Abundance Male R 1 late Sept. to early Oct. few HG 2 mid- to late Aug. 2 yr > 1 yr HG 3 late July 3 yr > 2 yr HG 4 mid-July — HG 5 late June to early July — HG 6 late June — R >7 late June — Female R NP 1 Oct. to early Nov. few HG,R NP 2 mid- to late Sept. 2 yr > 1 yr HG NP >3 mid-Aug. 3 yr > 2 yr HG P >A mid-Aug. — R P >4 mid-July — 1R = rookery; HG = hauling grounds. 2NP = nonpregnant; P = pregnant. 3 Date when essentially all seals would have arrived. 390 BIGG: ARRIVAL OF NORTHERN FUR SEALS the hauling grounds at this time (Figs. 4, 5). The rate was probably biased downward by the fact that non- pregnant females stayed on land for a slightly shorter period of time than nursing females. Using data given by Gentry and Holt (in press), nonnursing females appeared to stay on shore for only about 64% as long as nursing females. Nonnursing females make about half as many visits to land as nursing females, but stay about one-third longer for each visit. A gradual increase in the nonpregnancy rate took place on Polivina rookery during early to mid-July: 1 July - 0% (n = 280), 6 July = 2% (734), 11 July = 1% (198), 16 July = 3% (148), and 21 July = 6% (173). When weighted for the shorter period of stay on land by nonpregnant females, the rates increased from 0% by 1 July to 10% by 21 July. Presumably, the increasing rate during July resulted from the arrival of more nonpregnant females age >A. Numbers of nonpregnant females began to increase particularly by mid-July. DISCUSSION Northern fur seals arriving on St. Paul Island can go first to rookeries located on beaches just above high tide, or to hauling grounds more inland. The typical arrival sequence (Jordan and Clark 1898; Kenyon and Wilke 1953; Peterson 1965, 1968) is for the bulls to establish territories for breeding on rookeries in May-June. Pregnant females arrive next on rookeries to pup, mate, and nurse in harems within the territories. Subadult males arrive main- ly during the pupping season and go to hauling grounds rather than rookeries. Although young males of different sizes (i.e., ages) tend to arrive in successive waves with time, studies of marked seals (Gentry et al. 1979) indicate that arrival times of in- dividual subadult males can be quite variable be- tween years. In early August, harem bulls abandon their territories, and the social structure of the rookery disintegrates. Nursing cows then tend to disperse more widely on land, and nonterritorial bulls and some subadult males move on rookeries from hauling grounds. The mixing of seals between rookeries and hauling grounds after July results in less site distinction. The literature is unclear as to the arrival times of subadult and nonpregnant adult females after July, and whether these seals go first to rookeries or to hauling grounds, or go to both simultaneously. Age 2 females arrive later in the season, and go to rookeries and hauling grounds, while yearlings of both sexes arrive last, and go mainly to rookeries. Seals begin leaving St. Paul Island for the southern migration in October to November (Roppel et al. 1965a; Kenyon and Wilke 1953). Few remain on the hauling grounds after mid- October, and few on rookeries after early November. Table 4 summarizes the age-specific arrival times and relative numbers of seals seen on rookeries and hauling grounds, based on information given in the Results. Two arrival times existed for pregnant females, one by mid-July on rookeries and the other by mid- August on hauling grounds. The second date no doubt resulted from the movement of some post- partum females from the rookeries to the hauling grounds after the harems disintegrated. Thus, the arrival time on St. Paul Island was by mid-July, rather than mid-August. The arrival times for nonpregnant females at >2> yr on to St. Paul Island was less certain than for pregnant females because age-specific data on ar- rival times existed from hauling grounds up to mid- August, but not from rookeries after mid-July. Also it was not known whether nonpregnant females went first to rookeries or to hauling grounds. The main arrival time was probably by mid- August, a's was found on hauling grounds. This was likely because nonpregnant females began to increase in numbers on rookeries in early to mid-July, and an interval of 1-1 V2 mo was probably needed for essen- tially all arrivals to be completed. Also, Abegglen et al. (1956) felt that most females on the hauling grounds during August came directly from the sea, although some came from rookeries. From the cur- rent study, some postpartum females go from rookeries to hauling grounds. Perhaps most non- pregnant females go first to the hauling grounds. Nonpregnant females >3 yr arrived about 1 mo later than pregnant females. According to R. Gen- try (fn. 4), marked adult females on St. George Island also arrived later when nonpregnant, al- though only about 10 d later. The reason for the dif- ferences in length of delay caused by nonpregnan- cy found in the two studies is unclear at this time. The answer may come when details of the study by Gentry are reported, or perhaps when more is known about movement patterns of adult females between rookeries and hauling grounds. The finding that nonpregnant females arrived after pupping suggests nonpregnancy delayed the date of mating. A delay in mating has been reported previously for maturing females, but not for non- pregnant cows. Because parous females pup about 1 d after arrival, and mate 5-6 d after pupping (Peterson 1968; Gentry and Holt in press), essen- tially all females that pup will have mated by mid- 391 FISHERY BULLETIN: VOL. 84, NO. 2 to late July. Assuming a similar interval between arrival and mating for nonpregnant females, most nonpregnant females would be mated by mid- to late August. Jordan and Clark (1898) stated that young females were impregnated in early August, after old females, and Abegglen et al. (1958) observed that females ages 3 and 4 bred after the harems dis- banded. Also, Craig (1964) reported females ovulated for the first time in late August or Septem- ber. The only evidence that I could find of late mating in a nonpregnant cow was by Osgood et al. (1915), who observed a harem bull mating a female that was "not very young" on 21 August. A comparison of the age-specific arrival times for each sex on St. Paul Island (Table 4) largely con- firms the comments by Kenyon and Wilke (1953) and Fiscus (1978) that arrival began progressively earlier with increasing age. However, the current study in- dicated that this phenomenon was obvious only for young ages. It was seen in nonpregnant females ages 1-3 and in males ages 1-6. Although no dif- ferences in arrival times were shown for older males and nonpregnant females, differences could exist, but would be small. The differences in arrival times became progressively less with age for males be- tween 1 and 6 yr and apparently for females between 1 and 3 yr. A comparison of the relative numbers returning to St. Paul Island (Table 4) suggests that progres- sively more males and females returned between ages 1 and 3. The cumulative effect of the kill on males of 2 and 3 yr prevented comparisons of abun- dance with males >4 yr. For females, the number of 4-yr-olds returning was probably not greater than 3-yr-olds, as suggested by the similarity in the number of 3- and 4-yr-olds killed on hauling grounds by mid-August (Figs. 3, 4). However, pregnancies complicate comparisons of abundance on hauling grounds between females 3 yr and older. Between ages 4 and 10, an increasing proportion of females become pregnant (Lander 1981) and thus go to rookeries rather than hauling grounds. The data collected in this study suggest that, with age, young seals of both sexes arrive progressively earlier, and in progressively larger numbers. The reason for these changes in arrival schedules lies in an understanding of the mechanism that controls the migration schedule. However, little is known about this mechamism in the northern fur seal. The mechanism, if it is like that of other vertebrates (see Gauthreaux 1980; Baker 1978), is probably complex. It could involve selective factors, such as food supply and climate, and numerous environmental and physiological factors, such as photoperiod, reproduc- tive hormones, and endogenous rhythms. For north- ern fur seals, learned and innate components are likely to be involved. There are several examples of where learning has been suggested to be involved in migration. When the species leaves the Pribilof Islands for the southern migration, juveniles tend to disperse widely in the North Pacific Ocean, preg- nant females tend to travel to the coastal waters off California, and adult males generally remain in the northern Gulf of Alaska (Baker et al. 1970; Fiscus 1978). Baker (1978) has suggested that the juvenile northern fur seals may explore the habitat, and, with age, eventually learn the best wintering areas. Also, an increasing proportion of immature seals return to their natal sites on Pribilof Islands with age (Ken- yon and Wilke 1953), although sometimes the natal site is abandoned and a new colony is established, such as at San Miguel Island, CA (Peterson et al. 1968). Baker (1978) has proposed that site recog- nition may be learned shortly after birth, and with time, the site is usually relocated. However, other components of migration may be innate. For exam- ple, the annual timing of arrival for pregnant females on St. Paul Island is remarkably precise. Peterson (1968) calculated the mean arrival date to be 30 June for each of 3 years. Such precision seems unlikely to be the result of only learning. Keyes et al. (1971) examined the pineal gland of this species for seasonal variations in hydroxy-indole levels for various ages of males and females, and postulated photoperiodic regulation of the reproductive cycle. A physiological event in the lives of young males and females which coincides with the cessation of arriving earlier and returning in greater numbers is the attainment of sexual maturity. Baker (1978) pointed out that sexual maturation controls the ini- tiation of migration in many vertebrates. While a few male northern fur seals begin to produce sperm at 3 yr, most do not do so until about 5 yr (Kenyon et al. 1954; Murphy 1969, 1970). The average female conceives for the first time on her 5th birthday, although typically ovulates for the first time on her 4th (Craig 1964; York 1983). Thus, it was during the years of immaturity that young seals gradually syn- chronized their arrival schedules with that of the adults. Perhaps the gradual process of gonad maturation in both sexes over several years plays a role in inducing a cohort to migrate progressively earlier in the year and in causing a greater propor- tion to return to breeding sites. A relationship between sexual maturity and changes in arrival times on St. Paul Island could ex- plain two other arrival phenomena noted in this study. In the first case, considerable annual varia- 392 BIGG: ARRIVAL OF NORTHERN FUR SEALS tion was noted in the seasonal pattern of arrival for 4-yr-old males, ranging from the typical pattern seen in 3-yr-olds to that seen in 5-yr-olds. Such dif- ferences in the arrival pattern may indicate that the age at which males reach sexual maturity differs between cohorts, a possibility worth further investi- gation. Variations in the age at sexual maturity could result from annual variations in body growth rate caused in turn by fluctuations in food supply. In the second case, pregnant females at >A yr may have arrived slightly earlier with increasing age. This would take place if the first conception resulted in a later date of parturition than in subsequent years. This is a possibility because, according to Craig (1964), the first ovulation appears to be later than subsequent ovulations. The age of primiparous females spans mainly between 4 and 10 yr (York 1983), and thus the age at first ovulations presum- ably also spans a similar number of years. Arrival times would tend to be slightly earlier with age from the increased proportion of mature females. An alternate explanation for seals arriving in pro- gressively larger numbers, may lie in the energetic costs of the return migration from the North Pacific Ocean to the Bering Sea. For yearlings, the energetic costs may be too large for all but a few individuals to return. With age, the relative costs may be more favorable and permit an increased pro- portion to return. For each age, males tended to arrive before females. This situation could result if, through selec- tion or learning, the time of the return migration was ultimately established for each sex by the adults. The mechanism controlling the timing of migration in young seals would gradually shift arrival times with age to eventually synchronize with those of the adults. However, because the arrival times of adult males was earlier than that of cows, the arrival times of immature males would also be before those of im- mature females. The fact that nonpregnant adult females arrived after parous females could be the result of nonpregnant females gaining some advan- tage in the energetic costs of migration. Since pre- sumably competition exists for food around the Pribilof Islands during the summer, perhaps survival of nonpregnant adult females is enhanced by feed- ing elsewhere, thus delaying the return migration by 1 mo. ACKNOWLEDGMENTS I am grateful to P. Olesiuk for preparing the Probit plots, and I. Fawcett for collating much of the data on kills and pregnancy rates. I thank P. Olesiuk, T. Quinn, and two journal reviewers for useful comments on the manuscript. LITERATURE CITED Abegglen, C. E., A. Y. Roppel, and F. Wilke. 1956. Alaska fur seal investigations, Pribilof Islands, Alaska. Report of field activities, June-September 1956. U.S. Fish Wildl. Serv., Bur. Commer. Fish., Sect. Mar. Mammal Res., 1956, 145 p. 1957. Alaska fur seal investigations, Pribilof Islands, Alaska. Report of field activities, June-September 1957. U.S. Fish Wildl. Serv., Bur. Commer. Fish., Sect. Mar. Mammal Res., 1957, 163 p. 1958. Alaska fur seal investigations, Pribilof Islands, Alaska. Report of field activities, June-September 1958. U.S. Fish Wildl. Serv., Bur. Commer. Fish., Sect. Mar. Mammal Res., 1958, 187 p. 1959. Alaska fur seal investigations, Pribilof Islands, Alaska. Report of field activities, June-September 1959. U.S. Fish Wildl. Serv., Bur. Commer. Fish., Sect. Mar. Mammal Res., 1959, 132 p. Abegglen, C. E., A. Y. Roppel, A. M. Johnson, and F. Wilke. 1961. Alaska fur seal investigations, Pribilof Islands, Alaska. Report of field activities, June-September 1961. U.S. Fish Wildl. Serv., Bur. Commer. Fish., Sect. Mar. Mammal Res., 1961, 148 p. Baker, R. C, F. Wilke, and C. H. Baltzo. 1970. The northern fur seal. U.S. Fish. Wildl. Serv., Circ. 336, 19 p. Baker, R. R. 1978. The evolutionary ecology of animal migration. Holmes and Meier Publ. Inc., N.Y., 1012 p. Bartholomew, G. A., Jr., and P. G. Hoel. 1953. Reproductive behavior in the Alaska fur seal, Callo- rhinus ursintis. J. Mammal. 34:417-436. Craig, A. M. 1964. Histology of reproduction and the estrus cycle in the female fur seal, Callorhinus ursinus. J. Fish. Res. Board Can. 21:773-811. Fiscus, C. H. 1978. Northern fur seal. In D. Haley (editor), Marine mam- mals of eastern North Pacific and Arctic waters, p. 152-159. Pac. Search Press, Seattle. Fowler, C. W. 1982. Interactions of northern fur seals and commercial fisheries. Trans. 47th North Am. Wildl. Natl. Resour. Conf., p. 278-292. Gauthreaux, S. A., Jr., (editor). 1980. Animal migration, orientation, and navigation. Acad. Press, N.Y., 387 p. Gentry, R. L. 1981. Land-sea movements of northern fur seals relative to commercial harvesting. In J. A. Chapman and D. Pursley (editors), Worldwide Furbearer Conference Proceedings, Vol. 2, p. 1328-1359. The Worldwide Furbearer Conference Inc., Frostburg, MD. Gentry, R. L., and J. R. Holt. In press. Attendance behavior of northern fur seals. In R. Gentry and G. L. Kooyman (editors), Fur seals: maternal strategies on land and at sea. Princeton Univ. Press, Princeton, NJ. Gentry, R. L., J. Holt, and J. Francis. 1979. Land-sea movements of juvenile males. In Marine 393 FISHERY BULLETIN: VOL. 84, NO. 2 Mammals Division, Fur seal investigations. 1978, p. 19-32. National Marine Mammal Laboratory, Northwest and Alaska Fisheries Center, National Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA. Johnson, A. M. 1968. Annual mortality of territorial male fur seals and its management significance. J. Wildl. Manage. 32:94-99. Jordan, D. S., and G. A. Clark. 1898. The history, condition, and needs of the herd of fur seals resorting to the Pribilof Islands. In D. S. Jordan, L. Stejneger, F. A. Lucas, J. F. Moser, C. H. Townsend, G. A. Clark, and J. Murrary (editors), The fur seals and fur-seal islands of the North Pacific Ocean, Part 1, 249 p. Treas. Dep. Doc. 2017, 4 parts. Kenyon, K. W., and F. Wilke. 1953. Migration of the northern fur seal, Callorhinus ur- sinus. J. Mammal. 34:86-98. Kenyon, K. W., V. B. Scheffer, and D. G. Chapman. 1954. A population study of the Alaska fur-seal herd. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Wildl. 12, 77 p. Keyes, M. C, R. J. Wurtman, C. A. Elden, C. Chou, and R. McGuire. 1971. Seasonal fluctuations of hydroxy-indole levels in the pineal gland of the northern fur seal. [Abstr.] J. Wildl. Dis. 7:235. King, J. E. 1983. Seals of the world. British Museum (Natural History), Oxford Univ. Press, Oxford, 240 p. Kozloff, P. 1982. Fur seal investigations, 1981. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-37, 88 p. 1985. Fur seal investigations, 1982. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-71, 127 p. Lander, R. H. 1979. Role of land and ocean mortality in yield of male Alaskan fur seal, Callorhinus ursinus. Fish. Bull., U.S. 77:311-314. 1980 (editor). Summary of northern fur seal data and collec- tion procedures. Vol. 1: Land data of the United States and Soviet Union (excluding tag and recovery records). U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-3, 315 p. 1981. A life table and biomass estimate for the Alaska fur seals. Fish. Res. 1:55-70. Murphy, H. D. 1969. Microscopic studies on the testis of the northern fur seal, Callorhinus ursinus, (Part Two). In Proceedings of the Sixth Annual Conference on Biological Sonar and Diving Mammals, p. 15-18. Stanford Research Institute, Menlo Park, CA. 1970. Microscopic studies on the testis of the northern fur seal, Callorhinus ursinus, (3). In Proceedings of the Seventh Annual Conference on Biological Sonar and Diving Mammals, p. 97-103. Stanford Research Institute, Menlo Park, CA. North Pacific Fur Seal Commission. 1984a. Proceedings of the 27th Annual Meeting April 9-13, 1984, Moscow, U.S.S.R. North Pac. Fur Seal Comm., Wash., DC, 50 p. 1984b. North Pacific Fur Seal Commission report on in- vestigations during 1977-80. North Pac. Fur Seal Comm., Wash., DC, 198 p. Osgood, W. H., E. A. Preble, and G. H. Parker. 1915. The fur seals and other life of the Pribilof Islands, Alaska, in 1914. U.S. Bur. Fish. Bull. 34:1-172. Peterson, R. S. 1965. Behavior of the northern fur seal. Ph.D. thesis, The John Hopkins Univ., Baltimore, 233 p. 1968. Social behavior in pinnipeds, with particular reference to the northern fur seal. In R. J. Harrison, R. C. Hubbard, R. S. Peterson, C. E. Rice, and R. J. Schusterman (editors), The behavior and physiology of pinnipeds, p. 3-53. Apple- ton-Century-Crofts, N.Y., 411 p. Peterson, R. S., B. J. Le Boeuf, and R. L. Delong. 1968. Fur seals from the Bering Sea breeding in California. Nature 219:899-901. Roppel, A. Y. 1984. Management of northern fur seals on the Pribilof Islands, Alaska, 1786-1981. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 4, 26 p. Roppel, A. Y., and S. P. Davey. 1965. Evolution of fur seal management on the Pribilof Islands. J. Wildl. Manage. 29:448-463. Roppel, A. Y., A. M. Johnson, and D. G. Chapman. 1965a. Fur seal investigations, Pribilof Island, Alaska, 1963. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 497, 60 p. Roppel, A. Y., A. M. Johnson, R. E. Anas, and D. G. Chapman. 1965b. Fur seal investigations, Pribilof Islands, Alaska 1964. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 502, 46 p. 1966. Fur seal investigations, Pribilof Islands, Alaska, 1965. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish 536, 45 p. Wilke, F. 1953. Alaska fur seal investigations, Pribilof Islands, Alaska, summer of 1953. U.S. Fish Wildl. Serv., Bur. Commer. Fish., Sect. Mar. Mammal Res., 34 p. York, A. E. 1983. Average age at first reproduction of the northern fur seal (Callorhinus ursinus). Can. J. Fish. Aquat. Sci. 40: 121-127. York, A. E., and J. R. Hartley. 1981. Pup production following harvest of female northern fur seals. Can. J. Fish. Aquat. Sci. 38:84-90. 394 MODELING LIFE-STAGE-SPECIFIC INSTANTANEOUS MORTALITY RATES, AN APPLICATION TO NORTHERN ANCHOVY, ENGRAULIS MORDAX, EGGS AND LARVAE Nancy C. H. Lo1 ABSTRACT Life-stage-specific instantaneous mortality rates (IMRs) are often estimated individually for each life stage of an organism using regression analysis. A single estimation procedure for all life stages may be preferable because it would increase the overall precision of the IMRs and also provide a more realistic mortality model. Two such procedures were developed in this paper. One is single-equation model where regression estimates of all IMRs are obtained by fitting a single survivorship function to the entire data set. The other is the maximum likelihood estimator. These models were compared using northern an- chovy egg and larval data. The survivorship functions of each were, respectively, exponential and Pareto functions. The mortality of marine fish can be described by its survival probability S(t) = P(T > t) - exp [- f •Jo X(u)du], where T is the age of the fish and X(t) is the instantaneous mortality rate (IMR) at age t. Dur- ing their early life history, pelagic marine fishes pass through a series of life stages: eggs, yolk-sac lar- val, feeding pelagic larval, juvenile and adult stages. The IMR X(t) could be different for some life stages. Therefore, for / life stages, there may be G distinc- tive IMRs where Gt) will be m - Kit) ug-l < t < un MO UG-1 < t < UG where ug is the maximum age of mortality stanza g. Xg(t) # Xg(t) for g ± g. For example, Xx(t) may be the IMR for egg and yolk-sac larval stages, even though each is a different life stage, and X2(t) the IMR for feeding larvae. As a result, the conditional survival probability, Sg(t) = P (T > t\T > ug_x) cor- responding to Xg(t), will also be different from Sg(t) 'Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. Sl(ul)S2(t) ux < t < u2 S(t) = G-l n Sd(ud)SG(t) Uq.^Kug k f°r k < ux), and Equation (5b) is fitted to data set (Info), \n(klux) for ux < k < 20 d) to estimate a and SINGLE-EQUATION MODEL (SEM) The SEM consolidates all the conditional survival probabilities (Sg(t)) from each mortality stanza into a single equation. It not only eliminates discontin- uities at transitions between life stages, but also im- proves the precision of overall mortality estimates because of the large sample size Moreover, the SEM makes it possible to estimate the IMR for life stages where data are scarce. Based upon Equation (2), S(k) of anchovy eggs and larvae is m = or S(k) = Sife) k < u, S^u^iti) Ui < ^ < 20 s^tdSM k < ux Sx(ux)S2(k) ux ux\T > 0) = e~M\ S2(ux) = P(T> ux\T > ux) = 1, and ux = tys = 4.5 d. Thus by creating two new independent variables xxi and x2i such that Xli - and x2i - k k < ux ux ux < ti < 20 ux k < ux k % < t{ < 20 it follows that S(k) = Sx(xxi)S2(x2i) and the mortal- ity curve can be written as E(yd = e^x(xXl)S2(x2l) = e,e-°*i> Xj*, and \n(x2ilu{}) is then used to estimate a and (1 through linear least squares regression. MAXIMUM LIKELIHOOD ESTIMATOR (MLE) The MLE is presented here as an alternative method of estimating IMRs. Because the data used for mortality estimators are grouped by age, I fol- lowed the procedures described by Kulldorff (1961) and McDonald and Ransom (1979) for grouped data. Here, N{ = Y{_x - Y{ (number of deaths between ages ti_i and t£ of a single cohort are multinomial variables, each with probability Pi = Sfo-i) - SiU). L{N%,Pt{z');i = 1, ...,/) ex n p&y, 1=1 (8) where z is the parameter vector in X(t). The derivatives of the logarithm of likelihood function with respect to the parameters z's are set equal to zero. Solutions to the simultaneous equations 31nL dZs = 0 are MLEs of z, if certain conditions are satisfied (Kulldorff 1961). In marine fish only the IMRs of a few life stages are considered because of the lack of data. It is then necessary to compute the condi- tional probability Px = %14 h (0.17 d) and <20 d, the conditional probabilities are computed from a truncated exponential and Pareto survival proba- bility (Equation (2)) (Gross and Clark 1975, p. 128-132): P% = P(ti_1 0, and can also be expressed as function of sample proportions NJN (Equations (Al) and (A2)), which are equal to the ratios of differences of sam- ple mean daily productions (^_j - y^K^i - Vk)-n^n (see Appendix). The quantity n^ = NJm is the sample mean daily death between two adjacent groups. The MLEs require N{ > 0. Due to sampling error, it is possible to observe more individuals in the older group than the adjacent younger group, i.e., yi_x < y{. If so, some adjacent groups 401 FISHERY BULLETIN: VOL. 84, NO. 2 (i/i, ti) have to be combined so that y{ > y{- for t{ < tf. The ratio n{ln can be used in place ofNJN to compute the MLEs. This correction is inappropriate if the reason for y{ > y{; for t{ < t? is that in- dividuals were evicted from the sampling area or immigrated into it, as such movements violate the assumption of a stationary population. Although %s are sufficient for computing point estimates of the MLE, the total number of deaths between ages tx and tk (N = m(yl - yk) is required for computation of the ASVAR-COV of the MLEs. N can then be used to determine minimum number of tows (mx) for the youngest stage through m1yl = N for a given precision of the MLE. Although the sample size for eggs may dif- fer from that of larvae, an equal number of sample sizes is assumed to compute the ASVAR-COV. The minimum number of egg tows can be determined by m, = Nlyv RESULTS Both the MEM and the SEM were fitted to the basic data (yi7 t(, 0.17 d < t{ < 20 d) collected from 1980 to 1983, using NR and LR (Table 1, Fig. 2). The point estimates and their asymptotic standard errors are listed in Table 2 and Figure 3. NR and LR produced similar estimates of the IMRs for the MEM. When the SEM was applied to the combined egg and larval data, the WNR was also used to com- pute the IMRs in addition to NR and LR because of the inequality of the variances among life stages. The variance of egg counts was higher than that of larvae because eggs were more patchily distributed than larvae. Because of this, the inverse of the variances of sample means of eggs and larvae was used as the weights for the WNR. The estimates from the WNR were similar to those from LR and the standard errors from both methods were lower than those from NR. The WNR estimates of egg IMRs from the SEM were more precise than estimates from the MEM, whereas the most precise estimates of larval IMRs were provided by the MEM using NR. The SEM was more precise than the MEM for eggs but not for the larvae, because the variance of eggs was larger than that of larvae. Thus, when eggs and larvae were combined in an SEM, the variance around the single equation was smaller for the eggs and larger for the larvae. Nevertheless, the SEM produced larval IMRs with reasonable precision when the WNR was used. Therefore, the SEM WNR is suitable for ap- UNWEIGHTED SEM 2.0 r 1983 10 15 20 25 0 5 AGE IN DAYS 10 15 20 25 10 15 20 25 Figure 2.— Observed daily anchovy egg and larval production/0.05 mz (O = eggs, • = larvae), and the mortality curves from the MEM (two short curves) and the SEM (one single curve) using unweighted and weighted nonlinear regression for 1980-83 field collected data. 402 LO: MORTALITY RATES OF NORTHERN ANCHOVY Table 2.— Estimates from multi-equation model (MEM), single-equation model (SEM), and maximum likelihood estimator (MLE) for anchovy egg and larval mor- tality {a and /?), and their standard error (SE) based upon 1980-83 field data where K is number of age groups and m is number of tows used in each model. For both MEM and SEM, nonlinear regression (NR), linear regression (LR) and weighted nonlinear regression (WNR) estimates are given. Egg mortality Larval mortality a SE Km P SE Km 1980 MEM NR 0.39 0.103 5(961) 1.22 0.0314 7(199) LR 0.35 0.13 1.32 0.06 SEM NR 0.32 0.05 12(1,160) 1.06 0.41 12(1,160) WNR 0.25 0.02 1.33 0.06 LR 0.24 0.05 1.36 0.13 MLE 0.36 0.012 11(961) 1.28 0.09 11(961) 0.016 (500) 0.12 (500) 0.02 (300) 0.16 (300) 0.03 (199) 0.27 (199) 1981 MEM NR 0.13 0.16 5(1,134) 1.53 0.032 7(403) LR 0.13 0.15 1.54 0.06 SEM NR 0.13 0.07 12(1,537) 2.19 0.96 12(1,537) WNR 0.33 0.06 1.70 0.18 LR 0.20 0.05 1.64 0.15 MLE 0.24 0.008 10(1,134) 0.96 0.06 10(961) 0.01 (500) 0.08 (500) 0.02 (300) 0.11 (300) 0.01 (403) 0.10 (403) 1982 MEM NR 0.17 0.26 5(992) 1.81 0.036 6(96) LR 0.19 0.24 1.87 0.065 SEM NR 0.14 0.10 11(1,088) 1.77 1.46 11(1,088) WNR 0.13 0.04 1.83 0.36 LR 0.12 0.07 1.85 0.20 MLE 0.24 0.008 8(992) 1.20 0.08 8(992) 0.01 (500) 0.11 (500) 0.015 (300) 0.14 (300) 0.03 (100) 0.25 (100) 1983 MEM NR 0.23 0.29 5(850) 2.05 0.11 7(78) LR 0.27 0.25 1.80 0.10 SEM NR 0.26 0.19 12(928) 2.45 2.71 12(928) WNR 0.30 0.05 2.23 0.28 LR 0.33 0.08 1.84 0.22 MLE 0.32 0.007 11(850) 2.48 0.10 11(850) 0.01 (500) 0.14 (500) 0.013 (300) 0.18 (300) 0.02 (80) 0.35 (80) plications where it is preferable to estimate IMRs for egg and larvae simultaneously (e.g., simulation studies of mortality at all life stages). The SEM is preferable for modeling the mortality curves through all life stages because it eliminates the multiple estimates that occur at the endpoint of each life stage (Fig. 2). In addition, the SEM allows estimation of the IMRs for all life stages even when data for some life stages are inadequate for indepen- dent estimation of a life-stage-specific IMR. In com- paring NR and LR, the estimates of IMRs from these two procedures were similar, despite the dif- ferent assumptions about the error term. One com- plication of using LR is that the abundance for any 403 FISHERY BULLETIN: VOL. 84, NO. 2 0.5 0.4 > ~ 0.3 o S 0.2 - en o> Ui 0.1 h 0.0 1979 0.5 r 2. 0.4 H 1980 x MEM-NR • MEM-LR o SEM-NR A SEM-WNR A SEM-LR + MLE 1981 1982 1983 1983 ^ 3.0 oa c 2.5 ^ 20 O o > 1.5 r i.o o « 0.5 (0 0.0 x MEM-NR • MEM-LR o SEM-NR A SEM-WNR ▲ SEM-LR + MLE 1980 1981 1982 1983 1983 Figure 3.— Estimated anchovy egg mortality (a), larval mortality coefficient (/3), and their standard error (SE) using multi-equation model (MEM), single-equation model (SEM), and maximum likelihood estimator (MLE) for 1980-83. specific age needs to be transformed back to the original unit. Direct inverse transformation may bias the estimates. Thus, the LR may not be ap- propriate for biomass estimation or other applica- tions where a transformation back to original units is required. In addition to the above regression models, the MLEs of egg and larval IMRs were also computed based on the data set n{ = y^\ - yit i = 1, . . .k (Equations (Al) and (A2), Table 1). The ASCOV- VAR for anchovy egg and larval mortality rates re- quires the total number of eggs and larvae that died between ages 4 h and 20 d from the sample (N). It is not possible to obtain N directly from my^ (i.e. N = myx) because eggs and larvae are sampled with different nets and in different regions. Anchovy eggs have a more concentrated and patchy distri- bution than larvae which are less numerous and distributed more uniformly throughout the entire survey area because of the diffusion of larvae after hatching (Hewitt 1982). Zero density of eggs was assumed for the offshore regions where eggs were not sampled to compute the weighted average egg production y% = Z. wr yir. I then divided m^ by r the proportion of area sampled (q = Z. wr where wr's are summed over the regions where egg tows were taken) to obtain sample daily death N in [tlt tk). Thus, N = mtfjjq; q ranges from 0.53 to 0.82 for 1980-83. Four sets of sample sizes were con- sidered: m = m1, 500, 300, m2 where mx is the ac- tual number of egg tows and m2, actual number of larval tows (Table 2). For any given N, one obtains the ASVAR-COV of a and p by dividing a{j by N where a^'s are the elements in matrix A of Equa- tion (12). The MLE point estimates a and /?, were between the estimates yielded by the SEM and the MEM in most cases. The precision of the MLE for egg IMR was higher than that of the regression estimates. The standard error of the MLE of the larval IMR was between those of the MEM and SEM regres- sion estimates (Table 2, Fig. 3). 404 LO: MORTALITY RATES OF NORTHERN ANCHOVY DISCUSSION All the estimates of instantaneous mortality rates (IMR) discussed in this paper were computed from age (stage) frequency data. To ensure the unbiased- ness of the estimates, three assumptions have to be met: a stationary population, reliable growth curves, and accurate samplers. Any violation of these assumptions will cause biases in the mortality estimates. Nets usually do not retain fish of all sizes because some small fish extrude through the net and some large fish avoid the net. Thus the estimates of size-specific retention rates are essential correc- tion factors for the catch. If fish migrate at a signifi- cant rate, either the migration rate should be estimated or the sampling area should be expanded to eliminate migration problems, for migration violates the assumption of a stationary population and thus biases the mortality. Because growth curves are normally used to assign age to stage of eggs and larvae, biased growth curves would lead to inaccurate age assignments which definitely would bias the mortality estimates. Although modeling the mortality rates of the early life stages of anchovy is the focus of this paper, I have shown that the SEM (Fig. 2) can be applied to any continuous process whose parameters are life- stage specific and generally estimated separately. For example, many allometric relations such as the growth curves may have different instantaneous growth rates for different life stages. A single con- tinuous growth curve for the whole life cycle is possi- ble using the SEM which allows greater latitude of modeling life-stage-specific growth rates than modeling the instantaneous growth rate for the whole life cycle as proposed by Schnute (1981). How- ever, the SEM does require knowledge of the forms of instantaneous rates and the endpoint of each mor- tality stanza (or life stage). In this study, the determination of a cutoff point between life stages was based upon examination of the empirical data and biological implications. It is conceivable to include the cutoff point (%) as one of the parameters in both SEM and MLE (Matthews and Farewell 1982). The cutoff point can then be estimated directly through the models. Matthews and Farewell considered the exponential mortality curve with one cutoff point and obtained MLE of the cutoff point (change point). For anchovy egg and larvae, the cuttoff point for the eggs and larvae up to 20 d old was easily determined from the IMR and age data (Lo 1985). Estimation of the cutoff point through SEM or MLE would be laborious and any improvement may be minimal. However, the estimates through the models would eliminate the problem of whether ux should be hatching time or the age of yolk-sac larvae. Comparison of these two regression models with the MLEs based on anchovy egg and larval data in- dicated that the point estimates of the IMRs were similar. The SEM using WNR provided the most precise egg IMR which was nearly the same as the MLE. The MEM, using NR, provided the most precise estimates of larval IMR's. The regression estimators of the IMR's are easier to compute than the MLEs, yet they require larger sample sizes than the MLEs. If money is not a constraint, the SEM is preferred to the MLE. Otherwise, the MLE should be used. Based upon 1980 anchovy egg and larval data, 300 tows for eggs and larvae each (a total of 600 tows) could guarantee MLEs of a and (i with cv = 0.10. The current sampling design (egg tows ~ 1,000) seems to use an excessive number of egg tows for the MLEs of egg and larval IMRs. If the larval IMR is the only parameter to be estimated, the MEM is recommended. ACKNOWLEDGMENTS I thank J. Hunter of Southwest Fisheries Center, National Marine Fisheries Service, and C. J. Park of San Diego State University for valuable discus- sions through the writing of the manuscript, the referee for constructive comments, and Mary Ragan and Larraine Prescott for typing the manuscript. LITERATURE CITED Dixon, W. J., M. B. Brown, L. Engleman, J. W. Frane, M. A. Hill, R. J. Jennrich, and J. D. Toporek. 1983. BMDP statistical software. Univ. Calif. Press, Berkeley. Gross, A. J., and V. A. Clark. 1975. Survival distributions: reliability applications in the biomedical sciences. John Wiley and Sons, N.Y., 331 P- Hewitt, R. P. 1982. Spatial pattern and survival of anchovy larvae: implica- tions of adult reproductive strategy. Ph.D. Thesis, Univ. California, San Diego, 207 p. Hewitt, R. P., and G. D. Brewer. 1983. Nearshore production of young anchovy. CalCOFI (Calif. Coop. Oceanic Fish. Invest.), Rep. 24, 235-244. Johnson, N. L., and S. Kotz. 1976. Distributions in statistics: continuous univariate distributions - 2. John Wiley and Sons, N.Y., 306 p. KULLDORFF, G. 1961. Contributions to the theory of estimation from grouped and partially grouped samples. John Wiley & Sons, Inc., N.Y., 144 p. Lo, N. C. H. 1983. Re-estimation of three parameters associated with an- 405 FISHERY BULLETIN: VOL. 84, NO. 2 chovy egg and larval abundance: Temperature dependent hatching time, yolk-sac growth rate and egg and larval reten- tion in mesh nets. U.S. Dep. Commer., NOAA NMFS SWFC-31, 33 p. 1985. Egg production of the central stock of northern an- chovy 1951-83. Fish. Bull., U.S. 83:137-150. Matthews, D. E., and V. T. Farewell. 1982. On testing for a constant hazard against a change-point alternative. Biometrics 38:463-468. McDonald, J. B., and M. R. Ransom. 1979. Alternative parameter estimators based upon grouped data. Commun. Stat.-Theory, Method A8(9)899-917. Parker, K. 1980. A direct method for estimating northern anchovy, Engraulis mordax, spawning biomass. Fish. Bull., U.S. 78:541-544. Schnute, J. 1981. A versatile growth model with statistically stable parameters. Can. J. Fish. Aquat. Sci. 38:1128-1140. Seber, G. A. F. 1980. Some recent advances in the estimation of animal abun- dance. Tech. Rep. WSG 80-1, 101 p. Smith, P. E. 1972. The increase in spawning biomass of northern anchovy, Engraulis mordax. Fish. Bull., U.S. 80:849-974. ZWEIFEL, J. R., AND P. E. SMITH. 1981. Estimates of abundance and mortality of larval an- chovies (1951-75). Rapp. P.-v. Reun Cons. int. Explor. Mer. 178:248-259. APPENDIX The two partial deviations of In L (Equation (11)) are a InL N -N I da i=2 N -U-i- k«-aLi 1 - e -oA. Ni % Z -z + tx - (% - h) I— I - = 0 (Al) d InL dp -N I N i=c+i N In ux + \-j~\ ln li 1 - / u W k-i - In '201 Wi i— r = 0 (A2) where A^ = t{ - ti_1 and ux ~ 3. Both Equations (Al) and (A2) depend on the proportion NJN rather than the absolute counts (iV/s). In order to have a unique solution of a and (1, it is necessary to have a2 In L „ - , 32 InL ^ — n o < 0 and — nn9 < 0. da2 dft2 Moreover, the conditions (A3) and hm — > 0, hm — < 0 o — o da o— oo da .. d InL .. d InL hm — — — > 0, hm — — -— < 0 is— o dp p~oo dp (A4) guarantee a positive solution of a and p. Equation (A3) leads to the following constraints 406 1 < g^-t,) |20r _ ± | — I .2 JV 1- e-^i-^-i) (A5) and 1 < I — I / u \-p In '20^ MT W e°(ui-fi) i=c+i N ti-it In ti t- i-l Mi « i-l, (A6) After algebraic manipulation, it was easy to see that Equation (A4) was true for this truncated exponential and the Pareto MLE. We used an iterative procedure to select the MLE of a and /J, which satisfies not only Equation (A3) but also the constraints of Equations (A5) and (A6). The partial derivations in each entry of matrix A (Equation (12)) are '201" 32lnP, da2 e«(«i-*i) -5— - e**i + (Ml _ tf- u. e.<.,-.,. g ' . i 32 In ^ dp2 In '201 Wi 201" Ui e"(»i-(i) g-d^-y j^j'3 _ i and d2\nP, V J v\uJ \ux dadp e-dH-t,) I?0-]" _ i 407 METHODOLOGICAL PROBLEMS IN SAMPLING COMMERCIAL ROCKFISH LANDINGS A. R. Sen1 ABSTRACT The present sample survey plan, for the estimation of age and species composition of California rockfish landings, which is stratified two-stage with port-month group as a stratum, poses serious operational problems in data collection. A revised plan is suggested which is workable. Formulas have been developed for estimating total catch and its error by species-sex-age groups; optimum sampling and subsampling fractions have been obtained for a given cost function and the precision of the estimator is compared with two other estimators. The method developed has been extended to cover situations other than rockfish. The paper also deals with double-sampling for specified cost for the estimation of age composition of a species, which is important to predict the status of a stock in future years, the inherent problems in data collection in commercial fisheries, and the measurement errors involved in the survey. Estimates of the total catch (in terms of number) by species-sex-age and by area of landing and dur- ing a given time for commercial rockfish caught in California north of point Arguello are currently based on a probability sample of landings. The com- mercially important species of rockfish taken by California's fishery with mixed species are widow rockfish, Sebastes entomelas; bocaccio, Sebastes paucispinis; and chilipepper, Sebastes goodei. A study was undertaken during 1983 under agree- ment between the present author, the Humboldt State University Foundation, and the Tiburon Laboratory of the National Marine Fisheries Ser- vice, NOAA, to determine if the present sampling plan for the estimation of species and age-composi- tion of California rockfish landings is workable. The study revealed that the current plan is not opera- tionally feasible. A revised plan is proposed which is workable and would provide efficient estimates of the parameters based on existing catch data within the usual limitations of budget and person- nel and under the assumptions made in the plan. Formulas have been developed for the ratio estimators of mean and total catch and their errors. Optimum sampling and subsampling fractions have been obtained for a given cost function and the preci- sion of the estimator is compared with two other estimators. For most theoretical population work and for management purposes, the knowledge of the age 'Department of Mathematics and Statistics, Queen's University, Kingston, Ontario, Canada K7L 3N6; present address: 67 Ranch- Ridge Way N.W., Calgary, Alberta, Canada T3G 1Z8. composition is important to predict the status of the stock in future years. Fridricksson (1934) developed the age-length key method for determing age com- position from a large number of length measure- ments. Fridricksson's approach was improved by Ketchen (1950) who provided more accurate results for age groups at the extremities of the distribution. Kutkuhn (1963) mentioned the limitations of the age- length key approach except in situations where price differentials may demand sorting of landings by size criterion. Westrheim and Ricker (1978) pointed out that the age-length key approach will almost always give biased estimates. Clark (1981) and more recent- ly Bartoo and Parker (1983) dealt with methods for control or elimination of bias. Following the method of Tanaka (1953) in which stratification occurs after subsampling for age, Kutkuhn (1963) estimated absolute age composition of California salmon land- ings by port-month groups. He showed that the sam- pling procedure is not effective unless the age sam- ple is at least five times costlier than the length sample. Mackett (1963) found double sampling more effi- cient than simple random sampling with fixed sam- pling costs for estimating relative age composition of Pacific albacore landings. Southward (1976) found that a sample of otoliths proportional to the length frequency of sampled fish from each port was preferable to fixed sample size procedure for estimating age composition of Pacific halibut. Kimura (1977) arrived at the same conclu- sion as Southward by following a somewhat dif- ferent approach. We will present some of the important considera- Manuscript accepted August 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 409 MSttfcKX JBULL^IIJN: VUL. 84, INU. Z tions in sampling for estimating age composition of rockfish landings based on recent widow rockfish data from the California coast. Finally, we will describe some of the measurement errors, which would normally occur in simple random sample of individual fish and which are taken care of in cluster sampling adopted in our approach. The sampling plan arrived at may produce usable results under the assumptions stated, though some of the assumptions have been under attack during recent years. DESIGN OF THE SURVEY Rockfish are being landed at 14 points on the California coast. Of these, three cater only to com- mercial fishing, four to sport fishing, and seven to both sport and commercial fishing. The 10 commer- cial ports are grouped into 6 port groups with a sam- pler (six in all) assigned to each of the 6 ports— Eureka, Fort Bragg, Bodega Bay, San Francisco, Monterey, and Morro Bay. The commercial trawlers make trips varying in length from 1 to 8 d. These vessels maintain log books to keep records of area fished and appropriate catch for each tow. Sampling by tow is generally not feasible because it is not possible for the sampler to be on board during haul time. For the same reasons no estimates of fish being rejected and returned to the sea are obtained because this would involve collection of discarded fish from randomly selected tows within sampled trips. Selection Procedure A two-stage stratified random sampling plan was adopted with port-month group as a stratum and boat trips within a stratum as first-stage sampling units. Fish are sorted at sea into market categories. The first stage sampling units are poststratified into categories and at least one cluster of a given weight is subsampled within each sort-type from a first- stage sampling unit. Categories are based upon species composition, size, and quality, but in other contexts they could be strictly size or species categories. Cluster (box) of 25 lb is taken when sampling small fish, or any time small rockfish are landed such that there would be more than 20 fish in the 50-lb cluster. In all other cases 50-lb standard cluster size is selected. A cluster is next separated by number of each species and its weight, which are recorded along with sex, total length, and otolith of each member of a species in the cluster. The instructions are to "sample all market categories (sorts) from a boat, and from as many boats as possible and select: "(i) 1 cluster per 20,000 lb of widow rockfish landed by each boat, up to 4 clusters, "(ii) 1 cluster for all other species, if less than 5,000 lb landed, and "(iii) 2 clusters for all species if more than 5,000 lb are landed. "The second cluster should not be taken if this precludes sampling another boat." Estimation with Poststratification of Sample Trips by Categories Consider the problem of estimation of total catch of a given species for a port-month stratum. Equa- tions for estimation of other characteristics for fisheries with mixed species are straightforward and can be obtained by substituting the value of the characteristic for the catch of the species. Totals across strata are formed by simple additon. Notation For a given species, let N = total number of trips, n = number of randomly sampled trips, W = total weight of fish caught from all trips, W{ = weight of fish caught on trip i, Wy = weight of fish for sort j caught in trip i, my = number of clusters sampled from sort j on trip i, mi = number of clusters sampled on trip i, m = number of clusters sampled over n trips, Wi = Z. Wy where L{ is the number of sorts 3 in trip i, yv-k = number of fish of the species in cluster k from sort j of trip i, Yy- = total number of the species caught from sort j of trip i, Y = total number of species caught from all trips, Y = mean catch per cluster for the species, yv- = 2. yyklrriy = unbiased estimate of Yv-, wijk = weight of the A;th cluster from the jt\\ sort of the ith trip, Wi Mi = -5=-^ where w,- = 2. 2. Wi^lZ. m^ = w{ j k l3k j y 410 dEjri. osxmr LiLlvkj ov^iuiYiii*n.v^in.u r 1011 um^iyiiiuo average weight of sampled clusters in the ith trip. If Wx is a constant, its estimate w will be given by w = 2 X 5! wilkIZ. 2. mvj. In practice, N and Mt i j k J i j J will not be known and will be estimated by N = wW - - W = WIW; M{ = -zr respectively, if Wl is a " w constant = w (say). Ratio Estimates of Mean and Total The ratio estimate of mean catch (Y) per cluster is n n I mm I wt y, YP = (1) I Mi I Wi where yx = 1 M^-/Z My = I W^/I ^ California Fish and Game. The reasons for failure to collect the data are discussed in the section on Collection of Representative Data-Measurement Er- rors. The above estimators are, however, recom- mended for use in situations where the problem does not exist and, in particular, for single species where the categories are based on size. The estimates of error are given in Equations (4) and (5). Estimation Ignoring Category Variation Within Sampled Trips Assume that a cluster is selected at random from all possible clusters in a sampled trip. In other words, we ignore categories altogether both in sam- ple selection as well as in estimation. Valid ratio estimates Y^ of Y and Y^ of Y are respectively given by Y,„ = 1 Wi LR I Wi Vi „ w = W (3) The ratio estimate of total catch Y is W - Y« - i Y°- (2) The above estimators recommended for use are not workable in rockfish sampling because the sampler failed in almost all cases to subsample from more than one category in a sampled trip as would be seen from a sample of basic data for 1982 (Table 1) available for Eureka from the Department of Table 1 .—Distribution of landing weights (lb) from all categories and from the sampled category for Eureka for 1982. Number of Weight of Weight of all clusters Market all fish fish for the Sample no. sampled category (IV,-) in a category in (boat trip) (m,) sampled1 given trip a given trip 1528 1 269 26,550 24,176 1529 1 250 4,133 445 1530 2 269 59,218 58,239 1531 1 269 20,511 15,987 1533 1 269 35,022 14,661 1534 1 269 20,757 20,705 1535 1 269 15,812 8,436 1536 1 250 1,975 1,010 1537 1 250 16,055 1,075 1541 3 269 65,837 65,837 'Shows the code number of categories which are based on species, size, and quality. Note: In all cases, only one of the categories could be sampled from a given trip. In boat 1541 there was only one category (269) of fish. Note these equations are essentially the same as Equations (1) and (2) except that we now assume that a cluster is randomly selected from all possible clusters in a sampled trip where W{ is the total landing weight from all categories for the ith boat trip in the sample (W = X W{). In practice, the X sampler would tend to subsample from a category which is accessible and is preponderant. This may lead to some bias in the estimate though its contribu- tion to the total error will be negligible, since this would occur at the second stage of sampling. The estimates of variance of estimated total and mean are approximately given by HYW) = JL (1 . /l)8f + MLzJ^i n nm v(Ym) ± (^)2 v{Ym) (4) (5) where s2h = Z. K\2®i- Ym)2. W n - 1 n — m i n w, 2 „2 s2i m,- (6) 411 nSHfcKY BULLETIN: VUL. 84, NU. Z and si = Z (ylk - ^/(m, - 1); W = I W^n; /! - Z W,IW;f2 = W,- w (7) We will consider an operationally feasible plan in which sample trips at a port during a month are poststratified into categories and clusters are sub- sampled from each category; where one or more categories are missed due to inadequate field staff and/or management problems, clusters should be selected from other boat trips containing the missed categories. Assuming that the cluster weight of the unequal cluster size varies over trips, i.e., w^ = Z Z wijlc/ J k Z m„ estimates of mean and total are j 3 n n £ Z wA Z wA (8) Z Wilwt Z Wi where Rt = -=?- ; viY^) and viY^) can be obtained similar to Equations (4) and (5). Yj = Vj 5 W v(Y3R) = Z 3 + 17? y W^ (11) j 4) is selected for each category and (b) the landing weights are available by categories after the season to serve as weights at the estimation stage. The minimum number in (a) is mainly based on limitations of field staff and budget restrictions. The ratio estimates of mean catch per cluster, total catch, and their errors, assuming clusters of equal size and using categories as domains of study are given by Y3R = Z W^y/Z Wfi Y3R = Z Yj (9) where y3 = Z W^lJ. Wl} (10) and y, = Z wdtdl. WJw. ir"iy-r xy^i) where iL = !*; A = Z iL-WyZ Wu. y w..' J y lJ i lJ (15) (16) If yij is small compared to N3 and if the same sub- sampling strategy is applied to each of the sample landings, we have, ignoring contribution due to second-stage sampling units, Vjffij) = nj(nj ~ 1) W (R^ - R3)2. (17) Another estimator v2(R3) is the jackknife v&lj) = ■ - Z (Ri3- - R3f (18) thi 412 SEN: SAMPLING COMMERCIAL FISH LANDINGS where R , + 72 W and i? l^y + ... + W{l_1)rlj + W(l+l)Ki + ... + W%i (19) i-^z*;, Thus i? y is obtained by omitting trip i from the sample for sort j and calculating &.■ instead of Rrj as in Equation (16). Hence, for category j of a species or v(y3.) = w^o^ (20) where vx(Rj) and v2(-R/) are given by Equations (17) and (18). For estimate of total over all sort groups for a species ya„ = I y, AR (21) v(?4*) = I i;(iy + 2 1 1 cov(fv fy (22) A simpler formula viY^) = Z. v(Y„) can be used 'j ' where subsamples from different categories are from different boat trips and are, therefore, in- dependent. It is, however, more reasonable to assume that the frequency distribution of fish caught is more uniform within a category so that cluster weight would be approximately a constant within a category. If so, the estimates of mean and total are given by Y5R = I W&lL Wfi Y5R = I Y} (23) J 3 3 where y3 = X W^fL WtJ; Y3 = ^W^ WJ 1 W, (24) Wa and Wj is the simple mean weight of clusters in the jth group. Where the assumption of constant cluster weight within a category is not valid, the more general results given in Equations (14) and (15) should be used. Comparison of Methods: Ignoring Category Variation Versus Poststratification by Categories We will compare the efficiency of the estimators (3), ignoring variation due to categories, with the estimators (9), based on poststratification of land- ings by categories at a port during a month. The analyses were based on Eureka and Monterey data for 1982. The coefficients of variation (c.v.) of mean catch per cluster for a species based on categories as domains of study (method 2) were in almost all cases lower (Table 2) than ignoring category varia- tion (method 1). Since method 1 results in under- estimation of c.v.'s because sampling is actually based on a stratified random sample instead of a simple random sample, the increased precision of method 2 is all the more striking. The c.v. of the estimated mean catch by sex-age groups for a species for which the number of sam- ple landings were MO (Table 3) were in all cases less for method 2 than for method 1. It may, however, be pointed out the c.v.'s are likely to be affected by factors such as growth, maximum age, and max- imum size of fish. These have not been considered in this study. Thus, estimates based on categories as domains of study proved more efficient than ignoring categories altogether. Besides, method 2 has the added advantage of providing estimates by Table 2.— Coefficient of variation (c.v., in percent) of mean catch by species at Eureka and Monterey based on the two methods during 1982. Location and species Sample size (number of boat trips sampled) c.v. (0/0) Method 11 Method 22 Eureka Widow rockfish Chilipepper Bocaccio Monterey Widow rockfish Chilipepper Bocaccio 88 88 88 54 54 54 11.48 30.83 26.01 18.31 15.68 12.57 7.33 32.12 24.40 6.62 13.92 10.32 1Method 1, based on random categories (i.e., ignoring stratification by categories). 2Method 2, based on categories as domains of study. 413 FISHERY BULLETIN: VOL. 84, NO. 2 Table 3.— Coefficient of variation (c.v., in percent) of mean catch by species-sex-age1 group at Eureka and Monterey based on the two methods during 1982. Eureka Monterey Number of boat Sex Age (yr) c.v. (%) Number of boat trips sampled Sex Age (yr) c.v. (%) trips sampled Method 1 Method 2 Method 1 Method 2 17 18 11 11 15 19 M F F F M F 7 7 13 12 6 6 19.71 13.50 39.98 34.77 30.10 35.87 Widow rockfish 18.83 10 10.94 10 Chilipepper 24.89 24 31.21 21 Bocaccio 19.82 14 32.45 20 F F F F M F 13 12 9 7 7 7 39.98 35.16 18.48 22.09 27.46 24.34 24.29 20.49 7.63 9.81 12.45 10.06 1Age-sex groups for which primary sampling units (landings) are >10. market categories which is of considerable economic importance. COST FUNCTION The components cx and c2 were estimated at Consider the cost function C = cxn + c2nm (25) where cx is the average cost (in minutes) per boat trip due to transport, contact, and delay in making a contact, c2 the average cost in data collection (identification of species, sex, length, otoliths, etc.) per cluster within clusters per boat trip and C is the total cost involved in visiting the primary sampling units (boat trips) and collecting data from the n boats with an average of m clusters per boat sampled. Data collected at Tiburon by the California Depart- ment of Fish and Game and the National Marine Fisheries Service show that c = 111.80 min, c2 = 58.3 mm so that — = 2 apply. However, from more C2 recent studies conducted — = 3. c2 Activity Transport Contact Delay (off loading, etc.) Data collection Species1 Sex, length Otolith Preparation time Percent 50.0 5.0 13.0 68.0 Percent 7.7 5.8 10.8 7.7 32.0 Mean (in minutes) 81.7 8.7 21.4 111.8 Mean (in minutes) 14.0 10.6 19.7 14.0 58.3 Excluding samples dominated by single species. Minimizing Equation (4) subject to Equation (25) for the optimum allocation we have ™opt = (26) Table 4. — Optimum values of m for estimating species catch per cluster by categories for different variance and cost ratios, 1978. Species Category1 n si s2 m clc - ^as C1'C2 - 2 cjc2 = 3 Eureka Bocaccio 250 25 1.80 3.01 2.16 3.86 4.73 Chilipepper 250 13 24.45 3.13 1.92 0.52 0.64 Widow rockfish 250 11 59.49 8.71 2.46 0.56 0.68 Monterey Bocaccio 253 31 95.15 4.20 1.97 0.63 0.77 Chilipepper 253 33 43.71 4.16 1.94 0.45 0.55 Widow rockfish 253 12 22.38 4.66 2.00 0.68 0.84 'Code numbers of categories which are based on size, species and quality. 414 SEN: SAMPLING COMMERCIAL FISH LANDINGS The variation among clusters (sf) in different land- ings at Eureka and Monterey for 1978 was in almost all cases greater than between clusters within the same landings (Table 4); also the optimum number of clusters per boat for estimating species number was mostly unity. Data from other ports follow the same pattern. Since a minimum of two clusters is needed to provide an estimate of between cluster within trip variation, a subsample of two clusters per category per trip is recommended. In practice, it is preferable to select a systematic sample of clusters separated in time. VARIANCE COMPONENTS: SPECIES-AGE AND LENGTH GROUPS A two-level nested analysis of variance for length and age with unequal sample size for species based on sample landings at ports during 1979 (Table 5) shows that both the variation, because of length and age, was generally high among sample landings compared with clusters within landings. Also, varia- tion between clusters was generally of the same order as within clusters, and the optimum number of clusters was <2. Data for other ports and years (not shown in the table) mostly supported the findings. On the whole, both the variation in species number (Table 4) as well as in length and age (Table 5) was consistently high among sample landings relative to between clusters within landings; also, variation among clusters was not significant compared with variation within clusters. Hence, for precise estima- tion of species number, length, and age„composition for a category at a port during a season, data should be collected from a large number of landings and from few clusters (two) from a category within a sample landing. RELATIVE EFFICIENCY OF ESTIMATORS USING POSTSTRATIFICATION Consider the three estimators of total catch for a sort of a species at a port during a year. We will use the same selection procedure with poststratifica- tion by sorts but different estimation procedures. Y, = 7I» nj i=1 Vij (27) ^ W&q Wj Yj i w* W: Yk = RjWj (28) (29) where Rj is given by Equation (16), y{j is the simple mean of species number per cluster for sort j from the ith sample, Yj is the same as Equation (24) with a constant cluster weight within a sort group, and Yj is a more general estimator based on the assumption that cluster weight varies among trips. For v(Yj) use W2jv2{R]) where v2(Rj) is the jack- Table 5.— Two-level nested ANOVA of length and age of pie sizes by ports during 1979. MS = mean square; F observed probablity level. species with unequal sam- = F-RATIO, Statistic; P = Age Length Source df MS F P df MS F P Widow rockfish at Eureka Samples 15 34.45 4.75 <0.005 37.86 3.09 <0.025 Clusters (within samples) 13 7.25 1.19 0.35 12.27 1.43 ~0.18 Within clusters 320 6.09 8.58 Chilipepper at Monterey Samples 43 31.74 4.05 <0.001 48 145.20 4.02 <0.001 Clusters 39 7.84 1.80 ~0.001 44 36.10 1.43 ~0.035 Within clusters 320 4.35 971 25.25 Bocaccio at San Francisco Samples 10 84.97 6.95 <0.001 10 317.88 6.98 <0.001 Clusters 15 12.23 1.20 ~0.30 16 45.55 0.80 ~0.75 Within clusters 225 10.20 227 57.11 415 FISHERY BULLETIN: VOL. 84, NO. 2 knife estimator of Equation (18) and for v(Y.) see Sukhatme (1954). Yj is generally subject to considerable bias. The c.v. of total catch of bocaccio, chilipepper, and widow rockfish for different categories by port-year groups (Table 6) show that the estimators Yj and Yj are highly efficient compared with Yj', also, Yj turns out to be slightly superior to Yj since the jackknife estimator v2(Yj ) is an underestimate and does not take into account the contribution of the within component of variance. Thus, the empirical evidence supports strongly the use of the estimator Yr Table 6.— Coefficient of variation (in percent) of estimates of total catch of bocaccio, chilipepper, and widow rockfish per cluster by ports during 1978 and for different categories for the three estimators 9\, Yh and V", . Port Number of boat trips Category sampled *7 % \ Bocaccio San Francisco Fort Bragg Monterey Eureka 253 250 253 250 20 86 31 25 Chilipepper 13.51 16.21 12.07 40.11 10.24 7.36 17.93 26.00 11.64 8.14 19.51 29.84 Eureka 250 13 Widow rockfish 37.66 34.52 42.33 Monterey Eureka 250 250 12 11 111.20 72.69 43.47 27.81 68.29 33.90 AGE-COMPOSITION: DOUBLE SAMPLING Studies mentioned in the Introduction section have shown that since aging from otoliths of each individual fish in a sample is more expensive than an easily measured quantity such as length, it may pay 1) to choose a random subsample from the whole sample of length measurements for age determina- tion or 2) stratify the sample according to length classes and choose a subsample from each class for age determination. The technique is profitable only if the correlation between length and age is fairly high. It may be recalled that considerable bias is in- troduced by applying age-length keys developed dur- ing a year to subsequent years. Both Kimura (1977) and Westrheim and Ricker (1978) showed that age- length keys can yield most inefficient estimates of numbers-at-age with substantial overlap of lengths between ages. In the latter case the correlation be- tween length and age will be low for the larger and the very small sizes. Consequently, we will need a higher sampling intensity at the tails to provide reliable estimates of age for such sizes. In the construction of length strata for selection of the subsample, additional questions arise on 1) number of strata to choose, 2) strata boundaries to decide, and 3) the number of sampling units to be allocated to each stratum for deriving maximum gain from double sampling. These are discussed as follows. Number of Strata The values of V(yst)/V(y) (Cochran 1977) are given below as a function of L, the number of strata using the linear model y = a + fix + £ (30) where y is the length, x the age of female widow rockfish and ViVst) V(y) L2 + (1 - p2) (31) where P is the correlation between length and age in the unstratified sample and L the number of strata. It can be shown for this model that when L > 6 and p > 0.95, there is hardly any gain due to stratification (Table 7). The improvement in stratification is highest for data set 1 for which p2 = 0.7004 and lowest for set 3 for which P2 = 0.5278. The results for the regression model indicate that unless p exceedes 0.95, little reduction in variance is to be expected beyond L = 6. Data sets 1, 2, and 3 support this conclusion. In fact, there does not seem to be any profit resulting from in- crease in strata beyond L = 5. Strata Boundaries For the length-age strata on 239 females (widow rockfish) landed during 1982 at San Francisco and the rule based on the cumulative of \Jf{y) (Cochran 1977) where y denotes the length in centimeters, the nearest available points for the two strata are Stratum Boundaries 36-47 cm 48-55 cm Intervals on cum \fi 18.70 23.72 416 SEN: SAMPLING COMMERCIAL FISH LANDINGS Table 7.— V(yst)IV(y) as a function of L for the linear regression and for some actual data. Linear regression nodel p = Data set L 0.99 0.95 0.90 0.85 1 2 3 2 0.265 0.323 0.392 0.458 0.4747 0.5114 0.6041 3 0.129 0.198 0.280 0.358 0.3774 0.4209 0.5308 4 0.081 0.154 0.241 0.323 0.3434 0.3892 0.5052 5 0.059 0.134 0.222 0.306 0.3276 0.3746 0.4933 6 0.047 0.123 0.212 0.298 0.3154 0.3740 0.4890 oo 0.020 0.098 0.190 0.277 Type of data X Age y Length Set Data (yr) (cm) Source 1 Female widow rockfish (532) 1982 1982 Department Monterey, San Fi ancisco (Jan. -Mar. ) (Jan.-Mar.) of and Bodega Bay California Fish 2 Female widow rockfish (444) 1981 1981 and Eureka (Jan. -Sept.) (Jan. -Sept.) Game and 3 Female widow rockfish (328) 1980 1980 Tiburon Eureka (Apr.-Dec.) (Apr.-Dec.) Laboratory It turns out that the division point is approximate- ly the same for young as well as old widow rockfish. For length-age data (1981) based on 444 females (widow rockfish) landed at Eureka, the boundaries using 2 and 3 strata are Stratum Boundaries Intervals on cum y£ 1 31.5-47 cm 17.70 46.5-55 cm 29.01 Stratum 1 2 3 oundaries 31.5-46 cm 46.5-49 cm 49.5-55 cm itervals on cum \/f 17.70 13.12 15.89 Optimum Allocation Plan Double sampling with regression is more efficient than single sampling (when the first sample is measured for age alone) for the same cost if p2> 44 c 1 + (32) where P is the correlation between length and age of fish, c and c are respectively the costs of aging and measuring a fish. Assuming that the average cost of aging a rockfish (including small and large fish) is 6 min and of measuring it is 1.2 min (estimates based on measurements by W. Lenarz of Tiburon Laboratory), we have from Equation (32) or p2 > 0.5555 p > 0.7453. For the three data sets (Table 7) the values of p2 are respectively 0.7004, 0.6515, and 0.5278 so that Equation (32) is approximately satisfied. However, neither p nor — - are large enough to suggest that double sampling will be much more efficient than single sampling. We will illustrate the use of double sampling for stratification by analyzing 1981 length-age data at Eureka to estimate the proportion of female in age group 11, based on a sample of 444 fish. For the three length strata, h = 1, 2, 3 with stratum bound- aries based quadratic fit of length on age are 31.5-43, 43.5-49, 49.5-55. (Note this is different than bound- aries based on length only.) Also 417 FISHERY BULLETIN: VOL. 84, NO. 2 = 1.2 min, Cj = w1 = Si = 3.8 min, c2 = 3.8 min, and c3 = 8 min 0.0653, w2 = 0.5451, and w3 = 0.3896 0.1825, s2 = 0.4966, s3 = 0.1503, and s = 0.4343 where wlf w2, and w3 are the proportions of fish in the sample, c0 is the cost of measuring a fish and c1( c2, c3 are respectively the costs of aging them in the three length groups. From Cochran (1977, p. 331) we have (Pst) = J_ C* I whshy/ch + (S2 - Z whsl)m\f& (33) = 0.8915/C* where pst is the estimated proportion and C* = E(c) = £"(c0 n + Z. chnh) with nj = 14, n2 = 120, n3 = 48 and ri = 444. The efficiency of double sampling with respect to single sampling is given by Vsr,(p)IVmm(Pst) = 1-21 where vsrs(p) = 0.1885/-^-, i.e., double sampling is 27% more efficient than single sampling. How- ever, as noted by Ricker (1975) the increase in ac- curacy achieved by combining a length sample with a smaller age sample may not be great unless fish used for age determination is taken from the same stock, during the same season and using gear having the same selective properties as the length-fre- quency samples. This point will generally be met if fish are subsampled systematically for age from fish arranged in increasing (or decreasing) order of length from a port-month stratum. Our studies have shown that the best length-age fit does not change significantly if age determination is made on every other fish arranged in ascending order of length. It is difficult to obtain reliable estimates of the numbers at age for the extremely small or larger sizes because lengths cannot be used for estimating age. There is need for search for other auxiliary variables (other than length) associated with age and for increase in sampling rate at the tails. In double sampling where lengths are obtained in the first phase, a number of small clusters may be used separated in space and time to provide a large number of fish at the tails for estimating numbers at age. The extent of bias in estimation of numbers at age through length-age key approach may be tested by Monte Carlo simulation. COLLECTION OF REPRESENTATIVE DATA-MEASUREMENT ERRORS Owing to uncertainty of arrival times and vary- ing unloading procedures, no objective method is available to ensure random sampling of the trips. When the vessels return to port, they are usually available for sampling except when they are tran- shipped immediately due to inclement weather, lack of processing facilities, uncooperative buyers, or unscheduled deliveries at short notice. It is, how- ever, not unreasonable to regard a set of sample landings during a week at a port as random and representative of the totality of all landings at the port for the month. Although rockfish are landed by categories, which are mostly determined by market agreement based on size, composition, and condition of the catch, the number of categories per delivery cannot be pre- determined. This number would vary from delivery to delivery and from dealer to dealer. Also, there are no guarantees that a complete boat sample, covering clusters from each category, can be taken on any sampling day and some of the categories are actually missed in sampling. Some of the possible reasons for missing the categories are 1) when landing weight would not occur during regular hours, one of the sorts may have already been shipped before the sample could arrive at the spot; 2) often one of the sorts may be quite small and there may be a buyer at the dock waiting for the fish to be taken away; 3) while the sampler is working on a sort, the other sort(s) will have either been pro- cessed or shipped away; and 4) the sampler may 418 SEN: SAMPLING COMMERCIAL FISH LANDINGS be prevented from taking a sample from another sort by the skipper who may not like some of his fish being cut and otoliths removed for biological studies. This may happen at ports where either pro- cessing facilities are inadequate or fish are bought by local merchants immediately after landing. The question arises if failure to sample from all cat- egories of a sample landing as originally planned would cause appreciable bias and loss in efficiency in the estimates of species catch and its distribution and whether a more efficient method could be developed that is operationally feasible. This point has been examined in the present paper. The present technique of selecting a cluster (box) of fish as second stage sampling unit is preferred to random selection of a specified number of in- dividual fish because in practice the potential of per- sonal bias of the sampler could be considerable. Often fish chosen by the latter technique are ones closest to the sampler or those that fell in a certain position. Tomlinson (1971) felt that in this approach the sampler may tend to choose a fish with certain qualities and thus may introduce procedural bias. The selection of a representative cluster would de- pend whether samples after sorting on the vessel come from bins, strap boxes, or off conveyor belts. Buyers from small markets occasionally select fish from the top of bins. Hence, to avoid this bias, it is preferable to select the cluster from the conveyor belt which exposes unsorted fish from the lower por- tion of the bin. However, where small market buyers do not buy fish, a cluster may be selected from a bin. Where many bins are present a systematic sam- ple of two clusters, preferably from the beginning and end of the trip may be selected. Where fish are graded on a conveyor belt before they enter the plant (e.g., Fieldslanding at Eureka) the sampler should try to intercept the landings prior to sec- ondary sorting or obtain separate weights for each subsort category. In general, selection of a cluster for a market category should be done before any presorting is done at the port. It has been pointed out earlier that bias may result from personal selection of fish within a cluster. If the sampler were to select a number of clusters with few fish per cluster, a cluster will on the average contain more big fish. This would lead to high non- sampling bias. Sometimes, the top few fish in a bin are selected and put there to impress small buyers. The resulting bias in selection can be avoided by taking all the fish in a cluster (e.g., 50 lb) from one side of the box. For obtaining reliable and comprehensive infor- mation on population characteristics, it is essential for the sampler to maintain good relationships with both the skipper and the buyer; this will depend to a large extent on the expertise of the sampler gained in the course of the field work. SUMMARY 1. The sampling scheme at a port during a month with poststratification of sampled trips into categories and subsampling of clusters from each category (see sections on Estimation with poststratification and Estimation ignoring category variation) is not workable for esti- mating rockfish catch since some of the categories may be missed in sampling due to in- adequate field staff and/or management problems. 2. For other commercial fish where the above problem does not exist and landing weights by categories are not available at the end of the season, the methods (see sections on Estima- tion with poststratification and Estimation ig- noring category variation) are recommended, e.g., for single species where the categories are based on size. 3. For estimating the catch of rockfish, a two- stage sampling plan is recommended with boat trips as first stage units poststratified into categories and clusters subsampled from a category; estimates are based on categories as domains of study with landing weights available for each category. A minimum of four landings or boat trips should be used for each category, to provide efficient estimates. With few categ- ories, this number is likely to be large. Where only one category is subsampled for each boat in the sample, v(Y3R) = ^ V(YX In 3 all other cases Equation (13) should be used. 4. The design described in the above paragraph is recommended for use in other fisheries where landing weights are available for each category. Equations (9) and (21) are recommended for the estimation of catch according as the clusters are of equal or unequal size. Equations have been provided for the more practical case when cluster weight can be treated as constant with- in a category but different among catego- ries. 5. Estimates of species catch by sex and age based on method 1 are less efficient than those based on method 2 which is based on categories as do- mains of study (Tables 2, 3). 6. Method 2 is preferred to method 1 when there 419 FISHERY BULLETIN: VOL. 84, NO. 2 is variation among categories. This is true for all fish. 7. With few categories (species-size-qualities) the chance of missing a category is reduced. Equa- tions (9) and (13) should be used for clusters of equal size and Equations (21) and (22) for un- equal size clusters. This result is, of course, ap- plicable to all commercial fish. 8. As far as practicable, selection of a cluster for a market category should be done before any presorting is done at the port either from bins, strap boxes, or off conveyor belts. 9. Variation (within categories) in length and age for a species was considerably higher among boat trips than among clusters within boat trips. Also, variation among clusters was not signifi- cant, compared with variation within clusters (Table 5). Hence, for precise estimation of species number, length, and age composition for a category at a port during a season data should be collected from a large number of landings and from few clusters from a category within a sample landing. This result should hold for all commercial fish. 10. For the cost function C = cxn + c2nm where Cj is the average cost (in minutes) per boat trip due to transport, contact, and delay in making a contact, c2 the average cost of data collection (identification of species, sex, length, otoliths, etc.) per cluster per boat trip and C is the total cost involved in visiting the primary sampling units (boat trips) and collecting data, the opti- mum number of clusters per sampled trip for a fixed cost for a category is two (Table 4). This should provide valid estimates of error as re- quired in Equations (13) and (22). 11. The principal contribution of the paper is that a minimum of four sample landings be sub- sampled for each category from a port-month stratum, i.e., about 1 per week and two clusters of 50 lb (25 lb for small fish) each should be sampled to provide port-year estimates with a reasonable degree of accuracy. If a category is infrequently landed, sampling should be directed towards the infrequent category, as long as the number of landings for the category is less than four per month. 12. The efficiency of the ratio estimator (Equation (28)) based on poststratification by categories at port-year level and using constant cluster weight within a category was compared with two other estimators, including the ratio esti- mator based on jackknife. Empirical evidence indicated that the ratio estimator using constant cluster weight within a category proved most efficient for estimation of species catch. 13. Age-length keys can yield most inefficient estimates of the numbers at age for extremely small and large fish. It is suggested that cluster sampling for length be based on a number of clusters separated in space and time; also, sam- pling for age should be intensified for small and large fish. This approach is applicable to all fish. 14. Double-sampling was adopted for estimating proportion of widow rockfish in 11-yr age group. A sample of fish was divided into 3 strata and optimum allocation for age was adopted within strata. The estimated proportion was 27% more efficient than if single sampling were adopted. The best length-age did not change significantly if age determination is made on every other fish selected in ascending order of length. The method is general and is applicable to all fish. ACKNOWLEDGMENTS Thanks are due to William Lenarz of Tiburon Laboratory for providing information on problems related to widow rockfish landings on the Califor- nian coast, to Candis Cooperider and Mark Allen for the computations done on data collected, to the field staff of the California Department of Fish and Game responsible for collection of relevant data, and to Norman Abramson, Director, Tiburon Laboratory for all the assistance rendered to me during my work in the Laboratory. My thanks are also due to Pat Dalgetty, Department of Mathematics and Statistics, University of Calgary, for assistance in typing the paper and finally to the referees for helpful comments. LITERATURE CITED Bartoo, N. W., and K. R. Parker. 1983. Stochastic age-frequency estimation using the von Ber- talanffy growth equation. Fish. Bull, U.S. 81:91-96. Clark, W. G. 1981. Restricted least-squares estimates of age composition from length composition. Can. J. Fish. Aquat. Sci. 38: 297-307. Cochran, W. G. 1977. Sampling techniques. 3d ed. Wiley and Sons, N.Y., 428 p. Fridriksson, A. 1934. On the calculation of age-distribution within a stock of cod by means of relatively few age-determinations as a key to measurements on a large scale. Rapp. P. -v. R6un. Cons. Perm. int. Explor. Mer 86:1-14. 420 SEN: SAMPLING COMMERCIAL FISH LANDINGS KETCHEN, K. S. 1950. Stratified subsampling for determining age determina- tions. Trans. Am. Fish. Soc. 79:205-212. Kimura, D. K. 1977. Statistical assessment of the age-length key. J. Fish. Res. Board Can. 34:317-324. KUTKUHN, J. H. 1963. Estimating absolute age composition of California salmon landings. Calif. Dep. Fish Game, Fish. Bull. 120, 47 p. Mackett, D. J. 1963. A method of sampling the Pacific albacore (Thunnus germo) catch for relative age composition. F.A.O. Fish. Rep. 3:1355-1366. Ricker, W. E. 1975. Computation and interpretation of biological statistics offish populations. Fish. Res. Board Can., Bull. 191, 382 p. Southward, G. M. 1976. Sampling landings of halibut for age composition. Int. Pac. Halibut Comm. Sci. Rep. 58:1-31. SUKHATME, P. V. 1954. Sampling theory of surveys with applications. Iowa State College Press, Ames, 491 p. Tanaka, S. 1953. Precision of age-determination of fish estimated by double sampling method using the length for stratification. Bull. Jpn. Soc. Sci. Fish. 19:657-670. Tomlinson, P. K. 1971. Some sampling problems in fishery work. Biometrics 27:631-641. WESTRHEIM, S. J., AND W. E. RlCKER. 1978. Bias in using an age-length key to estimate age- frequency distributions. J. Fish. Res. Board Can. 35:184- 189. 421 A VARIABLE CATCHABILITY VERSION OF THE LESLIE MODEL WITH APPLICATION TO AN INTENSIVE FISHING EXPERIMENT ON A MULTISPECIES STOCK Jeffrey J. Polovina1 ABSTRACT A variable catchability version of the Leslie model is developed which permits the catchability of one species to vary inversely with the abundance of competing species. This model is used to fit data from an intensive fishing experiment conducted on a multispecies bottom fish stock in the Marianas where catchability of a subordinate species is inversely related to the abundance of a more dominant species. Analysis of this multispecies intensive fishing experiment produced estimates of exploitable bottom fish density in the 150-275 m depth range of 10,156 fish per nmi2 or 1,354 fish per nmi of 183 m (100-fathom) contour. Intensive fishing of a closed population can produce data to estimate the initial population size and the catchability coefficient of fish stocks. Two frequently used models applied to intensive fishing data are the Leslie model and the Delury model (Ricker 1975). The Leslie model expresses catch per unit effort (CPUE) at any point during the period of intensive fishing as a linear function of the cumulative catch to that point, whereas the Delury model expresses the logarithm of CPUE at any point during the in- tensive fishing experiment as a linear function of the cumulative effort. From a statistical viewpoint the Leslie model is often preferable to the Delury model, since a predictive linear regression is used to estimate the parameters of both models and since typically catch is measured more accurately than effort. Both the Leslie and Delury models assume that catchability is constant during the period of inten- sive fishing. However, experience indicates that this assumption may not always be satisfied (Pope and Garrod 1975; Schaaf 1975; MacCall 1976; Ulltang 1976; Garrod 1977; Peterman and Steer 1981; Fox2). Several authors have found that competition for baits between fish of different size or species can alter catchability (Allen 1963; Rothschild 1967). In this paper a variable catchability Leslie model will be developed for multispecies application where, due 1 Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 3830, Honolulu, HI 96812. 2Fox, W. W. 1974. An overview of production modelling. U.S. National Marine Fisheries Service, Southwest Fisheries Center, Administrative Report LJ-74-10, La Jolla, CA. to species interactions, the catchability of one species is altered by the presence of other species. This variable catchability Leslie model will be ap- plied to multispecies intensive fishing data from snapper (family Lutjanidae) populations where the application of the constant catchability Leslie model leads to biologically untenable results. VARIABLE CATCHABILITY LESLIE MODEL The CPUE during a time interval t (CPUE(O) is defined as the product of catchability (q) and the mean population size (number of individuals) pres- ent during the period t (N(t)), thus CPUE(0 - qN(t). (1) Suppose that up to the beginning of period t, K(t) fish have been caught and removed. If the period t is relatively short, the population of fish closed or isolated, and the fishing pressure heavy enough so that it can be assumed that mortality from other fac- tors is negligible, then N(t) can be expressed as N(t) = N(0) - K(t), where N(0) is the initial population size at the begin- ning of the experiment (t = 0). Inserting this ex- pression for N(t) in Equation (1) produces the Leslie model: CPUE(0 = q(N(0) - K(t)). Manuscript accepted August 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. (2) 423 FISHERY BULLETIN: VOL. 84, NO. 2 Henceforth, this model will be referred to as the con- stant catchability Leslie model. In a multispecies situation, competition between species for baited hooks may produce a dominance hierarchy where some species are more aggressive feeders than others and effectively out compete the less aggressive feeders for baited hooks. The catch- ability of the species at the top of the dominance hierarchy, is independent of the presence of more subordinate species, while the catchability of those species not at the very top of the hierarchy will vary inversely with the abundance of the more dominant species. A simple model which describes the catch- ability of a subordinate species (q(s,t)) as a function of the cumulative catch and initial population size of the more dominant species, K(d,t) and N(d,0) respectively is q(s,t) = q(s)(K(d,t)/N(d,0)) (3) where q(s) is the catchability of the subordinate species in the absence of the dominant species. Com- bining Equations (2) and (3) produces CPUE(s,0 = q(s)(K(d,t)/N(d,0)) x (N(s,Q) - K(s,t)) (4) where V( ) and E( ) represent the variances and means, respectively. APPLICATION OF MULTISPECIES LESLIE MODEL TO SNAPPER INTENSIVE FISHING A 13-d intensive fishing experiment covering the period 10-19 April and 5-7 May 1984 was conducted at Pathfinder Reef (lat. 16°30'N, long. 143°05'E) in the Mariana Archipelago. Pathfinder Reef is a cir- cular pinnacle rising steeply from a depth of about 1,600 to 16 m beneath the surface At the 200 m con- tour, the diameter is about 0.8 nmi (Fig. 1). The snap- per population at Pathfinder Reef is a closed popula- tion for purposes of the intensive fishing since the closest bank is a small pinnacle 40 nmi to the north. Intensive fishing was conducted from the NOAA ship Townsend Cromwell using four bottom hand- lines on hydraulic gurdies targeting species in the 150-275 m depth range. Each day during the 13-d experiment, fishing was conducted around the en- tire perimeter of the bank. During the experiment 1,467 bottom fish were caught. Three lutjanids, Pristipomoides zonatus, P. auricilla, and Etelis car- bunculus, accounted for 1,317 fish or about 90% of the catch (Table 1). Fishing effort was measured in and by defining K(ds,t) q{s)(N(s,0)/N(dM^dB2 (3) becomes K(d,t)K(s,t), B\ = q(s)IN(d,0) Equation CPUE(s,0 = BlK(d,t) - B2K(ds,t). Estimates of SI and .62 are obtained from multiple linear regression and the estimates of N(s,0) and q(s) are computed as N(s,0) = B1IB2, and q(s) = N(d,0)B2. The estimate ofN(d,0) is determined from the con- stant catchability model. As is evident from Equa- tion (4), the estimate of N(s,0) is independent of the estimate oiN(d,0). Estimates of the variance of the estimate of N(s,0) are obtained from estimates of the means and variances of the estimates of 61, and B2 and an exact expression for the variance of a ratio (Frishman 1975). Thus, V(N(s,0j) = V01IB2) V(Bl)[E(B2)f - V(B2)[E(Bl)f (E(B2)f [V(B2) + [E(B2)f] (5) Figure 1— Bathymetric chart of Pathfinder Reef showing the segments of the 100-fathom (183 m) contour used to partition daily fishing effort. 424 POLOVINA: CATCHABILITY VERSION OF LESLIE MODEL Table 1 .—Species composition of bottom fish catch at Pathfinder Reef. Percent Species Number caught of catch Lutjanidae Aphareus rutilans 4 0.27 Aprion virescens 1 0.07 Etelis carbunculus 314 21.40 Pristipomoides auricilla 262 17.86 P. filamentosus 16 1.09 P. flavipinnis 7 0.48 P. zonatus 741 50.51 Carangidae Caranx lugubris 83 5.66 Seriola sp. 32 2.18 Serranidae Cephalopholis igarasiensis 2 0.14 Epinephelus cometae 2 0.14 Saloptia powelli 3 0.20 Total 1,467 100.00 line-hours. As is indicated in Figure 1, the circum- ference of the reef can be divided into three segments— north, west, and south-southeast, each having similar species composition (Table 2). Further, an attempt was made daily to allocate a consistent proportion of the day's fishing effort to each seg- ment. The proportion allocated to each segment was influenced by the length of each segment and wind Table 2. — Species composition for the three segments of the cir- cumference of Pathfinder Reef (see Figure 1). South- Southeast North West Species No. % No. % No. Pristipomoides zonatus P. auricilla Etelis carbunculus 358 51 170 24 171 25 160 68 37 16 39 17 223 55 104 58 14 27 and current conditions. On the average, the propor- tion of the total daily effort allocated to each seg- ment was 0.45 on the south-southeast, 0.21 on the north, and 0.34 on the west. A chi-squared test ap- plied to the daily allocation of fishing effort indicates that there was no significant departure (P = 0.89) from this allocation during the course of the fishing experiment. Since the effort was reasonably con- stant over the duration of the experiment and the entire reef was fished each day, catch, effort, and CPUE computed on a daily basis were used in the analysis. An adjustment to cumulative catch sug- gested by Chapman (1961) was subsequently shown to improve the model fit in the Delury model (Braaten 1969). This adjustment computes cumula- tive catch for interval i as the cumulative catch to interval i plus one half the catch during interval i. This adjustment compensates for the decline in CPUE within each time interval. The adjusted cumulative catch is used as the independent variable in all subsequent analyses (Table 3). Plots of CPUE against adjusted cumulative catch for each of the three species of snappers show a decline in CPUE for P. zonatus, a slight decline for E. carbunculus, and an increase for P. auricilla (Fig. 2). A regression line fitted to these data results in negative slopes for P. zonatus (P = 0.0007) and E. carbunculus (P = 0.05) and a positive slope for P. auricilla (P = 0.008). The constant catchability Leslie model fitted the P. zonatus data well and resulted in an R2 of 0.71 and a pattern of residuals which supports the linear model. The estimates of N(0) and q for P. zonatus from this fit are 1,066 fish and 0.0038 per line-hour. Due to the selectivity of the fishing gear, N(0) estimated from this intensive fishing data does not represent total population size Table 3.— Daily catch, effort, catch per unit of effort (CPUE), and adjusted cumulative catch for Pristipomoides zonatus, P. auricilla, and Efe//'s carbunculus Effort Tota I Pristipomoides zonatus Adjusted P. auricilla £fe//s carbunculus Adjusted Adjusted Adjusted Date (line- Catch cumulative Catch cumulative Catch cumulative Catch cumulative 1984 hours) (no.) CPUE catch (no.) CPUE catch (no.) CPUE catch (no.) CPUE catch Apr. 10 27.5 152 5.53 76 98 3.56 49 12 0.44 6 42 1.53 21 Apr. 11 23.7 150 6.33 227 111 4.68 153.5 17 0.72 20.5 22 0.93 53 Apr. 12 21.3 100 4.67 352 47 2.21 232.5 12 0.56 35 41 1.93 84.5 Apr. 13 29.7 139 4.68 471.5 91 3.06 301.5 29 0.98 55.5 19 0.64 114.5 Apr. 14 29.3 112 3.82 597.0 66 2.25 380 17 0.58 78.5 29 0.99 138.5 Apr. 15 17.5 84 4.80 695.0 50 2.86 438 13 0.74 93.5 21 1.20 163.5 Apr. 16 30.7 129 4.20 801.5 67 2.18 496.5 26 0.85 113 36 1.17 192.0 Apr. 17 21.4 65 3.04 897.5 38 1.78 548 12 0.56 132 15 0.70 217.5 Apr. 18 22.4 81 3.62 970.5 41 1.83 587.5 15 0.67 145.5 25 1.12 237.5 Apr. 19 21.6 60 2.78 1,041 28 1.30 622.0 17 0.78 161.5 15 0.69 257.5 May 5 20.3 82 4.04 1,112.5 40 1.97 656 29 1.43 184.5 13 0.64 271.5 May 6 22.8 91 3.99 1,199.0 35 1.54 693.5 35 1.54 216.5 21 0.92 288.5 May 7 24.1 72 2.99 1,281 30 1.25 726.0 27 1.12 248.5 15 0.62 306.5 425 FISHERY BULLETIN: VOL. 84, NO. 2 5.0 1 r 4.0 UJ a. u 3.0 2.0 1.0- -i r Pristipomoides zonotus Table 4.— Percent of catch by depth (in fathoms, 1 fathom = 1 .83 m). UJ a. u I00 200 300 0 100 CUMULATIVE CATCH 200 300 400 Figure 2.— Daily catch per unit effort (CPUE) and adjusted cumulative catch for Pristipomoides zonatus, P. auricilla, and Etelis carbunculus. but rather the population size of those fish that can be caught by the fishing gear which will be termed the exploitable population. Although the constant catchability Leslie model does not explain as much of the variation for E. carbunculus (R2 = 0.35) as it does for P. zonatus, the regression is significant and the pattern of residuals supports the linear fit. The estimates for catchability and initial exploitable population size for E. carbunculus from the fit of this model are 0.0025 per line-hour and 583 fish. The positive slope for the regression of CPUE on cumulative catch for P. auricilla does not make sense biologically under the constant catchability Leslie model. The depth of capture data show that P. zonatus and P. auricilla were caught in the same depth range, whereas E. carbunculus was typically caught at somewhat greater depths (Table 4). Thus, species interactions would most likely occur between P. zonatus and P. auricilla. If P. zonatus is more ag- gressive than P. auricilla in pursuing fishing baits or in some other way affects the behavior of the lat- ter, then the initial catchability for P. auricilla will Depth Species <100 100-120 >120 Pristipomoides zonatus P. auricilla Etelis carbunculus 15.1 12.6 1.9 71.7 79.0 46.5 13.2 8.4 51.6 be low but will rise as the population of P. zonatus is reduced. Applying the variable catchability Leslie model to the P. auricilla data, with the assumption that P. zonatus is the dominant species and that P. auricilla is the subordinate species so that the catch- ability of P. auricilla depends on the population size of P. zonatus, results in the following relationship: CPUE(a,0 = q(a)(K(z,t)/N(z,0)) x (N(a,Q) - K(a,t)), (6) where q(a) is the catchability of P. auricilla in the absence of P. zonatus and N(z,0) and N(a,0) are the initial exploitable population sizes of P. zonatus and P. auricilla, respectively, and K(z,t) andK(a,t) are the cumulative catch of P. zonatus and P. auricilla to time t, respectively. Using the estimate of N(z,0), 1,066 fish, from the fit of the constant catchability model to P. zonatus data, Equation (6) has two unknowns to be esti- mated— q(a) and N(a,0). A multiple linear regression model estimates the initial exploitable population size of P. auricilla, N(a,0), at 2,007 fish and q(a) at 0.00087. The variable catchability Leslie model fits the P. auricilla CPUE data well and produces an R2 of 0.89 (Fig. 3). The estimates of initial popula- tion sizes for the three species are summarized in Table 5 together with their 95% confidence inter- vals. For the constant catchability model, the population size confidence interval is computed from a relationship derived by Delury (1958), whereas the confidence interval for the variable catchability model is computed from the variance expression given in Equation (5). DISCUSSION The constant catchability Leslie model fit the P. zonatus and E. carbunculus data well but was not appropriate for the P. auricilla data. The variable catchability Leslie model fit the P. auricilla data well and provided a plausible explanation for the observed increase in CPUE. Given that there was a time delay between the first 10 d of the intensive 426 FOLOV1NA: CATCHABILITY VEKS1UN OE LESLIE MODEL fishing (10-19 April) and the last 3 d (5-7 May), and that the greatest increase in the catchability of P. auricilla occurred after the time delay, it is possible that the increase in catchability might have a time lag component associated with it. However, given the short time series of data, it would be difficult to test the appropriateness of a more complicated time lag model. Based on the fit of these two models the initial exploitable population of the three species in the 150-275 m depth range at Pathfinder Reef is esti- mated at 3,656 fish (Table 5). If we assume, based on the species composition data (Table 1), that these three species represent 90% of the exploitable population then the total exploitable population at the beginning of the intensive fishing is 4,062 fish. 18 1.6 Q. O Pristipomoides ouricillo J I I L 0 25 50 75 IOO 125 150 175 200 225 250 275 CUMULATIVE CATCH Figure 3.— Daily catch per unit effort (CPUE) and predicted CPUE based on the variable Leslie model as a function of adjusted cumulative catch for Pristipomoides auricilla. From Figure 1 the length of the 183 m (100-fathom) contour is estimated at 3.0 nmi, and the area in the 180-300 m depth range is estimated to be 0.4 nmi2. With these area measures, density estimates of 1,354 fish per nmi of (183 m) 100-fathom contour and 10,156 fish/nmi2, are obtained for Pathfinder Reef. Estimates of bottom fish densities based on visual observation from a submersible at Johnston Atoll were 57,281 fish/nmi2 for the 92-183 m (50-100 fathom) depth range and 66,199 fish/nmi2 for the 1983-274 m (100-150 fathom) depth range (Ralston et al. 1986). These figures are considerably larger than both the point and interval estimates presented here. Significantly, the study of Ralston et al. (1986) also employed the Townsend Cromwell, and the catch rates were comparable at Pathfinder and Johnston (e.g., 3.18 bottom fish/line-hour for the latter). Thus the difference between estimates of standing stock is likely not due to differences in ab- solute abundance but rather to differences between exploitable population size and total population size. For example, at Johnston Atoll at least 69 species of fish were observed from the submersible, whereas only 10 species were taken by fishing gear in the same depth (Ralston et al. 1986). If the constant catchability Leslie model is applied to the pooled data for the three species, an estimate of exploitable population size of 2,689 is obtained, about 71% of the estimate of the exploitable popula- tion size for the three species when they are estimated separately (Table 5). Size-specific behavior has been raised as a factor which might affect catchability (Allen 1963). For all three species, there is no evidence of intraspecies size-specific behavior affecting catchability since for two of the species the constant catchability model fits well and for the third species, catchability depends only on the population size of an interact- ing species. Further, under the hypothesis that within a stock catchability is size-specific across the Table 5.— Estimates of population size and catchability for three species. Species Model f? Catch- ability SE Initial Confidence population interval size (95%) Pristipomoides Constant zonatus catchability 0.71 0.0038 0.0075 Efefe Constant carbunculus catchability 0.35 0.0025 0.0010 P. auricilla Variable catchability 0.89 0.00087 0.00031 Three species Constant pooled catchability 0.66 0.0022 0.0047 1,066 (803-1,691) 583 (361-3,011) 2,007 (261-5,727) 2,689 (1,955-4,535) 427 FISHERY BULLETIN: VOL. 84, NO. 2 range of exploitable size, intensive fishing would produce a substantial change in the population size structure. A plot of the mean fork length by day of fishing for the three species (Fig. 4) shows very lit- tle change in fork length even for P. zonatus where 68% of the exploitable stock is estimated to have been removed. Thus, the mean size of the fish in a catch may be a much less sensitive indicator of changes in the population size than catch rates, at least over the short term. ACKNOWLEDGMENTS I wish to thank Alec D. MacCall and William E. Schaaf whose reviews resulted in an improvement in the formulation of the variable catch Leslie model. This paper is a result of the Resource Assessment Investigation of the Mariana Archipelago at the Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA. 41 40 39 E o 38 x K O 5 37 O 36 35 h ! r Pristipomoides zonotus Pristipomoides ouricillo i i i i i _i_ 23456789 10 DAY 12 13 14 Figure 4.— Mean fork length for each day of fishing for Pristi- pomoides zonatus, P. auricilla, and Etelis carbunculus. LITERATURE CITED Allen, K. R. 1963. The influence of behavior on the capture of fish with baits. In The selectivity of fishing gear, Vol. 5, p. 5-7. Proceedings of Joint ICNAF/ICES/FAO, Special Scientific Meeting, Lisbon, 1957, Special Publications No. 5. Braaten, D. 0. 1969. Robustness of the Delury population estimator. J. Fish. Res. Board Can. 26:339-355. Chapman, D. G. 1961 . Statistical problems in dynamics of exploited fisheries populations. Proc. Berkeley Symp. Math. Stat. Probab. 4:153-168. Delury, D. B. 1958. The estimation of population size by a marking and recapture procedure. J. Fish. Res. Board Can. 15:19-25. Frishman, F. 1975. On the arithmetic means and variances of products and ratios of random variables. In G. P. Patil et al. (editors), Statistical distributions in scientific work, Vol. I, p. 401-406. Garrod, D. J. 1977. The North Atlantic cod. In J. A. Gulland (editor), Fish population dynamics, p. 216-242. John Wiley & Sons, N.Y. MacCall, A. D. 1976. Density dependence of catchability coefficient in the California Pacific sardine, Sardinops sagax caerula, purse seine fishery. Calif. Coop. Oceanic Fish. Invest. Rep. 18: 136-148. Peterman, R. M., and G. J. Steer. 1981. Relation between sport-fishing catchability coefficients and salmon abundance. Trans. Am. Fish. Soc. 110:585-593. Pope, J. G., and D. J. Garrod. 1975. Sources of error in catch and effort quota regulations with particular reference to variation in the catchability coef- ficient. Int. Comm. Northwest Atl. Fish. Res. Bull. 11: 17-30. Ralston, S., R. M. Gooding, and G. M. Ludwig. 1986. An ecological survey and comparison of bottom fish resource assessments (submersible versus handline fishing) at Johnston Atoll. Fish. Bull., U.S. 84:141-155. Richer, W. E. 1975. Computation and interpretation of biological statistics of fish populations. Fish. Res. Board Can., Bull. 191, 382 p. Rothschild, B. J. 1967. Competition for gear in a multiple-species fishery. J. Cons. 31:102-110. Schaaf, W. E. 1975. Fish population models: potential and actual links to ecological models. In C. S. Russell (editor), Ecological modeling in a resource management framework, p. 211-239. Johns Hopkins Univ. Press, Bait. Ulltang, 0. 1976. Catch per unit of effort in the Norwegian purse seine fishery for Atlanto-Scandian (Norweigian spring spawning) herring. FAO Fish. Tech. Pap. 155:91-101. 428 EARLY DEVELOPMENT OF THE LOPHIID ANGLERFISH, LOPHIUS GASTROPHYSUS Yasunobu Matsuura and Nelson Takumi Yoneda1 ABSTRACT Using larval specimens collected in bongo nets in southern Brazilian waters (between lat. 23° and 29 °S), early development of the lophiid anglerfish, Lophius gastrophysus, is described and compared with other lophiid species. Larval morphology of L. gastrophysus is very similar to that of L. americanus, having three conspicuous melanophores on the trunk and caudal region, but the former can be easily distinguished from the latter by the presence of two melanophores on the preopercular and suborbital regions and positions of the melanophores on the elongate ventral fin. The peculiar larvae of Lophius have been known since the description of the early developmental stage of L. americanus by Agassiz (1882). Their characteristic form with elongate dorsal and ven- tral fin rays makes them easily identifiable. Of the 25 species of the Lophiidae (Caruso 1981), larvae have been repeatedly described and discussed for L. piscatorius (Prince 1891; Williamson 1911; Stiasny 1911; Allen 1917; Lebour 1919, 1925; Bow- man 1920; Taning 1923; Arbault and Boutin 1968 Russel 1976) and for L. americanus (Agassiz 1882 Connolly 1920, 1922; Taning 1923; Berrill 1929 Dahlgren 1928; Procter et al. 1928; Bigelow and Schroeder 1953; Martin and Drewry 1978; Fahay 1983; Pietsch 1984). The larvae of two other species also have been described: L. budegassa (Stiasny 1911; Padoa 1956) and L. litulon (Tanaka 1916; Mito 1966). There is no literature on larval morphology of L. gastrophysus. During ichthyoplankton surveys along the south- ern Brazillian coast, many Lophius larvae were col- lected and identified as L. gastrophysus. This report gives a detailed comparative description of larval development based on 136 specimens collected dur- ing the past 13 years. MATERIALS AND METHODS Larval specimens used in this report were ob- tained from the collections of ichthyoplankton at the Instituto Oceanografico da Universidade de Sao Paulo. These samples were collected from the south- ern Brazillian coast using a 61 cm bongo net follow- ing the sampling method of Matsuura (1979) and preserved in 10% Formalin2 solution. Notochord length (NL) was taken from the tip of the upper jaw to the tip of the notochord. A total of 136 larvae (3.3-15.7 mm NL) of L. gastrophysus was used in this study. Specimens were measured with a micrometer in a stereoscopic dissecting microscope and illustrations were made with the aid of a camera lucida. MORPHOLOGY OF LARVAE The smallest identified specimens which were col- lected with plankton nets as free-living forms were about 3.3 mm NL, but they still had a large yolk sac. Fahay (1983) showed that the newly hatched larvae of L. americanus was as small as 2.5 mm long, and they were still encased in the egg veils (Fahay3). The reported size of newly hatched larvae of L. pisca- torius was 4.5 mm TL (Lebour 1925). Since the 3.3 mm larvae were not in perfect con- dition, we used larger specimens for the morpho- logical description. Preflexion larvae of L. gastro- physus have a slender body (Fig. 1A, B, C, D), but they later become robust form (Fig. IE, F). This change of body shape is partly a result of increase in body depth and partly due to enlargement of subepidermal space (Fig. 1C, D, E, F), which ap- pears, firstly, on the head region and later becomes larger and extends posteriorly, giving the larvae a balloonlike appearance. This subepidermal space consists of transparent, gelatinous connective tissue and is considered an adaptation to planktonic life 'Instituto Oceanografico da Universidade de Sao Paulo, Butanta, Sao Paulo 05508, Brasil. 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 3M. P. Fahay, Northeast Fisheries Center Sandy Hook Labora- tory, National Marine Fisheries Service, NOAA, Highlands, NJ 07732, pers. commun. July 1985. Manuscript accepted August 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 429 FISHERY BULLETIN: VOL. 84, NO. 2 430 Figure l.—Lophitis gastrophysus larvae from southern Brazil: A. 3.8 mm NL, B. 4.5 mm NL. Scale bar is 1.0 mm. MATSUURA and YOXEDA: EARLY DEVELOPMENT OF LOPHIID ANGLERFISH Figure 1.— Continued— Lophius gastrophysus larvae from southern Brazil: C. 5.6 mm NL, D. 7.8 mm NL. Scale bar is 1.0 mm. 431 rionr^ivi dullliu1!. vul, o1*, n\j. c Figure 1.— Continued— Lophius gastrophysus larvae from southern Brazil: E. 9.2 mm NL, F. 14.9 mm NL. Scale bar is 1.0 mm. (Tarring 1923). Notochord flexion starts at about 9 mm NL (Fig. IE). As shown in L. piscatorius larvae (Taning 1923), the laterally compressed larval form changes gradu- ally during their planktonic stage toward the dorso- ventrally depressed shape of juvenile and adults. The largest larvae examined, 15.7 mm NL, had not yet achieved the juvenile stage, but a similar tendency was observed. For example, the maximal breadth of the head in 3.5 mm larva is only 22%, but that in 15.7 mm larva is about 40% of body length. The proportion of body depth also shows a similar tendency, i.e., it starts at 30% at 4 mm and attains 45% of body length at 15.7 mm. The proportion of head length starts at about 23% at 4.5 mm and at- tains almost 45% at 15.7 mm NL. Statistics describing regressions of different body parts in relation to body length are shown in Table 1. The regressions lines of head length and body depth showed an inflexion at the size of 7.6 mm NL, while those of other body parts were linear for the size range 3.2-15.7 mm NL. Thus, the regressions 432 MATSUURA and YONEDA: EARLY DEVELOPMENT OF LOPHIID ANGLERFISH Table 1 .—Statistics describing regressions relating notochord length with length of different body parts of Lophius gastrophysus larvae, a and b = constant (y = a + bx), r = correlation coeffi- cient, n = number of specimens. Characters (x) Size range of notochord length (V) (mm) a b r n Head length 3.2- 7.5 -0.18334 0.27117 0.68827 97 7.7-15.7 -2.11501 0.56058 0.94483 27 Body depth 3.2- 7.5 0.10498 0.28333 0.74299 99 7.7-15.7 -2.49286 0.65397 0.92825 27 Preanal distance 3.2-15.7 - 1 .54475 0.77254 0.96482 27 Predorsal distance 3.2-15.7 - 1 .65336 0.74641 0.96717 27 Eye diameter 3.2-15.7 - 0.00095 0.10383 0.94353 124 Length of the second dorsal spine 3.2-15.7 -3.19397 0.99987 0.90707 96 Length of the third ventral fir i ray 3.2-15.7 -3.47484 1.14083 0.90497 102 lines of the former were calculated in two size ranges. PIGMENTATION Lophius gastrophysus larvae develop a distinct pattern of melanophores. Since early stage (Fig. 1 A), there are three large pigment bars on the trunk and caudal region and they remain at the same posi- tion during larval stage. The larva of 14.9 mm NL (Fig. IF) has a heavily pigmented body, but the three large pigment bars on the trunk and caudal region are still visible. There are dense melanophores over the occipital region of the head and shoulder (Fig. 1A). Pigments on the elongate ventral fin ray is also visible in the smallest specimen, but the positions and number of them change gradually. In the earliest stage (3.8 mm NL) there are two melano- phores on the ventral fin: one at the fin ray base and another at the middle of the ventral fin. At the size of 4.5 mm NL (Fig. IB), there appears another small melanophore at one-third the length of the fin ray. The melanophore at the fin ray base remains at the same position, but the distal large one moves to the position three-fourths the length of the fin ray. After this size, positions and number of melan- ophores on the elongate third ventral fin ray remain the same up to 15.7 mm NL. When distal part of other ventral fin rays start to separate from the third one, there appears some melanophores on the distal edge of each fin ray. There appears a patch of melanophores on the preopercular region at 3.8 mm NL and another small one appears on the suborbital region at 4.5 mm NL. The small melanophore, which appears on the tip of the elongate second dorsal spine at 4.5 mm NL, will later become a large pigment bar (Fig. 1C, D). FIN DEVELOPMENT The most remarkable change can be seen in lengths of the dorsal and ventral fins. Since the earliest stage (Fig. 1A), the larvae have elongate dorsal spine and ventral fin ray, which later become the second dorsal spine and the third ventral fin ray, respectively. The length of the second dorsal spine relative to body length changed from 28% at 3.3 mm NL to 90% at 8.3 mm NL (Fig. 2A). In larger lar- vae the proportion of the second dorsal spine length relative to body length decreased gradually to 70% at 15.7 mm NL. A similar tendency was observed for the length of the third ventral fin ray: it varied from 45% of body length at 3.3 mm NL to 121% at 11.6 mm NL (Fig. 2B). Unfortunately, these fin rays are in many cases lost or damaged at the distal tip, making it difficult to say whether we measured the total length of fin rays or the partial length of a damaged ray. In any case, the figure shows a clear tendency of rapid increase of fin rays during larval stage. The number of fin rays increases during larval stage. For example, the origin of the first dorsal spine firstly appears anterior to the elongate sec- ond dorsal spine in 9.2 mm NL larva (Fig. IE). The tip of the first dorsal spine which will transform in- to the illicium in the adult fish, emerges from the epidermal skin at about 10 mm NL. At this size, all fin rays are well developed and number of fin rays on the second dorsal, anal, and caudal fins attains the adult number. Another remarkable change in fin development is a forward advancement of the dorsal spines. At 3.3 mm NL larva, the elongate second dorsal spine lies behind the head (Fig. 1A) and it moves gradu- ally forward during larval stage; at 14.9 mm NL, 433 FISHERY BULLETIN: VOL. 84, NO. 2 % 80 60 40 20 0 120 100 80 60 40 20 ••• • • * ••< •: •* • •• . .• * • . •••• i i i t ' i ' ' ' L IT? ' oMv ••if*. • i i i i i j i i_ _i i i_ 5 7 9 II BODY LENGTH 13 15 mm Figure 2.— Relationships between changes of proportion of second dorsal spine (A) and third ventral fin ray (B) and body length (NL) of Lophius gastrophysus. it becomes the position anterior to the eyes (Fig. IF). DISCUSSION Based on a study of world-wide collections, Caruso (1981, 1983) recently concluded that the Lophiidae is represented by 4 genera and 25 species, of which only 2 species inhabit the western Atlantic: Lophius americanus in the western North Atlantic and L. gastrophysus in the western Central and South Atlantic. The geographic ranges of the two species overlap between Cape Hatteras, NC, and Florida. The two western Atlantic species are very similar, but they can be easily distinguished by differences in dorsal and anal fin ray counts, size of the third and fourth dorsal spines, and differences in pigment pattern (Caruso 1983). It is well known that lophiid anglerfishes spawn over deep water producing large gelatinous ribbons of spawn which often contain more than a million eggs (Berrill 1929). Spawning behavior is not known, but some authors have suggested that it may occur at or near the bottom (Taning 1923; Dahlgren 1928). After hatching, the larvae emerge from the gela- tinous capsules and pass a long planktonic stage. Upon attaining a length of about 60 mm TL, young fish probably take to the bottom (Connolly 1922; Taning 1923; Bigelow and Schroeder 1953). As shown previously, Lophius larvae can be easily distinguished from those of other species. Because there is only one species in the western South Atlan- tic, there is no doubt about the identification of our larvae as L. gastrophysus. Therefore, we have documented morphological differences in early developmental stages of our specimens and com- pared them with those of other well-known species (Table 2). Meristic characters and adult forms of L. ameri- canus and L. piscatorius are very similar, but their larval forms are quite different (Taning 1923). The most remarkable difference is the presence of three large pigment bars on the trunk and caudal region in L. americanus from the yolk-sac stage. He also pointed out that the larval development of L. ameri- canus was more rapid than that of L. piscatorius. The larvae of L. gastrophysus are very similar to that of L. americanus. Both species have three large pigment bars on the trunk and caudal region from the very earliest stages. Larval development of L. gastrophysus is more rapid than that of L. ameri- canus, e.g., formation of the bases of the second dor- sal and anal fins and the five dorsal spines occurs at sizes 8.1 mm, 8.5 mm, and 11.5 mm, respective- ly, for L. gastrophysus, L. americanus, and L. piscatorius. In the same way, the first appearance of canine teeth on both jaws occurs at sizes of 4.2 mm, 6.5 mm, and 9.8 mm, respectively, in the same order for the three species. Another difference is in the position of the mela- nophore of the ventral fin, present on the distal part of this fin in larvae of L. americanus and L. pisca- torius, but at three-fourths the length of the fin in L. gastrophysus larvae. The presence of pigmenta- tion in the preopercular and suborbital regions is also peculiar to L. gastrophysus larvae. ACKNOWLEDGMENTS The authors wish to thank Edward D. Houde of the University of Maryland for revision and critical reading of the manuscript. They are also grateful to June Ferraz Dias and Kazuko Suzuki for sorting and drawing the larvae. The financial support of this work came from the 434 MATSUURA and YONEDA: EARLY DEVELOPMENT OF LOPHIID ANGLERFISH Table 2.— Comparison of development stages of three Atlantic species of lophiid anglerfishes. Characters L gastrophysus L. americanus L. piscatorius General development Formation of bases of second dorsal and anal fins, the five dorsal spines, and the elongate ventral fin First dorsal spine Completion of anal fin rays Completion of soft dorsal fin rays Size at first appearance of canine teeth in both jaws Size of newly hatched larva Pigment on distal edge of the second dorsal spine Position of pigment on distal part of the ventral fin Pigment bars on the trunk and caudal region Meristic characters7 Dorsal fin rays Anal fin rays Pectoral fin rays Vertebrae 8.1 about 10-11 mm 9.3 mm 9.3 mm 4.2 mm about 3.5 mm since 5.2 mm 3/4 of ventral fin three bars since early stage 9-11 8-9 22-26 (24.6) 26-27 (26.2) 8.5 mm' about 12-14 mm1'2 10.5 mm2 10.5 mm2 6.5 mm2 about 2.5 mm4 no pigment ' 2 far distal edge2, 6 three bars since2 early stage 11-12 9-10 25-28 (26.1) 28-30 (29.1) 11.5 mm2 about 15-16 mm2 3 16 mm3 16 mm3 9.8 mm2 about 4.5 mm3 since 6 mm3 5 far distal edge5 anterior two bars5 since 11 mm 11-12 9-10 23-27 (25.2) 30-31 (30.4) 'Martin and Drewry 1978; 2Taning 1923; 3Russel 1976; "Fahay 1983; 5Lebour 1925; 6Agassiz 1882; 7Caruso 1983. Note: For comparative purpose, the body length was given in total length for all species. Notochord length of L gastrophysus larvae was converted to total length with an equation: TL = 1.024 mm NL + 0.1168 (r = 0.999), for larvae smaller than 10.0 mm NL. Financiadora de Estudos e Projetos (FINEP). The senior author received the research fellowship of the Conselho Nacional de Desenvolvimento Cientifico e Tecnologico (CNPq) and the junior author received the scholarship of the Fundacao de Amparo a Pes- quisa do Estado de Sao Paulo (FAPESP). This is contribution n? 620 of the Institute Oceanografico da Universidade de Sao Paulo. LITERATURE CITED Agassiz, A. 1882. On the young stages of some osseous fishes. Part III. Proc. Am. Acad. Arts Sci. 17:271-303. Allen, E. J. 1917. Post-larval teleosteans collected near Plymouth during the summer of 1914. J. Mar. Biol. Assoc, U.K. 11:207-250. Arbault, S., and N. L. Boutin. 1968. Ichthyoplancton. Oeufs et larves de poissons t6l6o- steens dans le Golfe de Gascogne en 1964. Rev. Trav. Inst. Peches Marit. 32:413-476. Berrill, N. J. 1929. The validity of Lophius americanus Val. as a species distinct from L. piscatorius Linn., with notes on the rate of development. Contrib. Can. Biol. (N.S.) 4:145-151. Bigelow, H. B., and W. C. Schroeder. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53, 577 p. Bowman, A. 1920. The eggs and larvae of the angler (Lophius piscatorius) in Scottish waters. A review of our present knowledge of the life history of the angler. Sci. Invest. Fish. Board Scotl. 1919, No. 1, p. 1-42. Caruso, J. H. 1981. The systematics and distribution of the lophiid angler- fishes: I. A revision of the genus Lophiodes with the descrip- tion of two new species. Copeia 1981:522-549. 1983. The systematics and distribution of the lophiid angler- fishes: II. Revisions of the Genera Lophiomus and Lophius. Copeia 1983:11-30. Connolly, C. J. 1920. Histories of new food fishes. III. The angler. Bull. Biol. Board Can., Ottawa, No. 3, 17 p. 1922. On the development of the angler (Lophius piscatorius L.). Contrib. Can. Biol. 1921(7):113-124. Dahlgren, U. 1928. The habits and life history of Lophius, the angler fish. Nat. Hist. 28:18-32. Fahay, M. P. 1983. Guide to the early stages of marine fishes occurring in the western North Atlantic Ocean, Cape Hatteras to the southern Scotian shelf. J. Northwest Atl. Fish. Sci. 4:1-423. Lebour, M. V. 1919. Feeding habits of some young fish. J. Mar. Biol. Assoc, U.K. 12:9-21. 1925. Young anglers in captivity and some of their enemies. A study in a plunger jar. J. Mar. Biol. Assoc, U.K. 13: 721-734. Martin, F. D., and G. E. Drewry. 1978. Development of fishes of the Mid-Atlantic Bight: an atlas of eggs, larval and juvenile stages. Vol VI, Strom- ateidae through Ogcocephalidae. U.S. Fish Wildl. Serv. Off. Biol. Serv., 78/12, 416 p. Matsuura, Y. 1979. Distribution and abundance of eggs and larvae of the Brazilian sardine, Sardinella brasiliensis, during 1974-75 and 1975-76 seasons. Bull. Jpn. Soc Fish. Oceanogr. 34: 1-12. Mito, S. 1966. Fish eggs and larvae. [In Jpn.] In S. Motoda (editor), Illustrations of the marine plankton of Japan, Vol. 7, 74 p. Soyo-Sha, Tokyo. 435 Padoa, E. 1956. Triglidae, Peristediidae, Dactylopteridae, Gobiidae, Echneidae, Jugulares, Gobiesocidae, Heterosomata, Pedicu- lati. [In Ital.] In Uova, larve, e studi giovanili di Teleostei, p. 627-880. Fauna Flora Golfo Napoli 38. PlETSCH, T. W. 1984. Lophiiformes: development and relationships. In H. G. Moser et al. (editors), Ontogeny and systematics of fishes, p. 320-322. Am. Soc. Ichthyol. Herpetol., Spec. Publ. No. 1. Prince, E. E. 1891. Notes on the development of the angler-fish (Lophius piscatorius). 9th Ann. Rep. Fish. Board Scotl., (1890), p. 343-348. Procter, W., H. C. Tracy, E. Helwig, C. H. Blake, J. E. Morrison, and S. Cohen. 1928. Fishes— a contribution to the life history of the angler (Lophius piscatorius). In Biological survey of the Mount Desert region, Part 2, p. 1-29. Philadelphia. Russel, F. S. 1976. The eggs and planktonic stages of British marine fishes. Acad. Press, Lond., 524 p. Stiasny, G. 1911. Uber einige postlarvale Entwicklungsstadien von Lo- phius piscatorius L. Arb. Zool. Inst. Univ. Wien 19:57-74. Tanaka, S. 1915-1919. Figures and descriptions of the fishes of Japan. J. Coll. Sci., Imp. Univ. Tokyo 24:419-440. Taning, A. V. 1923. Lophius. Rep. Dan. Oceanogr. Exped. 1908-1910 Mediterr. Adjacent Seas 2, Biol. A. 10, 30 p. Williamson, H. C. 1911. Notes on the eggs of the angler (Lophius piscatorius), halibut (Hippoglossus vulgaris), Conger vulgaris and tusk (Brosmius brosme), a young Arnoglossus sp.; abnormalities in Lophius, Gadus, Raia; diseases in Gadus, Pleuronectes, Onos, Zoarces; occurrence of Himantolophus rheinhardti, and Clupea pilchardus; the effectiveness of a seine-trawl in a small pond. 28th Ann. Rep. Fish. Board Scotl, (1909), Part III, p. 46-66. 436 EX-VESSEL PRICE LINKAGES IN THE NEW ENGLAND FISHING INDUSTRY Dale Squires1 ABSTRACT This study examines the direction of ex- vessel price linkages between the three New England ports of Boston, New Bedford, and Gloucester. Within-sample, bivariate tests of Granger causality are applied for monthly data from 1965 through 1981. It is found that cod and haddock prices are formed in New Bedford, that pollock prices are simultaneously formed between Boston and Gloucester, and that a spurious relationship exists for flounder prices between the three ports. The hypothesis is advanced that this spurious relationship may be due to flounder price leadership from outside the region, most probably the New York Fulton Fish Market. The direction of price linkages between various market and production centers in an industry is im- portant to studies of marketing and prices. Although these spatial and hierarchical relationships are generally well understood in domestic agriculture, they have received little or no attention in natural resource utilization and in the domestic commercial fishing industry in particular. This study therefore examines the spatial characteristics of round ex- vessel price linkages of the most important species in the New England fishing industry from 1965 through 1981. Three ports— New Bedford, Boston, and Glouces- ter—dominate the New England fishing industry, as both home ports or production centers and as marketing centers. By both volume and value of landings, New Bedford is the most important port, followed by Gloucester and then Boston. The most important species of groundfish in New England are cod; haddock; yellowtail, winter, and other flounders; ocean perch or red fish; and pollock. Sea scallops and lobsters also provide a significant con- tribution to the industry in both value and volume of landings. This study accordingly focuses upon the ports of Boston, New Bedford, and Gloucester, and the species of cod, haddock, yellowtail and winter flounders, and pollock. Additional attention is given to ocean perch and sea scallops, though rigorous con- clusions are not possible. In New Bedford and Boston, fishermen sell their catches to the highest bidder in an open auction. The New Bedford auction begins at 8:00 a.m. and ends at 8:22 a.m. The Boston market begins at 7:00 a.m., and invariably overlaps with the New Bedford market. There is significant communication between the two markets during the auctions. The volume and total value of fish harvested is substantially greater in New Bedford than in Boston. Bidders pur- chase an entire vessel's landings in New Bedford, while in contrast, purchasers offer individual bids for each species in Boston. In most of the ports other than Point Judith in Rhode Island (where an impor- tant fishermen's cooperative exists), the catch is sold directly to fish processors or by prior arrangements between individual vessels and purchasers. Further, it is generally believed that Gloucester prices for most fresh groundfish species are set in Boston, and differ only by a transportation cost. Fishermen of all ports are free to land their harvests at any port offering the highest prices, which, however, must be balanced against steam- ing time. Few vessels land exclusively at a single port, since the distances between the three are not great. A definite limit exists to port switching due to the prevalence of market transactions costs. Wilson (1980) indicated that personal and financial relationships tend to bind particular fishermen and fish buyers. In contrast to many other natural resource and primary production industries, a futures market does not exist for fresh fish.2 Different ports and markets have developed singular reputations. These specializations are based in large part upon proximity to resource stocks. New Bedford has developed a reputation as a flounder and sea scallop port, while Boston has become known as a cod, haddock, and, to a lesser extent, 'Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. 2Trading on futures markets involves buying and selling stan- dardized contracts for the future delivery of a specific grade of a commodity at a specific location(s). Manuscript accepted August 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 437 KISHLKY BULLETIN; VUL. 84, NO. Z pollock port. Although Gloucester fishermen direct much of their effort towards cod, haddock, and flounders (generally joint products), Gloucester has developed a reputation as a port for both pollock and ocean perch. Conventional wisdom in the New England ground- fishery market holds that New England round (fish as harvested) ex- vessel prices of fresh flounders are formed in the New Bedford auction market, while fresh cod, haddock, and pollock round ex-vessel prices are set in the Boston auction. These widely held beliefs serve as the null hypotheses to be tested in this study of the ex-vessel groundfish price link- ages in New Bedford, Gloucester, and Boston. Knowledge of ex-vessel price linkages has a num- ber of applications. Efforts at improving market ef- ficiency would find this information useful. The broadcasting of daily ex- vessel fish prices by the Na- tional Marine Fisheries Service can properly focus upon the most crucial markets. Infrastructural or institutional improvements can be more judicious- ly targeted, an important consideration in a time of tight public and private budgets. Price forecasts to improve industry functioning can concentrate upon those prices formed in markets which demonstrate price leadership. Fishermen may want to land their harvests in the market in which ex-vessel prices are first formed, should fishermen want to affect the pricing process, be less dependent upon the land- ings of others, or capture advantageous prices. Similar considerations apply to buyers. Knowledge of the price formation process allows government price policies to target the appropriate markets. Finally, price linkage information is crucial to studies of marketing margins, length of price trans- mission, and asymmetric pricing. THE DATA The data are taken from the vessel weighout files of the National Marine Fisheries Service. After every trip of a commercial fishing vessel of any gear type, port agents in each port obtain the value and volume of landings for each species harvested. The entire collection of this information constitutes the weighout file. The output vector from the weighout file is then linearly aggregated over vessels and trips to form monthly round ex-vessel prices for each port. The resulting nominal prices are subsequent- ly deflated by the consumer price index for food. As Sims (1974) and Feige and Pierce (1980) noted, the use of seasonally adjusted data may confound lag distributions and causality relationships. Conse- quently, the data are left in their unseasonalized state. However, to account for seasonal differences, quarterly dummy variables are employed. The time domain of the data set extends from 1965 through 1981. METHOD OF ANALYSIS Granger (1977) provided a definition of causality among a set of variables that is based upon predic- tability as well as the fact that the effect of a change in an exogeneous variable upon an endogeneous variable requires time. A variable X causes another variable Y, with respect to a given universe or in- formation set that includes X and Y, if present Y can be better predicted by using past values of X than not doing so, all other information in the past of the universe being used in either case. Causality from Y and X is defined in the same manner. Feed- back occurs if X causes Y and Y causes X. A causal relationship between X and Y does not exist if causality does not run from X to Y or from Y to X, and feedback does not occur. Causality tests may be classified into two funda- mental types at their most basic level, within-sample and out-of-sample tests. The within-sample test is widely applied and is the first one developed. This test is developed over the full-time domain of the data set, and essentially relies upon a measure of fit. The definition of causality in the out-of-sample test requires evidence of improved forecasts. This approach is implemented by identifying and esti- mating different models using the first part of the sample and then comparing their respective fore- casting abilities on the latter part of the sample. This study utilizes the within-sample test, the one most commonly applied, since the properties of the out- of-sample test have yet to be systematically examined. Two basic approaches have been advanced by which to apply empirically the within-sample bivariate Granger criterion to time series. The first approach is represented by the test proposed by Pierce (1977) based upon Haugh (1976). The proce- dure first estimates whitening filters for each time series, then subsequently estimates the cross- correlation function for the first step's residuals.3 However, Sims (1977) and Geweke (1981) indicated that this approach may be limited.4 A second basic 3Whitening filters remove serial correlation from a time series. Each time series used in a test of causality will be a white noise process, and any relationships will be based on actual, systematic relationships between the two time series, instead of a spurious relationship caused by the common serial correlation. 4Prefiltering each time series with separate autoregressive inte- grated moving average (ARIMA) filters biases the test toward 438 SQUIRES: EX-VESSEL PRICE LINKAGES approach relying directly upon distributed lag rela- tionships between dependent and independent variables has led to three widely used tests: those suggested by Sims (1977), the direct Granger test forwarded by Sargent (1976), and the Modified Sims test advanced by Geweke et al. (1983). The small-sample properties of the Sims (1972), direct Granger, and Modified Sims tests have recently been examined within Monte-Carlo frame- works by Guilkey and Salemi (1982) and Geweke et al. (1983). Although the two studies differ somewhat in their specifications, both found that the Sims test was outperformed by the other two. Since the Sims test is more time-consuming and expensive to employ and requires more decisions about param- eterizations, both studies unequivocally recommend against its use. The two studies reach slightly different conclu- sions on the efficacy of the direct Granger and Modified Sims test. These contradictory results can be attributed to differences in research design. Geweke et al. (1983) concluded that the two tests essentially perform equally well. In contrast, Guilkey and Salemi (1982) determined that the direct Granger test consistently outperforms the Modified Sims procedures by small amounts. Since the direct Granger test is computationally the least expensive of the three and results in the fewest degrees of freedom lost from formation of leads and lags, Guilkey and Salemi recommend its use over the Modified Sims and Sims procedures. Nonetheless, they do note that the Granger procedure's advan- tage over the other two diminishes with increases in sample size. Several additional findings of Guilkey and Salemi (1982) are also worth reporting. They observed that for sample size <200, the shorter versions of all three tests are superior to the longer versions.5 They further noted that in their Monte-Carlo study the direct Granger and Modified Sims procedures ac- curately recover the coefficients of the relevant population projections of the statistical model used to generate experimental time series in small samples. Consequently, it may be unlikely to observe "large" coefficient estimates arising spuriously. Finally, test performance is extremely sensitive to sample size, strength of causation, and length of test parameterization employed. The direct Granger test as applied in this study is based upon ordinary least squares regression of the current observation of the time series of round ex-vessel prices from one port upon its own past observations and the past observations of the other port's round ex-vessel prices for species k: 4 J P2k(t) = a, + I bkl D% + cLT + I dkj P2k J x(t-j)+ Z fkj Plk (t - j) + ekt. (1) Here, LT refers to a linear time trend, Dt is the zero-one variable for quarter i, Plk(t) is the round ex- vessel price of species k in month t in port 1, J is the number of periods lagged, and ekt is a vector of stochastic, white noise residuals. The presence of lagged dependent variables in Equation (1) is counted on to remove serial correlation from the estimated residuals.6 The test of the null hypothesis that P\k does not cause P2k is a test that fkj = 0, j = 1,2,. . .,J. Guilkey and Salemi (1982) indicated that the F-test statistic is calculated by estimating Equation (1) in both constrained (fkj = 0, j = 1,2,. . . ,J) and un- constrained forms, and may be written as7 F = (SSEC - SSEJJ SSEJ(T - (2J + 2)) (2) where SSEU and SSEC are the residual sum of squares from the unconstrained and constrained regressions, respectively, and T represents the number of monthly observations on round ex- vessel prices. Under the null hypothesis, F is an F-test statistic with J and T - (2J + 2) degrees of free- dom. This procedure is then repeated reversing the roles of P\k and P2k to test the null hypothesis that P\k does not cause P2k. The direct Granger test requires selection of a lag length, J, large enough to purge serial correlation from estimated residuals. Several factors require consideration before specifying the lag length. Chilled fresh fish is a commodity that rapidly deteriorates in quality. Consequently, definite limits exist to the length of time which inventories of failing to reject the null hypothesis of independence of the two series more often than the specified level of significance suggests. Because of this limitation, the second basic approach is applied. 5Longer versions of these tests include additional lead and lag variables. "Serial correlation exists when the error terms from different observations in a time series are correlated. Serial correlation tends to give unbiased but inefficient estimators, and a biased sampling variance, which then affects the results from significant tests such as the F- or -t-tests. 7A constrained F-test includes one or more restrictions, such as one or more coefficients constrained to zero. An unconstrained F- test does not include these restrictions. 439 chilled fresh fish can be held. Since most ground- fish harvested in New England waters are not pro- cessed into frozen fish products, long-term storage of New England groundfish is unlikely, and fresh fish prices are likely to adjust more quickly than those of most other food commodities. In addition, previous exploratory analysis with adaptiver filter- ing methods on the weighout file suggests that two sets of round ex-vessel prices for any species k are particularly important, the previous month's price and the price within one month on either side of the previous year. In order to account for these charac- teristics and to provide both short and long versions of the test, lags of 8 and 14 mo were specified. These lag lengths are sufficiently long to encompass price lags with monthly data. The diagnostic Q test of Box and Pierce (1970) is used to detect serious serial correlation. EMPIRICAL RESULTS The empirical results from the direct Granger causality tests lead to somewhat unexpected con- clusions for most species. The null hypothesis that monthly round ex-vessel prices of cod and haddock in all three ports are first formed in the Boston auc- tion market is rejected in almost all instances. The findings in Table 1 instead suggest that the cod and haddock prices established in the New Bedford auc- tion lead the prices formed in the Boston market. Several factors may account for this. The New Bed- ford auction's volume of landings is substantially higher than that of Boston. In addition, the two market times ordinarily overlap, and frequent com- munication occurs between economic agents during the auctions. Further, the proximity of New Bed- ford to Boston allows fresh fish to be easily trucked to Boston from New Bedford. The markets are thus physically linked, before the auctions by fishermen and after the auctions by fish buyers. One element of conventional wisdom may perhaps be substan- tiated, however. Although the Q-test statistic in- dicates severe serial correlation (and thereby possibly refuting the F-test statistic), the empirical results indicate that Boston cod prices do lead Gloucester cod prices at the ex-vessel level for the shorter lag length parameterization. Rejection of the null hypothesis that Boston prices lead New Bedford and probably Gloucester cod and haddock round ex-vessel prices and the finding that New Bedford prices lead Boston prices suggest a second null hypothesis for consideration. This sec- ond hypothesis states that Gloucester cod and had- dock prices are directly led by New Bedford prices. In addition, the possibilities that Boston prices lead Gloucester prices and that New Bedford prices lead Boston prices suggest an additional, indirect price linkage between Gloucester and New Bedford via Boston. The results for this second null hypothesis are also given in Table 1. Since this is an unplanned com- parison, a Scheffe interval is used.8 Strictly followed, 8An unplanned comparison occurs when in the course of exam- ining results a hypothesis is tested which was not specified prior to the experiment. The initial region is altered by the additional information, so that the level of significance has changed. A Scheffe interval allows for a more cautious test by providing a larger critical value than that given by a t or F table. This pre-test bias is accounted for by a conservative test. The F-test statistic now Table 1 .—Direct Granger causality tests for monthly fresh round ex-vessel cod and haddock prices. Cod Haddock Direction1 Lags2 F-test3 Q-test4 Direction1 Lags2 F-test3 Q-test4 B B G G B B NB NB G G NB NB ->G ->G ->B ->B ->NB ->NB ->B ->B ->NB ->NB ->G ->G 8 14 8 14 8 14 8 14 8 14 8 14 2-37* 1.17 1.59 1.02 1.67 0.74 2.52* 1.89* 1.78 0.91 52.96 1.40 35.67* 2.61 10.98 12.35 13.52 21.85* 8.82 16.27 16.46 28.42 7.84 6.86 B B G G B B NB NB G G NB NB ->G ->G ->B ->B ->NB ->NB ->B ->B ->NB ->NB ->G ->G 8 14 8 14 8 14 8 14 8 14 8 14 1.69 1.09 1.07 1.29 0.08 0.09 3.15* 1.76* 0.56 1.36 52.90 1.65 9.60 4.49 10.98 15.07 12.54 7.06 8.62 8.74 11.71 15.09 11.76 8.37 'Variable abbreviations are B (Boston), G (Gloucester), NB (New Bedford). 2J indicates J months lagged. 3Null hypothesis that past values of the causal variable do not significantly affect current values of the dependent variable. An asterisk indicates rejection of the null hypothesis at the 5% level. 4Null hypothesis that regression residuals are white noise. An asterisk indicates rejection of the null hypothesis at the 5% level. 5F-test statistic is significant at the 5% level, but not significant at the 5% level when a Scheffe interval is used. 440 SQUIRES: EX-VESSEL PRICE LINKAGES the results indicate that the cod and haddock price linkage does not run from New Bedford to Glouces- ter. If a Scheffe interval is not used, then the New Bedford cod and haddock prices do lead those of Gloucester. Therefore, with this caveat, New Bed- ford auction market monthly round ex-vessel cod and haddock prices lead the prices of Gloucester and Boston, and Boston prices may lead those of Gloucester. In any case, it appears that the New Bedford auction market dominates the formation of round ex-vessel prices for cod and haddock. The empirical results for yellowtail and winter flounder of Table 2 also contradict the null hypoth- esis that monthly fresh round ex-vessel prices for both species are formed first in New Bedford. In- stead, the findings indicate that pricing feedback ex- ists between both New Bedford and Gloucester and between New Bedford and Boston. These conclu- sions must be tempered by the significant Q-test statistics for several relationships. These conclusions lead to a second null hypothesis between the prices of New Bedford and Gloucester, New Bedford and Boston, and possibly between Gloucester and Boston rests with a spurious rela- tionship. Although New Bedford is the most impor- tant flounder port by landings in New England, New York City is even more important on the eastern seaboard by volume of consumption. New York City's Fulton Fish Market is primarily a wholesale market without substantial landings. Much of the New England flounder harvested is sent to Fulton on consignment without an ex-vessel price being established in New England. The Fulton Fish Market also begins much earlier in the morning than New Bedford's auction market. Thus the apparent feedback among the ex-vessel yellowtail and winter flounder prices in the New England ports is prob- ably due to their following of the wholesale prices set in the Fulton Fish Market. Table 3 presents the results for pollock. As with the other species, consistent results are obtained for different lag lengths. Again, the null hypothesis dic- Table 2.— Direct Granger causality tests for monthly fresh round ex-vessel yellowtail and winter flounder prices. Yellowtail flounder Winter flounder Direction1 Lags2 F-test3 Q-test4 Direction1 Lags2 F-test3 Q-test4 NB NB G G NB NB B B G G B B ->G ->G ->NB ->NB ->B ->B ->NB ->NB ->B ->B ->G ->G 8 14 8 14 8 14 8 14 8 14 8 14 2.05* 2.27* 1.96* 2.48* 2.02* 1.76* 0.75 2.83 2.275 2.50s 1.48 1.27 20.19 19.42 23.56* 26.78* 10.56 9.89 17.63 21.83* 9.41 12.74 16.86 18.82 NB NB G G NB NB B B G G B B ->G ->G ->NB ->NB ->B ->B ->NB ->NB ->B ->B ->G ->G 8 14 8 14 8 14 8 14 8 14 8 14 7.52* 8.71* 2.88* 2.97* 11.87* 7.66* 23.57* 5.25* 3.875 4.375 4.15s 3.56s 5.68 1.78 16.98 22.19* 18.45 7.92 9.94 3.99 12.43 4.70 18.16 17.34 'Variable abbreviations are B (Boston), G (Gloucester), NB (New Bedford). 2J indicates J months lagged. 3Null hypothesis that past values of the causal variable do not significantly affect current values of the dependent variable. An asterisk indicates rejection of the null hypothesis at the 5% level. "Null hypothesis that regression residuals are white noise. An asterisk indicates rejection of the null hypothesis at the 5% level. 5F-test statistic is significant at the 5% level, but not significant at the 5% level when a Scheffe interval is used. to be tested on the yellowtail and winter flounder price linkages between Gloucester and Boston. Since this test is also an unplanned comparison, a Scheffe interval is required. Again, the strict test results in- dicate that neither port's prices lead the other, nor that feedback exists. The most probable explanation for the feedback becomes significant only if it exceeds in magnitude ((a-l]Fh)'k, where F is the b ■ 100% critical value for F (a-1, N-a) and N is the number of observations. See Snedecor and Cochran (1976, p. 271) for more details. Table 3.— Direct Granger causality tests for monthly fresh round ex-vessel pollock prices. Direction1 Lags2 F-test3 Q-test4 B >G 8 4.34* 12.54 B >G 14 4.99* 10.15 G >B 8 5.28* 18.27 G >B 14 4.77* 22.96* 'Variable abbreviations are B (Boston), G (Gloucester), NB (New Bedford). 2J indicates J months lagged. 3Null hypothesis is that past values of the causal variable do not significantly affect current values of the dependent variable. An asterisk indicates rejec- tion of the null hypothesis at the 5% level. "Null hypothesis that regression residuals are white noise. An asterisk in- dicates rejection of the null hypothesis at the 5% level. 441 FISHERY BULLETIN: VOL. 84, NO. 2 tated by widely held industrial perceptions is re- jected. The results indicate that feedback exists between the monthly fresh round ex-vessel prices of pollock in both Gloucester and Boston. Both ports dominate pollock landings and are close to one another. A complete time series of prices for sea scallops exists only for New Bedford. Since New Bedford greatly dominates this fishery by both volume and value of landings, it may be safely concluded that monthly round ex-vessel sea scallop prices are formed in New Bedford. Finally, Gloucester is the only one of these ports to possess a complete time series of prices and landings of ocean perch or red fish. Since Gloucester dominates this fishery, monthly fresh round ex-vessel ocean perch prices appear to be formed in this port, at least among these three. CONCLUDING COMMENTS The within-sample bivariate direct Granger causality tests of monthly round ex-vessel price linkages for the three most important New England ports (Boston, New Bedford, and Gloucester) and the most important groundfish species lead to unex- pected results. Conventional wisdom considers the round ex-vessel cod and haddock prices formed in the Boston auction market to lead the comparable prices of the other New England ports. However, the empirical results indicate that New Bedford's prices lead those of the other ports, although in cer- tain cases Boston's cod prices may lead those of Gloucester as well. The common industry perception also holds that the yellowtail and winter flounder round ex-vessel prices are first formed in New Bedford and lead those of Boston and Gloucester. Instead, the em- pirical findings suggest that feedback and simul- taneous price formation occur among all three ports for both species. Since flounder landings in Boston and Gloucester are negligible in comparison to those of New Bedford, a spurious relationship due to the leading wholesale prices formed in the even earlier and more flounder-important Fulton Fish Market of New York City is suggested. Feedback is likely for fresh round pollock ex-vessel price formation in Boston and Gloucester. Finally, it is suggested that the New Bedford auction market dominates fresh ex-vessel sea scallop price formation and that Gloucester dominates among these three ports for ocean perch. New Bedford thus generally dominates ex-vessel price formation among the major New England ports for the most important species harvested. ACKNOWLEDGMENTS Helpful comments from David Bessler, Joseph Mueller, Robert Reidman, Patricia Kurkul, and DanieJ Huppert and an anonymous reviewer are gratefully acknowledged. LITERATURE CITED Box, G. E. P., and D. Pierce. 1970. Distribution of residual autocorrelations in autoregres- sive-integrated moving average time series models. J. Am. Stat. Assoc. 65:1509-1526. Feige, E. L., and D. K. Pearce. 1980. The casual causal relationship between money and in- come: some caveats for time series analysis. SSRI Pap. No. 7809, 39 p. Social Syst. Res. Inst., Univ. Wisconsin, Madison. Geweke, J. 1981. The approximate slopes of econometric tests. Econo- metrica 49:1427-1442. Geweke, J., R. Meese, and W. Dent. 1983. Comparing alternative tests of causality in temporal systems: analytic results and experimental evidence. J. Econ. 21:161-194. Granger, C. W. J. 1977. Investigating causal relations by econometric models and cross-spectral methods. Econometrica 37:424-438. Guilkey, D. K., and M. K. Salemi. 1982. Small sample properties of three tests for granger- causal ordering in a bivariate stochastic system. Rev. Econ. Stat. 64:668-680. Haugh, L. D. 1976. Checking the independence of two covariance-station- ary time series: a univariate cross correlation approach. J. Am. Stat. Assoc. 71:378-385. Pierce, D. A. 1977. Relationships - and the lack of thereof - between economic time series, with special reference to money and interest rates. J. Am. Stat. Assoc. 72:11-21. Sargent, R. J. 1976. A classical macroeconomic model for the United States. J. Polit. Econ. 84:207-238. Sims, C. 1972. Money, income, and causality. Am. Econ. Rev. 62: 540-552. 1974. Distribution lags. In M. D. Intriligator and D. A. Ken- dricks (editors), Frontiers of quantitative economics, Vol. II, p. 289-336. Amsterdam: North-Holland Publ. Co. 1977. Comment to Pierce. J. Am. Stat. Soc. 72:23-24. Snedecor, G., and W. Cochran. 1976. Statistical methods. 6th ed. State Univ. Press, Ames, 593 p. Wilson, J. A. 1980. Adaption to uncertainty in small numbers exchange: the New England fresh fish market. Bell J. Econ. 11, p. 491-504. 442 COMMUNITY STUDIES IN SEAGRASS MEADOWS: A COMPARISON OF TWO METHODS FOR SAMPLING MACROINVERTEBRATES AND FISHES1 Kenneth M. Leber2 and Holly S. Greening3 ABSTRACT r The effectiveness of using an otter trawl for estimating macrofaunal species ranks and abundances in seagrass meadows is unknown. In this study, we compare the catch effectiveness of the commonly used 5 m otter trawl with that of a 0.9 m wide epibenthic crab scrape for fishes, decapod crustaceans, molluscs, and echinoderms, using data from both day and night collections from a northeast Gulf of Mexico sea- grass meadow. The crab scrape collected significantly more individuals and species of all taxa except (water- column) fishes. Clear discrepancies existed between trawl and scrape estimates of species ranks and relative abundances, with trawl collections estimating a higher degree of dominance within groups of shrimps and demersal fishes, and lower dominance among crabs. Whereas the crab scrape was clearly superior to the trawl for sampling macroinvertebrates and demersal fishes, the trawl was the better device for collecting water-column fishes. Explanations for observed differences in the sampling effectiveness of these gears are discussed. Sampling was considerably more productive at night than during the day. The combined approach of day-night sampling with both a crab scrape (for demersal fishes and epibenthic invertebrates) and an otter trawl (for water-column fishes) is recommended for community-wide studies in seagrass meadows. Hypotheses concerning ecological community dynamics should be based upon accurate descriptions of the habitats and species involved. It is thus essen- tial that collection methods maximize sampling ef- ficiency in "community" (sensu Pielou 1977) studies. Because estimates of species composition, relative abundances, and biomass in aquatic environments may vary with different sampling devices (eg., Lewis and Stoner 1981; Stoner et al. 1983), knowledge of sample gear effectiveness allows a more rigorous ap- proach to sampling design and interpretation of results from studies of aquatic communities. Seagrass community studies often employ a small, semiballoon otter trawl (try net) for sampling fishes and epibenthic invertebrates (Kikuchi 1966; Living- ston 1975, 1976, 1982; Heck 1976, 1977, 1979; Hooks et al. 1976; Heck and Wetstone 1977; Weinstein and Heck 1979; Heck and Orth 1980; Orth and Heck 1980; Ryan 1981; Dugan and Livingston 1982; Dugan 1983). Although a small otter trawl may be one of the most effective samplers for estimating relative abundances of juvenile and small pelagic Contribution No. 439 of the Harbor Branch Foundation, Ft. Pierce, FL 33450. department of Biological Science, Florida State University, Tallahassee, FL 32306; present address: The Oceanic Institute, Makapuu Point, Waimanalo, HI 96795. department of Biological Science, Florida State University, Tallahassee, FL 32306; present address: Martin Marietta Environ- mental Systems, 9200 Rumsey Road, Columbia, MD 21045. Manuscript accepted January 1985. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. fishes in shallow nonvegetated waters (Kjelson and Johnson 1978; Orth and Heck 1980), there are few published accounts of its effectiveness in sampling benthic fishes or epibenthic invertebrates in vege- tated habitats. Greening and Livingston (1982) noted that a Chesapeake Bay crab scrape appeared to col- lect more invertebrate species per sample effort in vegetated habitats than did an otter trawl. Miller et al. (1980) found a crab scrape to be more effective than either an otter trawl or a push net for collect- ing juvenile blue crabs, Callinectes sapidus, in the Chesapeake Bay area. Blue crab fishermen routine- ly use crab scrapes, rather than trawls, in grassbeds in Chesapeake Bay (Warner 1976). In this study, the catch effectiveness of a 5 m otter trawl is compared with that of a 0.9 m epibenthic scrape in the shallow grassbeds of Apalachee Bay, FL. Species richness and abundance are examined within four taxonomic groups (decapod crustaceans, molluscs, echinoderms, and fishes). Because many grassbed organisms are more susceptible at night to certain sampling methods (Ryan 1981; Greening and Livingston 1982), both day and night samples are considered. METHODS Day and night samples were taken in about 1.7 m of water from seagrass beds in Apalachee Bay, FL. 443 FISHERY BULLETIN: VOL. 84, NO. 2 The sample site was located 5 km southwest of the E confirm River mouth (permanent station E-12 (Livingston 1975)). This site is characterized by relatively uniform, dense stands of the seagrasses, Thalassia testudinum and Syringodium filiforme, with seasonal occurrence of red drift algae (mean annual macrophyte biomass = 320 g dry wt/m2; see Zimmerman and Livingston 1979 for a description of macrophytes). Station E-12 was polyhaline, with salinities during collections ranging from 22 to 30 ppt (x = 27.0 ppt). Water temperatures ranged from 12.0° to 31.0°C (x = 19.9°). Depth varied from 1.6 to 2.1 m. Physical characteristics are summarized in Table 1. Table 1.— Physical characteristics of the sampling station for collection dates, Apalachee Bay, FL. Temp. Salinity Depth (°C) (PPt) (m) Jan. 1979 Day 12 31 2.0 Night 10 30 1.8 Apr. 1979 Day 22 23 2.1 Night 21 22 1.6 July 1979 Day 31 25 1.7 Night 30 25 2.1 Oct. 1979 Day 17 30 2.1 Night 16 30 1.7 A 90 cm wide commercial Chesapeake Bay crab scrape (Miller et al. 1980) was fitted with the cod end of a 5 m otter trawl (6 mm mesh liner). The crab scrape was towed at about 1.4 knots for 1 min (after Greening and Livingston 1982; Leber 1983), yielding a standardized tow of 42 m (mean of 10 preliminary measured 1-min tows). A 42 m weighted line was then used to standardize scrape tows during collec- tions. A 5 m otter trawl (19 mm mesh wings, 6 mm mesh liner in the cod end) was towed at the same speed for 2 min (as in Livingston 1975, 1982; Hooks et al. 1976; Heck 1977, 1979; Orth and Heck 1980; Stoner 1980; Stoner and Livingston 1980; Dugan and Livingston 1982; Dugan 1983), covering an average measured distance of 84 m. Under tow, the trawl mouth tickler chain fished a 2.1 m wide path over the substratum (Leber, pers. obs.). Hence, each individual trawl tow fished over 4.6 times the sub- stratum surface area sampled by each tow of the crab scrape (176 m2 vs. 38 m2). Because the scrape collected larger amounts of dead vegetation, it was logistically difficult to sample as much surface area with it as was sampled by the trawl. Collections were made quarterly (January, April, July, and October). On each sampling date eight scrape and four trawl tows were taken (in the se- quence two trawls, eight scrapes, two trawls) dur- ing the day, and again beginning 1 h after dark. Greening and Livingston (1982) determined that eight 1-min scrapes were sufficient for sampling >95% of the species of macroinvertebrates at our sample site in Apalachee Bay. Because each scrape was towed for only half the 2-min towing time used for each trawl (scrape tows lasting longer than 1 min often resulted in clogging the net with red drift algae), only four trawls were taken during each sam- pling period. Thus, the combined length of the eight scrape tows (8 x 42 m = 336 m) matched that of the four trawl tows. All samples were collected from a 0.25 km2 area immediately south of the station marker. Replicate tows were taken along transects spaced at least 30 m apart to prevent overlapping samples. Organisms were preserved in 10% Formalin4 (buf- fered with seawater) in the field, then identified, counted, and measured in the laboratory. A two-way, Model II, factorial ANOVA design for unequal but proportional cell sizes (Sokal and Rohlf 1969) was used to compare mean numbers of species and in- dividuals of each taxon group in scrape vs. trawl (Factor 1) and day vs. night (Factor 2) samples. Log10 transformations were used where F-max tests indicated heterogeneity of variance Rather than extrapolating our data to numbers per unit area, we compared the collections made with these two gears using absolute numbers per tow in our calculations (which are biased in favor of the trawl by a factor of 4.6). We used these absolute abun- dances because 1) we wanted a strongly conservative test of our premise that the scrape is the more ef- fective of these two sample gears in vegetated aquatic habitats, and 2) we believe that extrapola- tions of semiquantitative data to abundances per unit area yield highly unrealistic results, which may be misinterpreted by readers as accurate densities (cf. Howard 1984, who determined that a towed beam trawl was only 4.7% efficient in estimating densities of shrimp in an Australian seagrass meadow). RESULTS Factor 1: Trawl vs. Scrape Although the surface area sampled by the otter "Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 444 LEBER AND GREENING: COMMUNITY STUDIES IN SEAGRASS MEADOWS trawl during each tow exceeded that sampled with the crab scrape by a factor of 4.6, mean numbers of individuals collected in scrape samples were signifi- cantly greater than those in trawl samples in 44% of the 16 scrape-trawl comparisons (Table 2). The trawl was a significantly more effective collecting device for number of individuals of fishes (Table 2; April, July, and October fishes), but interaction terms were significant for April and October analyses (see Interactions, below). Mean numbers of individuals were greater in trawl, than in scrape, samples in two other cases (Fig. 1, January and July decapods in night samples); however, scrape-trawl differences on those dates were nonsignificant (Table 2). The crab scrape was clearly the better gear for sampling epibenthic individuals. Species numbers were never significantly greater in trawl, than in scrape, samples (Fig. 1). In contrast, the crab scrape collected significantly more species than the trawl in 75% of the scrape-trawl com- parisons (Table 2). Because the scrape often sampled greater numbers of individuals than the trawl, the presence of more species in scrape, than in trawl, samples may be simply a sampling phenomenon. By chance alone, one would expect to encounter more rare species in larger samples. Using rarefaction analysis (Simberloff 1978), we have factored out the influence of sample size on species number for a better comparison of scrape vs. trawl sampling ef- fectiveness (Fig. 2). Eight of the 12 cases in which the scrape sampled significantly more species than the trawl (Table 2) can be attributed to a sampling phenomenon; there were generally more species in scrape samples because so many more individuals were collected in each scrape tow. However, it is clear in Figure 2 that the greater numbers of decapod species in January and July scrape samples, and fish species in April and October scrapes, represent real differences in the catch effectiveness of these gears for species within these two taxa. Factor 2: Day vs. Night Day-night differences were clear. None of the com- bined (scrape-trawl) daytime collections contained significantly more species or individuals than night collections. But nocturnal samples contained signifi- cantly more individuals than daytime samples in 69%, and more species in 62%, of the 16 day-night comparisons (Table 2). Interactions Significance of an interaction term indicates dependence of one factor upon the other; in this case, when sampling differences between scrape and trawl exist but are dependent upon time of day. Scrape- trawl vs. day-night interactions were significant in 8 of the 32 ANOVAs in Table 2. For these eight cases, either the trawl sampled better only at night for a certain taxon/month combination (one of the eight interactions), or the scrape sampled better only dur- ing the day (five of the eight cases), or both of these events occurred (two of the eight cases, scrape was better during the day but the trawl was better at night). Although fish were taken in greater abundances by the trawl on three of the four sampling dates, interactions were significant on two of those dates (April and October, Table 2). With the exception of July collections, fish were equally as abundant in day- time scrape samples as in trawls (see Figure 1). Table 2.— Two-way ANOVA, F-values. Underlined values indicate trawl samples significantly larger, all other significant values are scrape samples. All significant day-night values indicate night significantly larger than day samples. Decapods Molluscs Echinoderms Fishes No. No. No. No. No. No. No. No. Date Sample indiv. species indiv. species indiv. species indiv. species Jan. Day Night 0.48 35.81*** 0.02 0.31 0.73 1.22 57.98*** 42.14*** 1979 Scrape Trawl 0.02 27.77*** 56.71*** 58.48*** 1.92 4.29 0.44 0.08 Interaction 0.00 5.07* 0.01 0.15 1.73 0.03 0.35 0.33 Apr. Day Night 37.72*** 31.16*** 63.17*** 21.41*** 0.00 1.01 103.02*** 29.93*** 1979 Scrape Trawl 106.26*** 68.13*** 206.89*** 55.30*** 111.27*** 29.71*** 61.55*** 27.47*** Interaction 5.24* 0.62 22.51*** 2.21 0.51 0.50 68.10*** 0.03 July Day Night 97.55* 139.64*** 16.93*** 24.75*** 6.64* 2.79 14.06** 4.00 1979 Scrape Trawl 4.16 66.94*** 70.39*** 30.56*** 3.06 1.72 6.93* 0.35 Interaction 55.35*** 3.67 4.32 0.29 2.57 2.48 0.03 0.09 Oct. Day Night 7.29* 45.12*** 8.03* 20.32*** 1.87 3.27 20.36*** 5.04* 1979 Scrape Trawl 0.42 32.14*** 34.46*** 99.21*** 10.12** 7.91* 5.04* 9.62** Interaction 10.63** 0.02 5.01* 1.43 8.28** 3.20 23.85*** 0.13 = P < 0.05. = P< 0.01. = P< 0.001. 445 FISHERY BULLETIN: VOL. 84, NO. 2 co — o LlI Q. CO 16 8 Day 161 8 Night J A J 0 DECAPODS Scrape Trawl o UJ 8s 12 6 Night \r— "—I -I J A J 0 MOLLUSCS co _i < Q > CO UJ o UJ a. co J A J 0 ECHINODE RMS 60" CO 30 Day | .01 — 30 Night CO UJ o UJ a. CO 8- Day 4- T ^-—^** ""■*- Night A J FISHES 0 Figure 1— Mean numbers of individuals and species (±1 SD) collected by the crab scrape (solid line) and trawl (dashed line) during day and night sampling in January, April, July, and October 1979. 446 LEBER AND GREENING: COMMUNITY STUDIES IN SEAGRASS MEADOWS 150 450 750 Decapods 10 Molluscs Echinoderms NUMBER OF INDIVIDUALS Fishes Figure 2— Rarefaction curves for crab scrape (closed circles) and trawl data (open circles) from 1979 night samples. Expected numbers of species ( ± 2 SD) are plotted against numbers of individuals. Length of curves indicates maximum number of individuals taken in any single tow. Hence, with only one exception (July fish abundance), the otter trawl never outperformed the scrape dur- ing daylight collections. The trawl was more effective in sampling a tax- onomic group other than fish in only one case Sig- nificantly more decapod individuals were taken in July trawl samples at night, reflecting high densities of two caridean shrimps, Tozeuma carolinense and Periclimenes longicaudatus, which appear to be more susceptable to night trawl, rather than scrape, sampling. However, decapod abundances were notably higher in July daytime collections made with the crab scrape (see Figure 1), thus the highly sig- nificant interaction term for the July analysis (decapod individuals, Table 2). Relative Abundance Numerical rankings of the most abundant or- ganisms in each taxonomic group (combined over all sample dates) taken in night scrape samples are com- pared with those from night trawl samples in Table 3. Clear discrepancies exist between scrape and trawl estimates of species ranks and relative abundances. Relative to scrape samples, trawl collections over- estimated the degree of dominance (DI = combined proportions of the two most abundant species, {nx + n2)/N, McNaughton 1967) contributed by the most abundant shrimp Tozeuma carolinense and demersal fish Gobiosoma robustum, and under- estimated dominance of the most important crab Pagurus maclaughlinae and mollusc Argopectin ir- radians in our samples (Table 3). Relative to trawl collections, the scrape underestimated dominance for the most abundant water-column fishes, Lagodon rhomboides and Bairdiella chrysura. Species ranks of subdominants in trawl samples also differed from rankings based on data from scrape samples. DISCUSSION Scrape-trawl and day-night differences in sampling effectiveness were conspicuous and generally con- stant throughout the year. Although more (by a fac- 447 FISHERY BULLETIN: VOL. 84, NO. 2 Table 3.— Species ranks, relative abundances, and dominance for each taxonomic group. Combined night samples, x = mean number of individuals per sample (per group), Dl = dominance (McNaughton 1967). Scrape Trawl Scrape Trawl Relative Relative Relative Relative Rank abundance Rank abundance Rank abundance Rank abundance Shrimp Molluscs 1 0.324 Tozeuma carolinense 1 0.667 1 0.413 Argopectin irradiens 1 0.383 2 0.157 Penaeus duorarum 4 0.027 2 0.145 Modulus modulus 4 0.118 3 0.143 Periclimenes longicaudatus 2 0.191 3 0.130 Cerithium muscarum 6 0.077 4 0.127 Hippolyte zostericola 3 0.066 4 0.096 Anachis avara 2 0.169 5 0.099 Thor dobkini 6 0.016 5 0.086 Columbella rusticoides 3 0.131 6 0.049 Latreutes fucorum 5 0.018 6 0.064 Turbo castanea 5 0.101 7 0.049 Ambidexter symmetricus 8 0.003 7 0.025 Urosalpinx perrugata 7 0.009 8 0.038 Alpheus normanni 10 0.0002 8 0.013 Nassahus vibex 8 0.006 9 0.009 Palaemon floridanus 7 0.010 9 0.008 Hyalina veliei — 0 10 0.006 Periclimenes americanus 9 0.001 10 0.007 Fasciolaria hunteri — 0 X = 219.98 X = 423.38 X = 48.92 X = 13.32 Dl = 0.481 Crabs Dl = 0.858 Dl = 0.558 Demersal Fishes Dl = 0.501 1 0.735 Pagurus maclaughlinae 1 0.578 1 0.360 Gobiosoma robustum 1 0.544 2 0.117 Neopanope packardii 3 0.101 2 0.291 Opsanus beta 4 0.097 3 0.039 Epialtus dilatatus 4 0.055 3 0.246 Paraclinus fasciatus 2 0.194 4 0.032 Libinia dubia 5 0.048 4 0.086 Centropristis melana 3 0.106 5 0.027 Podochela riisei 6 0.041 5 0.017 Ophidion beani 5 0.058 6 0.026 Metaporaphis calcerata 2 0.133 X = 7.2 X = 2.6 7 0.016 Neopanope texana 9.5 0.007 Dl = 0.651 Dl = 0.738 8 0.004 Pitho anisodon 9.5 0.007 9 0.003 Pilumnus sayi 7 0.018 Water-Column Fishes 10 0.002 Pilumnus dasypodus 8 0.011 1 0.345 Lagodon rhomboides 1 0.621 X = 75.1 X = 10.9 2 0.158 Monacanthus ciliatus 4 0.044 Dl = 0.852 Dl = 0.711 3 0.154 Syngnathus floridae 5 0.042 4 0.151 Orthopristis chrysoptera 3 0.099 Echinoderms 5 0.067 Hippocampus zosterae 7 0.007 1 0.659 Echinaster sp. 1 0.824 6.5 0.052 Micrognathus crinigerus 8.5 0.002 2 0.255 Ophiothrix angulata 2 0.176 6.5 0.052 Haemulon plumieri 6 0.013 3 0.056 Lytechinus variegatus — 0 8 0.015 Bairdiella chrysura 2 0.168 4 0.027 Ophioderma brevispinum — 0 9 0.004 Monacanthus hispidus 8.5 0.002 X = 8.42 X = 5.12 X = 11.5 X = 31.7 Dl = 0.914 Dl = 1.00 Dl = 0.503 Dl = 0.789 tor of 4.6) substratum surface area was sampled per tow by the otter trawl, the crab scrape collected more species and individuals per tow, across taxa, with few exceptions. The trawl was the better faunal collect- ing gear in this seagrass habitat only for numbers of individuals of certain water-column fishes and for two species of caridean shrimps. The scrape was notably more effective than the trawl (day and night) for collecting penaeid, alpheid, and processid shrimps, brachyuran and pagurid crabs, molluscs, echinoderms, syngnathid fishes, and demersal fishes (Opsanus, Paraclinus, Gobiosoma, and Centropris- tis). The otter trawl appears to collect fewer species and individuals of demersal animals in grassbeds than does the scrape because the weighted (tickler) chain on the trawl is not in contact with the sub- stratum. Under tow, the cylindrical bottom crossbar of a crab scrape bends grassblades flat against the substratum, sweeping demersal and epifaunal organisms over the bar and into the net, whereas the otter trawl tickler chain is generally supported 8-10 cm above the substratum by the buoyant vege- tation (Leber, pers. obs.). Grassblades do not yield as much to the relatively light weight of a tickler chain (as they do to a scrape crossbar), and any organisms remaining close to the substratum as the chain passes over them evade capture Most epi- benthic inhabitants of grassbeds, including several fishes, are more closely associated with seagrasses and red drift algae than with the water column above the vegetation or bare patches within beds (Hooks et al. 1976; Heck and Wetstone 1977; Stoner 1980; Stoner and Livingston 1980; Gore et al. 1981). The crab scrape is more effective because it samples more grassblade surface area, including an addi- tional microhabitat, the region <10 cm above the substratum (Leber, pers. obs.). The greater effectiveness of both devices at night is probably accounted for, in part, by nocturnal in- creases in faunal activity on the substratum, on blade tips, and in the water column above vegetation. 448 LEBER AND GREENING: COMMUNITY STUDIES IN SEAGRASS MEADOWS Several crustaceans emerge from the substratum and forage at night in grassbeds, including pink shrimp, Pendens duorarum, some majid crabs (notably Pitho and adult Libinia at our site), and alpheid and processid shrimps (Fuss 1964; Fuss and Ogren 1966; Hughes 1968; Kikuchi and Peres 1977; Saloman 1979; Greening and Livingston 1982; Leber 1983). Emergence of nocturnal organisms from the substratum after dark would explain some of the variability between day and night collections of in- vertebrates. Higher densities of diurnally active animals in night samples may be due to nocturnal vertical migrations up grass-blades. Animals located near the tips of blades are clearly more vulnerable to capture by either device; even the scrape misses individuals trapped between grass-blades and substratum by the crossbar, an event less likely to occur to an individual near a blade tip. Fishes were probably less abundant in daytime trawl collections because of avoidance reactions to the clearly visible net. Emergence and vertical migration do not account for all of the increases in invertebrate abundance in night samples. The case of the arrow shrimp, Tozeuma carolinense, is interesting in this regard. We expected no day-night sampling differences for Tozeuma with either device, based on evidence that Tozeuma inhabit the region near tips of grass-blades, both during the day and at night (Main in press). As expected, Tozeuma were collected in roughly equal numbers in both day and night scrape samples. However, almost an order of magnitude more Tozeuma were taken in night trawl samples than dur- ing daytime collections (Leber and Greening, unpubl. data). It appears that Tozeuma may be capable of avoiding the trawl, which is highly visible during the day. These shrimp have keen vision in daylight and are capable of rapid movement (up to 30 cm) via a caridoid escape response (Main in press). They need only move down blades, closer to the substratum, to avoid the trawl net. This study suggests that many demersal fishes and epibenthic invertebrates may be more important numerically in seagrass communities than indicated by collections made with an otter trawl. Species ranks and relative abundances of these organisms determined from trawl collections in seagrass beds should be interpreted with care Whereas trawl col- lections may be satisfactory for monthly or year-to- year comparisons of single species abundances within a seagrass habitat, application of such data to examination of predatory-prey relationships (e.g, energy flow and optimal-diet models) or other biotic interactions in grassbeds may lead to erroneous interpretations. The combined approach of day-night sampling with both an otter trawl (for water-column fishes) and a crab scrape (for demersal organisms) is recommended for seagrass studies. ACKNOWLEDGMENTS We thank B. J. Freeman, J. Gerritsen, R. Howard, C. Koenig, F G. Lewis, K. Main, G. Morrison, and J. Ryan for comments and reviews of earlier drafts of this manuscript. R. J. Livingston provided technical support and M. Babineau provided help with the graphics. LITERATURE CITED DUGAN, P. J. 1983. Seasonal and geographic distribution of seven decapod crustaceans in Apalachee Bay, Florida. Contrib. Mar. Sci. 26:65-79. Dugan, P. J., and R. J. Livingston. 1982. Long-term variation of macroinvertebrate assemblages in Apalachee Bay, Florida. Estuarine Coast. Shelf Sci. 14: 391-403. Fuss, C. M., Jr. 1964. Observations on burrowing behavior of the pink shrimp, Penaeus duorarum Burkenroad. Bull. Mar. Sci. Gulf Caribb. 14(l):62-73. Fuss, C. M., and L. H. Ogren. 1966. Factors affecting activity and burrowing habits of the pink shrimp, Penaeus duorarum Burkenroad. Biol. Bull. (Woods Hole) 130:170-191. Gore, R. H., E. E. Gallaher, L. E. Scotto, and K. A. Wilson. 1981. Studies on decapod Crustacea for the Indian River region of Florida. XI. Community composition of structure, biomass and species areal relationships of seagrass and drift algae-associated macrocrustaceans. Estuarine Coastal Shelf Sci. 12:485-508. Greening, H. S., and R. J. Livingston. 1982. Diel variation in the structure of seagrass-associated epibenthic macroinvertebrate communities. Mar. Ecol. Prog. Ser. 7:147-156. Heck, K. L., Jr. 1976. Community structure and the effects of pollution in sea- grass meadows and adjacent habitats. Mar. Biol. 35:345-357. 1977. Comparative species richness, composition, and abun- dance of invertebrates in Caribbean seagrass (Thalassia testudinum) meadows (Panama). Mar. Biol. 41:335-348. 1979. Some determinants of composition and abundance of motile macroinvertebrate species in tropical and temperate turtlegrass (Thalassia testudinum) meadows. J. Biogeogr. 6:183-200. Heck, K. L., Jr., and R. J. Orth. 1980. Structural components of eelgrass (Zostera marina) meadows in the lower Chesapeake Bay-decapod Crustacea. Estuaries 3:289-295. Heck, K. L., Jr., and G. S. Wetstone. 1977. Habitat complexity and invertebrate species richness and abundance in tropical seagrass meadows. J. Biogeogr. 4:135-142. Hooks, T. A., K. L. Heck, and R. J. Livingston. 1976. An inshore marine invertebrate community: structure 449 FISHERY BULLETIN: VOL. 84, NO. 2 and habitat association in the northeastern Gulf of Mexico. Bull. Mar. Sci. 26:99-109. Howard, R. K., and K. W. Lowe. 1984. Predation by birds as a factor influencing the demo- graphy of an intertidal shrimp. J. Exp. Mar. Biol. Ecol. 74:35-52. Hughes, D. A. 1968. Factors controlling the emergence of pink shrimp (Penaeus duorarum) from the substrate Biol. Bull. (Woods Hole) 134:48-59. Kikuchi, T. 1966. An ecological study on animal communities of the Zostera marina belt in Tbmioka Bay, Amakusa, Kyushu. Amakusa Mar. Biol. Lab., Kyushu Univ., Vol. 1, No. 1. Kikuchi, T., and J. M. Peres. 1977. Consumer ecology of seagrass beds. In C. P. McRoy and C. Helfferich (editors), Seagrass ecosystems; a scientific perspective, p. 147-193. Marcel Dekker, Inc. Kjelson, M. A., and G. N. Johnson. 1978. Catch efficiencies of a 6.1-meter otter trawl for estu- arine fish populations. Trans. Am. Fish. Soc. 107:246- 254. Leber, K. M. 1983. Feeding ecology of decapod crustaceans and the in- fluence of vegetation on foraging success in a subtropical seagrass meadow. Ph.D. Thesis, Florida State Univ., Talla- hassee, 166 p. Lewis, F G., Ill, and A. W. Stoner. 1981. An examination of methods for sampling macrobenthos in seagrass meadows. Bull. Mar. Sci. 31:116-124. Livingston, R. J. 1975. Impact of kraft pulp-mill effluent on estuarine coastal fishes in Apalachee Bay, Florida, USA. Mar. Biol. 32:19- 48. 1976. Diurnal and seasonal fluctuation of organisms in a north Florida estuary. Estuarine Coast. Mar. Sci. 4:373-400. 1982. Trophic organization of fishes in a coastal seagrass system. Mar. Ecol. Prog. Ser. 7:1-12. Main, K. L. In press. Predator avoidance in seagrass meadows: An ex- amination of prey behavior, microhabitat selection, and cryp- tic coloration. Ecology. McNaughton, S. J. 1967. Relationships among functional properties of Califor- nia grasslands. Nature (Lond.) 216:168-169. Miller, R. E., D. W. Campbell, and P. J. Lunsford. 1980. Comparison of sampling devices for the juvenile blue crab, Callinectes sapidus. Fish. Bull., U.S. 78:196-198. Orth, R. J., and K. L. Heck, Jr. 1980. Structural components of eelgrass (Zostera marina) meadows in the lower Chesapeake Bay - fishes. Estuaries 3:278-288. PiELOU, E. C. 1977. Mathematical ecology. 2d ed. Wiley, N.Y. 385 p. Ryan, J. D. 1981. Diel predator-prey relationships in a subtropical sea- grass meadow: Apalachee Bay, Florida. MS Thesis, Florida State Univ., Tallahassee Saloman, C. H. 1979. New records of Caridean shrimps (Decapoda, Caridea) from the nearshore area of Panama City Beach, Florida, USA. Crustaceana(Suppl.)5:147-152. SlMBERLOFF, D. 1978. Use of rarefaction and related methods in ecology. Am. Soc Testing Materials, Spec. Tech. Publ. 652, p. 150-164. SOKAL, R. R., AND F. J. ROHLF. 1969. Biometry. W. H. Freeman and Co., San Franc, 776 p. Stoner, A. W. 1980. Feeding ecology of Lagodon rhomboides (Pisces: Spar- idae): Variation and functional responses. Fish. Bull., U.S. 78:337-352. Stoner, A. W, H. S. Greening, J. D. Ryan, and R. J. Livingston. 1983. Comparison of macrobenthos collected with cores and suction sampler in vegetated and unvegetated marine habitats. Estuaries 6:76-82. Stoner, A. W, and R. J. Livingston. 1980. Distributional ecology and food habits of the banded blenny Paraclinus fasciatus (Clinidae): a resident in a mobile habitat. Mar. Biol. 56:239-246. Warner, W. W. 1976. Beautiful swimmers; watermen, crabs and the Chesa- peake Bay. Penguin Books, N.Y. Weinstein, M. D, and K. L. Heck, Jr. 1979. Ichthyofauna of seagrass meadows along the Caribbean Coast of Panama and in the Gulf of Mexico: composition, structure and community ecology. Mar. Biol. 50:97-107. Zimmerman, M. S., and R. J. Livingston. 1979. Dominance and distribution of benthic macrophyte assemblages in a north Florida estuary (Apalachee Bay, Florida). Bull. Mar. Sci. 29:27-40. 450 NOTES A PRELIMINARY INVESTIGATION OF THE STOCK STRUCTURE OF THE DOLPHIN, CORYPHAENA HIPPURUS, IN THE WESTERN CENTRAL ATLANTIC Dolphin, Coryphaena hippurus, are fast swimming, migratory, pelagic fish, which support commercial and sport fisheries throughout the western central Atlantic (Erdman 1956; Zaneveld 1961; Beardsley 1967; Rose and Hassler 1969; Sacchi et al. 1981; Olsen and Wood 1982). In terms of weight and revenue, they are the most important large pelagic fish landed by the commercial fisheries in the south- eastern Caribbean (Mahon et al. 1981). In the north- west, they are the most important sport fish, being taken on more trips and in greater numbers by charter boats in Florida (Ellis 1957; Iversen 1962) and in North Carolina (Hassler and Hogarth 1977; Rose and Hassler 1969) than any other species. Rapid expansion of the dolphin fishery fleets is cur- rently underway in the eastern Caribbean, but the biological data necessary for management have not been gathered. For example, we remain ignorant of the number and distribution of stocks of C. hippurus in the western central Atlantic. Regional dolphin fisheries are markedly seasonal and this presumably results from migration; but migration patterns remain largely unknown (Palko et al. 1982). However, Beardsley (1967) believed that dolphin migrate northwards during spring and sum- mer, and Gibbs and Collette (1959) suggested that the spring abundance of C. hippurus in the Carib- bean may be a prespawning migration, mostly by females. A preliminary survey of regional catch records indicates a staggering of the peak fish- ing seasons, which supports the assumption that migration is large-scale (Hunte and Mahon 1982). In the present paper, we take three approaches to our investigation of C. hippurus in the western cen- tral Atlantic: 1) We use commercial and sport fishing data from several countries to examine seasonality and size structure of catch throughout the region; 2) we compare growth, age/size at sexual maturity, fecundity, and egg size of dolphin from different parts of the region; 3) we use electrophoretic tech- niques to compare dolphin sampled from Miami and Barbados, two widely spaced fisheries in the region. Electrophoretic techniques, combined with histo- FISHERY BULLETIN: VOL. 84, NO. 2, 1986. chemical staining for isozymes, are now widely recognized as a useful tool for examining genetic af- finities between fish stocks (Iwata 1975; Allendorf 1979; McGlade 1981; Ihssen et al. 1981; Ferris et al. 1982). By these means, we address the question of whether the dolphin fisheries in the western central Atlantic exploit a single stock migrating through the region or distinct units located in geographically con- tiguous areas. Resolution of this question will affect the extent to which individual territories should ex- pand their dolphin fisheries, will determine whether management programs need be regional or territory- specific, and will identify which territories need to collaborate for joint management of stocks. Methods Dolphin monthly catch data, recorded by commer- cial or sport fisheries, were obtained either by let- ter, personal visit to fisheries departments, and/or published literature (Table 1). The catch data, re- corded as numbers, weights, catch per day or per boat, and over time periods of 1 to 12 years, were standardized and plotted as percentages of total an- nual catch landed each month. Where more than 1 year's data were available, the average catch each month was calculated. Tissue samples for the electrophoretic survey were collected off Barbados between December 1982 and March 1983, and off Miami in May and June 1983. Samples of eye, heart, liver, gonad, and white mus- cle were taken from a total of 1,669 freshly landed dolphin and were deep frozen for later analysis. A survey of 22 enzymes encoded by 55 presumptive loci was conducted to identify polymorphic enzyme systems. The allelic frequencies of the highly poly- morphic isocitrate dehydrogenase, Idh-2, locus were compared in Miami and Barbados dolphin. The horizontal starch gel electrophoresis methodology follows that of May et al. (1979) and McGlade et al. (1983). Allelic nomenclature follows that of Allen- dorf and Utter (1979). Life history data were obtained from the literature, from records of length and weight of specimens caught in the Bahamas, Bermuda, and North Carolina, and from our own studies of 624 dolphin landed during the peak of the sport fishery in Miami and 3,126 dolphin landed by the commercial fishery in Barbados. 451 Table 1.— Countries from which catch data on the dolphin, Coryphaena hippurus, were obtained, with the data source for each country. Territory Data source Time period Territory Data source Time period Curacao Zaneveld (1961) 1957-58 Puerto Rico Erdman (1956) 1951-56 Grenada J. Finlay, Fisheries Officer, Ministry of Agriculture, National Resources and Industrial Development, St. George's, 1981-83 0. Munoz-Roure, Executive Director, Caribbean Fisheries Management Council, Hato Rey, Puerto Rico. Grenada. Bahamas P. Major, Fisheries Biologist, 1976, 1978 St. Vincent K. Morris, Fisheries Officer, Min- istry of Agriculture and Fish- eries, Kingstown, St. Vincent. 1975-81 Ministry of Agriculture, Fisheries, and Local Government, Nassau, Bahamas. Barbados R. Hastings and P. McConney, Fisheries Officers, Fisheries Division, Bay Street, Bridge- 1973-82 Florida A. Jones, Fisheries Scientist, Southeast Fisheries Center, NMFS, NOAA, Miami, Florida. 1970-80 town, Barbados. Fable et al. (1981) 1971-79 St. Lucia P. Murray, Fisheries Biologist, 1978, Georgia A. Jones, see Florida 1978-79 Ministry of Agriculture, Lands, 1980-82 South Carolina A. Jones, see Florida 1976-80 Fisheries, and Cooperatives, Fisheries Division, Castries, St. North Carolina A. Jones, see Florida 1978-80 Lucia. Manooch and Laws (1979) 1977 Martinique and Rose and Hassler (1969) 1961 Guadeloupe Sacchi et al. (1981) 1980 Bermuda B. Luckhurst, Fisheries Officer, 1973-80 Virgin Isles R. Wood, Fisheries Biologist, Department of Conservation and Cultural Affairs, Division of Fish and Wildlife, St. Thomas, Virgin Islands. Olsen and Wood (1982) 1967-78 Ministry of Fisheries and Agriculture, Naval Base, Southampton, Bermuda. RESULTS AND DISCUSSION Seasonality and Size Structure of Catch The seasonality of dolphin catch in 14 territories is shown in Figure 1. Martinique and Guadaloupe supplied no data, but information was given on the duration and peak of the dolphin season. It should be noted that the U.S. Virgin Islands is the only ter- ritory with a distinctly bimodal catch pattern. The peak months of catch in each territory are superimposed on a map of the western central Atlan- tic in Figure 2. Grenada peak catch is in February/ March; Barbados, St. Vincent, and St. Lucia in March/April; Martinique and Guadeloupe in April; and the Virgin Islands in April/May, giving the Virgin Islands their first and largest annual peak. This pattern of catch seasonality is suggestive of a stock (subsequently called the southern stock) moving northwest through the island arc. If the stock then turned west and moved past Puerto Rico, we would expect peak catch there to be between June, July, and August; but this is when Puerto Rico catches the least dolphin (see Figure 1). We therefore suggest that, on leaving the Virgin Islands, the stock moves northeasterly into the Atlantic, completing a circuit and returning to Grenada by February/March of the following year. This implies that there is a sec- ond stock (subsequently called the northern stock) located in the northwest region of the western cen- tral Atlantic. It occurs near Puerto Rico between December and February. It next moves northwest- erly past the Bahamas in April/May Florida and Georgia in May/June, South and North Carolina in June/July and Bermuda in July/August. It then com- pletes its circuit by passing through the Virgin Islands, giving that territory its second and smaller peak in November and returning to Puerto Rico by December/February. The mean size of fish caught in five territories dur- ing peak fishing season is shown in Figure 3, and the size structure of the catch throughout the fishing season in Barbados is shown in Figure 4. The data are consistent with the migration circuits proposed. In the northern stock, small presumably young-of- the-year fish are predominant during the summer when the stock is near Florida, North Carolina, and Bermuda. The mean size taken by the sport fishery in Florida is 1.69 kg; in North Carolina, where they occur 1 mo later, it is 2.92 kg and in Bermuda, where they occur 2 mo later, it is 3.85 kg. These differences presumably reflect growth within the cohort. The largest fish are taken by Puerto Rico, where Erd- man (1956) reported that dolphin up to 23 kg in weight occur during the peak winter fishing season, and by the Bahamas where the mean weight during the peak fishing months is 6.45 kg. This suggests 452 40 - 20 - 40 20 >- o a LU a: LL. 6-S 40 20 40 40 20 CURACAO GRENADA ST, VINCENT BARBADOS ST. LUCIA MARTINIQUE & GUADELOUPE US VIRGIN ISLES PUERTO RICO RGIA SOUTH CAROLINA NORTH CAROLINA BERMUDA J F M A M J J A S O N D JFMAMJJASOND i — i — i — i — i — i — i — i — i — i — i — r JFMAMJJASOND MONTHS Figure 1.— Seasonality of the dolphin, Coryphaena hippurus fisheries in the western central Atlantic, shown in geographical order from south to north. Note that raw catch data were not available from Martinique and Guadeloupe, but the duration of season and peak month were known. 453 ATLANTIC OCEAN Figure 2— Months of peak catch of the dolphin, Coryphaena hippurus, and proposed migration circuits for northern and southern dolphin stocks in the western central Atlantic. Letter symbols (eg., A-M) indicate months of peak catch. M^t indicate proposed migration. i S indicate proposed migration where catch data were not available • indicate locations from which samples for electrophoresis were collected. continued growth of the cohort as it leaves Bermuda and returns southwards into the northern Caribbean for the winter. Note that since dolphin are serial spawners and since fecundity is proportional to size (Beardsley 1967; Oxenford and Hunte in press), most spawning by a cohort will occur when the dolphin comprising it are large For the northern dolphin, this would be when the stock is near Puerto Rico, i.e, at the southeastern or up-current limit of their range Peak spawning near Puerto Rico is reported to occur in early spring (Erdman 1976) and presum- ably produces the small young-of-the-year fish caught near Florida during the summer. The size structure of dolphin caught at Barbados (Fig. 4) is consistent with the proposed migration for the southern stock. In February, the main cohort is composed of fish about 5V2 mo old with a mean standard length of 812.24 mm. Growth within this cohort occurs throughout the fishing season to June, when the average fish size is 1,007.83 mm SL (Oxen- ford and Hunte 1983). After this, abundance drops sharply (Fig. 1) as the cohort leaves Barbados 10 -, en u s 6 - 2 - BARBADOS FLORIDA NORTH BERMUDA BAHAMAS CAROLINA Figure 3— Mean weights of individuals of the dolphin, Coryphaena hippurus, landed during peak fishing seasons at five locations in the western central Atlantic. migrating northwards. During early summer (June/ July) and early autumn (October), the presence of 454 a V o h O) Q. J3 CO 01 (A 1 a. 0/ 3 S o I a) ■£ to o iS g k. r. 0) o> I T3 C 0> « — i J .SF t5 SB e8 5 M &s v & 1*3 g i 3 o ■5" o 1 c I AONBnoayd % i I Ed g 455 a few very small dolphin (<2% of the annual catch by weight), landed as bycatch of the flying fish fishery (see Figure 4), indicates the arrival of the first of the young-of-the-year group, with a few very large mature adults from the previous year. Note that many of these young-of-the-year are already mature on reaching Barbados in November, and all are ripe by the time the cohort leaves Barbados in June Note too, that aging of the cohort (Oxenford and Hunte 1983) suggests that the cohort was spawned between September and January, when the parent stock would be towards the southeastern, up- current extreme of the proposed migration circuit. Life History Comparisons Data on life history parameters of dolphin from northern and southern circuits are summarized in Table 2.— Life history characteristics of the dolphin Coryphaena hippurus in the western central Atlantic. Life history characteristics Location Data Source Northern area: Average 1st year growth k 1.64 N. Carolina Rose and Hassler (1968) rate (mm SL/d) = 1.82 Florida Beardsley (1967) Length-weight relationship Males: in the form y = ax6 y = 0.05 X 10 -8 ^.75 N. Carolina Rose and Hassler (1968) (y is weight (kg) y ~ 1 .45 x ■ 10 -7^.58 Florida Beardsley (1967, fig. 7) x is SL (mm)) Females: y = 1.27 X 10 -7x2.59 N. Carolina Rose and Hassler (1968) y ~ 5.75 x • 10 -8 ^.71 Florida Beardsley (1967, fig. 7) Fecundity-length relationship y = 2.52 x 10 -4 K312 Florida Beardsley (1967, fig. 11) in the form y = ax*3 (y is mature egg numbers x is FL (mm)) Size at first maturity Males: I 393 (mm SL) Females ! 324 Florida Beardsley (1967) Age at first maturity « 6-7 Florida Beardsley (1967) (months) Mature egg size range 1-1.7 Florida Beardsley (1967, fig. 9) (mm diameter) Mean mature egg size 1.3 N. Carolina Hassler and Rainville (mm diameter) (1975) Spawning season Extended Atlantic Florida Current Shcherbachev (1973) Fahay (1975) Johnson (1978) Gibbs and Collette (1959) Beardsley (1967) Age structure (% which are <2 yr) 96 98 N. Carolina Florida Rose and Hassler (1968) Beardsley (1967) Southern area: Average 1st year growth = 4.17 Barbados Oxenford and Hunte rate (mm SL/d) (1983) Length-weight relationship Males: in the form y = ax*3 y = 1.24 X 10 -8 ^.94 Barbados Oxenford and Hunte (y is weight (kg) Females: (in press) x is SL (mm)) y 2.22 x 10 -8 x284 Fecundity-length relationship in the form y = ax6 y = 2.7 x 10 -6 x367 Barbados this study (y is mature egg numbers x is FL (mm)) Size at first maturity Males: 735 Barbados this study (mm SL) Females: 610 Age at first maturity ss 4 Barbados this study (months) Mature egg size range 0.86-1.00 Barbados this study (mm diameter) Mean mature egg size 0.97 Barbados this study (mm diameter) Spawning season Extended Barbados this study Age structure 100 Barbados Oxenford and Hunte (% which are <2 yr) (1983) 456 Table 2 and are not supportive of a single stock hypothesis. Dolphin in Barbados waters appear to grow faster (Oxenford and Hunte 1983) than those in North Carolina (Rose and Hassler 1968) and Florida (Beardsley 1967). Note that scale annuli are found in northern dolphin but not in southern dolphin; a difference supportive of the assertion that the two groups are distinct. Beardsley (1967) sug- gested that the formation of the dolphin scale an- nuli at Florida was correlated with the temperature reduction occurring in the Florida Current during winter. Dolphin from Barbados are larger but younger at first sexual maturity than those from Florida. Fecundity increases with fish size in both groups, but Florida dolphin have higher fecundity at size than Barbados dolphin (Oxenford and Hunte in press). Mature eggs taken from Florida and North Carolina dolphin are apparently larger than those from Bar- bados dolphin. Intraspecific variation in egg size is seldom environmental and is typically a function of fish age (Bagenal 1971; Kazakov 1981). Mature egg size does not increase with fish size/age for Barbados dolphin (linear regression, r = 0.353 b = 0.0001). Therefore, assuming that the differences observed in egg size of southern and northern dolphin do not result merely from differences in investigators' methodologies, they are suggestive of separate stocks as shown for different spawning groups of herring (Blaxter and Hempel 1963; Cushing 1967) and sockeye salmon (Foerster 1968; Bagenal 1971). Electrophoretic Comparisons In the electrophoretic survey, 55 presumptive loci could be consistently scored. Of these, 39 were fixed for the same alleles in both samples, and a further 12 were close to fixation. Two isocitrate dehydro- genase loci (Idh-2,3) and two esterase loci (Est-1,2) had alternate alleles at a frequency >0.05, i.e. were significantly polymorphic Seven phenotypes were observed at the Idh-2 locus expressed in heart tissue (Fig. 5). The pattern of ac- tivity at this locus is typical of an active dimeric en- zyme with disomic inheritance (Darnall and Klotz 1972; Kirpichnikov 1981) and four alleles with relative mobilities to 100, 123, 86, and 68. Thus, putative genotypes could be assigned to the observed phenotypes as indicated in Figure 5, and allelic fre- quencies calculated (Table 3). Unequivocal assigna- tion of genotypes to the phenotypes, observed at the remaining polymorphic loci, was not possible in the absence of inheritance data, since the loci have alleles with overlapping mobilities. Idh-3 and Idh-2, ex- pressed together in liver tissue, both have alleles with relative mobilities to 100, 123, and 86, and although the asymmetrically banded phenotypes could be easily read, the presence of a null allele at Idh-3 meant that certain phenotypes could have been pro- duced by a number of different genotypes. Est-1 and Est-2 share all or some of four alleles, but the band- ing intensity ratios of individual phenotypes could not be determined. Hence, assignation of genotypes to phenotypes at these loci was not possible In sum- mary, only the Idh-2 locus, expressed indepen- dently from Idh-3 in heart tissue, was considered suitable for a comparison of Miami and Barbados dolphin. The frequencies of alleles at Idh-2 differed signifi- cantly in the two populations (chi-square 2x4 con- tingency test: x2 = 12.725, df = (r - 1) (C - 1) = 3, 0.01 > P > 0.0005; Table 3). Note that the varia- CO I— Q O £ 123 _ o + 100 86 68 — ' PHENOTYPE ca o 123/123 100/100 86/86 123/100 100/86 123/86 100/68 (1) (1) (1) (1-2-1) (1-2-1) (1-2-1) (1-2-1) GENOTYPE BANDING INTENSITIES Figure 5.— A starch-gel zymogram of the dimeric enzyme isocitric dehydrogenase, showing the phenotypes observed and putative genotypes at the Idh-2 locus in heart extracts of the dolphin, Coryphaena hippurus, from the western central Atlantic. Values in parentheses are ratios of allele products. 457 Table 3— Observed allelic frequencies (obs. freq.) and number (obs. no.) of alleles at the ldh-2 locus in heart tissue of the dolphin, Coryphaena hippurus. from Miami and Barbados. Expected values (exp. no.) refer to the number expected if the samples do not differ. Sample location No. of fish 539 Alleles 68 86 100 123 Miami obs. freq. 0.0009 0.3154 0.6660 0.0176 obs. no. 1 340 718 19 exp. no. (0.47) (304.14) (751.56) (21.83) Barbados 597 obs. freq. 0.0000 0.2521 0.7253 0.0226 obs. no. 0 301 866 27 exp. no. (0.53) (336.86) (832.44) (24.17) tion observed at ldh-2 did not differ from that predicted under Hardy-Weinberg equilibrium for either population (chi-square goodness of fit: for Bar- bados, x2 = 6.337, df = 3, 0.25 > P > 0.1; for Miami, X2 = 9.9145, df = 6, 0.25 > P > 0.1; Table 4). The differences in life history traits of Miami and Barbados dolphin could in principle be environmen- tal. The genetic differences observed at the ldh-2 locus suggest that there may be little gene flow be- tween the northern and southern groups; but could in theory result from a regional cline. The primary evidence supporting our suggestion of more than one dolphin stock in the western central Atlantic is there- Table 4.— The number of each phenotype observed (obs.) at the ldh-2 locus in heart tissue of the dolphin, Coryphaena hippurus, from Barbados and Miami. Expected values (exp.) refer to the numbers expected if the populations are in Hardy-Weinberg equilibrium. Sample location Putative genotype for ldh-2 Barbados (n = 597) Miami (n = 539) 86/86 obs. 47 64 exp. (37.94) (53.62) 100/86 obs. 199 205 exp. (218.31) (226.46) 100/100 obs. 325 251 exp. (314.05) (239.11) 100/123 obs. 17 10 exp. (19.58) (12.65) 123/123 obs. 1 1 exp. (0.31) (0.17) 123/86 obs. 8 7 exp. (6.81) (5.99) 68/68 obs. 0 exp. (0.00) 100/68 obs. 1 exp. (0.67) 123/68 obs. 0 exp. (0.02) 86/68 obs. 0 exp. (0.32) fore the seasonality of catch data and the mean size of dolphin landed in each territory. Taken together, the three data sets certainly suggest that the assumption of a single stock may be unjustified. Ef- forts should now be made to test the two stock hypothesis proposed and to investigate the possible presence of additional dolphin stocks, particularly in the western Caribbean Sea and in the Gulf of Mexico. Acknowledgments This project was supported by an Inter-University Council (British Council) grant to Hazel Oxenford, a University of the West Indies research grant to Wayne Hunte and a Manual Noreiga Morales Science Prize from the Organization of American States to Wayne Hunte. The electrophoresis was carried out at the Bedford Institute of Oceanography, Dart- mouth, Nova Scotia, with technical supervision by J. McGlade, and C. Annand, and with technical assistance from D. Beanlands. We thank C. Limouzy for collecting specimens in Miami, R. Mahon for assistance in compiling regional catch records, and J. Horrocks and J. Marsden for comments on the manuscript. Cooperation of fisheries officers and fisheries biologists in the region is gratefully acknowledged. Literature Cited Allendorf, F. W. 1979. Rapid loss of duplicate gene expression by natural selec- tion. Heredity 43:247-258. Allendorf, F. W., and F. M. Utter. 1979. Population genetics. Fish. Physiol. 8:407-454. Bagenal, T. B. 1971. The interrelation of the size of fish eggs, the date of spawning and the production cycle J. Fish. Biol. 3:207-219. Beardsley, G. L., Jr. 1967. Age, growth, and reproduction of the dolphin, Cory- phaena hippurus, in the Straits of Florida. Copeia 1967: 458 441-451. Blaxter, J. H. S., AND G. Hempel. 1963. The influence of egg size on herring larvae (Clupea harengus L.). J. Cons. Perm. Int. Explor. Mer 28:211- 240. Cushing, D. H. 1967. The grouping of herring populations. J. Mar. Biol. Assoc U.K. 47:193-208. Darnall, D. W., and I. M. Klotz. 1972. Protein subunits: a table (revised edition). Arch. Bichem. Biophys. 149:1-14. Ellis, R. W. 1957. Catches of fish by charter boats on Florida's east coast. Univ. Miami, Mar. Fish. Res. Spec Ser. Bull. 14, 6 p. Erdman, D. S. 1956. Recent fish records from Puerto Rico. Bull. Mar. Sci. Gulf Caribb. 6:315-348. 1976. Spawning patterns of fishes from the northeastern Caribbean. Agric Fish. Contrib. Off. Pub. Spec Serv. 7(2): 10-11. Fable, W. A., Jr., H. A. Brusher, L. Trent, and J. Finnegan, Jr. 1981. Possible temperature effects on charter boat catches of king mackerel and other coastal pelagic species in northwest Florida. Mar. Fish. Rev. 43(8):21-26. Fahay, M. P. 1975. An annotated (sic) list of larval and juvenile fishes cap- tured with surface-towed meter net in the South Atlantic Bight during four RV Dolphin cruises between May 1967 and February 1968. U.S. Dep. Commer., NOAA Tech. Rep. NMFS, SSRF-685, 312 p. Ferris, S. D., D. G. Buth, and G. S. Whitt. 1982. Substantial genetic differentiation among populations of Catostomus plebeius. Copeia 1982:444-449. Foerster, R. E. 1968. The sockeye salmon Oncorhynchus nerka. Bull. Fish. Res. Board Can. 162:1-422. Gibbs, R. H., Jr., and B. B. Collette. 1959. On the identification, distribution, and biology of the dolphins, Coryphaena hippurus and C. equiselis. Bull. Mar. Sci. Gulf of Caribb. 9:117-152. Hassler, W. W., and W. T. Hogarth. 1977. The growth and culture of dolphin, Coryphaena hip- purus, in North Carolina. Aquaculture 12:115-122. Hassler, W. W., and R. P. Rainville. 1975. Techniques for hatching and rearing dolphin, Cory- phaena hippurus, through larvae and juvenile stages. Univ. N.C., Sea Grant Program Publ. UNC-SG-75-31, 16 p. HUNTE, W., AND R. MAHON. 1982. How important are migratory patterns of pelagic fishes in the Caribbean? FAO Fish. Rep. 278, p. 165-175. Ihssen, P. E., H. E. Booke, J. M. Casselman, J. M. McGlade, N. R. Payne, and F. M. Utter. 1981. Stock identification: materials and methods. Can. J. Fish. Aquat. Sci. 38:1838-1855. Iversen, E. S. 1962. The dolphin fish. Sea Front. 8:167-172. Iwata, M. 1975. Genetic identification of walleye pollock Theragra chalcogramma populations on the basis of tetrazolium ox- idase polymorphism. Comp. Biochem. Physiol. 50(B):197- 201. Johnson, G. D. 1978. Development of fishes of the Mid-Atlantic Bight. An atlas of egg, larval, and juvenile stages. Vol. IV, Carangidae through Ephippidae U.S. Dep. Inter., Fish Wildl. Ser., Biol. Serv. Program, FWS/OBS-78/12, Jan. 1978:123-128. Kazakov, R. V 1981. The effect of the size of Atlantic salmon, Salmo salar L., eggs on embryos and alevins. J. Fish. Biol. 19:353-360. Kirpichnikov, V S. 1981. Genetic bases of fish selection. Springer-Verlag, Berlin, 410 p. (Translated from Russian by G. G. Gause.) Manooch, C. S., Ill, and S. T. Laws. 1979. Survey of the charter boat troll fishery in North Carolina, 1977. Mar. Fish. Rev. 41(4):15-27. Mahon, R., W. Hunte, H. Oxenford, K. Storey, and F. E. Hastings. 1981. Seasonality in the commercial marine fisheries of Bar- bados. Proc Gulf Caribb. Fish. Inst. 34:28-37. May, B., J. E. Wright, and M. Stoneking. 1979. Joint segregation of biochemical loci in Salmonidae: Results from experiments with Salvelinus, and review of the literature on other species. Can. J. Fish. Aquat. Sci. 36: 1114-1128. McGlade, J. M. 1981. Genotypic and phenotypic variation in the brook trout, Salvelinus fontinalis (Mitchill). Ph.D. Thesis, Univ. Guelph, Guelph, Ontario, 280 p. McGlade, J. M., M. C. Annand, and T. J. Kenchington. 1983. Electrophoretic identification of Sebastes and Helicolenus in the northwestern Atlantic. Can. J. Fish. Aquat. Sci. 40:1861-1870. Olsen, D A., and R. S. Wood. 1982. The marine resource base for marine recreation fish- eries development in the Caribbean. Proc Gulf Caribb. Fish. Inst. 35:152-160. Oxenford, H. A., and W Hunte. 1983. Age and growth of dolphin, Coryphaena hippurus, as determined by growth rings in otoliths. Fish. Bull., U.S. 81: 906-909. In press. Migration of the dolphin (Coryphaena hippurus) and its implications for fisheries management in the Western Central Atlantic. Proc Gulf Caribb. Fish. Inst. 37. Palko, B. J., G. L. Beardsley, and W J. Richards. 1982. Synopsis of the biological data on dolphin-fishes, Cory- phaena hippurus Linnaeus and Coryphaena equiselis Lin- naeus. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Cir. 443, 28 p. Rose, C. D., and W W. Hassler. 1968. Age and growth of the dolphin Coryphaena hippurus (Linnaeus), in North Carolina waters. Trans. Am. Fish. Soc 97:271-276. 1969. Application of survey techniques to the dolphin, Cory- phaena hippurus, fishery of North Carolina. Trans. Am. Fish. Soc 98:94-103. Sacchi, J., A. Lagin, and C. Langlais. 1981. La peche des especes pelagiques aux Antilles Francais. Etat actuel et perspective de developpement. Bull. Inst. Peches Marit. 312:1-15. ShcherbaChev, Y. N. 1973. The biology and distribution of the dolphins (Pisces, Coryphaenidae). J. Ichthyol. 13:182-191. Zaneveld, J. S. 1961. The fishery resources and the fishery industries of the Netherlands Antilles. Proc Gulf Caribb. Fish. Inst. 14:137- 171. Hazel A. Oxenford Department of Biology 459 University of the West Indies Cave Hill Campus, P.O. Box 6U Barbados, West Indies Bellairs Research Institute ofMcGill University St. James Barbados, West Indies ascent in the early morning and maintenance of a deeper distribution at night. This pattern was similar to that observed by Kiefer and Lasker (1975) for this Wayne Hunte species in the Gulf of California. Vertical chlorophyll a profiles indicated the cells rose in the morning and descended in the evening. The present study was undertaken to measure swimming speeds of G. splendens under different temperature conditions. The observed speeds vary with temperature and are similar to those calculated from field studies. EFFECTS OF TEMPERATURE ON SWIMMING SPEED OF THE DINOFLAGELLATE GYMNODINIUM SPLENDENS Dinoflagellate blooms or red tides frequently occur in a stratified water column having low nutrients near the surface (Huntsman et al. 1981). Under these conditions dinoflagellates have a competitive advan- tage over other phytoplankton due to their motility and diel vertical migration pattern. In the absence of turbulence, active swimming allows them to over- come sinking and thereby, remain close to the sur- face The normal diel vertical migration consists of an ascent to some minimum depth during the day and descent to a maximum depth at the night (reviewed by Forward 1976). Through this pattern they have access to nutrients over the area covered by migration and they can migrate to the surface during the day to obtain more light for photosyn- thesis (Ryther 1955; Margalef 1978; Huntsman et al. 1981). The success of dinoflagellates depends to a great extent upon their swimming capability. There have been few measurements of actual swimming speeds of individual dinoflagellates (eg., Hand et al. 1965) or estimates of speeds from population movements during migration (Eppley et al. 1968; Kamykowski and Zentara 1977). This is unfortunate because such measurements are necessary to estimate the depth of the water column available to dinoflagellates for nutrients during migration. The most pronounced and widespread dinoflagel- late blooms off the coast of Peru are caused by Gym- nodinium splendens Lebour. Blooms occur most frequently during the summer and are usually asso- ciated with the phenomenon of El Nino (Rojas de Mendiola 1979). At the beginning of the 1976 El Nino, there was a major bloom of G. splendens. Blasco's (1979) surface measurements during this bloom indicated the dinoflagellate vertically migrated with the suggested pattern involving an Materials and Methods The dinoflagellate Gymnodinium splendens Lebour was cultured as described previously (For- ward 1974) in a Sherer1 environmental chamber (Model CEL-44) on a 14:10 LD cycle at a salinity of about 34 ppt. All experiments were performed in the middle 4 h of the light phase with cultures having densities of about 2,000 cells/mL. This cell density was used because it was similar to that used in past studies (Forward 1974, 1977) and thus past results can be applied to the present study. Swimming speed during phototaxis was only measured during a specific time interval because there is a circadian rhythm in phototaxis (Forward 1974). Gymnodinium splendens shows about average levels of phototaxis during the middle 4 h of the light phase It is not known whether there is a similar rhythm in swim- ming. Subcultures were exposed to two sets of temper- ature conditions to test for the effects of 1) tem- perature acclimation and 2) acute temperature changes upon swimming speed. In the first tests cells were acclimated to selected temperatures from 13° to 25 °C for at least 5 d prior to swimming speed measurements. These temperatures were used because they encompass the range in which the cells grow at reasonable rates (Thomas et al. 1973). For the second tests, cultures were acclimated to 19°C for at least 5 d. At the time of testing cultures were exposed to an acute temperature change by placing the flasks in a water bath set at selected tempera- tures for 0.5 h, after which time swimming speed was measured. Room lights were on during this 0.5-h period. The temperature of the room in which swim- ming was measured was regulated to approximate- ly each test temperature Each test was performed on four separate subscultures. To measure swimming speeds, a sample of cells was removed from a subculture and placed in a clear 1 Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 460 FISHERY BULLETIN: VOL. 84, NO. 2, 1986. cuvette The cells were viewed by the closed circuit video system described by Forward (1974). During random swimming the cells can move in any direc- tion and are not necessarily moving in the plane of view of the video camera. Thus measurements of swimming speeds during random swimming tend to underestimate true speed. To prevent this problem, cells were stimulated horizontally with light and speed measured during oriented swimming toward the light (positive phototaxis). Room lights were off during light stimulation. The light stimulus was a tungsten light source filtered with a 4-96 Corning filter. The spectral com- position of the light was similar to the spectral sen- sitivity of phototaxis of G. splendens (Forward 1974). A constant light intensity of 4.79 x 102 Wm~2, as measured with an EG and G radiometer (Model 550) and calculated at 465 nm, was used for all tests because maximum positive phototaxis occurs in this intensity range, and it was necessary to measure swimming speed during phototaxis. Swimming was recorded on video tape and speed analyzed using previous techniques (Forward 1974). Results Swimming speed varied greatly with temperature (Fig. 1) as mean speeds approximately double upon acclimation to temperatures between 13° and 25 °C (0.56 h_1 to 1.16 mh_1). The dinoflagellate seems capable of limited temperature acclimation. If the cells were acclimated to 19°C and suddenly exposed to other temperatures, there was always a signifi- cant difference (Student's t test; P < 0.001) between these mean speeds and those upon acclimation. At a lower temperature of 13 °C the acclimation speed was higher; while at temperatures above 19 °C, the acclimation speeds were lower. This trend is expected with acclimation. The effects of temperature can be further assessed by calculating the temperature coefficients upon ac- climation and exposure to acute temperature changes (Table 1). The Q10 values for acute changes are always higher than those upon acclimation, which is expected if swimming rates are adjusted through acclimation. When acclimated to temperatures be- tween 13° and 19°C, the cells showed total compen- sation (Q10 = 0.98). In contrast, they were less able to adjust their rates upon acclimation to higher tem- peratures between 19° and 25°C (Q10 = 3.42). Par- tial acclimation occurred over this temperature Table 1.— Temperature coefficient values for the dinoflagellate Gymnodinium splendens upon tem- perature acclimation and exposure to acute changes in temperature. Temperature range Acute Q10 Acclimation Q10 13°-19°C 19°-25°C 1.47 4.68 0.98 3.42 13 15 17 19 21 23 25 Temperature (°C) Figure 1.— Swimming speeds of the dinoflagellate Gymnodinium splendens upon acclimation to various temperatures (solid line). The effect of acute temperature change was measured by acclimating the animals to 19°C and measuring speeds upon exposure to other temperatures (dashed line). The number beside the points are the sam- ple sizes and the vertical bars are standard errors. 461 range since the acclimation Q10 is lower than the acute Q10 (Table 1). levels can affect migration patterns (Kamykowski 1981). Discussion Blooms of G. splendens occur off the coast of Peru in temperatures ranging from 17° to 23 °C with op- timum being 18°-21°C (Rojas de Mendiola 1979). The lower temperature agrees with laboratory measurements of vertical migration, as Kamykowski (1981) found migration in the laboratory occurred at temperatures above 16°C. In the laboratory, this dinoflagellate can survive and divide at temperatures from 12° to 29°C. The most rapid growth rates (0.4 divisions/d), however, occur at 20°-27°C (Thomas et al. 1973). Within the optimum temperature range suggested from these studies (18°-26°C), swimming speed of G. splendens approximately doubles (Fig. 1). These speeds and their change with temperature are similar to those reported for other dinoflagellate species (Hand et al. 1965). The speeds of movement calculated from field studies of vertical migration of G. splendens agree with the speeds found in the present study. Blasco (1979) calculated that a speed of 1 m/h was sufficient to account for the migration off Peru during the 1976 El Nino. In the Gulf of California, G. splendens migrated over a depth of about 9 m and had a calculated descent velocity at sunset of 1.7 m/h (Kiefer and Lasker 1975). The present study pre- dicted this speed would occur at temperatures above 25°C. Unfortunately Kiefer and Lasker (1975) did not state the water temperature at the time of migration. An objective of the present study was to use the measured swimming speeds to determine the distance over which G. splendens should be capable of migrating. A conservative estimate of distance can be calculated from the speeds upon acclimation to optimum temperatures (19°-25°C) and assuming the dinoflagellate 1) swims continuously in either the up- ward or downward direction for half of the migra- tion cycle (12 h) and 2) does not have a diel rhythm in swimming speed. At 19°, 22°, and 25°C the cal- culated distances are 6.6, 11.3, and 13.9 m respec- tively. These distances would increase slightly if a temperature gradient existed because speed is ap- proximately constant at 19°C and lower tempera- tures, and acute exposure to higher temperatures, which would occur high in the water column, would elevate speeds above those upon acclimation (Fig. 1). In addition, these values would probably vary if G. splendens is exposed to different environmental con- ditions, since salinity, light intensity, and nutrient Acknowledgments This work was supported by an International Oceanographic Commission travel grant to BRM and a National Science Foundation Grant No. OCE- 8110702. Literature Cited Barber, R. T., and F. P. Chavez. 1983. Biological consequences of El Nino. Science 222:1203- 1210. Blasco, D. 1979. Changes in surface distribution of a dinoflagellate bloom of the Peru coast related to time of day. In D. L. Taylor and H. H. Seliger (editors), Toxic dinoflagellate blooms, p. 209- 214. Elsevier, N.Y. Eppley, R. W., O. Holm-Hansen, and J. D. H. Strickland. 1968. Some observations on the vertical migration of dino- flagellates. J. Phycol. 4:333-340. Forward, R. B., Jr. 1974. Phototaxis by the dinoflagellate Gymnodinium splendens Lebour. J. Protozool. 21:312-315. 1976. Light and diurnal vertical migration: photobehavior and photophysiology of plankton. In K. Smith (editor), Photo- chemical and photobiological reviews, Vol. 1, p. 157-209. Ple- num Press, N.Y. 1977. Effect of neurochemicals upon a dinoflagellate photo- response J. Protozool. 24:401-405. Hand, W. C, P. A. Collard, and D. Davenport. 1965. The effects of temperature and salinity change on swim- ming rate in the dinoflagellates, Gonyaulax and Gyrodi- nium. Biol. Bull. 128:90-101. Huntsman, S. A., K. H. Brink, R. T. Barber, and D. Blasco. 1981. The role of circulation and stability in controlling the relative abundance of dinoflagellates and diatoms over the Peru shelf. Coastal Upwelling Coast Estuarine Sci. 1:357- 365. Kamykowski, D. 1981. Laboratory experiments on the diurnal vertical migra- tion of marine dinoflagellates through temperature gradients. Mar. Biol. (Berl.) 62:57-64. Kamykowski, D., and S. J. Zentara. 1977. The diurnal vertical migration of motile phytoplankton through temperature gradients. Limnol. Oceanogr. 22:148- 151. Kiefer, D. A., and R. Lasker. 1975. Two blooms of Gymnodinium splendens, an unarmored dinoflagellate Fish. Bull., U.S. 73:675-678. Margalef, R. 1978. Life-forms of phytoplankton as survival alternatives in an unstable environment. Oceanol. Acta 1:493-509. Rhyther, J. H. 1955. Ecology of autotrophic marine dinoflagellates with reference to red water conditions. In F. H. Johnson (editor), The luminescence of biological systems, p. 387-414. Am. Assoc Adv. Sci., Wash., D.C. Rojas de mendiola, B. 1979. Red tide along the Peruvian coast. In D. L. Taylor and H. H. Seliger (editors), Toxic dinoflagellate blooms, p. 183-190. 462 Elsevier, N.Y. Thomas, W. H., A. N. Dodson, and C. A. Linden. 1973. Optimum light and temperature requirements for Gym- nodinium splendens, a larval fish food organism. Fish. Bull., U.S. 71:599-601. Richard B. Forward, Jr. Duke University Marine Laboratory Beaufort, NC 28516-9721 and Zoology Department, Duke University Durham, NC 27706 Blanca Rojas de Mendiola Instituto del Mar del Peru P.O. Box 22, Callao, Peru Richard T. Barber Duke University Marine Laboratory Beaufort, NC 28516-9721 MORPHOLOGY AND POSSIBLE SWIMMING MODE OF A YELLOWFIN TUNA, THUNNUS ALBACARES, LACKING ONE PECTORAL FIN In September of 1982, the Mexican bait boat, Paesa, fishing off Baja California, captured a 36.5 cm fork length (861.2 g wet weight) yellowfin tuna, Thunnus albacares, that lacked a left pectoral fin (Fig. 1). The fish was frozen and was brought to the Inter-Ameri- can Tropical Tuna Commission, La Jolla, CA, for study by W. H. Bayliff. Pectoral fins provide virtually all hydrodynamic lift in scombrids and are essential for stable and effi- cient swimming at sustained speeds (Magnuson 1973, 1978). A specimen with only one pectoral fin raises questions on what ways the fish might have compensated for an asymmetrical decrease in hydro- dynamic lift and how the presence of only one pec- toral fin might have affected its locomotion. We ex- amined the fish to determine what may have caused fin loss and whether morphology was noticeably altered in a manner suggesting some compensation. Skin in the area where the left pectoral fin should have been was thin, smooth, and silvery in appear- ance (Fig. 1). There was neither a trace of pectoral fin remnants nor a skin groove for it, suggesting the fin had never formed. On the other hand, the ap- pearance of the skin and the presence of variably sized scales in the area around the normal fin posi- tion is compatible with a healed wound, and we thus could not rule out the possibility that the fin had been bitten off cleanly. Methods The specimen was X-rayed and maximum body height and width measured. We measured and traced its median fins, caudal keel, pectoral fin, and both pelvic fins, and estimated their surface areas with a planimeter. The same body and fin measure- ments were made on similarly sized, preserved yellowfin tuna in the Scripps Institution of Ocean- ography Fish Collection (SIO). Morphometric data were compared with values derived from the litera- ture (Gibbs and Collette 1967; Fierstine and Walters 1968; Magnuson 1973, 1978; Magnuson and Wein- inger 1978, app. II). Although some of the specimen's caudal rays were bent (Fig. 1), all rays were present, and the fin was extended to a more natural position before its span was measured and area (which was well defined) traced. Also, to avoid measurement er- rors noted by Fierstine and Walters (1968) and Magnuson (1978), care was taken not to overextend caudal fins during span measurement. Density of the thawed fish was determined by water displacement (density = wet weight/displace- ment volume). The right and left pectoral girdles were then removed and the gas bladder was in- spected. Transverse sections were cut (see Graham et al. 1983), concentric myotomal rings on the right and left sides were counted, and red and white muscle were weighed for each section. Results The abundance of comparative morphometric and anatomical data for the yellowfin tuna permits a nearly complete assessment of the morphologic and hydrodynamic status of the one-finned specimen. The length (L; 36.5 cm)/weight (861.2 g) relationship and the density (1.080 g-mL-1) agree with values published for yellowfin tuna by Magnuson (1973, tables 1, 4). Also, the maximum thickness value (i.e., max. height + max. width/2 = 21.6% L) is within the range (20.5-23.0% L) measured for four SIO specimens (L from 28.5 to 42.5 cm) and near the value given by Magnuson (1973, table 7, 22.3% L). Finally, the point of maximum body thickness in the study fish (39.7% L) and that of SIO fish (36-40% L) are near Magnuson's value of 41.2% L (for fish from 28 to 45 cm L). The dorsal fin of this fish is normal in shape, with 13 spinous rays, a maximum height of 3.5 cm and a surface area of 9.5 cm2. The second dorsal fin is FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 463 Figure 1— Left- and right-side close-ups and a full-length, left-side photo of the Thunnus albacares with only one pectoral fin. 464 1 cm high and has an area of 2.0 cm2. The anal fin is also 1 cm high and has an area of 2.2 cm2. The combined total surface area of both sides of the sec- ond dorsal and anal fins is 8.4 cm2, which is larger than predicted (7.2 cm2) by the Magnuson and Weininger equation (1978, app. II). The total num- ber of second dorsal and anal fin rays and dorsal and ventral finlets agrees with that for other yellowfin tuna (Gibbs and Collette 1967, table 1). Table 1 compares caudal keel area and caudal and right pectoral fin dimensions of the study specimen and seven SIO fish of differing L. Also shown are values calculated for several of the same parameters using allometric equations for T. albacares (Magnuson 1978, table X; Magnuson and Weininger 1978, app. II). The caudal keel area of the one-finned fish (6.2 cm2) is smaller than the value expected from the equation (6.7 cm2) but is well within (i.e, 93%) the range of variation (77-102%) seen in the SIO specimens (Table 1). Comparison of the measured and the equation-derived caudal data for the one-finned fish with the same set of values for the next smallest (32.5 cm) and largest (37.0 cm) SIO fish indicates that the caudal fin of the one-finned fish has a slightly smaller span but larger area than would be expected for its L. This is further reflected in its aspect ratio (AR; 4.63), which is lower than that of any of the SIO specimens. This lower value prob- ably does not represent an artifact of preservation because in the other material caudal span and area increased directly with L. There is also general agreement between the measured and calculated values for each, showing that neither preservation nor measurement protocols affected caudal fin data. As would be expected from the underlying formulae, caudal AR calculated from the equations increases with L. However, among the measured data, there is no correlation between AR and L. It is also note- worthy that both the mean and predicted AR values of all of these small yellowfin (5.64, 5.34, Table 1) are in good agreement but well below the summary range (6.8-7.2) given for larger T. albacares by Magnuson (1978, table IX). This serves to empha- size that while AR may differ between species of scombrids (Magnuson 1978), it also varies within each species as a function of body size Both the length and area of the right pectoral fin of the one-finned fish are much less than those of the 37 cm SIO specimen (Table 1). When measured and computed pectoral fin areas are compared, there is good agreement between both values for the 37 and 42.5 cm L fishes but not for the 36.5 cm L one- Table 1.— Comparative caudal and right pectoral fin measurements for the one-finned yellowfin tuna (36.5 cm L) and seven specimens of different lengths (L) from the SIO collection. Data for each fish includes the actual measured values (m) and values calculated (c) from equations in the footnotes (Magnuson and Weininger 1978, app. II). Caudal k< Area1 (cm2) ;el Caudal fin Right pectoral fin Fork length (cm) Length Span2 (cm) Area3 (cm2) Aspect ratio4 Area5 (cm) (°/oL) (cm2) 25.8 m c 2.7 3.1 9.5 6.8 12.7 9.9 7.11 4.67 5.63 (21.8) 6.7 9.4 28.5 m c 3.8 3.8 8.0 7.7 12.3 12.1 5.20 4.90 6.00 (21-0) 5.3 11.3 31.5 m c 3.7 4.8 9.0 8.8 15.8 14.8 5.13 5.23 7.71 (24.5) 11.1 13.5 32.5 m c 4.8 5.2 10.0 9.1 15.7 15.8 6.37 5.24 7.25 (22.3) 10.6 14.2 636.5 m c 6.2 6.7 10.0 10.5 21.6 20.0 4.63 5.51 7.50 (20.5) 12.8 17.5 37.0 m c 5.3 6.9 11.0 10.7 21.6 20.6 5.60 5.56 9.67 (26.1) 17.8 17.9 40.0 m c 8.5 8.3 12.5 11.7 25.4 24.1 6.15 5.68 10.40 (26.0) 714.3 20.6 45.0 m c 8.8 10.8 12.2 13.5 30.3 30.7 4.91 5.92 11.00 (25.9) 25.3 25.4 m c x, SE 5.64, 0.30 x, SE 5.34, 0.15 'Caudal keel area = 2Caudal span = 3Caudal area = 4Aspect ratio = 5Pectoral fin area = 6One-finned fish. 7Fin was torn. 0.00198 L226. -2.27 + 0.35 L 0.013 L204 Span2/area 0.116 L178/4. 465 finned fish. In general, application of the pectoral area equation to the smaller SIO fish (Table 1) does not result in close correspondence between estimated and observed areas, suggesting that the relationship derived from larger individuals does not fit smaller yellowfin tuna. The relative length of the pectoral fin in yellowfin tuna changes abruptly with size In fish between about 35 and 42 cm L, pectoral fin length should normally be about 25% L (Gibbs and Collette 1967, fig. 26). This contrasts with the value for the one-finned fish of 20.5% L. The left pectoral girdle is present, but clearly ab- normal in gross examination. The posttemporal is reduced in overall size; the upper (pterotic) fork is somewhat reduced and lower (epiotic) fork weakly developed and without a flattened articular surface The rear margin of the supracleithrum is eroded, and the lateral surface rough. The cleithrum is almost as large as that of the right side, but the lateral groove for muscle attachment is reduced, and the upper process that normally curves out over the scapula is absent. The scapula is a block of bone with- out an articular facet for the first pectoral ray, and the scapular foramen is represented by a slit in the lateral surface The coracoid is much reduced pos- teriorly, and its reduced lower process is tightly ap- plied to the cleithrum so that the interosseus space is almost absent. The pectoral actinosts may be represented by a small lump of bone that is tightly attached to the scapula. A number of bone chips were embedded in the tissue overlying the pectoral girdle The postcleithra appear to be essentially normal. Elements in the left side of the pelvic girdle are larger and have a different orientation from those of the right. Also, the left pelvic fin is both smaller in area and shorter than the right (Fig. 2). Pelvic fin lengths and areas in the one-finned fish are left 2.9 cm, 3.2 cm2; right 3.5 cm, 4.7 cm2. Comparable values for the 37.0 L SIO fish are left 3.5 cm, 4.5 cm2; right 3.7 cm, 4.9 cm2. X-rays showed that the centra of vertebrae 19 and 20 are abnormal (Fig. 3). They lie parallel to one another and overlap by about 80% in the horizontal axis. There is considerable ero- sion of the adjoining surfaces of the two centra and their neural and haemal spines are displaced. This deformity, together with the reduced left pelvic fin, the absence of a left pectoral fin, and a deformed left pectoral girdle, suggests the presence of a con- genital malformation. As would be expected from our density findings, the gas bladder of the one-finned fish was small (17 x 5 mm, length x diameter), but about the same size as that of other yellowfin tuna (Magnuson 1973, 1978). Finally, we found no differences in the left and right body myotomes. The total red muscle was estimated to be 6.7% of wet weight, which is with- in the 95% confidence limits of the value reported for yellowfin tuna (5.2-7.8%) by Graham et al. (1983). Figure 2— Anterior ventral view showing the reduced size of the left pelvic fin. 466 h Figure 3— Tbp: Right-side X-ray of the vertebral column showing the impacted vertebrae and the neural and haemal spine displacement. Bottom: Dorsal X-ray of the same vertebrae Arrow indicates anterior. Scale is 2.5 cm. 467 Discussion Our study suggests that congenital defects led to the absence of a left pectoral fin, the formation of a small right pectoral and left pelvic fins, and to the impaction of two vertebrae A smaller caudal span may also be a result of such defects. On the basis of age studies (Uchiyama and Struhsaker 1981) we estimate that this fish (36.5 L) was about 9 mo old when captured. (But, because of the vertebral damage, the fish is shorter than it should be and 9 mo is a conservative age estimate) Thus in spite of significant locomotory handicaps, this fish had been swimming and feeding effectively at the time it was taken by hook and line. Morphological comparisons with SIO specimens and with equation-derived values for similarly sized yellowfin tuna did not indicate any major structural differences in the one-finned fish that can be inter- preted as having facilitated its swimming. However, since the absence of one pectoral fin doubtlessly af- fects the minimum speed required for hydrostatic equilibrium, the horizontal stability, and the maneu- verability of a tuna, it is instructive to consider how the loss might have been compensated. Magnuson (1973, 1978) has amply demonstrated the role of the paired fins in providing lift and reducing minimum equilibrium speed. Total lift (Lt) is calculated as Lt (dynes) = M[l - - (g)\ (1) where M is fish wet weight, Pe is seawater density, Pf is fish density, and g is the acceleration of gravi- ty (980 cm -sec-2). The amount of lift needed by the one-finned fish (M = 861 g, Pf = 1.08, Pe = 1.02 at 25°C) is 47,203 dynes. The minimum speed for hydrostatic equilibrium U100 is determined by U 100 PJ2 (CLAp + CLAk_ % (2) where CL is the coefficient of lift for the pectoral fins (p) and caudal keel (A;) and Ap and Ak are their respective areas (Magnuson 1973). Pectoral fin lift area includes both fins and the flat section of body between them (Magnuson 1978, fig. 4). This can be calculated from an allometric relationship (Mag- nuson 1973, table 4). Av = 0.0609 L187, (3) and, for a 36.5 cm L yellowfin, Ap = 50.8 cm2. With this value, a measured keel area (Table 1) of 6.2 cm2, and assuming a lift coefficient of 1.0 for both sur- faces (Magnuson 1973, table 4) the calculated (Equa- tion (2)) minimum speed for a 36.5 cm yellowfin tuna is 40.3 cm-s-1. The same calculation for the one- finned fish (Ap = 25.4 cm2) yields a minimum speed of 54.1 cm-s-1, a 34.3% increase The one-finned fish would need to swim faster, and thus expend more energy. Its higher speed would also probably have required it to make continuous velocity and position changes in order to keep pace with a school of, on average, similarly sized and thus slower swim- ming yellowfin tuna. Alternatively the fish might have assumed a pitched (i.e, head up) swimming mode in an attitude such that its body surface would have contributed to hydrodynamic lift by having a positive angle of attack relative to the direction of motion, and the CL of the caudal keel would be increased (Magnu- son 1978). Of course this would result in increased pressure drag and require more swimming power, but it might have enabled the fish to swim more slowly. Under any conditions, it seems likely that this fish was not highly maneuverable and would have diffi- culty remaining upright (i.e, not rolling to the left). It, of course, could not use its left pectoral for braking and left turns, and its left pelvic fin, which would also contribute to these actions, was less ef- fective than normal because of its small size Tunas normally accelerate with their first dorsal, pectoral, and pelvic fins appressed (Magnuson 1978), but as this fish slowed and needed lift it would have likely began to roll to its left as soon as its right pectoral fin was extended. This could be countered somewhat by its dorsal fin, but the necessity for unilateral use of the right pectoral fin should have always resulted in some amount of leftward roll and a tendency to turn to the right. Both the sharpness of the turn and the net upward or downward spiral movement of the fish would depend upon the degree of fin extension and swimming velocity. Finally, to compensate for the tendency to roll it is possible that the fish habitually swam with its body tilted as much as 80° to the right. In this position it would retain the largest possible pectoral lift area and might gain sufficient additional lift from the dor- sal, second dorsal, anal fins and the body surface to more than compensate for loss of keel lift. It is note- worthy that the second dorsal and anal fin areas of this fish are larger than predicted (see above). The fish would be able to roll from its side to an upright position merely by extending its pectoral fin a bit 468 farther. Also, side swimming would place both pelvic fins in a position where they could facilitate rapid left (now ventral) turns while possibly adding lift. Acknowledgments This study benefited from funds provided by the Foundation for Ocean Research at SIO, and by University of California, San Diego, Biomedical and Academic Senate research grants. We thank Cap- tain Jesus Yamamoto for saving the specimen for study and W. Bayliff for providing it to us. We also thank W Klawe, W Bayliff, and A. Dizon for com- menting on the manuscript. Literature Cited FlERSTINE, H. L., AND V. WALTERS. 1968. Studies in locomotion and anatomy of scombroid fishes. Mem. South. Calif. Acad. Sci. 6:1-29. GIBBS, R. H., AND B. B. COLLETTE. 1967. Comparative anatomy and systematics of the tunas, genus Thunnus. Fish. Bull., U.S. 66:65-130. Graham, J. B., F. J. Koehrn, and K. A. Dickson. 1983. Distribution and relative proportions of red muscle in scombrid fishes: consequences of body size and relationships to locomotion and endothermy Can. J. Zool. 61:2087-2096. Magnuson, J. J. 1973. Comparative study of adaptations for continuous swim- ming and hydrostatic equilibrium of scombroid and xiphoid fishes. Fish. Bull., U.S. 71:337-356. 1978. Locomotion by scombrid fishes. In W. S. Hoar and D J. Randall (editors), Fish physiology, Vol. 7, p. 239-313. Acad. Press, N.Y. Magnuson, J. J., and D. Weininger. 1978. Estimation of minimum sustained speed and associated body drag of scombrids. In G. D. Sharp and A. E. Dizon (editors), The physiological ecology of tunas, p. 293-311. Acad. Press, N.Y. UCHIYAMA, J. H., AND P. STRUHSAKER. 1981. Age and growth of skipjack tuna, Katsuwonus pelamis, and yellowfin tuna, Thunnus albacares, as indicated by daily growth increments of sagittae Fish. Bull., U.S. 79:151-162. Jeffrey B. Graham Richard H. Rosenblatt Darcy L. Gibson Physiological Research Laboratory and Marine Biology Research Division Scripps Institution of Oceanography La Jolla, CA 92093 CHROMOSOMAL ANALYSIS OF ALBACORE, THUNNUS ALALUNGA, AND YELLOWFIN, THUNNUS ALABACARES, AND SKIPJACK, KATSUWONUS PELAMIS, TUNA Chromosomal analysis is being used as part of an investigation of the population stock structure of the North Pacific albacore, Thunnus alalunga. There is a growing body of evidence (Brock 1943; Laurs and Lynn 1977; Laurs and Wetherall 1981; Laurs 1983) that North Pacific albacore are not as homogeneous as usually assumed (Clemens 1961; Otsu and Uchida 1963). Results from recent tagging studies suggest that northern and southern substocks constitute the North Pacific albacore population and that these proposed substocks have different migratory pat- terns (Laurs and Nishimoto 19791; Laurs 1983). Laurs and Wetherall (1981) also found that the growth rates were significantly different in the two proposed substocks. In addition, the differences in growth rate are consistent with differences in length frequencies of albacore caught in commercial fish- eries off North America (Brock 1943; Laurs and Lynn 1977). In this paper we report results from chromosomal analysis using C-banding for albacore (from the pro- posed North Pacific southern substock) and compare them with similar results obtained for yellowfin, Thunnus alabacares, and skipjack, Katsuwonus pela- mis, tuna. We demonstrate that there is a chromo- somal basis for placing the albacore and the yellow- fin tuna in the genus Thunnus and that recognizable chromosomal differences exist between the genera Thunnus and Katsuwonus. These findings corrobo- rate the taxonomy of the albacore and the yellowfin and skipjack tuna based on comparative anatomy (Gibbs and Collette 1967; Collette 1978). The results reported here are from part of a larger study, which is helping us to evaluate if genetic heterogeneity exists in the North Pacific albacore population. Information on chromosome character- istics is scarce for fishes, and to our knowledge this is the first time chromosome analyses have been reported for scombrid fishes. Materials and Methods All blood samples were collected from freshly caught fish either aboard the NOAA RV David Starr Jordan (August 1983) or aboard fishing boats »Laurs, R. M., and R. N. Nishimoto. 1979. Results from North Pacific albacore tagging studies. U.S. Dep. Commer., Natl. Mar. Fish. Serv., SWFC Admin. Rep. LJ-79-17, 9 p. FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 469 (October-November 1983). Because albacore have a high titer of red blood cells (Alexander et al. 1980), it was expedient to separate the lymphocytes from the erythrocytes. The lymphocytes were isolated from the blood on a density gradient of ficoll-sodium diatrizoate solution using a modification of the tech- nique developed by Boyum (1968), which is specific for the concentration of lymphocytes. We found that it was necessary to isolate the lymphocytes and place them in culture within a couple of hours after blood samples were collected. The ficoll gradient pro- cedure was not successful using undiluted hepari- nized blood that was retained for more than a few hours. Two albacore, three skipjack tuna, and four yellowfin tuna were sampled. All fish were juveniles which have virtually no sexual dimorphic character- istics, and no sex determinations were made. The estimated fork lengths of the fish ranged from 65 to 85 cm for albacore, 80 to 120 cm for yellowfin, and 45 to 55 cm for skipjack. From each fish, an 8-10 mL sample of blood was withdrawn via sterile intracardial puncture into a syringe coated with 1,000 units/mL of heparin. Two mL aliquots of blood were pipetted into each of the four 15 mL centrifuge tubes, and 4 mL of cell culture medium2 was added. The mixture was centrifuged at 20 g for 5 min, and the white cells and plasma were transferred to another centrifuge tube. This procedure for the separation of the plasma and white cell mixture was repeated three times following the suggestions given by Blaxhall (1981). Five mL of the white cell-plasma mixture were layered over 3 mL of ficoll-sodium diatrizoate solu- tion and centrifuged at 572 g for 30 min. The over- laying plasma was removed carefully with Pasteur pipets, and the lymphocytes below were transferred to a culture tube containing 5 mL of marine teleost cell culture medium (Michael and Beasley 1973). This procedure resulted in an erythrocyte free culture of lymphocytes having a higher mitotic index. The cultures were incubated at 25 °C for 3-5 d, at which time they were terminated and the cells harvested. The techniques for chromosomal analysis were pat- terned after those of Nowell (1960) for mammals because tuna are also endothermic (Graham and Dickson 1981). This work is an extension of the pro- cedures developed by Kelly and Laurs (19833). Prior to harvesting the cells, 0.5 \ng colcemid was added to 5 mL of culture medium and incubated for 2 h at 25° C. The culture was then centrifuged for 5 min at 180 g and the supernatant was replaced with 5 mL 0.075 M KC1 for 10 min. The culture tubes were centrifuged again for 5 min at 180 g, and the supernatant was replaced with 3 mL of freshly prepared cold fixative which consists of 3 parts methyl alcohol to 1 part glacial acetic acid and mixed for 1 h. The tubes were again centrifuged at 180 g for 5 min, decanted, and fixed. The cell pellet plus 0.5 mL of fixative was retained for slide preparation. Precleaned slides dipped in methanol and then in deionized water were used for slide preparations. Two drops of cell suspension were placed on the slide and 4 drops of fixative were immediately added. The slide was dried on a slide warmer at 37 °C and stored at room temperature for 24-72 h prior to C-banding. The C-banding procedures were patterned after the work of Pardue and Gall (1970) and Arrighi and Hsu (1971). In preparation for C-banding, the slides were placed in 0.2 N HC1 for 15 min at 37°C, rinsed in deionized water, treated with saturated Ba(OH)2 at room temperature for 7 min, and rinsed in deionized water. They were then immediately dipped again in 0.2 N HC1 for 10 s and rinsed in deionized water. After the final rinsing the slides were incubated in 2x sodium chloride-sodium citrate solution at 60 °C for 90 min and then stained for 90 min in Giemsa diluted with 1:10 Sorenson's buffer pH 6.8. Suitable metaphase figures were photographed at 1,008 x magnification using a Zeiss4 microscope equipped with a phase planapochromat 63/1.4 oil immersion lens. Results Chromosome Numbers Kelly and Laurs (fn. 3) found that the diploid number of chromosomes for albacore was 48. We have confirmed this observation and have found that the diploid numbers for yellowfin and skipjack tuna are also 48. The modal frequencies of about 90 cells containing 48 chromosomes were 82.2% for alba- core, 92.6% for yellowfin, and 80.5% for skipjack. Kelly and Laurs also observed that 85% of albacore cells had 48 chromosomes. Two polyploid cells with 2RPMI-1640 Sigma Cat. No. R6504. 3Kelly, Raymond M., and R. Michael Laurs. 1983. Summary of methods developed for investigations of albacore chromosomes and of findings made on number of chromosomes. Unpubl. field and laboratory notes and results (April 1983). [Raymond M. Kelly, School of Medicine, University of California, La Jolla, CA; R. Michael Laurs, National Marine Fisheries Service, NOAA, La Jolla, CA.] 4Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 470 96 chromosomes were observed in skipjack, and one polyploid cell with 96 chromosomes was observed in albacore. No polyploid cells were observed in yellowfin. Chromosome Morphology The albacore and the yellowfin and skipjack tuna were observed to have the same diploid chomosome number; however, their karyotype differed with respect to chromosome morphology. In this study, the chromosome pairs were arranged according to the morphology index (M), developed by Giannelli and Howlett (1967), which is obtained by dividing the length of the total haploid chromosome set p + q by the arm ratio (q/p). Based on our evaluation of 256 metaphase cells (Table 1), we found that the chromosome morphology of the yellowfin (Fig. 1) is more similar to that of the albacore (Fig. 2) than the skipjack (Fig. 3). The differences in chromosome morphology were most apparent in the three largest pairs of chromosomes (Table 1). The morphology index (M) places the metacentric and submetacentric chromosomes of the albacore and yellowfin in the number 1 and 2 positions respectively. Chromosome 3 of the albacore is also submetacentric while chro- mosome 3 of the yellowfin is referred to as sub- telocentric. The subtelocentric category is used to describe chromosomes in which the centromeres are displaced more towards the telomere when com- pared with submetacentrics. The metacentric chromosome of the albacore was consistently larger than the metacentric of the yellowfin. The remain- ing 42 chromosomes were telocentric in the albacore and yellowfin. All of the chromosomes of the skip- jack were telocentric. C-Banding Patterns C-banding determinations were done to differen- tiate individual chromosome characteristics among the three species of tunas (Table 2). The centromeric regions of most of the chromosomes of all three species contained C-band constitutive heterochroma- tin. However, there were differences in the inten- sities of staining on comparable chromosomes among the three species. Intercalary C-banding was observed only in the skipjack tuna and there was variability in terminal banding among the three species. In the albacore all chromosomes, except pair 10, showed C-banding in the centromeric regions with intense, prominent bands notably apparent in chromosome pairs 2 and 3 (Fig. 2). Terminal band- ing was restricted to chromosome pair 1 which had obscure C-bands on one arm of each homologue. No intercalary C-banding was observed in the albacore. There were some minor differences in the C-band- ing patterns between albacore and yellowfin tuna. In the yellowfin, the centromeric regions of all chromosomes were banded, the intensity of the banding in the centromeric region was uniform Table 1.— Classification of chromosome morphology for albacore and yellowfin and skipjack tuna. Chromosome number Albacore Yellowfin Skipjack 1 2 3 4-48 metacentric submetacentric submetacentric telocentric metacentric submetacentric subtelocentric telocentric telocentric telocentric telocentric telocentric Table 2.— Summary of C-banding characteristics for albacore and yellowfin and skipjack tuna. Location of bands Albacore Yellowfin Skipjack Centromeric Present on all chromosomes Present on all chromosomes Present on all chromosomes region Terminal bands Intercalary except pair 10; intensely prominent on pairs 2 and 3 Present on one arm of each homologue on pair 1; weakly developed None present with uniform prominent intensity Weakly developed on chromosome pairs 1, 3, 7, 8, 14, 15, 21, & 24 None present except pairs 10 and 19, great variability in inten- sity most prominent on pairs 1, 3, 4, 7, and 18 Notably prominent in pair 4 Present on all chromosome pairs except 17 and 24 471 1 H M , - V V Id ■ t» (18 DC 53 4' 8 1 4 • * r 10 11 12 || Lh |0 #5 II Itf 13 14 15 16 17 18 II 60 10 II & Aft 19 20 21 22 23 24 Figure 1.— Giemsa stained karyotype (upper row) and C-banding karyotype (lower row) of the same yellowfin tuna. among all chromosomes, and terminal banding was weakly developed on eight pairs of chromosomes (Fig. 1). As in the albacore, no intercalary banding was observed in the yellowfin. The following sig- nificant differences were observed in the C-banding patterns between the skipjack and the other two species: 1) all chromosomes except 10 and 19 had C-banding in the centromeric region, 2) there was great variability in the intensity of staining in the 472 centromeric region, 3) terminal banding was notably prominent in chromosome pair 4, and 4) there were intercalary bands on all chromosomes except pairs 17 and 24. Discussion Our results assist in understanding speciation pro- cesses that have occurred in the evolution of the 1 7 13 19 (a) I | 2 3 4 5 6 U 8 9 10 11 12 14 15 16 17 18 R 20 21 22 23 24 • • 3 8 10 11 12 13 14 15 16 17 18 19 (b) 20 21 22 23 24 Figure 2.— Giemsa stained karyotype (a) and C-banding karyotype (b) from two different fish of the North Pacific albacore. tuna. Gibbs and Collette (1967) proposed that seven species of tuna be included in the genus Thunnus on the basis of external morphological and internal anatomical characters. Our results demonstrate that there is a genetic basis for placing the albacore, T. alalunga, and the yellowfin tuna, T. albacares, in one genus Thunnus and the skipjack tuna, Kat- suwonus pelamis, in a separate genus. These relationships are based on the assumption that closely related species will share certain karyotypic characteristics. The determination that the albacore, yellowfin tuna, and skipjack tuna have the same number of chromosomes suggests that speciation of the genera of Thunnini might have occurred by intrachromo- somal rearrangement as opposed to Robertsonian changes as hypothesized for the rainbow trout, Salmo gairdneri (Thorgaard 1976). If speciation had 473 3 »" * 8 m 2 I ft i M II ! f i § 10 I H 11 •■ It 12 H 13 14 15 16 17 18 i 19 I ft* ft ii •* 20 21 22 23 24 Figure 3.— Giemsa stained karyotype (upper row) and C-banding karyotype (lower row) of the same skipjack tuna. involved a reduction in uniarmed chromosomes to form biarmed chromosomes, we would have ex- pected to find a difference in the chromosome number between Katsuwonus and Thunnus. It is probable that speciation within the genus Thunnus might also be related to chromosome rear- rangement because the number of chromosomes is the same. Pericentric inversion is a type of intra- chromosomal rearrangement that could result in the displacement of the centromere to convert a telo- centric chromosome into a metacentric one. Zenzes and Voiculescu (1975) suggested that pericentric in- version was involved in the chromosomal organiza- tion of the brown trout, Salmo trutta. The extent to which this mechanism has been related to the speciation of genera Thunnus and Katsuwonus is 474 uncertain. However, the occurrence of terminal C- bands on chromosome 1 of the albacore and chromo- somes 1 and 3 of the yellowfin tuna is consistent with the hypothesis that these biarmed chromosomes were derived from a uniarmed condition. Indeed, White (1951) believed that, in grasshoppers, telo- centric chromosomes are more primitive than the metacentric condition. Absence of terminal bands on chromosomes 2 and 3 of the albacore and chromo- some 2 of the yellowfin tuna does not preclude the suggested derivation of metacentric chromosomes. It is possible that in the metacentric chromosomes lacking terminal bands, centrometric heterochroma- tin either was not moved or was lost. It is also possi- ble that chromosome rearrangement in the specia- tion of the albacore and yellowfin occurred through changes in the euchromatic portions of chromo- somes. To test this hypothesis it will be necessary to use G-banding techniques (Rishi 1978) to conduct analysis of these portions of the chromosomes. In contrast to the albacore and yellowfin tuna, the telocentric chromosomes of the skipjack tuna showed a variety of intercalary and terminal C-band- ing in addition to those of the centromeric regions. An interesting condition was the polymorphic ter- minal heterochromatic block that occurred in chro- mosome pair number 4 of the skipjack, but not in the albacore or yellowfin. While the four specimens of skipjack analyzed had this polymorphism, it is not possible to comment on the frequency with which it might occur in the population. This type of dif- ferential banding also occurs in other fishes as demonstrated by Zenzes and Voiculescu (1975) who observed a difference in the size of C-bands in Salmo trutta. The C-band polymorphism we observed in skipjack could be related to the sex determining mechanism of the fish. However, we do not have any information on the sex of the skipjack used in this study and most fish do not have heteromorphic sex chromosomes (Zenzes and Voiculescu 1975; Thor- gaard 1976; Kligerman and Bloom 1977). An excep- tion occurs in the eels which have highly hetero- morphic sex chromosomes (Park and Grimm 1981). Analysis of C-banding patterns associated with the morphological differences in chromosomes has per- mitted us to identify all of the chromosome pairs of the albacore, yellowfin tuna, and skipjack tuna. We have demonstrated that karyotype analysis may pro- vide a chromosomal basis for placing albacore and yellowfin in Thunnus and skipjack in Katsuwonus. Although C-banding techniques did not allow a detailed evaluation of the Thunnus chromosomes, we believe that the use of multiple banding pro- cedures could provide important information on the speciation and cytotaxonomy of the species of this commercially important genus. In addition, use of G-banding procedures will be an important next step in determining if genetic heterogeneity exists in the North Pacific albacore population. Acknowledgments We wish to thank A. Dean Stock (City of Hope Hospital, Duarte, CA) and James Mascarello (Children's Hospital, San Diego, CA) for many helpful suggestions, Raymond Kelly (University of California Medical School, San Diego, CA) for help in procurement of materials, and the personnel of the San Diego Sportsfishing Association, Captains Ed McEwen and Buzz Brizendine for space aboard their boats to make this work possible. Literature Cited Alexander, N., R. M. Laurs, A. McIntosh, and S. N. Russell. 1980. Haematological characteristics of albacore, Thunnus alalunga (Bonnaterre), and skipjack Katsuwonus pelamis (Linnaeus). J. Fish. Biol. 16:383-395. Arrighi, F. E., and T. C. Hsu. 1971. Localization of heterochromatin in human chromo- somes. Cytogenetics 10:81-86. Brock, V. E. 1943. Contribution to the biology of the albacore (Germo alalunga) of the Oregon coast and other parts of the North Pacific. Stanford Ichthyol. Bull. 2:199-248. Blaxhall, P. C. 1981. A comparison of methods used for the separation of fish lymphocytes. J. Fish. Biol. 18:177-181. Boyum, A. 1968. Separation of leukocytes from blood and bone marrow. Scand. J. Clin. Lab. Invest., 21, Suppl. 97:77. Clemens, H. B. 1961. The migration, age, and growth of Pacific albacore (Thunnus germo), 1951-1958. Calif. Dep. Fish Game, Fish Bull. 115, 128 p. Collette, B. B. 1978. Adaptations and systematics of the mackerels and tunas. In G. D. Sharp and A. E. Dizon (editors), The physiological ecology of tunas, p. 7-88. Acad. Press, N.Y. GlANNELLI, F., AND R. M. HOWLETT. 1967. The identification of the chromosomes of the E group (16-18 Denver): an autoradiographic and measurement study. Cytogenetics (Basel) 6:420-435. Gibbs, R. H., Jr., and B. B. Collette. 1967. Comparative anatomy and systematics of the tunas Genus Thunnus. U.S. Fish Wildl. Serv., Fish. Bull. 66: 65-130. Graham, J. B., and K. A. Dickson. 1981. Physiological thermoregulation in the albacore tuna Thunnus alalunga. Physiol. Zool. 54:470-486. Kligerman, A. D., and S. E. Bloom. 1977. Distribution of F-bodies, heterochromatin, and nucleo- lar organizers in the genome of the central mudminnow, Um- bra limi. Cytogenet. Cell Genet. 18:182-196. 475 Laurs, R. M. 1983. The North Pacific albacore - an important visitor to California Current water. Calif. Coop. Oceanic Fish. Invest. Rep. 24:99-106. Laurs, R. M., and R. J. Lynn. 1977. Seasonal migration of North Pacific albacore, Thun- nus alalunga, into North American coastal waters: distribu- tion, relative abundance, and association with Transition Zone waters. Fish. Bull., U.S. 75:795-822. Laurs, R. M., and J. A. Wetherall. 1981. Growth rates of North Pacific albacore, Thunnus alalunga, based on tag returns. Fish. Bull., U.S. 79:293- 302. Michael, S. M., and A. R. Beasley. 1973. Marine teleost fish tissues. In P. F. Kruse, Jr. and M. K. Patterson, Jr. (editors), Tissue culture methods and ap- plication, p. 134. Acad. Press, N.Y. Nowell, P. C. 1960. Phytohaemagglutinin: an initiator of mitosis in culture of normal human leukocytes. Cancer Res. 20:462-466. Otsu, T., and R. N. Uchida. 1963. Model of the migration of albacore in the North Pacific Ocean. U.S. Dep. Inter., Fish. Wildl. Serv., Fish. Bull. 63:33-44. Pardue, M. L., and J. G. Gall. 1970. Chromosome localization of mouse satellite DNA. Science 168:1356-1358. Park, E. H., and H. Grimm. 1981 . Distribution of C-band heterochromatin in the ZW sex chromosomes of European and American eels (Anguillidae, Teleostomi). Cytogenet. Cell Genet. 31:167-174. Rishi, K. K. 1978. Giesma-banding in fish chromosomes. Current Science 47:393-394. Thorgaard, G. H. 1976. Robertsonian polymorphism and constitutive hetero- chromatin distribution in chromosomes of the rainbow trout (Salmo gairdneri). Cytogenet. Cell Genet. 17:174-184. White, M. J. D. 1951. A cytological survey of wild populations of Trimero- tropis and Circotettix (Orthoptera, Acrididae). II. Racial dif- ferentiation in T. sparsa. Genetics 36:31-53. Zenzes, M. T., and I. Voiculescu. 1975. C-banding patterns in Salmo trutta, a species of tetra- ploid origin. Genetica 45:531-536. San Diego State University San Diego, CA 92182 San Diego State University San Diego, CA 92182 Present address: Wuhan University Wuhan, Peoples Republic of China Southwest Fisheries Center National Marine Fisheries Service, NOAA P.O. Box 271, La Jolla, CA 92038 F. J. Ratty Y. C. Song R. M. Laurs ABUNDANCE, SIZE, AND SEX RATIO OF ADULT SEA-RUN SEA LAMPREYS, PETROMYZON MARMUS, IN THE CONNECTICUT RIVER1 Populations of sea-run sea lampreys, Petromyzon marinus, occur in many rivers on the east coast of North America from Labrador to Florida (Bigelow and Schroeder 1953). The Connecticut River in the northeastern United States is believed to have the largest population (Beamish 1980). Although the historical, upstream range of the sea lamprey in the Connecticut River is not known, it probably was similar to American shad, Alosa sapidissima, which migrated 280 km upstream to Bellows Falls, VT (Moffitt et al. 1982). Upstream migration of anadromous fish species in the Connecticut River main stem was first re- stricted in 1798 by the construction of Turners Falls Dam at km 197, and further in 1849 by the construc- tion of Holyoke Dam at km 140. The first upstream fish passage facility for anadromous fish was a fish lift at Holyoke Dam that began operating in 1955. Until 1969 the sea lampreys using the fish lift were counted and either killed or thrown back. From 1969 to 1984, they have been passed upstream each year. Sea lampreys have also used the fish ladders that were completed in 1980 and 1981 at Turners Falls and Vernon Dams, respectively. With the comple- tion of the fish ladder at Bellows Falls Dam in 1984, migrants now have access to 350 km of main-stem river and many additional tributaries (Fig. 1). The present report summarizes the annual counts of sea lampreys from 1958 to 1984 at the two Holy- oke fish lifts (a second fish lift was added in 1976). We also examined the sex ratio, total length, and weight of adults in 1981-82 and compared these characteristics with those of the population in the St. John River, New Brunswick. Beamish et al. (1979) sampled the St. John River population at km 140, at a fish lift located at Mactaquac Dam. Methods Sea lampreys that were lifted above the dam were counted each year from 1958 to 1984, except for the period from 1969 to 1974. From 1958 to 1968, sea lampreys were counted by personnel of the Holyoke Water Power Company (the owner of the dam), and Contribution No. 95 of the Massachusetts Cooperative Fishery Research Unit, which is supported by the U.S. Fish and Wildlife Service, Massachusetts Division of Fisheries and Wildlife, Mass- achusetts Division of Marine Fisheries, and the University of Massachusetts. 476 FISHERY BULLETIN: VOL. 84, NO. 2, 1986. Figure 1.— Map of the Connecticut River showing the location of Holyoke Dam and the other dams with fishways on the lower 350 km of the main stem and major tributaries. Dams that sea lampreys can pass are designated by an open bar; dams they cannot pass are designated by a solid bar. from 1975 to 1984, they were counted by person- nel from either the Massachusetts Division of Fish- eries and Wildlife or the Massachusetts Cooperative Fishery Research Unit. Until 1975, fish of all species were lifted, deposited into small carts, carried across the dam, and counted as they were released. Begin- ning in 1975, all fish were sluiced directly from the fish lift bucket into a large flume and were counted through a glass window in the side of the flume as they swam upstream. The accuracy of these counts has not been experimentally determined. However, the counts are probably very accurate because the sea lampreys are large and swim slowly through the flume. We collected sea lampreys daily at the fish lift trap from 1 May to 10 June 1981, and from 10 May to 30 June 1982 for determination of total length (TL) and sex. The number of sea lampreys sampled each day was proportional to the number lifted the previous day. The number of sea lampreys lifted and (in parenthesis) the number collected follow: 0-50 (2); 51-100 (4); 101-200 (6); 201-400 (8); 401-800 (10); 801-1,000 (15); 1,001-2,000 (25); 2,001-3,000 (30); 3,001-5,000 (40); >5,000 (50). in both years, total 477 length was measured to the nearest millimeter and sex.was determined by dissection. We determined the sex ratio for each day of the run to observe changes during the migration. In*1982 each sea lam- prey was also weighed to the nearest gram. Chi- square tests were used to compare the sex ratios for differences from a 1:1 frequency. Student's t-test was used to compare the males and females for mean length and weight. We compared males and females for the length-weight relationship by cal- culating a separate regression for each sex using the logarithmic equation: log w = log a + (b) (log 1) (Ricker 1975). Results and Discussion Abundance The numbers of sea lampreys lifted from 1958 to 1967 were relatively few, and probably reflected the inefficiency of the fish lift rather than a small popula- tion (Fig. 2). After the flume and second lift were added in 1975 and 1976, respectively, 22,000-53,000 adults have been passed upstream each year. The 53,000 counted in 1981 was the largest number ever passed at Holyoke and the largest run documented in any river. In 1981, 59% of the total run was lifted during the week of 24-30 May; and in 1982, 68% were lifted during the week of 28 May-3 June. Beamish (1980) reported that about 8,600 sea lam- preys are lifted annually in the fish lift at Macta- quac Dam. He estimated the spawning populations in other northern streams at <8,000. The sea lampreys that reach Holyoke Dam are only a portion of the total run, because several tributaries below the dam support populations (Whitworth et al. 1976). The sea lamprey popula- tion may increase as adults gain access to additional spawning and rearing habitat in headwater streams by using fish passage facilities constructed for Atlan- tic salmon, Salmo salar, and American shad (Mof- fitt et al. 1982). Thus, the restoration program designed primarily for Atlantic salmon and Ameri- can shad is also restoring the sea lamprey to addi- tional habitat. Since 1975, over 20,000 sea lampreys have been passed each year at Holyoke Dam and given access to new spawning and rearing habitat. The estimated life span of sea lampreys in the St. John River is estimated at 9-12 yr (Beamish and Pot- ter 1975). Therefore, if the Connecticut River population returns to their natal stream and has a similar life cycle, and if the strength of the year classes after 1975 was enhanced by the additional rearing habitat above Holyoke, then beginning in 60,000- 50,000- 40,000 (Z 30,000 LU CD •^ 20,000 600- not counted 58 60 62 64 66 68-74 76 78 80 82 84 YEAR Figure 2.— Number of adult sea lampreys lifted in the Holyoke fish lifts each year, 1958-84. 478 1984 there should be increased returns of adults at Holyoke. The return of sea lampreys at Holyoke in 1984 was not a record return, but this could be due to the high discharge caused by the 50-yr flood that occurred in early June 1984, when most sea lam- preys are lifted. If the sea lamprey population in- creases, the wound frequencies should increase on host species of marine and anadromous fish. Sex Ratio Sex ratios for both years were skewed from 1:1 in favor of males, but the ratio was only significant in 1982: in 1981, 56% were males (ratio: 1.3:1; x2 = 3.4, P > 0.05; in 1982, 62% were males (ratio: 1.6:1; x2 = H.6, P < 0.005, Table 1). Sex ratios also changed during the spawning migration with the proportion of males increasing late in the run. The percent of males in the early and late periods were 55 and 59% in 1981 and 59 and 67% in 1982. The increase in the proportion of males was not signifi- cantly different from a 1:1 ratio in 1981, but the in- crease was significant in 1982 (x2 = 7.6:P < 0.01). Applegate (1950) found that males in landlocked sea lamprey populations increased to about 75% in the late part of the run. The reason for this phenomenon is unknown. Males are the most abundant sex in stable popula- tions of sea-run and landlocked sea lampreys. Beamish et al. (1979) reported 55% males (ratio: 1.36:1) in nearly mature adults in the St. John River in 1974-77 (Table 1). Davis (1967), who collected anadromous sea lampreys for 5 yr from Barrows Stream, ME, reported a male:female ratio of 1.9:1; however, the sample size was very small (N = 66). Potter et al. (1974) found an excess of males in land- Table 1. — Mean total length and weight (SE in parenthesis), and percent males in sea lampreys sampled at Holyoke Dam, Connec- ticut River, compared with samples collected from the Mactaquac Dam, St. John River. Dam & N Mean length (cm) Mean length (cm) Male Female year Male Female male Holyoke 1981 464 71.3 (2.8) 71.5 (2.9) — — 56 1982 404 71.4 71.1 1794 2806 62 (2.7) (3.6) (8.2) (12.01) Mactaquac3 1974-77 341 72.4 72.9 868 885 55 (4.7) (5.1) (18.1) (18.3) '249 males were weighed. 2155 females were weighted. 3Data from Beamish et al. (1979); ± 95% confidence limits in parenthesis. locked sea lamprey (ratio: 1.26:1). The sex ratio in stable populations (where males are more abundant than females) is different from the ratio in popula- tions from the upper Great Lakes, where an excess of females is typical of populations being eradicated or controlled (Smith 1971). Sex ratios in sea lam- preys also vary with cycles of abundance (Wigley 1959; Smith 1971), and temperature and nutrition may differentially affect growth and survival of male and female ammocoetes (Hardisty 1954). Total Length and Weight In 1981, 464 sea lampreys (0.9% of the number lifted) were measured for total length; in 1982 the number examined was 404 (1.5% of the number lifted). There was no significant difference between the mean length of males and females during either year or for both years (Student's t-test: P > 0.05, Table 1). Length of females and males ranged from 60 to 85 cm in both years. The similarity in mean total length of adults in the consecutive spawning runs of 1981-82 suggests relative stability of the sea lamprey population. This differs greatly from the unstable sea lamprey populations in the Great Lakes where body length decreased from 1950's to 1960's— changes related to decreases in food supply and changes in the en- vironment (Smith 1971). The mean weight of females was not significant- ly different from the mean weight of males (Stu- dent's t-test: P > 0.05, Table 1). We determined the length-weight relationship by using the regression equations: log w = -3.42 ± (2.21) (log 1), (r2 = 0.75, P < 0.01) for females and log w = -3.11 ± (2.10) (log 1), (r2 = 0.76, P < 0.01) for males. There was no significant difference between the slopes of the regression lines, consequently we combined males and females (N = 404). Using the equation y = b + mx or weight = b + (slope) (length), a highly significant correlation (r2 = 0.76, P < 0.01) was found for the regression equation: weight = 521.9 + (0.23890) (length). The length-weight relationship is linear, rather than sigmoid, as it is in most fishes. Because the body is attenuate, the weight of sea lam- preys does not increase as rapidly with length as it does in most other fishes. This relationship is less evident in females, possibly because of the additional weight of their eggs. Generally, in landlocked populations, females are slightly heavier than males because of their high fecundity (Applegate 1950). We also found this was true. Although the sea lampreys at Holyoke Dam were similar in length to those in the St. John River, 479 the average weight of Connecticut River fish was considerably less (Table 1). The difference in average weight between sea lampreys in the two populations is not due to the difference in location of upstream sampling sites, but possibly to differences in energetic requirements, food supplies, or some aspect of the environment during the oceanic parasitic phase. A difference in weight between populations has previously been found in landlocked sea lampreys in the Great Lakes (Smith 1971). Acknowledgments This project was supported by Federal Aid Pro- ject AFS-4-R-21 and D-J Project 5-29328 to the Massachusetts Division of Fisheries and Wildlife and the Massachusetts Cooperative Fishery Research Unit. We thank P. Eschmeyer for a valuable review of the manuscript. Literature Cited Applegate, V. C. 1950. Natural history of the sea lamprey, Petromyzon marinus, in Michigan. U.S. Fish Wildl. Serv., Spec. Sci. Rep.-Fish. 55, 237 p. Beamish, F. W. H. 1980. Biology of the North American anadromous sea lam- prey, Petromyzon marinus. Can. J. Fish. Aquat. Sci. 37: 1924-1943. Beamish, F. W. H., and I. C. Potter. 1975. The biology of the anadromous sea lamprey (Petro- myzon marinus) in New Brunswick. J. Zool. (Lond.) 177: 57-72. Beamish, F. W. H., I. C. Potter, and E. Thomas. 1979. Proximate composition of the adult anadromous sea lamprey, Petromyzon marinus, in relation to feeding, migra- tion and reproduction. J. Anim. Ecol. 48:1-19. BlGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53:1-577. Davis, R. M. 1967. Parasitism by newly-transformed anadromous sea lam- preys on landlocked salmon and other fishes in a coastal Maine lake. Trans. Am. Fish. Soc. 96:11-16. Hardisty, M. W. 1954. Sex ratio in spawning populations of Lampetra planeri. Nature (Lond.) 173:874-875. Moffitt, C. M., B. Kynard, and S. G. Rideout. 1982. Fish passage facilities and anadromous fish restoration in the Connecticut River basin. Fisheries (Bethesda) 7(6): 2-11. Potter, I. C, F. W. H. Beamish, and B. G. H. Johnson. 1974. Sex ratios and lengths of adult sea lampreys (Petro- myzon marinus) from a Lake Ontario tributary. J. Fish. Res. Board Can. 31:122-124. Ricker, W. E. 1975. Computation and interpretation of biological statistics offish populations. Fish. Res. Board Can., Bull. 191:1-382. Smith, B. R. 1971. Sea lampreys in the Great Lakes of North America. In M. W. Hardisty and I. C. Potter (editors), The biology of lampreys, Vol. 1, p. 207-247. Acad. Press, Lond. Whitworth, W. R., P. L. Berrien, and W. T. Keller. 1976. Freshwater fishes of Connecticut. State Geol. Nat. Hist. Serv. Conn., Dep. Environ. Prot., Bull. 101, 134 p. WlGLEY, R. L. 1959. Life history of the sea lamprey of Cayuga Lake, New York. U.S. Fish Wildl. Serv., Fish. Bull. 59:560-617. Kathleen Stier Boyd Kynard Massachusetts Cooperative Fishery Research Unit 204 Holdworth Hall University of Massachusetts Amherst, MA 01003 AN IMPROVED OTTER SURFACE SAMPLER Field trials using a neuston sampler described by Sameoto and Jaroszynski (1969) revealed serious sampling problems associated with coastal waters of British Columbia. Due to extensive freshwater runoff in the vicinity of large rivers, sampling con- ditions including choppy surface waters of lowered salinity and vertically depressed distributions of near-surface larval and juvenile fishes. Under such conditions, the S-J sampler behaved erratically, throwing considerable spray, and, when adjusted to increase depth of tow, the body and control surfaces deformed at speeds in excess of 5 knots. The modifi- cations described here reflect our objectives of im- proving performance, increasing durability, and ease of handling, without increasing costs other than those incurred by adding a flowmeter to provide quantitative catches. The complete unit is depicted in Figure 1. Detailed Description Sampler Box Constructed of 1/8" marine aluminum, this alu- minum is folded into a body with one welded seam (Fig. 2). The leading edges are reinforced with 1/4" aluminum for attaching the bridles and depressor. The square mouth opening was sized to accomo- date 0.25 m2 bongo nets having a circumference of 185 cm. Body dimensions are 46 x 46 x 60 cm. 480 FISHERY BULLETIN: VOL. 84, NO. 2, 1986. PERSPECTIVE - NEUSTON SAMPLER swivel Figure 1.— Neuston sampler with net cod-end attached. Net Attachment We replaced the grommet and bolt-through net fastening system of the S-J sampler with an alu- minum channel clamp (Fig. 2). Net slippage is pre- vented by sewing a 1/4" rope into the net collar. Stainless steel bolts remain permanently attached to the sampler body so that, to mount or replace the net, it is merely slid over the box and the channel placed over the bolts and secured. One man can replace the net in 5 min. Lateral Wings Individual fins bolt directly to the sides of the body and are made of 1/8" aluminum with the inside edge bent at 90° for an attachment face (Fig. 3). The outer edge is bent downward 15° to stiffen it and to reduce side slippage under tow. The wings pivot on a bolt anteriorly and are adjusted through a series of holes in the sampler body (Figs. 1, 2). Depressor Bolted directly to the body and adjusted as for the wings (Figs. 1, 2), the depressor is made from 1/4" marine aluminum bent at right angles on either end for attachment (Fig. 3). It serves also as the lower towing point and stiffens the body. Tow Points The sampler is adjusted in relation to the towing vessel by a stainless steel turnbuckle on the upper bridle (roll aspect), and by selecting the lower tow point (depressor) and upper tow point (leading top corner of the body) from a series of holes (Figs. 1, 2, 3). The tow point fastening is a threaded U- bolt, fastened on both sides of the sampler frame (Fig. 3). Flotation A streamlined float constructed of fiberglassed, polyurethane foam which bolts to the upper face of the body (Figs. 1, 2). At neutral buoyancy the sampler floats with the mouth opening just below the water surface. As with the S-J sampler, vertical positioning under tow is the balanced outcome of downward depressor force and lift from the lateral fins. These adjustments are made to maintain an 8-10 cm headspace of air in the sampler while under tow. 481 towing points 0 towing points tow direction 51cm -• p— 47 cm - — h / PLAN VIEW FLOAT 6.3mm holes 10 cm centres float bolts to top of box 63mm x 25mm stainless bolts SIDE ELEVATION 6.3 x 25mm stainless bolt 1.2 x 3cm alum, channel net collar with rope sewn in 6.3mm rope stainless nut I5cm 47 cm 51 cm NET ATTACHMENT DETAIL L.SIDE ELEVATION FLOAT to accept standard 1/4 m2 (184 cm circum) plankton net EF 46cm -46cm B- 15x5 x 0.5cm reinforcing pads welded both sides 9.5mm x 25mm stainless bolt 5cm — ► wing depressor ^ 32mm marine alum, folded a welded at one corner. 9.5mm x 38mm stainless bolt 0 6.3mm holes 19mm o.c. 09.5 mm holes 61cm 2.5cm 06.3mm holes BOX - FRONT ELEVATION R=IOcm- BOX- L.SIDE ELEVATION Figure 2.— Scale drawings of the sampler body and float, and net attachment detail. Flowmeter A General Oceanics meter is attached inside the body by means of a hinged strut which folds forward to facilitate reading the meter (Figs. 1, 3). The meter is free-pivoting in the horizontal plane and offset 17 cm from the center of the mouth opening. Evaluation This modified version of the otter neuston sampler has been used extensively since 1981, offshore to Station Papa (Mason et al. 1983) and in inside waters under all weather conditions, including a full gale. It performs best when towed into or across the wave direction at 4-6 knots. At higher speeds, disturbance due to backsplash from the fins and bridal may cancel out potential advantage of further increase in tow speed. Sampling efficiency is deemed to be relatively high when using a 500 pjm mesh net at night. Catches of juvenile fishes in the Strait of Georgia are quantitatively comparable with those 482 51cm 5cm 15cm I— 09.5mm holes 10cm -—5cm 50cm L WING -SIDE ELEVATION strut bend at 15° from horiz. -30° WING -FRONT ELEVATION strut 2.5cm wide 7 x 3.2mm thick / weld bend h— 10cm f— 10cm 45cm 45.75cm heat bene 90* ,' 6.3mm marine alum. RHOcm- 2.5cm ~*l\ 5cm -J 0 9.5mm holes 4cm o.c. 0 6.3mm holes ^_ 19mm o.c. L.WING PLAN VIEW DEPRESSOR- FRONT ELEVATION SIDE ELEVATION 0 9.5mm threaded brass rod lock nut 012mm alum, pipe Top half hinge threaded to accept brass rod Hole cut in lower half of hinge and in box to accomodate lock nut. 0 6.3mm holes 6.3 mm U bolt DETAIL OF TOW POINTS J Leading edge, bottom of sampler stainless flathead bolts c/w nuts tow bridle attaches here FLOWMETER ATTACHMENT DETAIL Figure 3.— Scale drawings of the depressor and wings, and tow point and flowmeter details. made with a large volume, two-boat surface trawl as employed by Barraclough et al. (1966). We found no significant difference (student's £-test) between mean total catch (12.9 and 12.1 fish/100 m3) for nine taxa common to both gears in eight pairs of tows made locally in the Strait of Georgia, British Columbia, during March-April. Among the fish sam- pled by this gear in offshore and shelf waters are juvenile Pacific salmon to 14 cm, Pacific saury to 25 cm, juvenile sablefish, rockfish, greenlings, and squid, in addition to the routine catches of ichthyo- plankton and general zooplankton. 483 Literature Cited Baraclough, W. E. 1967. Number, size and food of larval and juvenile fish caught with a two-boat surface trawl in the Strait of Georgia, April 25-29, 1966. Fish. Res. Board Can., Biol. Stn. Nanaimo, B.C., Manuscr. Rep. Ser. 922, 54 p. Mason, J. C, R. J. Beamish, and G. A. McFarlane. 1983. Sexual maturity, fecundity, spawning, and early life history of sablefish (Anoplopoma fimbria) off the Pacific coast of Canada. Can. J. Fish. Aquat. Sci. 40:2126-2134. Sameoto, D. D., and L. 0. Jaroszynski. 1969. Otter surface sampler: a new neuston net. J. Fish. Res. Board Can. 26:2240-2244. J. C. Mason A. C. Phillips Pacific Biological Station Fisheries Research Branch Department of Fisheries and Oceans Nanaimo, British Columbia V9R 5K6, Canada MORPHOLOGICAL EVIDENCE FOR STARVATION AND PREY SIZE SELECTION OF SEA-CAUGHT LARVAL SABLEFISH, ANOPLOPOMA FIMBRIA One of the major causes of larval mortality is star- vation, this being related to the patchiness of food resources (Hunter 1981). While starvation has been induced under laboratory conditions [e.g., herring, Clupea harengus, and plaice, Pleuronectes platessa (Ehrlich et al. 1976); northern anchovy, Engraulis mordax (O'Connell 1976); jack mackerel, Trachurus symmetricus (Theilacker 1978, 1981)], starved lar- vae have rarely been observed in nature (northern anchovy, O'Connell 1980; jack mackerel, Theilacker 1986). Various methods have been used to charac- terize starvation in fish larvae, including condition factor (Blaxter 1971), chemical analyses (Ehrlich 1974), histological analyses (Umeda and Ochiai 1975; O'Connell 1976, 1980; Theilacker 1978, 1986), and morphological analyses (Shelbourne 1957; Nakai et al. 1969; Ehrlich etal. 1976; Theilacker 1978, 1981, 1986). While histological and chemical analyses are based on qualitative changes in tissues that result from starvation, their methodologies require special preservation techniques, negating their application to samples preserved without these techniques in mind. To characterize starvation in samples that have not been specially preserved, measures of mor- phology and/or condition factor are more appropri- ately applied. In the present study, in the absence of special preservation techniques, the occurrence of starvation in sea-caught larval sablefish, Anoplo- poma fimbria, was examined using morphological measures. The sablefish inhabits the continental shelf of the North Pacific Ocean and is the subject of an inten- sifying fishery off the west coast of North America, yet little is known about the early life history of the species. Recent evidence obtained off Canada sug- gests that sablefish spawn in water deeper than 300 m, with spawning activity peaking in February. Eggs (1.8-2.2 mm in diameter) descend while devel- oping, and hatching probably occurs at depths in ex- cess of 400 m (Mason et al. 1983). Although size at hatching and the size at first feeding have not been clearly defined, Mason et al. (1983) reported collect- ing recently hatched larvae of 5-6 mm. After hatch- ing, larvae ascend to surface waters and become neustonic (Kendall and Clark 19821). Juveniles ap- parently remain in shallow water until they mature. Beyond reports of distribution (Kendall and Clark fn. 1; Clark 19842) and descriptive work (e.g., Koba- yashi 1957; Ahlstrom and Stevens 1976), studies of larval and early juvenile sablefish have concentrated on aging and growth (Boehlert and Yoklavich 1985; Shenker and Olla in press). Our aim in the present study was to detect the possible occurrence of starvation in larval sablefish collected off Washington and Oregon during April and May 1980 (Kendall and Clark fn. 1), using selected morphological measurements to determine variability in larval condition. Further, to elucidate the possible relationship between larval condition and feeding requirements, prey size-selection and diet were analyzed. Methods Sablefish larvae were collected by using a 0.5 m neuston net (Sameoto and Jaroszynski 1969) with 0.505 mm mesh, towed for 10 min from the RV Tikhookaenskiy , during the first cooperative U.S.- U.S.S.R. ichthyoplankton survey off the Washing- ton and Oregon coast in 1980 (Kendall and Clark fn. 1). Larvae from stations 20, 24, 25, 34, 38, 50, Kendall, A. W., and J. Clark. 1982. Ichthyoplankton off Washington, Oregon and Northern California, April-May 1980. Processed Rep. 82-11, 44 p. Northwest and Alaska Fisheries Center, National Marine Fisheries Service, NOAA, Seattle, WA 98112. 2Clark, J. B. 1984. Ichthyoplankton off Washington, Oregon and Northern California, May-June 1981. Processed Rep. 84-11, 46 p. Northwest and Alaska Fisheries Center, National Marine Fisheries Service, NOAA, Seattle, WA 98112. 484 FISHERY BULLETIN: VOL. 84, NO. 2, 1986. 54, 70, and 71 (Fig. 1), collected between 22 April and 4 May 1980, formed the basis for this study. All larvae were preserved in 10% Formalin3 at sea. After sorting, larvae were switched into 5% For- malin, where they remained until their examination in 1983. The following body measurements were recorded: standard length (SL), head length (HL), eye diameter (ED), body depth at pectoral (BD.P), and body depth at anus (BD.A) (after Theilacker 1981). Standard length was measured to the nearest 0.1 mm. All other measurements were made to the nearest 0.05 mm using an ocular micrometer. Because body proportions change dramatically with size of larvae, it was necessary to restrict any com- parisons to samples which were not statistically dif- ferent in terms of the distribution of SL values. Also, to minimize ambiguities attributable to slight dif- ferences in size, comparisons of body measurements were made using a ratio of the body measurement to SL (e.g., HL/SL) as well as the absolute measure- ment (mm). Because a number of larvae were dam- aged prior to the time measurements were made (e.g., eyes were missing, the gut was separated from 6Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. the body) the sample size (n) varied within a station. To classify larval condition, statistical comparisons of the body measurements were made using the Mann- Whitney test (Zar 1974), a nonparamotric rank procedure. Food particle-size selection was examined by measuring the widths of prey items ingested by 84 larvae from stations 24, 34, 50, 70, and 71. Soft- bodied prey items were not measured due to the dif- ficulty in accurately assessing their effective width. All measurements were made using an ocular micrometer at 40 x. Prey widths were originally plotted for five size classes of larvae: 8.2-12.5, 12.6-16.5, 16.6-20.5, 20.6-24.5, and 24.6-28.5 mm SL. The prey-size selection curve of larvae 12.6-16.5 mm closely approximated the curve of larvae 16.6-20.5 mm, and so these size classes were com- bined. Similarly, the curves of larvae 20.6-24.5 mm and 24.6-28.5 mm were essentially superimposed one upon the other, and as a result these size classes were also combined. This yielded three functional sablefish size classes for particle-size analysis: 8.2-12.5, 12.6-20.5, and 20.6-28.5 mm SL. The incidence of empty guts was recorded, and diet was analyzed in terms of numerical percent composition and frequency of occurrence of copepod nauplii. I30°W -48°N Figure 1.— Map of the Washington and Oregon coast where lar- val sablefish were collected in 1980. Results Morphological Measurements Out of a total of 56 larvae collected at station 25, 48% (27 larvae) appeared emaciated, in marked con- trast to larvae collected at all other stations. This emaciated condition, which we interpreted as evi- dence of starvation, was present in 82% of the lar- vae <12.5 mm SL (27 out of 32) collected at this sta- tion but was absent in fish larger than 12.5 mm SL. To test whether this interpretation, which was based on a gross visual examination of these larvae, was statistically verifiable, the morphology of the ema- ciated larvae from station 25 was compared with lar- vae of the same size from stations 20, 24, 34, 38, 50, 54, 70, and 71. The size range, 8.2-12.5 mm SL, was selected as the broadest range over which the distributions of SL values of these two groups were equivalent, and excluded the two smallest larvae col- lected at station 25 from the comparisons. Signifi- cant differences were observed in seven of eight body measurements, indicating that distinct dif- ferences were present in the larvae from station 25 when compared with larvae of similar size from all other stations (Table 1). 485 Table 1 .—A comparison of median values of body measurements of Anoplopoma fim- bria larvae from station 25 with larvae from stations 20, 24, 34, 38, 50, 54, 70, and 71. The size range was 8.2-12.5 mm SL. Stations 20, 24, 34, 38, 50, 54, Station 25 70, and 71 P1 Standard length, SL (mm) 10.0 10.25 >0.20 95% Confidence Interval, C.I. (9.2-10.4) (10.0-10.5) n2 = 25 118 Head length, HL (mm) 1.5 2.1 <0.001 95% C.I. (1.4-1.8) (2.0-2.1) HL/SL 0.165 0.200 <0.001 95% C.I. (0.147-0.177) (0.194-0.206) n = 25 114 Eye diameter, ED (mm) 0.7 0.85 <0.001 95% C.I. (0.6-0.7) (0.8-0.9) ED/SL 0.069 0.082 <0.001 95% C.I. (0.066-0.073) (0.081-0.084) n = 23 113 Body depth at pectoral, BD.P (mm) 1.0 1.3 <0.001 95% C.I. (0.9-1.15) (1.3-1.3) BD.P/SL 0.109 0.128 <0.001 95% C.I. (0.098-0.118) (0.125-0.131) n » 13 100 Body depth at anus, BD.A (mm) 1.0 1.2 <0.005 95% C.I. (0.7-1.2) (1.15-1.25) BD.A/SL 0.104 0.116 >0.10 95% C.I. (0.082-0.139) (0.112-0.118) n » 11 84 1P = probability that body measurements of station 25 larvae were equivalent with larvae from stations 20, 24, 34, 38, 50, 54, 70, and 71, as determined by the Mann-Whitney test. 2The sample size was not constant within each group because some larvae were damaged prior to the time measurements were made (e.g., some had lost eyes, the gut was separated from the body). Analysis of Gut Contents Examination of the gut contents of larvae <12.5 mm SL provided further evidence as to the starved condition of the larvae at station 25. At this station 75% of the larvae (24 out of 32) had no food in their guts, and 9% (3 larvae) had ingested 2 or fewer prey items. In addition to being empty, the guts of lar- vae collected at station 25 were shrunken, which is reflective of poor feeding conditions (Nakai et al. 1969). At all other stations the incidence of empty guts for larvae <12.5 mm SL was <1%, as was the incidence of larvae ingesting 2 or fewer prey items. Circumstantial evidence as to the cause of star- vation comes from food analyses. It was apparent that while sablefish larvae selected increasingly larger prey as they grew larger, the minimum size of prey eaten did not increase appreciably. By ex- amining the widths of all prey items ingested by larvae of different lengths (Fig. 2), three general patterns emerged: 1) Larvae 8.2-12.5 mm SL prin- cipally ingested the narrowest prey (0.01-0.10 mm in width), 2) larvae 12.6-20.5 mm SL ingested slightly larger prey (0.11-0.20 mm in width), and 3) sablefish 20.6-28.5 mm SL primarily ingested the largest prey (0.21-0.30 mm in width), although they also ingested a broad range of prey sizes. Copepod nauplii were the dominant small prey, and were all <0.20 mm wide. They accounted for 88.3% of the diet (by number) of small larvae (<12.5 mm). Based on prey-size selection alone (Fig. 2), it appears that copepod nauplii may have also contrib- uted substantially to the diet of larvae 12.6-20.5 mm SL, but not to the diet of fish 20.6-28.5 mm SL. Dietary analysis confirmed this, with nauplii com- prising 26.9% of the diet of larvae 12.6-20.5 mm, but merely 1.4% of the diet offish 20.6-28.5 mm SL. Considering the relative importance of copepod nauplii in the diet of larvae 12.6-20.5 mm SL and the fact that this size class continued to ingest nauplii although capable of ingesting larger prey, the frequency of occurrence of copepod nauplii in the guts of these larvae was examined at each sta- tion as inferential evidence of the abundance or availability of copepod nauplii (Table 2). At station 25 only 27% of larvae 12.6-20.5 mm SL ingested nauplii compared with 60-100% at all other stations; the low frequency of occurrence of nauplii in guts of these larvae at station 25 was obtained even though no guts were empty. These data indicate that copepod nauplii may not have been abundant or 486 bT o 50 \ • A ^•-^8.2 -12.5 mm SL ° — ° 12.6-20.5 mm SL o t \ • •20.6-28.5 mm SL t 40 CO o Q_ O 30 O - i 1 1 ~ 1 'V A \ \ \ \ \ o PERCENT o o 1 -1 1 1 > / / \ \ \ • \ • \ \\ \\ i V~- \ _ T-mQ, 0< 8 A i i 0 .01- 0.1 1- 0.21- 0.31- 0.41- 0.51- 0.61- 0.71- 0.81- 0.91 0 .10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 PREY WIDTH (mm) Figure 2.— The size of prey selected by larval sablefish, plotted for three size classes of larvae: 8.2-12.5 mm SL (n = 43), 12.6-20.5 mm SL (n = 25), and 20.6-28.5 mm SL (n = 16). Table 2. — Frequency of occurrence of copepod nauplii found in the guts of larval sable- fish, by size class and station. Station number Size class 20 24 25 34 38 50 54 70 71 <12.5 mm 12.6-20.5 mm 38/38 1 00% 6/10 60% 6/6 100% 2/2 100% !7/32 22% 4/15 27% 9/9 32/32 6/6 100% 100% 100% 8/8 11/11 1/1 100% 100% 100% 15/16 94% 3/3 100% 10/10 100% 10/11 91% 11/12 92% 2/3 67% 1This includes 24 larvae <12.5 mm SL with empty guts. readily available at station 25, with the high in- cidence of starvation at this station suggesting a cause-and-effect relationship between these two factors. Discussion There is no definitive way of discerning whether the sablefish larvae that we categorized as starv- ing had starved to the "point of no return". To ascertain whether sea-caught larvae have starved beyond recovery requires rearing larvae from eggs in the laboratory under different feeding regimes, and using these as standards of comparison for sea- caught specimens. Unfortunately, this has been done in only a few cases. For example, O'Connell (1976) established histological criteria for starvation under laboratory conditions for the northern anchovy. These criteria were then employed to identify starv- ing larvae collected in the Southern California Bight (O'Connell 1980). The proportion of starving larvae was estimated to be 8%, for larvae <7.5 mm SL, with this representing 40% of the daily rate of mor- tality. In a more recent and comprehensive study, Theilacker (1986) utilized both histological and mor- phological criteria (Theilacker 1978, 1981) to ex- amine starvation of sea-caught first-feeding jack mackerel in the Southern California Bight. She determined that starvation varied with habitat. In the open ocean, the number of larvae <3.5 mm dying 487 of starvation per day was 57-70%, whereas only 6-12% of the first-feeding larvae collected near islands and banks were starving. Until techniques are developed for rearing sable- fish from eggs, we are limited to utilizing com- parisons of sea-caught larvae to infer the importance of starvation in the early life history of this species. While starving larvae were observed at only one sta- tion, our finding confirms that sablefish larvae do encounter suboptimal environmental conditions in the sea. However, neither the transience nor geo- graphic extent of this phenomenon can be assessed in the absence of an intensive sampling scheme designed specifically to answer these questions. Although definitive plankton composition data are lacking, the occurrence of starving larvae at station 25 appears to reflect a paucity of copepod nauplii. While appropriate prey concentrations (Laurence 1974; Lasker 1975; Houde 1978), particle size (Lasker 1975; Hunter 1981), and prey species com- position (Lasker 1975; Scura and Jerde 1977) all relate to the survival and growth of marine fish lar- vae, not all larvae are able to maintain associations with suitable prey patches. Lasker (1975) empha- sized the transient nature of optimal feeding condi- tions in the sea, noting that northern anchovy lar- vae which had been associated with a good feeding patch (a bloom of Gymnodinium splendens that per- sisted for 18 d) would probably die of starvation after a wind storm broke up the bloom. Patchiness of food resources has also been suggested by the station-to-station variability in growth rates of northern anchovy (as determined from daily incre- ments of otoliths) (Methot and Kramer 1979). Similarly, after monitoring larval development in both good and bad plankton patches, Shelbourne (1957) reported that a scarcity of appropriate food resulted in a deterioration of the physical condition of plaice larvae. Where morphological measurements of larvae are concerned, changes in body measurements which result from handling and preservation techniques must be considered. Net abrasion results in mech- anical damage to the larvae (Blaxter 1971) as well as shrinkage (Blaxter 1971; Theilacker 1980), with the amount of shrinkage depending on whether death preceded fixation (Blaxter 1971), and the ex- tent of handling (Theilacker 1980). The type of fix- ative used (Theilacker 1980), its concentration, salinity, and temperature (Hay 1982) also affect the degree of shrinkage. In the present case, shrinkage most likely occurred during the 3 yr these larvae were held in Formalin. However, absolute lengths may not be critical to evaluating the significance of our findings, and the differences that were seen be- tween stations could not have resulted simply from differences in shrinkage. This was clear from the qualitative differences in gut appearance seen between stations (i.e., shrunken and empty guts ver- sus guts filled to distention). Further, since the sablefish larvae we examined were all caught and preserved during the same cruise, we assumed that whatever shrinkage that may have resulted from handling and preservation techniques is constant throughout the samples. Larval fishes are limited in the prey that they con- sume by their ability to capture and process it. As they grow, larvae become very successful predators, caused in part by an increase in mouth size. As a result, the size of prey selected increases as devel- opment proceeds. Prey width was used to examine prey-size selection because prey width appears to be the critical dimension for the successful inges- tion of oblong prey by larval fishes (Blaxter 1965; Arthur 1976; Hunter 1981). For sablefish, definitive shifts in the size of prey consumed occurred at about 12.5 and 20.5 mm SL. The diet of the larger larvae was more diverse than the diet of small larvae. This expansion of the range of prey selected is not un- common (e.g., Hunter 1981) and is adaptive inas- much as it enables larvae to ingest suboptimal prey items at times when optimal or preferred prey are not available. Smaller fish appear limited in the size of prey they can exploit. This limitation, combined with larvae <12.5 mm SL being associated with an unsuitable prey patch at station 25, may have been responsible for the high incidence of empty guts and starvation. Acknowledgments We wish to thank Kevin Bailey, George Boehlert, and two anonymous reviewers for their comments on drafts of this manuscript. Thanks also to Michael Davis and Steve Ferraro for their advice on statis- tical analyses, and Art Kendall for valuable discus- sions and continual encouragement. This work was supported by the Northwest and Alaska Fisheries Center, National Marine Fisheries Service, NOAA Contract No. 83-ABC-00045. Literature Cited Ahlstrom, E. H., and E. Stevens. 1976. Report of neuston (surface) collections made on an ex- tended CalCOFI cruise during May 1972. Calif. Coop. Oceanic Fish. Invest. Rep. 18:167-180. Arthur, D. K. 1976. Food and feeding of larvae of three fishes occurring 488 in the California Current, Sardinops sagax, Engraulis mor- dax, and Trachurus symmetricus. Fish. Bull., U.S. 74: 517-530. Blaxter, J. H. S. 1965. The feeding of herring larvae and their ecology in rela- tion to feeding. Calif. Coop. Oceanic Fish. Invest. Rep. 10: 79-88. 1971. Feeding and condition of clyde herring larvae. Rapp. P.-v. Reun. Cons. int. Explor. Mer 160:128-136. BOEHLERT, G. W., AND M. M. YOKLAVICH. 1985. Larval and juvenile growth of sablefish, Anoplopoma fimbria, as determined from otolith increments. Fish. Bull., U.S. 83:475-481. Ehrlich, K. F. 1974. Chemical changes during growth and starvation of her- ring larvae. In J. H. S. Blaxter (editor), The early life history of fish, p. 301-323. Springer- Verlag, Berlin. Ehrlich, K. F., J. H. S. Blaxter, and R. Pemberton. 1976. Morphological and histological changes during the growth and starvation of herring and plaice larvae. Mar. Biol. (Berl.) 35:105-118. Hay, D. E. 1982. Fixation shrinkage of herring larvae: effects of salin- ity, formalin concentration, and other factors. Can. J. Fish. Aquat. Sci. 39:1138-1143. Houde, E. D. 1978. Critical food concentrations for larvae of three species of subtropical marine fishes. Bull. Mar. Sci. 28:395- 411. Hunter, J. R. 1981. Feeding ecology and predation of marine fish larvae. In R. Lasker (editor), Marine fish larvae: morphology, ecology, and relation to fisheries, p. 33-77. Sea Grant Pro- gram, Univ. Wash. Press, Seattle. KOBAYASHI, K. 1957. Larvae and youngs of the sablefish, Anoplopoma fim- bria (Pallas), from the sea near the Aleutian Islands. [In Jpn., Engl, abstr.] Bull. Jpn. Soc. Sci. Fish. 23:376-382. Lasker, R. 1975. Field criteria for survival of anchovy larvae: The rela- tion between inshore chlorophyll maximum layers and suc- cessful first feeding. Fish. Bull., U.S. 73:453-462. Laurence, G. C. 1974. Growth and survival of haddock Melanogrammus aegle- finus larvae in relation to planktonic prey concentration. J. Fjsh. Res. Board Can. 31:1415-1419. Mason, J. C, R. J. Beamish, and G. A. McFarlane. 1983. Sexual maturity, fecundity, spawning, and early life history of sablefish (Anoplopoma fimbria) off the Pacific coast of Canada. Can. J. Fish. Aquat. Sci. 40:2126- 2134. Methot, R. D., Jr., and D. Kramer. 1979. Growth of northern anchovy, Engraulis mordax, lar- vae in the sea. Fish. Bull., U.S. 77:413-423. Nakai, Z., M. Kosaka, M. Ogura, G. Hayashida, and H. Shimo- zono. 1969. Feeding habit, and depth of body and diameter of digestive tract of shirasu, in relation with nutritious condi- tion. [In Jpn., Engl, abstr.] J. Coll. Mar. Sci. Technol., Tokai Univ. 3:23-34. O'Connell, C. P. 1976. Histological criteria for diagnosing the starving con- dition in early post yolk sac larvae of the northern anchovy, Engraulis mordax Girard. J. Exp. Mar. Biol. Ecol. 25: 285-312. 1980. Percentage of starving northern anchovy, Engraulis mordax, larvae in the sea as estimated by histological methods. Fish. Bull., U.S. 78:475-489. Sameoto, D. D., and L. O. Jaroszynski. 1969. Otter surface sampler: a new neuston net. J. Fish. Res. Board Can. 26:2240-2244. Scura, E. D., and C. W. Jerde. 1977. Various species of phytoplankton as food for larval northern anchovy, Engraulis mordax, and relative nutri- tional value of the dinoflagellates Gymnodinium splendens and Gonyaulax polyedra. Fish. Bull., U.S. 75:577-583. Shelbourne, J. E. 1957. The feeding and condition of plaice larvae in good and bad plankton patches. J. Mar. Biol. Assoc. U.K. 36:539-552. Shenker, J. M., and B. L. Olla. In press. Laboratory feeding and growth of early juvenile sablefish, Anoplopoma fimbria. Can. J. Fish. Aquat. Sci. Theilacker, G. H. 1978. Effect of starvation on the histological and morpho- logical characteristics of jack mackerel, Trachurus sym- metricus, larvae. Fish. Bull., U.S. 76:403-414. 1980. Changes in body measurements of larval northern an- chovy, Engraulis mordax, and other fishes due to handling and preservation. Fish. Bull., U.S. 78:685-692. 1981. Effect of feeding history and egg size on the morphol- ogy of jack mackerel, Trachurus symmetricus, larvae. Rapp. P.-v. Reun. Cons. int. Explor. Mer 178:432-440. 1986. Starvation-induced mortality of young sea-caught jack mackerel, Trachurus symmetricus, determined with histological and morphological methods. Fish. Bull., U.S. 84:1-17. Umeda, S., and A. Ochiai. 1975. On the histological structure and function of digestive organs of the fed and starved larvae of the yellowtail, Seriola quinqueradiata. [In Jpn., Engl, abstr.] Jpn. J. Ichthyol. 21:213-219. Zar, J. H. 1974. Biostatistical analysis. Prentice-Hall, Englewood Cliffs, NJ, 620 p. Jill J. Grover College of Oceanography Oregon State University Hatfield Marine Science Center Newport, OR 97365 Bori L. Olla Cooperative Institute for Marine Resources Studies Northwest and Alaska Fisheries Center National Marine Fisheries Service, NOAA Hatfield Marine Science Center Newport, OR 97365 489 NOTICES NOAA Technical Reports NMFS published during last 6 months of 1985 Technical Report NMFS 31. Shark catches from selected fisheries off the U.S. east coast. July 1985, iii + 22 p. Analysis of various sources of pelagic shark catches in the Northwest and Western Central Atlantic Ocean and Gulf of Mexico with com- ments of other large pelagics. By Emory D. Anderson, p. 1-14, 3 figs., 17 tables. Estimated catches of large sharks by U.S. recreational fishermen in the Atlantic and Gulf of Mexico. By John G. Casey and John J. Hoey p. 15-19, 5 tables. The incidental capture of sharks in the Atlantic United States Fishery Conservation Zone reported by the Japanese tuna longline fleet. By W. N. Witzell, p. 21-22, 3 tables. 32 . Nutrient distributions for Georges Bank and adjacent waters in 1979. By A. F. J. Draxler, A. Matte, R. Waldhauer, and J. E. O'Reilly. July 1985, iii + 34 p., 32 figs., 2 tables. 33. Marine flora and fauna of the Northeastern United States. Echinoder- mata: Echinoidea. By D. Keith Serafy and F. Julian Fell. September 1985, iii + 25 p., 42 figs. 34. Additions to a revision of the shark genus Carcharhinus: synonymy of Aprionodon and Hypoprion, and description of a new species of Car- charhinus (Carcharhinidae). By J. A. F. Garrick. November 1985, iii + 26 p., 14 figs., 4 tables. 35. Synoptic review of the literature on the Southern Oyster Drill Thais haemastomafloridana. By Philip A. Butler. November 1985, iii + 9 p. 36. An egg production method for estimating spawning biomass of pelagic fish: application to the northern anchovy, Engraulis mordax. Reuben Lasker (editor). December 1985, iii + 99 p. Introduction: an egg production method for anchovy biomass assess- ment. By Reuben Lasker, p. 1-3, 1 fig. Biomass model for the egg production method. By Keith Parker, p. 5-6. Parameter estimation for an egg production method of northern an- chovy biomass assessment. By Susan Picquelle and Gary Stauffer, p. 7-15, 8 figs., 6 tables. Sea survey design and analysis for an egg production method of an- chovy biomass assessment. By Paul E. Smith and Roger P. Hewitt, p. 17-26, 4 figs., 8 tables. The CalCOFI vertical egg tow (CalVET) net. By Paul E. Smith, William Flerx, and Roger P. Hewitt, p. 27-32, 5 figs., 1 table. Procedures for sorting, staging, and ageing eggs. By Gary Stauffer and Susan Picquelle, p. 33-35. Staging anchovy eggs. By H. Geoffrey Moser and Elbert H. Ahlstrom, p. 37-41, 2 figs. A model for temperature-dependent northern anchovy egg develop- ment and an automated procedure for the assignment of age to staged eggs. By Nancy C. H. Lo, p. 43-50, 2 figs., 6 tables. A protocol for designing a sea survey for anchovy biomass assess- ment. By Robert P. Hewitt, p. 51-53. Sampling requirements for the adult fish survey. By Susan Picquelle, p. 55-57, 1 fig. Spawning frequency of Peruvian anchovies taken with a purse seine. By Jurgen Alheit, p. 59-61, 1 table. Preservation of northern anchovy in formaldehyde solution. By J. Roe Hunter, p. 63-65, 1 fig., 1 table. 490 Batch fecundity in multiple spawning fishes. By J. Roe Hunter, Nancy C. H. Lo, and Roderick J. H. Leong, p. 67-77, 6 figs., 5 tables. Measurement of spawning frequency in multiple spawning fishes. By J. Roe Hunter and Beverly J. Macewicz, p. 79-94, 7 figs., 1 table. Comparison between egg production method and larval census method for fish biomass assessment. By Roger P. Hewitt, p. 95-99, 2 figs., 1 table. Some NOAA publications are available by purchase from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. 491 > 4 S3 - $6 8 «*•-©**■■ INFORMATION FOR CONTRIBUTOR TO THE FISHERY BULLETIN Manuscripts submitted to the Fishery Bulletin will reach print These are not absolute requirements, of course, but desiderata. if they conform to the following instructions. CONTENT OF MANUSCRIPT The title page should give only the title of the paper, the author's name, his affiliation, and mailing address, in- cluding ZIP code The abstract should not exceed one double-spaced page In the text, Fishery Bulletin style for the most part, follows that of the U.S. Government Printing Office Style Manual. Fish names follow the style of the American Fisheries Society Special Publication No. 12, A List of Com- mon and Scientific Names of Fishes from the. United States and Canada, Fourth Edition, 1980. Text foowotes should be typed separately from the text. Figures and tables, with their legends and headings, should be self-explanatory, not requiring reference to the text. Their placement should be indicated in the right-hand margin of the manuscript. Preferably figures should be reduced by photography to 5% inches (for single-column figures, allowing for 50% reduction in printing), or to 12 inches (for double-column figures). The maximum height, for either width, is 14 inches. Photographs should be printed on high quality or glossy paper. Do not send original drawings to the Scientific Editor; if they, rather than the photographic reductions, are ne> by the printer, the Scientific Publications Office will re< . them. Each table should start on a separate page Consistency in headings and format is desirable Vertical rules should be avoided, as they make the tables more expensive to print. Footnotes in tables should be numbered sequentially in arabic numerals. To avoid confusion with powers, they should be placed to the left of numerals. Acknowledgments, if included, are placed at the end of the text. Literature is cited in the text as: Lynn and Reid (1968) or (Lynn and Reid 1968). All papers referred to in the text should be listed alphabetically by the senior author's sur- name under the heading "Literature Cited." Only the author's surname and initials are required in the literature cited. The accuracy of the literature cited is the respon- sibility of th». Abbreviations of names of periodicals and s form to Serial Sources for the BIOSIS Data 1 stracts also uses this system, which was developed I . ner Standards Association.) Common abkrevi* n 1 symbols, such as mm, m, g, L, mL, m, .). %, °/00, and so forth, should be used. measure only when used with numeni rarely used with abbrevia- tions. We prefer that measurements be given in metric units; other equivalent uni4j may be given in parentheses. FORM OF THE MANUSCRIPT The original of the manuscript should be typed, double- spaced, on (Vhite bond paper. Please triple space above headings. We would rather receive good duplicated copies of manuscripts than carbon copies. The sequence of the mat >uld be: TITLE PAGE ABSTRACT TEXT LITERATURE CITED TEXT FOOTNOTES APPENDIX TABLES (Each table should be numbered with an arabic numeral and heading provided). LIST OF FIGURES (Entire figure legends) FIGURES (Each figure should be numbered with an arabic numeral; legends are desired) ADDITIONAL INFORMATION Send the ribbon copy and two duplicated or carbon copies of the manuscript to: Dr. William J. Richards, Scientific Editor Fishery Bulletin Southeast Fisheries Center Miami Laboratory National Marine Fisheries Service, NOAA 75 Virginia Beach Drive Miami, FL 33149-1099 Fifty separates will be supplied to an author free of charge and 50 supplied to his organization. No covers will be supplied. Content LO, NANCY C. H. Modeling life-stag itaneous mortality rates, an application to northern anchovy, Engt • eggs and larvae 395 SEN, A. R. Methodological problem? ommercial rockfish landings . . . 409 POLOVINA, JEFFREY J. A varial v version of the Leslie model with application to an intensive fishi; i a multispecies stock 423 MATSUURA, YASUNOBU, and TAKUMI YONEDA. Early development of the lophid anglerfish, Lophv; ophysus 429 SQUIRES, DALE. Ex-vessel pric in the New England fishing industry .. 437 LEBER, KENNETH M., and I I ( & GREENING. Community studies in seagrass meadows: A comparison of t or sampling macroinvertebrates and fishes 443 Notes OXENFORD, HAZEL A. AYNE HUNTE. A preliminary investigation of the stock structure of the dolphin, Coryphaena hippurus, in the western central Atlantic 451 FORWARD, RICHARD B., JR., BLANCA ROJAS de MENDIOLA, and RICHARD T. BARBER. Effects of temperature on swimming speed of the dinoflagellate Gym- nodinium splendens 460 GRAHAM, JEFFREY B., RICHARD H. ROSENBLATT, and DARCY L. GIBSON. Morphology and possible swimming mode of a yellowfin tuna, Thunnus aWacares, lack- ing one pectoral fin 463 RATTY, F. J., Y. C. SONG, and R. M. LAURS. Chromosomal analysis of albacore, Thun- nus alalunga, and yellowfin, Thunnus albacares, and skipjack, Katsuwonus pelamis, tuna 469 STIER, KATHLEEN, and BOYD KYNARD. Abundance, size, and sex ratio of adult sea-run sea lamprey, Petromyzon marinus, in the Connecticut River 476 MASON, J. C, and A. C. PHILLIPS. An improved otter surface sampler 480 GROVER, JILL J., and BORI L. OLLA. Morphological evidence for starvation and prey size selection of sea-caught larval sablefish, Anoplopoma fimbria 484 Notices 490 • GPO 593-096 MBL WHOI LIBRARY UH nwB G J? ^TO'ca C \ / —__--_-_-__-«-_-_-—-——_ Vol. 84, No. 3 July 1986 PRINCE, ERIC D., DENNIS W. LEE, CHARLES A. WILSON, and JOHN M. DEAN. Longevity and age validation of a tag-recaptured Atlantic sailfish, Istiopfwrus platypterus, using dorsal spines and otoliths 493 PARRISH, RICHARD H, DONNA L. MALLICOATE, and RICHARD A. KLINGBEIL. Age dependent fecundity, number of spawnings per year, sex ratio, and maturation stages in northern anchovy, Engraulis mordax 503 PENNINGTON, MICHAEL. Some statistical techniques for estimating abundance indices from trawl surveys 519 REILLY, STEPHEN B., and JAY BARLOW. Rates of increase in dolphin population size 527 ATKINSON, C. ALLEN. Discrete-time difference model for simulating interacting fish population dynamics 535 PARSONS, D. G., and G. E. TUCKER. Fecundity of northern shrimp, Pandcdus borealis, (Crustacea, Decapoda) in areas of the Northwest Atlantic 549 WAHLEN, BRUCE E. Incidental dolphin mortality in the eastern tropical Pacific tuna fishery, 1973 through 1978 559 SOMERTON, DAVID A, and ROBERT S. OTTO. Distribution and reproductive biology of the golden king crab, Lithodes aequispina, in the eastern Bering Sea 571 POWER, JAMES H. A model of the drift of northern anchovy, Engraidis mordax, larvae in the California Current 585 JOHNSON, PHYLLIS T Parasites of benthic amphipods: dinoflagellates (Duboscquodinida: Syndinidae) 605 YANG, M. S., and P. A LIVINGSTON. Food habits and diet overlap of two congeneric species, Antheresthes stomias and Atheresthes evermanni, in the eastern Bering Sea 615 SHEPARD, ANDREW N., ROGER B. THEROUX, RICHARD A. COOPER, and JOSEPH R. UZMANN. Ecology of Ceriantharia (Coelenterata, Anthozoa) of the northwest Atlantic from Cape Hatteras to Nova Scotia 625 POTTHOFF, THOMAS, SHARON KELLEY, and JOAQUIN C. JAVECH. Cartilage and bone development in scombroid fishes 647 REIS, ENIR GIRONDI. Age and growth of the marine catfish, Netuma barba (Siluri- formes, Ariidae), in the estuary of the Patos Lagoon (Brasil) 679 DANDONNEAU, YVES. Monitoring the sea surface chlorophyll concentration in the tropical Pacific: consequences of the 1982-83 El Nino 687 (Continued on back cover) Seattle, Washington U.S. DEPARTMENT OF COMMERCE Malcolm Baldrige, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION Anthony J. Calio, Administrator NATIONAL MARINE FISHERIES SERVICE William G. Gordon, Assistant Administrator Fishery Bulletin The Fishery Bulletin carries original research reports and technical notes on investigations in fishery science, engineering, and economics. The Bulletin of the United States Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates were issued as documents through volume 46; the last document was No. 1103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared as a numbered bulletin. A new system began in 1963 with volume 63 in which papers are bound together in a single issue of the bulletin instead of being issued individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued quarterly. In this form, it is available by subscription from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. It is also available free in limited numbers to libraries, research institutions, State and Federal agencies, and in exchange for other scientific publications. SCIENTIFIC EDITOR, Fishery Bulletin Dr. William J. Richards Southeast Fisheries Center Miami Laboratory National Marine Fisheries Service, NOAA Miami, FL 33149-1099 Editorial Committee Dr. Bruce B. Collette Dr. Jay C. Quast National Marine Fisheries Service National Marine Fisheries Service Dr. Reuben Lasker Dr. Carl J. Sindermann National Marine Fisheries Service National Marine Fisheries Service Mary S. Fukuyama, Managing Editor (ISSN 0090-0656) is published quarterly by the Scientific Publications Office, National Marine Fisheries Service, rid Point Way NE, BIN CI 5700, Seattle, WA 98115. Second class postage is paid in Seattle, Wash., and additional offices. TER send address changes for subscriptions to Superintendent of Documents, U.S. Government Printing Office, Washington, ■102. Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated. Fishery Bulletin CONTENTS Vol. 84, No. 3 July 1986 PRINCE, ERIC D., DENNIS W. LEE, CHARLES A. WILSON, and JOHN M. DEAN. Longevity and age validation of a tag-recaptured Atlantic sailfish, Istiophorus platypterus, using dorsal spines and otoliths 493 PARRISH, RICHARD H., DONNA L. MALLICOATE, and RICHARD A. KLINGBEIL. Age dependent fecundity, number of spawnings per year, sex ratio, and maturation stages in northern anchovy, Engraulis mordax 503 PENNINGTON, MICHAEL. Some statistical techniques for estimating abundance indices from trawl surveys 519 REILLY, STEPHEN B., and JAY BARLOW. Rates of increase in dolphin population size 527 ATKINSON, C. ALLEN. Discrete-time difference model for simulating interacting fish population dynamics 535 PARSONS, D G., and G E. TUCKER. Fecundity of northern shrimp, Pandalus borealis, (Crustacea, Decapoda) in areas of the Northwest Atlantic 549 WAHLEN, BRUCE E. Incidental dolphin mortality in the eastern tropical Pacific tuna fishery, 1973 through 1978 559 SOMERTON, DAVID A., and ROBERT S. OTTO. Distribution and reproductive biology of the golden king crab, Lithodes aequispina, in the eastern Bering Sea 571 POWER, JAMES H. A model of the drift of northern anchovy, Engraulis mordax, larvae in the California Current 585 JOHNSON, PHYLLIS T Parasites of benthic amphipods: dinoflagellates (Duboscquodinida: Syndinidae) 605 YANG, M. S., and P. A. LIVINGSTON. Food habits and diet overlap of two congeneric species, Antheresthes stomias and Atheresthes evermanni, in the eastern Bering Sea 615 SHEPARD, ANDREW N, ROGER B. THEROUX, RICHARD A. COOPER, and JOSEPH R. UZMANN. Ecology of Ceriantharia (Coelenterata, Anthozoa) of the northwest Atlantic from Cape Hatteras to Nova Scotia 625 POTTHOFF, THOMAS, SHARON KELLEY, and JOAQUIN C. JAVECH. Cartilage and bone development in scombroid fishes 647 REIS, ENIR GIRONDI. Age and growth of the marine catfish, Netuma barba (Siluri- formes, Ariidae), in the estuary of the Patos Lagoon (Brasil) 679 DANDONNEAU, YVES. Monitoring the sea surface chlorophyll concentration in the tropical Pacific: consequences of the 1982-83 El Nino 687 I M«rtn« BMogteat Uboraton (Continued on next page) Seattle, Washington 1986 For sale by the Superintendent of Documents, U.S. Government Printing Office, W DC 20402— Subscription price per year: $21.00 domestic and $26.25 foreign. Cost per single issue: $6.50 domestic and $8.15 foreign. OCT 6 1986 Woods Hole. Mass. Contents— Continued ROGERS, S. GORDON, HIRAM T. LANGSTON, and TIMOTHY E. TARGETT. Ana- tomical trauma to sponge-coral reef fishes captured by trawling and angling .... 697 QUAST, JAY C. Annual production of eviscerated body weight, fat, and gonads by Pacific herring, Clupea harengus pallasi, near Auke Bay, southeastern Alaska . . . 705 WENNER, CHARLES A., WILLIAM A. ROUMILLAT, and C. WAYNE WALTZ. Con- tributions to the life history of Black sea bass, Centropristis striata, off the south- eastern United States 723 Notes LENARZ, WILLIAM H., and TINA WYLLIE ECHEVERRIA. Comparison of visceral fat and gonadal fat volumes of yellowtail rockfish, Sebastes flavidus, during a normal year and a year of El Nino conditions 743 SORENSEN, PETER W, MARCO L. BIANCHINI, and HOWARD E. WINN. Diel foraging activity of American eels, Anguilla rostrata (Lesueur), in a Rhode Island estuary 746 KILLAM, KRISTIE, and GLENN PARSONS. First record of the longfin mako, Isurus paucus, in the Gulf of Mexico 748 STIER, KATHLEEN, and BOYD KYNARD Movement of sea-run sea lampreys, Petro- myzon marinus, during the spawning migration in the Connecticut River 749 HOGANS, W E., and P. C. F. HURLEY. Variations in the morphology of Fistulicola plicatus Rudolphi (1802) (Cestoda:Pseudophyllidea) from the swordfish, Xiphias gladius L., in the Northwest Atlantic Ocean 754 The National Marine Fisheries Service (NMFS) does not approve, recommend or en- dorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to NMFS, or to this publication furnished by NMFS, in any advertising or sales promotion which would indicate or imply that NMFS ap- proves, recommends or endorses any proprietary product or proprietary material mentioned herein, or which has as its purpose an intent to cause directly or indirect- ly the advertised product to be used or purchased because of this NMFS publication. LONGEVITY AND AGE VALIDATION OF A TAG-RECAPTURED ATLANTIC SAILFISH, ISTIOPHORUS PLATYPTERUS, USING DORSAL SPINES AND OTOLITHS Eric D. Prince,1 Dennis W. Lee,1 Charles A. Wilson,2 and John M. Dean3 ABSTRACT A tagged female Atlantic sailfish, Istiophorus platypterus, of 24.6 kg (54 lb) was recaptured on 14 January 1984, after being at large for 10 yr and 10 mo (4,025 d). Approximate age based on tagging records ranged from at least 13 to 15 + yr. Maximum estimated longevity of this species was therefore revised upwards from previously reported >7 yr to at least 13-15+ yr. Estimates of age based on sections of dorsal spine numbers 3-6 ranged from 2 to 8 yr and substantially underestimated the range in age known from tagging records (13-15 + yr). This discrepancy was due to enlargement of the porous, vascularized core of spine sections which obscured zonations associated with early growth history. Thus, dorsal spines do not appear to be useful in ageing older sailfish (i.e., >5 yr). Age estimates from sagittae (otoliths) were 13 yr based on scanning electron microscope counts of external ridges and analysis of internal otolith microstructure. Otolith age, therefore, agreed with age known from tagging records. The relatively large size of the sagitta (7.84 mg) also provides additional evidence that the otolith could be from a very old sailfish. These data strongly suggest that in older, larger sailfish (>5 yr, 22.7 kg), sagittae, rather than dorsal spines, should be used as the source of age and growth information. The Atlantic sailfish, Istiophorus platypterus, is one of the most popular recreational fishes along the U.S. Atlantic coast, Gulf of Mexico, and Caribbean Sea. In fact, this species has been described as the most sought after fish by southeast marine charter boat anglers, particularly in south Florida (Ellis 1957). Although most landings of Atlantic sailfish in the southeastern United States are made by recreational anglers, many are also taken inciden- tally by domestic and foreign commercial longline vessels (Lopez et al. 1979). The biological informa- tion presently used in stock assessments of Atlan- tic sailfish (Conser 1984) consists of age and growth data derived exclusively from analysis of dorsal spines (Jolley 1974, 1977; Hedgepeth and Jolley 1983). However, uncertainties remain concerning Atlantic sailfish age structure, longevity, choice of skeletal structure for ageing, and rate of growth because of inconsistencies reported in the literature. In addition, the accuracy of age and growth esti- 1 Southeast Fisheries Center Miami Laboratory, National Marine Fisheries Service, NOAA, 75 Virginia Beach Drive, Miami, FL 33149-1099. 2BeIle W. Baruch Institute for Marine Biology and Coastal Research, Department of Biology and Marine Science Program, University of South Carolina, Columbia, SC 29208; present ad- dress: Coastal Ecology and Fisheries Institute, Louisiana State University, Baton Rouge, LA 70803-7503. 3Belle W. Baruch Institute for Marine Biology and Coastal Research, Department of Biology and Marine Science Program, University of South Carolina, Columbia, SC 29208. mates from skeletal structures and length-frequency analyses have not been validated for all age classes (de Sylva 1957; Jolley 1974, 1977; Radtke and Dean 1981; Hedgepeth and Jolley 1983). One problem in using spines as a source of age and growth information is the tendency of the vascularized core to obscure zonations associated with early growth history. The enlargement of the vascularized core and subsequent reabsorption of tissues are most severe in the largest and oldest specimens (causing underestimates of true age) and have contributed to the lack of detailed information for older age classes. Several studies have also reported difficulty in interpreting the double and tri- ple bands often observed in Atlantic sailfish spines (Jolley 1977; Hedgepeth and Jolley 1983). These problems are not unique to sailfish (Casselman 1983; Compean- Jimenez and Bard 1983) and have resulted in an unusually large proportion of spine samples (as much as 76%) being rejected for age and growth analysis (Jolley 1977). Radtke and Dean (1981) reviewed this problem and suggested that otoliths (sagittae) may be a better skeletal structure for age and growth assessment in sailfish because these structures do not have the disadvantages associated with the spinal core. For example, 98% of the oto- lith samples examined by Radtke and Dean (1981) were reportedly suitable for age and growth estima- tion. Even though these preliminary findings were Manuscript accepted October 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 493 FISHERY BULLETIN: VOL. 84, NO. 3 encouraging, use of otoliths to resolve age and growth discrepancies for Atlantic sailfish has not been reported, and no conclusive evidence is avail- able to validate the accuracy of age estimates for this species using any method. We present an analysis of dorsal spines and otoliths obtained from one tag-recaptured Atlantic sailfish, where age was very closely approximated from tagging records, to help resolve the problems associated with ageing this species. METHODS The Cooperative Gamefish Tagging Program of the Southeast Fisheries Center Miami Laboratory recovered a tag from a female Atlantic sailfish, which had been recaptured on 14 January 1984, off Boynton Beach, FL (Prince and Lee 1984). This fish was originally tagged and released off the Florida Keys (Islamorada) on 5 March 1973, at an estimated weight of 18.2 kg (about 40 lb). When recaptured it weighed 24.6 kg (54 lb) and had a lower jaw fork length (LJFL) of 176.5 cm. The sailfish appeared to have a healthy external appearance when caught and body proportions and overall morphology were within the normal range for a specimen of this size. The entire fish was made available to us by J. T. Reese Taxidermist, Inc. (Ft. Lauderdale, FL), and both sagittae and the first six dorsal spines were sampled for age determination. Dorsal Spine Analysis Dorsal spines were collected from the tagged Atlantic sailfish following the procedures of Prince and Lee (1982). Past efforts to age sailfish using dor- sal spines have relied on spine number 4 as the source of age and growth information (Jolley 1974, 1977; Hedgepeth and Jolley 1983). We collected the first six anterior dorsal spines to insure that the number assigned to each spine was accurate for identification and analysis and to gain information about possible differences between spines. The first two anterior dorsal spines of sailfish are greatly reduced in size compared with spines 3-6 and were not used to estimate age. In addition, spines pos- terior to spine number 6 have a smaller diameter and were not used for age determination. This deci- sion was based, in part, on a report by Robins4 and Robins and de Sylva (1963) who believed that the 4Robins, C. R., Professor, Rosenstiel School of Marine and At- mospheric Sciences, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149, pers. commun. 1982. posterior dorsal spines of billfish do not grow throughout their entire lifetime and recommended that only anterior spines be used for age and growth studies. Dorsal spines 3-6 were cleansed of tissue, labeled with a collection number, and preserved in isopropyl alcohol (98%). The methods of sectioning dorsal spines given by Hedgepeth and Jolley (1983) and Prince et al. (1984) were used in this study. Dorsal spine number 4 was sectioned by M. Y. Hedgepeth at the laboratory of the Florida Department of Natural Resources (FDNR), West Palm Beach, FL, to ensure that processing of this spine was identical with methods previously reported. We sectioned spines 3, 5, and 6 using a Buehler ISOMET5 saw and a 10.16 cm diameter diamond wafer blade. At least 2 or 3 sections (0.44-0.46 mm thick) were taken from each spine. Additional sections were taken from spine number 4 after it had been processed by FDNR personnel. All spine sections were placed into labeled vials with isopropyl alcohol (98%) for storage and extraction of oil. A single section was selected and allowed to air dry before microscopic examina- tion. Dorsal spine sections were examined initially using a compound stereoscope (6.0 x) with trans- mitted light in order to assess that portion of the section not affected by the vascularized core. Mea- surements (in millimeters, mm) of the solid bone area in the distal portion of the right lobe of each section were taken along a straight-line counting path from the focus to the outside margin of the structure. We assigned an age to each spine by counting only concentric translucent bands that were continuous around the circumference of the entire section. In transmitted light, the zonations consisted of a dark opaque zone followed by a light translucent zone. D. W. Lee made three repeated counts of translu- cent zones using a compound stereoscope at 12.0 to 25.0 x magnification. Otolith Analysis The general methods of Radtke and Dean (1981) and Wilson and Dean (1983) were used to extract and prepare the sagittae for examination by scan- ning electron microscopy (SEM) and light micro- scopy. The sagittae were removed from the tagged Atlantic sailfish, cleaned with sodium hypochloride solution, and rinsed in xylene and then 95% ethanol. 5Reference to trade names and products does not imply endorse- ment by the National Marine Fisheries Service. 494 PRINCE ET AL.: LONGEVITY AND AGE OF ATLANTIC SAILFISH The weight of one air-dried otolith was measured to 0.001 mg (±5%) using a Perkin Elmer AD2Z ultra-microbalance. The sagitta was attached to an aluminum stub, coated with gold, and examined by SEM at 15.0-1500 x to observe the surface mor- phology. External ridges on the rostral lobe of sailfish sagittae, first described by Radtke and Dean (1981), was one of the features used in this study for age estimation. Following the methods of Haake et al. (1982) and Wilson and Dean (1983), the other member of the pair of sagittae was embedded in epoxy resin, and a section was made in the transverse plane by polish- ing both sides to 0.5 mm thickness with 600 grit sandpaper and 0.3 ^m alumina polish. The internal structure of the sectioned sagitta was examined with an Olympus BH2 compound microscope at 4.0 to 1200 x to aid overall orientation and understanding of the growth of the structure and to interpret the external ridges used for age estimation. imum longevity of this species by a considerable margin. Although Jolley (1977) speculated that sail- fish may live as long as 9 or 10 yr because the one age 8 individual was not the largest specimen in his sample, his estimated ages did not exceed 8 yr. In addition, the maximum estimated age reported in other recent studies was >7 yr (Radtke and Dean 1981; Hedgepeth and Jolley 1983). An Atlantic sail- fish of estimated age 7 or 8 from the above sources corresponds to an average size of about 25 kg (55 lb). Since our records indicate the age of the tag- recaptured 24.6 kg (54 lb) sailfish was 13-15+ yr, it appears that maximum longevity of Atlantic sail- fish could be considerably older, perhaps over 20 yr, because numerous specimens exceeding 45.5 kg (100 lb) have been caught during the last decade (Beards- ley 1980). This reasoning assumes that sailfish have indeterminate growth throughout their entire life- time and that their size is proportional to age. It also appears from tagging data that Atlantic sailfish may RESULTS AND DISCUSSION Our tagging records indicate that the tagged Atlantic sailfish recaptured on 14 January 1984, was at-large for 10 yr and 10 mo or 4,025 d. An experi- enced charter boat captain estimated its size when tagged to be 18.2 kg (40 lb). Bias in overestimating the size of billfish during tagging has been a com- mon problem since the inception of the Cooperative Gamefish Tagging Program in 1954 (Prince 1984). However, we feel that such an error would probably not exceed ± 4.6 kg (10 lb) in a fish of this size, par- ticularly when the experience of the captain making the estimate is considered. The estimated age of a sailfish of about 18.2 kg (40 lb) would be 2-4 yr based on dorsal spine analysis (Jolley 1974, 1977; Hedge- peth and Jolley 1983) and 3-5 yr based on otolith analysis (Radtke and Dean 1981). Therefore, the ap- proximate range in age of this sailfish based on tag- ging information is 13-15+ yr. We feel these are conservative figures based on the available informa- tion and it is highly unlikely that this fish could be younger than 13 yr. Maximum longevity of Atlantic sailfish was first inferred by de Sylva (1957) to be at least 3 or 4 yr based on length-frequency analysis (Fig. 1). A modal group beyond 4 yr was indicated in his analysis but year class designation was not discussed. Since 1957, estimated longevity of Atlantic sailfish has been revised upwards (Fig. 1) to ^7 yr. Our tagging records indicate, however, that the oldest Atlantic sailfish aged by dorsal spine analysis (Jolley 1977; estimated age 8) probably underestimates the max- 14 r of 12 > x en Ll. 5'° CO li. o > t 8 > UJ o 6 UJ I- < I- CO A _ UJ ^ 2E 3 2 x < 2 STUDIES 1 deSylva (1957) 2 Jolley (1974) 3 Jolley (1977) 4 Radtke ond Deon (1981) 5 Hedgepeth ond Jolley (1983) 6 Prince et al. (This paper) 13-15+ 9-10 3-4 >7 957 974 1977 1981 1983 984 (1) (2) (3) (4) (5) (6) YEAR (STUDY) Figure 1.— Estimates of maximum longevity (yr) for Atlantic sailfish from six different studies, 1957-84. 495 FISHERY BULLETIN: VOL. 84, NO. 3 grow very slowly after sexual maturity (sexual maturity for female Atlantic sailfish reported at 13-18 kg, Jolley 1977). For example, tagging records indicate that this fish, which was tagged at 18.2 kg, gained only about 6.4 (14 lb) while being at-large almost 1 1 yr. Our analysis of spines and otoliths sup- port these findings. Dorsal Spines Examination of sections from dorsal spines 3-6 (Fig. 2) indicated that the vascularized core com- prised an extensive area in all sections. The solid bone area where zonations were not disrupted varied in size and comprised 14, 19, 30, and 37% of the right lobe of spine sections 3, 4, 5, and 6, respectively (Table 1). The vascularized core severe- ly restricted the zonation counts because increments associated with early growth history were totally disrupted and could not be enumerated. Counts of zonations on the four spine sections ranged from 2 to 8 (Table 1). This suggests that spine number 4, which had been used in past studies to assign ages, may not necessarily be the best choice for ageing sailfish, particularly for the larger, older specimens. For example, spines 5 and 6 both had a higher per- centage of solid bone, and counts of zonations in these spines were proportionately higher than in spines 3 and 4 (Fig. 2, Table 1). However, all spines substantially underestimated the age of this sailfish, where approximate age (13-15+ yr) was known from tagging records. Hedgepeth6 reports that spine number 4 would not have been included in the data sets of previous published studies because of the ex- tensive size of the vascularized core area. We con- clude from these data that dorsal spine sections are probably only useful for ageing sailfish from > 1 to 6Hedgepeth, M. Y., Fisheries Biologist, Florida Department of Natural Resources, 727 Belvedere Rd., West Palm Beach, FL 33405, pers. commun. 1984. Table 1 .—Mean count of zonations (3 repetitions) and percentage solid bone in the distal portion of the right lobe of sections taken from dorsal spines 3-6 of Atlantic sailfish (see text and Fig. 2). Measurements and counts were taken along a straight line count- ing path bisecting the spine laterally from the focus to the edge of each section. Dorsal Mean Solid Solid bone Total spine count bone measurement measurement number (range) (0/0) (mm) (mm) 3 2.0 14 1.89 13.52 4 3.7(3-4) 19 3.55 18.76 5 5.0 30 4.90 16.56 6 7.3(7-8) 37 6.08 16.39 5 yr. Although there may be some bias associated with ageing these young sailfish because of the vascularized core, this bias is probably minimal. However, for sailfish older than estimated age 5 and about >22.7 kg (50 lb), the bias substantially under- estimates age and this bias increases with an in- crease in size and age of the fish. In addition, spines have not been shown to be useful in ageing sailfish 05 mm LATERAL SURFACE 498 PRINCE ET AL.: LONGEVITY AND AGE OF ATLANTIC SAILFISH ' S **? ' urn ■T'^W VENTRAL SURFACE 10 9 8 ^ f 4 -TT+-T- • T^~>* LATERAL Ik '*■''. X -SURFACE 5W > t FIRST 2 RIDGED COVERED BY/ CaCOj^fvER- BURDEN ^ / CORE NSID Figure 4.— Lateral and ventral view of the sagitta rostrum from the tag- recaptured Atlantic sailfish showing the overall pattern of otolith growth and the approximate location and number of the first 10 external ridges. 499 FISHERY BULLETIN: VOL. 84, NO. 3 zone exists between the boundary of many of the external ridges shown in Figure 3. This zone extends from the surface to deep within the internal struc- ture of the rostrum (Fig. 3). The distinct change in optical density of the first of these prominent zones marks what we believe is the boundary between the 1st and 2d ridges and suggests that the first major growth zone is located between the core and the bend (Fig. 3). Further, using the SEM we counted 150-200 finely spaced increments between the core and the first prominent translucent zone. This count also supports our interpretation of the location of the first annual zone if these increments are as- sumed to form daily, the fish was born sometime in late spring or early summer (as reported by Beardsley et al. 1975), and the annual zonations are being formed in the winter. Jolley (1977) reported that annual zones in Atlantic sailfish spines tend to be formed in late fall or winter and he also spec- ulated that sailfish may form the first annuli on spines prior to a full year's growth. The location of the second translucent zone, based on similar evi- dence, appears to be at the beginning of the bend (Fig. 3). The width of the first two major growth zones (^0.5 mm) are considerably larger than the zones beyond the end. Wide spacing of year marks during early growth have been observed in many fishes when growth rates are most rapid (Dean et al. 1983). Therefore, these data support our conten- tion that at least two ridges should be accounted for as occurring within the boundaries of the lateral surface. Rostral ridges 3 through 10 were easily distin- guished and counted on the sagitta's ventral surface within the same plane of focus (Fig. 5, bottom). After the 10th ridge, however, the rostrum changes direction slightly (Fig. 3), and it was necessary to refocus to observe ridges 11 through 13 (Fig. 5, top). We feel that potential sources of error in our counts of rostral ridges would have most likely occurred at the beginning and end of the counting path. In addition, we feel that if errors were made at these locations, they would have increased the count. Therefore, otolith age of the tagged Atlantic sailfish was estimated to be 13 yr. However, it should be recognized that potential errors in this estimate could have resulted if one or two ridges were un- accounted for on the lateral surface or on the tip of the rostrum on the ventral surface. Otolith age under these circumstances should be presented con- servatively as ranging from 13 to 15+ yr. The weight of one sagitta from the tagged Atlan- tic sailfish (7.84 mg) was extremely heavy for an istiophorid of comparable size. For example, it was 1.24 mg heavier than the sagitta from a 29.6 kg (65 lb) sailfish caught in 1985 off Miami and was 1.18 mg heavier than the largest sagitta from Pacific blue marlin reported by Radtke (1983). In addition, the tagged sailfish sagitta was in the upper range in weight (0.51-8.16 mg) of more than 500 blue and white marlin sagittae examined by Wilson (1984). Since the relationship between the size of otoliths and the age of fishes has been shown to be positive- ly correlated for some teleosts (Somerton 1985), we feel that the relatively large size of this sagitta pro- vides additional indirect evidence that this structure could be from a very old sailfish. CONCLUSIONS Our tagging records indicate that estimates of maximum longevity for Atlantic sailfish should be revised upwards to at least 13-15+ yr, and that sailfish of this age can grow at a very slow rate (about 0.59 kg/yr during its time at large). Dorsal spines do not appear to be an accurate source of age and growth information for older, larger sailfish (>5 yr, ^22.7 kg or 50 lb), while sagittae do provide more accurate estimates of age for these older age groups. Since current stock assessments of Atlantic sailfish (Conser 1984) rely exclusively on dorsal spine ageing data as input, these assessments offer little insight into the more mature segments of the population. If skeletal structures from the larger, older fish are systematically rejected for ageing analyses, an underestimate of age and longevity and an overestimate of growth rate can occur (Nammack et al. 1985). Therefore, future assessments should be revised using otolith ageing methods to clarify that portion of the age structure that can not be reliably appraised using dorsal spines. ACKNOWLEDGMENTS We thank J. T. Reese Taxidermist, Inc., Ron Har- rison (angler), and Captain Bud Carr for providing us with biological samples and other information from the tagged Atlantic sailfish. Personnel from the Florida Department of Natural Resources in West Palm Beach, FL, sectioned and analyzed dor- sal spine number 4. Dana Dunkleberger (University of South Carolina) assisted in preparing scanning electron micrographs. LITERATURE CITED Beardsley, G. L. 1980. Size and possible origin of sailfish, Istiophonts 500 PRINCE ET AL.: LONGEVITY AND AGE OF ATLANTIC SAILFISH Figure 5.— Scanning electron micrograph of the ventral view of the sagitta rostrum from the tag-recaptured Atlantic sailfish. A count of external ridges 3-10 (bottom) and 10-13 (top) were used to assign a numeric otolith age of 13 yr. Bar on bottom = 1.0 mm, bar on top = 0.1 mm. 501 FISHERY BULLETIN: VOL. 84, NO. 3 platypterus, from the eastern Atlantic Ocean. Fish. Bull., U.S. 78:805-808. Beardsley, G. L., Jr., N. R. Merrett, and W. J. Richards. 1975. Synopsis of the biology of the sailfish, Istiophorus platypterus (Shaw and Neddor, 1791). In R. S. Shomura and F. Williams (editors), Proceedings of the international billfish symposium, Kailua-Kona, Hawaii, 9-12 August 1972, part 3, p. 95-120. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-675. Casselman, J. M. 1983. Age and growth assessment of fish from their calcified structures— techniques and tools. /« E. D. Prince and L. M. Pulos (editors), Proceedings of the international work- shop on age determination of ocean pelagic fishes: tunas, billfishes, and sharks, p. 1-18. U.S. Dep. Commer., NOAA Tech. Rep. NMFS-8. COMPEAN-JlMENEZ, G., AND F. X. BARD. 1983. Growth increments on dorsal spines of eastern Atlan- tic bluefin tuna, Thunnus thynnus, and their possible rela- tion to migration patterns. In E. D. Prince and L. M. Pulos (editors), Proceedings of the international workshop on age determination of oceanic pelagic fishes: tunas, billfishes, and sharks, p. 77-86. U.S. Dep. Commer., NOAA Tech. Rep. NMFS-8. CONSER, R. J. 1984. Yield per recruit analysis of sailfish, Istiophorus platy- pterus, in the western Atlantic Ocean. Int. Comm. Conserv. Atl. Tunas, Collect. Vol. Sci. Pap., Madrid 20:448-464. Dean, J. M., C. A. Wilson, P. W. Haake, and D. W. Beckman. 1983. Microstructural features of teleost otoliths. In P. Westbroek and E. W. deJong (editors), Biomineralization and biological metal accumulation, biological and geological perspectives, p. 353-359. D. Reidel Publ. Co., Holland. de Sylva, D. P. 1957. Studies on the age and growth of the Atlantic sailfish, Istiophorus americanus (Cuvier), using length-frequency curves. Bull. Mar. Sci. Gulf Caribb. 7:1-20. Ellis, R. W. 1957. Catches of fish by charter boats on Florida's east coast. Univ. Miami Mar. Lab. Spec. Serv. Bull. 14, 6 p. Haake, W., C. A. Wilson, and J. M. Dean. 1982. A technique for the examination of otoliths with ap- plications to larval fishes. In C. F. Bryan, S. V. Conner, and F. M. Truesdale (editors), Proceedings of the 5th annual larval fish conference, p. 12-15. Baton Rouge, LA. Hedgepeth, M. Y., and J. W. Jolley, Jr. 1983. Age and growth of sailfish, Istiophorus platypterus, using cross sections from the fourth dorsal fin spine. In E. D. Prince and L. M. Pulos (editors), Proceedings of the international workshop on age determination of oceanic pelagic fishes: tunas, billfishes, and sharks, p. 131-136. U.S. Dep. Commer., NOAA Tech. Rep. NMFS-8. Jolley, J. W., Jr. 1974. On the biology of Florida east coast Atlantic sailfish (Istiophorus platypterus). In R. S. Shomura and F. Williams (editors), Proceedings of the international billfish sym- posium, Kailua-Kona, Hawaii, 9-12 August 1972, part 2, Review and contributed papers, p. 81-88. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-675. 1977. The biology and fishery of Atlantic sailfish, Istiophorus platypterus, from southeast Florida. Fla. Mar. Res. Publ. 28, 31 p. Lopez, A. M., D. B. McClellan, A. R. Bertolino, and M. D. Lange. 1979. The Japanese longline fishery in the Gulf of Mexico, 1978. Mar. Fish. Rev. 41(10):23-28. Nammack, M. F., J. A. Musick, and J. A. Colvocoresses. 1985. Life history of spiny dogfish off the northeastern United States. Tran. Am. Fish. Soc. 114:367-376. Prince, E. D. 1984. Angler participation in oceanic pelagics programs, Southeast Fisheries Center, Miami Laboratory. (Abstr.) In 114th annual meeting of the American Fisheries Society, Ithaca, NY, August 12-16, 1984, p. 74-75. Prince, E. D., and D. W. Lee. 1982. Bioprofiles sampling manual for oceanic pelagic fishes, 1982-83. U.S. Dep. Commer., NOAA Tech. Memo. NMFS SEFC-103, 26 p. 1984. Research on age and growth. In Oceanic pelagic pro- gram summary 1983, p. 45-62. NOAA, NMFS, Miami Lab. Ann. Rep. Prince, E. D., D. W. Lee, C. A. Wilson, and J. M. Dean. 1984. Progress in estimating age of blue marlin, Makaira nigricans, and white marlin, Tetrapturus albidus, from the western Atlantic Ocean, Caribbean Sea, and Gulf of Mexico. Int. Comm. Conserv. Atl. Tunas, Collect. Vol. Sci. Pap., Madrid 20:435-447. Radtke, R. L. 1983. Istiophorid otoliths: extraction, morphology, and possi- ble use as ageing structures. In E. D. Prince and L. M. Pulos (editors), Proceedings of the international workshop on age determination of oceanic pelagic fishes: tunas, bill- fishes, and sharks, p. 123-129. U.S. Dep. Commer., NOAA Tech. Rep. NMFS-8. Radtke, R. L., and J. M. Dean. 1981. Morphological features of the otoliths of the sailfish, Istiophorus platypterus, useful in age determination. Fish. Bull., U.S. 79:360-367. Radtke, R. L., L. M. Collins, and J. M. Dean. 1982. Morphology of the otoliths of the Atlantic blue marlin (Makaira nigricans) and their possible use in age estima- tion. Bull. Mar. Sci. 32:498-503. Robins, C. R., and D. P. de Sylva. 1963. A new western Atlantic spearfish, Tetrapturus pflue- geri, with a redescription of the Mediterranean spearfish Tetrapturus belone. Bull. Mar. Sci. Gulf Caribb. 13:84- 122. Somerton, D. 1985. Indirect aging of fish using otolith weights and cross- section areas. (Abstr.) In International symposium on age and growth of fish, Des Moines, Iowa, June 9-12, 1985, p. 93-94. Wilson, C. A. 1984. Age and growth aspects of the life history of billfish. Ph.D. Thesis, Univ. South Carolina, Columbia, SC, 180 p. Wilson, C. A., and J. M. Dean. 1983. The potential use of sagittae for estimating age of Atlantic swordfish, Xiphias gladius. In E. D. Prince and L. M. Pulos (editors), Proceedings of the international work- shop on age determination of oceanic pelagic fishes: tunas, billfishes and sharks, p. 151-156. U.S. Dep. Commer., NOAA Tech. Rep. NMFS-8. 502 AGE DEPENDENT FECUNDITY, NUMBER OF SPAWNINGS PER YEAR, SEX RATIO, AND MATURATION STAGES IN NORTHERN ANCHOVY, ENGRAULIS MORDAX Richard H. Parrish,1 Donna L. Mallicoate,1 and Richard A. Klingbeil2 ABSTRACT Maturity stage data from fishery sampling programs and ovarian histological data from research cruises were used to develop a method for determining the age-specific number of spawnings per year and annual fecundity of the central stock of northern anchovy, Engraulis mordax. The sex ratio was found to be size and age dependent in both the fishery and trawl surveys with females increasingly dominant in the larger and older size and age classes. The overall sex ratio in trawl surveys was nearly 1:1; the fishery data favored females 1.48:1. The magnitude and duration of maturity stages were size and age dependent with peak spawning occurring earlier in the season in younger fish. Maturity stages and histological classes with hydrated eggs showed essentially the same diurnal pattern for nightly spawning activity indicating that the presence of hydrated eggs could be used as an index of daily spawning. The daily spawning incidence and total annual fecundity were found to be heavily age dependent. Females in their first spawning season had an average of 5.3 spawnings, while those in their fourth had an average of 23.5 spawnings. When combined with the higher batch fecundity of larger fish this results in 4 + year-old females producing nearly 5 times as many eggs per unit of weight as 1-year-olds. When the age-specific fecundity and sex ratio in the fishery are combined it is apparent that the catch of a ton of 4 + year-old northern anchovy reduces the reproductive potential of the stock 7.3 times as much as the catch of a ton of 1-year-olds. It was concluded that age-specific fecundity in multiple spawning fishes is of greater significance for management than previously thought. It is also significant that much of the observed variance in stock-recruitment relationships for multiple-spawning fishes may be due to the fact that spawning biomass is a poor index of the egg production and reproductive potential of the stock. Age-specific variation in life history rates is a major factor in population and management models of ex- ploited fishes, and variation in reproductive effort is of great significance in such models. Size and age- specific batch fecundity estimates have been avail- able for many species for decades, and for species which spawn only once per spawning season these are readily incorporated into models. However it has been impossible to determine the age-specific repro- ductive effort of species which spawn many times during a spawning season because there has been no way to determine the number of spawnings per year. Recent research on the histology of the ovaries of northern anchovy, Engraulis mordax, and an- choveta, Engraulis ringens, suggest that they spawn approximately once a week during peak spawning months (Hunter and Goldberg 1980; 'Southwest Fisheries Center, Pacific Fisheries Environmental Group, National Marine Fisheries Service, NOAA, P.O. Box 831, Monterey, CA 93942. California Department of Fish and Game, 245 W. Broadway Street, Long Beach, CA 90802. Hunter and Macewicz 1980; Alheit et al. 1983; Alheit et al. 1984). Hunter and Leong (1981), in their study of the spawning energetics of the northern anchovy, found that northern anchovy spawn about 20 times per year. Hunter and Leong (1981) and Alheit et al. (1983) suggested that annual fecundity per unit of parental biomass may be highly variable and depen- dent upon the nutritional state and size structure of the stock. Potentially the recently developed histological techniques could be utilized to determine age-specific annual fecundity; however, this would be very ex- pensive as it would require a very large data set which would necessarily be stratified by age and time of year. The objectives of this report are 1) to demonstrate a method for combining the high resolution reproductive information from the histology of the ovaries with inexpensive, lower resolution reproductive information derived from resource surveys and fishery sampling programs to determine the age-specific reproductive potential of a multiple spawning species, and 2) to evaluate the gross anatomical maturity stages which have been Manuscript accepted January 1986. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 503 FISHERY BULLETIN: VOL. 84, NO. 3 utilized in field sampling programs for northern an- chovy and to use the historical data from these pro- grams in conjunction with ovarian histological data to describe age-dependent annual fecundity in the central stock of northern anchovy. DATA SOURCES There are three major sources of biological data for adult northern anchovy in California: samples taken from the commercial fishery (Collins and Spratt 1969), samples taken from midwater trawl hauls carried out by the Sea Survey Program (Mais 1974), and samples taken primarily by midwater trawl during egg production cruises (Picquelle and Hewitt 1983). The first two sources are the result of long-term research and monitoring programs carried out by the California Department of Fish and Game, and the third is the result of research cruises carried out by the National Marine Fisheries Ser- vice. The fishery data used in this analysis consist of biological information for 60,661 northern an- chovy sampled from the San Pedro purse seine fleet during the period of 1966-80 and 4,904 northern an- chovy sampled from the Monterey fleet during 1966-78. All northern anchovy in the fishery samples were aged and nearly all were assigned maturity stages. We used a geographically restricted subset of the 1966-83 sea survey data (lat. 29.5°-34.5°N; 54,457 northern anchovy). Maturity stages were not recorded for males in the sea survey data and age determinations were made on only a portion (19,031) of the fish sampled. In both data sets age determina- tions were made from otoliths with methods described by Collins and Spratt (1969). The third source of data, provided to us by B. Macewicz3, con- sists of histological information for the gonads of 8,672 females sampled during the months of February to April from 1977 to 1984. Age deter- minations were not made and maturity stages were not taken on these fish. The gross anatomical maturity stage description used for northern anchovy is a slightly modified version of the system developed by Bowers and Holliday (1961) for herring. The system has seven maturity stages which are primarily based on the portion of the body cavity occupied by the gonads and, in the later stages, by the appearance of trans- lucent eggs or milt (Table 1). Herring are consider- ably larger than anchovy, and they are generally not 3B. Macewicz, Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, 8604 La Jolla Shores Drive, La Jolla, CA 92038, pers. commun. August 1984. considered to be multiple spawners; therefore, there are some difficulties in applying the maturity stages to anchovy. The most obvious problem is that a con- siderable proportion of the anchovy sampled had gonads so small that sex determinations were not made as they would have required magnification. There was also a small proportion of fish in which physical deterioration made sex determination im- possible. The California anchovy fishery is primar- ily for reduction to fish meal, and the quality of the fish was occasionally very poor when the fish were sampled. Another major difficulty is that it is not possible to distinguish between anchovy gonads that are resting (i.e., stage 2) between multiple spawn- ings in the same season and those resting between spawning seasons. A comparable problem exists with spent fish (stage 7). Table 1 .—The international (Hjort) scale of maturity stages of the gonad. From Bowers and Holliday (1961). Stage 1 : Virgin individuals: very small sexual organs close under vertebral column; ovaries wine-colored, torpedo-shaped, about 2-3 cm long and 2-3 cm thick, eggs invisible to naked eye; testes whitish or greyish-brown, knife- shaped, 2-3 cm long and 2-3 cm broad. Stage 2: Maturing virgins or recovering spents: ovaries some- what longer than half the length of ventral cavity, about 1 cm diameter, eggs small but visible to naked eye; milt whitish, somewhat bloodshot, of same size as ovaries, but still thin and knife-shaped. Stage 3: Sexual organs more swollen, occupying about half of ventral cavity. Stage 4: Ovaries and testes occupying almost two thirds of ven- tral cavity; eggs not transparent, milt whitish, swollen. Stage 5: Sexual organs filling ventral cavity; ovaries with some large transparent eggs; milt white, not yet running. Stage 6: Roe and milt running (spawning). Stage 7: Spents: ovaries slack with residual eggs; testes baggy, bloodshot. SEX RATIO Description of the sex ratio in northern anchovy was confounded by the presence of fish for which the sex could not be determined. The relationship between size and the percentage of these unsexed fish is similar for both the commercial purse seine and midwater trawl data. In both data sets, a large percentage of the fish smaller than 100 mm stand- ard length (SL) are of unknown sex, about 10% of the 101-110 mm fish are of unknown sex, and only a small percentage of the fish larger than 110 mm are of unknown sex (Table 2). The percentages of fish with unknown sex at sizes larger than 110 mm in the purse seine data are somewhat higher than those in the midwater trawl data. This is probably due to the occasional occurrences of fish in which 504 PARRISH ET AL.: AGE DEPENDENT FECUNDITY IN NORTHERN ANCHOVY the condition was too poor to allow sex identifica- tion. The relationship between age and the percent- age of unsexed fish is similar to that described for size (Table 2). It should be noted that both data sets are biased towards larger and older fish. The age composition of anchovy in the midwater trawl data is biased because of the fact that otoliths were often not taken when trawl hauls were dominated by young-of-the-year fish (Parrish et al. 1985). The purse seine data contain a much smaller percentage of small or young fish than the midwater trawl data. This is primarily caused by the fact that a 5-in (about 108 mm SL) minimum size limit was in effect for most of the 1966-80 period. To evaluate the seasonal cycle of the occurrence of northern anchovy for which the sex could not be identified, the monthly percentages of females, males, and anchovy with unknown sex were deter- mined by age group for the San Pedro fishery (Fig. 1). In all age groups the minimum percentages of fish which could not be sexed occurred from about January to May in association with the spawning season. Higher percentages occurred both before and after the spawning season, particularly in the first potential spawning season. This implies that a significant percentage of anchovy mature, or at least partially mature, and then reabsorb their gonadal tissue to the point that their gonads are so small that they cannot be sexed without magnifica- tion. It also implies that a bias due to the unsexed fish exists. This bias is at a minimum from January to May and it primarily affects the analyses of fish in their first potential spawning season. Klingbeil (1978) found the female:male ratios of northern anchovy sampled in the Sea Survey Pro- gram and in the commercial fishery to be 1.03:1 and 1.60:1 respectively. The additional years of infor- mation from these two sources, in our data sets, pro- Table 2.— Proportion of northern anchovies with unknown sex and sex ratio by size (A) and by age (B). San Pedro anchovy fishery Sea survey (lat 29.5°-34.5°N) Length Unknown Females Mean Unknown Females Mean (mm) sex male SL Number sex male SL Number 61- 70 1.000 A 1 0.996 0.00 273 71- 80 0.676 0.71 — 37 0.941 0.94 — 597 81- 90 0.437 0.67 — 252 0.810 0.78 — 1,396 91-100 0.283 0.77 — 2,261 0.500 0.84 — 2,208 101-110 0.114 1.06 — 8,684 0.100 0.82 — 2,882 111-120 0.057 1.30 — 17,186 0.016 0.78 — 4,141 121-130 0.029 1.53 — 19,396 0.007 1.07 — 4,016 131-140 0.014 2.10 — 10,010 0.003 1.43 — 2,439 141-150 0.007 2.90 — 2,354 0.003 2.56 — 772 151-160 0.009 4.43 — 438 0.005 3.09 — 185 161 + 0.024 7.20 — 42 0.000 9.67 — 32 Total 0.057 1.48 — 60,661 B Annual 0.187 1.02 — 19,031 Age 0 0.202 0.83 104.2 1,862 0.779 0.81 88.7 3,616 1 0.090 1.15 112.8 16,167 0.122 0.88 107.2 4,812 2 0.046 1.49 121.1 20,885 0.022 0.86 119.4 4,733 3 0.036 1.76 126.3 14,174 0.010 1.15 126.4 3,543 4 + 0.022 2.01 134.0 7,573 0.003 1.66 135.0 2,327 Total 0.057 1.48 121.2 60,661 0.187 1.02 113.7 19,031 Age 0 1 February-April 0 2,271 0.093 0.95 106.5 4,646 0.153 0.79 101.0 2 0.027 1.33 116.2 4,279 0.004 0.78 117.3 2,035 3 0.018 1.45 123.3 3,410 0.002 1.04 125.3 1,928 4 + 0.014 1.69 134.4 3,620 0.001 1.55 134.6 1,382 Total 0.041 1.30 119.0 15,955 0.048 0.96 117.6 7,616 Age August-December 0 0.259 0.77 104.2 1,204 0.831 0.78 88.6 3,216 1 0.083 1.19 117.6 6,411 0.104 0.97 112.2 2,198 2 0.060 1.59 123.1 12,082 0.035 0.91 120.5 2,311 3 0.051 1.92 127.7 7,648 0.021 1.28 128.1 1,255 4 + 0.039 2.33 132.9 2,532 0.007 1.57 135.4 560 Total 0.069 1.58 123.2 29,877 0.316 1.02 109.7 9,540 505 FISHERY BULLETIN: VOL. 84, NO. 3 40 O30 20 10- - 1 1 1 f 1 1 1 1 1 1 r JUL RUG SEP OCT NOV DEC JAN FEB MflR RPR HAY JUN MONTH Figure 1.— The monthly percentages of northern anchovies with unknown sex, by age group, in samples from the San Pedro fishery. duced essentially no change in the sex ratio in the sea survey data (1.02:1). However, there was a reduction in the proportion of females in the fishery data (1.48:1) which was associated with a reduction in the average age of anchovy in the catch after 1977 (Mais 1981). Sunada and Silva (1980) also found a female:male sex ratio greater than unity in the northern Baja California purse seine fishery, 2.15:1 in 1976 and 1.44:1 in 1977. Alheit et al. (1984) reported a sex ratio of 1.30:1 in purse seine caught Peruvian anchoveta sampled during their spawning season. Klingbeil (1978) and Alheit et al. (1984) reported that during the spawning season there were unexpectedly large numbers of samples in which sex ratios were heavily dominated by either males or females. Alheit et al. (1984) suggested that "hydrated females segregate, either by depth or area, from the 'normal' school, taking a high per- centage of males with them forming 'spawning schools' dominated by males." Analysis of the sex ratio by size and age groups shows that there are increasingly larger proportions of females in the larger and older groups (Table 2). This trend is evident in both the fishery and sea survey data. In the fishery data there are more males than females identified in the fish smaller than 100 mm SL and in age group 0. The proportion of females increases until there are more than twice as many females as males among fish larger than 130 mm and in age group 4 + . There is a similar trend in the sea survey data; however, females do not outnumber males until the fish are larger than 120 mm and 3 yr of age. The sex ratio in age group 4+ is 1.66:1. The apparent dominance of females in the larger size classes may be partially caused by sex related differences in growth rates; however, their dominance in the older age classes of both the purse seine and midwater trawl samples cannot be explained by differences in growth. We grouped our data sets into the spawning months (February- April) and nonspawning months (August-December) in order to evaluate features which might be caused by behavioral differences that may occur during the spawning season. This analysis shows that the over- all sex ratio in northern anchovy taken by midwater trawl is close to 1:1 in both nonspawning and spawn- ing seasons (Table 2B). It also shows that the sex ratios in younger fish are dominated by males and those in older fish are dominated by females. The overall sex ratio in northern anchovy sampled in the purse seine fishery is heavily dominated by females; however, the sex ratio is higher in the nonspawn- ing season (1.58:1) than in the spawning season (1.30:1). The crossover from male to female domi- nance of the sex ratio occurs between age group 2 and 3 in the sea survey data and at age 1 in the fishery data. MATURITY STAGES IN NORTHERN ANCHOVY Seasonal Variation in Maturity Stages To determine which of the various data sets avail- able for northern anchovy were best suited for evaluating maturity stages in the central stock, we examined the seasonality of three grouped matur- ity stages of four data subsets. The grouped stages included immature or resting females (stages 1 and 2); females just beginning to mature (stage 3); and the highly mature, spawning, and spent females (stages 4-7). The data consisted of two sets of samples from the commercial fishery (Monterey and San Pedro) and the sea survey samples from south- ern California (lat. 32.5°-34.5°N) and northern Baja California (29.5°-32.5°N). The seasonal patterns of the grouped maturity stages of females sampled in the San Pedro fishery (Fig. 2A), the sea survey in southern California (Fig. 2B), and the sea survey in northern Baja California (Fig. 2C) are quite similar. The pattern in the Mon- terey fishery differs from that in the other data sets in that spawning is at the highest levels in April and September (Fig. 2D). It cannot presently be deter- mined if there are one or two peaks of spawning in 506 PARRISH ET AL.: AGE DEPENDENT FECUNDITY IN NORTHERN ANCHOVY 90 80 70 £ 60H t— (§ 50 DC 40- 30- 20- 10- — i 1 1 1 1 i 1 1 1 1 i — JUL AUG SEP OCT NOV DEC JAN FEB MAR RPR MAY JUN MONTH 90 -\ 80 70 g B0H I— § 50 ie 40- 30- 20- 10- STAGE3 .4 i — •*-. ; * \ V \ •A \ /■. \ 1 STAGES t-7 '• \ •A -V — i 1 : l 1 1 1 1 1 r JUL RUG SEP OCT NOV DEC JAN FEB MAR RPR MAY JUN MONTH 90 80 72 -\ g 60 S 5H 40 30- 20- 10- STAGE 3 /r "-r --< \ / \ ^ STAGES 4-7 • ._r^ I I I 1 1 1 1 1 1 1 1 JUL RUG SEP OCT NOV DEC JAN FEB MAR RPR MAY JUN MONTH 90 H 80 70 g 60 u 50 H £ 40 30- 20- 10- — i 1 1 — JUL RUG SEP OCT NOV DEC JAN FEB MAR RPR MAY JUN MONTH Figure 2.— The monthly percentages of grouped maturity stages for female northern anchovies sampled in A, San Pedro fishery; B, Sea survey (lat. 32.5°-34.5°N); C, Sea survey (29.5°-32.5°N); D, Monterey fishery. the Monterey area due to the lack of data from May to July. Because of the different seasonal pattern of the grouped maturity stages in the Monterey fishery data and because there is more than one stock in this region (Vrooman et al. 1981), it was decided to exclude this area from further analysis. The analyses that follow are based on the combined San Pedro fishery and southern California-northern Baja California sea survey data sets (52,352 females of which 41,930 were aged). To obtain a first approximation of the magnitude and duration of maturity stages in the central stock of northern anchovy, the monthly percentages of fish with each maturity stage were calculated for females and for males. The seasonal patterns were found to be essentially the same for males and females; however, the males tended to have some- what larger percentages of fish in the higher matur- ity stages. Our presentation is limited to informa- tion on females. 507 FISHERY BULLETIN: VOL. 84, NO. 3 The proportion of females classified as stage 1, virgin individuals, is at a minimum during the spawning season, comprising <10% of the females sampled during February, March, and April. How- ever, during the summer more than half of the females were classified as stage 1 (Fig. 3A). Stage 2 females, maturing virgins and recovering spents, are at a minimum in August (when most females are classified as stage 1). From September until January between 40 and 60% of the females are classified as stage 2. During the spawning season, and just after, the percentage of stage 2 females dropped between 30 and 40%. The August to September decline in the percentage of stage 1 females is primarily caused by the sharp increase in the pro- portion of stage 2 females. Thus, the combined per- centage of stages 1 and 2 females is probably a reasonable inverse indicator of the seasonality of spawning. However, during the spawning season an unknown proportion of those classified as stages 1 and 2 are females that have recently spawned and are between multiple spawnings. Stage 3 females (ovary enlarged, occupying about half of the length of the ventral cavity) have a con- siderably different pattern. There is a gradual in- crease from about 5% in August to about 30% in January. This percentage is maintained until April, i.e., through the spawning season; it then drops to about 5% in June. The monthly percentages of the higher maturity stages (4,5, and 6) clearly delineate the spawning season as primarily a January-May event (Fig. 3B). The relatively constant low level of stage 7 females is unexpected as the maximum pro- portion of spent fish would be expected to occur just after the peak of spawning. Maturity Stage Relationships with Size and Age To examine potential relationships between the size and age of northern anchovy and the duration and magnitude of maturity stages we calculated the monthly percentages of grouped maturity stages for four size classes (81-100, 101-120, 121-140, and 141-160 mm SL) and four age groups (1, 2, 3, and 4 + ). Age group 1 includes fish prior to and after their first potential spawning season (i.e., young-of- the-year fish in July through the following June). Age group 4 + includes fish in their fourth and sub- sequent spawning seasons. The grouped maturity stages (1, 2, 3, and 4-7) are the same as those pre- sented earlier. Size has a large effect on both the duration and magnitude of maturity stages in northern anchovy. With the exception of those sampled from February to April nearly all of the 81-100 mm SL females were classified as immature or resting (Fig. 4A). In addi- tion, the majority of this size anchovy have gonads too small to determine their sex without magnifica- tion (Table 2). As the size class increases the per- centage of stages 1 and 2 decreases; this occurs in all months; however, the minimum percentage of 90- 80- 70 & 60 § 50H p 40- 30- 20- 10 STAGE 1 / \ STAGE 2 \ .• J N /» ' '■ \ t STAGE 3 -••.-■ —1—1 1 1 1 1 t ' 1 » / \ — V JUL HUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN MONTH -i 1 ■! 7 *-'i JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN MONTH Figure 3.— The monthly percentages of individual maturity stages for female northern anchovies. A. Stages 1-3. B. Stages 4-7. 508 PARRISH ET AL.: AGE DEPENDENT FECUNDITY IN NORTHERN ANCHOVY 90 80 70 U 60 t— £ 50 pe 40 30 20 1BH M-160 JUL RUG SEP OCT NOV DEC JflN FEB MRR RPR MAY JUN MONTH 90 80- 70- UJ g M1 i 50H 40- 30- 20- 10 B M-160 / ** \/x //">* M-160 / V / 101-120 \/ //•••- \ N — i — i — i — i — \ — JUL RUG SEP OCT NOV DEC JflN FEB MAR RPR MAY JUN MONTH Figure 4.— The monthly percentages of grouped maturity stages for female northern anchovies, by size group. A. Stages 1 + 2 B. Stage 3. C. Stages 4-7. stages 1 + 2 occurs in February to April. The per- centage of females just beginning to mature (i.e., stage 3) has an abrupt peak in February-March in the smallest size class (Fig. 4B). This peak becomes increasingly spread out in the larger size classes. The higher maturity stages (4-7) are most abundant from February to April in all size classes (Fig. 4C). The larger size classes have much larger percent- ages of females in the higher maturity stages than the smaller size classes, and there is a minor peak in the percentage of the higher maturity stages dur- ing the fall in the two largest size classes. Analysis of the data by age group showed, as would be ex- pected, that increased age has essentially the same effect as increased size on the magnitude and dura- tion of maturation stages. SPAWNING INCIDENCE AND FECUNDITY Studies by Hunter and Goldberg (1980) in Califor- nia and Alheit et al. (1983) in Peru examined post- ovulatory follicles to determine the spawning frequency of female anchovies (i.e., the time inter- val between spawnings). A second method would be to determine the percentage of females with 509 FISHERY BULLETIN: VOL. 84, NO. 3 hydrated eggs. Hunter and Macewicz (1980) showed that northern anchovy begin hydrating eggs at about 0600 in the morning and by sunset about 14% of the females have hydrated eggs. They felt that the best indicator of the time of spawning was the occurrence of both hydrated eggs and new postovulatory follicles. This occurred in a low percentage of their samples indicating that spawning was completed rapidly; the time of maximum spawning occurred between 2100 and 0200 with a peak between 2200 and 2300. Hunter and Macewicz (1980) divided the nightly pattern of spawning in anchovy into three periods: "early spawning period (1800 to 2100 hours), some spawning occurs but the ovaries of most reproductively active females are in the hydrated stage; maximum spawning (2100-0200 hours), most females spawn (females with hydrated eggs decline to 0 and females with new postovulatory follicles reach the maximum number for the night); and post- spawning (0200-0600 hours), little or no spawning occurs and females destined to spawn the next night begin hydration." A considerable amount of new histological data is now available as a result of a series of egg production biomass surveys for north- ern anchovy. B. Macewicz (fn. 3) has analyzed the histology of the ovaries of 8,672 anchovy sampled in these surveys, and our analysis of this new infor- mation verifies the temporal patterns which Hunter and Macewicz (1980) described from a much smaller sample (Fig. 5). Comparison of Maturity Stages and Histology Classes The histological data show that during the early evening the percentage of females with hydrated eggs could be an indicator of the percentage of females spawning per day (i.a, the spawning in- cidence). To use the extensive maturity stage data available for northern anchovy it is necessary to determine the relationships between the histology of the gonads and the maturity stages used in the California Department of Fish and Game's sampling programs. To date histological and field maturity stage data have not been taken on the same individ- uals; therefore, analysis is limited to comparisons of the two data sets. In the following comparisons the sea survey and histology data sets were limited to samples taken during the period 1977-84 and dur- ing the principal spawning season (i.a, February- April). Since nearly all of the trawls were taken at night, the data were limited to those taken from 1800 to 0500 h. The midwater trawl hauls were normally 15 min in duration, and about 30% of the fish in the JO en o in si is 10- 5- UJ d 20 15- £ 10- 5 - HYDRATED EGGS (H) t r DAY 0 P0STGVULAT0RY FOLLICLES (POF) I I -1 ' 1800 2000 2200 2400 200 400 TIME OF NIGHT Figure 5— The percentages of female northern anchovies with ovaries in three histological classes, by time of night. histological data set and 30% in the maturity stage data sets were taken in the same trawl hauls during cooperative cruises. The histological data are divided into six classes (B. Macewicz fn. 3): 1. Ovaries with hydrated eggs and no day-0 post- ovulatory follicles. 2. Ovaries with hydrated eggs and day-0 postovul- atory follicles. 3. Ovaries with day-0 postovulatory follicles and no hydrated eggs. 4. Ovaries with day-1 postovulatory follicles. 5. Mature ovaries with no hydrated eggs, no day-0 nor day-1 postovulatory follicles. 6. Immature ovaries, few or no yolked oocytes, no atresia present in the ovary other than late-stage corpora atretica. Northern anchovy, spawning on the night they were sampled (day 0), include the first three classes; those that spawned on the night before they were sampled (day 1) are class four. A comparison of the percentages of hydrated females in the sea survey data (i.e, stages 5 + 6) with that in the histological data (i.e, classes 1 + 2) shows that they have essentially the same pattern from the onset of spawning in the early evening until spawn- 510 PARRISH ET AL.: AGE DEPENDENT FECUNDITY IN NORTHERN ANCHOVY ing is completed in the early morning (Fig. 6). This implies that in the early evening maturity stages 5 + 6 can be used to estimate the spawning incidence; however, within a few hours after sunset the per- centage of females with hydrated eggs (i.e., stages 5 + 6) rapidly becomes an underestimate of the in- cidence of spawning due to the completion of spawn- ing. If only the females (n = 2,161) sampled between the hours of 1800 and 2000 are considered, then the percentage in maturity stages 5 + 6 (15.3%) is quite close to the percentage of day-0 females calculated for the total histology data set (15.9%). The variation throughout the night of the percent- ages of the other maturity stages is also of interest as it offers some insight into the meaning of maturity stages in anchovy. Hunter and Macewicz (1980) showed that spawning primarily occurs between the hours of 1800 and 0200. In the sea survey data the percentage of stages 5 + 6 falls from 15.3 to 1.6% over this time period (Fig. 7). The expected matur- ity stage that should increase over this time period is stage 7 (i.e., spents: ovaries slack with residual eggs). This, however, is not the case The percentage of stage-7 females has very little variation over the 1900-0200 period; going from 2% at 1900 to 3.6% at 0200. This suggests that residual eggs occur in only a small percentage of anchovy and that stage 7 cannot be used to determine if an anchovy has spawned within 24 h. This is consistent with Stauf- fer and Picquelle's (1980) observation that field- spawned northern anchovy were found to release nearly 100% of their hydrated eggs. The percentages of the other maturity stages show considerable varia- tion from 1900 to 0200 h. Stages in which the ovary is small (i.e., 1 + 2) occur in about 37% of the females in the early evening. This increases rapidly after 2300 and by 0200 these stages comprise about 46% of the females. Stages 3 + 4, in which the ovaries occupy from one half to two thirds of the ventral cavity, occur in about 46% of females in the early evening. This rises to a peak of about 54% at 2300- 2400 and then declines to about 49% at 0200. Our interpretation of the patterns exhibited by the sea survey data is that the percentage of females at stages 5 + 6 in the early evening (i.e., 15.3%) is a valid estimate of the percentage of sampled females with hydrated eggs. However, as the night progresses the percentage of stages 5 + 6 declines. At the peak of spawning, just before midnight, many females ap- pear to be misidentified as stages 3 + 4. This could occur if they had spawned part of their eggs before they were captured and if the person making the maturity stage determinations used the size of the ovary, rather than the presence of hydrated eggs, to determine the maturity stage After midnight an increasing percentage of spawning females have CO Ll_ O CO * s 10- 5- 20- LU CD F 15 en £ 10 HYDRATED EGGS STAGES 5+6 1 I 1 1 1 1 1 1 — T 1800 2000 2200 2400 200 TIME OF NIGHT 400 x: ■*• 20- u. o CO 10- cs ^ 50- i i i i i i i i i i i • — •■. ,.-'' "•«■.. STAGES 3+4 " -• ^-> At* CD 40 - d 1— z LlJ o 30- UJ Q_ /•STAGES 1+2 20- STAGES 5+6 10- •v. STAGE 7 \ ^ i ■ i i i i-i ■ ■ ■ • 1800 2000 2200 2400 200 TIME OF NIGHT 400 Figure 6.— The percentages of female northern anchovies with hydrated eggs and with maturity stages 5 + 6, by time of night. Figure 7.— The percentages of grouped maturity stages for female northern anchovies by time of night. (Two hour moving average.) 511 FISHERY BULLETIN: VOL. 84, NO. 3 apparently spawned all of their hydrated eggs and they are then classified as stage 1 or 2. The mean- ing of stage 7 in female anchovy remains a mystery. The seasonal and diurnal patterns described above indicate that, with the exception of stages defined by the presence of hydrated eggs, gross anatomical maturity stages have little utility other than describ- ing the seasonality of spawning. However, field iden- tifications of the presence of hydrated eggs, if they are calibrated with histological data and if the diur- nal pattern of hydration is known, can potentially be used to determine spawning incidence. Several authors have pointed out that females with hydrated eggs and actively spawning females were more numerous than females with day-1 postovula- tory follicles, and have suggested that hydrated and actively spawning females may be more susceptible to capture due to behavioral or physiological factors (Hunter and Goldberg 1980; Stauffer and Picquelle 1980; Alheit et al. 1984). Previous workers have therefore used the percentage of day-1 females as the index of the daily spawning incidence The over- all percentages of day-0 and day-1 females in the histology data set (8,672 females) used in our analyses are 15.9 and 11.5%. Alheit et al. (1983) found the overall percentages of day-1 and day-2 Peruvian anchoveta females to be 17.26 and 14.81%. Hunter and Goldberg (1980), and subsequent workers on the northern anchovy, took their samples at night whereas Alheit et al. (1983) took their Peru- vian anchoveta samples primarily during the day. Therefore the definition of day-1 is somewhat dif- ferent in studies of the two species. In our analysis both day-0 and day-1 females appear to be more susceptible to capture by midwater trawl in the early evening than later at night. The percentages of both decline as the night progresses; however, the decline is more extreme in the day-0 females (Fig. 8). The use of maturity stages 5 + 6 could result in several sources of bias that would tend to produce overestimates of the spawning incidence of northern anchovy. If females with hydrated eggs are more susceptible to capture, there will be a tendency to produce biased estimates. However, this bias would not be expected to be size or age dependent, nor would it be expected to vary during the spawning season. The same bias would be expected to occur in 1- and 4-yr-old hydrated females and the same bias would be expected in February and April. Therefore the use of the percentages of hydrated females or maturity stages 5 + 6 females may result in over- estimates of the total spawning incidence or annual fecundity, but the relative spawning incidence or relative annual fecundity of the different age groups would not be biased. A second source of bias is that an unknown number of females have ovaries so small that visual determination of sex is impossible with- out magnification. Therefore, the incidence of spawn- ing is overestimated because it is calculated by dividing the number of stages 5 + 6 females by less than the total number of females. This bias is size and age dependent, being much more common in smaller and younger anchovy, but not month depen- dent. Note that the various studies of the spawning incidence in northern anchovy and Peruvian an- choveta have defined the spawning incidence to be the number of females spawning per day divided by the number of mature females, i.e., these studies ex- clude immature females, which are primarily the smaller and younger fish, from the calculation. There are also several sources of bias that would tend to produce underestimates of the spawning in- cidence The anchovy fishery in southern California primarily occurs at night during the fall months and during the daylight hours in the spring. A period of low availability to the commercial fishery is asso- ciated with the spawning season. Mais (1974) asso- ciated this phenomenon with variation in schooling behavior and showed that acoustic surveys detect relatively few "commercial-sized" anchovy schools during the spawning season. If low availability to the commercial fishery is associated with spawning ac- t 1 — i 1 1 1 1 1 1 1 r 1800 2000 2200 2400 200 400 TIME OF NIGHT Figure 8.— The percentages of day-0 and day-1 female northern anchovies by time of night. 512 PARRISH ET AL.: AGE DEPENDENT FECUNDITY IN NORTHERN ANCHOVY tivity, it is probable that the fishery undersamples the active spawners. In addition, a proportion of the commercial catches occur during the time of day when the females do not have hydrated eggs. The fishery data will therefore tend to underestimate the spawning incidenca The total sea survey data will also produce an underestimate as it includes samples taken throughout the night. The combined fishery-sea survey data used in our analyses will therefore provide only an index of the daily spawning incidenca To evaluate the potential net bias of this index we calculated the percentage of females with maturity stages 5 + 6, in the com- bined fishery-sea survey data, and the percentage of females with day-1 postovulatory follicles, in the histological data. To make the data comparable we used the period 1977-84 and the months February- April. The percentage of females with maturity stages 5 + 6 and the percentage of females with day-1 postovulatory follicles was 10.9 and 11.5%. Use of the maturity stage data will therefore slightly underestimate the daily spawning incidence (i.a, 10.9/11.5 = 0.948). Size Dependent Batch Fecundity Annual fecundity in the northern anchovy is dependent upon the batch fecundity and the number of spawnings per year. Batch fecundity is size depen- dent and the best average estimate over six seasons (Hunter et al. 1985) is given below. Note that Hunter et al. found significant variation (ANOVA) among years. batch fecundity = -1,104 + 614 (WT) where WT = female wet weight, minus ovaries, in grams. During the spawning season ovary free weight of northern anchovy is equal to 95% of the total wet weight (Hunter and Leong 1981). Batch fecundity, with the above relationships, for a typical 1-yr-old (12 g) and a typical 4-yr-old (24 g) are 5,896 eggs and 12,895 eggs. On a per unit weight basis the 24 g fish would produce only 9.4% more eggs than the 12 g fish. Age-dependent variations in batch fecundity are therefore of only minor signifi- cance in the relationship between spawning biomass and annual fecundity. There is the possibility that batch fecundity could vary over the spawning season, and since we have shown an age-dependent seasonality in the spawning incidence of northern anchovy, this could potentially contribute to age- dependent differences in annual fecundity. Hunter and Leong (1981), however, found average batch fecundity to be essentially the same in samples taken in January-February and in March- April. Size-Dependent Histology Classes Hunter and Macewicz (1980) found no relationship between size and the percentage of mature female northern anchovy with day-1 postovulatory follicles. Later work by Picquelle and Hewitt (1984) showed that weight and spawning incidence were highly cor- related in the northern portion of the central stocks range. They stated that this implied that the larger females spawned more frequently or that the smaller females had a much shorter spawning season. We analyzed the larger histology data set now available and found that the percentages of females with hydrated oocytes or with day-1 post- ovulatory follicles, as well as the percentage of females with maturity stages 5 + 6, were depen- dent upon the size of the females (Fig. 9). CO E 40- u. o tn 20- HISTOfJGY MATURITY — -^— /STAGES ca *\ 20- "1 1 HYDRATE) 1 i PERCENTflGE ca en •' DAY1 *~— ^r^" STAGES 5+6 5 - . r , 1 i 80 81- 100 101- 120 121- 140 141- 160 LENGTH SLCMN; Figure 9.— The percentages of female northern anchovies with hydrated eggs, with maturity stages 5 + 6, and day-1 histological classes by size group. Age-Dependent Spawning Incidence and Annual Fecundity To assess age-dependent, annual fecundity in the central stock of northern anchovy we calculated the 513 FISHERY BULLETIN: VOL. 84, NO. 3 number of spawnings and the fecundity on a monthly basis for age groups 1, 2, 3, and 4 + . Average monthly wet weight by age was taken from Mallicoate and Parrish (1981). The number of spawnings per month was calculated from the num- ber of days per month and the index of daily spawn- ing (i.e., the proportion of stages 5 + 6). Note that the bias due to the unknown sex problem discussed earlier would tend to cause an overestimation of the daily spawning incidence: particularly in females in their first spawning season. Also note that the in- dex of daily spawning underestimates the spawn- ing incidence by about 5%. Our analysis shows that there are large, age- dependent variations in the proportions of female northern anchovy spawning as the spawning season progresses (Fig. 10). From July until January all age groups have a very low daily spawning index. Inten- sive spawning commences in February and all age groups have roughly the same spawning index (9-12%). In March the spawning index of age group 1 declines to about 2%; it increases slightly in April and declines to about 1% in May. In age group 2 the spawning index increases to 13% in March and then declines to about 2% by May. Age groups 3 and 4 + have peak spawning indices in March (25 and 27%), considerable spawning in April (10 and 17%) and lesser amounts in May (3 and 6%) and June (3 and 6%). Older females have a much larger number of spawnings per spawning season than younger ~i 1 1 1 r JUL RUG SEP OCT NOV DEC JflN FEB MflR RPR MAY JUN MONTH Figure 10.— The monthly percentages of female northern anchovies with maturity stages 5 + 6, by age group. females (Table 3). In their first spawning season females have an average of 5.3 spawnings. In their second spawning season this rises to 11.9 and in their third and fourth plus seasons the number of spawnings rises to 19.2 and 23.5. The increase in the number of spawnings associated with increas- ing age appears to be primarily due to the increase in the length of the spawning season that occurs in older fish. The average number of spawnings per season for all females sampled was 15.1. This is less than the estimate that Hunter and Leong (1981) developed from the energetics of female northern anchovy (i.e., 20 spawnings per year). Their calcula- tions indicated that mature female northern anchovy spawned on the average 15 times between February and September; their calculation of the number of spawnings from October to January (5) was esti- mated indirectly from the relative monthly larval abundance in 1953-60. Our estimate of the number of spawnings from February to September (14.3) is very close to the Hunter and Leong (1981) estimate which was based on a smaller histology data set. However, our estimate of the number of spawnings from October to January is only 0.8 and is much less than their indirect estimate based on the relative seasonal larval abundance for the 1953-60 period. The central stock of northern anchovy was at a much smaller population size in 1953-60 than it was in 1966-84 (MacCall 1980) and northern fish, with a seasonal spawning pattern similar to that occurring in the Monterey data, may have comprised a larger proportion of the anchovy population off California during the 1953-60 period than at present thus in- flating Hunter and Leong' s estimate for the October-January period. Our analysis indicates that annual fecundity in the central stock of northern anchovy is heavily age dependent; the average 4 + yr-old female produces nearly 10 times as many eggs as a 1-yr-old female (Table 3). Our calculations show that central stock, female anchovy produce 2,803, 6,550, 11,434, and 13,861 eggs/g of body weight per spawning season in their 1st, 2d, 3d, and 4th plus spawning seasons. Females 4 yr of age and older produce nearly 5 times as many eggs per unit of weight as 1-yr-olds. DISCUSSION Over the last decade it has become apparent that recruitment failure is the major threat to many of the world's largest fisheries. In addition, variation in recruitment is a significant causal factor in the interyear variation of the annual catches of many fisheries. Stocks of small pelagic fishes appear to 514 PARRISH ET AL.: AGE DEPENDENT FECUNDITY IN NORTHERN ANCHOVY Table 3.— Proportion of maturity stages 5 + 6, number of spawnings and fecundity of female northern anchovies sampled in the Sea Survey Program (lat. 29.5°-34.5°N) and San Pedro fishery. July1 Aug.1 Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June Total2 Eggs/g3 Prop. 5 + 6 Spawnings Wt. (g) No. eggs Prop 5 + 6 Spawnings Wt. (g) No. eggs Prop. 5 + 6 Spawnings Wt. (g) No. eggs Prop. 5 + 6 Spawnings Wt. (g) No. eggs Prop. 5 + 6 Spawnings 0.000 0.000 0.000 0.000 0.002 0.000 0.062 0.000 15.5 15.5 492 0 0.022 0.024 0.682 0.744 18.3 18.3 6,527 7,120 0.021 0.016 0.651 0.496 20.1 20.1 6,914 5,268 0.017 0.527 0.021 0.651 0.000 0.000 11.2 0 0.005 0.150 17.4 1,357 0.005 0.150 19.1 1,506 0.004 0.120 20.9 1,330 0.008 0.240 0.005 0.155 11.1 832 0.007 0.217 16.8 1,887 0.010 0.310 19.3 3,148 0.013 0.403 21.8 4,680 First spawning season 0.000 0.000 0.005 0.087 0.023 0.000 0.000 0.155 2.436 0.713 12.0 11.0 11.4 11.6 12.8 0 0 860 13,793 4,536 Second spawning season 0.001 0.001 0.015 0.110 0.132 0.030 0.031 0.465 3.080 4.092 17.2 16.3 16.2 15.6 16.5 268 261 8,881 24,626 34,866 Third spawning season 0.002 0.002 0.008 0.124 0.251 0.060 0.062 0.248 3.472 7.781 19.2 19.3 19.1 18.0 20.7 606 630 2,489 33,836 85,360 Fourth-plus spawning seasons 0.003 0.003 0.008 0.115 0.271 0.090 0.093 0.248 3.220 8.401 22.3 22.2 23.6 23.3 26.6 1,071 1,102 2,952 37,390 110,293 All spawning seasons combined 0.011 0.002 0.002 0.010 0.107 0.151 0.341 0.060 0.062 0.310 2.996 4.681 0.036 0.011 1 .080 0.341 13.7 15.4 7,438 2,687 0.065 0.021 1.950 0.651 17.7 18.3 17,980 6,230 0.101 0.031 3.030 0.961 22.2 20.9 35,891 10,655 0.166 0.065 4.980 2.015 26.5 25.7 67,123 26,454 0.004 0.120 13.6 819 0.020 0.600 17.5 5,462 0.026 0.780 22.7 9,467 0.056 1.680 25.7 18,136 0.094 0.044 0.012 2.820 1.364 0.360 5.3 32,514 2,803 11.9 102,174 6,550 19.2 205,819 11,434 23.5 322,957 13,861 15.1 'Missing data estimated from adjacent months, includes 5% correction for spawning incidence bias. 3Total eggs/February weight. be particularly susceptible to collapse; however, per- turbations of recruitment is a potential threat to any fishery in which one or two year classes comprise the bulk of the landings. The stock-recruitment ap- proach to understanding or predicting recruitment has fallen into disfavor, at least in the small pelagic fishes, because stock size has not proven to be a good predictor of recruitment. In its pure form (Bever- ton and Holt 1957; Cushing 1971; Ricker 1975) the stock-recruitment concept is based on two factors: 1) Parent stock size is a measure of the reproduc- tive potential of the stock, and 2) there are compen- satory mechanisms which reduce the number of recruits per spawner as the size of the parent stock increases. This compensation occurs through some mix of reduced fecundity of the parent stock, reduced growth of the recruiting cohort and in- creased mortality of the recruiting cohort. Recruit- ment variations are usually attributed to changes in environmental conditions, usually unknown, and the causal mechanisms, also usually unknown, are thought to occur during the early life history stages. The present emphasis of recruitment research is on the growth and mortality of the early life history stages. Potential variations of stock fecundity as a factor in recruitment variations has largely been ignored. There are now 6 years of egg production estimates available for the central stock of northern anchovy (Bindman 1985). The mean spawning incidence for these years is 0.124 and the spawning incidence varied from 0.094 in the El Nino year of 1983 to 0.160 in 1984. This implies that the central stock produced 70% more eggs, per unit of spawning biomass, in 1984 than in 1983. Santander (1980) showed that the Peruvian anchoveta had both re- duced spawning and an alteration of the seasonal- ity of spawning during the 1972 El Nino. The results presented here, which show that fecundity is strong- ly age dependent, suggest that the reduction in age composition caused by heavy exploitation will great- ly reduce the average fecundity per unit of biomass and also result in a reduction in the length of the spawning season. It appears that interyear varia- tions in the age composition of a stock or in en- vironmental factors associated with energy reserves or egg production are likely to alter greatly a stock's reproductive potential. If this is the case in other species which have multiple spawning, much of the variance in the stock-recruitment relationships of these fishes may be due to the fact that spawning biomass is a poor index of the reproductive poten- tial of the stock. To date information concerning age-specific reproductive potential has not been available for multiple spawning fishes because of the difficulty 515 FISHERY BULLETIN: VOL. 84, NO. 3 of determining the number of spawnings per year. The pioneering work by Hunter and Goldberg (1980) and later studies based on this work clearly demon- strate that, at least for many clupeids, the spawn- ing incidence or spawning frequency can be deter- mined with properly designed histological studies. Unfortunately a research program designed to determine age-specific reproductive potential would be very expensive as it would require a field sam- pling progam extending over the whole spawning season, in many cases the entire year; and it would require histological analysis and aging of a large number of females, both quite labor intensive. It appears that the only way it may be possible to determine age-specific reproductive potential for many fishes is to use the approach developed here which combines two methodologies: histological assessment of ovaries because it unambiguously and accurately measure spawning rate and a traditional fishery sampling program which utilizes an inexpen- sive rapid index of reproductive condition, such as the maturity stage system or a gonado-somatic index, in which thousands of specimens can be processed. Whichever anatomical grading system is used, its principal purpose would be to determine the percentage of hydrated females. Most of the maturity stages (i.e., 1-4, 7) in the system used for northern anchovy are only of value in describing the seasonality of spawning. The only stages (i.e., 5 and 6) which can be used to determine the number of spawnings are those in which the eggs are hydrated, and they can be directly used as an index in north- ern anchovy because it is known that the duration of hydrated eggs in the ovary is <24 h. The tradi- tional fishery sampling program may, as in the case for northern anchovy, already be available. If this is the case the principal work will be to calibrate properly the maturity stage or gonado-somatic in- dex with the histological analysis. For this approach to work the fishery must, of course, take hydrated females. CONCLUSIONS It is important for those managing fisheries which are susceptible to recruitment overfishing to realize that the alteration in the age structure of a stock that occurs under heavy exploitation may have greater effects on the total fecundity and seasonality of spawning than previously recognized. Manage- ment strategies which decrease the exploitation of older, more fecund females could increase yields and also provide increased protection against recruit- ment overfishing. In northern anchovy there is the additional factor that the sex ratio in the fishery is age dependent (i.e., the female:male ratio for 1-yr- old anchovy in the San Pedro fishery is 0.83:1, whereas that for 4+ yr-olds is 2.01:1). When this factor is multiplied by the difference in the fecun- dity of the two age groups, it is apparent that the catch of a ton of 4 + yr-old northern anchovy reduces the reproductive potential of the stock 7.3 times as much as the catch of a ton of 1 -yr-old fish. ACKNOWLEDGMENTS We gratefully acknowledge John Hunter and Beverly Macewicz for allowing us to use their exten- sive data on the histology of the ovaries of northern anchovy and Carol Kimbrell for providing us with the computer files. John Hunter also provided con- siderable input to the development of the work and edited the manuscript. We would also like to thank Eric Knaggs, Eugene Fleming, and John Sunada for providing us with the anchovy fishery data and Ken- neth Mais for providing the sea survey anchovy data. LITERATURE CITED Alheit, J., B. Alegre, V. H. Alarcon, and B. J. Macewicz. 1983. Batch fecundity and spawning frequency of various an- chovy (Genus: Engraulis) populations from upwelling areas and their use for spawning biomass estimates. FAO Fish. Rep. 291, 3:977-985. Alheit, J., V. H. Alarcon, and B. J. Macewicz. 1984. Spawning frequency and sex ratio in the Peruvian an- chovy, Engraulis ringens. CalCOFI Rep. 25:43-52. Beverton, R. J. H., and S. J. Holt. 1957. On the dynamics of exploited fish populations. Fish. Invest. Lond. Ser. 2, 19:1-533. Bindman, A. G. 1985. The 1985 spawning biomass of the northern anchovy. U.S. Dep. Commer., NOAA, NMFS, SWFC, Admin. Rep. LJ-85-21, 21 p. Bowers, A. B., and F. G. T. Holliday. 1961. Histological changes in the gonad associated with the reproductive cycle of the herring (Clupea harengus L.). Mar. Res. Scot. 5, 16 p. Collins, R. A., and J. D. Spratt. 1969. Age determination of northern anchovies, Engraulis mordax, from otoliths. Calif. Dep. Fish Game, Fish Bull. 147, p. 39-55. Cushing, D. H. 1971. The dependence of recruitment on parent stock in dif- ferent groups of fishes. J. Cons. Perm. Int. Explor. Mer 33:340-362. Hunter, J. R., and S. R. Goldberg. 1979. Spawning incidence and batch fecundity in northern anchovy, Engraulis mordax. Fish. Bull., U.S. 77:641-652. Hunter, J. R., and R. Leong. 1981. The spawning energetics of female northern anchovy, Engraulis mordax. Fish. Bull, U.S. 79:215-230. Hunter, J. R., N. C. H. Lo, and R. J. H. Leong. 1985. Batch fecundity in multiple spawning fishes. In R. 516 PARRISH ET AL.: AGE DEPENDENT FECUNDITY IN NORTHERN ANCHOVY Lasker (editor), An egg production method for estimating spawning biomass of pelagic fish: application to the northern anchovy, Engraulis mordax, p. 67-77. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 36. Hunter, J. R., and B. J. Macewicz. 1980. Sexual maturity, batch fecundity, spawning frequency, and temporal pattern of spawning for the northern anchovy, Engraulis mordax, during the 1979 spawning season. CalCOFI Rep. 21:139-149. Klingbeil, R. A. 1978. Sex ratios of the northern anchovy, Engraulis mordax, off southern California. Calif. Fish Game 64:200-209. MacCall, A. D. 1980. Population models for the northern anchovy (Engraulis mordax). Rapp. P. -v. Reun. Cons. Perm. int. Explor. Mer 177:292-306. Mais, K. F. 1974. Pelagic fish surveys in the California Current. Calif. Dep. Fish Game, Fish Bull. 162. 79 p. 1981. Age-composition changes in the anchovy, Engraulis mordax, central population. CalCOFI Rep. 22:82-87, Mallicoate, D. L., and R. H. Parrish. 1981 . Seasonal growth patterns of California stocks of north- ern anchovy, Engraulis mordax, Pacific mackerel, Scomber japonicus, and jack mackerel, Trachurus symmetricus. CalCOFI Rep. 22:69-81. Parrish, R. H., D. L. Mallicoate, and K. F. Mais. 1985. Regional variations in the growth and age composition of northern anchovy, Engraulis mordax. Fish. Bull., U.S. 83:483-496. Picquelle, S. J., and R. P. Hewitt. 1983. The northern anchovy spawning biomass for the 1982-83 California fishing season. CalCOFI Rep. 24:16-28. 1984. The 1983 spawning biomass of the northern anchovy. CalCOFI Rep. 25:16-27. Ricker, W. E. 1975. Computation and interpretation of biological statistics offish populations. Fish. Res. Board Can. Bull. 191, 382 p. Santander, H. 1980. Fluctuaciones del desove de anchoveta y algunos fac- tores relacionados. In IOC Workshop Rep. 28, UNESCO, Paris, p. 255-274. [Workshop on the effects of environmen- tal variation on the survival of larval pelagic fishes, Lima, Peru. 20 April-5 May 1980.] Stauffer, G. D., and S. J. Picquelle. 1980. Estimates of the 1980 spawning biomass of the cen- tral subpopulation of northern anchovy. U.S. Dep. Com- mer., NOAA, NMFS, SWFC, Admin. Rep. LJ-80-09, 24 p. SUNADA, J. S., AND S. SlLVA. 1980. The fishery for northern anchovy, Engraulis mordax, off California and Baja California in 1976 and 1977. CalCOFI Rep. 21:132-138. Vrooman, A. M., P. A. Paloma, and J. R. Zweifel. 1981. Electrophoretic, morphometric, and meristic studies of subpopulations of northern anchovy, Engraulis mordax. Calif. Dep. Fish Game 67:39-51. 517 SOME STATISTICAL TECHNIQUES FOR ESTIMATING ABUNDANCE INDICES FROM TRAWL SURVEYS Michael Pennington1 ABSTRACT Methods are presented for estimating an index of relative abundance from trawl survey catch per tow data. The estimated variance of the index takes into account the within survey variability in catch and possible yearly changes in catchability. Applying the techniques to a series of surveys for yellowtail flounder, Limanda ferruginea, off the northeast coast of the United States yields an abundance index with a variance which is 40% lower than the variance of the original survey index for the current value and 57% lower for values not near the ends of the survey series. The average number of fish caught per tow during a trawl survey is often used as an index of a species's relative abundance (Grosslein 1969; Clark 1979). Catch per tow data are usually quite variable because of the heterogeneous distribution of many fish stocks (Byrne et al. 1981). A further source of variability for survey indices of abundance is that the catchability of a particular species with respect to the survey trawl may change from year to year (Byrne et al. 1981; Collie and Sissenwine 1983). As a result, the observed time series of abundance in- dices reflects changes in the population, within survey sampling variability, and varying catchabil- ity over time. This paper uses various statistical methods to con- struct from the catch per tow data an index of abun- dance which more closely tracks the population than does the original (average catch per tow) series. Specifically, since the distribution of catch per tow data is often highly skewed and contains a propor- tion of zeros, estimates of the mean catch per tow for each survey are made based on the A-distribution (Aitchison and Brown 1957). Next, time series tech- niques are used to estimate the component of the series generated by the actual changes in the population. The methods are applied to data for yellowtail flounder, Limanda ferruginea, from a series of groundfish trawl surveys conducted off the north- east coast of the United States as part of the National Marine Fisheries Service's MARMAP pro- gram. The resulting index of abundance is substan- tially more precise than the original index. Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. STATISTICAL METHODS Sources of Variability Let yt denote the observed average catch per tow for the survey conducted in year t and z\ = E[yt], the expected value of yt. Since a species catchabil- ity may change from year to year with respect to the survey trawl, let z = E[z'\p] denote the expected value of z given a population level p. Then y, = zt + et. The error term, et, can be expressed as et = (Vt ~ z't) + (z't ~ Zt), where the first error component is due to the within survey variability and the second is due to changes in catchability. In order to construct an index of abundance, it is necessary to assume a functional relationship be- tween zt and pt. A reasonable assumption made in practice (and in this paper) is that zt = apt. If the relationship is not linear, then the unadjusted catch per tow index will be a biased measure of relative abundance. Estimating the Mean Catch per Tow The distribution of marine survey data often can be described by what is called a A-distribution (Ait- chison and Brown 1957). That is, the data contain Manuscript accepted October 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 519 FISHERY BULLETIN: VOL. 84, NO. 3 a proportion of zeros and the nonzero values are distributed lognormally. The minimum variance un- biased estimates of the mean, c, and its variance, var(c), for the A-distribution are given by (Penning- ton 1983), c = and var(c) = ra n n' 0, exp(^) Gm(s2/2), ra > 1, ra = 1, ra = 0, (1) f ™ exp(2^) = <*<**> •■feff x Gr Im - 2 m - 1 , m > 1, ra = 1, (2) EFFICIENCY OF x Figure 1.— The efficiency of x and s2 (the sample mean and variance, respectively) for the A-distribution with 50% zeros. o, ra = 0, where n is the number of tows, m is the number of nonzero values, y and s2 are the sample mean and variance respectively of the nonzero logf values, x1 is the single (untransformed) nonzero value when ra = 1, and GJx) = 1 + + I ra 1 ra x (m - l)2-?-1 x> j=2 jyO (ra + 1) (ra + 3). . .(ra + 2j> - 3) j\ ' The series defining Gm{x) is a function of x [e.g., # = s2/2 in Equation (1)] and ra which is easily evaluated for particular values of x and ra using a computer. Figure 1, which is an extension of a graph in Ait- chison and Brown (1957, p. 98), shows the large sam- ple efficiency of the ordinary sample statistics as compared with their most efficient estimates for the A-distribution with 50% zeros. Estimates of a2, the variance of the nonzero loge values, are often be- tween 1 and 2 for trawl surveys. Thus (Fig. 1) the sample mean is a fairly efficient estimator of the mean for trawl surveys, but the sample variance is highly inefficient. Though for larger values of o2, which, for example, are common for egg surveys (Pennington and Berrien 1984), the sample mean is also very inefficient. It does not follow that the variance of c is necessarily small, but it is smaller, and as o2 increases, much smaller than the variance of the sample mean. However, it should be noted that if the sample variance is used to estimate the variance of the sample mean for moderate sample sizes because of the inefficiency of the sample variance, the estimated variance of c will often be greater than the estimated variance of the sample mean. Estimating the Index of Abundance As an index of abundance, the series of yearly catch per tow estimates, yh (based, e.g., on the A- distribution theory if appropriate) has two draw- backs. First, its estimated variance when derived from the within survey variance can be an under- estimate since catchability may vary from year to year. The second and more serious deficiency is that the index for a particular year is based only on that year's survey which disregards relevant information contained in the surveys for other years. 520 PENNINGTON: TECHNIQUES FOR ESTIMATING ABUNDANCE One method to construct an abundance index based on the entire survey series is briefly as follows (more details can be found in Pennington (1985)). Suppose the population (or zt) can be represented by the autoregressive integrated moving average process (Box and Jenkins 1976, Chap. 4) 0(5) zt = 0(5) at. where the at's are independently identically distrib- uted {iid) and normally distributed (N) with mean zero and variance o2 [iid N(0, o2)]. If yt = zt + et, and the et's are assumed iid N(0, o2), then yt will follow the model 0(5) yt = r,(B) ct, (3) Suppose the factors causing the change in popula- tion from year t - 1 to year t (such as recruitment, fishing mortality, natural mortality, and migrations) produce at's which are approximately iid N{0, o2). If the measurement errors are multiplicative, then In yt = In zt + et. (8) Assuming the e/s are iid N(0, o2) and independent of the a/s, then it follows as above that yt can be represented by the model (1 - 5) In yt = (1 - 05) ct. (9) where the ct's are iid N(0, of) For model (9) [generated by Equations (7) and (8)] where the ct's are iid N(0, o2). Now if model (3) and the ratio of/of are known, then the maximum and likelihood estimate of zt is given by 0 = o*M 2/„2 (10) (1 - 0)2 = olio &t = Vt ~ -z(Ct - nj ct+1 - TIo C 2 W + 2 ~ ) • • • » nT_t cT), (4) where T denotes the last year of the series, the ct's are the estimated residuals generated by model (3), and the n values are calculated using the identity 0(5) = (1 - nxB - n252 - . . .) r]{B). (5) The variance of zt is given approximately by varfo) = o ^ „2 1 - (n§ + 4 2 \ °e: (6) where rc0 = 1. The model for yt [Equation (3)] is usually ob- tained in practice by fitting a model to the observed series using procedures described in Box and Jenkins (1976). If catchability is constant over time, the within survey sampling variance provides an estimate of oez. But if catchability varies, another approach is necessary. Toward this end, consider the expression "t-i or (1 - 5) In zt = at. (7) Therefore, assuming the above approximations to the population dynamics, fitting model (9) to the observed survey series provides an estimate, 0, of o^lol and an estimate of o2. The it-weights for the model are from Equation (5) given by = (1 - 0) 0< i > 1. (11) It may be noted that if model (9) is valid and catch- ability is constant over time then the estimate of o2 given by 0 d2 [from Equation (10)] would approx- imately equal the estimate of o2 based on the within survey sampling variance. AN APPLICATION The Northeast Fisheries Center conducts an extensive groundfish trawl survey as part of its MARMAP program two times a year: in the fall since 1963 and in the spring since 1968 (Grosslein 1969). The survey region is divided into sampling strata based on geographic boundaries and depth contours (Fig. 2). For each survey, trawl stations are chosen randomly within each stratum. One of the objectives of the surveys is to provide indices of abundance for the many species of commercial value in the region. Yellowtail flounder is an important New England fishery resource whose population has fluctuated considerably over the survey period (Clark et al. 1984). Commercial catch statistics exist for yellow- tail flounder, but age data suitable for a VPA (Vir- tual Population Analysis) are unavailable. Major 521 FISHERY BULLETIN: VOL. 84, NO. 3 CD > 3 PL, < s OS > 03 CD g C El) H I w a! D o 522 PENNINGTON: TECHNIQUES FOR ESTIMATING ABUNDANCE yellowtail flounder fisheries are off southern New England (strata 5, 6, 8, 9) and on Georges Bank (strata 13-21). The two stocks are fairly distinct but with some intermixing (Clark et al. 1984). The nonzero catch per tow survey data for yellow- tail flounder are approximately lognormally distrib- uted within a stratum. Therefore, the estimators based on the A-distribution [Equations (1) and (2)] were used to estimate the mean catch per tow and its variance in each stratum. The regional estimates for southern New England and Georges Bank were then calculated in the usual manner for each survey (see, e.g., Pennington and Brown 1981). Model (9) was fit to each series (spring 1968-84 and fall 1963-84 in both regions) and the model's adequacy checked (Box and Jenkins 1976, Chap. 8). Table 1 contains summary statistics and parameter Table 1.— Summary statistics and parameter estimates for the yellowtail flounder survey series. The first three sample autocorre- lations (rv r2, and r3) are for the first differenced logged series. No. of Survey years '1 r2 r3 0 SE(fl) oi Southern New En gland Spring 17 -0.23 0.12 -0.18 0.21 0.28 0.57 Fall 22 -0.26 0.07 -0.31 0.40 0.22 0.71 Georges Bank Spring 17 -0.32 0.00 -0.09 0.61 0.23 0.36 Fall 22 -0.30 -0.06 0.18 0.36 0.23 0.33 Average -0.28 0.03 -0.10 0.40 ^.12 0.50 'Assuming the estimates of 8 are independent. estimates for the four series. Since the series are relatively short, the averages of the areal and seasonal estimates are used as the final estimates of 0 and o2c (last line in Table 1). Abundance indices for the two regions and seasons were calculated by applying to each series Equation (4) with 6 = 0.4, the rc-weights given by Equation (11), and the ct's (for each series) gener- ated by model (9). An estimate of 6] equal to 0.20 and of o\ equal to 0.18 were obtained from Equa- tion (10). The estimated variance of the index equals, from Equation (6), 0.12 for the current value and declines to 0.09 for values not near the series' end points. This compares with a variance of 0.20 ( = of) for the original index. Figures 3 (log scale) and 4 (linear scale) show plots of the estimated index and the observed catch per tow series for the fall sur- veys off southern New England. DISCUSSION The major advantage of estimating an index of abundance from the entire survey series is that it can produce an index with a variance considerably smaller than the variance of the observed series. But the application also demonstrates that estimates of the accuracy of an index based only on the within survey sampling variance can be misleading. For ex- ample, the 1972 survey value for yellowtail flounder off southern New England is considered an anom- o 1 — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i 1963 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 YEAR Figure 3.— Logged average catch per tow and the estimated index of abundance for southern New England yellowtail flounder. 523 FISHERY BULLETIN: VOL. 84, NO. 3 5 o 80-1 70 60- o 50 a. .e o T3 40 0) ro 30 > < 20 10 Survey catch perjow _ Survey index o( abundance N / "1963 64 6 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 YEAR Figure 4.— Average catch per tow and the estimated index of abundance for southern New England yellowtail flounder. aly (Collie and Sissenwine 1983). It does appear anomalous if comparisons are made using 0.11, the estimated variance based on the within survey variance, but not if the estimate of 0.20 (= of) is considered (Fig. 3). Assessing the accuracy of an index of abundance for marine stocks is difficult since the true levels are never known with certainty. But they can be compared with other indicators of abundance. The methods were applied to the haddock stock on Georges Bank (Pennington 1985) for which a VPA exists. It was found that model (7) adequately describes the dynamics of the VPA series, and the survey series follows model (9). The resulting index of abundance is quite similar to the VPA estimates. Collie and Sissenwine (1983) give a method for estimating the relative abundance of a fish stock using survey data and commercial catch statistics. They observe that their method produces estimates which compare favorably with VPA estimates. Figure 5 shows plots of Collie and Sissenwine's estimate of the relative abundance of southern New England yellowtail flounder and the index based only on the survey data. Finally, it should be noted that the purpose of the modeling stage in the estimation procedure is not necessarily to develop a realistic model for the population, but to describe the important stochastic properties of the series. As the observed series becomes longer, more precise estimates can be made. For shorter series, given the large variabil- ity inherent in marine trawl surveys, a preliminary estimate of between 0.3 and 0.4 for the smoothing parameter 6 appears to be an appropriate initial value to use for estimating an abundance index un- til more information becomes available. LITERATURE CITED AlTCHISON, J., AND J. A. C. BROWN. 1957. The lognormal distribution. Cambridge Univ. Press, Lond., 176 p. Box, G. E. P., and G. M. Jenkins. 1976. Time series analysis: forecasting and control. Rev. ed. Holden-Day, San Franc, 575 p. Byrne, C. J., T. R. Azarovitz, and M. P. Sissenwine. 1981 . Factors affecting variability of research trawl surveys. Can. Spec. Publ. Aquat. Sci. 58:238-273. Clark, S. H. 1979. Application of bottom trawl survey data to fish stock assessment. Fisheries 4:9-15. Clark, S. H., M. M. McBride, and B. Wells. 1984. Yellowtail flounder assessment update. U.S. Dep. Commer., NOAA, Natl. Mar. Fish. Serv., Woods Hole Lab. Ref. Doc. No. 84-39, 29 p. Collie, J. S., and M. P. Sissenwine. 1983. Estimating population size from relative abundance data measured with error. Can. J. Fish. Aquat. Sci. 40: 1871-1879. Grosslein, M. D. 1969. Groundfish survey program of BCF Woods Hole. Comm. Fish. Rev. 31(8-9):22-30. Pennington, M. 1983. Efficient estimators of abundance, for fish and plank- ton surveys. Biometrics 39:281-286. 1985. Estimating the relative abundance of fish from a series 524 PENNINGTON: TECHNIQUES FOR ESTIMATING ABUNDANCE 80-1 YEAR Figure 5.— Survey index of abundance (solid line) and Collie and Sissenwine's index (broken line) for southern New England yellowtail flounder. of trawl surveys. Biometrics 41:197-202. Pennington, M., and P. Berrien. 1984. Measuring the precision of estimates of total egg pro- duction based on plankton surveys. J. Plankton. Res. 6: 869-879. Pennington, M., and B. E. Brown. 1981 . Abundance estimators based on statified random trawl surveys. Can. Spec. Publ. Aquat. Sci. 58:149-153. 525 RATES OF INCREASE IN DOLPHIN POPULATION SIZE Stephen B. Reilly and Jay Barlow1 ABSTRACT Annual finite rates of increase in dolphin population size were estimated to vary up to a maximum of 1.09, using simulation, based on ranges in vital rates. Vital rate ranges were defined from values reported in the literature where possible, otherwise by making assumptions about biological or logical limits. Given information on current values, or limits, of one or more vital rate, one can use the figures presented to determine ranges of possible rates of increase in population size. The highest rates estimated here (up to 1 .09) are probably unrealistic, because of the unlikely combinations of high fecundity and low mor- tality needed to achieve them. Rates of increase in population size are important in determining management strategies for fish and wildlife subject to exploitation. A common manage- ment approach for setting incidental mortality or harvest quotas is to use a stock-production model (Schaeffer 1957; Allen 1976) with an assumed max- imum rate of increase. For dolphins and other ceta- ceans, rates of increase have proven extremely dif- ficult to measure directly. Nonetheless, estimates of this parameter are sometimes necessary, e.g., in setting incidental mortality quotas for dolphin populations involved in the eastern tropical Pacific purse seine fishery for yellowfin tuna (Smith 1983). In such situations, even a range, when rigorously defined, can contribute substantially to delineating the management options. In this paper we define a range of reasonable values of rate of increase (hereafter also referred to as ROI) in dolphin population size, given what is known or can be inferred about their age-specific survival and fecundity distributions, or "vital rates". We estimate rates of increase using population pro- jection matrices for various parameter combina- tions. We also suggest how the resulting ranges in ROI can be further narrowed, given specific infor- mation for an individual population. There are many slightly different definitions for rate of increase, but all share the commonsense no- tion of change in population size over time. Caughley (1977) reiterated the distinction between exponen- tial and finite rates: finite rates, here symbolized A, are related to exponential rates, here symbolized r, by the simple conversion A = er. (We use the term Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. "finite rates of increase" for A following Birch 1948.) Further, within exponential rates Caughley distin- guished among "intrinsic" (rm), "survival-fecun- dity" (rs) and "observed" (r), rates. In this paper we compute a series of rs values, resulting from ranges of survival-fecundity distribu- tions. The highest value of rs resulting from the range of vital rates considered is our best estimate of dolphin rm, or "r-max". We define the ranges in vital rates based on the literature for dolphins where possible. Otherwise, we rely on information for other large mammals and what appear to be logical or biological limits. There are two previous studies of a similar nature for delphinids. As part of a general review of life history analysis of large mammals, Goodman (1981) examined the relationships among rate of increase, juvenile and adult survival rates. He looked at single values for calving interval and age at first reproduc- tion across ranges of survival rates. We take a broader look at these relationships, examining ranges for all four parameters. Polacheck (1984) examined interparameter rela- tionships for eastern tropical Pacific (ETP) dolphins, Stenella spp., given specific vital rate estimates available as of 1981, showing the values were not consistent with a positive population growth rate. Since then, revised estimates have become available for some relevant parameters, and this specific case has been reanalyzed, with similar general conclu- sions. The only reported dolphin rates of increase are for Stenella coeruleoalba. For the year 1974, Kasuya (1976) estimated a rate of 0.024 for the population off Japan. This value was computed in a complex manner, based on an observed fishing mortality, assumed natural mortality, and estimated popula- Manuscript accepted October 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 527 FISHERY BULLETIN: VOL. 84, NO. 3 tion size, calving interval and sex ratio. Assuming that calving interval was density dependent, Kasuya (1976) estimated a maximum annual rate of increase of 0.044 for this population of 5. coeruleoalba. METHODS The Model Population growth rates are estimated here using the familiar Leslie matrix model (Leslie 1945). A simplified parameterization is used for which sur- vival rates and fecundities remain constant over many age classes. Four parameters are required: 1) calving interval for reproductively mature females, 2) average age at first birth for females, 3) annual adult (noncalf) survival rate, and 4) annual calf sur- vival rate. This degree of detail corresponds to the practical limitations in collecting data on wild dolphin stocks. The model is constructed with the assumption that age class 1 corresponds to newly born calves (i.e., censuses occur immediately after the calving sea- son). In fact, the model is not dependent on discrete calving seasons, but this assumption helps in con- ceptualizing some elements of the model. The fecun- dities (elements of the first row of the Leslie matrix) represent the number of female calves born in one year per female of a given age class in the previous year. Fecundities for mature age classes are esti- mated as the annual pregnancy rate (the inverse of calving interval) multiplied by the adult survival rate (the probability that a [pregnant] female will sur- vive to the calving season) multiplied by 0.5 (the frac- tion of female offspring). The annual pregnancy rate is estimated as the percent of sexually mature females which are pregnant, divided by the gesta- tion period (in years). The choice of only two different survival rates for all life stages was made because of data limitations for dolphins. Perhaps a more biologically reasonable assumption would be that dolphins have a U-shaped mortality curve which is characteristic of mammals in general (Spinage 1972; Caughley 1977; Siler 1979; Smith and Polacheck 1981). Barlow2 incorporated this typical mammalian survivorship curve in models of growth for spotted dolphins, Stenella attenuata. Our choice of a separate survival rate for calves was based on the common observation of higher mortal- ity in juvenile mammals (Caughley 1977; Siler 1979). For convenience, juvenile mortality factors are com- pressed into the first year's survival rates. This simplification is justified because population growth rates do not depend on the age at which juvenile mortality actually occurs. We recognize that juvenile mortality factors probably extend past the first year of life, but insufficient data exist to justify including this in our model. Higher mortality in old age was not incorporated in our model, but maximum age was limited to 50 yr. The survival rate at age 50 was thus zero. We calculate population growth rates for a range of the four vital rate parameters mentioned above. Finite population growth rates, A, that are associ- ated with these parameter values were calculated by solving Lotka's characteristic equation, using Newton's method. The explicit form of Lotka's equa- tion used is 50 1 = Z A~*l x=l x ™>x 2Barlow, Jay. 1986. Biological limits on current growth rate of a spotted dolphin population (Stenella attenuata). Unpubl. manuscr. Southwest Fisheries Center La Jolla Laboratory, Na- tional Marine Fisheries Service, NOAA, 8604 La Jolla Shores Drive, La Jolla, CA 92038. where lx is the survivorship from birth to age class x and mx is the fecundity of age class x. Below, we define the ranges used for the four population parameters and describe how they were selected. Survival Rates Ranges in Noncalf Survival Rates Few estimates of adult survival rates for dolphins are available in the literature, primarily because ade- quate data are difficult to collect. Kasuya (1976) pre- sented annual survival rate estimates of 0.925 and 0.882 for exploited populations of Stenella attenuata and S. coeruleoalba, respectively; however, his method (log-linear regression) is biased (Barlow 1982), and he did not adjust for the effect of popula- tion growth on age structure. A range of 0.85 to 0.97 was chosen for survival rates in this study. Values <0.85 do not allow population growth for the ranges of other parameters appropriate here, hence these values were not considered. Values higher than 0.97 result in more than 22% of the population being over 50 yr old. This is inconsistent with estimates of longevity for delphinids based on tooth layer counts [58 yr in S. coeruleoalba (Sacher 1980), 38 yr in S. attenuata (Hohn and Myrick3)], hence values 3Hohn, A. A., and A. C. Myrick, Jr. 1986. Age distribution of the kill of spotted dolphins in the eastern tropical Pacific. 528 REILLY and BARLOW: INCREASE IN DOLPHIN POPULATION >0.97 are untenable as mean per-capita survival rates. Ranges in Calf Survival Rate Again little information is available on calf sur- vival for dolphins. Kasuya (1976) estimated a juven- ile survival rate that was higher than that of adults, based on a balance equation. His methods assume that populations are neither growing nor declining, and he did not show that this assumption was met. Also his juvenile period included all sexually im- mature age classes. The overwhelming body of evidence from terrestrial mammals is that very early juvenile mortality is higher than adult mortality (Spinage 1972; Caughley 1977; Siler 1979). Even human populations had a first year survival rate of <0.88 prior to modern antibiotics (Fruehling 1982, data for U.S. circa 1900). An upper limit on calf sur- vival rates was generated by assuming a calf is ab- solutely dependent on its mother for 1 yr. A calf has the same risk of dying as an adult, plus the addi- tional risk of dying of starvation if its mother dies before completing 1 yr of lactation. The upper limit on calf survival would thus equal the square of the adult survival rate. The lower limit on calf survival rates was chosen as 0.50, a value that seems typical of pinnipeds (Smith and Polacheck 1981) and long- lived terrestrial mammals (Spinage 1972). Fecundity-Related Rates Ranges in Calving Interval Observed calving intervals for dolphins general- ly range from 2 to 4 yr (Perrin and Reilly 1984); con- sequently, we have used this range in our computa- tions. Intervals reported for killer whales (which are also delphinids, but not "dolphins") are considerably longer, up to 8 yr (e.g., Jonsgard and Lyshoel 1970). The literature includes reports of calving inter- vals <2 yr for dolphins. These reports do not appear to be valid. Reevaluation of data for three of these reports4 indicates that sampling was biased to- ward pregnant females (Perrin and Reilly 1984), a result of what may be a general tendency for Unpubl. manuscr. Southwest Fisheries Center La Jolla Labora- tory, National Marine Fisheries Service, NOAA, 8604 La Jolla Shores Drive, La Jolla, CA 92038. 4Three reported cases of dolphin calving intervals <2 yr, later found to be biased due to age and sex segregation, are Black Sea Delphinus delphis and Tursiops truncatus (KJeinenberg 1956) and Western Pacific Stenella coeruleoalba (Miyazaki and Nishiwaki 1978). dolphins to segregate by age/sex groupings5. The remaining reports of calving intervals <2 yr are from very small sample sizes.6 Gestation periods for dolphins are at minimum 10 mo, and intraspecific variation is small. Reported lactation periods range from 1 yr to over 2 yr (Perrin and Reilly 1984). Sum- ming these two periods gives another indication that dolphin calving intervals are not likely to be <2 yr. An exception to the 2-yr minimum calving inter- val would possibly be in a population experiencing very high calf mortality, causing premature cessa- tion of lactation, and allowing females the opportun- ity to begin a new calving cycle (assuming there was no seasonality to breeding which could require a resting period before the next breeding season). To include consideration of this case we would need to devise an arbitrary function relating low calf sur- vival to short calving intervals. The net result would again be low rates of increase. To avoid such com- plications we have simply used 2 yr as the minimum average calving interval. Ranges in Age at First Birth The available data suggest a range in age at at- tainment of sexual maturity of 6 to 12 yr for dolphins (Perrin and Reilly 1984). Early reports of Black Sea common dolphins, Delphinus delphis, attaining sex- ual maturity at an average of 3 yr (Kleinenberg 1956) are almost certainly due to faulty age determina- tion7. Because of the recent findings for S. attenuata from the ETP (Myrick et al. 1986), we considered the ages at first birth up to 15 yr. In our formula- tion of the Leslie model, if females mature and first conceive at an average age of 10 yr, the first nonzero fecundity would be in age class 11 (Table 1). 6Hohn, A. A., and M. D. Scott. 1983. Segregation by age in schools of spotted dolphins in the eastern tropical Pacific. Fifth Biennial Conf. Biol. Mar. Mammals, Abstr., p. 47. 6Henderson, J. R., W. F. Perrin, and R. B. Miller. 1980. Rate of gross annual reproduction in dolphin populations (Stenella spp. and Delphinus delphis) in the eastern tropical Pacific, 1973-78. Southwest Fisheries Center, La Jolla, California, Admin. Rep. LJ-80-02, 51 p. 7Myrick, A. C. Jr., Southwest Fisheries Center La Jolla Labor- atory, National Marine Fisheries Service, NOAA, 8604 La Jolla Shores Drive, La Jolla, CA 92038, pers. commun. June 1984. Table 1 . — Parameters used and values included in the computa- tion of rates of increase in dolphin population size. Parameter Values Calving interval Age at first birth Calf survival rate Noncalf survival rate (Sa) 2 yr 3 yr 4 yr 7 yr 9 yr 11 yr 13 yr 15 yr 0.50 0.52 0.54 . . . (Sa)2 0.850 0.855 0.860 0.865 . . . 0.970 529 FISHERY BULLETIN: VOL. 84, NO. 3 RESULTS Figures 1 through 5 give finite rates of increase (displayed as (A - 1) • 100) for the above ranges of age at first birth, calving interval, and calf and non- calf survival. The lower left corner of each panel is blank because we did not consider cases where calf survival exceeded the square of noncalf survival, for the reason discussed in Methods. The maximum finite rates of increase which would result from the parameter ranges included here are 1.08 to 1.09. Rates as low as 0.89, i.e., decreaseof 11%/yr. also resulted from the parameter ranges used. Within the ranges of parameters examined here, rate of increase is most sensitive to calving inter- val and noncalf survival rate, followed by age at first birth, and is relatively insensitive to changes in calf survival rate. This is an expected result following the reports by Eberhardt and Siniff (1977) and Good- man (1981). An increase in calving interval of 1 yr results in a decrease in ROI of about 0.02, holding other parameters constant. For example, the max- imum ROI for a 9 yr age at first birth is about 1.07 with a 2 yr calving interval. This ROI drops to 1.05 with a 3 yr calving interval. A decrease of 0.01 in noncalf survival rate results in a 0.01 decrease in ROI, while a 0.10 decrease in calf survival rate decreases ROI by <0.01. Age at first birth appears to be nonlinearly related to ROI over the ranges ex- amined here. An increase in this age from 7 to 9 yr results in a 0.02 decrease in ROI, while an increase c o o Q. O CO DC "cO > > CO CO o Figures 1-5.— Contours of percent rate of increase in dolphin population size ((A - 1) • 100), as a function 0.50 0.60 0.96 0.85 0.90 0.95 0.97 0.95 0.97 0.96 0.85 0.90 0.95 0.97 Noncalf Survival Rate (proportion) c O "+- i_ o a o i_ Q. (D -t-> CO DC "co > "> i_ CO M— CO O 0.95 0.97 0.96 0.85 0.90 0.95 0.97 0.85 0.90 0.95 0.97 Noncalf Survival Rate (proportion) Figure 1.— First reproduction of dolphin age class 7 yr: a) 2-yr calving interval (upper panel); b) 3-yr calving interval (middle panel); c) 4-yr calving interval (lower panel). Figure 2.— First reproduction of dolphin age class 9 yr: a) 2-yr calving interval (upper panel); b) 3-yr calving interval (middle panel); c) 4-yr calving interval (lower panel). 530 REILLY and BARLOW: INCREASE IN DOLPHIN POPULATION from 11 to 13 yr causes only a 0.01 decrease in ROI. DISCUSSION The ranges of rate of increase estimated here are potentially useful in bracketing possible ROIs for delphinids in general. For any particular population it should be possible to further narrow the range of likely values of ROI, given available estimates for vital rates. For example, Tursiops truncatus from the northeast coast of Florida reportedly attain sex- ual maturity at 12 yr on the average (Sergeant et al. 1973) and have a 12-mo gestation period (Essa- pian 1963), giving an estimated age at first birth of 13 yr. Knowledge of this single parameter can nar- row consideration to Figure 4. Here the estimated range in ROI is up to a maximum of 1.05, for the extreme case of an average calving interval of 2 yr, and noncalf survival >0.96. Additional knowledge of, say, minimal calving interval for Tursiops could further narrow consideration to one of the three panels of Figure 4, and establish minimal survival rates for positive growth rates, or the maximum rate of increase possible, given the above constraints on age at first birth and calving interval. We assume that the ranges defined here also en- compass the limits within which vital rates for any one dolphin species might change in response to changes in population density. This obviously entails making simplistic assumptions about density depen- dence in vital rates, and therefore in rate of increase. of calf and noncalf survival rates, for the following combinations of calving interval and age at first reproduction: c O O Q. O CD ■♦-• CO DC "cO > > CO "co O 0.95 0.97 •■£ 0.95 0.97 0.85 0.90 0.95 0.97 Noncalf Survival Rate (proportion) Figure 3.— First reproduction of dolphin age class 11 yr: a) 2-yr calving interval (upper panel); b) 3-yr calving interval (middle panel); c) 4-yr calving interval (lower panel). C o o a o CD CO DC "c0 > ■> Z5 CO "cO O 0.95 0.97 0.96 0.85 0.90 0.95 0.97 0.85 0.90 0.95 0.97 Noncalf Survival Rate (proportion) Figure 4.— First reproduction of dolphin age class 13 yr: a) 2-yr calving interval (upper panel); b) 3-yr calving interval (middle panel); c) 4-yr calving interval (lower panel). 531 FISHERY BULLETIN: VOL. 84, NO. 3 These assumptions are implicit in the concept of r-max. There is no evidence that the highest rates of in- crease calculated here can be achieved by any real dolphin population. Trade offs may exist between survival and reproduction. Because of this, some of the parameter combinations examined here are probably unlikely, especially combinations of the ex- treme values, i.e., those producing the highest rates of increase. Although our figures also present minimum values based on parameter combinations we used, we do not believe that these will be useful in setting lower bounds on finite rates of increase. Catastrophic events can always lead to rapid extirpation of a population. In fact, it is clear that dolphins (and other animals with similar life histories) can c o '■*-• o Q. O 0 CO DC "cO > > CO CO O 0.95 0.97 0.95 0.97 0.85 0.90 0.95 0.97 Noncalf Survival Rate (proportion) Figure 5.— First reproduction of dolphin age class 15 yr: a) 2-yr calving interval (upper panel); b) 3-yr calving interval (middle panel); c) 4-yr calving interval (lower panel). decrease in number much faster than they can increase. ACKNOWLEDGMENTS This study benefited greatly from reviews by J. Breiwick, D. Chapman, D. DeMaster, D. Goodman, J. Hedgepeth, F. Hester, G. Sakagawa, D. Siniff, T. Smith, and an anonymous reviewer. We sincere- ly thank these people for their contributions. LITERATURE CITED Allen, K. R. 1976. A more flexible model for baleen whale populations. Rep. Int. Whaling Comm. 26: App. Report and Papers of the Scientific Committee, p. 247-263. Barlow, J. 1982. Methods and applications in estimating mortality and other vital rates. Ph.D. Thesis, Univ. California, San Diego, 177 p. Birch, L. C. 1948. The intrinsic rate of natural increase of an insect population. J. Anim. Ecol. 17:15-26. Caughley, G. 1977. Analysis of vertebrate populations. Wiley-Inter- science, N.Y., 234 p. Eberhardt, L. L., and D. B. Siniff. 1977. Population dynamics and marine mammal management policies. J. Fish. Res. Board Can. 34:183-190. Essapian, F. S. 1963. Observations on abnormalities of parturition in captive bottlenosed dolphins, Tursiops truncatus, and concurrent behavior of other porpoises. J. Mammal. 44:405-414. Fruehling, J. A. (editor). 1982. Sourcebook on death and dying. Marquis Prof. Publ., Chicago, 788 p. Goodman, D. 1981. Life history analysis of large mammals. In C. W. Fowler and T. D. Smith (editors), Dynamics of large mam- mal populations, p. 415-436. Wiley, N.Y. JONSGARD, A., AND P. B. LYSHOEL. 1970. A contribution to the knowledge of the biology of the killer whale Orcinus orca (L.). Nytt. Mag. Zool. (Oslo) 18:41-48. Kasuya, T. 1976. Reconsideration of life history parameters of the spotted and striped dolphins based on cemental layers. Sci. Rep. Whales Res. Inst. Tokyo 28:73-106. Kleinenberg, S. E. 1956. Miekopitauishchenie Chernogo i Azovskogo Morei (Mammals of the Black Sea and Sea of Azov). Akad. Nauk., Moscow, 288 p. (Fish. Mar. Serv., Quebec, 1978, Transl. ser. 4319, 428 p.) Leslie, P. H. 1945. On the use of matrices in certain population mathe- matics. Biometrika 33:183-212. MlYAZAKI, N., AND M. NlSHIWAKI. 1978. School structure of the striped dolphin off the Pacific coast of Japan. Sci. Rep. Whales Res. Inst., Tokyo 30: 65-115. 532 REILLY and BARLOW: INCREASE IN DOLPHIN POPULATION Myrick, A. C. Jr., A. A. Hohn, J. Barlow, and P. A. Sloan. 1986. Reproductive biology of female spotted dolphins, Stenella attenuata, from the eastern tropical Pacific. Fish. Bull., U.S. 84:247-259. Perrin, W. F., and S. B. Reilly. 1984. Reproductive parameters of dolphins and small whales of the family delphinidae. In W. F. Perrin, D. P. DeMaster, and R. L. Brownell, Jr. (editors), Cetacean reproduction, p. 181-185. Rep. Int. Whaling Comm. Spec. Issue 6. Sacher, G. A. 1980. The constitutional basis for longevity in the Cetacea: Do the whales and the terrestrial mammals obey the same laws? In W. F. Perrin and A. C. Myrick, Jr. (editors), Age determination of toothed whales and sirenians, 229 p. Rep. Int. Whaling Comm., Spec. Issue 3. Schaefer, M. B. 1957. Some aspects of the dynamics of populations important to the management of commercial marine fisheries. Inter- Am. Trop. Tuna Comm. Bull. 1:27-56. Sergeant, D. E., D. K. Caldwell, and M. C. Caldwell. 1973. Age, growth and maturity of bottlenosed dolphins (Tur- siops truncatus) from northeast Florida. J. Fish. Res. Board Can. 30:1009-1011. Siler, W. 1979. A competing-risk model for animal mortality. Ecology 60:750-757. Smith, T. D. 1983. Changes in size of three dolphin (Stenella spp.) popula- tions in the eastern tropical Pacific. Fish Bull., U.S. 81: 1-13. Smith, T., and T. Polacheck. 1981. Reexamination of the life table for northern fur seals with implications about population regulatory mechanisms. In C. W. Fowler and T. D. Smith (editors), Dynamics of large mammal populations. Wiley, N.Y. Spinage, C. A. 1972. African ungulate life tables. Ecology 53:645-652. 533 DISCRETE-TIME DIFFERENCE MODEL FOR SIMULATING INTERACTING FISH POPULATION DYNAMICS C. Allen Atkinson1 ABSTRACT The dynamics of interacting fish populations are modeled using a coupled set of discrete-time difference equations. The basic equations describe predator-prey and competitive relationships analagous to the first-order expressions used in standard differential equation models. Population births and aging are represented using a modified Leslie matrix. A spatial representation is also incorporated and consists of a number of separate compartments, each containing interacting population groups which can be inter- changed between compartments during a given time period. The potential applicability of the discrete- time formulation is demonstrated via a simulation of the multispecies fish populations within the Califor- nia Current during the sardine population collapse of 1930-60. Numerous mathematical models of interacting multi- species fish populations are found in the literature (Riffenburgh 1969; Saila and Parrish 1972; May et al. 1979; Steele 1979). Depending on the nature of a particular ecosystem and the desired resolution level for its components and processes, these models can become extremely complex (Parrish 1975; Anderson and Ursin 1977; Laevastu and Favorite 1978). The major limitation in practical fisheries ap- plications is the lack of sufficient field data to ade- quately estimate many of the model parameters, particularly the population interaction terms in com- plex multispecies models (Goodall 1972). The two objectives in the present multispecies model development are 1) to establish a general mathematical form applicable to a variety of prac- tical fisheries problems and 2) to provide an efficient computational tool for simulating complex multi- species systems. The latter feature has implications for dealing with the problem of model parameter uncertainty via specialized Monte Carlo and non- linear programming procedures as discussed by Atkinson (1985). The proposed formulation consists of a unique set of discrete-time difference equations that describe first-order dynamic processes affecting some ar- bitrary number of interacting fish populations at one or more trophic levels. The discrete equations are particularly well suited for computer implementa- tion. There are no requirements for sophisticated integration routines (e.g., Runge-Kutta, Adams- Moulton), and the equations have inherent numerical 'System Science Applications, Inc., 121 Via Pasqual, Redondo Beach, CA 90277. Manuscript accepted January 1986. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. stability. Difference equations are also compatible with fisheries data sets (e.g., eggs and larvae sur- veys) which are usually sampled seasonally. The essential biological processes represented in the model are spawning, growth, mortalities, age class structure, nonuniform spatial distributions, and migrations. Certain of these features, such as spawning, sexual maturation, and migrations, are often most conveniently described in a discrete form as assumed in the model. Seasonal time steps are natural increments for consideration as the values of appropriate model parameters can then be easily changed to relate seasonal fish behavior. The mathematical details of the discrete-time difference model are developed below. The special problem of estimating model parameters in practical applications is also briefly discussed. The dynamics of the California Current fish populations are then modeled and simulation runs performed correspond- ing to the period of the sardine collapse in 1930-60. Comparisons are made between the simulation results and the actual (estimated) population responses. DEVELOPMENT OF THE DISCRETE-TIME DIFFERENCE EQUATIONS The dominant first-order ecological processes af- fecting fish populations are modeled by discrete-time difference equations. For convenience in the mathe- matical development, these processes are assumed to occur in the following sequence during a given time period: 1) individual growth and mortalities; 2) spatial redistributions of the surviving members; and 3) births and age class changes of the surviving, 535 FISHERY BULLETIN: VOL. 84, NO. 3 redistributed populations. Consistent with the first- order nature of the formulas, certain simplifications are expected to be incorporated in the ecological representation including implicit modeling of lower trophic levels (e.g., phytoplankton and zooplankton) and functional groupings of less important species as competitors, predators, and prey. Growth and Mortalities First-order differential equations of the following general form are typically used to describe the growth and mortalities of a population P, under competitive and predator-prey influences with itself and other populations: d t (rt - u%P - v,P + WjP)Pi (1) where r, = survival/growth parameter P = population vector = \P\ > " 2 t • • • t "it • ■ • ) Pn) u{ = competition coefficient vector — {tin , u,{ 2) ■ U; uin) Vi = W: = predation coefficient vector prey coefficient vector. The coefficient vectors ut , v{ , and wx contain ap- propriate zeros such that only the active interactions between populations are defined. (Note that vector multiplication is implied by the forms such as u^P.) The competition terms correspond to the standard Gause model, while the predator-prey terms corres- pond to the simple Lotka-Volterra model (Pielou 1977). The population variables P{ can be expressed in units of either numbers of individuals or total biomass, with the coefficients defined accordingly. Assuming a small time step (At) relative to the characteristic time of the system (1/r), a discrete- time approximation is found directly by integrating Equation (1) to give Pt(M) = er>M -u PM -v P&t 0w, PM ■Pr(0) (2) These exponential terms form the basis of the dif- ference model. However, some modification and in- terpretation of terms is required in order to describe a general form appropriate over a range of popula- tion levels. The most obvious inadequacy of Equation (2) is the positive exponential prey term, ew2 bm(t);j ^ l,n = j ain{t);j ^ 2,n = j -1 0 ; otherwise. n„ (o = (10) The model parameters in Equations (9) and (10) consist of maximum survival/growth rates (S), star- vation mortality rates (R), transport terms (g), fecun- dity factors (/), age class changes (a and b), and population interaction coefficients (a, ft, and y). Time dependency is indicated for all parameters except the interaction terms. Space dependency is assumed to apply to all but age class changes and interaction terms. If the parameters are described by probabi- listic functions, the model becomes a stochastic representation. The above difference model represents a com- prehensive description of coupled fish population dynamics and is proposed for general application. The form of Equation (9) is particularly well suited for computer implementation; it provides an effi- cient time-step simulation capability without requir- ing a numerical integration scheme. The model can be conveniently programmed on a mini-computer system and used to simulate complex multispecies population dynamics. MODEL PARAMETER ESTIMATION IN PRACTICAL APPLICATIONS The predictive power of the difference model in practical applications is obviously dependent on the knowledge of the ecosystem processes and the abil- ity to estimate the associated parameters used in the modeling. This situation is true for any eco- system model whether it consists of difference equations, differential equations, or any other for- mulation. In fact, I (1980) showed that difference equations representing multispecies populations can be used to approximate the complex response modes of differential equations by relating parameters and choosing suitably small differencing time steps. I also showed that the difference model suffers from a similar sensitivity to the parameter estimates; the problem becomes more severe with increasing eco- system complexity. Certain parameters in either difference or dif- ferential equation models can be roughly estimated from field and/or laboratory studies. Examples in- clude fecundity and growth rates of individual fish which can be observed directly. Population-level parameters, such as interaction and transport terms, are more difficult to estimate given the dynamic, wide-ranging nature of fish behavior. Even with extensive field sampling and the use of multi- variate statistical techniques to sort out stochastic environmental features (Reid and Mackay 1968; Mobley 1973; Poole 1976), these parameter esti- mates will typically have a large degree of uncertainty. The potential advantage of difference models in dealing with parameter uncertainty is related to their computational efficiency. When parameter uncertainty is represented in a probabilistic frame- work, Monte Carlo procedures can be applied to statistically describe population response character- istics based on large numbers of simulation runs. Probabilistic descriptions of parameter uncertain- ty can express both the inherent stochastic nature of the ecosystem and the parameter estimation er- ror. One problem is that the stochastic ecosystem features, which are of primary interest, will typically be masked in the statistics by the large parameter estimation errors if realistic values for the latter are included. I (1980, in press) used nonlinear programming (NLP) techniques to treat parameter uncertainty in dynamics models for a general class of ecosystem problem. My approach is summarized below; it has been used for resolving parameter estimates in the difference model application discussed in the section that follows. An NLP problem can be stated in the following general form: minimize M subject to g(x) = 0 X0 < X < xm where x is the variable vector with upper and lower bounds of x0 and xm, respectively; f(x) is the so- called objective function; and g{x) is a vector func- tion of implicit constraints. 539 FISHERY BULLETIN: VOL. 84, NO. 3 The problem scenario for my NLP formulation is that of predicting the dynamic response of eco- system populations to a given perturbation. The response is characterized over some period of in- terest by the objective function which, depending on the particular problem, can be equated to average population numbers, final population levels, worst- year fishery catch, or some other dynamic feature. The ecological parameters in the dynamics model become the variables with bounds corresponding to the estimated parameter uncertainty range. Implicit parameter constraints are added to the formulation based on available population history data, ecosystem stability observations, or any known or postulated relationships between parameters. The historical population data are substituted directly into the difference equations, or other assumed dynamics equations. In effect, such constraints force the response modes of the dynamics model to include past population observations, albeit ones that oc- curred under different (known) conditions than those of interest in the future. Stability observations also infer conditions on the dynamics equations and, hence, model parameters. However, there are prac- tical issues in formulating such conditions. Lyapunov stability analysis techniques (Brogan 1974), while applicable to nonlinear system analysis, are not readily defined for the complex difference equations. Efficient NLP computational procedures have been applied by me (1980) to solve the special eco- system formulation described above. A search takes place through bounded parameter space for extreme (minimum and maximum) objective function values while maintaining the equality of the implicit con- straints, i.e., the search proceeds on the "constraint surface" in parameter space. The key to an effec- tive problem solution is the computational require- ments of the dynamics model which is used in both constraint formulation and for evaluating the objec- tive function at each search step. While the NLP approach does not give definitive estimates of in- dividual model parameters, it strongly delimits their range of values via the interrelationships established by the implicit constraints (Atkinson 1980). ECOSYSTEM SIMULATIONS USING THE DIFFERENTIAL EQUATION MODEL The discrete-time multispecies dynamics model given by Equation (9) has been implemented as a FORTRAN computer program and used to perform a variety of simulations of theoretical and applied fisheries scenarios (Atkinson 1980). A case of some practical interest, the collapse of the sardine popula- tion within the California Current region, will be described and used to demonstrate the potential model utility. General Description of the Sardine Population Collapse off California The waters of the California Current flow south- ward along the west coast of North America cover- ing the general region are illustrated in Figure 2. While the California Current supports a diverse group of fish, the sardine fishery was by far the most important in the early years of this century until the dramatic collapse of the sardine population in 1930-60. A large increase in fishing effort took place during this time and apparently caused, or at least was associated with the sardine population collapse. The estimated history of the sardine population from 1930 to 1960 as derived by Murphy (1966) is shown in Figure 3. Two sets of anchovy population estimates for the 1930-60 time frame are also presented in Figure 3. Although these data are confused by significant gaps and strong fluctuations from year to year, there does appear to be a significant population increase from levels in the 1940's and early 1950's to that near the end of the 1950's. Since the anchovy is the chief competitor of the sardine with similar food require- ments and overlapping habitat boundaries, the general indication is that the anchovy replaced the sardine within the trophic structure (Murphy 1966; Gulland 1971). Murphy's (1966) 3-yr averaged data provides the clearest evidence of this increasing trend. Smith's (1972) yearly estimates show that the anchovy population actually declined from 1940-41 to 1950 (the next year in which data was available), before a sharp rise occurred. The significant varia- tions evident in both anchovy and sardine data are probably caused by random environmental in- fluences on recruitment success (Lasker 1978; Par- rish et al. 1981; Methot 1983). Soutar and Isaacs (1974) presented some interest- ing longer term data on the sardine and anchovy (plus other pelagic fish) as derived from sedimen- tary scale depositions in anaerobic basins off South- ern California and Baja California. The deposition rate, which is averaged by 5-yr periods, provides a relative picture of the population variations over the last 150 yr (up to 1970). The data for the 1930-60 time frame indicate similar trends to that above, i.e., decreasing sardine levels and increasing anchovy levels. However, significant sardine and anchovy 540 ATKINSON: FISH POPULATION DYNAMICS r^T" % '# *o^ BRITISH COLUMBIA 50° « WL i§N WASH.V^^ 45° — / OREGON V >» AVERAGE YEARLY SARDINE CATCH (1920- 1950)* PACIFIC NORTHWEST 44K TONS \ 40° — \ N.CALIFORNIA S. CALIFORNIA = 208KTONS = 115KTONS fs ^Y BAJA CALIFORNIA / = 0 TONS** 35° 4 / \ CALIF. i "'•v'\ U* 1 1 \7 \ V 30° mm SARDINE ^ 1 1 DISTRIBUTION F^Tl FISHING Hii LOCALITIES mm 25° 'FROM DATA IN MURPHY (1966) **BAJA FISHERY NOT SIGNIFICANT UNTIL AFTER 1950 I I i I 125° 120c 115< 110° Figure 2.— Map of the California Current region showing sardine distribution and major fishing localities in the period before 1950 (from Murphy 1966). 541 FISHERY BULLETIN: VOL. 84, NO. 3 4000 z 2 3000 z o _j 2000 h a. O a. 1000 - r i ANCHOVY ' ™— A V / « ▲ / / f l # ft / SARDINES' 1 4\ ^ — I / \ / *• \ / V v A • \ A* 1 1 1 1920 30 40 YEAR 50 1960 Figure 3.— Estimated adult populations of sardine and anchovy during the 1930-60 sardine collapse period. The solid line cor- responds to yearly sardine estimates by Murphy (1966). The dashed line with triangles corresponds to 3-yr average anchovy esti- mates also by Murphy; the initial point is a 2-yr estimate with a data gap until 1951. The circles correspond to yearly anchovy estimates by Smith (1972); a data gap exists between 1941 and 1950. variations are also evident in earlier times before fishing pressure became a significant factor in the ecosystem. For example, the sardine history showed extremely low levels in 1865-80 comparable to the levels after 1940. The earlier anchovy record, while also having periods of relatively high and low sedimentation rate, appears to have been at con- sistently higher levels before 1930-60, even higher than the recent increase of the late 1950's. Soutar and Isaacs (1974) stated that relatively unproduc- tive conditions have apparently existed for the past 30 yr or so and have generally affected fish popula- tions of the California Current. Model Formulation subsystem defined by Riffenburgh (1969) and shown in Figure 4. While not a comprehensive description of this ecosystem, I use this representation to demonstrate the application of the difference model in a reasonably complex fishery situation. The sar- dine ecosystem will be simulated during the period from 1932 to 1952 spanning the years of the major sardine collapse. The sardine population is divided into three age groups: larval-year stages, yearlings, and adults. The larval year is the most vulnerable period of the sardines' development during which it goes through many fundamental changes. The yearlings are the in-between stage to the sexually mature adult members of the population, which are defined to be 2-yr-olds and above. Early stages of the sardine feed on phytoplankton while the adults feed primarily on zooplankton (Huppert et al. 1980). The adults are also predators of their own larval stages and those of the anchovy as indicated in Figure 4. The anchovy population is divided into two groups, larvae and adult, which have similar intergroup rela- tionships and feeding habits to the corresponding sardine groups. Competitor and predator groups to the sardine and anchovy are defined as lumped assemblages, both encompassing a broad range of diverse fish species; the competitor group also con- tains many invertebrates. The pelagic fish com- petitors (e.g., jack mackerel) are assumed to behave similarly to the sardine and anchovy except that some of the larger members feed on the sardine yearling stage (Riffenburgh 1969). The predators (e.g., hake and baracuda) feed on the adults of the sardine-anchovy-competitor trophic level and also have other prey that have been decoupled from the modeled subsystem. Phytoplankton and zooplankton groups are modeled implicitly as carrying capacity terms. Additional model assumptions are that 1) spatial features are not critical (i.e., one spatial compart- ment is used), and 2) seasonal effects can be ignored (i.e., a yearly time step is defined). These two assumptions are probably not justifiable in the time period after 1950 or so, because of the shift of dominance from the northern sardine subpopulation to the southern one. Important differences in such factors as natural survival rates, maturation charac- teristics, and fishing effort exist for these subpopula- tions (Murphy 1966). The waters of the California Current region, with their chemical and biological constituents, can be viewed as an ecological system (Sette 1969). The present model focuses on the sardine and anchovy Discrete-Time Difference Equations The difference model representing the seven inter- acting populations of the sardine ecosystem is pre- 542 ATKINSON: FISH POPULATION DYNAMICS c 0 SARDINE LARVAE (© ANCHOVY LARVAE (3) SARDINE ADULT (D ANCHOVY ADULT (6) COMPETITOR GROUP © PREDATOR GROUP (ADDITIONAL FOOD SUPPLY) ) Figure 4.— Schematic showing interactions between sardine ecosystem groups as modeled by Riffenburg (1969). Competitive relationships are indicated by the connecting lines with dual arrowheads, while predator-prey relationships are defined by arrows pointing to the predator. sented in Table 1. These equations reflect the general form of Equation (9) for a single spatial com- partment. Parameters are defined for all processes other than transport, including competition, pred- ator-prey, survival/growth, births, and fecundity. These parameters are assumed to be independent of the year during the 1932-52 simulation period, except for 1) the sardine fishing rate, 63(t), and 2) a sardine larvae survival factor, E^t). The latter are related to the parameters presented earlier by 63(t) = 1 - Sf3(t) E,(t) = S.iD/S, where Sf3 is defined in Equation (5) and S: is the average (reference) sardine larvae survival rate dur- ing 1932-52. The time-varying fishing rate and lar- vae survival factor represent the "drivers" perturb- ing the ecosystem during the sardine collapse period. Time-varying representations may also be ap- Table 1— Difference equations describing biomass dynamics of the sardine ecosystem populations. Note that age sub-groups are indexed as separate populations to simplify the nomenclature. Also, all populations in exponentials are assumed to be at time f. Population 1 - P,(t + 1) Population 2 - P2(t + 1) Population 3 ■ P3(t +1) Population 4 P*(t + 1] Population 5 P5(t +1) Population 6 P6{t + 1) Population 7 - P7(t + 1) sardine larvae = f3S3.o[1 - 0.4 in 1936. The fishing rate is assumed to remain constant for the remainder of the simulation period. The assumed model for the sardine larvae survival term, E^t), is presented in Figure 6 along with Sette's (1969) data from which it was derived. These data represent numbers of fish at age class two ver- sus the year spawned. The survival rate model assumes that these observed fluctuations in the data primarily reflect random survival effects during the first year of life. Ex{t) was obtained by normaliz- ing Sette's data with respect to the spawning population biomass and defining a relative scale such that the integrated value over the 20-yr period from 1932 to 1952 was equal to one. The remaining model parameters, which repre- sent the great majority of those in the equations of Table 1, could not be directly estimated to any degree of accuracy from available literature data. Instead, these estimates were derived from the special nonlinear programming analysis of mine (1980, in press) mentioned previously. I treated these ecosystem model parameters as variables with upper and lower bounds reflecting their uncertain- ty ranges. The bounds established by me for the sar- dine ecosystem parameters were typically an order of magnitude. Implicit parameter constraints were 544 ATK1NSUN: 1-1SH i'Ui'ULAilUIN DflNAMlOS 1.0 ESTIMATED FISHING RATE DERIVED FROM MURPHY'S (1966) DATA FISHING RATE MODEL 1930 34 38 42 46 50 YEAR Figure 5.— Model of sardine fishing rate, 63(t), used in the sardine ecosystem simulations. SURVIVAL FACTOR MODEL YEAR CLASS SURVIVAL ESTIMATE BY SETTE (1969) 24 « (A Z o 20 16 1930 34 38 42 YEAR 46 50 S o t 8) < < W O a. >■ at z i < (0 Figure 6.— Model of sardine larvae survival, E^t), used in the sardine ecosystem simulations. defined by the assumed equilibrium condition prior to 1932-52. Setting the time-varying fishing rate at its pre-1932 value (d3 = 0.10) and fixing the time- varying larval survival factor at its reference value (E1 = 1.0), a set of seven equality constraints were specified corresponding to the seven population equations in Table 1 with P(t + 1) = P(t) = P. While there is still significant degrees-of -freedom in the model (i.e., more parameters than equality constraints), I was able to greatly resolve their values based on my nonlinear programming pro- cedures. The parameters in Table 2 represent the "nom- inal" estimates presented by me (1980) based on my NLP analyses. In searching for minimum and max- imum population response levels throughout bounded parameter space, a series of intermediate search steps were taken that produced suites of interdependent parameter values satisfying the pre-1932 equilibrium condition. Population response levels were equated to the average sardine popula- tion during the 1932-52 simulation period in this analysis. The selected nominal parameter suite in Table 2 gives response levels approximately midway between the determination of minimum and max- imum levels. Note that the parameter values in Table 2 were not derived from statistical procedures using the 545 FISHERY BULLETIN: VOL. 84, NO. 3 Table 2.— Estimated values of the sardine ecosystem model parameters (from Atkinson 1980). Parameter Nominal Parameter Nominal Population type Symbol value Population type Symbol value 1 Sardine larvae Survival/growth s, 7.26 5 Anchovy adult Survival/growth s5 1.30 Competition aii 5 x 10~6 Competition °53 2.0 x 10"5 Competition au 2.5 x 10-6 Competition °55 3.0 x 10-5 Predation 013 7.6 x 10-4 Competition a56 1.0 x 10'5 Predation 015 3.8 x 10-4 Predation 057 1.0 x 10"4 Predation 016 7.6 x 10~5 Fecundity ^5 0.432 2 Sardine yearling Survival/growth s2 2.10 6 Competitor group Survival/growth s6 1.65 Competition o22 3.7 x 10"5 Competition a63 5.0 x 10~5 Competition a25 1.8 x 10-5 Competition a65 5.0 x 10-5 Predation 026 1.8 x 10"5 Competition a66 5.0 x 10"5 3 Sardine adult Survival/growth $3.0 1.40 Predation 067 5.0 x 10~5 Competition a33 1.5 x 10~5 7 Predator group Survival/growth s7 1.23 Competition °35 1.0 x 10-5 Mortality *7 0.5 Competition a36 5.0 x 10~6 Competition <*77 5.2 x 10~5 Predation 037 1.0 x 10-4 Prey Y73 2.5 x 10-4 Fecundity '3 0.468 Prey Y75 2.5 x 10"4 4 Anchovy larvae Survival/growth Competition Competition Predation Predation Predation S4 "41 Q'44 043 045 046 0.50 2.5 x 5.0 x 1.5 x 3.0 x 3.0 x 10-6 10-6 10-4 10-4 10~5 Prey Y76 1.25 x 10"4 population data during the simulation period (Fig. 3). The estimates are uncoupled from these data and, hence, reflect strictly a priori knowledge as would exist in applications where predictions are required. Furthermore, the parameter values are not pro- posed as best estimates of these parameters, but simply provide a consistent set of values for use in the simulation demonstration. The nonlinear pro- gramming approach of mine is structured in general to bound future ecosystem response characteristics given only a priori population data. Ecosystem Simulations The simulated sardine ecosystem histories are presented and compared with estimated sardine and anchovy population data in Figure 7. The adult sar- dine population simulation is in reasonably good agreement with the data of Murphy (1966) giving the many approximations and simplifying assump- tions used in the modeling. The major dynamic features of the adult sardines decline are consistent, including the sharp rebounds associated with the favorable conditions for sardine larvae survival in 1938 and 1939 and again in 1947 (Fig. 6). The simulated anchovy response in Figure 7, which ignores any fluctuating larvae survival com- ponent, appears to track the 3-yr averaged estimates of Murphy (1966). The anchovy population increases along with the competitor group to fill the ecological void in this trophic level. The predator biomass decreased slightly because the decline of the sardine results in a less desirable food supply, at least ac- cording to estimated input parameters. Unfortun- ately, there are no available data for comparing with the predicted competitor and predator group responses. Another simulation run was made to investigate the speculation that fluctuating larval survival rates, by themselves, might have caused the sardine col- lapse. The sardine fishing rate was held at the relatively low levels that existed before 1932 (d3 = 0.10), and the fluctuating larvae survival model in Figure 6 was applied. The resulting simulation run is presented in Figure 8 and shows the predicted history of the adult sardine population, along with that of the anchovy, competitor, and predator groups. The adult sardine population again fluc- tuates markedly but now remains at relatively high levels, in no apparent danger of collapsing. It would appear from these runs that the added fishing pressure is necessary to explain the actual event dur- ing this period. CONCLUSIONS A general set of discrete-time difference equations have been developed for use in simulating the im- portant dynamic processes effecting fish popula- tions, including • interactions between competors, predators, and prey • birth, growth, and aging processes within a 546 ATKINSON: FISH POPULATION DYNAMICS 4000 ESTIMATED FROM SARDINE DATA BY MURPHY (1966) ESTIMATED FROM ANCHOVY DATA BY MURPHY (19661 1932 36 40 44 48 52 YEAR (a) SARDINES AND ANCHOVY 8000 — _ 6000 V) z o I- 8 < g CQ 4000 — 2000 — — COMPETITOR GROUP I PREDATOR GROUP ^^^^^^^^^^ I I I 1932 36 48 52 40 44 YEAR (b) COMPETITOR AND PREDATOR GROUPS Figure 7.— Simulation run for assumed models of increased sar- dine fishing rate and fluctuating sardine larvae survival rate. single population group • spatial and temporal variations. The sardine subsystem within the California Cur- rent region was modeled using the multispecies dif- ference model and simulations computed for the sardine's collapse period of 1932-52. Input drivers perturbing the system included representations of the increased sardine fishing pressure and the fluc- tuating sardine larvae survival rates during this period. Simulation results were shown to compare favorably with the available population history data. The increased fishing pressure was indicated to be 1000 ANCHOVY 1932 36 40 44 YEAR 48 52 Figure 8.— Simulation run for assumed constant pre-1932 fishing rate but with fluctuating sardine larvae survival rate. the fundamental cause for the sardine collapse; the estimated yearly fluctuations in sardine larvae sur- vival could not by themselves have caused this sud- den event. These simulation results demonstrate the use of the discrete-time difference model as an efficient simulation tool. There appear to be many applica- tions for the model in theoretical and applied multi- species fisheries studies. ACKNOWLEDGMENTS This work was based on a part of a dissertation submitted in partial satisfaction of the requirements for the Ph.D. degree at the University of Califor- nia, Los Angeles. S. E. Jacobsen, chairman of the dissertation committee, provided guidance and en- couragement throughout these studies. D. A. Kiefer of the Department of Biological Sciences, Univer- sity of Southern California, reviewed early versions of this paper and made helpful comments. LITERATURE CITED Anderson, K. P., and E. Ursin. 1977. A multispecies extension to the Beverton and Holt theory of fishing, with accounts of phosphorus circulation and primary production. Medd. Dan. Fisk. Havunders. 7:319-435. Atkinson, C. A. 1980. Analysis of perturbed dynamic systems under param- 547 FISHERY BULLETIN: VOL. 84, NO. 3 eter uncertainty - a nonlinear programming approach with applications to marine ecosystems. Ph.D. Thesis, Univ. California, Los Angeles, 198 p. In press. A nonlinear programming approach to the analysis of perturbed marine ecosystems under model parameter uncertainty. Ecological Modeling. Beverton, R. J. H., and S. H. Holt. 1957. On the dynamics of exploited fish populations. Fish. Invest. Minist. Agric. Fish. Food (G.B.). Ser. II, 19, 533 p. (See p. 21-135.) Brogan, W. L. 1974. Modern control theory. Quantum Publishers, Inc., N.Y., 393 p. (See p. 262-283.) Clark, F. N., and J. B. Phillips. 1952. The northern anchovy in the California fishery. Calif. Fish and Game 38:189-207. Cushing, D. H. 1975. Marine ecology and fisheries. Cambridge Univ. Press, Cambridge, 278 p. (See p. 85-247.) Goodall, D. W. 1972. Building and testing ecosystem models. In J. N. R. Jeffers (editor), Math models in ecology. The 12th sym- posium of the British Ecological Society, Grange-Over- Sands, Lancashire, 23-26, March 1971. p. 173-183. Black- well Sci. Publ. Lond. Gulland, J. A. 1971. Ecological aspects of fishing research. Adv. Ecol. Res. 7:115-176. Huppert, D. D., A. D. MacCall, G. F. Stauffer, J. A. McMillan, and H. W. Frey. 1980. California's northern anchovy fishery: biological and economic basis for fishery management. U.S. Dep. Com- mer., NOAA Tech. Memo. NMFS-SWFC-1. Ivlev, V. S. 1961. Experimental ecology of the feeding of fishes. Yale Univ. Press, New Haven, Conn., 302 p. Laevastu, T., and F. Favorite. 1978. Fluctuations in Pacific herring stock in the Eastern Bering Sea as revealed by ecosystem model (DYNUMES III). Rapp. P.-v. Reun. Cons. Perm. int. Explor. Mer 177:445-459. Lasker, R. 1978. The relation between oceanographic conditions and lar- val anchovy food in the California Current: identification of factors contributing to recruitment failure. Rapp. P.-v. Reun. Cons. int. Explor. Mer 173:212-230. Lefkovitch, L. P. 1965. The study of population growth in organisms grouped by stages. Biometrics. 12:1-18. Leslie, P. H., and J. C. Grower. 1960. The properties of a stochastic model for the predator- prey type of interactions between two species. Biometrika 476:219-234. Levine, S. H. 1976. Competitive interactions in ecosystems. Am. Nat. 110:903-910. MacCall, A. D. 1979. Population estimates for the waning years of the Pacific sardine fishery. Calif. Coop. Oceanic Fish. Invest. Rep. 20, p. 72-82. May, R. M. 1973. Stability and complexity in model ecosystems. Prince- ton Univ. Press, Princeton, NJ, 265 p. (See p. 13-171.) May, R. M., J. R. Beddington, C. W. Clark, S. J. Holt, and R. M. Laws. 1979. Management of multispecies fisheries. Science 205: 267-277. Methot, R. D., Jr. 1983. Seasonal variation in survival of larval northern an- chovy, Engraulis mordax, estimated from the age distribu- tion of juveniles. Fish. Bull., U.S. 81:741-750. Mobley, C. D. 1973. A systematic approach to ecosystems analysis. J. Theor. Biol. 42:119-136. Murphy, G. I. 1966. Population biology of the Pacific sardine (Sardinops caerulea). Proc. Calif. Acad. Sci. 34:1-84. Parrish, J. D. 1975. Marine trophic interactions by dynamic simulation of fish species. Fish. Bull, U.S. 73:695-716. Parrish, R. H., C. S. Nelson, and A. Bakun. 1981. Transport mechanisms and reproductive success of fishes in the California Current. Biol. Oceano. 1:175- 203. Parsons, T. R., and M. Takahashi. 1973. Biological oceanographic processes. Pergamon Press, Oxford, 186 p. (See p. 149-155.) Pielou, E. C. 1977. Mathematical ecology. John Wiley and Son, N.Y., 385 p. (See p. 8-110.) Poole, R. W. 1976. Empirical multivariate autoregressive equation predic- tors of the fluctuations of interacting species. Math. Biosci. 28:81-97. Reid, M. H., and R. S. Mackay. 1968. A review and discussion of some methods for sys- tem identification and modeling. Med. Biol. Eng. 6:231- 268. RlCKER, W. E. 1958. Handbook of computations for biological statistics of fish populations. Fish. Res. Board Can. Bull. 119, 300 P- Riffenburgh, R. H. 1969. A stochastic model of interpopulation dynamics in marine ecology. J. Fish. Res. Board Can. 26:2843-2880. Saila, S. B., and J. D. Parrish. 1972. Exploitation effects upon enterspecific relationships in marine ecosystems. Fish. Bull., U.S. 70:383-393. Sette, O. E. 1969. A perspective of a multi-species fishery. Calif. Coop. Oceanic Fish. Invest. Rep. 13, p. 81-87. Smith, P. E. 1972. The increase in spawning biomass of northern anchovy, Engraulis mordax. Fish. Bull., U.S. 70:849-874. SOUTAR, A., AND J. D. ISAACS. 1974. Abundance of pelagic fish during the 19th and 20th cen- turies as recorded in anaerobic sediment off the Californias. Fish. Bull., U.S. 72:257-273. Steele, J. H. 1979. Some problems in the management of marine re- sources. Appl. Biol. 4:103-140. Vance, R. R. 1978. Predation and resource partitioning in one predator- two prey model communities. Am. Nat. 112:797-813. 548 FECUNDITY OF NORTHERN SHRIMP, PANDALUS BOREALIS, (CRUSTACEA, DECAPODA) IN AREAS OF THE NORTHWEST ATLANTIC D. G. Parsons and G. E. Tucker1 ABSTRACT Fecundity of the northern shrimp, Pandal-us borealis, and relationships between number of eggs and carapace length were determined from 15 samples taken in 9 areas of the Northwest Atlantic. The sam- pling area extended from Davis Strait to the south coast of Newfoundland. Comparisons of samples suggested that fecundity levels can vary between seasons, years, and areas. A relationship between egg production and environmental temperature was not evident from available samples. The northern or pink shrimp, Pandalus borealis, is a protandric hermaphrodite with a circumboreal distribution. In the Northwest Atlantic, it occurs from about lat. 75 °N at West Greenland to about lat. 42°N at Georges Bank (Squires 1970). Fecun- dity of this species in the North Atlantic has been studied in southern Norway (Rasmussen 1953), northern Norway (Thomassen 1977), the North Sea (Allen 1959), Iceland (Skuladottir et al. 1978), West Greenland (Horsted and Smidt 1956), Barents Sea (Teigsmark 1983), and Gulf of Maine (Haynes and Wigley 1969). Bottom water temperatures recorded at depths where shrimp samples were collected dur- ing these studies varied considerably between areas but were within the range of tolerance for survival of adults as reported by Allen in 1959 (-1.68° to 11.13°C). This paper provides information on the fecundity of P. borealis in the Northwest Atlantic. Samples were collected in areas of known shrimp concentra- tion off Baffin Island, in the eastern Hudson Strait and Labrador Sea, and off the south coast of New- foundland. Bottom temperatures at sampling sites also varied between these areas but were confined to the lower half of the tolerance range (<7°C). Com- parisons are made between selected combinations of the data sets presented. The possible effects of ambient temperature on fecundity levels also are considered. MATERIALS AND METHODS Samples of ovigerous female shrimp were col- lected opportunistically during various research cruises conducted by or for the Department of Fisheries and Oceans, St. John's, Newfoundland, Canada, between 1971 and 1982. A total of 15 samples was selected for analysis. These were taken from the Baffin Island area (east of Cumberland Sound); Hudson Strait; North Labrador Sea; Hope- dale, Cartwright, and Hawke Channels (on the Labrador Shelf); St. Mary's Bay; Fortune Bay; and the Southwest Newfoundland coast (Fig. 1). For some areas, only one sample was available while for others, samples were obtained in different months and/or different years (Table 1). Only animals in good condition were selected from the trawl catches for the study (i.e., no noticeable damage and egg mass undisturbed). Individuals were selected over the complete size range of females, preserved in 10% Formalin2 and returned to the laboratory. It was assumed that within any length group the selection (in terms of number of eggs) was random. Oblique carapace lengths were measured to the nearest 0.1 mm using Vernier calipers. This measurement is the distance between the posterior margin of the orbit of the eye and the posterodorsal margin of the carapace (Rasmussen 1953). All eggs were removed from the pleopods, spread in a Petrie dish, and oven dried overnight at 60°C. After drying, eggs were further separated and counted. Accuracy of the counts was determined by re- counting the eggs from 49 animals. Differences from the initial counts in 48 cases varied between - 5.75% 'Fisheries Research Branch, Department of Fisheries and Oceans, P.O. Box 5667, St. John's, Newfoundland A1C 5X1, Canada. 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Manuscript accepted December 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 549 FISHERY BULLETIN: VOL. 84, NO. 3 56° M° 52° 50° 550 PARSONS and TUCKER: FECUNDITY OF NORTHERN SHRIMP Table 1 .—Regression equations for fecundity (F) vs. length (L) for Pandalus borealis in the Northwest Atlantic. Temp. Date N Regression equation r2 °C Sample Baffin Island 14 Aug. 1978 48 iog10 F = 3.0955 log10 L- -1.1417 0.75 0.7-1.8 Hudson Strait 13 Sept. 1982 24 iog10 F = 3.8880 log10 L- -2.3967 0.33 0.6 North Labrador Sea 22 Sept. 1982 43 iog10 F = 3.2715 log10 L- - 1 .4550 0.45 0.5 Hopedale Channel 28 Sept. 1978 46 iog10 F = 2.8045 log10 L- - 0.7202 0.70 3.0 Hopedale Channel 11 , 25 Sept. 1982 96 log10 F = 2.8884 log10 L- -0.8893 0.74 3.2 Cartwright Channel 20 Sept. 1978 45 iog10 F = 3.1824 log10 L- - 1 .3059 0.74 3.0 Cartwright Channel 11 , 26 Sept. 1982 87 login F = 2.5240 log10 L- -0.3750 0.68 2.0-2.4 Hawke Channel 24 Aug. 1974 20 iog10 F = 3.4614 log10 L- - 1 .6670 0.70 — Hawke Channel 30 Nov. 1974 24 iog10 F = 1.4613 log10 L+ 1.1015 0.31 2.9 Hawke Channel 23 Sept. 1975 27 iog,0 F = 3.0106 log10 L- -1.0147 0.68 2.7 St. Mary's Bay 18 Mar. 1971 48 login F = 2.3954 log10 L- -0.3691 0.53 — St. Mary's Bay 28 Feb. 1974 44 iog,0 F = 2.5290 log10 L- -0.5476 0.47 — Fortune Bay 17 Mar. 1978 48 login F = 3.0413 log10 L- -1.1187 0.57 1.0 Fortune Bay 30 Mar. 1979 47 iog,0 F = 2.6870 log10 L- -0.6428 0.70 — SW Newfoundland Coast 27 Feb. 1978 48 iog10 F = 2.8557 log10 L- -0.7396 0.78 6.2 and +5.65%. The difference between the total number of eggs counted and recounted was only - 0.22% of the initial count. A recount of eggs from one female indicated a difference of -9.38%. It is possible that, in this case, some of the eggs were inadvertently lost between counts. Parameters for the relationship between number of eggs and carapace length for each sample were determined by linear regression using log-log (base 10) transformation. Some data sets were compared by analysis of covariance, assuming homoscedas- ticity. All statistical analyses were performed using the REG (regression) and GLM (general linear models) procedures of SAS (Statistical Analysis System). It must be stressed that samples were obtained opportunistically and not according to a predeter- mined sampling design. Consequently, the statistical analyses were performed based on a practical ap- proach rather than attempting methods for which strict sampling procedures are required. It was anti- cipated that differences in fecundity-length relation- ships could be due to seasonal, annual, and areal effects. Our data only permitted simple compari- sons, investigating each factor separately. Bottom temperatures at most sample locations were recorded to the nearest 0.1 °C using either manual or expendable bathythermographs. RESULTS The parameters of the fecundity-carapace length relationships for all 15 samples are given in Table Figure 1.— Positions of stations in the northwest Atlantic where northern shrimp fecundity samples were collected. 1. Data and the fitted line for each sample are displayed in Figure 2. Coefficients of determination ranged from 0.31 to 0.78 and all relationships were significant (differences from zero slope were highly significant). Intercepts for the log transformed data were less than zero in all but one case. Slopes ranged from 2.4 to 3.9 except for the sample with positive intercept (1.5). Only two samples were available (Hawke Chan- nel, August and November 1974) for comparison of fecundity between seasons. Analysis of covariance on the log of both variables indicated a highly sig- nificant difference in slopes (Table 2). The data showed that larger females (>24 mm), on average, carried more eggs in August whereas smaller females showed higher fecundity in November (Fig. 3). Samples from specific areas and seasons were compared to determine similarities or differences between years. Five simple comparisons were possi- ble: St. Mary's Bay - March 1971 vs. February 1974, Hawke Channel - August 1974 vs. September 1975, Fortune Bay - March 1978 vs. March 1979, Cart- wright Channel - September 1978 vs. September 1982, and Hopedale Channel - September 1978 vs. September 1982. No significant differences in either the rate of in- crease in fecundity with increasing size (slope) or mean number of eggs produced (intercept) were detected between years in three of the five areas compared (Table 2). These were St. Mary's Bay, 1971 and 1974; Hawke Channel, 1974 and 1975; and Fortune Bay, 1978 and 1979 (Fig. 4a, b, and c, respectively). Samples from Cartwright Channel from September 1978 and 1982 showed a significant difference in slopes at a = 0.05 (Fig. 4d) whereas samples from the Hopedale Channel for the same 551 FISHERY BULLETIN: VOL. 84, NO. 3 BAFFIN ISLAND AUG. 1978 • • . CARTURISHT SEPT 1878 HUDSON STRAIT SEPT . 1 982 CARTURISHT SEPT. 1982 • N. LABRADOR SEA SEPT 1 982 • • 1 * * ^ • • • •^""'^^ * ■ . *^^"^% . « • — — ~^"^ • • • HAUKE AUG. 1974 SBB-I 28 HOPEDALE SEPT. 1978 HOPEDALE SEPT . 1 982 ^eae- HAUKE • • • • • * isee- NOV . 1 ©74 • • • • • • • • laoe- • see- 22 sae- HAWKE SEPT . 1 975 552 PARSONS and TUCKER: FECUNDITY OF NORTHERN SHRIMP ST , MARYS BAY MARCH I 97 \ ST. MARYS BAY FEB. 1974 see-l 20 2500- FORTUNE BAY MARCH 1078 FORTUNE BAY MARCH 1 07O 2 B 22 ZA 26 28 4000- S.W. COAST 3000 - FEB. 1978 • ■ * • • • 2000- . V^^ « - — ^» a • 1 000- • years were similar in slope but different in eleva- tion (Fig. 4e). Average fecundity at length was higher in 1978 than in 1982 in the latter area. Three comparisons were possible to detect differ- ences between areas. In 1982, four areas were sampled during September: Hudson Strait, North Labrador Sea, Hopedale Channel, and Cartwright Channel. Analysis of the data indicated no differ- ence in the rate of increase in fecundity with increas- ing size but a highly significant difference in the mean number of eggs produced (Table 2, Fig. 5a). T-Tests for sample means showed that the sample from Hudson Strait was different from those taken off the Labrador coast (Table 3). Fecundity in the former was less at comparable sizes. Three areas were sampled in August and Septem- ber 1978: east of Baffin Island (August), Hopedale and Cartwright Channels (September). The data from these samples also were similar in slope but different in elevations (Table 2, Fig. 5b). T-tests showed that the lower fecundity observed in the Cartwright Channel was significantly different (a = 0.05) from that observed in the other two areas (Table 3). Two samples were taken off the south coast of Newfoundland early in 1978: one from the south- west coast in February and the other from Fortune Bay in March. Eggs in both samples were "eyed", indicating late stage development. The data showed that average egg production was higher off the southwest coast over the range of sizes compared (Fig. 5c). The statistical analysis indicated similar- ity in slopes but a highly significant difference in elevations (Table 2). DISCUSSION Loss of eggs over the ovigerous period has been reported in previous studies on fecundity of P. borealis (Elliot 19703; Ito 1976; Skuladottir et al. 1978; Stickney and Perkins 1979; Stickney 1981). This loss could be incidental or due to incomplete fertilization and/or disease. Egg diameter also in- creases between spawning and hatching (Haynes and Wigley 1969; Ito 1976), and some eggs that are 3Elliot, D. L. 1970. Fecundity of the northern shrimp, Pan- dalus borealis. Unpubl. manuscr., 32 p. Bowdoin University, Brunswick, ME 04011. Figure 2.— Number of eggs (vertical axis) vs. carapace length in mm (horizontal axis) for 15 samples of female northern shrimp taken from areas of the northwest Atlantic. 553 FISHERY BULLETIN: VOL. 84, NO. 3 Table 2.— Analyses of covariance for fecundity-length relationships. Slopes Intercepts Effect Comparison F-value Prob. F-value Prob. Season Hawke Channel 08/74 vs. Hawke Channel 11/74 8.19 0.0067 Year St. Mary's Bay St. Mary's Bay Hawke Channel 03/71 vs. 02/74 08/74 vs. 0.06 0.8024 0.18 0.6748 Hawke Channel 09/75 0.42 0.5215 2.30 0.1367 Fortune Bay Fortune Bay 03/78 vs. 03/79 0.52 0.4707 1.23 0.2709 Cartwright Channel Cartwright Channel 09/78 vs. 09/82 4.27 0.0408 0.43 0.5141 Hopedale Channel Hopedale Channel 09/78 vs. 09/82 0.07 0.7861 26.31 0.0001 Area Hudson Strait North Labrador Sea 09/82 vs. 09/82 vs. Hopedale Channel Cartwright Channel 09/82 vs. 09/82 1.78 0.1490 10.69 0.0001 Baffin Island 08/78 vs. Hopedale Channel Cartwright Channel 09/78 vs. 09/78 0.49 0.6140 8.51 0.0003 SW Newfoundland Coast 02/78 vs. Fortune Bay 03/78 0.17 0.6792 61.32 0.0001 2888- AUGUST.1974 NOVEMBER, 1974 22 23 24 25 LENGTH CMM5 28 Figure 3.— Comparisons of northern shrimp fecundity between seasons for the Hawke Channel, based on predicted values from equations in Table 1. close to the periphery and loosely attached may be simply "crowded out". The evidence of egg loss described in previous studies is sufficient to suggest that combining data from different times of year is not appropriate. The two samples compared in this study produced incon- clusive results in that average fecunity was not con- sistently lower over the complete size range in November compared with the August sample. Annual variation in fecundity-length relationships occurred in two of five areas sampled in different years. The rate of increase in number of eggs with Table 3. — Paired comparisons for area differences when k (no. of samples) >2. P values for H0: Mean, = meany Date/sample No. September 1982 Hudson Strait 1 North Labrador Sea 2 Hopedale Channel 3 Cartwright Channel 4 August 1978 Baffin Island 1 Hopedale Channel 2 Cartwright Channel 3 0.0002 0.0001 0.0001 0.5709 0.0121 0.2141 0.0695 0.0002 0.4525 increasing size only differed significantly in one case, however. The reasons why fecundity differs between years are not known but could be related to changes in environmental conditions and/or egg disease (Stickney 1981). In support of the latter, it is noted that the proportion of nonviable eggs in the 1982 Hopedale Channel sample was higher than in the 1978 sample by an order of magnitude (D. G. Par- sons unpubl. data). Fecundity was significantly higher in the 1978 data. Teigsmark (1983) found that variation within a population during successive years is as great as the variation between populations in a single year and was unable to make a conclusive statement about fecundity of different populations of P. borealis in the Barents Sea. He speculated that such differences could be related to availability of food and popula- tion density. 554 PARSONS and TUCKER: FECUNDITY OF NORTHERN SHRIMP •T. MARYS BAY. 1974 ST. MARYS BAY. 1971 CARTVRXSHT CHANNEL. 1979- -CARTWRIBMT CHANNEL. I 902 01 i » 3 CHANNEL. 1992 FORTUNE BAY. 1979- FORTUNE BAY. I 979 Carapace length (mm) Figure 4.— Comparisons of northern shrimp fecundity be- tween years from five different areas, based on predicted values from equations in Table 1. Carapace length (mm) Based on the comparison of samples taken in 1982, it was shown that the fecunity-length relationships in three areas off Labrador were similar. Similar- ity was not apparent in 1978 samples which showed that fecundity in the Cartwright Channel was lower than in the Hopedale Channel. This discrepancy in results from Labrador is due to annual differences demonstrated for both channels in 1978 and 1982 samples. The comparison by area for the 1978 data also implied similarity between the Baffin Island and Hopedale Channel samples. However, the size ranges compared were not the same. Female shrimp ranged in size from 23.7 to 34.5 for the Baffin Island sample in contrast to 21.7 to 29.0 for the Hopedale Channel sample. These differences in size likely reflect separate rates of growth and maturity in the two areas. Therefore, from a biological viewpoint, all three areas sampled in 1978 exhibited different fecundity-length relationships. The differences between areas, described above, can be considered in relation to the temperatures present in these areas. The bottom temperature at the sampling station off the southwest coast of New- foundland was 6.2°C, the warmest of all areas sampled (Table 1). The temperature recorded in 555 FISHERY BULLETIN: VOL. 84, NO. 3 INZ NORTH LABRADOR SEA- HUDSON STRAIT CARTWBIBMT " HOPEDALE CHANNEL .Q 4sae- b I97« yf 3sae- BAFfTN ISLANO-^^ zeae- ^^•"^-"^CARTWRIOHT CHANNEL isee- •*}PEDALE CHANNEL— ^55^ eea- NCWTOUNDLAND 33 35 Carapace length (mm) Figure 5.— Comparisons of northern shrimp fecundity between areas in 1982 and 1978, based on predicted values from equations in Table 1. March 1978 in Fortune Bay was 1.0°C, one of the coldest areas. According to Squires (1968), the penetration of Atlantic water into the former area accounts for these warmer temperatures which per- sist throughout the year. In Fortune Bay, however, the deep bottom water is of mixed Atlantic and Arc- tic origin resulting in much colder temperatures. Thus, the lower fecundity in the Fortune Bay sam- ple is likely linked with an overall reduction in pro- ductivity in a cold water environment. Reduced productivity has been observed previously in the cold water habitats of le Fjord du Saguenay, Quebec (Couture 1971) and the Barents Sea (Berenboim 1982). The sample taken east of Baffin Island showed relatively high fecundity in cold water (0.7°-1.8°C) compared with other cold water areas. Also, average size of females was larger than encountered else- where with largest females carrying clutches in ex- cess of 4,300 eggs. This is similar to a situation in the Sea of Japan where female shrimp carried similar numbers of eggs as those (at comparable lengths) off Labrador. Again, greater sizes were attained and egg counts as high as 4,900 were en- countered (Ito 1976). Growth and maturation are delayed in colder water (Allen 1959; Rasmussen 1969; Butler 1971) and shrimp in these two cold water environments likely live longer than conspe- cifics on the Labrador Shelf. Dupouy et al. (1981) concluded that shrimp off Baffin Island spawned intermittently based on the high proportion of nonspawning females observed during a survey in 1979. If all females do not spawn annually, more time is available for growth. (Oviger- ous females do not molt.) This can account for the larger sizes attained in the colder area. Failure to spawn annually reduces reproductive potential but is compensated to some degree by the large sizes females attain (larger females carry more eggs) and the apparently increased longevity. Samples taken in 1982 in the Hudson Strait and North Labrador Sea came from waters of 0.6° and 0.5°C, respectively, but only data from the former were significantly different (a = 0.05) from samples taken in the warmer Hopedale and Cartwright Channels. Data from Haynes and Wigley (1969) showed higher fecundity in warmer water (~5°C) of the Gulf of Maine where a 28 mm female can pro- 556 PARSONS and TUCKER: FECUNDITY OF NORTHERN SHRIMP duce around 2,800 eggs compared with 1,900-2,000 in the Cartwright Channel (2°-3°C). In the Gulf of St. Lawrence, temperatures were similar to those in the Gulf of Maine but fecundity in 1970 (E. J. Sandeman4 unpubl. data) was comparable with levels observed in the colder Labrador channels. Allen (1959) reported smaller shrimp and fewer eggs for P. borealis in the North Sea (~9°C) compared with the colder area off Southern Norway (7°C). CONCLUSIONS Fecundity of Pandalus borealis in the areas of the Northwest Atlantic considered in this study was generally lower than observed previously in the Gulf of Maine (Haynes and Wigley 1969) and off South- ern Norway (Rasmussen 1953). Fecundity can vary seasonally, annually, and between areas, making conclusions based on such data difficult. Skuladot- tir et al. (1978) concluded that fecundity does not seem to be a useful characteristic for distinguishing between populations unless it is certain that no egg loss or hatching has taken place. The results of the present study concur with these findings and those of Teigsmark (1983) which also showed that annual variation within areas also must be considered. In some comparisons between areas, there ap- pears to be reduced egg production in areas with low environmental temperature. In others, this is not at all apparent, especially at extremely cold and warm temperatures. Thus, there is no clear relation- ship between fecundity and environmental temper- ature, especially at the extremes of the range of temperature tolerance. Squires (1968) described warm water areas as areas of high reproductive potential for shrimp and colder regions as areas of low reproductive poten- tial. The cold water bays of Newfoundland and the eastern Hudson Strait fit into the latter category in terms of shrimp fecundity. Other cold water con- centrations of shrimp appear to be better adapted such as those off Baffin Island, in the North Lab- rador Sea and Sea of Japan. In these cases, en- vironmental conditions other than temperature (e.g., availability of nutrients) may be more impor- tant. ACKNOWLEDGMENTS We are grateful to the many technicians and 4E. J. Sandeman, Fisheries Research Branch, Department of Fisheries and Oceans, P.O. Box 5667, St. John's, Newfoundland A1C 5X1, Canada. casual employees who assisted in collecting the data over the years and performed the laborious task of counting the eggs. In this regard, the services of W. Edison are particularly appreciated. Assistance in the statistical analyses was provided by D. Stansbury. LITERATURE CITED Allen, J. A. 1959. On the biology of Pandalus borealis Kr^yer, with reference to a population off the Northumberland coast. J. Mar. Biol. Assoc. U.K. 38:189-220. Berenboim, B. I. 1982. Reproduction of the shrimp Pandalus borealis popula- tions in the Barents Sea. Okeanologiya 22(1):118-124. Butler, T. H. 1971. A review of the biology of the pink shrimp, Pandalus borealis Krtfyer 1838. Can. Fish. Rep. 17:17-24. Couture, R. 1970. Reproduction de Pandalus borealis Krtfyer (Crustacea, Decapoda) dans le fjord du Saguenay. Nat. Can. 97:825- 826. Dupouy, H., C. Leroy, and J. Frechette. 1981. Etude des Stocks de Crevette Pandalus borealis du Detroit de Davis. Sci. Peche, Bull. Inst. Peches Marit. 311, mars 1981, 21 p. Haynes, E. B., and A. L. Wigley. 1969. Biology of the northern shrimp Pandalus borealis in the Gulf of Maine. Trans. Am. Fish. Soc. 98:60-76. Horsted, Sv. Aa., and E. Smidt. 1956. The deep sea prawn (Pandalus borealis Kr.) in Green- land waters. Meddelelser fra Danmarks Fiskeri-og Havun- ders^gelser. Ny Serie, Bind I, Nr. 11, 118 p. Ito, H. 1976. Some findings concerning Pandalus borealis Kr^yer originating in the Sea of Japan. Bull. Jpn. Sea Reg. Fish. Res. Lab. 27, p. 75-89. Rasmussen, B. 1953. On the geographical variation in growth and sexual development of the deep sea prawn (Pandalus borealis Kr.). Norweg. Fish. Mar. Invest. Rep. 10(3):1-160. 1969. Variations in protandric hermaphroditism of Pandalus borealis. FAO Fish. Rep. 57:1101-1106. Skuladottir, U., E. Jonsson, and I. Hallgrimsson. 1978. Testing for heterogeneity of Pandalus borealis popula- tions at Iceland. ICES CM. Doc. 1978/K:27, 41 p. Squires, H. J. 1968. Relation of temperature to growth and self-propogation of Pandalus borealis in Newfoundland. FAO Fish. Rep. 57:243-250. 1970. Decapod crustaceans of Newfoundland, Labrador and the Eastern Canadian Arctic. Fish. Res. Board Can., Manuscr. Rep. Ser. No. 810, 212 p. Stickney, A. P. 1981. Laboratory studies on the development and survival of Pandalus borealis eggs in the Gulf of Maine. In T. Frady (editor), Proceedings of the International Pandalid Shrimp Symposium, Kodiak, Alaska, 1979, p. 395-406. Sea Grant Rep. 81-3. Stickney, A. P., and H. C. Perkins. 1979. Environmental physiology of northern shrimp, Pan- dalus borealis. Completion Report. Maine Dep. Mar. Res., Proj. 3-277-R, 66 p. 557 FISHERY BULLETIN: VOL. 84, NO. 3 Teigsmark, G. Thomassen, T. 1983. Populations of the deep-sea shrimp (Pandalus borealis 1977. Comparisons of growth, fecundity and mortality be- Kr^yer) in the Barents Sea. Fiskeridir. Skr. Serv. Havun- tween two populations of Pandalus, borealis in Northern Nor- ders. 17:377-430. way. ICES CM. Doc. 1977/K:38, 16 p. 558 INCIDENTAL DOLPHIN MORTALITY IN THE EASTERN TROPICAL PACIFIC TUNA FISHERY, 1973 THROUGH 1978 Bruce E. Wahlen1 ABSTRACT Since the late 1950's, large numbers of dolphins have been killed incidentally in the yellowfin tuna purse seine fishery in the eastern tropical Pacific. Estimates of numbers of dolphins killed incidentally in this fishery from 1973 through 1978 were made previously using a stratified ratio estimator. Previous estimates were revised by reducing the number of strata and incorporating revisions in the data. Revised estimates of total mortality, which are consistently more precise than previous estimates, declined from about 100,000 dolphins per year from 1973 through 1976 to about 25,000 and 15,000 during 1977 and 1978. The decline in estimated mortality between 1976 and 1977 was primarily the result of a decline in the kill rate which coincided with a significant management action in late 1976. Other examples dur- ing the 1964 through 1982 period of such a temporal correspondence between a change in the number or distribution of dolphins killed and legal or management actions are discussed. Since the late 1950's, tuna purse seine fishermen operating in the eastern tropical Pacific Ocean (ETP) have exploited several dolphin species— pri- marily spotted dolphins, Stenella attenuata, and spinner dolphins, S. longirostris, and also striped dolphins, 5. coeruleoalba, and common dolphins, Delphinus delphis— to locate and catch yellowfin tuna, Thunnus albacares. Perrin (1969) described the process of deploying, or setting, the net around the tuna and dolphins, and then releasing the dolphins while retaining the tuna. During this pro- cess, however, large numbers of dolphins have been killed incidentally by becoming entangled in the purse seines (Smith 1983). The U.S. Marine Mammal Protection Act of 1972 mandated the Secretary of Commerce to make periodic assessments of the condition of dolphin populations involved in this ETP fishery. As a result of a 1976 ruling by a U.S. District Court regarding regulations promulgated under the Act, the Federal Government established annual dolphin mortality limits for the U.S. registered fleet (Fox 1978). Esti- mates of annual dolphin mortality have been an in- tegral component of periodic assessments (Smith 1983). Estimates of cumulative dolphin mortality made throughout the year are used to monitor mortalities relative to the annual limits (Lo et al. 1982). When a particular limit is reached, regulations prohibit U.S. registered vessels from fishing on the affected populations for the remainder of the year. In Octo- ber 1976, the National Marine Fisheries Service (NMFS) issued a prohibition notice for the first time (Federal Register 1976), but because of litigation the notice did not become effective until November 1976. In recent years, researchers have published sev- eral estimates and revisions of estimates of dolphin mortality incidental to this fishery. For the period 1959-78, estimates have been made by Smith (19792), Lo et al. (1982), Smith (1983), and Lo and Smith (1986); for the years 1979-83, see Allen and Goldsmith (1981, 1982), Lo et al. (1982), Hammond and Tsai (1983), Hammond (1984), and Hammond and Hall (1985). Lo et al. (1982) suggested that previous estimates of dolphin mortality incidental to this fishery were based on a stratification scheme with an unneces- sarily large number of strata. In this paper, I revise the 1973-78 estimates for U.S. registered purse seiners by reducing the number of strata and by in- corporating revisions in the data. DATA Sample data were obtained from recorded obser- vations of scientific observers who had been placed by the NMFS aboard selected U.S. registered tuna purse seine vessels fishing in the ETP. Data re- corded by these observers included the type, date, 'Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, P.O. Box 271, La Jolla, CA 92038. Manuscript accepted September 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 2Smith, T. D. (editor). 1979. Report of the status of porpoise stocks workshop, La Jolla, Calif., 27-31 August 1979. Southwest Fish. Cent. La Jolla Lab., Natl. Mar. Fish. Serv., NOAA, Admin. Rep. LJ-79-41, 120 p. 559 FISHERY BULLETIN: VOL. 84, NO. 3 location, and estimated tuna catch of each set, and other information describing the fishing operation. For sets involving dolphins (dolphin sets), the ob- servers collected additional data, including the number of dolphins killed by species. In 1973, all sampled trips were arranged with vessel captains under a voluntary sampling pro- gram. Beginning in 1974, trips to be sampled were determined from a randomly ordered list of vessels. Which trips were actually sampled during 1974 and 1975 depended on several factors, including the cooperation of the captains. Because of uncertain- ties about the cooperation of the captains and the number of fishing trips that would be made in a year, observers were placed on vessels as soon as possi- ble. Thus, before 1976, the planned number of sam- pled trips was frequently obtained in the first half of the year. The sampling process became more ran- dom starting in 1976, when participation in the sampling program became mandatory for captains making sets on dolphins. I extracted independent information for the population of all fishing trips by U.S. registered tuna purse seiners in the ETP from the Inter- American Tropical Tuna Commission (IATTC) logbook data base. This data base contains abstracts of vessels' logbooks obtained by IATTC personnel. An in- dividual entry in the logbook data base provides information about one or more sets, including num- ber of sets, set type, date and location, estimated tuna catch by species, but not numbers of dolphins killed. The logbook data are incomplete in that number of sets may be missing, set type may not be re- corded, and information for some sets of a trip and for all sets of some trips may be omitted (Punsly 1983). To compensate for these omissions from the logbook data, Punsly (1983) estimated the total number of dolphin sets made by all U.S. and non-U. S. seiners in the ETP. He then modified this procedure to estimate the total number of dolphin sets made by U.S. seiners only (Table 1). METHODS I stratified the data to allow for potential differ- ences in dolphin kills. If kills do indeed differ among strata, then a stratified estimator may be superior to an unstratified estimator in two respects. First, a stratified estimator will have a smaller standard error and thus be more precise. Second, if the sam- ple data are unrepresentative of the population with respect to these strata, a stratified estimator will be less biased. Therefore, estimates of incidental dolphin mortal- ity by species or species grouping were computed using a stratified kill-per-set ratio estimator, follow- ing the general approach described by Lo et al. (1982). I excluded trips which made no dolphin sets and experimental-gear trips from the sample and the population. However, I added dolphin kills in- cidental to the experimental-gear trips as constants to the mortality estimates. I stratified the dolphin set data by four factors used in previously published estimates: 1) year of the set, 2) fish-carrying capacity of the vessel, or simply vessel capacity, 3) period within year of the set, and 4) geographic location of the set. Vessel capacity was divided into two categories— small and large. The breakpoint between categories was deter- mined by examining the cumulative distribution of sampled trips by capacity. Periods were defined to be quarters of the year, considering the results of Wahlen and Smith (1985). The ETP was divided in- to three geographic areas— North Inside, North Out- side, and South (Fig. 1)— because mean kill (per set) after 1978 has been shown to differ among these areas.3 In previous estimates, the amount of tuna caught in the set was included as a stratification factor. However, Hammond and Tsai (1983) found that stratification by this factor made very little differ- ence in their estimates. For that reason and to avoid possible overstratifi cation, I omitted amount of tuna caught as a stratification factor. I pooled dolphin set data over strata when it was determined that between- strata differences in mean kill were not statistically significant or that sample sizes were otherwise too small. I prorated the esti- mated numbers of dolphin sets made by U.S. seiners (Table 1) among the pooled strata according to pro- 3K. -T. Tsai, Inter-American Tropical Tuna Commission, c/o Scripps Institute of Oceanography, La Jolla, CA 92093, pers. com- mun. December 1983. Table 1.— Estimated number of dolphin sets made by U.S. purse seiners fishing in the eastern tropical Pacific, by year.1 Year Number of dolphin sets 1973 1974 1975 1976 1977 1978 8,341 7,475 7,902 7,126 7,239 4,214 'Peterson, C. L. (editor). 1984. The quarterly report October-December 1983 of the Inter-American Tropical Tuna Commission. Inter-Am. Trop. Tuna Comm., c/o Scripps Inst. Oceanogr., La Jolla, CA 92093. 560 WAHLEW: INCIDENTAL UUL^HIIN MUKiALlTK 40°N 20c - 20e 40"S 1 Si o to CM r ' ' ir7 ' " NORTH 20° N. \-V7 OUTSIDE 5 o o 0* 5eN. northVI^) INSIDE ^^ny^^' Si (j o V i SOUTH ^ - i i i \ 160»W 140* 120c 100' 80« Figure 1.— The three areas of the eastern tropical Pacific used to stratify the data, bounded by lat. 40°N, long. 160°W, lat. 40°S, and the western coastline of the North and South American continents. portions of known dolphin sets which were cal- culated from logbook data. I tested for significant between-strata differences in mean total kill using analysis of variance (ANOVA) methods of BMDP programs P7D and P2V (Dixon 1983). Violation of the ANOVA assumption of equal cell variances may seriously distort significance probabilities in unbalanced models such as in this study (Glass et al. 1972). Because such distortion could be great, test results were considered to be inconclusive when significance probabilities were close to 0.05. I was unable to test for the combined effect of all four stratification factors using the whole data set because data were sparse or unavailable in many of the 144 cells of the proposed four-factor stratifica- tion. Thus, ANOVA results for restricted subsets of the data containing adequate sample sizes were assumed to hold for subsets with inadequate sam- ple sizes. To eliminate significant between-factor interactions, I tried logarithmic and power trans- formations of the dependent variable, total number of dolphins killed. When these transformations failed to eliminate the interactions, I partitioned the analysis into individual levels of the interacting variables. When it was necessary to determine where the within-factor differences occurred, £-tests for differ- ences between all pairs of cell means were made. Since I tested for differences between all pairs rather than between a few preselected pairs, a dif- ference was considered significant if its significance probability was less than the quotient of 0.05 and the number of pairs. This Bonferroni adjustment to the significance level of each test assured a level of 0.05 simultaneously across all tests (Snedecor and Cochran 1980). I computed lvalues for pairwise differences using separate rather than pooled variance estimates because the cell variances were unequal according to Levene's test; this test was selected because it is more robust under nonnormality than either the common F-ratio or Bartlett's test (Brown and For- sythe 1974). Degrees of freedom were calculated with Satterthwaite's approximation, so that sig- nificance probabilities could be obtained from an ordinary ^-distribution (Snedecor and Cochran 1980). ' 561 FISHERY BULLETIN: VOL. 84, NO. 3 RESULTS Table 2.— Cumulative distributions of number of sampled U.S. purse seine trips, by vessel capacity (tons) for the years 1973-78, with relative frequencies (percents) in parentheses. Vessel Year capacity (tons) 1973 1974 1975 1976 1977 1978 <200 0 0 0 0 0 0 (0) (0) (0) (0) (0) (0) <300 0 1 1 0 0 0 (0) (3) (3) (0) (0) (0) <400 1 3 2 3 0 0 (4) (8) (7) (7) (0) (0) <600 5 9 5 11 15 8 (22) (24) (17) (24) (15) (8) <800 15 15 8 20 34 27 (65) (41) (28) (43) (35) (26) <1,000 16 20 16 24 39 33 (70) (54) (55) (52) (40) (32) <1,200 20 30 26 34 70 67 (87) (81) (90) (74) (71) (66) Total 23 37 29 46 98 102 (100) (100) (100) (100) (100) (100) Setting the breakpoint between small and large vessel capacities at 600 tons (or lower) or at 1,200 tons would, in each case, create small and large vessel categories with severely unbalanced sample sizes (Table 2). The percent of sampled vessels with capacity < 1,000 tons was more stable over years than the percent of sampled vessels <800 tons, especially from 1973 through 1976. Therefore, the breakpoint between vessel categories was set at 1,000 tons. Because of data sparseness, particularly in the North Outside and South areas (Table 3), I made three multiway ANOVA tests restricted to subsets of the whole data set: 1) two-way test of year and vessel capacity, restricted to the North Inside area and the second quarter, 2) three-way test of year, quarter, and vessel capacity, restricted to the North Inside area and the first two quarters during 1973 through 1976, and 3) four-way test, restricted to the North Inside and North Outside areas, and the Table 3.— Number of dolphin sets made during sampled trips, by year, vessel capacity (tons), quarter of the year, and area. Vessel capacity (tons) Quarter Area Total Year Vessel capacity (tons) Quarter Area Year North Inside North Outside South North Inside North Outside South Total 1973 <1,000 1 2 3 4 325 167 0 0 0 47 0 0 9 0 0 0 334 214 0 0 1976 <1,000 1 2 3 4 155 41 18 20 0 27 71 13 0 0 0 4 155 68 89 37 Total 492 47 9 548 Total 234 111 4 349 >1,000 1 2 3 4 116 43 0 0 0 2 0 0 5 0 0 0 121 45 0 0 >1,000 1 2 3 4 129 42 17 16 0 0 80 6 118 0 1 0 247 42 98 22 Total 159 2 5 166 Total 204 86 119 409 1974 <1,000 1 2 3 4 459 88 0 0 0 0 6 0 0 0 0 0 459 88 6 0 1977 <1,000 1 2 3 4 3 239 433 53 0 67 134 0 1 0 0 0 4 306 567 53 Total 547 6 0 553 Total 728 201 1 930 £1,000 1 2 3 4 362 57 31 0 0 0 12 0 0 0 0 0 362 57 43 0 5*1,000 1 2 3 4 0 563 1,034 356 0 86 218 30 2 16 23 31 2 665 1,275 417 Total 450 12 0 462 Total 1,953 334 72 2,359 1975 <1,000 1 2 3 4 404 106 0 0 0 0 0 0 3 4 0 0 407 110 0 0 1978 < 1,000 1 2 3 4 170 75 86 59 0 50 138 6 13 0 0 0 183 125 224 65 Total 510 0 7 517 Total 390 194 13 597 >1,000 1 2 3 4 268 111 47 0 0 0 0 0 0 5 0 0 268 116 47 0 > 1,000 1 2 3 4 320 117 104 165 0 77 255 8 52 1 2 15 372 195 361 188 Total 426 0 5 431 Total Total 706 6,799 340 1,333 70 305 1,116 8,437 562 WAHLEN: INCIDENTAL DOLPHIN MORTALITY second and third quarters during 1977 and 1978. Additionally, pairwise i-tests were made to isolate annual and quarter differences detected by the above tests. Tests for Year Differences Sample statistics of mean kill by year and vessel capacity (Test 1) revealed an unbalanced design and suggested that cell variances were related to cell means (Table 4). For each of several transformations of total kill, neither the interaction between vessel capacity and year nor the difference between vessel capacities was significant, but the difference among years was significant. To determine where the yearly differences oc- curred, the data were pooled over vessel capacity so that t-tests for differences between each pair of yearly means could be made. The resulting one-way classification by year was unbalanced and charac- terized by significantly different cell variances (P < 0.001), and suggested that the means from 1973 Table 4.— Mean of total number of dolphins killed (k), standard deviation (s), and number of dolphin sets (d) for sampled trips, by year and vessel capacity (tons) for all sets made in the North In- side area during quarter two. Significance probabilities (P) obtained for 2-way ANOVA's on transformed values of total kill: interaction (P > 0.1 180), year (P < 0.001), and vessel capacity (P > 0.7851). Vessel Year (tons) tistic 1973 1974 1975 1976 1977 1978 <1,000 k 8.62 6.20 16.33 8.58 3.08 2.35 s 18.84 10.94 40.73 20.08 11.64 7.34 d 167 88 106 41 239 75 >1,000 k 11.49 5.25 8.79 14.57 2.53 1.18 s 25.96 14.08 17.43 32.44 7.84 3.31 d 43 57 111 42 563 117 Pooled k 9.21 5.83 12.47 11.61 2.69 1.64 s 20.46 12.23 31.23 27.05 9.13 5.28 d 210 145 217 83 802 192 Table 5.— Matrix of significance probabilities associated with f-tests for differences between pairs of annual means of total number of dolphins killed. Signifi- cant values, required by the Bonferroni adjustment to be <0.0033, are indicated by "*", and nearly signifi- cant values are indicated by " + ". Data are from sam- pled dolphin sets made during quarter two in the North Inside area. Year Year 1974 1975 1976 1977 1978 1973 0.0526 0.2007 0.4659 0.0000* 0.0000* 1974 0.0050 0.0681 0.0037 + 0.0002* 1975 0.8139 0.0000* 0.0000* 1976 0.0037 + 0.0013* 1977 0.0344 through 1976 were larger than the means from 1977 and 1978 (Table 4). Results from pairwise i-tests indicated that means were not significantly different within each of the two periods from 1973 through 1976 and from 1977 through 1978 (Table 5). However, each of the means from 1973 through 1976 was significantly different (or nearly so) from each of the means from 1977 and 1978 (Table 5). Based on these results, I divided the data into two periods, 1973 through 1976 and 1977 through 1978, for further tests within each period. Tests for Differences Within the 1973-76 Period The three-way ANOVA table by year, quarter, and vessel capacity (Test 2) was unbalanced and sug- gested that cell variances were unequal (Table 6). Furthermore, no between-f actor interactions were significant. The test for year differences was incon- clusive; however, there is evidence for a year effect within only one cell (1976, quarter 1, small vessels), and there is no consistent yearly pattern of means within rows of the table. Therefore, I concluded that annual means during 1973-76 were not significant- ly different and, hence, I pooled over years within this period. I also pooled over vessel capacity since it was not a significant effect during these years. Before 1976, sample data were unrepresentative of quarter and area, since nearly all data were ob- tained from trips made during the first half of the year and, thus, within the North Inside area. Data sparseness in the North Outside and South areas, Table 6.— Mean of total number of dolphins killed (k), standard deviation (s), and number of dolphin sets (d) for sampled trips, by year, quarter, and vessel capacity (tons) for all sets made in the North Inside area during the first two quarters of 1973-76. Signifi- cance probabilities (P) obtained for a 3-way ANOVA on total kill: interactions (P> 0.1374), year (P = 0.0534), quarter (P = 0.0501), and vessel capacity (P = 0.8219). Vessel capacity (tons) Year <1,000 >1,000 itistic 1973 1974 1975 1976 k 23.15 14.10 16.45 3.47 s 69.70 40.39 47.31 8.12 d 325 459 404 155 k 8.62 6.20 16.33 8.59 s 18.84 10.94 40.73 20.08 d 167 88 106 41 k 15.75 11.23 15.45 10.86 s 27.64 24.85 34.37 24.35 d 116 362 268 129 k 11.49 5.25 8.79 14.57 s 25.95 14.08 17.43 32.44 d 43 57 111 42 563 FISHERY BULLETIN: VOL. 84, NO. 3 resulting from the unrepresentative areal sample, precluded testing for an area effect during the 1973-76 period (Table 3); however, after 1978 mean kills were shown to differ among the three areas, as noted earlier. Therefore, to minimize the amount of bias which might be introduced into the estimates from a sample which was unrepresentative of area, I retained the three-area stratification. Small sample sizes during 1973-76 dictated pool- ing over all quarters within both the North Outside and South areas and over quarters 3 and 4 within the North Inside area (Table 3). The test result for quarter 1 and 2 differences in the North Inside area was inconclusive (Table 6). Based on bias considera- tions similar to those above, I did not pool data from quarters 1 and 2 in the North Inside area in case their means did indeed differ. After pooling over year, vessel capacity, and quarter as indicated above, five strata remained for the 1973-76 data: 1) North Inside, quarter 1, 2) North Inside, quarter 2, 3) North Inside, quarters 3 and 4 pooled, 4) North Outside, all quarters pooled, and 5) South, all quarters pooled. Tests for Differences Within the 1977-78 Period Interpretation of the four-way ANOVA (Test 3), restricted by data sparseness to the North Inside and North Outside areas during the second and third quarters of 1977-78 (Table 3), was complicated by significant interaction between quarters and each of the other three factors (P < 0.0159, for total kill and all transformations of total kill). Therefore, the four-way table was partitioned into two, three-way tables, one for each level of quarter (Tables 7, 8), and decisions about between-strata differences dur- ing these years were based on results obtained separately for each quarter. For second quarter data, interactions were not significant (Table 7). However, for third quarter data, interaction between year and vessel capacity was significant (P < 0.001) primarily because of the two large means recorded in the North Inside and North Outside areas by small vessels during 1978 (Table 8). Omitting the data from one extraordi- narily large kill set in each of these two cells reduces their means from 11.27 to 4.68 for the North Inside and from 9.77 to 5.46 for the North Outside. Results from the second and third quarter data were inconsistent for both year and vessel capacity. Neither effect was significant during the second quarter (Table 7), but during the third quarter (Table 8) the large means in the two cells noted above pro- vided some evidence of both a year and vessel capacity effect. Since the evidence for both a year and vessel capacity effect was confined to two, third quarter cells whose means were each strongly in- fluenced by only one set, I concluded that year and vessel capacity were not significant effects during 1977-78. Hence, I pooled over year and vessel capacity within this period. The evidence regarding an area effect during 1977-78 was also inconsistent between quarters (Tables 7, 8); however, I retained area as a stratification factor since it was shown to be significant after 1978. Beginning in 1976 when the sampling program became mandatory the sample data became more representative of quarter. Thus, for the 1977-78 data, bias considerations were of lesser importance Table 7. — Mean of total number of dolphins killed (k), standard deviation (s), and number of dolphin sets (d) for sampled trips, by area, vessel capacity (tons), and year for all sets made in the North Inside and North Outside areas during quarter two of 1977-78. Significance probabilities (P) obtained for a 3-way ANOVA on total kill: interactions (P > 0.2095), area (P = 0.0060), vessel capacity (P = 0.8857), and year (P = 0.6050). Vessel capacity (tons) Statistic Area Year North Inside North Outside 1977 <1,000 k 3.08 2.94 s 11.64 8.68 d 239 67 >1,000 k 2.53 5.66 s 7.84 13.82 d 563 86 1978 <1,000 k 2.35 4.82 s 7.34 16.00 d 75 50 > 1,000 k 1.18 4.26 s 3.31 21.19 d 117 77 Table 8.— Mean of total number of dolphins killed (k), standard deviation (s), and number of dolphin sets (d) for sampled trips, by area, vessel capacity (tons), and year for all sets made in the North Inside and North Outside areas during quarter three of 1977-78. Vessel capacity (tons) Statistic Area Year North Inside North Outside 1977 <1,000 k 2.08 2.33 s 5.72 7.92 d 433 134 >1,000 k 2.85 2.89 s 9.81 6.45 d 1034 218 1978 <1,000 k 11.27 9.77 s 61.74 52.52 d 86 138 >1,000 k 2.00 2.81 s 4.34 10.97 d 104 255 564 WAHLEN: INCIDENTAL DOLPHIN MORTALITY in stratification decisions than for the 1973-76 data. The four- way test on 1977-78 data (Test 3) was not helpful in resolving the question of quarter differ- ences because of the interactions between quarter and each of the other three factors. However, pair- wise t-tests for differences between quarterly means in the North Inside area during 1977-78, pooled over Table 9.— Matrix of significance pro- babilities associated with f-tests for dif- ferences between pairs of quarterly means of total number of dolphins killed. No significant values, required by the Bonferroni adjustment to be <0.0083, were attained. Data are from sampled dolphin sets made in the North Inside area from 1977 through 1978. Quarter Quarter 2 3 4 1 2 3 0.7062 0.2417 0.2623 0.0684 0.0730 0.2831 year and vessel capacity, detected no significant quarter differences (Table 9). Based on that result, I pooled over all quarters in the North Inside and North Outside areas. Finally, I pooled over all quarters in the South area because of the small sam- ple sizes (Table 3). Thus, after pooling over year, vessel capacity, and quarter as indicated above, only three strata re- mained for the 1977-78 data: 1) North Inside, 2) North Outside, and 3) South. Estimates I obtained annual estimates of the total number of dolphins killed by summing estimates for each of three or five strata, depending on the year. Esti- mates for a stratum were computed as the product of (a) total number of dolphin sets (Table 10) and (b) the corresponding total kill-per-set ratio (Table 11), increased by (c) the observed total number of dolphins killed during experimental-gear trips (Table Table 10.- -Estimated number of dolphin sets (D) and number of trips (N) for the population of trips, by year within strata. North Inside Quarters North Year Statistic Quarter 1 Quarter 2 3&4 Total Outside South Total 1973 D 3,203 1,670 501 2,591 330 8,295 N 172 104 69 117 34 1974 D 3,486 1,176 242 2,453 12 7,369 N 126 92 38 93 4 1975 D 3,069 1,749 434 2,495 53 7,800 N 119 96 40 96 23 1976 D 1,618 1,520 716 2,001 729 6,584 N 127 98 92 90 76 1977 D N 5,722 186 1,128 76 252 37 7,102 1978 D N 2,811 206 1,153 58 162 27 4,126 Table 11.— By-trip means of total number of dolphins killed (k) and of number of dolphin sets (d), total kill-per-set ratio (R), estimated standard error of the total kill- per-set ratio (s), and number of sampled trips (n), by strata. North Inside North Quarters Years Statistic Quarter 1 Quarter 2 3&4 Total Outside South 1973-76 k 354.51 157.32 155.50 265.06 239.85 d 24.11 15.98 9.31 16.50 7.45 R 14.70 9.85 16.70 16.06 32.19 s 1.19 1.15 4.38 4.21 4.17 n 92 41 16 16 20 1977-78 k 58.66 57.55 22.12 d 19.88 13.71 4.73 R 2.95 4.20 4.68 s 0.23 0.51 1.03 n 190 78 33 565 FISHERY BULLETIN: VOL. 84, NO. 3 12). For example, the total estimated kill for 1977 cept that values for the species or species grouping (Table 13) was obtained as a sum of estimates of the were substituted for the totals in (b) and (c) above, total for three strata as [(5,722)(2.95) + 175] + Similarly, I estimated the variance of the number [(1,128)(4.20) + 15] + [(252)(4.68) + 0]. Estimates of dolphins killed during any year by summing for each species or species grouping were obtained variance estimates for each stratum. The estimated in the same manner as estimates of the total, ex- variance of total kill for a stratum was computed Table 12.— Total number of dolphins killed (/c), number of dolphin sets (d), and number of experimental-gear trips (n), by year within strata. These data were excluded from all sample and population statistics. North Inside North Quarters Year Statistic Quarter 1 Quarter 2 3 & 4 Total Outside South Total 1973 k 0 0 513 0 0 513 d 0 0 46 0 0 46 n 0 0 2 0 0 1974 k 0 0 497 192 0 689 d 0 0 70 36 0 106 n 0 0 2 1 0 1975 k 0 0 512 271 0 783 d 0 0 76 26 0 102 n 0 0 2 1 0 1976 k 139 1,400 111 1,886 547 4,083 d 35 256 92 153 6 542 n 2 16 7 5 2 1977 k 175 15 0 190 d 129 8 0 137 n 4 2 0 1978 k 226 27 0 253 d 77 11 0 88 n 6 2 0 Table 13.— Estimates of dolphin mortality incidental to U.S. purse seiners, by species grouping and year, with coefficients of variation in parentheses. Year Species grouping 1973 1974 1975 1976 1977 1978 Spotted 70,000 61,000 63,000 61,000 14,000 9,000 (0.12) (0.13) (0.13) (0.11) (0.08) (0.08) Spinner Eastern1 12,000 1 1 ,000 1 1 ,600 9,500 1,300 700 (0.16) (0.11) (0.11) (0.12) (0.12) (0.11) Whitebelly 20,000 16,000 17,000 19,000 3,600 2,300 (0.17) (0.19) (0.19) (0.15) (0.11) (0.10) Unidentified 8,700 7,600 7,700 7,500 60 40 (0.24) (0.26) (0.25) (0.25) (0.18) (0.17) Total 41,000 35,000 36,000 36,000 5,000 3,000 Common 8,500 7,000 8,300 6,600 3,000 1,500 (0.22) (0.25) (0.22) (0.20) (0.23) (0.24) Striped 640 380 500 800 200 130 (0.30) (0.34) (0.35) (0.33) (0.26) (0.24) Unidentified 5,000 3,700 4,000 5,400 450 300 (0.19) (0.20) (0.19) (0.26) (0.12) (0.12) Other 180 90 100 280 180 100 (0.45) (0.26) (0.26) (0.64) (0.22) (0.26) Total* 125,000 107,000 112,000 110,000 23,000 14,000 (0.10) (0.10) (0.10) (0.09) (0.06) (0.06) 'May include small number of Costa Rican spinner dolphins. 2Sum of estimated kills over species grouping not exactly equal to total estimated kill because of rounding error. 566 WAHLEN: INCIDENTAL DOLPHIN MORTALITY as the square of the product of (a) the number of dolphin sets4 (Table 10) and (b) the corresponding estimated standard error of the total kill-per-set ratio (Table 11). I computed the estimated stratum variance for each species or species grouping, sub- stituting values for the species or species grouping for the total in (b) above. I estimated the standard errors (Table 11) using mean number of dolphin sets per trip calculated from the sample rather than from the population (Lo et al. 1982). My annual estimates of the total number of dolphins killed incidentally in the U.S. purse seine fishery of the ETP ranged from a maximum of 125,000 dolphins in 1973 to a minimum of 14,000 dolphins in 1978, with coefficients of variation (CV) no greater than 10% (Table 13). The estimated mor- talities of the two species most often exploited, spotted and spinner dolphins, together accounted for about 80-90% of each annual total. DISCUSSION AND CONCLUSIONS The kill-per-set ratio, or mean kill, declined from 15 dolphins/set during the 1973-76 period to 3 dolphins/set during the 1977-78 period (Table 11, pooled over quarter and area). Many changes affect- ing dolphin kill were made during these periods, including improvements in fishing gear and dolphin- release procedures and introduction of federal regulations. The change in mean kill between 1973-76 and 1977-78 coincided with the first NMFS notice in late 1976 prohibiting fishing on dolphins for the remainder of the year. This one example of a correspondence between a change in the number or distribution of dolphins killed during purse seine sets and an identifiable legal or management action is not necessarily indicative of a cause and effect relationship. There are, however, two other ex- amples of such a temporal correspondence present in the data from 1964 through 1982. In the second example, the data prior to 1973, while sparse, suggest that the mean kill was substan- tially higher than during the 1973-76 period. Lo and Smith (1986) reported a mean kill of 46 dolphins/set based on 1964 through 1972 data, pooled over vessel capacity and catch of tuna. They found no consis- tent differences in annual mean kill during that period. The decline in the mean from 46 dolphins/ set during the 1964-72 period to 15 dolphins/set dur- 4Numbers of dolphin sets were treated as constants since no variances were provided for these estimated quantities. Therefore, my variances of estimated mortality are underestimated to an unknown, though likely small, degree. ing the 1973-76 period coincided with the passage of the Marine Mammal Protection Act in late 1972. In the third example, Wahlen and Smith (1985) demonstrated a difference between the two periods from 1979 through March 1981 and from April 1981 through 1982 in the frequency distributions of number of dolphins killed during purse seine sets. While the difference in mean kill during these two periods was not significant, the percent of dolphin sets in which no dolphins were killed (zero-kill sets) decreased significantly. This decrease coincided with a court order in March 1981 which prohibited using data collected by NMFS observers to monitor com- pliance of vessel captains with dolphin-release procedures. These three examples suggest that significant legal or management actions can affect kill rates, measured by the kill-per-set ratio or by the percent of zero-kill sets. Furthermore, such effects on the kill-per-set rate are reflected in the series of esti- mates of total numbers of dolphins killed presented here. For example, between 1976 and 1977 the num- ber of dolphin sets increased slightly (Table 10), yet total estimated mortality decreased by nearly 80% (Table 13), due primarily to the significantly lower kill rate after 1976. The further decline in estimated mortality to 14,000 dolphins in 1978 reflects both the lower kill rate and a decline in the number of dolphin sets. Thus, the decrease in the kill rate following the first enforcement of dolphin kill limits in 1976 is reflected in the decrease in the estimates of total mortality after 1976. My estimates of total annual dolphin mortality in U.S. purse seine fishing from 1973 through 1978 (Table 13) are lower, except for 1976, and more precise than those in the Status of Porpoise Stocks Workshop Report (SOPS) (Table 14). However, for each year except 1973, my estimated total is con- tained inside an approximate 95% confidence inter- val around the estimated total (t) in SOPS, where the confidence interval is computed as T ± 2 • CV • T. Thus, the differences between my point estimates and those in SOPS are small when the imprecision (large CV) of the SOPS estimates is considered. The lower precision of the SOPS estimates may be due to overstratification because of a concern that the sample might not be representative of the population, particularly during the 1973-75 period. Thus, in order to minimize bias, a large number of strata (32 per year) were defined. Tests for between- strata differences in mean kill were not made, but strata were pooled to the degree that each pooled stratum contained some sample data. However, even after pooling, some strata during the years 1973-76 567 FISHERY BULLETIN: VOL. 84, NO. 3 Table 14.— Estimates of dolphin mortality incidental to U.S. purse seiners, by species and year, from Status of Porpoise Stocks Workshop Report (Smith text footnote 2). Coefficients of variation of totals in parentheses. Year Species 1973 1974 1975 1976 1977 1978 Spotted 114,000 75,000 84,000 57,000 12,000 13,000 Spinner 65,000 62,000 70,000 39,000 5,000 4,400 Common 18,000 3,000 2,300 5,400 6,600 1,000 Striped 40 140 900 2,000 100 250 Unidentified1 0 0 0 0 0 0 Other 0 80 160 690 80 50 Total2 197,000 140,000 157,000 104,000 24,000 19,000 (0.17) (0.41) (0.17) (0.13) (0.15) (0.20) 1Kills of unidentified dolphins prorated among known species. 2Sum of estimated kills over species not exactly equal to total estimated kill because of rounding error. contained only one dolphin set. Such small sample sizes within some strata account for the lower preci- sion of the SOPS estimates relative to my estimates. While the differences between my point estimates and those in SOPS are small in a statistical sense, my estimates are consistently lower except for 1976. However, these new estimates are to be preferred on methodological grounds. I tested for statistical- ly significant between-strata differences in mean kill and pooled over strata when significant differences were not detected or when sample sizes were other- wise too small. Pooling of the data produced esti- mates which were more precise than the SOPS estimates because it resulted in fewer strata with larger sample sizes. However, to minimize the possibility of introducing bias into the estimates dur- ing the 1973-76 period, when the sample was un- representative of area and quarter, I did not pool over area and I pooled over quarter only in the event of small sample sizes. ACKNOWLEDGMENTS I am grateful to K. E. Wallace, C. J. Orange, and G. Ver Steeg for providing data, and to R. G. Punsly for modifying his procedure to estimate numbers of sets for U.S. seiners only. I also appreciate the helpful reviews by two anonymous individuals as well as those by F. G. Alverson, I. Barrett, D. G. Chapman, P. S. Hammond, R. S. Holt, N. C. H. Lo, J. M. Michalski, G. T. Sakagawa, and K.-T. Tsai. Finally, I am especially indebted to T. D. Smith for his suggestions and encouragement. LITERATURE CITED Allen, R. L., and M. D. Goldsmith. 1981. Dolphin mortality in the eastern tropical Pacific in- cidental to purse seining for yellowfin tunas, 1979. Rep. Int. Whaling Comm. 31:539-540. 1982. Dolphin mortality in the eastern tropical Pacific in- cidental to purse seining for yellowfin tuna, 1980. Rep. Int. Whaling Comm. 32:419-421. Brown, M. B., and A. B. Forsythe. 1974. Robust tests for the equality of variances. J. Am. Stat. Assoc. 69:364-367. Dixon, W. J. (editor). 1983. BMDP statistical software. Univ. Calif. Press, Berkeley, CA, 733 p. Federal Register. 1976. Dep. Commer., NOAA, Regulations governing the taking and importing of marine mammals. Prohibition on en- circling marine mammals in the course of fishing operations for yellowfin tuna. Fed. Regist. 41(201):45569. Fox, W. W., Jr. 1978. Tuna-dolphin program: five years of progress. Oceans ll(3):57-59. Glass, G. V., P. D. Peckham, and J. R. Sanders. 1972. Consequences of failure to meet assumptions under- lying the fixed effects analyses of variance and covariance. Rev. Educ. Res. 42(3):237-288. Hammond, P. S. 1984. Dolphin mortality incidental to purse-seining for tunas in the eastern tropical Pacific, 1982. Rep. Int. Whaling Comm. 34:539-541. Hammond, P. S., and M. A. Hall. 1985. Dolphin mortality incidental to purse-seining for tunas in the eastern tropical Pacific inflicted by the US fleet in 1983 and non-US fleet in 1979-1983. Rep. Int. Whaling Comm. 35:431-433. Hammond, P. S., and K.-T. Tsai. 1983. Dolphin mortality incidental to purse-seining for tunas in the eastern Pacific Ocean, 1979-81. Rep. Int. Whaling Comm. 33:589-597. Lo, N. C. H., J. E. Powers, and B. E. Wahlen. 1 982. E stimating and monitoring incidental dolphin mortality in the eastern tropical Pacific tuna purse seine fishery. Fish. Bull, U.S. 80:396-401. Lo, N. C. H., and T. D. Smith. 1986. Incidental mortality of dolphins in the eastern tropical Pacific 1959-1972. Fish. Bull., U.S. 84:27-34. Perrin, W. F. 1969. Using porpoise to catch tuna. World Fish. 18(6):42- 45. Punsly, R. G. 1983. Estimation of the number of purse-seiner sets on tuna 568 WAHLEN: INCIDENTAL DOLPHIN MORTALITY associated with dolphins in the eastern Pacific Ocean dur- Snedecor, G. W., and W. G. Cochran. ing 1959-1980. [In Engl, and Span.] Inter-Am. Trop. Tuna 1980. Statistical methods. 7th ed. Iowa State Univ. Press, Comm. Bull. 18:229-299. Ames, 507 p. Smith, T. D. Wahlen, B. E., and T. D. Smith. 1983. Changes in size of three dolphin (Stenella spp.) popula- 1985. Observer effect on incidental dolphin mortality in the tions in the eastern tropical Pacific. Fish. Bull., U.S. 81: eastern tropical Pacific tuna fishery. Fish. Bull., U.S. 83: 1-13. 521-530. 569 DISTRIBUTION AND REPRODUCTIVE BIOLOGY OF THE GOLDEN KING CRAB, LITHODES AEQUISPINA, IN THE EASTERN BERING SEA David A. Somerton1 and Robert S. Otto2 ABSTRACT The golden king crab is a large anomuran that supports a new, rapidly expanding fishery in the eastern Bering Sea and Aleutian Islands. Based on size, sex, and abundance data collected by U.S. observers aboard foreign trawlers and by National Marine Fisheries Service personnel aboard research vessels, we examined latitudinal and depth variation in mean size (carapace length), size at maturity, weight at size, and relative abundance. Mean size decreases by 6.2 mm for males and 4.6 mm for females with each 1 degree increase in latitude. Size at maturity decreases with increasing latitude from 130 mm for males and 111 mm for females in the southern area to 92 mm and 98 mm in the northern area. These decreases may be due to a temperature induced decrease in growth rate. Weight at size increases by 10% from the southern to the northern area owing to a latitudinal change in body shape. Mean size and relative abundance of both sexes increase with a decrease in depth, suggesting that an onshore ontogenetic migration occurs and that adult males migrate into somewhat shallower water than adult females. Fecun- dity (number of uneyed embryos) of northern females increases with size according to -24815 + 323 CL, where CL is carapace length. This relationship changes with latitude and southern females carry about 1,330 fewer eggs than equal-sized northern females. Mean length of uneyed eggs is 2.2 mm. Based on the lack of a clear seasonal change in the occurrence of eyed and uneyed clutches, golden king crab appear to have protracted, or perhaps year-round, breeding. The golden king crab, Lithodes aequispina, is a large anomuran that inhabits the upper continental slope from southern British Columbia, Canada, northward to the Bering Sea and westward to Suruga Bay, Japan (Butler and Hart 1962; Suzuki and Sawada 1978). Although similar in size to red king crab, Paralithodes camtschatica, and blue king crab, P. platypus, the traditional species harvested by Alaskan crab fisheries, golden king crab have not been intensively harvested because they live in deeper water than red and blue king crabs and are therefore more difficult and expensive to capture (McNair 1983). Since 1980, however, precipitous declines in abundance of red and blue king crabs have stimulated growth of directed fisheries for golden king crab. These fisheries expanded rapidly in the eastern Bering Sea and Aleutian Islands, and between 1981 and 1983 the catch of golden king crab increased from 50 t to 4900 t or 44% of the total king crab catch from these areas. Northwest and Alaska Fisheries Center Seattle Laboratory, Na- tional Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA; present address: Southwest Fisheries Center Hawaii Laboratory, National Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI 96822-2396. 2Northwest and Alaska Fisheries Center Kodiak Laboratory, Na- tional Marine Fisheries Service, NOAA, P.O. Box 1638, Kodiak, AK 99615. snt -sr* Manuscript accepted September 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. The fisheries for golden king crab have been managed according to regulations designed for red and blue king crabs (Alaska Department of Fish and Game 1983; Miller 1976) because little biological in- formation was available to establish more specific regulations. Although golden king crab have been studied before, published reports either concern Asian stocks (Hiramoto and Sato 1970; Suzuki and Sawada 1978; Rodin 1970) or are restricted to tax- onomy (Benedict 1895; Makarov 1938), distribution (Butler and Hart 1962; Slizkin 1974), or early life history (Haynes 1981). In 1981, the National Marine Fisheries Service (NMFS) began collecting biological data on golden king crab necessary for establishing minimum size limits and fishing seasons. We summarize these data here, focusing our attention on latitudinal and depth gradients in mean size, size at maturity, weight at size, and sex ratio as well as various aspects of the reproductive biology. We then examine the manage- ment implications of our findings. MATERIALS AND METHODS Data Sources Golden king crab were sampled by NMFS re- 571 FISHERY BULLETIN: VOL. 84, NO. 3 search personnel on stock assessment and tagging cruises and by NMFS observers aboard foreign fishing vessels. In the following we distinguish be- tween survey data and observer data because the sampling designs differed considerably. In both cases, however, crabs were measured with calipers to the nearest millimeter according to the descrip- tions in Wallace et al. (1949). Survey Data From 1981 to 1983, NMFS conducted 10 survey cruises sampling the eastern Bering Sea population of golden king crab with either bottom trawls or commercial king crab pots (Table 1). All crabs were measured for carapace length, and females were classified into one of four categories of reproduc- tive condition: 1) Non-ovigerous - no embryos or empty egg cases attached to the pleopod setae. 2) uneyed embryos - embryos without conspicu- ous dark eyes. 3) eyed embryos - embryos with dark eyes. 4) empty egg cases - empty egg cases and funi- culi attached to the pleopod setae. When opportunity occurred, one or more of the following were also collected: 1) height of the right chela of males (excluding males with partially regenerated right chela). 2) Total wet body weight of males, measured to the nearest gram on a triple beam balance or to the nearest 5 g on a handheld spring scale (excluding males with damaged exoskeletons or missing appendages). 3) egg masses from females (stored in Formalin3 diluted to 10% with seawater). Observer Data Golden king crab, like most of the other large Alaskan crabs, is classified as a prohibited species and, as such, may not be retained if captured by foreign fisheries. Because of this status, NMFS fishery observers routinely record the carapace length, sex, and number of golden king crab that are incidentally caught by foreign vessels during their normal fishing operations for other species (Nelson et al. 1981). To delineate the distribution Table 1.— Inclusive dates, latitude and depth ranges, number of crabs sampled and type of sampling gear are shown for each of the 10 golden king crab research cruises conducted by the Na- tional Marine Fisheries Service. Latitude Depth Number Year Dates (degrees N) (m) males females Gear 1981 2/12-2/23 54.4-55.1 346-472 4 6 trawl 1982 7/12-8/4 58.3-60.9 168-849 292 341 trawl 1983 1/31-2/8 52.3-52.5 183-366 188 123 pot 2/22-2/24 54.4-55.7 362-461 24 17 trawl 5/9-5/10 56.0-56.1 365-421 288 1,114 pot 5/12-5/15 57.8-58.5 329-365 489 1,753 pot 7/8-7/10 57.7-57.7 347-365 1,073 741 pot 7/14-7/18 55.9-56.3 311-384 1,285 1,012 pot 10/2-10/4 56.2-56.2 347-365 596 1,035 pot 11/15-11/21 52.4-52.6 110-283 376 404 pot 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. of golden king crab in the eastern Bering Sea, we chose to examine the 1981 and 1982 observer data obtained from Japanese small stern trawlers that fish for turbot (Reinhardtius hippoglossoides) be- cause 1) these vessels use trawls designed to remain in direct contact with the bottom and are therefore likely to catch crabs, 2) these vessels operate year- round along nearly the entire length of the continen- tal slope of the eastern Bering Sea, and 3) turbot have a depth distribution similar to that of golden king crab. Although these data are not necessarily a random sample of the golden king crab population, they are the most extensive data available and in- clude samples from the entire depth range of golden king crab during all four seasons. The number of crabs measured and the number of trawl hauls sampled are summarized by year, month, latitude, and depth (Table 2). Due to a lack of Japanese fishing effort for turbot, observer data were unavail- able for areas south of lat. 54°15'N. Both survey and observer data, in some instances, were partitioned into three latitudinal strata or subareas (Fig. 1): northern (north of lat. 58°30'N), central (between lat. 58°30'N and 54°15'N), southern (south of lat. 54°15'N and east of long. 173°00'W), which correspond approximately to the crab management districts used by the Alaska Department of Fish and Game. In addition, the observer data were partitioned into two depth strata separated at the approximate median depth (500 m) of the samples (nearly the entire depth range of golden king crab is bounded by the 200 m and 1,000 m isobaths). Methods of Analysis Size-frequency distributions by sex were con- structed from the combined 1981 and 1982 observer 572 SOMERTON and OTTO: DISTRIBUTION AND REPRODUCTION OF GOLDEN KING CRAB Table 2. — Number of trawl hauls sampled (ex- cluding hauls without crabs) and number of crabs sexed and measured by U.S. observers aboard Japanese small trawlers within the study area during 1981 and 1982. Data are summarized by depth, latitude and month. 1981 1982 Hauls Crabs Hauls Crabs Depth (m) 100 0 0 3 3 200 7 81 23 86 300 30 323 33 848 400 217 2,475 339 3,551 500 456 6,885 548 4,380 600 201 2,065 192 1,112 700 16 97 27 81 800 6 24 1 2 900 2 13 0 0 1,000 1 6 0 0 Latitude (degrees N) 53 12 18 0 0 54 7 135 132 678 55 34 455 43 184 56 165 1,582 163 879 57 59 376 75 552 58 284 3,936 151 899 59 175 2,995 166 2,019 60 200 2,472 436 4,852 Month 1 18 55 19 125 2 65 443 19 79 3 75 977 66 528 4 114 1,688 41 73 5 104 1,398 102 332 6 112 2,027 136 1,681 7 72 1,528 75 306 8 89 1,121 124 1,215 9 106 927 192 992 10 99 814 194 2,436 11 68 931 150 1,654 12 14 60 48 654 Total 936 1 1 ,969 1,166 10,063 data for each of the two depth strata within the northern and central subareas to help illustrate depth and latitude trends in the size distributions (see Figure 2). Potential bias because of the varia- tion of fishing effort with depth was minimized by first partitioning the data into 100 m depth inter- vals. Within each depth interval, a size-frequency distribution and an average catch per hour (CPH) were calculated. Size-frequency distributions, weighted by the appropriate mean CPH, were then summed over all 100 m depth intervals within each of the two depth strata. Variations in mean size, CPH, and proportion male with latitude and depth were also examined using multiple regression. Two normalizing trans- formations were used: 1) CPH was transformed to the natural log scale and 2) proportion male was transformed to the arcsine-square root scale after replacing 0 with 0.25/N and 1 with (N - 0.25)/JV, where N is the number of crabs within each trawl haul (Bartlett 1947). Egg size was estimated by randomly selecting 10 eggs from each preserved egg mass and measuring their maximum lengths (eggs are oval) to the nearest 0.1 mm with an ocular micrometer. The remainder of each egg mass was air dried and, after separating the eggs from the pleopods and setae, weighed to the nearest 0.1 mg. Two subsamples of about 200 eggs each were randomly selected from each dried egg mass and then weighed and counted. Fecundity was then estimated by dividing the total weight of an egg mass by the average of the two estimates of individual egg weight that were obtained from that egg mass. Male size at maturity was estimated from the allometric growth of the right chela. When king crab chela measurements are plotted against carapace measurements on log-log axes, the data conform to two straight lines that intersect at the average carapace length at maturity (see Figure 3) (Somer- ton 1980; Somerton and Macintosh 1983). To estimate this size, we used the computer method described in Somerton and Macintosh (1983) which fits a pair of intersecting straight lines by iteratively varying the carapace length at the intersection point until the residual sum of squares about the lines is minimized. Variance of the male size at maturity was estimated using a computer technique known as bootstrapping (Efron and Gong 1983). In our ap- plication, the method consisted of randomly choos- ing, with replacement, 50 subsamples equal in size to the original data set. For each subsample, the size at maturity was estimated by fitting the two line model. Variance of the estimated size at maturity was then computed as the variance among the 50 independent estimates. Although we attempted to detect and exclude par- tially regenerated chelae in the field, we were not always successful. Measurements from partially regenerated chelae can increase the variance of estimates of male size at maturity; therefore, these measurements were removed from the data set before analysis using a sequential outlier elimina- tion technique described in Somerton and Macin- tosh (1983). Golden king crab females were considered to be mature, if they had eggs or empty egg cases at- tached to the pleopod setae. Although we are not certain that this is always true, for red and blue king crabs, adult females extrude eggs soon after every molt and the empty egg cases remain attached to the pleopod setae until the next molt (Somerton and Macintosh 1985). 573 FISHERY BULLETIN: VOL. 84, NO. 3 61 ST. MATTHEW ISLAND /l S fV MAT ^V-ISL NORTHERN PRIBILOF ISLANDS PftlBILOF CANYON CENTRAL 59 -57 ■55 -53 *fr/^' SOUTHERN 51 180 178 176 174 172 170 168 Figure 1.— Areas of the Bering Sea and eastern Aleutian Islands where golden king crab were sampled by U.S. fishery observers and National Marine Fisheries Service research cruises. Golden king crab occur primarily in a region bounded by the 200 m (solid line) and 1,000 m (dashed line) isobaths. The dark lines indicate the separation of this region into the three latitudinal strata discussed in the text. Female size at maturity was estimated as the size at which 50% of the crabs were mature. Weighted nonlinear regression (weights equal to the inverse of the binomial variance at each size) was used to fit a logistic equation to the percentage mature within 5 mm size intervals. The fitted logistic equa- tion was then evaluated to determine the carapace length corresponding to 50% maturity. Variance of this size was estimated using the formula provided in Somerton (1980). 574 SOMERTON and OTTO: DISTRIBUTION AND REPRODUCTION OF GOLDEN KING CRAB BIOLOGICAL VARIATION WITH DEPTH AND LATITUDE Mean Size Size-frequency distributions of golden king crab, based on the combined 1981 and 1982 observer data, are shown by sex, area, and depth strata in Figure 2. Linear trends in mean size with depth and latitude were examined statistically using multiple regres- sion. For each sex in each year, when carapace length was regressed against latitude and depth simultaneously, ignoring interaction, both the latitude coefficient and the depth coefficient were negative and highly significant (P < 0.001). Aver- aged over both years, mean size decreased by 6.2 mm for males and by 4.6 mm for females for each 1 degree increase in latitude, and mean size de- creased by 7.9 mm for males and by 6.2 mm for females with each 100 m increase in depth. The latitudinal decrease in size probably reflects a latitudinal decrease in growth rate. Two shallow- water Bering Sea crabs, Chionoecetes bairdi and C. opilio, also show a latitudinal decrease in size, and this decrease was correlated with a latitudinal decrease in maximum summer water temperature (Somerton 1981a). Although we lack sufficient tem- perature data from the depths inhabited by golden king crab to allow a statistical test, it is likely that mean annual bottom temperature also decreases 0.25 -I MALES NORTHERN AREA 0.35 FEMALES NORTHERN AREA 0.00 0 25 50 75 100 125 150 175 200 0.00 0 25 50 75 100 125 150 175 200 0.25 MALES CENTRAL AREA 0.00 0 25 50 75 100 125 150 175 200 CARAPACE LENGTH (MM) 0.35 FEMALES CENTRAL AREA 0-30 - 0.25 - < o 0.20 o I UJ CL x 0.15- o 0. 10 0.05 - 0.00 <500 M >500 M CARAPACE LENGTH (MM) Figure 2.— Size-frequency histograms for males and females of golden king crab, by depth strata and subarea. Due to differences in the sampling intensity with depth (Table 2), frequencies have been standardized to catch per hour of trawling. 575 FISHERY BULLETIN: VOL. 84, NO. 3 with increasing latitude along the slope. If this is true, then it is reasonable to assume that growth rates are lower in higher latitudes. Part of the lat- itudinal decrease in mean size, however, is due to the greater relative abundance of small (25-50 mm) crabs in the northern area (Fig. 2). Since we have only two years of data, we do not know if the greater abundance of small crabs in the northern area is a persistent feature of the distribution. But if it is, it may indicate that greater larval settlement occurs in the northern area because of the advection of lar- vae by the northwesterly currents over the continen- tal slope (Kinder and Schumacher 1981). The decrease in size with depth may reflect an ontogenetic upslope migration. Another slope dwell- ing crab, Chionoecetes tanneri, also displays a de- crease in size with depth, and this was attributed to an offshore advection of pelagic larvae followed by an onshore migration of juveniles (Pereyra 1968). Although an onshore migration might explain the size variation with depth of golden king crab in the eastern Bering Sea, offshore advection depends on local oceanographic conditions and may not occur everywhere ovigerous golden king crab occur. For example, studies of golden king crab in other areas indicated that adults could be found in shallower water than juveniles (Hiromoto and Sato 1970), or at similar depths as juveniles but in different areas (Rodin 1970) or in deeper water than juveniles (N. Sloan4). Size at Maturity The change in the relative growth of a male's chela which occurs at maturity is more pronounced for golden king crab than it is for either blue king crab (Somerton and Macintosh 1983) or red king crab (Somerton 1980), and this allows greater precision in the estimates of size at maturity (Fig. 3). Never- theless, the estimates of male size at maturity are less precise than those for females (Fig. 4). For both sexes, however, the estimated sizes at maturity dif- fer significantly between areas and progressively decrease with increasing latitude (Fig. 4). The decrease in the size at maturity is consistent with a latitudinal decrease in growth rate; however, the decrease is greater for males than it is for females (Fig. 4). If golden king crab are similar to red king crab (Weber 1967) in that males and females grow identically while they are immature, 4N. Sloan, Department of Fisheries and Oceans, Pacific Bio- logical Station, Nanaimo, British Columbia, V9R 5K6, Canada, pers. commun. 1984. 576 then the greater latitudinal decrease in male size at maturity implies that female age at maturity in- creases, relative to that of males, with latitude. This could occur if females and males have different life history strategies to maximize their reproductive values (Bell 1980). The reproductive value of a female is largely determined by her lifetime fecun- dity. Since fecundity increases markedly with size and somatic growth decreases abruptly at matur- ity, under conditions of reduced growth, female reproductive value might be increased by delaying maturity until some optimum size is reached. The reproductive value of a male, however, is largely determined by the number of females he is able to mate with over his lifetime. Unless access to females is strictly limited to the largest males, male repro- ductive value is unlikely to be increased by delay- ing maturity. Along a gradient of decreasing growth rate, such strategies would lead to a divergence between male and female sizes and ages at maturity. Weight at Size Weight-size relationships of males were deter- mined for each of the three subareas by regressing body weight on carapace length after transforming both variables to natural logarithms. Analysis of covariance showed that the slopes of the regression lines were not significantly different (F = 0.49, df = 2, 1,079, P = 0.613), but that the intercepts were significantly different between areas (F = 19.03, df = 2, 1,081, P < 0.001). Pairwise £-tests further showed that the intercept for each area differed significantly from the other two (Bonferroni critical values; maximum P < 0.05) and that the intercepts progressively increased with increasing latitude. Males in higher latitudes are therefore propor- tionately heavier than equal-sized males from lower latitudes. This proportionate change in weight with latitude might be due to changes in body shape, such as the relative size of the chelae, that are coincident with the onset of maturity. Since the rate of chela growth increases, relative to carapace growth, at maturity, and since the size at maturity decreases with lat- itude, mature males in northern areas should have larger chelae than equal-sized males in southern Figure 3.— For the golden king crab males, chela heights, carapace lengths, and the best fitting two line model are shown for each subarea. For the females, percentage mature, within 5 mm size intervals, and the fitted logistic equation are shown for each sub- area. Estimated sizes (carapace length) at maturity are indicated by dotted lines. SOMERTON and OTTO: DISTRIBUTION AND REPRODUCTION OF GOLDEN KING CRAB MALE FEMALE 80-, 60 NORTHERN SM = 92.0 SD = 2.4 N = 205 o UJ 00 90 80 - 70 - 60 - 50 - 40 - 30 20 10 i 0 NORTHERN SM = 97.7 SD = 0.5 N = 324 200 10 30 50 70 80-| 60 CENTRAL SM = 107.0 SD = 4-6 N = 1866 UJ CD < O a: UJ 200 80]S0UTHERN eo^SM = 1 30.0 .SD = 4.0 N = 299 40 20- 10 8 40 60 80 100 CARAPACE LENGTH (MM) 200 UJ UJ o < o UJ Q. 00 90 80 70 60 50 40 30 20 10 0 SOUTHERN SM = 110.7 HSD = 0.8 N = 527 50 70 90 110 130 150 CARAPACE LENGTH (MM) 577 FISHERY BULLETIN: VOL. 84, NO. 3 1 40 1 30 X 120 I— CD -z. UJ _J 1 1 0 UJ o < 100 < en < o 90 80 -• — MALE -— FEMALES 1 1 1 1 SOUTH CENTRAL NORTH AREA Figure 4.— For both sexes of golden king crab, estimated sizes at maturity, and their 95% confidence intervals, are plotted against area. areas. To test whether this is true, chela height and carapace length relationships for adult males were compared between areas. Analysis of covariance showed that the slopes did not differ (F = 0.14, df = 2, 1,998, P = 0.87), but the intercepts differed significantly (F = 146.7, df = 2, 2,000, P < 0.001). Pairwise i-tests further showed that each intercept differed significantly (Bonferroni critical values; maximum P < 0.05) from the other two and, similar to the weight-size relationships, that the intercepts progressively increased with latitude. Thus north- ern males, which are the heaviest, have the largest chelae. By itself, chela size is unlikely to be responsible for latitudinal differences in weight because chela weight is only a small proportion of total body weight. However, chela size may be correlated with other body dimensions (for example, length of walk- ing legs) that also increase relative to carapace length at maturity. We therefore used chela height as a proxy for these dimensions and examined whether the difference in chela height could account for the difference in weight-size relationships. This was done by comparing the weight-size relationships between areas including the logarithm of chela height as a covariate. Two additional modifications of the previous weight-size comparison were made. First, since weights and chela measurements were not obtained from the same crabs in the southern area, the comparison was restricted to the northern and central areas. Second, since chela height and carapace length are linearly related only over the adult (or juvenile) size range, the comparison was restricted to males greater than the size at matur- ity in each area. When the northern (N = 129) and central (N = 614) areas were compared consider- ing only carapace length as a covariate, the slopes were not significantly different (F = 0.06, df = 2, 739, P = 0.81), but the intercepts were significant- ly different (F = 7.36, df = 1, 740, P = 0.007). When chela height was included as a covariate, however, neither the slopes (P = 0.316) nor the intercepts (P = 0.430) differed significantly between areas. This indicates that latitudinal changes in chela size, and perhaps other body measurements that also increase at maturity, account for the observed latitudinal in- crease in body weight. Juvenile weight-size relationships were also com- pared between the northern (N = 10) and central (N = 207) areas and neither the slopes (F = 0.06, df = 1, 213, P = 0.938) nor the intercepts (F = 0.19, df = 1, 214, P = 0.664) were significantly different. The weight-size relationship for male golden king crabs is therefore described by one equation for juveniles and three equations for adults. Trans- formed back to a linear scale, these relationships are Juveniles W = 0.000365 CL3-099 (N = 217, R2 = 0.88) Adults Northern W = 0.000225 CL3206 (N = 139, R2 Central W = 0.000219 CL3206 (N = 632, R2 Southern W = 0.000204 CL3206 (N = 100, R2 0.93) 0.91) 0.91) where W is body weight in grams and CL is cara- pace length in millimeters. Within the adult size range, males from the northern area are 10.3% heavier and males from the central area are 9.8% heavier than equal-sized males from the southern area. Relative Abundance and Proportion Male Relative abundance, or catch per hour (CPH), based on combined 1981 and 1982 observer data, is shown by sex, latitude, and depth in Figure 5. Linear trends in CPH with depth and latitude were examined statistically using multiple regression (depth and latitude were considered simultaneous- ly; interaction was ignored). The latitude coefficient for males was not significant (P > 0.05) in either year, but the latitude coefficient for females was positive and highly significant (P < 0.01) in both years. The depth coefficient for males was negative 578 SOMERTON and OTTO: DISTRIBUTION AND REPRODUCTION OF GOLDEN KING CRAB 3.0 -, . 3.0 54 55 56 57 58 59 60 700 0.7 0.7 55 56 57 51 LATITUDE 59 60 100 300 500 700 DEPTH Figure 5.— Catch per hour, by sex, and the proportion of males of golden king crab are shown as a function of latitude (left panels) and depth (right panels). and highly significant in both years (P < 0.01), but the depth coefficient for females, although negative in both years, was significant (P < 0.05) in only one. Although male CPH decreases significantly with depth whereas female CPH decreases significantly with latitude, CPH is not a strict linear function of depth and latitude; therefore, linear approximations mask aspects of the variability. The important point is that both male and female CPH generally increase with an increase in latitude or a decrease in depth, but at more southerly latitudes or at the shallowest depth, male CPH is considerably higher than female CPH (Fig. 5). Different trends in CPH between sexes suggested that the sex ratio of golden king crab varied spatial- ly. To investigate this further we examined the variation in proportion of males within trawl hauls having at least five crabs. The proportion of males, based on combined 1981 and 1982 observer data, is shown by latitude and depth in Figure 5. When the proportion of males was regressed against latitude and depth (using weights equal to the num- ber of crabs within each trawl haul), the latitude coefficient was negative and highly significant (P < 0.01) in both years; and the depth coefficient, although negative in both years, was significant (P < 0.05) in only one. From a biological perspective, the latitudinal decrease in the proportion of males is difficult to ex- plain; therefore, we considered possible sampling bias that could lead to an apparent change in the proportion of males. Since males are considerably larger than females in the central area but nearly the same size as females in the northern area, the proportion of males might vary with latitude due to size selectivity of the trawls. This hypothesis was tested by comparing the proportion of males be- tween the northern and central areas considering only crabs within an equal size range. To eliminate a possible confounding effect due to a sexual differ- ence in growth rate that begins at maturity, we restricted the comparison to crabs <90 mm. Based 579 FISHERY BULLETIN: VOL. 84, NO. 3 on the combined 1981 and 1982 observer data, the proportion of males was 0.51 (N = 1,375) in the cen- tral area and 0.43 (N = 8,271) in the northern area. Since the proportion of males still differed signifi- cantly between areas (2x2 contingency table, x2 = 30.7, df = 1, P < 0.001), it is unlikely that the change in the proportion of males was due to size selectivity. Furthermore, since the difference in the proportion of males appears to be established before maturity, biological explanations such as sexual dif- ferences in migratory behavior or natural mortal- ity are also unlikely. Although we cannot explain the latitudinal varia- tion in the proportion of males, we believe that the depth variation, especially the abrupt increase in the proportion of males in the shallowest depth zone, is due to sexual segregation. Sexual segregation by depth has been observed for another slope-dwelling crab, C. tanneri (Pereyra 1966). Adult female C. tan- neri occur within a rather narrow depth zone throughout the year while adult males undergo a seasonal migration from relatively shallow water in summer to the deeper water occupied by females during the winter mating period. To determine if golden king crab have a similar seasonal migration, we examined the proportion of males from the northern area at depths <400 m (the northern area had nearly equal sampling in all four seasons). Using pooled 1981 and 1982 data, analysis of variance showed that the proportion of males did not vary significantly between seasons (F = 0.13, df = 3, 179, P > 0.05). Although adult males of golden king crab probably congregate in somewhat shallower water than adult females, unlike C. tanneri this segre- gation appears to be maintained throughout the year. REPRODUCTIVE BIOLOGY significance. Since the coefficient was not signifi- cantly different from zero (F = 3.85, df = 1, 57, P = 0.06), we chose a linear relationship to describe the data. Fecundity-size relationships for females with uneyed embryos (N = 46) and eyed embryos (N = 19) from the central area were compared to deter- mine whether the relationships changed with stage of embryo development. Analysis of covariance showed that the slopes did not differ (F = 0.77, df = 1, 61, P = 0.38) but that the intercept for eyed embryos was significantly less (F = 4.89, df = 1, 62, P = 0.03) than that for uneyed embryos. At 114 mm, the median size of adult females in all areas combined, uneyed clutches were 18% greater than eyed clutches. Similar to other crab species (Wear 1974), golden king crab lose a significant number of embryos between egg extrusion and the appear- ance of embryonic eyes. Fecundity-size relationships were then compared between the northern (N = 59), central (N = 46), and southern (N = 24) areas considering only those clutches with uneyed eggs. Analysis of covariance showed that the slopes did not differ (F = 0.74, df = 2, 123, P = 0.48), but the intercepts differed significantly between areas (F = 4.38, df = 2, 125, P = 0.01). Pairwise i-tests indicated that southern and central intercepts did not differ (P = 0.99) from each other, but that both differed significantly (P = 0.01, P = 0.04) from the northern intercept. Data from the southern and central areas were therefore pooled and compared with those from the northern area. Again, the slopes did not differ (F = 1.25, df = 1, 125, P = 0.27), but the northern intercept was significantly greater (F = 8.83, df = 1, 126, P = 0.004) than the pooled central and southern inter- cept. Assuming equal slopes, the resulting fecundity- size relationships are Fecundity Fecundity-size relationships for golden king crab were estimated stagewise by examining 1) the form of the relationship, 2) whether the relationships varied with stage of embryo development, and 3) whether the relationships varied between areas. The fecundity of king crabs has been reported to increase as either a linear (Haynes 1968) or a curvi- linear (Somerton 1981b) function of carapace length. To determine which form was more appropriate for golden king crab, a second degree polynomial was fitted to the fecundity and size data from the north- ern area (all clutches contained uneyed embryos) and the coefficient of the quadratic term was tested for Northern Central-southern E = -24815 + 323 CL (N = 59, R2 = 0.79) E = -26145 + 323 CL (N = 68, R2 = 0.74) where E is number of uneyed eggs and CL is cara- pace length in millimeters. Females from the north- ern area carry, on average, 1,330 more eggs than equal-sized females from the central and southern areas. For 114 mm females, this represents a 12.6% difference in fecundity. Northern females may be more fecund than equal- sized central and southern females because they are older and size-specific fecundity often increases with age (Pianka and Parker 1975). But, it is also likely 580 SOMERTON and OTTO: DISTRIBUTION AND REPRODUCTION OF GOLDEN KING CRAB that the observed difference in fecundity is an arti- fact due to a difference in mean embryo age. We attempted to eliminate the effect of embryo age by considering only clutches with uneyed embryos, but this may not have been a sufficiently sensitive criterion of age and northern females could have had more embryos simply because they had younger em- bryos. Considering that for equal-sized females the percent difference in clutch size between eyed and uneyed stages was greater than the percent differ- ence in clutch size between areas, it is possible that the loss of embryos within the uneyed stage is suf- ficient to account for between-area differences. More precise embryo aging techniques are needed to clarify this. Egg Size To estimate the size of golden king crab eggs, we considered 1) whether egg size varied with stage of embryo development and 2) whether egg size varied between areas. When mean lengths of uneyed eggs (N = 42) and eyed eggs (N = 26) from the cen- tral area were compared, eyed eggs were found to be significantly larger than uneyed eggs (two sam- ple £-test, P < 0.001). Golden king crab eggs therefore appear to increase in size, as has been reported for other crab species (Wear 1974), dur- ing embryonic development. When mean length of uneyed eggs from the southern (N = 25) and cen- tral (N = 42) areas (no egg length data was collected from the northern area) were compared, no signifi- cant difference was found (two sample £-test, P = 0.25). Mean length of uneyed eggs, based on the pooled central and southern data, is 2.2 mm (SD = 0.1). Our estimate of egg length is similar to those reported for Asian populations of golden king crab (2.38 mm, Hiramoto and Sato 1970; 2.30 mm, Suzuki and Sawada 1978), and it is also similar to egg lengths reported for other Lithodes species (L. ant- arctica, 2.2 mm, Guzman and Campodonico 1972; L. couesi, 2.3 mm, Somerton 1981b). However, this size is more than twice as large as the egg lengths reported for Paralithodes species (P. camtschatica, 1.0 mm, Haynes 1968; P. platypus, 1.2 mm, Sasa- kawa 1975). The larger eggs of golden king crab are, in turn, reflected in the relatively large size of their first stage zoea (L. aequispina, 7.3 mm TL, Haynes 1981; P. camtschatica, 4.6 mm TL, Sato and Tanaka 1949; P. platypus, 4.9 mm TL, Hoffman 1968). The larger size of L. aequispina larvae may allow them to withstand starvation for a longer period or may allow them to capture a wider size range of prey than Paralithodes larvae. If this is true, golden king crab larvae may not need to ascend to the photic zone but instead stay at greater depths. Evidence supporting this hypothesis is provided by a study on crab larvae that sampled the upper 50 m near the edge of the eastern Bering sea continental shelf (Fig. 1). Although both P. platypus and P. cam- tschatica larvae were found, L. aequispina larvae were not (D. Armstrong5). Seasonality of Reproduction King crabs either can be synchronous and seasonal in their egg extrusion and embryo hatching, as reported for P. camtschatica (Powell et al. 1973), or they can be asynchronous and lack seasonal periodicity, as reported for L. couesi (Somerton 1981b). To determine which pattern better charac- terizes golden king crab, we tabulated the percent- age of mature females in each of the three reproduc- tive conditions by area and by quarter (Table 3). If the reproductive cycle were synchronous and seasonal, then each of the three categories of repro- ductive condition should predominate sequentially over the course of a year, but such a pattern is not evident. Regardless of the area or the season in which a sample was collected, all three reproduc- tive categories were always obtained. This sug- gests that golden king crab have an asynchronous reproductive cycle lacking distinct seasonal vari- ation. Table 3.— Percentage of adult females in each of three categories of reproductive condition: 1) uneyed embryos, 2) eyed embryos, and 3) empty egg cases, and total sample size (W) by subarea and quarter. South Central North Quarter 1 2 3 N 1 2 3 N 1 2 3 N 1 55 33 12 2 3 4 28 67 5 67 50 28 14 384 61 8 42 12 63 9 1,307 36 50 1,399 16 23 859 78 3 19 224 The apparent lack of seasonality conflicts with previous studies of golden king crab reproduction. Hiramoto and Sato (1970) reported that egg extru- sion occurs from July to October and embryo hatch- ing occurs from February to July along central Japan. However, Hiramoto and Sato found embryos in late stages of development throughout the year, 6D. Armstrong, College of Fisheries, University of Washington, Seattle, WA 98195, pers. commun. 1984. 581 FISHERY BULLETIN: VOL. 84, NO. 3 indicating that embryo hatching was probably occur- ring at times other than the peak season. Rodin (1970) reported that egg extrusion occurs from August to September based on the relatively high incidence of recently molted females with new em- bryos. However, Rodin based this on only one sum- mer sample and some of our samples, especially those from the northern area, if examined alone would have also incorrectly led to the same conclu- sion. Our findings, however, are consistent with those for other deep water crabs (L. couesi, Somer- ton 1981b; Geryon quinquedens, Haefner 1978) which have asynchronous or protracted spawning. Asynchronous spawning is also consistent with two of our other observations. First, the larvae of golden king crab, due to their large size and pre- sumably deep habitat, should be relatively insen- sitive to seasonal changes in primary production. Second, adult males and females of golden king crab appear to segregate by depth and this segregation appears to be maintained throughout the year. Such year-round sexual segregation is unlikely for a seasonally reproducing species; however, it is con- sistent with an asynchronous reproducing species if only the reproductively active individuals migrate between depth zones. IMPLICATIONS FOR FISHERY MANAGEMENT Two of our findings, the latitudinal decrease in the size at maturity and the asynchronous reproductive cycle, pertain to regulations used to manage the golden king crab fisheries in Alaska. Commercial harvest of king crabs is restricted to males larger than a minimum legal size (maximum carapace width including spines) which is specified for each species in each management area. These minimum sizes are set at the average size of a male three years after reaching maturity based on the ra- tionale that such a size would preserve sufficient males for breeding even when the exploitation rate is high (North Pacific Fishery Management Coun- cil 1981). Thus, to establish a minimum size limit that conforms to this rule, both an estimate of the size at maturity and an estimate of male growth rate are needed. Unfortunately, we lack sufficient data to estimate the growth rates of golden king crab in any of the three management areas considered here and therefore cannot determine appropriate minimum size limits. However, our estimates of male size at maturity can be used to judge, in a qualitative sense, the adequacy of the current minimum size limits. These size limits and the estimated sizes at matur- ity, expressed in terms of carapace length, are as follows: Minimum size limit Size at maturity (mm CL) (mm CL) Northern area 123 92 Central area 123 107 Southern area 134 130 The current minimum size limits decrease with increasing latitude, but not in proportion to the esti- mated sizes at maturity. Based solely on the relative magnitude of our estimates, we believe that the cur- rent minimum size limit in the southern area, and perhaps in the central area as well, is too low. However, we believe that the prolonged or year- round breeding of golden king crab would allow males more opportunities for mating than would be possible with a short breeding season; therefore, relative to seasonally breeding king crabs, fewer males would be sufficient for the breeding needs of the population. If this is true, then minimum size limits based on the criteria established for red and blue king crabs may be unnecessarily conservative for golden king crab. Commercial harvest of king crabs is also restricted to a legal fishing season specified for each species in each area. Although economic or logistic factors are considered when fishing seasons are established, of primary importance is the timing of the breeding and molting seasons. During the breeding season, females molt while aggregated together with the males (Powell et al. 1973); and if fishing were per- mitted at this time, not only would females be caught in greater numbers, they would also be injured by the fishing gear. During and soon after the male molting season, the recovery rate (ratio of recover- able meat to total body weight) is low; and if fishing were permitted at this time, the value of the crabs would also be low. Since the breeding seasons tend to occur in the late winter and early spring and the male molting seasons occur in late spring, the fish- ing seasons usually begin in the fall. For golden king crab, however, there is no clear seasonality in breed- ing; and adult males and females appear to be spatially segregated throughout the year. Although we lack sufficient data to determine if there is any seasonality in male molting, it appears that there is no compelling biological reason to restrict the golden king crab fisheries to any particular time of the year. Therefore, we believe that, at present, fishing seasons should be determined primarily by what is most convenient or beneficial to fishermen and processors. 582 SOMERTON and OTTO: DISTRIBUTION AND REPRODUCTION OF GOLDEN KING CRAB ACKNOWLEDGMENTS We thank Peter Cummiskey, Steven Meyers, and Kenneth Cronk (NMFS, Kodiak) for providing in- valuable assistance in both field and laboratory data collection; David Stanchfield (FV Morning Star), Joe Wabey (FV American Eagle), Scott Bowlden (FV Valiant), Edward Compton (FV Valiant), John Atwell (RV Miller Freeman), and Edward Gelb (RV Miller Freeman) for their cooperation and exper- tise in the operation of the research vessels; Kevin Bailey, James Balsiger, Nicholas Bax, Robert Fran- cis, and Nancy Pola for providing helpful reviews of the manuscript. In addition, we express our gratitude to Dennis Peterson, Barry Collier and the North Pacific Fishing Vessel Owner's Association for arranging and supporting the charter of the FV American Eagle. LITERATURE CITED Alaska Department of Fish and Game (ADFG). 1983. Commercial shellfish regulatons. Alas. Dep. Fish Game, Juneau, AK, 111 p. Bartlett, M. S. 1947. The use of transformations. Biometrics 3:39-52. Bell, G. 1980. The costs of reproduction and their consequences. Am. Nat. 116:45-76. Benedict, J. E. 1895. Scientific results of explorations by the U.S. Fish Com- mission Steamship Albatross No. XXXI: Descriptions of new genera and species of crabs of the family Lithodidae with notes on the young of Lithodes camtschatica and Lithodes brevipes. Proc. U.S. Natl. Mus. 17:479-488. Butler, T. H., and J. F. L. Hart. 1962. The occurrence of the king crab Paralithodes cam- tschatica (Tilesius), and of Lithodes aequispina Benedict in British Columbia. J. Fish. Res. Board Can. 19:401-408. Efron, B., and G. Gong. 1983. A leisurely look at the bootstrap, the jackknife and cross validation. Am. Stat. 37:36-48. Guzman, L. M., and I. G. Campodonico. 1972. Fecundidad de la centolla Lithodes antarctica Jacquinot (Crustacea, Decapoda, Anomura: Lithodidae). Ans. Inst. Pat. (Punta Arenas, Chile) 3:249-258. Haefner, P. A., Jr. 1978. Seasonal aspects of the biology, distribution and relative abundance of the deep-sea red crab, Geryon quin- quedens Smith, in the vicinity of the Norfolk Canyon, western North Atlantic. Proc. Natl. Shellfish. Assoc. 68: 49-62. Haynes, E. B. 1968. Relation of fecundity and egg length to carapace length in the king crab, Paralithodes camtschatica. Proc. Natl. Shellfish. Assoc. 58:60-62. 1981. Description of larvae of the golden king crab, Lithodes aequispina, reared in the laboratory. Fish. Bull., U.S. 80: 305-313. HlRAMOTO, K., AND S. SATO. 1970. Biological and fisheries survey on an anomuran crab, Lithodes aequispina, off Boso Peninsula and Sagami Bay, central Japan. Jpn. J. Ecol. 20:165-170. Hoffman, E. G. 1968. Description of laboratory-reared larvae of Paralithodes platypus (Decapoda, Anomura, Lithodidae). J. Fish. Res. Board Can. 25:439-455. Kinder, T. H., and J. D. Schumacher. 1981. Circulation over the continental shelf of the South- eastern Bering Sea. In D. W. Hood and J. A. Calder (editors), The eastern Bering Sea shelf: Oceanography and resources, Vol. 1. p. 53-75. U.S. Gov. Print. Off., Wash., DC. Makarov, V. V. 1938. Rakoobraznye, Anomura (Crustacea, Anomura). [In Russ.] Akad. Nauk SSSR, Zool. Inst. Fauna SSSR 10(3), N.S. 16, 325 p. Izd. Akad. Nauk SSSR, Moscow-Leningrad. [Transl. By Israel Program Sci. Transl., 1962, 283 p., avail- able from U.S. Dep. Commer., Natl. Tech. Inf. Serv., Spring- field, VA, as OTS 60-21822.] McNair, D. 1983. The new king crabbing: How much harder can this fishery get? Pac. Fish. 4:50-53. Miller, R. J. 1976. North American crab fisheries: regulations and their rationales. Fish. Bull., U.S. 74:623-633. Nelson, R., Jr., R. French, and J. Wall. 1981. Sampling by U.S. observers on foreign fishing vessels in the eastern Bering Sea and Aleutian Island region, 1977- 1978. Mar. Fish. Rev. 43(5):1-19. North Pacific Fishery Management Council. 1981. Bering Sea/ Aleutian Islands king crab fishery manage- ment plan. North Pacific Fishery Management Council, Anchorage, AK, 35 p. Pereyra, W. T. 1966. The bathymetric and seasonal distribution, and repro- duction of adult tanner crabs, Chionoecetes tanneri Rathbun (Brachyura: Majidae), off the northern Oregon coast. Deep- Sea Res. 13:1185-1205. 1968. Distribution of juvenile tanner crabs, Chionoecetes tan- neri Rathbun, life history model and fisheries management. Proc. Natl. Shellfish. Assoc. 58:66-70. Pianka, E. R., and W. S. Parker. 1975. Age-specific reproductive tactics. Am. Nat. 109: 453-464. Powell, G. C, B. Shafford, and M. Jones. 1973. Reproductive biology of young adult king crabs Para- lithodes camtschatica (Tilesius) at Kodiak, Alaska. Proc. Natl. Shellfish. Assoc. 63:77-87. Rodin, V. E. 1970. Novye dannye o ravnoshipom krabe (New data on the golden king crab). [In Russ.] Rybn. Khoz. 46:11-13. [Transl. by S. Pearson, 1982, 7 p.; available from Northwest and Alaska Fisheries Center, National Marine Fisheries Ser- vice, NOAA, 2725 Montlake Blvd. E., Seattle, WA 98112.] Sasakawa, Y. 1975. Studies on blue king crab resources in the western Ber- ing Sea III ovarian weights, egg numbers carried and diam- eters. Bull. Jpn. Soc. Sci. Fish. 41:941-944. Sato, S., and S. Tanaka. 1949. Study on the larval stage of Paralithodes camtschatica (Tilesius) I. Morphological research. [In Jpn.] Hokkaido Fish. Exp. Stn. Res. Rep. 1:7-24. [Transl. by I. M. Nakatsu, U.S. Fish Wildl. Serv., Seattle Biol. Lab., 1968, 24 p.; available from Northwest and Alaska Fisheries Center, Na- tional Marine Fisheries Service, NOAA, 2725 Montlake Blvd. E., Seattle, WA 98115.] 583 FISHERY BULLETIN: VOL. 84, NO. 3 SLIZKIN, A. G. 1974. Osobennosti raspredeleniya krabov (Crustacea, Deca- poda, Lithodidae et Majidae) v beringovom more (Specifics of the distribution of crabs (Crustacea, Decapoda, Lithodidae and Majidae) in the Bering Sea). Bonitet mirovogo okeana (Estimated productivity of the world ocean), part 5. [In Russ.] Tr. Vses. Nauchno-Issled. Inst. Morsk. Rybn. Khoz. Okeanogr. 99:28-37. [Transl. by A. L. Peabody, 11 p., available from Northwest and Alaska Fisheries Center, Na- tional Marine Fisheries Service, NOAA, 2725 Montlake Blvd. E., Seattle, WA 98112.] SOMERTON, D. A. 1980. A computer technique for estimating the size of sex- ual maturity in crabs. Can. J. Fish. Aquat. Sci. 37:1488- 1494. 1981a. Regional variation in the size of maturity of two species of tanner crab (Chionoecetes bairdi and C. opilio) in the eastern Bering Sea, and its use in defining management subareas. Can. J. Fish. Aquat. Sci. 38:163-174. 1981b. Contribution to the life history of the deep-sea king crab, Lithodes couesi, in the Gulf of Alaska. Fish. Bull., U.S. 79:259-269. SOMERTON, D. A., AND R. A. MACINTOSH. 1983. The size at sexual maturity of blue king crab, Para- lithodes platypus, in Alaska. Fish. Bull., U.S. 81:621-628. 1985. Reproductive biology of the female blue king crab, Paralithodes platypus near the Pribilof Islands, Alaska. J. Crustacean Biol. 5:365-376. Suzuki, Y., and T. Sawada. 1978. Notes on an anomuran crab, Lithodes aequispina Benedict, in Suruga Bay. Bull. Shizuoka Prefect. Fish. Exp. Stn. 12:1-10. Wallace, M., C. J. Pertuit, and A. R. Hvatum. 1949. Contribution to the biology of the king crab, Para- lithodes camtschatica. U.S. Fish Wildl. Serv., Fish. Leafl. 340, 50. p. Wear, R. G. 1974. Incubation in British decapod Crustacea, and the effects of temperature on the rate and success of embryonic devel- opment. J. Mar. Biol. Assoc. U.K. 54:745-762. Weber, D. G. 1967. Growth of the immature king crab, Paralithodes cam- tschatica (Tilesius). Int. North. Pac. Fish. Comm., Bull. 21:21-53. 584 A MODEL OF THE DRIFT OF NORTHERN ANCHOVY, ENGRAULIS MORDAX, LARVAE IN THE CALIFORNIA CURRENT James H. Power1 ABSTRACT The drift of northern anchovy, Engraulis mordax, larvae in the California Current to unfavorable offshore areas may be an important factor contributing to larval mortality, and hence it may affect recruitment and subsequent adult population size. A simulation model based on a finite-difference approximation to the advection-diffusion equation was developed to aid in the study of larval anchovy drift. Model com- ponents included the long-term mean geostrophic and wind-driven current velocities to 50 m depth, and turbulent diffusion. The model predicted larval distributions in the Southern California Bight and off- shore regions after 30 days of drift, and these distributions were used to assess the extent of cross-shore and alongshore larval transport that occurs when spawning takes place at different locations, seasons, and during times of increased offshore-directed Ekman transport. Offshore transport was minimal in most simulations. Simulations of drift starting from the location of peak spawning showed strongest seasonal effects, with currents during the season of peak northern anchovy spawning (March) resulting in reduced offshore dispersal when compared with currents at other times of the year. March currents also produced the greatest downshore (southeasterly) transport of larvae, and strong seasonal currents, such as the nearshore, northwesterly flowing California Counter- current, can greatly affect the alongshore 30-day larval distributions. Offshore directed Ekman trans- port, associated with upwelling, does not strongly affect the drift of larvae in the nearshore region, but large increases in overall Ekman transport, or extension of spawning into offshore regions, can result in significant seaward transport of larvae out of the Southern California Bight. The total population of northern anchovy, Engraulis mordax, a common pelagic fish off the west coast of North America, is comprised of three subpopula- tions (Vrooman et al. 1981): northern (found north of lat. 36°30'N); central (between lat. 29° and 38°N); and southern (south of lat. 29 °N). The central sub- population inhabits the Southern California Bight region, and in recent times has exhibited substan- tial changes in population biomass (e.g., Smith 1972). Analysis of northern anchovy scales deposited in sediments indicates that large northern anchovy population fluctuations have also occurred in the past few centuries (Soutar and Isaacs 1974). Histor- ically the central subpopulation of northern anchovy has supported a significant fishery (Messersmith and Associates 1969; Sunada 1975; Stauffer and Charter 1982), and although the U.S. fishery has recently declined, there is still a significant Mexican fishery. The northern anchovy fishery, the recent and histor- ical changes in anchovy population size, and the fish's important role in the marine ecosystem all pro- vide the motivation for studying the mechanisms 'Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038; present address: Coastal Fisheries Institute, Center for Wetland Resources, Louisiana State University, Baton Rouge, LA 70803-7503. that may cause interannual variations in northern anchovy stock size. Such changes in stock size may be a consequence of variations in the previous spawning stock size, or they can also arise as a result of interannual dif- ferences in mortality during prerecruit life history stages (Rothschild et al. 1982). Because the egg and larval stages have the highest mortalities, it seems possible that processes affecting the relative mor- tality during these stages can have a significant ef- fect on subsequent recruitment. Two major causes of larval mortality are starvation and predation (Smith and Lasker 1978; Hunter 1981). A factor that may contribute to these is larval drift. The northern anchovy eggs and larvae, lacking adequate motil- ity, can be involuntarily transported away from nearshore spawning areas. It is the nearshore regions in the Southern California Bight that most frequently contain adequate food concentrations for growth and survival of first feeding northern an- chovy larvae (Lasker 1978, 1981). Although eddies and other short-term mesoscale features are important in the Southern California Bight (Mooers and Robinson 1984; Simpson et al. 1984), the broad and relatively slow equatorward flow termed the California Current is the dominant feature in the region that persists on evolutionary Manuscript accepted September 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 585 FISHERY BULLETIN: VOL. 84, NO. 3 time scales. Hence, it seems plausible that northern anchovy spawning strategies have developed in response to the relatively predictable seasonal and spatial trends in the California Current. Possible relationships between time and location of fish spawning and the currents off the west coast of North America have been discussed by Parrish et al. (1981). They noted that in the Southern Califor- nia Bight the Ekman (wind-driven) currents are generally diminished relative to other areas along the coast. This reduced offshore transport is favor- able for the retention of fish eggs and larvae. However, some weak offshore directed Ekman transport is consistently present in the Southern California Bight year round (Nelson 1977; Parrish et al. 1981; Bakun and Parrish 1982). Smith (1972) analyzed historical records of north- ern anchovy larval distribution in the Southern California Bight and found that samples taken farther offshore had a higher proportion of older lar- vae than that of samples taken nearshore. Assum- ing a uniform spatial and temporal distribution of spawning, this result implied that a significant frac- tion of northern anchovy eggs and larvae were transported offshore after nearshore spawning. Bailey (1981) found that the average distance off- shore of Pacific hake, Merluccius productus, larvae north of Point Conception was positively correlated with offshore Ekman transport and that the magni- tude of subsequent Pacific hake recruitment was negatively correlated with offshore transport. Hewitt and Methot (1982) compared the distribu- tions of northern anchovy larvae sampled in 1978 and 1979 and found that the bulk of the larvae in 1979 were farther offshore than those in 1978 and that mortality of 0-group northern anchovy was greater in 1979 when compared with those spawned in 1978. The year 1979 was one of enhanced upwell- ing and colder temperatures (both concomitants of offshore Ekman transport) relative to 1978. The studies cited above suggest drift may play an important role in larval ecology, but the conclusions drawn from plankton sampling must be viewed with caution. Inferences drawn from field collections about the drift of larvae usually carry the assump- tion that both northern anchovy spawning and lar- val mortality were uniform in space and time, because the time and distance scales involved largely preclude synoptic sampling of eggs and larvae throughout the region. Hence, only correlative ex- planations for the observed distribution can be made, and other causal factors affecting the larval distribution may be hidden. For example, an obser- vation of greater proportions of older larvae in off- shore waters could also result from earlier spawn- ing or greater early mortality (possibly coupled with increased spawning activity) in those waters, and not drift. Additionally, the mesoscale variability present in the Southern California Bight and the considerable patchiness of early and late larvae (due to northern anchovy schooling behavior; Hewitt 1980, 1981a) further confound the conclusions drawn from plankton samples and diminish the value of interannual comparisons. Therefore, as an alter- native to field studies, a simulation model of north- ern anchovy drift in the California Current was developed to help evaluate the role of drift in larval ecology. The objective was to use the model to deter- mine the effect of differences in northern anchovy spawning location and time on the subsequent larval distribution and to evaluate the effects on larval distribution when offshore Ekman transport is in- creased above its normal mean value. METHODS The drift simulation was based on the two-dimen- sional (x,y) form of the advection-diffusion equation: dF d dt dx U-K,?l\+ 3 vF dx dy where F = the concentration of eggs and larvae; u and v = current velocities in the respective x and y directions; and Kx and Ky = eddy diffusivity coefficients for the x and y directions. An analytical solution to this equation cannot be evaluated relative to northern anchovy larval drift in the California Current, although a numerical ap- proximation that specifies larval concentration as a function of location and time can be determined. This was accomplished by approximating each of the derivatives in the equation by weighted finite- differences, so that the model was algebraically formulated as the current and diffusivity-mediated fluxes of larvae among geographic points in the Southern California Bight. Apart from the assump- tion that larvae continually maintained themselves in surface waters, the northern anchovy were assumed to be conservative and completely passive drifters, i.e., no mortality or movement due to lar- val swimming was incorporated into the model. Details of the numerical methods used are presented in Power (1984). The geographic grid for the model was defined 586 POWER: MODEL OF NORTHERN ANCHOVY DRIFT using the California Cooperative Oceanic Fisheries Investigation (CalCOFI) coordinate system. The CalCOFI grid is a regular coordinate system of cross-shore "lines" and alongshore "stations". The model and CalCOFI grids are oriented with respect to the coast so that increasing station number corresponds to increasing offshore distance and in- creasing line number implies the downshore (south- easterly) direction. Each line unit increment is spaced 12 nm apart, and each station unit represents 4 nm. The grid for the model was defined to form cells that were 37 km (20 nm) on a side, and the fluxes of larvae were among grid ceil centers. Model coverage was from CalCOFI lines 70 to 120, and ex- tended offshore to CalCOFI station 120 (Fig. 1). Unless northern anchovy utilize a strategy where- by spawning is initiated in response to the presence of an eddy or other short-term mesoscale features, it can reasonably be expected that spawning time and location have evolved partially in response to predictable current features. For this reason, seasonal currents based on interannual means were used in the model. Northern anchovy spawning behavior relative to eddies, etc., is presently un- known, and there are persistent seasonal trends in spawning, e.g., northern anchovy spawn through- out the year, but March is typically the peak time of spawning (Smith 1972; Methot 1981). Geostrophic currents for the model were cal- culated using the geopotential anomalies computed by Lynn et al. (1982). Lynn et al. used CalCOFI data collected between 1950 and 1978 to compute the average geopotential anomaly relative to 500 m for four seasonal periods (nominally January, April, July, and October) at 175 locations in the Califor- nia Current. Average geopotential anomalies were computed for an additional 23 locations for this study to augment the Lynn et al. (1982) coverage 125 120 115 35- 125 120 115 Figure 1.— Geographic grid used in the northern anchovy drift simulations, and the corresponding CalCOFI line and station coordinate system. Lettered locations are starting points for simulation presented in this paper. Mean geopotential anomolies (used for computing geostrophic currents) were calculated for the locations indicated by dots (Lynn et al. 1982; this study). 587 FISHERY BULLETIN: VOL. 84, NO. 3 (Fig. 1). The top 50 m of the water column is the predominant depth range of anchovy larvae (Ahlstrom 1959), therefore mean anomaly values for the surface and for 10, 20, 30, and 50 m of depth were used in this study. The anomalies at each standard depth were interpolated to model grid nodes using the bivariate interpolation algorithm of Akima (1978). Geostrophic current velocities normal to each grid cell interface were computed for each of the standard depths, and average geostrophic cur- rent velocities were then calculated for a layer ex- tending from the surface to 50 m. Wind speed and direction data used in this study were from the data base summarized and discussed by Nelson (1977). The raw wind observations were converted to surface wind stress (t) values using the relation where p Cd w T = RCdW2 air density (1.22 kg m drag coefficient; and wind speed. 3); The drag coefficient was computed as a function of wind speed using the empirical relation of Amorocho and DeVries (1980, 1981). The computed wind stress vectors were partitioned by month of observation and resolved into alongshore and cross-shore com- ponents. A monthly mean wind stress component for each model grid cell interface was then computed by averaging the appropriate component of the stress vectors in the 37 km by 37 km area bisected by the grid cell interface. Total Ekman or wind- driven transport in the direction 90° to the right of the wind can be approximated by dividing the wind stress by the Coriolis parameter (Neumann and Pier- son 1965), and this calculation was performed for the mean wind stress components. The mixed layer depth in the California Current is seldom >50 m, and is often <20 m in the Southern California Bight during the summer (Husby and Nelson 1982). It was assumed that Ekman transport occurring deeper than 50 m was negligible, and the Ekman transport values were converted to a mean wind-driven velocity for the surface to 50 m layer by dividing the transport by the 50 m layer thickness. The final current velocities were calculated as the vector sum of the seasonal geostrophic and appro- priate monthly Ekman components. Vector addition of the two components appears to be a reasonable assumption (Parrish et al. 1981), and no compensa- tion for redistribution of mass owing to sustained winds was performed. The final seasonal current fields for the simulations were January, March (April geostrophic velocities plus March Ekman velocities), July, and October currents. Figure 2 illustrates the general trends in the California Current for the January and March seasons. This figure should be interpreted with cau- tion. Apart from the large potential differences between actual synoptic conditions and the average pattern used in the simulations, the resultant vec- tor for a cell was necessarily computed for Figure 2 by averaging the current components of oppos- ing cell faces and then calculating the resultant. A distortion is introduced wherever components on op- posite faces of a cell differ in magnitude or sign, so that Figure 2 best represents features of the Califor- nia Current that are consistent over several model grid cells. The California Current is evident as two regions of intensified southeasterly flow at the left margins and midlines of the plots. During all parts of the year except spring, the current turns toward shore at the southern end of the Southern Califor- nia Bight. A northwesterly flow near the coast sub- sequently forms the inshore portion of a large cyclonic eddy (the Southern California Eddy; Owen 1980) that occupies most of the Southern Califor- nia Bight. During most of the year part of this eddy's northeasterly flow continues past Point Conception, to form the California Countercurrent (Hickey 1979; Fig. 2, January plot). In the spring the southeast- erly flow of the California Current moves closer to shore to obliterate the surface portion of the Countercurrent (Fig. 2, March). Tsuchiya (1980, fig. 2) gives a clear picture of the seasonal inshore- offshore movements of the California Current at CalCOFI lines 90 and 93. Close to shore in the south- ern half of the modeled region there is another region of intensified southeasterly flow, most evi- dent in the March current plot. Lynn et al. (1981) provided detailed illustrations of the geostrophic flow regimes used in the simulations, and Nelson (1977) presented graphical representations of the wind stress fields along the west coast of North America. Hickey (1979) presented a comprehensive review of seasonal and spatial variations of the California Current and the possible driving mech- anisms involved, and Owen (1980) reviewed the in- cidence and ecological consequences of eddies in the California Current system. Two additional current fields were calculated in order to assess the effects of increased offshore directed Ekman transport on larval northern an- chovy distribution. As mentioned earlier, the mean wind stress is consistently directed downshore during March in the modeled region, a condition 588 POWER: MODEL OF NORTHERN ANCHOVY DRIFT MARCH Figure 2.— Resultant mean current vectors for the normal January and the March seasonal current data used in this study. See text for cautions concerning figure interpretation. Length of arrow indicating north direction corresponds to a current velocity of 10 cm/s. producing offshore directed Ekman transport. Two current fields were obtained by increasing the cross- shore component of the mean March Ekman velocities by the factors 1.5 and 3.0, and then com- bining the April seasonal geostrophic and aug- mented March Ekman velocities. Wind stress, and hence transport, is proportional to the square of wind speed. This means that roughly a 22% increase in a downshore wind speed increases the corres- ponding offshore directed Ekman transport by the factor 1.5. A threefold increase in offshore Ekman transport results from about a 77% increase in the downshore wind speed. Bakun and Nelson (1976) presented extensive statistical analyses of an "up welling index" (defined as the offshore directed component of Ekman transport) for the location lat. 33°N, long. 119°W (this point is very close to loca- tion A used in the simulations; see below). Over an annual cycle the mean upwelling index for this loca- tion changes by at least a factor of two, with a rapid increase in both mean and standard deviation dur- ing the spring. The March mean index at this point was about 50 t/s per 100 m of coastline with a stan- dard deviation of roughly 80, hence upwelling at this particular time and location can be highly variable. Further, Bakun and Nelson (1976) found that en- hanced or diminished upwelling persists on a seasonal time scale, so incorporation in the model of prolonged increased Ekman transport was not unrealistic. Diffusion was incorporated into the model solely to parameterize subgrid scale mixing; including larger scale and more ephemeral mixing processes would obscure the broad seasonal trends the model was intended to illustrate. The eddy diffusivity parameter was computed using scale-dependent dif- fusion formulae of Okubo (1976) and a regression analysis of diffusion data presented by Okubo (1971). The finite-difference representation of diffusion re- quired the use of a pseudo-Fickian diffusivity coeffi- cient, so the mean scale-dependent diffusivity for the 37 km grid spacing (Kx = Ky = 101 m2/s) was 589 FISHERY BULLETIN: VOL. 84, NO. 3 used for all locations and all times in the model. The numerical method incorporated diffusion as the weighting factor coth [(uh)/(2K)] for the flux at each grid cell interface, where u is the current velocity at the interface and h is the 37 m grid spacing (see Power 1984 for further details). Hence, diffusion becomes important in regions of low current velocity, and at higher velocities diffusion is less im- portant and advection dominates the flux. For the current velocities in most of the modeled region, the above hyperbolic cotangent function is usually evaluated to a magnitude near unity, making the contribution of turbulent transport to larval drift minimal relative to advective (current velocity) transport. Simulations were carried out by starting an ini- tial point source of northern anchovy eggs or larvae at various locations historically known to be larval anchovy habitat (Hewitt 1980). Examples of simula- tions for four starting locations (Table 1; Fig. 1), which are representative of the overall patterns pro- duced by the simulations, are presented here. The four locations will be referred to in the text by their letter designations indicated in Figure 1 and Table 1. Northern anchovy larvae begin to school at about 27 d (Hunter and Coyne 1982); therefore larval distributions after 30 d of drift are presented. Thirty-d-old larvae are also rapidly increasing their "patchiness" (Hewitt 1981a), indicating that they could then exert significant control over their posi- tion. The time step in the simulations was 1 d. Results from a simulation using the actual northern anchovy egg distribution found in 1982 as the ini- tial condition can be found in MacCall (1983). Table 1.— Geographic and CalCOFI coordinates of start- ing locations for simulations presented in this paper. Letter designation corresponds to the same locations in Figure 1 . Starting location Coordinates CalCOFI Lat. N Long. W Line Station A B C D 33°08.4' 32°54.1' 31°59.3' 32°14.1' 118°51.4' 117°47.3' 118°05.4' 119°09.2' 89.17 92.5 95.83 92.5 42.5 32.5 42.5 52.5 Northern anchovy larval concentrations in the contour plots are relative to starting concentration; the unitless contour value of 10~2 represents a lar- val concentration two orders of magnitude below the starting concentration, and only concentrations down to 10 "7 are illustrated. Larvae were per- mitted to be advected out the borders of the modeled area, except for the border along the coast. Grid cells bordering the Santa Barbara Channel (at about lat. 34°N, long. 120°W; Fig. 1) between the Channel Islands and Point Conception were open, and lar- vae advected into this region were considered to be lost from the system. Larvae were not permitted to be transported across any of the islands in the modeled region. Because March is the peak spawn- ing time of northern anchovy, the effects of different starting locations on the 30-d larval distributions during March conditions will be presented first. The effects of spawning in different seasons and en- hanced offshore Ekman transport during March will then be presented for comparison. The simulation results nominally represent larval northern anchovy distributions, but the results also apply to any plank- tonic species that begin drift at the same locations and maintain themselves in the top 50 m of the California Current. The overall extent of onshore-offshore and along- shore transport was of major interest in this study. A convenient way of summarizing the simulated larval distributions relative to their cross-shore distribution was to sum all larval concentrations in the cells having the same CalCOFI station coordin- ates. These sums were converted to percentages of the total number of larvae at 30 d, and the cumula- tive percentage of larvae present as one progressed offshore was plotted versus CalCOFI station coor- dinates. A similar procedures using CalCOFI line coordinates was done to summarize alongshore transport. RESULTS Effects of Starting Location, Normal March Currents Northern anchovy larvae that began drift at loca- tion B, near the coast, were transported downshore by March currents (Fig. 3B). This was an effect of the nearshore southeasterly current (Fig. 2), and because of this flow only 15% of the larvae were at or upshore of the starting location after 30 d of drift Figure 3.— Distribution of northern anchovy larvae after 30 d of drift in March currents. Letter designation corresponds to a simula- tion with northern anchovy begun at the corresponding lettered location in Figure 1 and Table 1; starting location is marked in this and subsequent contour plots with asterisks. Locations A and C share the same CalCOFI station coordinate; points B and D have the same CalCOFI line coordinate. Concentration contour inter- vals are proportions of the starting concentration, decreasing in order of magnitude steps. Tic marks around perimeter are at whole degrees of latitude and longitude; dots are at intervals of 3.33 CalCOFI line units from lines 70 to 120 and intervals of 10 station units offshore to station 120 (i.e., every 74 km). 590 POWER: MODEL OF NORTHERN ANCHOVY DRIFT 591 FISHERY BULLETIN: VOL. 84, NO. 3 (Fig. 4). The alongshore distribution of larvae below the starting point was quite uniform, and the lower larval concentrations had reached the southern border of the modeled region (CalCOFI line 120). Dispersal offshore was minimal, and a majority of the larvae lay in a band near the coast with about equal proportions inshore and offshore of the start- ing point; 92% of the larvae were on or inshore of CalCOFI station 37.5. After initial southeasterly transport, some larvae were transported in an off- shore, southwesterly direction. Extensive downshore transport also occurred to northern anchovy larvae begun at location C, and in fact only 3% of the larvae remained at or upshore of the starting location after 30 d of drift (Figs. 3C, 4). The larvae begun at point C were also concen- trated in a narrow band along the coast, but unlike those started at point B most of the larvae begun at C moved inshore of the starting location after 30 d of drift. Northern anchovy larvae begun at the offshore location D showed much less extensive downshore transport than those begun at B or C (Figs. 3D, 4). Only 10% of the larvae remained at or upshore of the starting point, but 86% of the total remained at or between CalCOFI line 92.5 (location C's line coordinate) and line 102.5, a span of 222 km. Most larvae were inshore of location D, and the cross- shore distribution was slightly more uniform than those begun farther inshore. Starting point D's distance from the coastline permitted the slightly broader cross-shore distribution. Larvae begun at location A showed an alongshore cumulative percentage distribution after 30 d of drift which was similar to that of larvae that begin drift at point D, although it was displaced farther upshore (Fig. 4). Location A produced the greatest percent- age of larvae remaining at or upshore of the start- ing location, and there is a small patch of high (10 _1) larval concentrations present at the starting location (Fig. 3A). This reduced dispersal of larvae begun at A also produced the strongest cross-shore gradient of larvae. A majority of the larvae were again on or inshore of the starting location after 30 d of drift. In summary, the distributions of northern anchovy larvae that began drift at locations A through D and that were produced by March currents were formed as relatively strong cross-shore gradients, so that the 30-d distributions were bands (ca. 100 km wide) parallel to the coast. The results of starting larvae at locations A, C, and D were that more than 85% of the larvae were inshore of the starting location after 30 d of drift. Larvae that began drift at loca- STATION 120 105 90 75 60 45 30 100- 120 115 110 105 100 95 90 85 80 75 70 LINE Figure 4.— Cumulative percentages of northern anchovy larvae after 30 d of drift, progressing offshore (increasing CalCOFI sta- tion number) and downshore (increasing CalCOFI line number), for the four starting locations under March current conditions. Cross symbols are at the starting location's corresponding CalCOFI line or station coordinate. Distance between tic marks on the abscissae is equivalent to a distance of 111 km. Note that a steep curve implies a compact distribution of larvae, while more gradual slopes imply more widely dispersed larvae. tions B and C were extensively carried downshore of the starting location. Most of the larvae that started at points A and D also moved downshore from those locations, but the bulk of the larvae were not as widely dispersed from the starting location as those begun at points B and C. Effects of Seasonal Current Fields on Larval Distribution The distributions of northern anchovy larvae started at the same location but using different seasonal current regimes appear very different to the eye (Figs. 3, 5-7). Part of this effect is real, but part is also due to displacement of the contours for the lower larval concentrations (e.g., 10~7), which represent few larvae. The cumulative percentage plots (Fig. 8) indicate that, when summarized on a model-wide basis, the overall cross-shore distribu- tions of larvae begun at locations B, C, and D were not greatly different when currents from the four seasonal periods were used in the simulations. A fixed distance offshore there were some large differ- ences in the cumulative percentages among seasons 592 POWER: MODEL OF NORTHERN ANCHOVY DRIFT "7 . V I Figure 5.— Distribution of northern anchovy larvae after 30 d of drift in normal July currents. 593 FISHERY BULLETIN: VOL. 84, NO. 3 Figure 6.— Distribution of northern anchovy larvae after 30 d of drift in normal October currents. 594 POWER: MODEL OF NORTHERN ANCHOVY DRIFT Figure 7.— Distribution of northern anchovy larvae after 30 d of drift in normal January currents. 595 FISHERY BULLETIN: VOL. 84, NO. 3 STATION 120 105 90 75 60 45 30 100 STATION 120 105 90 75 60 45 30 120 115 110 105 100 95 90 85 80 75 70 LINE STATION 120 105 90 75 60 45 30 100- '120 115 110 105 100 95 90 85 80 75 70 LINE STATION 120 105 90 75 60 45 30 100- ~120 115 110 105 100 95 90 85 80 75 70 LINE "120 115 110 105 100 95 90 85 80 75 70 LINE Figure 8.— Cumulative percentage plots of northern anchovy larval concentrations after drift in the four seasonal current regimes. Letter designation corresponds to starting locations indicated in Figure 1 and Table 1. for larvae begun at the same location, but a com- parable percentage was usually present a short distance away, i.e., most curves in Figure 8 are closely spaced on the CalCOFI station abscissae. Most larvae begun at the offshore location D moved inshore regardless of season, and the seasonal dif- ferences were in the relative extent of inshore move- ment, the maximum occurring during July. In all simulations the cross-shore distributions of larvae formed strong gradients, regardless of season. Starting location A is within the Southern Califor- nia Bight proper, the region that most consistently has high larval concentrations of northern anchovy (cf. Hewitt 1980). Larval distributions started at location A did exhibit notable seasonal differences in their 30-d cross-shore distributions, with the greatest offshore dispersal occurring during July (Figs. 5A, 8A), and the largest inshore movement 596 POWER: MODEL OF NORTHERN ANCHOVY DRIFT occurring during March current conditions (Figs. 3A, 8A). January and October were intermediate between these two extremes. In all cases the cross- shore gradients of larvae were strong. In contrast, the alongshore distributions of larvae differed markedly when the simulations were done with currents from the four seasonal periods. Vir- tually all larvae were carried upshore of starting location A by the California Countercurrent in the January simulation (Figs. 7A, 8A), but when the model was run using March currents, a majority of the larvae were downshore of point A after 30 d of drift (Fig. 3A). The July and October simulation results for point A seemed to indicate an annual pro- gression between the March and January extremes (Fig. 8A). The seasonal differences in the overall alongshore distributions were even more dramatic for northern anchovy larvae begun at locations B and C. The uniform downshore distribution produced by March currents differed from the distributions formed in all other seasons. Larvae begun at location B were all transported upshore of the starting location dur- ing October current conditions (Fig. 6B). When January currents were used (Fig. 7B), the upshore movement had lessened, so that only 62% of the larvae were at or upshore of location B, and the lar- vae were more evenly distributed along the coast (Fig. 8B). March currents yielded the greatest down- shore movement, and the July distribution (Fig. 5B) was intermediate between that produced by March and October conditions, with the alongshore gradient of larvae again steepening. The changes on an annual basis between upshore, then down- shore transport were similar for larvae begun at location C, except that the July current conditions produced the greatest upshore transport (Figs. 5C, 8C); March again produced the maximum down- shore transport for larvae begun at point C (Fig. 3C). Larvae begun at location C formed a relative- ly compact distribution after 30 d of drift in the January currents (Fig. 7C). The overall alongshore distributions of northern anchovy larvae that started drift at point D ap- peared to be least influenced by seasonal changes in the currents, although March conditions again produced the greatest transport downshore of the starting point (Fig. 3D), with July currents again yielding the greatest upshore transport (Fig. 5D). January currents also produced a very compact distribution of larvae started at location D, similar to that of larvae begun at point C. In summary, only northern anchovy larvae begun at location A appeared to have notable differences in their model-wide, cross-shore distributions after 30 d of drift. Larvae begun at all four locations did have substantial seasonal differences in their along- shore distributions, with March currents consistent- ly producing the greatest downshore dispersal. The least downshore dispersal occurred during January, October, July, and July current conditions for larvae started at locations A, B, C, and D respectively. January currents generally seemed to produce the most compact 30-d distributions of larvae (least dispersal). Effects of Increased Offshore Ekman Transport, March Currents Increasing the March cross-shore Ekman trans- port by a factor of 1.5 had little effect on the 30-d distributions of northern anchovy larvae begun at locations B and C (Fig. 9); these curves are also closely spaced on the CalCOFI station abscissae. In- creasing the average or "normal" offshore Ekman component by a factor of three produced more noticeable changes in the cross-shore distributions of larvae begun at points B and C, but this effect was not substantial; the contours representing the lower concentrations extended far offshore (Fig. 10B, C), but the higher concentration contours, which delimit the majority of the larvae, were not greatly displaced from those of normal March cur- rents (Fig. 3B, C). This is also evident in the cum- ulative percentage, curves. In comparison, northern anchovy larvae begun at locations A and D underwent about the same in- crease in offshore dispersal with a 1.5 x offshore directed Ekman component increase as those begun at points B and C did with the 3 x offshore Ekman increase (Fig. 9). When the offshore directed Ekman transport was increased to three times its normal mean value, the effects on larvae begun at points A and D were substantial. A majority of the larvae were carried offshore of starting locations A and D, and a large fraction were transported a significant distance (Fig. 10 A, D), well seaward of the South- ern California Bight. The increase in offshore Ekman transport also noticeably affected the along- shore distributions of larvae begun at locations A and D (Fig. 9). The overall pattern of alongshore distribution is similar to that produced by the nor- mal mean conditions, but the larvae were general- ly farther downshore. DISCUSSION Models have inherent assumptions and simplifica- 597 FISHERY BULLETIN: VOL. 84, NO. 3 STATION 120 105 90 75 60 45 30 _1 1 I I I L STATION 120 105 90 75 60 45 30 100- 120 115 110 105 100 95 90 85 80 75 70 LINE 120 115 110 105 100 95 90 85 80 75 70 LINE STATION 120 105 90 75 60 45 30 STATION 120 105 90 75 60 45 30 100" 120 115 110 105 100 95 90 85 80 75 70 LINE 120 115 110 105 100 95 90 85 80 75 70 LINE Figure 9.— Cumulative percentage plots of northern anchovy larval concentrations after 30 d drift in the three March Ekman current regimes ("normal" or long-term mean, 1.5 x offshore directed Ekman transport, 3x offshore directed Ekman). tions, and thus approximate what occurs in nature. The geostrophic current information, while some of the best available, nonetheless constrained the spatiotemporal resolution of this model, and incor- porating Ekman transport required several assump- tions. Further, there can be considerable interannual variability in the modeled region (Mooers and Robin- son 1984; Simpson et al. 1984), and presumably the model is of "average" conditions and cannot be representative of any specific year; I felt that in- cluding such variability (assuming adequate data were available) would complicate the results without significantly contributing to biological insight. An Figure 10.— Distribution of northern anchovy larvae after 30 d of drift in March currents with three times the normal offshore directed Ekman transport. 598 POWER: MODEL OF NORTHERN ANCHOVY DRIFT 599 FISHERY BULLETIN: VOL. 84, NO. 3 additional confounding factor is the lack of biological information in the model; consistent larval behavior patterns (e.g., diurnal vertical migrations) and spatially heterogeneous mortality could produce distributional patterns differing from those pre- sented here. In spite of these caveats, the simula- tions do demonstrate that variations in northern anchovy spawning location and time, and changes in the magnitude of offshore directed Ekman trans- port, can have significant consequences for the subsequent larval distribution. By inference, these changes in distribution can result in increased or reduced larval mortality, and ultimately affect adult northern anchovy population size. Offshore transport was not significant in the sim- ulations done with the unaugmented or "normal" March (seasonal April geostrophic + March Ekman) currents. A majority of the northern anchovy lar- vae that began drift at the four starting locations were inshore of their starting points after 30 d of drift, and the cross-shore distributions indicated that most larvae occupied a relatively narrow range of distances close to shore. Mais (1974) and Methot (1981) reported that most juvenile northern anchovy occupy inshore areas in the fall, and the model in- dicates that this inshore movement could be facil- itated by passive drift. As mentioned earlier, the consensus is that nearshore regions provide more hospitable food conditions for the northern anchovy larvae. Lasker (1978, 1981) summarized the results of surveys of larval food distributions in the South- ern California Bight. His figures indicate that suit- able larval food concentrations decline rapidly as one progresses offshore. O'Connell (1980) reported the results of a survey for starving northern anchovy larvae in the Southern California Bight, the degree of starvation being defined by histological criteria. He found apparently healthy larvae at locations as far as about 250 km offshore (at lat. 32°30'N, long. 120°W), where model concentrations were <10~7 after 30 d in all March simulations except for lar- vae begun at A, where they were <10~4. Despite the good condition of these offshore larvae, the simulation results indicate a low likelihood of their being recruited to the nearshore juvenile population. The low offshore larval concentrations will also hinder the development of schooling (Hewitt 1981a). The minimal offshore transport situation found in the March current simulations was also generally true when simulations were done using currents from other seasons, except for northern anchovy lar- vae begun at location A. This point is the most in- terior starting location within the Southern Califor- nia Bight proper and is primary northern anchovy spawning habitat (Hewitt 1980) and where seasonal changes in the currents are especially important (Tsuchiya 1980). Spring is a time when currents in the Southern California Bight are not as well organized as other times of the year, and the Southern California Eddy is often absent (Hickey 1979; Owen 1980). It is interesting that currents during March, the peak spawning period, produced the least offshore transport of larvae begun at loca- tion A when compared with other seasons, even though March is the time of greatest overall Ekman transport (Bakun and Nelson 1976). There is signifi- cant spawning in January (Methot 1981), and the January simulations also had reduced dispersal of larvae. The model results support the hypothesis of Parrish et al. (1981) that northern anchovy spawn- ing in the Southern California Bight do so at a time and place that minimizes offshore transport of eggs and larvae. It is clear that the overall 30-d alongshore distribu- tions of northern anchovy larvae produced by normal March currents depended largely on the spawning location's proximity to the well-defined southeasterly current present near the coast in the southern half of the modeled region. Larvae that started drift near this current underwent extensive downshore transport. Larvae begun farther into the Southern California Bight (location A), and farther offshore (location D), were also transported down- shore, but to a much lesser extent. This again con- firms the role of the Southern California Bight as an area where minimal transport of spawning products takes place. The southwesterly, offshore transport that occurred in many of the simulations at the southern margin of the modeled region (between CalCOFI lines 110 and 120) is consistent with the evidence that this region forms a faunal boundary between species of the Southern Califor- nia Bight and those of Baja California to the south, and that this faunal boundary is created by current patterns (Hewitt 1981b). This is also a region of in- creased surface convergence (Parrish et al. 1981). The extent of alongshore transport was marked- ly different for northern anchovy larvae begun at the same starting location when currents from the different seasons were used. Depending on start- ing location, seasonal changes in currents could pro- duce almost complete reversals between predom- inantly upshore or downshore transport. March currents consistently produced the greatest down- shore transport. These effects were due to the pres- ence or absence of the Southern California Eddy and the Southern California Countercurrent. Because the Southern California Countercurrent is present 600 POWER: MODEL OF NORTHERN ANCHOVY DRIFT year-round, except during peak spawning in the spring, it is clear that the relationship between the time of northern anchovy spawning and the time that this countercurrent diminishes is critical. The simulations indicated that eggs and larvae from early spawning (i.e., January) are carried upshore into the Santa Barbara Channel and north of Point Conception, while those from later spawning (March) move in the opposite, southeasterly direc- tion. The sizes and birth dates of juveniles collected in the fall of 1978 and 1979 were in accordance with this pattern. Methot (1981) reported that juvenile northern anchovy collected during both fall seasons in the northern portion of the Southern California Bight had birth dates (as determined from daily growth increments in otoliths) in the preceding months of December and January, and these fish were generally larger than those collected farther to the south. The northern anchovy collected in the south had predominantly February and March birth dates. It may be that the northern group, contain- ing fish from early spawning, were advected to the north by the Southern California Countercurrent and that the southern group of fish from late spawn- ing were produced when the surface countercurrent had diminished. Future studies of the transport and distribution of northern anchovy larvae or other planktonic species in the Southern California Bight should incorporate as much information as is avail- able on the presence and magnitude of the South- ern California Countercurrent and the Southern California Eddy. Nearshore winds in the Southern California Bight are relatively weak, and downshore wind speeds generally increase farther offshore (Bakun and Nelson 1976; Nelson 1977; Dorman 1982). The implication, in terms of offshore transport, is that larvae closest to shore are affected least by offshore transport, while those farther offshore experience a much greater impact. Thus the areal extent of northern anchovy spawning interacts with offshore Ekman transport; in years when most northern an- chovy spawn close to shore there will be decreased offshore transport, because of weak inshore winds, than in years when northern anchovy spawn farther offshore. The impact on the products of offshore spawning will depend on the magnitude of the winds in the offshore areas in each particular year. North- ern anchovy larvae that began drift farthest north in the Southern California Bight (location A) and at the more offshore location (D) were most affected by increases in offshore directed Ekman transport, indicating southerly and inshore spawning are best for reduced dispersal in March. Hewitt and Methot (1982) stated that the area of northern anchovy spawning was more compact and more northerly in 1978 than in 1979. Survival of young larvae was about the same in both years, indicating that early mortality from starvation and predation was not substantially different in the two years. Survival through the juvenile stage was greater in 1978 than in 1979, however, and Hewitt and Methot (1982) cited increased offshore transport in 1979 as a possi- ble reason. Superimposed on the effects of spawning location is the interaction between the increase in downshore wind speeds (offshore directed Ekman transport) as one progresses offshore and the magnitude of inter- annual variations in the wind speeds. In the simula- tions the effects of the 3 x increase in Ekman trans- port were substantially greater than those of the 1.5 x increase. The 1.5 x change was not a great enough increase to carry many northern anchovy larvae into offshore regions of higher, offshore directed Ekman transport. The inshore 3 x increase carried a greater fraction of larvae farther offshore, and the 3x increase in the offshore region subse- quently operated on a greater proportion of the larval population. Thus there was an interaction between enhanced offshore directed Ekman trans- port in the nearshore area and increased Ekman transport farther offshore, the two of these acting together to produce the extensive drift evident in the simulation results. Years in which downshore winds increase in only the inshore or the offshore regions would not produce as much overall offshore dispersal. Bakun and Nelson's (1976) statistical analyses of the "upwelling index" indicates that pro- longed increased Ekman transport is feasible, although the 3 x condition would probably be a par- ticularly bad year. It should also be noted that Ekman transport was incorporated into the model as acting uniformly on the 50 m surface layer, and presumably the model depicts the drift of "average" larvae. Larvae that remain near the surface or at 50 m would undergo greater or lesser transport, respectively. Alternatively, it is known that winds in the Southern California Bight have a strong diur- nal periodicity (Bakun and Nelson 1976; Dorman 1982), and a diurnal vertical migration coupled with diurnal changes in the winds could significantly alter larval drift. In summary, the simulation results indicated that seaward dispersal of northern anchovy larvae is generally small, but that seasonal effects are strong- est in the area of peak spawning (location A) and that March spawning at this point minimizes off- shore dispersal. Spawning at locations or times near 601 FISHERY BULLETIN: VOL. 84, NO. 3 well-defined currents, such as the California Countercurrent, can produce major changes in larval distribution, and consequently may affect lar- val survival. The effect of offshore directed Ekman transport on the larval population depends on the areal extent of northern anchovy spawning, and the spatial distribution of any changes in wind stress and subsequent Ekman transport; an increase in Ekman transport in both the inshore and offshore regions will act together to produce maximum off- shore dispersal. ACKNOWLEDGMENTS This work was done while the author held a Na- tional Research Council Research Associateship. I thank Reuben Lasker and John Hunter for their hospitality and advice, as well as for the opportu- nity to conduct this research. Larry Eber and Craig Nelson were instrumental in providing dynamic height and wind speed data, without which this work could not have been done. I also thank Andy Bakun, Roger Hewitt, Ron Lynn, Alec MacCall, Rick Methot, Bob Owen, Dick Parrish, and Paul Smith for their stimulating discussions, advice, and review of this work. This manuscript was revised while the author held a CIMAS (Cooperative Institute for Marine and Atmospheric Sciences) postdoctoral fellowship at the University of Miami, and this sup- port is gratefully acknowledged. LITERATURE CITED Ahlstrom, E. H. 1959. Vertical distribution of pelagic fish eggs and larvae off California and Baja California. Fish. Bull., U.S. 60:107-146. Akima, H. 1978. A method of bivariate interpolation and smooth sur- face fitting for irregularly distributed data points. ACM Trans. Math. Software 4:148-159. Amorocho, J., and J. J. DeVries. 1980. A new evaluation of the wind stress coefficient over water surfaces. J. Geophys. Res. 85C:433-442. 1981. Reply to comment on 'A new evaluation of the wind stress coefficient over water surfaces'. J. Geophys. Res. 86C:4308. Bailey, K. M. 1981. Larval transport and recruitment of Pacific hake Merluccius productus. Mar. Ecol. Prog. Ser. 6:1-9. Bakun, A., and C. S. Nelson. 1976. Climatology of upwelling related processes off Baja California. Calif. Coop. Oceanic Fish. Invest. Rep. 19:107- 127. Bakun, A., and R. H. Parrish. 1982. Turbulence, transport, and pelagic fish in the Califor- nia and Peru Current systems. Calif. Coop. Oceanic Fish. Invest. Rep. 23:99-112. Dorman, C. E. 1982. Winds between San Diego and San Clemente Islands. J. Geophys. Res. 87C:9636-9646. Hewitt, R. 1980. Distributional atlas of fish larvae in the California Cur- rent region: northern anchovy, Engraulis mordax Girard, 1966 through 1979. Calif. Coop. Oceanic Fish. Invest. Atlas 28, 101 p. 1981a. The value of pattern in the distribution of young fish. Rapp. P.-v. Reun. Cons. int. Explor. Mer 178:229-236. 1981b. Eddies and speciation in the California Current. Calif. Coop. Oceanic Fish. Invest. Rep. 22:96-98. Hewitt, R. P., and R. D. Methot, Jr. 1982. Distribution and mortality of northern anchovy larvae in 1978 and 1979. Calif. Coop. Oceanic Fish. Invest. Rep. 23:226-245. HlCKEY, B. M. 1979. The California Current system - hypotheses and facts. Prog. Oceanogr. 8:191-279. Hunter, J. R. 1981. Feeding ecology and predation of marine fish larvae. In R. Lasker (editor), Marine fish larvae, p. 33-77. Univ. Wash. Press., Seattle-Lond. Hunter, J. R., and K. M. Coyne. 1982. The onset of schooling in northern anchovy larvae, Engraulis mordax. Calif. Coop. Oceanic Fish. Invest. Rep. 23:246-251. Husby, D. M., and C. S. Nelson. 1982. Turbulence and vertical stability in the California Cur- rent. Calif. Coop. Oceanic Fish. Invest. Rep. 23:113-129. Lasker, R. 1978. The relation between oceanographic conditions and lar- val anchovy food in the California Current: identification of factors contributing to recruitment failure. Rapp. P.-v. Reun. Cons. int. Explor. Mer 173:212-230. 1981. Factors contributing to variable recruitment of the northern anchovy (Engraulis mordax) in the California Cur- rent: Contrasting years, 1975 through 1978. Rapp. P.-v. Reun. Cons. int. Explor. Mer 178:375-388. Lynn, R. J., K. A. Bliss, and L. E. Eber. 1982. Vertical and horizontal distributions of seasonal mean temperature, salinity, sigma-t, stability, dynamic height, oxy- gen, and oxygen saturation in the California Current, 1950- 1978. Calif. Coop. Oceanic Fish. Invest. Atlas 30, 513 p. MacCall, A. D. 1983. Population models of habitat selection, with application to the northern anchovy. Ph.D. Thesis, Univ. California at San Diego, San Diego, 170 p. Mais, K. F. 1974. Pelagic fish surveys in the California Current. Calif. Dep. Fish Game Fish Bull. 162, 79 p. MESSERSMITH, J. D., AND ASSOCIATES. 1969. The northern anchovy (Engraulis mordax) and its fishery, 1965-1968. Calif. Dep. Fish Game Fish Bull. 147, 102 p. Methot, R. Jr. 1981. Growth rates and age distributions of larval and juvenile northern anchovy, Engraulis mordax, with infer- ences on larval survival. Ph.D. Thesis, Univ. California at San Diego, San Diego, 203 p. Mooers, C. N. K., and A. R. Robinson. 1 984 . Turbulent jets and eddies in the California Current and inferred cross-shore transports. Science 223:51-53. Nelson, C. S. 1977. Wind stress and wind stress curl over the California Current. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 602 POWER: MODEL OF NORTHERN ANCHOVY DRIFT SSRF-714, 87 p. Neumann, G., and W. J. Pierson, Jr. 1966. Principles of physical oceanography. Prentice-Hall, Englewood Cliffs, NJ, 545 p. O'Connell, C. P. 1980. Percentage of starving northern anchovy, Engraulis mordax, larvae in the sea as estimated by histological methods. Fish. Bull., U.S. 78:475-489. Okubo, A. 1971. Oceanic diffusion diagrams. Deep-Sea Res. 18:789- 802. 1976. Remarks on the use of 'diffusion diagrams' in model- ing scale-dependent diffusion. Deep-Sea Res. 23:1213-1214. Owen, R. W. 1980. Eddies of the California Current system: physical and ecological characteristics. In D. M. Power (editor), The California Islands: Proceedings of a Multidisciplinary Sym- posium, p. 237-263. Santa Barbara Mus. Nat. History. Parrish, R. H., C. S. Nelson, and A. Bakun. 1981. Transport mechanisms and reproductive success of fishes in the California Current. Biol. Oceanogr. 1:175-203. Power, J. H. 1984. A numerical method for simulating plankton transport. Ecol. Model. 23:53-66. Rothschild, B. J., E. D. Houde, and R. Lasker. 1982. Causes of fish stock fluctuations: Problem setting and perspectives. In B. J. Rothschild and C. Rooth (editors), Fish ecology III, p. 39-86. Univ. Miami Tech. Rep. Simpson, J. J., C. J. Koblinsky, L. R. Haury, and T. D. Dickey. 1984. An offshore eddy in the California Current system: preface. Prog. Oceanogr. 13:1-4. Smith, P. E. 1972. The increase in spawning biomass of northern anchovy, Engraulis mordax. Fish. Bull., U.S. 70:849-874. Smith, P. E., and R. Lasker. 1978. Position of larval fish in an ecosystem. Rapp. P. -v. Reun. Cons. int. Explor. Mer 173:77-84. SOUTAR, A., AND J. D. ISAACS. 1974. Abundance of pelagic fish during the 19th and 20th cen- turies as recorded in anaerobic sediments off the Califor- nias. Fish. Bull., U.S. 72:257-273. Stauffer, G. D., and R. L. Charter. 1982. The northern anchovy spawning biomass for the 1981- 82 California fishing season. Calif. Coop. Oceanic Fish. In- vest. Rep. 23:15-19. Sunada, J. S. 1975. Age and length composition of northern anchovies, Engraulis mordax, in the 1972-73 season California anchovy reduction fishery. Calif. Fish Game 61:133-143. Tsuchiya, M. 1980. Inshore circulation in the Southern California Bight, 1974-1977. Deep-Sea Res. 27A:99-118. Vrooman, A. M., P. A. Paloma, and J. R. Zweifel. 1981. Electrophoretic, morphometric, and meristic studies of subpopulations of northern anchovy (Engraulis mordax). Calif. Fish Game 67:39-51. 603 PARASITES OF BENTHIC AMPHIPODS: DINOFLAGELLATES (DUBOSCQUODINIDA: SYNDINIDAE) Phyllis T. Johnson1 ABSTRACT During a 2V2-yr survey, 13 species of benthic amphipods collected from the continental shelf of the north- eastern United States were found infected by dinoflagellates. Prevalences ranged from <1% to 67%, depending on amphipod species, time, and place of collection. The parasites are assigned to the order Duboscquodinida, family Syndinidae, based on similar life histories and a similar kind of mitosis ("mitose syndinienne"). Two types of organisms were involved, both apparently more closely related to Hematodinium Chatton and Poisson than to other described syndinids. Morphology and development of the parasites and host-parasite interactions are discussed. A cytochemical method used to determine presence or absence of basic nuclear proteins was strongly positive for basic proteins in spores and prespores but negative in most other stages. A few spores in four infections possessed a distinct flagellum, but in the absence of living material, shape of spores and whether they were biflagellate could not be determined. With three possible exceptions in the group of 303 infections studied, the syndinids were not recognized as foreign by their hosts, and in joint infections of syndinids and fungi, only the fungi were being attacked by host hemocytes. High prevalences in certain of the amphipod species suggest that the syndinids might be population regulators in these species. This paper is one of three that describe and discuss the more common parasites found in populations of benthic amphipods of the continental shelf of the northeastern United States. The other papers con- cern microsporidans and ciliates (Johnson 1985, 1986). Because my observations on the parasites dis- cussed in this paper were based on examination of histological sections, I could not determine whether spores were typical "dinospores". However, agree- ment with other developmental stages of well- studied species of syndinids from copepods and an amphipod, and the nuclear type, indicates that the parasites of benthic amphipods are related to species currently placed in the Syndinidae, order Dubosc- quodinida (sensu Chatton 1952 and Cachon 1964). Previously described syndinids occur intracellular- ly in radiolarians and in copepod and shrimp eggs (Chatton 1952; Stickney 1978) and extracellu- larly in the hemocoel of copepods, an amphipod, and portunid and cancrid crabs (Chatton and Pois- son 1931; Chatton 1952; Manier et al. 1971; Newman and Johnson 1975; MacLean and Ruddell 1978). The relationship of the Duboscquodinida to free- living dinoflagellates is in doubt (Cachon 1964; Ris and Kubai 1974; Siebert and West 1974; Hollande 1975; Loeblich 1976; Herzog et al. 1984). Lacking a definitive consensus, the parasitic protists dis- cussed here are provisionally referred to the Dino- flagellata. The data presented and discussed in this paper show that species of syndinids are probably ubiqui- tous hemocoelic parasites of benthic and epibenthic amphipods, and may be population regulators in some species. METHODS The data are based on material collected during monitoring surveys carried out over a 2V2-yr period by the Northeast Fisheries Center, National Marine Fisheries Service. The 35 stations where benthic am- phipods were collected during the surveys are shown in Figure 1. Amphipods were sampled during 11 cruises, but not all stations were visited on each cruise, so that stations were sampled from 1 to 10 times each during the survey. A Smith-Mclntyre2 grab and occasionally an epibenthic sled or scallop dredge were used to obtain the samples. The 11 sta- tions indicated by solid circles on Figure 1 had the most consistent and numerous populations of am- phipods, and were sampled at least five times each. They yielded the majority of data presented here. Amphipods were preserved in 10% seawater Formalin. Up to 30 individuals of each species pres- 'Northeast Fisheries Center, National Marine Fisheries Service, NOAA, Oxford, MD 21654. 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Manuscript accepted October 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 605 FISHERY BULLETIN: VOL. 84, NO. 3 50 100 150 KILOMETERS Project on a'tei IWXjpi 1965 I 68° tf C NMFS Sano, ► Figure 1.— Benthic stations where gammaridean amphipods were sampled during the survey. ent in a sample, and sometimes more, depending on numbers present, were prepared for histological study. Details of collecting procedures and histo- logical preparation of the amphipods are given by Johnson (1985). Sections were cut at 6 ^m. Stain- ing methods included Harris' hematoxylin and eosin (H&E), the Feulgen reaction, and Alfert and Gesch- wind's (1953) fast-green method for demonstration of basic nuclear proteins. Harris' hematoxylin and eosin is specified because this combination stains nuclei of the parasites purple during certain stages. Other hematoxylin solutions, used with eosin, will not necessarily impart the same distinct purple color. Unless otherwise indicated, references to staining properties of the organisms are to H&E- stained specimens. OBSERVATIONS Thirteen amphipod species were infected with syn- 606 JOHNSON: PARASITES OF BENTHIC AMPHIPODS dinids (Table 1). The organisms occupied the hemocoel and morphologically were most like Hema- todinium perezi Chatton and Poisson, which was described from European portunid crabs. There were two distinct types, based on morphology and development. There is not enough information about the life history stages of Hematodinium to warrant assigning either or both types to that genus, and they are identified casually in this paper as "Type A A" and "Type AV" (Table 2). The Type AA forms were similar in all the amphipod species they in- fected, but there was variation in forms assigned to Type AV, and probably more than one species was involved. Host and Geographic Distribution Juvenile and mature amphipods of both sexes were attacked. Only Type AA was found in Am- pelisca agassizi (Judd), Byblis serrata Smith, and Table 1 .—Amphipod species infected with Type AA and Type AV parasites. Species of amphipod Type of parasite Prevalence positive stations (%) Prevalence all stations (%) Ampelisca agassizi (Judd) Byblis serrata Smith Harpinia propinqua Sars Ampelisca vadorum Mills Ampelisca verrilli Mills Casco bigelowi (Blake) Leptocheirus pinguis (Stimpson) Melita dentata (Kr0yer) s. lat. Monoculodes edwardsi Holmes Protohaustohus wigleyi Bousfield Phoxocephalus holbolli Krflyer Rhepoxynius epistomus (Shoemaker) Unciola species (probably all U. irrorata Say and U. inermis Shoemaker) AA 7 (101/1468) 4 (101/2403) AA 14 (24/170) 8 (24/316) AA1 18 (3/17) 3 (3/116) AV 41 (74/181) 17 (74/448) AV 18 (7/38) 15 (7/48) AV 67 (6/9) 10 (6/60) AV 4 (7/163) 0.8(7/913) AV 8 (1/12) 2 (1/44) AV 27 (25/93) 23 (25/110) AV 20 (1/5) 0.9 (1/110) AV 27 (10/37) 14 (10/73) AA and AV 20 (7/35) 3 (7/249) AA and AV 9 (37/404) 3 (37/1365) 1 Parasites in two of the infections may not be either Type AA or Type AV. Table 2. — Main characteristics of Type AA and Type AV. Stage Characteristic Type AA Type AV I Nuclear diameter <3 to >5 \im 2.5 to 3 ^m Nuclear color Blue or purple Purple Chromosomes Usually condensed Not condensed Plasmodia Present, small Present, small Single cells Common Absent or uncommon Cytoplasm Scanty Abundant, faintly fibrous IA Dense bodies Not present Present, <2 ^m in diameter Nuclear diameter — 2.5 to 4 film Nuclear color — Purple Plasmodia — Present, small Single cells — Present II Nuclear diameter 4 to 5.5 ^m 3 to 4 ^m Nuclear color Purple Purple Chromosomes Indistinct, partly Distinct, partly condensed condensed Plasmodia Uncommon, small Very rare Cytoplasm Vacuolate Homogeneous III Nuclear diameter (spore) 2.5 to 3 urn <2 urn Nuclear color (spore) Deep blue Deep blue Chromosomes (spore) Always condensed Always condensed Cytoplasm (spore) Scanty Scanty Plasmodia Absent Present Nuclear diameter — 3.5 fjm (Plasmodia) Nuclear color (Plasmodia) — Purple 607 FISHERY BULLETIN: VOL. 84, NO. 3 Harpinia propinqua Sars. Both Types AA and AV occurred in Rhepoxynius epistomus (Shoemaker) and Unciola species (U. irrorata Say and U. inermis Shoemaker), and only Type AV occurred in the re- maining species (Table 1). Both types of syndinids were present in Unciola species taken in a single sample at station 35, but individual specimens were parasitized by only one type. There are not enough data to indicate whether or not incidence varies by time of year in any of the amphipod species infected with these parasites. Infected amphipods were not found at the most northern and southern of the sta- tions, but these stations were sampled fewer times than most of the "positive" stations (i.e., stations where amphipods with syndinid infections occurred). There were 18 positive stations. Only Type AA was found at stations 23, 37, and 50. Only Type AV occurred at stations 33, 40, 56, and 62. Both types were represented at stations 20, 27, 35, 38, 47, 48, 49, 51, 57, 63, and 64. Whether one or both types occurred at a single station depended variously on which amphipod species were present, and on unknown factors. Two species of Ampelisca, A. vadorum Mills and A. agassizi, were common at inshore station 33. Prev- alence of Type AV in A. vadorum was 35% (56/158). However, Type AA did not occur at station 33 although a favored host, A. agassizi, was abundant there. In contrast, only Type AA was found at station 23, no doubt because of 2,811 amphi- pods collected there, only 23 were not A. agas- sizi. Development and Morphology All forms were similar in that extensive plasmodia were never present and chromosomes were con- densed in the interphase nuclei of the spores. There were four, possibly five, chromosomes. There was no metaphase plate. At telophase the apices of the two sets of chromosomes were touching (see Figure 3), and at all stages of mitosis the chromosomes of each group were juxtaposed basally (where they presumably were attached to the nuclear membrane) and spread out apically to varying degrees, like the spokes of a parasol (Figs. 2-4). These events are typical of "mitose syndinienne" (Chatton 1921). Syn- dinid chromosomes are V-shaped, so that each has two arms. In tissue sections the V shape was best seen in cells that had lysed, leaving only the chro- mosomes (Fig. 5). During telophase there were often only four (sometimes five?) visible arms of chro- mosomes in each daughter nucleus. If sectioning artifact was not responsible for the small number w *< m W M Figures 2-3.— Mitosis in Type AV parasites in Ampelisca vadorum (arrowheads). Interphase parasites of Figure 3 are stage II. Figure 4.— Mitosis in a Type AA parasite in Ampelisca agassizi (arrowhead). Chromosomes form a rosette in the interphase nucleus to the right (asterisk). Figure 5.— Chromosomes in a lysed Type AA parasite from Byblis serrata (arrowhead). The V shape of the chromosomes is evident. Figures 2-5, x 1500. of visible arms, the cells might have been haploid. Before spore formation, chromatin disposition in nuclei was variable, depending on the type of parasite and the stage of development. Resting nuclei with unfolded chromosomes were granular or vesicular, and sometimes rimmed with chromatin (see Figures 8, 17). In nuclei with partially unfolded chromosomes, clumps of chromatin often were ar- ranged so that they created a dashed or dotted line in the position that would be occupied by a complete- ly condensed chromosome (see Figure 9). When seen in a polar view, chromosomes or chromatin clumps formed rosettes (Figs. 4, 5). Morphology of the per- sistent chromosomes of spores was variable and will be described later. 608 JOHNSON: PARASITES OF BENTHIC AMPHIPODS Staining characteristics of nuclei differed depend- ing on the stage. Except for spores, prespores, and some cells in early Type AA infections, nuclei tended to be purple, not blue, with both chromatin and the matrix staining similarly in some cases. When nuclei at these stages were in mitosis, the chromosomes were little, if at all, bluer than chromatin in resting cells, although sometimes chromatin was more deep- ly staining in the dividing cells. Types of chromo- somes that stained with fast green by the Alfert and Geschwind method (indicating presence of basic pro- teins on the chromosomes) would stain blue in H&E preparations. Chromatin and chromosomes that did not stain with fast green in the Alfert and Gesch- wind method would stain purple with H&E. A comparison of Types AA and AV, by develop- mental stage, is given in Table 2. Infections con- sisting of few parasites were considered to be the earliest ones and are here designated stage I infec- tions. Stage II infections consisted of more numer- ous and generally larger organisms, and stage III infections consisted of prespores and spores that usually filled the hemocoel. Type AA Most Type AA infections were in Ampelisca agassizi (Table 1). Type AA chromosomes of all developmental stages were usually thicker than those of Type AV (compare Figures 2 and 4), and the organisms and their nuclei were larger (Table 2). Stage I organisms were scattered through the hemocoel, never numerous, and variable in mor- phology and staining characteristics. The one com- mon attribute was scanty and poorly staining cyto- plasm. Chromosomes were usually distinct. The most usual stage I infection consisted of scattered single cells and small plasmodia with nuclei that measured 3 to 4 /urn and had rather distinct chro- mosomes or chromatin clumps that stained a clear blue. Mitotic figures were not frequent, but were more common than in the other stages. A few cells in a late stage I (or very early stage II) infection probably were polyploid. They had many rather long, tangled chromosomes that sometimes formed partially separated groups within the nuclear area. The nuclei of these cells measured more than 7 ptm in the greater dimension. Stage II organisms were more numerous and dis- tinguished by having voluminous vacuolate or foamy cytoplasm (Fig. 6). Chromosomes and chromatin clumps were often obscured because the nuclear matrix stained almost as strong a purple as the chromatin. The nuclear matrix did not stain in the Feulgen reaction. Plasmodia were uncommon, always small, and sometimes consisted of short chains of joined cells. Mitosis was rarely seen in stage II and stage III, and probably was closely syn- chronized, which would reduce the probability of finding mitotic figures in fixed material. As the spore stage was approached, nuclei became smaller and bluer, and chromatin clumps and chromosomes gained clear outlines, because the matrix no longer stained. By the time of spore formation (stage III), organ- isms filled the hemocoel, and infected amphipods in H&E-stained sections could be distinguished with the naked eye because of their overall dark-blue color. Spore nuclei were spherical, and chromosomes were condensed but tightly packed and impossible to count (Fig. 7). In one infection, synchronized nuclear division had apparently just taken place, and daughter cells had not yet separated, so that bi- and 6 4 • i f 4P •I Figures 6-7.— Type AA parasites in Ampelisca agassizi. 6: Stage II. Nuclei do not have distinct clumps of chromatin and the cytoplasm is vacuolate. 7: Stage III (spores) (arrowhead). An unidentified fungus was also infecting the amphipod (asterisk). Figures 6-7, x 1500. 609 FISHERY BULLETIN: VOL. 84, NO. 3 quadrinucleate plasmodia were common. Cytoplasm of spores was scant. Sometimes spores were shaped like teardrops but generally they had amorphous outlines. A flagellum was visible on a few spores in an individual of Unciola species. 8 ** *" • I * Type AV This description is based on the organisms that infected Ampelisca vadorum. Stage I consisted of scarce and scattered small plasmodia, typically each with 2 to 10 nuclei. Their cytoplasm was faintly fibrous. Chromatin and the nuclear matrix were always purplish and nuclei were often rimmed with chromatin (Fig. 8). The nuclear matrix was not Feulgen positive, and chromatin did not stain strongly by this method. Slightly more advanced in- fections, with more parasites, had irregularly shaped single cells as well as plasmodia. The single organ- isms were often elongate, their nuclei were as above, and their cytoplasm was faintly stained. Stage I A, which I presume follows stage I, and which did not occur in Type AA, had moderate num- bers of small plasmodia and single cells. Chromatin patterns were rather distinct in most nuclei, par- ticularly in the larger ones. Chromatin stained pur- ple. Stage IA was distinguished by the presence of small, densely staining bodies. They were usually spherical but sometimes oval, and were usually sur- rounded by thin rims of cytoplasm. The bodies were associated with the plasmodia (Fig. 9) and also scat- tered through the hemocoel. They were intensely Feulgen positive and stained bright green by the Alfert and Geschwind method. The dense bodies were never extremely abundant and were present only in the company of many stage IA cells. Chromosomes of stage II cells were partially con- densed, and chromosomes and chromatin clumps were distinct because there was minimal staining in the nuclear matrix, unlike Type AA parasites in stage II. The cytoplasm was usually densely and homogeneously stained (Fig. 3). Cells were often very numerous and closely packed, but were not plasmodial. Occasionally there were a few dense bodies like those associated with stage IA. Occasional stage III infections were not as heavy as some stage II infections. There was apparently an abrupt transition from stage II cells to stage III prespores and spores. In one infection, a mass of spores with distinct deep-blue chromosomes oc- cupied a circumscribed area in the hemocoel, and larger single cells with condensed chromosomes that stained purple, and were probably very late stage II, occupied the remainder of the hemocoel (Fig. 10). % -- 1 € 4 10 1 J, t ,;;;.,f!pP ' ill - r .- * r *4* tt ' «A *%♦.* Figures 8-10.— Type AV parasites in Ampelisca vadorum. 8: Stage I. Several nuclei in the plasmodia are rimmed with chrom- atin. 9: Stage IA. Plasmodia with associated spherical dense bodies. Nuclei are pale and chromosomes are partially unfolded in some nuclei (arrowhead). 10: Late Stage II (larger, pale nuclei to the left— arrowhead) and Stage III (smaller, deeply staining nuclei to the right— open arrow). A demonstration of syn- chronized division of the parasite. Larger host nuclei are also present. Figures 8-9, x 1500; Figure 10, x 600. Presumably, the mass of spores resulted from syn- chronized but circumscribed division of a part of the population of the larger cell type. The roughly spherical nuclei of the spores in this infection were <2 /urn in diameter; nuclei of the larger cells were slightly >3 ^m in diameter. 610 JOHNSON: PARASITES OF BENTHIC AMPHIPODS Cells presumed to represent spores had either elongate or spherical nuclei (Figs. 11, 12). The two types did not occur together. Mitosis took place in very small cells, and possibly cells with spherical nuclei were prespores. They might also have been spores that had not yet acquired their final form, because cells of an intermediate shape also occurred. Chromosomes of the spherical nuclei were short; those of elongate nuclei were longer, somewhat more slender, and beaded. Because the cytoplasm was usually indistinct or invisible, outlines of spores were also indistinct. It is probable that spores often ruptured during fixation, resulting in loss of all cell components except the chromosomes, as shown in Figure 11. A probable polyploid cell was present in one early stage III infection, and there were small plasmodia in all stage III infections (as in Figure 17). Nuclei in plasmodia had purple-staining chromatin and did not stain by the Alfert and Geschwind method, unlike chromosomes of the spores. The relationship of the small plasmodia to spore formation was not obvious. Numbers of Type AV-infected individuals of species other than A. vadorum and M. edwardsi were small, and all stages of development were not usually represented. Stage IA infections, as well as Figures 11-12.— Type AV, Stage III (spores), in Ampelisca vadorum. 1 1 : Elongate spores. Note the beaded appearance of the chromosomes in one spore (arrowhead). 12: Spherical spores. Figures 11-12, x 1500. some or all the other stages, were seen in Ampelisca verrilli Mills, Leptocheirus pinguis (Stimpson), Casco bigelowi (Blake), and Unciola species. Stage I A infections of A. verrilli and C. bigelowi differed from those of A. vadorum because the small dense bodies were often irregularly shaped or composed of two or three contiguous particles rather than be- ing single and spherical or oval. In one of two stage III infections in L. pinguis, spores had almost spherical chromosomes (Fig. 13). In the other, chromosomes were indistinct because they were closely packed, but were longer than in the first in- fection and apparently beaded. All stages of infec- tion were represented in Unciola species. Spore nuclei were round or oval and a flagellum was visi- ble on a few spores in two infections. The final divi- sions were just taking place in one of these infec- tions, and many cells were still binucleate. Most of the single spores had rounded outlines, but spores with a visible flagellum were oval. Monoculodes edwardsi had the highest overall prevalence of Type AV (Table 1). The 25 infections encompassed all stages except Stage IA. There were polyploid cells in stage II infections. Their nuclei were sometimes over 6 ^m in diameter, often had chromatin separated into several areas (Fig. 14), and their chromosomes were seldom completely con- densed, except in mitotic cells. Polyploid cells in mitosis had at least three sets of chromosomes. Out- lines of both the interphase nucleus and the entire cell were often highly irregular. Plasmodia that presumably resulted from nuclear division of the polyploid cells often had nuclei of two or more sizes (Fig. 15), suggesting that all chromosome sets did not divide at the same time, or that the genetic material was not distributed equally at the time of division, so that a single Plasmodium might have contained haploid, diploid, and polyploid nuclei. Nuclei of Type AV spores in M. edwardsi were about 1 /um in diameter (Fig. 16). A single flagellum (not pictured) was visible on some spores in the infec- tion presented in Figure 16. As typical of Type AV, plasmodia were present in all stage III infections (Fig. 17). Host Response Reactions against the syndinid parasites were ex- tremely rare. One Type AV-infected specimen each oiMelita dentata (Kr0yer) s. lot. and Unciola species had scattered, melanized, amorphous nodules in the hemocoel, but the nodules could not be definitely associated with the syndinid infections. In one specimen of L. pinguis, hemocytes were associated 611 FISHERY BULLETIN: VOL. 84, NO. 3 « «• *. 14 15 # m I 16 17 '■-' # • *'-, •A. % Figures 13-17.— Type AV parasites. 13: Stage III (spores) in Leptocheirus pinguis. The chromosomes are spherical (arrow- head). 14: Stage II in Monoculodes edwardsi. Two of the para- sites are polyploid (arrowheads). Note separate groups of chromosomes or chromatin clumps in both these parasites. 15: Plasmodium resulting from nuclear division of a polyploid parasite in M. edwardsi. Note differently sized nuclei. 16: Stage III (spores) in M. edwardsi. There were flagellated spores in this infection. 17: Prespores, some dividing, in M. edwardsi. A Plasmodium, with rimmed nuclei, is also present (arrow- head). Figures 13-17, x 1500. with Type AV organisms, and karyorrhexis had oc- curred in unidentified cells in the area. With the possible exception of the Type AV infection in L. pinguis, the syndinids were not being attacked by hemocytes at the time of fixation. There was another sign that the syndinid parasites successfully evaded detection by their hosts. Two specimens of A. agassizi, both collected at station 47 but at different times, were infected jointly and heavily with Type AA and an unidentified fungus (Fig. 7). Of the more than 7,000 examined micro- scopically, these were the only two amphipods that had systemic fungal infections. Fungi were being phagocytized by hemocytes and fixed phagocytes, and other groups of fungi were being transformed into melanized nodules. (Probably the latter fungi had originally been phagocytized and killed by hemocytes that did not survive the process them- selves.) Although hemocytes and fixed phagocytes were actively destroying fungi, there was no indica- tion that the accompanying syndinids were recog- nized as foreign. Numbers of hemocytes apparently decreased dur- ing syndinid infection, but even in heavy infections some hemocytes remained and were still functional as shown by their ability to phagocytize the fungi discussed above. It is probable that the two suc- cessful fungal infections in syndinid-infected amphi- pods resulted in part from the fungi multiplying more rapidly than they could be phagocytized and degraded by the few remaining hemocytes and the fixed phagocytes associated with the heart. The syndinid parasites did not castrate their hosts. Whether death ensues from every infection with these parasites is not known. However, the general lack of discernible host response makes it unlikely that amphipods could successfully combat the parasites. DISCUSSION Like species of Syndinium described from cope- pods, Types AA and AV have a small number of chromosomes which are permanently condensed in spores and partially condensed in certain other stages; plasmodia (small and multiple in the case of Types AA and AV) are present during some devel- opmental stages; and spore formation takes place in the hemocoel of the host. However, species of Syndinium in copepods differ from Types AA and AV in that they develop from a Plasmodium that is first applied to the wall of the gut and then expands to fill the entire hemocoel. The massive Plasmodium then fragments to form individual dinospores. By 612 JOHNSON: PARASITES OF BENTHIC AMPHIPODS the time of sporulation, the host is castrated (Chat- ton 1910, 1920). Types AA and AV resemble Hema- todinium, not Syndinium, in that apparently none of these organisms develop from a primary Plas- modium associated with the gut, but instead they multiply from a few single cells and small plasmodia in the general hemocoel and never form a single massive Plasmodium. Further, these parasites do not castrate their hosts (Newman and Johnson 1975; MacLean and Ruddell 1978; P. T. Johnson, unpubl. data). Syndinium gammari, like Types AA and AV, is perhaps more closely related to Hematodinium than to Syndinium. Syndinium gammari was assigned to Syndinium by Manier et al. (1971) on the assump- tion that a massive Plasmodium was present dur- ing development. However, none of the infections studied by these authors had either a primary Plasmodium associated with the gut or a later and massive Plasmodium throughout the hemocoel. The first stage of S. gammari observed consisted of small irregular plasmodia up to 15 pm in diameter, which Manier and coworkers assumed resulted from the splitting-up of a large plasmodium. The small plasmodia then divided to form "diplococcal" forms, and these divided to give round, single organisms which transformed into spores measuring 7-8 /urn by 3-3.5 ^m. In the later stages of division, typical "dinomitosis" and "dinokaryons" were present. Considering the course of development in the ap- parently related parasites of benthic amphipods, Types AA and AV, it is possible that S. gammari does not have a primary plasmodium associated with the gut wall and does not develop an extensive Plasmodium in the hemocoel. If early stages of S. gammari consist of a few single cells or small plasmodia, these could have escaped notice because the parasites were observed after their removal from the host amphipod, either alive or in fixed and stained smears (Manier et al. 1971). Scattered organisms could more easily be missed by this tech- nique than by inspection of paraffin-embedded and sectioned whole amphipods. Chromosomes of Solenodinium globiforme and three species of Syndinium, all parasites of radio- larians, stain with fast green in the Alfert and Geschwind method for demonstration of basic nuclear proteins (Ris and Kubai 1974; Hollande 1975). Ris and Kubai remarked that chromosomes of the Syndinium species they studied also stained brightly in the Feulgen reaction. Although not definitely stated by the above authors, apparently chromosomes of all developmental stages of the above parasites stained equally with fast green. Chromosomes of these species tend to remain con- densed through the entire developmental cycle. On the other hand, Hollande (1975) found that trophont nuclei of the duboscquodinids Amoebophrya ceratii and Duboscquella melo do not stain by the Alfert and Geschwind method. He pointed out that chromo- somes are not condensed in the trophont nuclei of these forms and that he did not investigate stain- ing properties of the condensed chromosomes of spores. Hollande did find that a portion of the nucleolus of A. ceratii stains with fast green in the Alfert and Geschwind method. Like the syndinid parasites of radiolarians, chromosomes of Type AA and AV spores stain brightly in both Alfert and Geschwind's technique and the Feulgen reaction. However, Feulgen staining is less intense in stages I and II nuclei and these nuclei do not stain at all with fast green. Eukaryotes have a greater quantity of histone in rapidly dividing cells than in quiescent ones (DuPraw 1968; Wu et al. 1982), and nonhistone basic nuclear proteins— although scarce at all times— are much more abundant in log-phase than in stationary-phase cultures of the free-living dinoflagellates Gyro- dinium cohnii and Peridinium trochoideum (Rizzo and Nooden 1974). It would be interesting to deter- mine the relative amounts of basic nuclear proteins through the developmental cycle of syndinids and other duboscquodinids, and to determine whether basic proteins of the amphipod parasites increase when cells are dividing rapidly; and whether these proteins are masked by other substances (acidic pro- teins?) in stages where both chromatin and nuclear matrix stain purple with H&E and do not stain in the Alfert and Geschwind method. Probably fixation and paraffin embedment not only damaged flagella and were responsible for ap- parent lack of flagella on most spores of Types AA and AV, but also distorted spores of these parasites. Cachon (1964) cautioned that because spores of parasitic dinoflagellates become distorted or rup- tured both on fixation and when physical conditions are not proper, their shapes must be determined in living material. Origin and function of the small dense bodies pres- ent in Type AV, stage IA infections were not evi- dent. These bodies might represent necrotic nuclei like those seen in Syndinium infections (Jepps 1936-37), discarded chromatin resulting from reduc- tion divisions, or, perhaps, nuclei of microspores (Cachon 1964). Numbers of Gammarus locusta (Linn.) infected with Syndinium gammari in the Etang de Thau, France, varied from few to all members of a popula- 613 FISHERY BULLETIN: VOL. 84, NO. 3 tion (Manier et al. 1971). The infected amphipods these authors examined were apparently unaffected by the parasite. However, before one could deter- mine the mortality rate due to syndinid infection, it would be necessary to examine moribund and dead amphipods found in the field for presence of syn- dinids, as well as to follow progress of infection in the laboratory. Syndinids appear to be unaffected by host defense mechanisms. Spores of syndinids that parasitize the hemocoel must exit through breaks in the exoskeleton or gut. Because hemocytes are in short supply by time of sporulation and other host resources can be expected to be depleted, host defense mechanisms probably would not be suffi- cient to prevent death by infection with other micro- organisms that would enter through the breaks or death by leakage of body fluids. Assuming, on the basis of evidence presented in this paper, that amphipods are unable to contain syndinid infections and that most infections would therefore progress to the spore stage, syndinid infection could serve as a population regulator in heavily parasitized species. Monoculodes edwardsi and Ampelisca vadorum, which had overall prevalences of syndinid infection of 23% and 17% respectively, are examples of species that might be affected in this manner. ACKNOWLEDGMENTS Sara V. Otto, Maryland Department of Natural Resources, Oxford, MD, aided in translation of ar- ticles from the French, and the following person- nel from the Oxford and Sandy Hook Laboratories of the Northeast Fisheries Center helped as follows: Frank Steimle, David Radosh, Linda Dorigatti, Gretchen Roe, and Sharon MacLean collected the amphipods; Ann Frame and Linda Dorigatti aided in their identification; and Gretchen Roe, Dorothy Howard, Cecelia Smith, and Linda Dorigatti prepared the specimens for histological examination. My thanks to all of the above. LITERATURE CITED Alfert, M., and I. I. Geschwind. 1953. A selective staining method for the basic proteins of cell nuclei. Proc. Nat. Acad. Sci. USA 39:991-999. Cachon, J. 1964. Contribution a l'etude des Peridiniens parasites. Ann. Sci. Nat. Zool. Fr., Ser. 12, 6:1-158. Chatton, E. 1910. Sur l'existence de Dinoflagelles parasites coelomiques. Les Syndinium chez les Copepodes pelagiques. C. R. Acad. Sci. Paris, 151:654-656. Chatton, E. 1920. Les Peridiniens parasites. Morphologie, reproduction, ethologie. Arch. Zool. Exp. Gen. 59:1-475. Chatton, E. 1921. Sur un mecanisme cinetique nouveau: la mitose syndi- nienne chez les Peridiniens parasites plasmodiaux. C. R. Acad. Sci. Paris, Ser. D, 173:859-862. Chatton, E. 1952. Classe des Dinoflagelles ou Peridiniens. In P. P. Grasse (editor), Traite de Zoologie, Vol. 1, p. 309-390. Masson et Cie, Paris. Chatton, E., and R. Poisson. 1931. Sur l'existence, dans le sang des Crabes de Peridiniens parasites: Hematodinium perezi n.g., n. sp. (Syndinidae). C. R. Soc. Biol. 105:553-557. Du Praw, E. J. 1968. Cell and molecular biology. Academic Press, N.Y. Herzog, M., S. von Boletzky, and M.-O. Soyer. 1984. Ultrastructural and biochemical nuclear aspects of eukaryote classification: independent evolution of the dino- flagellates as a sister group of the actual eukaryotes? Origins Life 13:205-215. Hollande, A. 1975. Etude comparee de la mitose syndinienne et de celle des Peridiniens libres et des Hypermastigines infrastructure et cycle evolutif des Syndinides parasites de Radiolaires. Protistologica 10:413-451. Jepps, M. W. 1936-37. On the protozoan parasites of Calanus finmarchicus in the Clyde Sea area. Q. J. Microsc. Sci. 79:589-662. Johnson, P. T. 1985. Parasites of benthic amphipods: microsporidans of Ampelisca agassizi (Judd) and some other gammarideans. Fish. Bull., U.S. 83:497-505. 1986. Parasites of benthic amphipods: ciliates. Fish. Bull., U.S. 84:204-209. Loeblich, A. R., III. 1976. Dinoflagellate evolution: speculation and evidence. J. Protozool. 23:13-28. Maclean, S. A., and C. L. Ruddell. 1978. Three new crustacean hosts for the parasitic dino- flagellate Hematodinium perezi (Dinoflagellata: Syndinidae). J. Parasitol. 64:158-160. Manier, J.-F., A. Fize, and H. Grizel. 1971. Syndinium gammari n. sp. peridinien Duboscquodinida Syndinidae, parasite de Gammarus locusta (Lin.) Crustace Amphipode. Protistologica 7:213-219. Newman, M. W., and C. A. Johnson. 1975. A disease of blue crabs (Callinectes sapidus) caused by a parasitic dinoflagellate, Hematodinium sp. J. Parasitol. 61:554-557. Ris, H., and D. F. Kubai. 1974. An unusual mitotic mechanism in the parasitic proto- zoan Syndinium sp. J. Cell Biol. 60:702-720. Rizzo, P. J., AND L. D. Nooden. 1974. Isolation and partial characterization of dinoflagellate chromatin. Biochim. Biophys. Acta 349:402-414. Siebert, A. E., and J. A. West. 1974. The fine structure of the parasitic dinoflagellate Haplo- zoon axiothellae. Protoplasma 81:17-35. Stickney, A. P. 1978. A previously unreported peridinian parasite in the eggs of the northern shrimp, Pandalus borealis. J. Invertebr. Pathol. 32:212-215. Wu, R. S., S. Tsai, and W. M. Bonner. 1982. Patterns of histone variant synthesis can distinguish G0 from Gj cells. Cell 31:367-374. 614 FOOD HABITS AND DIET OVERLAP OF TWO CONGENERIC SPECIES, ATHERESTHES STOMIAS and ATHERESTHES EVERMANNI, IN THE EASTERN BERING SEA M. S. Yang1 and P. A. Livingston2 ABSTRACT Stomachs of 196 arrowtooth flounder, Atheresthes stomias, and 152 Kamchatka flounder, A. evermanni, collected from the same area of the eastern Bering Sea in summer 1983 were examined. Each species was divided into four fork-length groups: less than 201 mm, 201-300 mm, 301-400 mm, and greater than 400 mm. The principle diet of both species was comprised of walleye pollock, Theragra chalcogramma, shrimp (mostly Crangonidae), and euphausiids. Pollock was the most important prey item for both species in all four size groups, ranging from 56 to 86% and 66 to 88% of the total stomach content weight of Kamchatka flounder and arrowtooth flounder, respectively. Schoener's indices of diet overlap were calculated between the two species for each size group. The high value of the indices (ranging from 0.67 to 0.90) indicate that these two congeneric species basically consume the same resources. The genus Atheresthes of the family Pleuronectidae has two species: Kamchatka flounder, A. evermanni (Jordan and Starks), and arrowtooth flounder (Nor- man, 1934), A. stomias (Jordan and Gilbert). Ather- esthes evermanni is distributed from northern Japan (Hokkaido) through the Sea of Okhotsk to the western Bering Sea north to Anadyr Gulf (Willimov- sky et al. 1967). Atheresthes stomias is distributed from Central California to the eastern Bering Sea. In the Bering Sea, it meets about on a line with Saint Matthew Island, overlaps with, and is replaced by A. evermanni (Hart 1973). Because the morphological differences between A. evermanni and A. stomias are subtle, they have been recorded as one species, A. stomias, in the eastern Bering Sea resource assessment surveys of the Northwest and Alaska Fisheries Center (NWAFC) (Smith and Bakkala 1982). Food habits of A. stomias have been studied by some researchers (Gotshall 1969; Kabata and Forrester 1974; Smith et al. 1978), but none of those studies covered the food habits of A. evermanni. Shuntov (1970) studied the feeding intensity of the two Atheresthes species, but he did not compare the diets of these species. Using electrophoretic examination, Ranck et al. (1986) have confirmed that these two types of Ather- esthes are separate species. The purpose of this study is to analyze stomach samples of these two con- 1 Fisheries Research Institute, University of Washington, WH-10, Seattle, WA 98195. 2Northwest and Alaska Fisheries Center, National Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA 98115. Manuscript accepted December 1985. FISHERY BULLETIN: VOL. 82, NO. 3, 1986. generic species collected in the area of their distri- butional overlap in the eastern Bering Sea and compare the diets of both fish species to calculate the degree of diet similarity to determine whether the two species can be considered trophically equivalent. COLLECTION AND PROCESSING OF SAMPLES Specimens were collected from 6 July to 16 July 1983 in the eastern Bering Sea aboard the Alaska, a research vessel participating in the annual sum- mer resource assessment survey conducted by the Resource Assessment and Conservation Engineer- ing (RACE) division of the NWAFC in Seattle, WA. Stomachs of arrowtooth flounder and Kamchatka flounder were taken at standard resource assess- ment stations where half-hour tows were made using an 83-112 Eastern bottom trawl net with an estimated 2.3 m vertical and 16.4 m horizontal mouth opening. The samples were collected in an area around and to the northwest of the Pribilof Islands at bottom depths ranging from 71 to 137 m (Fig. 1, Table 1). A random subsample of individuals of both arrow- tooth flounder and Kamchatka flounder was ob- tained at each station with a total collection of 348 stomachs from 19 stations. Individual fish were first checked for signs of re- gurgitation, i.e., food items in mouth or gill plates or flaccid stomach, and were discarded if any such signs were noted. Stomachs from the remaining fish 615 FISHERY BULLETIN: VOL. 84, NO. 3 63 00N ■- 61 00N 59 00N - 57 00N 55 00N - 53 00N 5) OON 179 OOE 1 76 OOW 171 OOW 166 OOW 161 OOW 56 OOW Figure 1.— Sampling locations for arrowtooth flounder, Atheresthes stomias, and Kamchatka flounder, A. evermanni, in summer 1983 in the eastern Bering Sea. were excised along with the anterior portion of the body (including head, stomach, and intestines), and these samples were sent to the laboratory for species identification. Each specimen was placed in a muslin bag with a specimen label bearing fork length, sex, and station information. All samples were preserved in 10% Formalin3. In the laboratory, two characters were used for species identification: the position of the left eye relative to the dorsal profile and gill rakers. Kam- chatka flounder has the upper eye completely on the right side of the head and 13 or fewer gill rakers on the first arch. Arrowtooth flounder has an up- per eye which interrupts the dorsal profile of the head and 15 or more gill rakers on the first arch (Norman 1934; Willimovsky et al. 1967). 3Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Stomachs were analyzed individually. Prey items were identified to the lowest taxonomic level prac- tical, counted, and weighed damp to the nearest milligram. The standard length of fish prey were also measured. DATA ANALYSIS Specimens of A. stomias and A. evermanni were divided into 100 mm fork-length groups for data analysis: <201 mm, 201-300 mm, 301-400 mm, and >400 mm. Percent of frequency of occurrence (% FO), percentage of total stomach content weight (% W), percentage of total prey number (% N) and the Index of Relative Importance [IRI = % FO (% N + % W)] (Pinkas et al. 1971) were calculated for ma- jor categories of prey items in the 100 mm size groupings of A. stomias and A. evermanni. Based on a review of dietary overlap measures 616 YANG and LIVINGSTON: FOOD HABITS AND DIET OF TWO CONGENERIC SPECIES Table 1 .—Station information and number of stomachs collected at each station of arrowtooth flounder (ATF) and Kamchatka flounder (KF) in the eastern Bering Sea for the summer 1983. Alaska Haul Bottom No. ATF1 No. KF1 daylight depth temp. Latitude Longitude stomachs stomachs Haul Date time (m) (°C) N W collected collected 100 7/6 1000 71.3 3.1 56°29.46' 169°15.61' 14 (6) 19(17) 101 7/6 1200 71.3 3.1 57°19.25' 168°59.13' 4 (5) 2(1) 102 7/6 1400 76.8 2.6 57°10.87' 169° 10.22' 14(6) 7(6) 105 7/7 0800 102.4 2.7 56°39.96' 168°55.22' 18 (9) 7(4) 106 7/7 1100 133.5 3.8 56°20.18' 168°53.24' 11 (2) 10 (3) 107 7/7 1600 111.6 3.5 58°20.01' 170°02.15' 12 (0) 12 (1) 108 7/8 0800 96.9 3.8 56°40.00' 170°04.47' 5(1) 3(1) 114 7/9 1400 75.0 2.7 57°39.23' 170°16.12' 13 (1) 2 (0) 135 7/13 1900 96.9 1.5 58°22.09' 171°37.39' 6(1) 0(0) 136 7/14 0700 100.6 2.8 58°01.90' 171°33.37' 5 (0) 1 (0) 137 7/14 1000 100.6 3.0 57°41.94' 171°31.16' 7(0) 0(1) 138 7/14 1300 102.4 3.7 57°21.05' 171°28.26' 8 (3) 6(1) 139 7/14 1500 109.7 3.9 57°02.54' 171°25.63' 11 (0) 6(1) 140 7/14 1800 120.7 3.7 56°43.42' 171°23.38' 6 (0) 9 (0) 141 7/15 0700 137.2 4.0 56°42.72' 172°32.33' 2(6) 7(2) 142 7/15 0900 124.4 3.6 57°00.84' 172°39.37' 5 (6) 7(2) 144 7/15 1500 122.5 3.7 57°40.10' 172°47.92' 3(1) 2 (0) 146 7/16 0600 111.6 2.6 58°20.10' 172°55.00' 2 (2) 9 (0) 147 7/16 0900 115.2 2.6 58°40.07' 172°59.18' 3(0) 3(0) Total 149 (47) 112 (40) 'Stomachs containing food, number of empties in parentheses. (Cailliet and Barry 1979; Linton et al. 1981), Schoener's (1970) index was chosen because it was found to measure overlap accurately over most of the range of potential overlap (Linton et al. 1981). Schoener's index, Cxy, is calculated as Cxy = 1.0 - 0.5(I|p^ - VJ) where px t and pyi are the estimated proportions by weight of prey i in the diets of species x and y, respectively (the percentage by weight of prey items in Table 2). The index ranges from 0 which indi- cates no dietary overlap to a maximum overlap of 1 when all prey items are found in equal propor- tions. communis, were the dominant shrimp consumed. Walleye pollock constituted the highest proportion of the diet for all size groups of flounder, ranging from 56% by weight of the diet for Kamchatka flounders 301-400 mm long to about 88% by weight for arrowtooth flounders >400 mm long. Miscella- neous food items consumed included polychaetes, copepods, cumaceans, hippolytid shrimps, ophi- uroids, and various fish species. Mean stomach content weight of those stomachs with food was similar between arrowtooth flounder and Kamchatka flounder for all but the largest size group. The mean stomach content weight ranged from about 1.4 g for the small flounders to over 20 g for the largest size group. RESULTS General Feeding Trends A total of 348 stomachs were analyzed; 87 stomachs (25%) were empty. Table 2 shows the per- centages by weight of all prey items found in the stomachs of both flounder species by size group. In general, both species consumed the same prey species or groups: euphausiids, pandalid and crangonid shrimps, and walleye pollock (Fig. 2). Thysanoessa inermis and T. raschii were the domi- nant euphausiids consumed. Some pandalid shrimps were eaten by smaller (<301 mm) flounders of both species, but crangonid shrimps, mainly Crangon Diet Comparisons Within Size Groups The principle diet of both Atheresthes species in the <200 mm size group was comprised of walleye pollock, euphausiids, and shrimps (Fig. 3). Walleye pollock comprised 58% and 65.5% by weight of the diet of Kamchatka flounder and arrowtooth flounder, respectively. Euphausiids comprised the highest percentage by numbers of the diet of both species, 53% for Kamchatka flounder and 69.4% for arrowtooth flounder. Shrimps, including Crangon communis, Pandalus goniurus, Pandalus tridens, and Eualus avinus, constituted 17.1% and 7.2% by weight of the diet of Kamchatka flounder and arrow- tooth flounder, respectively. Other less important 617 FISHERY BULLETIN: VOL. 84, NO. 3 Table 2.— Percentage by weight of prey items in the stomachs of arrowtooth flounder (ATF) and Kam- chatka flounder (KF) by 100 mm FL categories; and Schoener's indices (Cxy) of diet overlap between the two species. Predator size group (mm) <200 201-300 301-400 >400 Prey item KF ATF KF ATF KF ATF KF ATF Invertebrates Polychaeta — — — 0.28 — — — — Copepoda — 0.01 — — — — — — Mysidacea 0.45 0.34 0.12 — — — — — Cumacea — — 0.01 — — — — — Amphipoda 0.22 — 0.52 — 0.07 — — — Euphausiiacea Unidentified 5.64 8.99 0.23 9.10 0.54 3.86 — 0.22 Thysanoessa rachii — 2.76 — 1.35 — 2.32 — — T. inermis 4.28 10.67 3.55 9.33 7.40 10.03 4.09 6.55 Caridea Unidentified 1.05 1.31 — — — 0.08 — — Hippolytidae Eualus avinus 0.88 — — 0.05 — — — — Pandalidae Unidentified — 1.08 — — — — — — Pandalus goniurus 3.89 — — — — — — — Pandalus tridens 4.77 — — — — — — — Pandalus sp. 0.54 0.37 0.24 — — — — — Crangonidae Unidentified 0.34 2.98 2.15 0.10 — — 0.07 — Crangon dalli — — 0.85 0.61 — — — — C. communis 5.67 1.43 5.53 0.70 0.31 0.15 — 0.22 Paguridae — 0.01 — — 0.58 — — — Ophiuroidea — 0.19 — — 0.03 — — — Chaetognatha Sagitta sp. — 0.03 — 0.01 — — — — Pisces Gadidae Unidentified — — 5.40 4.62 2.70 5.65 8.20 — Theragra chalcogramma 58.03 65.51 81.55 71.69 55.78 76.99 85.87 87.96 Zoarcidae Unidentified — — — — 5.55 — — 5.06 Ly codes brevipes — — — — 8.20 — — — Cottidae — 1.15 — — — — — — Stichaeidae Unidentified 3.00 — — — — — — — Lumpenus maculatus — 3.16 — 1.97 9.47 — — — Pleuronectidae Unidentified 7.50 — — — 9.36 — 1.77 — Atheresthes sp. 3.73 — — — — — — — Unidentified organic material — — 0.05 0.19 — 0.92 — — No. of stomachs with food 32 40 43 53 20 40 19 14 Total weight of stomach content (g) 46.96 57.24 93.29 167.66 181.89 291 .43 383.32 467.91 Mean stomach content weight (g) 1.47 1.43 2.17 3.16 9.09 7.29 20.17 33.42 Mean fish length (mm) 187.80 184.60 250.10 260.70 350.50 341.30 441.10 450.00 Cxy 0.72 0.82 0.67 0.90 food items were stichaeids, pleuronectids, cottids, mysids, and amphipods. Walleye pollock, the dominant food of both Atheresthes species in the 201-300 mm size group (Fig. 3), constituted 81.6% and 71.7% by weight of the diet of Kamchatka flounder and arrowtooth flounder, respectively. Euphausiids comprised 20% by weight of the diet of arrowtooth flounder. How- ever, euphausiids only comprised 3.8% by weight (39.9% by number) of the diet of Kamchatka flounder. Shrimps (Crangonidae, Pandalidae) were more important food for Kamchatka flounder (8.8% by weight) than for arrowtooth (1.4% by weight). Unidentified gadoids comprised 5.4% and 4.6% by weight of the diet of Kamchatka flounder and arrow- tooth flounder, respectively. Other less important food items were polychaetes, mysids, amphipods, and the stichaeid Lumpenus maculatus; they were 618 YANG and LIVINGSTON: FOOD HABITS AND DIET OF TWO CONGENERIC SPECIES N = 32 40 43 53 20 40 19 14 100 r- 50 - > .a O 0 u yr. ■ Euphausiids Shrimp Other fish Pollock Other KAKAKAKA <200 201-300 301-400 > 400 Predator length group (mm) Figure 2.— Percentage by weight of major prey categories in the diet of arrowtooth flounder (A), Atheresthes stomias, and Kamchatka flounder (K), A. evermanni, for dif- ferent length groups of fish collected from the eastern Bering Sea in summer 1983. all <5% by weight of the diet of both Atheresthes species. The principle diet by weight of Kamchatka flounder in the 301-400 mm size group was com- prised of 55.8% walleye pollock, 13.8% zoarcids, 9.4% pleuronectids, 9.5% stichaeids, and 7.9% euphausiids (Table 2, Fig. 3). Walleye pollock also dominated the diet of arrowtooth flounder (77% by weight). The other two main items of arrowtooth flounder were euphausiids (16.2% by weight) and unidentified gadoids (5.7% by weight). Shrimps were not important food for either Atheresthes species of this size; they contributed <1% by weight of the diet. Other less important prey items were ophiuroids and pagurids. Numerically, euphausiids dominated the food for both species (90.7% for Kamchatka flounder, 96.0% for arrowtooth flounder). Walleye pollock dominated the food of the two Atheresthes species in the >400 mm size group (Fig. 3). It constituted 85.9% and 88.0% by weight of the diet of Kamchatka flounder and arrowtooth flounder, respectively (Table 2). Though euphausiids dominated the food by number (91.5% for Kam- chatka flounder, 97.0% for arrowtooth flounder), they only contributed 4.1% and 6.8% by weight of the diet of Kamchatka flounder and arrowtooth flounder, respectively. In addition to walleye pollock, unidentified gadoids comprised 8.2% and pleuronec- tids comprised 1.8% by weight of the diet of Kam- chatka flounder. Zoarcids comprised 5.1% by weight of the diet of arrowtooth flounder. Shrimps played a less important role in the food of both Atheresthes species (<1% by weight). Diet Comparison Among Size Groups There was not much difference in diets among size groups in the proportion by weight of the prey categories such as euphausiids and fish (Fig. 2). However, shrimps disappeared from the diets of flounders in the two larger size groups. The number of different species in the diet also changes with size. The <200 mm size group of flounders consumed about 11 or 12 different prey categories while the >400 mm size groups consumed only 3 or 4 differ- ent prey types (see Table 2). Even though the proportion by weight of fish in the diet remained fairly constant over flounder size groups, the size of individual fish consumed did change with flounder length. Figure 4 shows the fre- quency distribution of fish prey lengths found in the stomachs of different size A. evermanni. Most of the prey fish were age-0 juvenile pollock (<100 mm) for the two smaller size groups and age-1 juvenile 619 FISHERY BULLETIN: VOL. 84, NO. 3 A. evermanni A. s torn i as CM 2 %N 40 ii 20 in 0 20 £" %W 40- V/2 60- _ POL in MYS CRA PAN ^LZ^-HIP lz PLE NST, 1 11,1 i i i i 100 200 80 60 o %N 40 7 20 UJ o O ii Ml £ %W 0 20 40 60 80 - POL EUP MYS / STI ~ / COT PAN 1 1 1 1 i i i l l 100 200 %N 40 So 20 h 7 o o' o, 20 40 - ii %W 60 80 100 EUP AMP / t MYS i i I I I I I 1 1 1 I 0 100 80 %N 60 So 40 - 20 LLI 0 °,<£ %W 40 60 100 200 0LU So ^ CM AMP OPI I OP ~QV ZOA \ UGAD _1 I I I L_l I I 1 I I 0 100 200 " %N o ^ .— n %W >- II 8^ %N on - EUP 80 60 40 20 POL CRA / .UGAD 0 20 40 N STI 60 80 i i i i i i i i i i i %w 100 200 ion EUP 80 60 40 20 0 - U GAD POL /^AMP CRA 2U 40 _ 60 80 i i i i i i 100 200 (N II en o *- g " ioor 80 %N 60 %W 40 20- 0 20 40 60 80 100 / U GAD -1 1 1 L ro ii UJ O ' o %N %W 100 EUP 80 60 - 40 20 POL CRA / 0 - -1 20 40 \ ZOA 60 - 80 100 - i 1 1 1, 1 l l 1 l l 0 100 200 100 200 % F.O. % F.O. Figure 3.— Indices of Relative Importance of major prey items in the diets of Atheresthes evermanni and A. stomias of different size groups. % F.O., percent frequency of occurrence; % N, percentage of prey number; % W, percent- age of total stomach content weight; POL, pollock; EUP, Euphausiacea; CRA, Crangonidae; PAN, Pandalidae; AMP, Amphipoda; PLE, Pleuronectidae; MYS, Mysidacea; STI, Stichaeidae; HIP, Hippolytidae; ZOA, Zoarcidae; U. GAD, Unidentified Gadidae; COT, Cottidae; S, number of stomachs containing food; E, number of empty stomachs. 620 YANG and LIVINGSTON: FOOD HABITS AND DIET OF TWO CONGENERIC SPECIES (U Q- Pollock Other fish Predator length: • 400 mm n= 15 x= 147.2 1 > I i | I | ' I T Predator length 301-400 mm n = 15 x = 117.58 I 'T ' I ' I ' I i | ' | Predator length: 201-300 n = 10 x = 72.31 I l I ' I l I I I i I i I ' I Predator length: < 200 mm n = 47 x = 43.03 c"* I "i"l i I I I ' I I I ' I I 1 1 I i I 120 160 200 240 280 Prey (fish) length (mm) Figure 4.— Frequency distribution of standard lengths of prey fish found in the stomachs of Atheresthes species from the eastern Bering Sea in summer 1983. pollock (100-200 mm) for the two larger size groups. The fish prey length was plotted against the predator length (Fig. 5). Fish prey size appears to increase linearly with increasing predator size. Diet Overlap Values for Schoener's (1970) index of dietary over- lap were obtained from a comparison (by weight) between the diets of Kamchatka and arrowtooth flounder of the same size groups (Table 2). All the values obtained were >0.60, an indicator of high dietary overlap (Langton 1982). The <200 mm size group had an overlap value of 0.72 and the 201-300 mm size group had an overlap value 0.82. Within each of these two size groups, fairly similar propor- tions by weight of walleye pollock, euphausiids, and shrimps were consumed. The 301-400 mm size group had the lowest overlap value of 0.67. This is probably because Kamchatka flounder ate less walleye pollock by weight (56%) than did the arrowtooth flounder (77%). Most of the remainder of the diet for Kam- chatka flounder in this size group was composed of different fish groups, such as zoarcids, stichaeids, and pleuronectids, which were almost totally absent from the arrowtooth' s diet at this size. The largest size group of flounders (>400 mm) had the highest overlap value of 0.90. This size group ate very similar proportions by weight of walleye pollock and euphausiids. DISCUSSION From this study, it appears that both Kamchatka flounder and arrowtooth flounder are largely fish feeders. Walleye pollock was the most frequently observed prey and contributed the largest percent- age by weight to the diets, followed by euphausiids and shrimps (Table 2, Fig. 3). Gotshall (1969) found that ocean shrimp, Pandalus jordani, was the most common food item of arrowtooth flounder (because the stomachs were collected on commercial shrimp grounds), followed by fishes and euphausiids. Pacific sanddabs, Citharichthys sordidus, were the most numerous prey fish found in his study. Kabata and Forrester (1974) examined 753 arrowtooth flounder collected off the west coast of Vancouver Island. Their study showed that euphausiids, followed by fish were the predominant foods taken by arrow- tooth flounder. The most commonly found species of fish were eulachon, Thaleichthys pacificus, and Pacific herring, Clupea pallasii. Smith et al. (1978) found that fish constituted 41.09% and euphausiids 37.22% by volume of the food of 236 arrowtooth flounder collected from the northeast Gulf of Alaska. Walleye pollock were most commonly consumed fish prey. Moiseev (1953) found that Kamchatka flounder fed almost exclusively on pollock and only occa- sionally on herring and other fishes. The type of prey eaten by a fish is strongly corre- lated with the morphology of the alimentary tract of the fish (De Groot 1971; Ebeling and Cailliet 1974; Allen 1982). Structure of the digestive tract of arrowtooth flounder and Kamchatka flounder are very similar. Both have a very large terminal mouth that is nearly symmetrical with a wide gape; teeth are arrow-shaped and well developed on both sides of the jaws; gill rakers are long and strongly dentate; 621 FISHERY BULLETIN: VOL. 84, NO. 3 £ E en c > 300 r 250 200 - 150 100 - 50 - Y = 0.39X-26.33 r-0.85 • • • • _ • • • • • • • • • • • • • • • • • • — v*^""'^ * * ' • ■ 1 ■ i 150 200 250 300 350 400 450 500 Predator length (mm) Figure 5.— Scatter plot of prey fish length consumed by Atheresthes species from the eastern Bering Sea in summer 1983. and the esophagus and stomach are large with four large pyloric caeca and the intestine is a simple loop. All of these characteristics indicate that Atheresthes species are fish feeders as predicted by using De Groot's (1971) morphological criteria. He stated that large gill rakers with teeth are indispensable to fish feeders, since they prevent the prey, grasped alive, from struggling out of the mouth. The high per- centages of fish in the diet of the two Atheresthes species obtained in this study would be expected on the basis of the similarities in the digestive tracts of the two species. The results also indicate that Atheresthes species feed up in the water column. According to Allen (1982), flatfishes with large symmetrical mouths (Atheresthes species) probably use sight to locate prey. They are oriented up in the water column when foraging. The presence of pelagic fish (T. chalcogramma) and euphausiids or nektonic bentho- pelagic crustaceans such as shrimps in the diets of Atheresthes species supports Allen's generalizations concerning correlations between morphology and feeding behavior in flatfishes. The trend of the feeding habits of Atheresthes species with regard to predator length is toward piscivory; that is, when the predators are bigger, they take more fish (by weight) as food. Specimens from the <200 mm size group were found to ingest the greatest variety of prey items in comparison to other size groups. Specimens >400 mm long preyed mainly on other fishes, primarily on pollock. However, euphausiids were of importance in the diet of all size groups. One 460 mm arrowtooth flounder was found to have 838 Thysanoessa inermis in its stomach. Smith et al. (1978) also noted a change in food habits with increasing length in the arrowtooth flounder. In their study, specimens over 450 mm long preyed exclusively on pollock and other gadoids. Euphausiids were important food of the arrowtooth flounder up to 350 mm long; however, none were found among the stomach contents of specimens larger than 350 mm. Based on the results of this study and those of Smith et al. (1978) and Gotshall (1969), it appears that Atheresthes species are opportunistic feeders; they feed on those prey items that are most abundant— pollock and euphausiids in the Gulf of Alaska and eastern Bering Sea and ocean shrimp in northern California. In the eastern Bering Sea, the estimated abundance of age-0 pollock in 1982 is between 100 billion and 1,300 billion and, based on the results of the 1983 bottom trawl survey by NWAFC, this 1982 year class is the largest observed since the large 1978 year class (Traynor in press). 622 YANG and LIVINGSTON: FOOD HABITS AND DIET OF TWO CONGENERIC SPECIES In spite of the high diet overlap between Kamchatka flounder and arrowtooth flounder, there is probably no competition for food between these two species because they are exploiting abundant food sources. Finally, although Kamchatka founder and arrow- tooth flounder are genetically distinct, they can be considered trophically equivalent on the basis of their similar diets and high diet overlap. LITERATURE CITED Allen, M. J. 1982. Functional structure of soft-bottom fish communities of the southern California shelf. Ph.D. Thesis, Univ. California, San Diego, 577 p. Cailliet, G. M., and J. P. Barry. 1979. Comparison of food array overlap measures useful in fish feeding habits analysis. In S. J. Lipovsky and C. A. Simenstad (editors), Gutshop '78, fish food habits studies; Proceedings of the 2d Pacific Northwest Technical Work- shop, p. 67-79. Univ. Wash., Div. Mar. Resour., Wash. Sea Grant, WSG-WO-79-1. De Groot, S. J. 1971. On the interrelationships between morphology of the alimentary tract, food and feeding behavior in flatfishes (Pisces: Pleuronectiformes). Neth. J. Sea Res. 5:121-196. Ebeling, A. W., and G. M. Cailliet. 1974. Mouth size and predator strategy of midwater fishes. Deep-Sea Res. 21:959-968. Gotshall, D. W. 1969. Stomach contents of Pacific hake and arrowtooth founder from northern California. Calif. Fish Game 55:75- 82. Hart, J. L. 1973. Pacific fishes of Canada. Bull. Fish. Res. Board Can. 180, 740 p. Kabata, Z., and C. R. Forrester. 1974. Atheresthes stomias (Jordan and Gilbert 1880) (Pisces: Pleuronectiformes) and its eye parasite Phrixocephalus cin- cinnatus Wilson 1908 (Copepoda; Lernaeoceridae) in Canadian Pacific waters. J. Fish. Res. Board Can. 31:1589- 1595. Langton, R. W. 1982. Diet overlap between Atlantic cod, Gadus morhua, silver hake, Merluccius bilinearis, and fifteen other North- west Atlantic finfish. Fish. Bull., U.S. 80:745-759. Linton, L. R., R. W. Davies, and F. J. Wrona. 1981. Resource utilization indices: an assessment. J. Anim. Ecol. 50:283-292. Moiseev, P. A. 1953. Treska i Kambaly dalnevestochnykh morei (Cod and flounders of Far-Eastern seas). [In Russ.] Izv. Tikhook- ean. Nauchno-Issled. Inst. Rybn. Khoz. Okeanogr. 40:1-287. (Transl. by Transl. Bur. Can. Dep. Seer. State, 576 p., available as Fish. Res. Board Can. Transl. Ser. 119.) Norman, J. R. 1934. A systematic monograph of the flatfishes (Hetero- somata). Vol. I.: Psettodidae, Bothidae, Pleuronectidae. Trustees Br. Mus., Lond., 459 p. (Available from Johnson Reprint, N.Y., 1966.) Pinkas, L., M. S. Oliphant, and I. L. K. Iverson. 1971. Food habits of albacore, bluefin tuna, and bonito in California waters. Calif. Dep. Fish Game, Fish Bull. 152, 105 p. Ranck, C, F. Utter, G. Milner, and G. B. Smith. 1986. Genetic confirmation of specific distinction of arrow- tooth flounder, Atheresthes stomias, and Kamchatka flounder, A. evermanni. Fish. Bull., U.S. 84:222-226. Schoener, T. W. 1970. Non-synchronous spatial overlap of lizards in patchy habitats. Ecology 51:408-418. Shuntov, V. P. 1970. Sezonnoe respredelenie chernogo i strelozubykh paltusov v Beringovum more (Seasonal distribution of black and arrow-tooth halibuts in the Bering Sea). [In Russ.] Tr. Vses. Nauchno-Issled. Inst. Morsk. Rybn. Khoz, Okeanogr. 70 (Izv. Tikhookean. Nauchno-Issled. Inst. Rybn. Khoz. Okeanogr. 72):391-401. [Transl. by Isr. Program Sci. Transl., 1972, In P. A. Moiseev (editor), Soviet fisheries in- vestigations in the northeastern Pacific, Part 5, p. 397-408. Available from U.S. Dep. Commer., Natl. Tech. Inf. Serv., Springfield, VA, as TT 71-50127.] Smith, G. B., and R. Bakkala. 1982. Demersal fish resources of the eastern Bering Sea: spring 1976. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-754, 129 p. Smith, R. L., A. C. Paulson, and J. R. Rose. 1978. Food and feeding relationships in the benthic and demersal fishes of the Gulf of Alaska and Bering Sea. In Environmental assessment of the Alaskan Continental Shelf, Final Rep., Biol. Stud. 1:33-107. U.S. Dep. Commer., NOAA, Environ. Res. Lab., Boulder, CO. Traynor, J. J. In press. Midwater abundance of walleye pollock in the eastern Bering Sea, 1979 and 1982. Bull., Int. North. Pac. Fish. Comm. WlLLIMOVSKY, N. J., A. PEDEN, AND J. PEPPAR. 1967. Systematics of six demersal fishes of the North Pacific Ocean. Fish. Res. Board Can., Tech. Rep. 34, 95 p. 623 ECOLOGY OF CERIANTHARIA (COELENTERATA, ANTHOZOA) OF THE NORTHWEST ATLANTIC FROM CAPE HATTERAS TO NOVA SCOTIA Andrew N. Shepard,1 Roger B. Theroux,2 Richard A. Cooper,1 and Joseph R. Uzmann2 ABSTRACT Ceriantharia, tube dwelling anthozoans, were collected in grab samples and documented by direct obser- vations and photographs from research submersibles on the continental shelf and slope off the northeast United States coast (Cape Hatteras to Nova Scotia). Two species [{Cerianthus borealis Verrill and Cerian- theopsis americanus (Agassiz)] were identified from grab samples and four species, probably including C. borealis, were observed from submersibles. Ceriantharia distribution in relation to latitude, depth, temperature, and sediments was examined. They occurred throughout the study area, abundantly at depths of 0 to 500 m and less abundantly from 900 to 2,400 m. Ceriantharia habitats displayed an extreme range in bottom water temperature (sum- mer maximum minus winter minimum) of from 8° to 16°C, and had every sediment type, except 100% gravel and coarse shifting sand. Geographic and bathymetric zonation is attributed primarily to tem- perature and secondarily to food supply and substrate type. Ceriantharia distribution patterns, in submarine canyon heads at depths of <400 m, were determined from photographic transects run with submersibles; observed patchiness may be related to local differences in food supply, sediments, and microtopography. The motile megafauna associated with Ceriantharia "forest" areas and the infauna and epifauna inhabiting ceriantharian tubes were evidence to show that tubes may enhance local species diversity and abundance in featureless soft-bottom areas by 1) attracting motile species seeking cover and 2), acting as a stable, elevated substrate for tubiculous and suspension feeding macrofauna. The possibility of exploitation of energy reserves beneath the northwest Atlantic outer continental shelf and slope has prompted many new studies and the reexamination of past investigations for baseline information on the region's seafloor communities. Research submersible studies of potential oil lease tracts identified "indicator species" for assessing environmental changes owing to drilling activities. We considered Ceriantharia suitable for this purpose because they were abundant, passive suspension feeders, and nonmobile. Literature searches re- vealed that very little has been published on the Ceriantharia species occurring from Cape Hatteras to Nova Scotia. This is surprising in light of the group's significant contribution to the benthic bio- mass of the region (Wigley and Theroux 1981) and the important functional role [the effect a species has on the distribution and abundance of other residents (Sutherland 1978)] Ceriantharia may have in structuring communities inhabiting featureless soft-bottom substrate (O'Connor et al. 1977). Woods Hole Laboratory, Northeast Fisheries Center (NEFC), National Marine Fisheries Service (NMFS), personnel have reported on the general composition and distribution of invertebrate fauna of the New England and Mid- Atlantic Bight con- tinental shelf and slope (e.g., Wigley and Theroux 1981; Theroux and Wigley 19843; Cooper et al, in press). Data on Ceriantharia were collected during ecological studies pertaining to various kinds of demersal fishes and benthic invertebrates: 1) a grab sample survey (Fig. 1) done from 1955 to 1969 (Shepard and Theroux 19834), and 2) observations, photographs, and limited sample collections from research submersible studies. Dredge and trawl data were available (Shepard and Theroux fn. 4), but not analyzed since deep burrowing Ceriantharia (some- 'NOAA National Undersea Research Program, University of Connecticut, Avery Point, Groton, CT 06340. 2Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. 3Theroux, R. B., and R. L. Wigley. 1984. Quantitative com- position and distribution of macrobenthic invertebrate fauna of the New England Region. Unpubl. Manuscr. Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Ser- vice, NOAA, Woods Hole, MA 02543. "Shepard, A. N., and R. B. Theroux. 1983. Distribution of Cerianthids (Coelenterata, Anthozoa, Ceriantharia) on the U.S. East Coast Continental Margin, 1955-1969: Collection data and environmental measurements. Lab Ref. Doc. 83-12, 24 p. North- east Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. Manuscript accepted September 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 625 H i 1 1 FISHERY BULLETIN: VOL. 84, NO. 3 45° -40° ATLANTIC OCEAN Figure 1. -Chart of the northwest Atlantic from lat. 35° to 45°N (Cape Hatteras to Nova Scotia) showing stations where grab samples of macrobenthic invertebrates were obtained, and the location of submarine canyons visited with research submersibles (1 fm = 1.83 m). times more than 1 m; Sebens5) may be poorly sampled by dragged collection gear. BK. P. Sebens, Maritime Studies Center, Northeastern Univer- sity, Nahant, MA 01908, pers. commun. February 1985. The objectives of this study are to describe 1) the Ceriantharia species encountered, 2) their general distribution in relation to latitude, depth, tempera- ture, and sediments, 3) their local distribution pat- 626 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA terns, and 4) how they interact with other benthic species. CERIANTHARIA Ceriantharians represent a small, incompletely described order of Anthozoa. Species identification is difficult, and many species probably remain undescribed since twice as many larval forms as adults are known (Hartog 1977; Hartog6). Two northwest Atlantic species have been identified; Cerianthus borealis Verrill (1873) (see also Parker 1900; Kingsley 1904; Widersten 1976) and Cerian- theopsis americanus (Agassiz 1859) (see also Ver- rill 1864; McMurrich 1890; Parker 1900; Carlgren 1912; Field 1949; Widersten 1976). Two other un- identified species have been found on the continental slope (Grassle et al. 1975; Hecker et al. 1980; Valen- tine et al. 1980; Sebens in press). Table 1 sum- marizes the geographic and bathymetric ranges of the above four species. Ceriantharia live in permanent semirigid tubes composed of a type of cnidae peculiar to the Order (called ptychocysts by Mariscal et al. 1977), mucus, and adhering substrate debris (Emig et al. 1972). The feltlike tube is usually deep purple in coloration and distinct enough to be used alone as evidence of Ceriantharia presence. New England bottom trawl fishermen are familiar with nets fouled with cerian- tharian tubes (Rogers 1979). In contrast to other burrowing anemones which have a single whorl of tentacles, Ceriantharia have two distinct whorls (marginal and oral tentacles) which remain outside the tube during feeding and rapidly retract into the tube when disturbed. Ceriantharia are protandric hermaphrodites; gametes are produced in the mesenteries and fer- tilization is external. The larvae are pelagic and duration of the planktonic stage is variable (Carlgren 1912; Hyman 1940; Robson 1966; TRIGOM-PARC 1974). Adults are capable of oral disc regeneration by budding (Hyman 1940; Frey 1970). Asexual reproduction has been described for at least one species, Aracnanthus oligopodus (Cerfontaine 1909). Ceriantharia are carnivorous passive suspension or impingement feeders (Emig et al. 1972; Carac- ciola and Steimle 1983). Digestion may begin in the tentacles, and larger particles are primarily taken up in the endoderm of sterile septa (Tiffon and Daireaux 1974). Fish species inhabiting the region, including cod, haddock, flounder, scup, and skate are known predators of whole juvenile Ceriantharia (Bowman and Michaels7) and may graze the tenta- cles of adults (TRIGOM-PARC 1974). Off the U.S. west coast, a nudibranch, Dendronotus iris Cooper, preys on adult Ceriantharia (Wobber 1970). Previous documentation of Ceriantharia in the northwest Atlantic has come from grab samples (Sanders 1956; Wigley 1968; Pearce 1972; Pearce et al. 1976; Pearce et al. 1981; Reid et al. 1981; Wigley and Theroux 1981; Caracciola and Steimle 1983) and submersibles (Grassle et al. 1975; Rowe et al. 1975; Hecker et al. 1980; Valentine et al. 1980). However, no studies in the region report exclusive- ly on ceriantharian ecology. 6J. C. den Hartog, Curator of Coelenterata, Rijksmuseum van Natuurlijke Historie, Postbus 9517, 2300 RA Leiden, Netherlands, pers. commun. March 1983. 7Bowman, R., and W. Michaels. 1983. Unpubl. data. Food Habits Program, Northeast Fisheries Center Woods Hole Labor- atory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. Table 1. — Morphologic descriptions and geographic and bathymetric ranges of previously described Ceriantharia species inhabiting the study area. Species General morphologic description Geographic range Bathymetric range (m) Ceriantheopsis americanus see Verrill 1864 Cape Cod to Florida1 20-370 Cerianthus borealis see Verrill 1873 Arctic to Cape Hatteras1 10-4500 Unidentified species I5 small (<5 cm contracted), dark brown tentacles, tube flush to seafloor.5 Continental slope off New England 5'6'7>1,000 Cerianthid A8 larger than unidentified species I, uniformly dark tentacles, tube flush to seafloor5 Continental slope off New England 5,6,7,8>1500 1 Parker 1900. 2Field 1949. 3Pearce et al. 1981. "Miner 1950, p. 196. 5Sebens in press. 6Grassle et al. 1975. 7Hecker et al. 1980. Valentine et al. 1980. 627 FISHERY BULLETIN: VOL. 84, NO. 3 METHODS Grab sample methodology (gear description, sam- ple processing, data reduction, bathymetry, tem- perature, and sediments) is reported in Wigley and Theroux (1981). A chi-square (x2) test, employing contingency tables (Richmond 1964), was used to assess ceriantharian occurrence at grab sample sta- tions (relation to latitude, depth, bottom water temperature, and sediment type). Table 2 lists the submersibles used and sampling gear employed by each. Quantitative data were ob- tained with externally mounted 35 mm camera- strobe systems. Qualitative ecological and behavioral information was acquired with 35 mm hand-held cameras, audio tapes, and video tapes made with a hand-held or externally mounted video camera. In situ faunal and sediment collections were made with the submersibles' manipulator arms. Only those dives performed to assess the distribution of mega- benthos and associated habitat types were analyzed. The externally mounted 35 mm camera systems used on Nekton Gamma, Johnson-Sea-Link, and Alvin were quantitatively calibrated, assessing 3.6 m2, 7.0 m2, and 15.0 m2 of ocean floor per photo- graphic frame, respectively (Bland et al. 1976; Cooper and Uzmann 19818). Photographs were read on either a light table with a hand-held magnifying glass or motorized micro- film reader with a 36 x 36 cm screen and 15 x magnification lens. Each photograph was time- annotated, thus allowing correlation with depth, 8Cooper, R. A., and J. R. Uzmann. 1981. Georges Bank and Submarine Canyon living resources and habitat baselines in oil and gas drilling areas. Northeast Monitoring Program Annual Report for FY 80. Unpubl. manuscr., 34 p. Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. temperature, slope angle, substrate-habitat type, and current speed and direction documented on hand-held audio recorders during the dives. RESULTS Species Identification Ceriantharia occurred at 229 of the 1,295 grab sample stations; 990 anemones were caught at 139 stations, whole tubes only at 29 stations, and tube fragments at 61 stations (Fig. 2). Two species, Ceri- antheopsis americanus and Cerianthus borealis, were identified from grab samples (at four stations), the remaining anemones were identified only as Ceriantharia. The mean blotted wet weight of the 990 anemones was 5.0 g (95% C.L. = ±3.6); how- ever, more than 90% weighed less than the mean. Ceriantharia occurred at 82% of the submersible dive sites (Appendix Tables 1, 2) and at every ma- jor geographic feature visited (Fig. 2, Table 3). Sub- mersible samples have not yet yielded anemones suitable for identification to the species level. Figure 3 shows three of the four species (Cerianthids A, B, C, and D) photographed from submersibles, and Table 3 classifies the species by morphological features apparent in photographs. The minimum gross Ceriantharia size (height above seafloor or width of exposed tentacle crown and/or tube) visible in photographs was about 5 cm. It was not unusual to see large Cerianthid B or C tubes 20 cm above the seafloor. Based on laboratory examination of 61 anemones and a few specimens which were photographed in situ and then collected with the manipulator arm, a gross size of 5 cm cor- responds to an anemone wet weight of about 16 g (3 times the mean weight of anemones captured with grab samplers). Table 2.— Submersible, cruise year, and gear used for data col- lection. PC8 = Perry Model C8, NG = Nekton Gamma, AL = Alvin, and JSL = Johnson-Sea-Link. In situ Submersible/ Audio Video 35 mm photographs collections of fauna/ year tapes tapes Hand-held External substrate PC8/1971 X X NG/1973 X X X NG/1974 X X X NG/1979 X X X AL/1975 X X X X AL/1976 X X X AL/1978 X X X AL/1980 X X X X JSL/1980 X X X X JSL/1981 X X X X Relation to Latitude Ceriantharia occurrence at grab sample stations was not independent of latitude (x2, P < 0.05). Oc- currence was highest in three areas: off Chesapeake Bay Gat. 37° to 38°N); south of Cape Cod in the zone also including the southern half of Georges Bank (lat. 40° to 41°N); and on the shelf off Nova Scotia (lat. 44° to 45 °N) (Fig. 4). From submersibles, Cerianthid B was the only species seen on Georges Bank, or north of 41 °N [Wilkinson Basin (Gulf of Maine) and Corsair Canyon]; Cerianthids A, C, and D were all seen in canyons or on the slope south of Georges Bank (Table 3). 628 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA H 1 1 1 i TLANTIC OCEAN O GRAB SAMPLE STATIONS □ SUBMERSIBLE DIVE(S) ■/OOfm Figure 2.— Chart showing the submersible dive(s) sites and grab sample stations containing Ceriantharia. Symbols for submersible dive(s) sites often circumscribe more than one dive, since at this scale some dives were too close together to distinguish with separate symbols (1 fm = 1.83 m). Relation to Bathymetry In grab samples, Ceriantharia were found at depths from 6 to 2,329 m, but occurrence was not independent of depth (x2, P < 0.05). Occurrence was highest from 0 to 100 m, and no Ceriantharia were caught from 501 to 900 m (Fig. 4). Submersible dive depth range was 80 to 1,930 m (Appendix Tables 1, 2). Cerianthids B, C, and D were seen within the 80-400 m range, no species were 629 FISHERY BULLETIN: VOL. 84, NO. 3 Figure 3.— A, B, C, - black and white prints of 35 mm Ektachrome transparencies from a hand-held camera; D - from externally mounted 35 mm brow camera. A. Alvin dive 838, axis of Oceanographer Canyon, 1,740 m: Cerianthid A (dark anemones); white brittle stars, Ophiomusium sp.; sea urchins, Echinus affinus; and a grenadier (Macrouridae) on a calcareous silt-sand substrate. B. Nekton Gamma 1974 dive 30, head of Lydonia Canyon, 300 m: Cerianthid B with a blackbelly rosefish, Helicolenus dac- 630 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA tylopterus, at its tube base, on silt-clay substrate. C. Nekton Gamma 1974 dive 30, 300 m: Cerianthid B with a portunid crab, Bathynectes sp., at its tube base on silt-clay substrate. D. Nekton Gamma 1979 dive 3, head of Block Canyon, 150 m: Cerianthid C with tube epifauna (sponges and colonial white anemones), and redfish, Sebastes sp., just visible near center of the photograph on a silt-clay substrate, current direction was from left to right. 631 FISHERY BULLETIN: VOL. 84, NO. 3 Table 3.— Morphological features, apparent in photographs taken from submersibles, used to distinguish between four Ceriantharia species seen, and the geographic areas and bathymetric ranges in which they were found (cf. Fig. 3, Appendix Tables 1 and 2). Species Tube height in relation to seafloor Above Flush Characteristics of marg Length Arrangement nal tentacles Coloration Geo- graphic areas1 Depth range (m x 100) A X unequal multiplanar2 dark red, black 6,8 16-19 B X unequal multiplanar pale purple, pink, tan, or brown 1-8, 10 1-4 C X equal parabolic2 white with purple marks 4,9 2-4 D X unequal multiplanar greenish yellow 10 2-3 11 - Wilkinson Basin, Gulf of Maine; 2 - Georges Bank; 3 - Corsair Canyon; 4 - Lydonia Canyon; 5 - Gilbert Canyon; 6 - Oceanographer Canyon; 7 - Hydrographer Canyon; 8 - Veatch Canyon; 9 - Block Canyon; 10 - Hud- son Canyon. 2Used by Meyer (1980) to characterize feeding nets of other passive suspension feeders. U 0 23 1 3 U 20 IS IB S 0 30 23 20 IS 10 S 0 3 5 9 16 DEPTH (M X IBB) ■ p • n 1 1 \\^ sAA \V ,\v 4 8 12 16 20 TEMPERATURE RRNGE (AT.°C) _ra. 1 p 1 ^ s ^ ^ s 2 3 4 SEDIMENT TYPE observed at depths from 400 to 1,600 m, and Ceri- anthid A was seen at depths from 1,600 to 1,930 m (Table 3). Relation to Bottom Water Temperature Temperature observations were sparse for grab sample stations, so, the extreme range of temper- ature (A T), a commonly used measure of climatic variability (MacArthur 1975), was used to compare temperature with Ceriantharia distribution; A T equals the difference between extreme annual recorded temperatures (summer high minus winter low), obtained from various published sources, and measurements, made by the NEFC. Site ranges were grouped for plotting: 0° to 3.9°C, 4° to 7.9°C, 8° to 11.9°C, 12° to 15.9°C, 16° to 19.9°C, and >19.9°C. Temperature range changed significant- ly with latitude and depth. Largest A T's generally dominated shelf waters south of lat. 41°N, and in- shore waters (Fig. 5). Ceriantharia occurrence at grab sample stations was not independent of temperature range (x2, P < 0.05); occurrence was highest on the continental shelf where A T was from 8° to 15.9°C (Fig. 4). All submersible dives were performed in July or August. Bottom water temperatures (external Figure 4.— Ceriantharia occurrence (% of grab sample stations) in relation to latitude, depth, temperature range (AT = summer high minus winter low), and sediment type. Depth stratum size was determined by pooling, from shallow to deep, adjacent 100 m depth intervals until enough observations were available for a chi-square test. Sediment type codes are 1 - gravel; 2 - gravel/sand, silt, mud or clay; 3 - sand; 4 - silt/sand; 5 - silt/clay. 632 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA Figure 5.— Distribution of extreme range in bot- tom water temperature (summer maximum minus winter minimum) in the Middle Atlantic Bight (from Wigley and Theroux 1981) and New England region (Theroux and Wigley, text footnote 3). Cape Hatteras 633 FISHERY BULLETIN: VOL. 84, NO. 3 thermometer observations) decreased with depth; temperatures ranged from 5° to 13 °C at depths <500 m and declined gradually from 5°C at 500 m to 3.5°C at 1,900 m (Appendix Tables 1, 2). Depth- temperature profiles of three Alvin dives (Fig. 6) indicate depths of 500 to 600 m were a transition zone; deeper bottom water temperatures decreased little with depth, in comparison to shallower tem- peratures. Cerianthids B, C, and D were seen at temperatures of 5.3° to 13.0°C, and Cerianthid A was observed only in colder, deeper water, in the narrow range of3.5°to3.9°C (Appendix Tables 1, 2). 0? Q X Q. U O 10 12 • TEMPERATURE (°C) 2 4 S 8 10 i i i i 1 1 r 1 i * - #576 0 - #578 + - #579 Figure 6.— Depth-temperature profiles constructed from obser- vations (of external thermometers) made on the bottom during three Alvin dives in Veatch Canyon. Temperature stabilized at about 500-600 m. Relation to Sediments Ceriantharia occurrence at grab sample stations was not independent of sediment type (x2, P < 0.05); they rarely inhabited 100% gravel sediments (Fig. 4). However, when stations with 100% gravel sediments were not included, occurrence was in- dependent of sediment type (x2, P > 0.05). Al- though occurrence in silt-clay sediments was lower than in other unconsolidated sediments (Fig. 4), this may be a result of the large proportion of silt-clay sediment stations at depths >500 m, where Ceri- antharia were scarcer; if only silt-clay sediments from shallower than 500 m are analyzed, occurrence is more than 20%. Photographic transect profiles of submersible dives (Appendix Table 2, depths <400 m) provided information on Ceriantharia abundance with respect to substrate, depth, temperature, transect direction, and distance (Figure 7 shows one profile). Based on the number of sightings in various substrata (Ap- pendix Tables 1, 2) and the transect profiles, about 70% of the Ceriantharia inhabited silt-sand and silt- clay sediments. However, they also commonly occurred in rarer gravelly substrates (less than about 50% gravel cover on sand or clay; only about 20% of the total seafloor viewed). They were not seen in coarse sand sediments (usually rippled and/or in dune formations). The clay substrate observed from submersibles was actually a semiconsolidated mud (Cooper et al. in press); the term clay was used to differentiate it from sand substrates, but clay may only be a minor constituent. Spatial Pattern Ceriantharia density and biomass estimates from grab sample data were determined for comparison to other studies (e.g., Sanders 1956; Pearce et al. 1981; Reid et al. 1981; Caracciola and Steimle 1983). However, because no replicate sampling was done at over 90% of the stations, density and biomass were not analyzed further. For stations with anemones or whole tubes, mean density was 35.7 m"2 (N = 168, 95% C.L. = ±12.1, range = 1.7 to 1,370 m~2). Mean station biomass (anemone blotted wet weight) was 48.6 g m~2 (N = 139, 95% C.L. = ±35.4). On the quantitative submersible dives, Cerian- tharia density ranged from 0 to 0.414 m~2 dive-1 (Appendix Table 2). The maximum density in one photographic frame was 6.6 m~2. The photographic transect profiles (Fig. 7) showed Ceriantharia populations shallower than 400 m were spatially ag- gregated. No quantitative information was available for the Cerianthid A populations seen in the axes of Oceanographer and Veatch Canyons. The largest aggregation encountered (head of Lydonia Canyon, Fig. 7) was over 0.5 km wide and composed mostly of Cerianthid B, with some Cerian- thid C individuals. The dives were run over a perma- nent station marker (37 khz pinger) positioned on a 14-15 m high knoll. Substrate atop the knoll was gravel-sand, near the base and surrounding the knoll was silt-sand. Approximately half of the Cerian- tharia aggregation occupied the gravel-sand sedi- ments. Ceriantharia were the dominant megafauna in the area, other common megafauna were gala- theid crabs, Munida iris Milne-Edwards, and 634 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA IA O I a. o x < cc Id IB IA SUBSTRATE 152 M 151 M i i i i — i — T — r 200. 250. PHOTO # — i — i — i — r 400. 0.0 0.6* TRANSECT LOCATION (KM) 1.2 Figure 7.— One example, from 1980 Johnson-Sea-Link dives 15 and 16 in Lydonia Canyon, of the photograph-by-photograph transect profiles of Ceriantharia abundance constructed for quantitative submersible dives during which Ceriantharia density exceeded 0.1 m"2 dive"1, at depths of less than 400 m. Substrate codes: 1 - sand base, IA - silt veneer, IB - greater than 5% gravel cover. A permanent station marker (37 khz pinger) was located at 0.6 km into the transect, as denoted by the asterisk. asteroids on gravel-sand, and shell-less hermit crabs, Catapagurus sp., on silt-sand. Galatheids were also observed on silt-sand sediments, often near cerian- tharian tubes. A qualitative observation made on several submersible dives was that Ceriantharia "forests" (aggregations) were often associated with rises in seafloor topography. Functional Role Figure 8 (data from 1979 Nekton Gamma dive #3 in Block Canyon) shows Cerianthid C frequency of occurrence and number of associated species (diver- sity) plotted by photographic frame. The substrate throughout the dive was a low-relief silt-clay, and -i — i — i — i — | — i — n — | V i — i — | — i — i — i — | — i — i — i — | — i- 20 40 60 80 100 120 140 160 PHOTO n Figure 8.— Cerianthid C abundance and diversity (number of species) of associated fauna along a 1.0 km photographic transect from 1979 Nekton Gamma dive #3. Each data point represents the sum of 5 adjacent photographic frames: species diversity increased significantly in areas with Ceriantharia (Mann-Whitney test, P < 0.01). 635 FISHERY BULLETIN: VOL. 84, NO. 3 the depth and temperature ranges were 137 to 183 m, and 13.0° to 10.7°C. Mean number of species was significantly higher in photographs with Cerian- tharia (Mann-Whitney test, P < 0.01): Three groups of epifauna (hydroids, sponges, and small white anemones; Fig. 3D) were attached to Ceriantharia tubes only and not found on the surrounding sub- strate. Also, blackbelly rosefish, Helicolenus dacty- lopterus (De La Roche) (Fig. 3B), and redfish, Sebastes sp. (Fig. 3D), abundances were higher in the Cerianthid C patch (0.40/frame and 0.18/frame, respectively) than in the adjacent area (0.03/frame and 0.00/frame); about half of the fish were nestled at tube bases. At other dive locations, motile megafaunal species often seen nestled near tubes included portunid crabs (Bathynectes sp.) (Fig. 3C); jonah crabs, Cancer sp.; pandalid shrimps, Pandalus sp.; American lobsters, Homarus americanus Milne-Edwards; hakes, Urophycis spp.; and greeneyes Chloropthal- mus agassizii Bonaparte. Two Cerianthid B tubes (50 m apart) and adjacent sediments were collected with the grab sampler of the submersible Johnson-Sea-Link, in the head of Oceanographer Canyon at a depth of 293 m. The tubes were separated from the adjacent sediments immediately after the submersible surfaced. The volume of each tube was less than the volume of ad- jacent sediments (80% fine sand, <0.5 mm; 10% coarse sand; 10% silt) (Appendix Table 3). After preservation and staining, the macrofauna (>0.5 mm) were identified for each sample (Appendix Table 3): Polychaetes were dominant and the three most abundant polychaete species inhabiting the Ceriantharia tubes were absent or scarce in the ad- jacent sediments; Polycirrus eximius (Leily) (a tentacle feeder which sweeps the water and sub- stratum for food), Marphysa sp. (a jawed omnivore), and a filter-feeder, Potamilla neglecta (Sars) (Fauchald 1977; Fauchald and Jumars 1979). DISCUSSION Collection Gear Gear differences largely account for the differ- ences in Ceriantharia size and density estimates from grab samples versus photographs. Due to limitations in resolution, photographs provide valid data only on larger epifauna (Emery et al. 1965; Barham et al. 1967; Wigley and Emery 1967). How- ever, since the estimated depth of penetration of a 0.1 m2 Smith-Mclntyre grab sampler, the gear used most frequently in this study, is only 3 to 5 cm in unconsolidated substrates (Smith and Mclntyre 1954), and large ceriantharian tubes often extend much deeper than 5 cm below the seafloor (Sebens fn. 5), making them difficult to dislodge, if the primary objective is to sample large individuals and document the associations between tubes and other fauna, then photographs and direct observations are more useful than grab samples. Species Identification Ceriantheopsis americanus and Cerianthus borealis, identified from grab samples, occurred within the geographic and bathymetric ranges noted previously for these species (Table 1). Unfortunately, many Ceriantharia samples were discarded, and none of the available samples from depths greater than 500 m contained anemones for taxonomic identification. The morphological features used to distinguish between the four species seen from submersibles (Table 3) may not individually be reliable; tentacle coloration may vary noticeably within a species (Arai 1971; Uchida 1979). However, taken together, we feel the features were consistent enough to indicate we saw four species of adult Ceriantharia: C. borealis (probably Cerianthid B), two unidentified species (Cerianthids C and D) from depths shallower than about 500 m, and another unidentified species (Cerianthid A) living deeper down the continental slope. The conclusion that Cerianthid B is C. borealis is based on the similarities between our descriptions of Cerianthid B morphology and distribution (Table 3), and information from other studies on C. borealis (Table 1; Gosner 1979). The only other previously identified inhabitant of the study area, C. ameri- canus, was probably not encountered on our sub- mersible dives; the deepest record found for C. americanus was about 70 m (Pearce et al. 1981), whereas our shallowest submersible dive was to a depth of 80 m. Sebens (in press) described two unidentified Ceri- antharia species which occur at depths >1,000 m in the Northwest Atlantic: Unidentified Species II (seen at depths >1,500 m) resembles Cerianthid A (Table 3, Fig. 3A), Cerianthid A in Valentine et al. (1980), and a photograph of unidentified Cerian- tharia taken by Grassle et al. (1975) at depths of 1,550 to 1,830 m just south of New England. The distinction Sebens (in press) makes between Uniden- tified Species I (seen at depths of >1,000 m) and Unidentified Species II (Cerianthid A) is that Species II is smaller (Table 1). Grassle et al. (1975) and 636 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA Hecker et al. (1980) also reported seeing small unidentified Ceriantharia at about 1,300 and 1,000 m, respectively. We saw (from submersibles) no Ceriantharia from 1,000 to 1,600 m for comparison. In addition to the six documented species above, other Ceriantharia sighted in or near the region in- clude two possible species photographed by Hecker9: one at depths of 1,800 to 2,800 m (from Lydonia Canyon to Cape Lookout, NC), which resembles a stout black Cerianthid B, and another resembling Cerianthid A (except its tube extends above the seafloor) at depths of 500 to 1,000 m off Cape Hat- teras. Rowe and Menzies (1969) photographed Ceriantharia on the continental slope (at depths of 400 to 3,000 m) south of Cape Hatteras (about lat. 34 °N) which they guessed to be Ceriantheomorphe brasiliensis Carlgren. However, they presented no photographs for comparison and collected no voucher specimens. The C. brasiliensis specimens identified by Carlgren (1931) were from Brazil, South America, and its resemblance to other slope species is uncertain. Submersible dive time devoted to in situ documentation and collection of specimens is obviously needed in order to identify the deep- water species10. Relation to Latitude North of Cape Cod and Georges Bank (lat. 42° to 44 °N) the continental shelf is dominated by the Gulf of Maine, a feature unlike the rest of the shelf in the region because of its topographic irregularity and because it reaches depths of more than 100 m closer to shore. The lack of tidal mixing below 100 m over much of the gulf, and the fact that the prin- cipal source of its bottom water is thermally stable continental slope water introduced through the Northeast Channel, results in water temperature stratification which keeps the gulf bottom water temperatures virtually constant throughout the year (TRIGOM-PARC 1974; Rowe et al. 1975; Ingham et al. 1982, p. 43). The narrow extreme range of bot- tom water temperature (A T) dominant from lat. 42° to 43°N (Fig. 5) may account for low Ceriantharia occurrence at grab sample stations there (Fig. 4), while peaks in occurrence are evident at lat. 40° to 41 °N (shelf just south of Cape Cod, including south- ern Georges Bank), and from 44° to 45°N (shelf off 9B. Hecker, Lamont-Doherty Geological Observatory, Columbia University, Palisades, NY 10964, pers. commun. October 1984. 10For all photographed, but unidentified slope species, we know of only one voucher specimen (of Unidentified Species I), present- ly located at the Harvard Museum of Comparative Zoology, Cam- bridge, MA. Nova Scotia) may be associated with more favorable intermediate temperature ranges which prevail there (8° to 15.9°C). High Ceriantharia occurrence at grab sample stations between 37° to 38 °N is in part due to high occurrence at stations in the lower half of Chesapeake Bay; occurrence was 56% at nine Bay Stations and 23% at 52 shelf/slope stations. However, our data is too sparse and inconclusive to make a bay versus non-bay comparison, or explain the high occurrence at shelf stations in this area. According to Gosner (1971), the continental margin from Cape Hatteras to Nova Scotia is divided into two faunal provinces with respect to benthic invertebrates: a Boreal (cold-temperate) province north of Cape Cod, and a Virginian (warm- temperate) province of Cape Cod, MA. Theroux (in press) considers the situation to be more complex and to depend on the species considered, but agrees that Cape Cod and Georges Bank are the beginning of a rapid transition from cold to warm temperate fauna, and suggests that the transition is associated with Georges Bank and Nantucket Shoals thermal fronts (Fig. 5; Ingham et al. 1982, p. 40-41). Using Gosner's (1971) faunal province descrip- tions, our submersible data indicate that, in addi- tion to C. americanus, at least two other warm- temperate species inhabit the northwest Atlantic continental shelf (Cerianthids C and D). The only cold-temperate shelf species, Cerianthid B (probably C. borealis) ranges south to Cape Hatteras (Tables 1, 3). The last species we saw (Cerianthid A) is bathyal. Relation to Bathymetry Bathymetric zonation of benthic fauna has been previously described for the continental shelf-slope region of the northwest Atlantic (Wigley and Emery 1967; Rowe and Menzies 1969; Sanders and Hessler 1969; Rowe 1972; Grassle et al. 1975; Haedrich et al. 1975, 1980; Hecker et al. 1980; Valentine et al. 1980; Wigley and Theroux 1981). Rowe et al. (1982) cautioned, "'zones' that previous investigations have described apparently are a function both of the animal groups studied and distribution of samples with depth". Thus, our discussion of Ceriantharia zonation is limited to depths <2,000 m, since below that depth there were no submersible data to sup- port the grab sample data. Ceriantharia distribution, as determined from the grab sample data (Fig. 4), our submersible obser- vations (Table 3), and data from other investigations (Table 1) imply boundaries (defined here as depths characterized by distinct changes in the benthic com- 637 FISHERY BULLETIN: VOL. 84, NO. 3 munity's species composition) to Ceriantharia distribution exist at about 500, 900, and 1,600 m. Our submersible data indicate that shelf species were confined to depths of less than about 400 m, and the bathyal species (Cerianthid A) was seen between 1,600 and 2,000 m. Published reports in- dicate another unidentified species lives deeper than about 1,000 m (Grassle et al. 1975; Hecker et al. 1980; Sebens in press). Similar depth zonation of slope fauna inhabiting the study area have been reported for isopods (Menzies et al. 1973), demer- sal fishes (Musick11), and megafauna captured in trawls (Haedrich et al. 1980). Some environmental factors, suggested as causes for observed distribu- tions, are temperature, sedimentation rates, and substrate types (summarized by Haedrich et al. 1975, 1980). The depth interval between about 400 and 600 m on the continental slope south of New England is a temperature transition zone; shallower bottom waters experience larger seasonal temperature variations than stable deeper waters (Sanders and Hessler 1969; Haedrich et al. 1975). Depth-temper- ature profiles (Fig. 6) made on A Ivin dives in Veatch Canyon showed larger depth related temperature variations also occurred shallower than 500 to 600 m. The shelf species (Cerianthids B, C, and D) may not be able to tolerate and/or thrive in the cold stable conditions below 500 m. The Cerianthid A population, we saw deeper than 1,600 m in the axis of Oceanographer Canyon, in- habited sediments high in biogenic carbonates; canyon axes may act as settling basins for suspended matter being funneled downcanyon (Valentine et al. 1980). Rowe and Menzies (1969) attributed increases in suspension-feeder concentration, in photographs from the upper slope (200-800 m) and at the slope base (3,000 m) off North Carolina, to increased detritus accumulation resulting from downslope movement and concentration by the prevailing bot- tom currents. Haedrich et al. (1980) stated, in reference to the depth zonation of megabenthic fauna on the slope off southern New England, that "zonation must result to some degree from vary- ing strategies that promote success along a food resource gradient". Haedrich et al. (1975) suggested boundaries to zones of larger epifauna, at about 400 and 1,000 m nMusick, J. A. 1976. Community structure of fishes on the continental slope and rise off the Middle Atlantic Coast of the U.S. Manuscript presented at Joint Oceanographic Assembly, Edinburgh, September. (Copies available from: J. A. Musick, Virginia Institute of Marine Science, Gloucester Point, VA 23062, USA). on the continental slope south of New England, result from physical changes in the slope environ- ment. Macllvaine (1973, p. 30-70) reported on the physical environment in the same area (sediment type, suspended sediments, and slope gradient). The zone between 400 and 1,000 m consists largely of homogeneous silt-sand substrate, near-bottom sus- pended sediments at 520 m were 50 to 60 \xglh (about 25% organics), and the slope gradient is about 1.4°. Deeper than about 1,000 m there are more variable sediment features (stiff clayey silt sedi- ments which are smooth or hummocky, talus slopes, and rock outcrops), suspended sediments were 20 ^g/L (about 45% organics) at 1,000 m and 80 /ig/L (about 80% organics) at 1,670 m, and the slope gradient is steeper (7.6°). Suspension feeders rely on current velocity and nutrient load for their food supply. Substrate vari- ability deeper than 1,000 m may enhance Cerian- tharia occurrence down to 2,000 m: Features such as hummocks may act as perches for suspension feeders, placing them up higher where current is swifter and their food supply is replenished more rapidly (Hughes 1975; Dyer 1980; Sebens 1984). Higher suspended sediments and percentage of organics may further enhance Ceriantharia occur- rence below 1,600 m, as compared with 1,000 or 520 m. The lesser slope gradient between 400 and 1,000 m probably results in lower near bottom current velocities; near the shelf-slope break in Ocean- ographer Canyon, bottom currents are swifter at 105 to 300 m than at 650 m, due primarily to a dif- ference in slope gradient (Valentine in press). Thus, increased slope gradient may enhance Ceriantharia occurrence below 1,000 m. Other mechanisms may affect ceriantharian depth zonation such as the direct effects of pressure (Siebenallar and Somero 1978), or predators (Paine 1966; Rex 1976); however, data were not available to evaluate these factors. Submarine canyons received particular attention during submersible dive activities because of the potential entrainment of discharges from oil explora- tion activities into productive canyon environments (Cooper and Uzmann fn. 8). Bathymetric zonation of slope fauna may be altered and/or species abun- dance enhanced by submarine canyons (Rowe 1971; Haedrich et al. 1975). The conduitlike nature and substrate heterogeneity of canyons have both been implied as explanations for observed faunal enrich- ment in canyons as opposed to adjacent noncanyon slope areas (Rowe and Menzies 1969; Rowe 1971, 1972; Haedrich et al. 1975; Hecker et al. 1980; Valentine et al. 1980; Rowe et al. 1982). Although 638 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA we had no adjacent slope dives to compare with the canyon dives, Ceriantharia were common in canyons and have been suggested to be canyon "indicator" species (Rowe 1972). In the future, we hope a canyon-slope comparison of Ceriantharia species' diversity and abundance will be made. Relation to Bottom Water Temperature Wigley and Theroux (1981) found that total macro- faunal density in the Middle Atlantic Bight generally increased directly with increasing temperature range (A T). Ceriantharia occurrence at grab sam- ple stations followed this trend until A T reached 15.9°C, after which it decreased (Fig. 4). Why an intermediate temperature range may be favorable to Ceriantharia is unknown. Wide ranges might en- tail harmful extremes of temperature, while nar- rower ones may be too constant at an unfavorable level, or larval stages may benefit from some degree of fluctuation for maximal development ( Andrewar- tha and Birch 1954, p. 129-205). Information on how temperature affects ceriantharian metabolism, activity patterns, and development is lacking. Marine organism distributions are largely con- trolled by temperature (Hutchins 1947; Crisp 1965; Gosner 1971). The most obvious effect of tempera- ture on invertebrate distributions is exclusion of species from areas with unsuitable thermal regimes (Kinne 1970). Submersible data on ceriantharian geographic and bathymetric distribution demon- strate allopatric speciation which we believe is primarily a response to temperature. Relation to Sediments The presence of silt is characteristic of deposi- tional areas which may be favorable to suspension feeders (Rowe and Menzies 1969). Wigley (1968) described Ceriantharia as common inhabitants of silty-sand sediments on Georges Bank. Through resuspension, surficial deposits are potential food for Ceriantharia (Rhoads 1974). In addition to low deposition, substrate instability may account for the scarcity of Ceriantharia in 100% gravel and rippled coarse sand substrate. Shifting substrates, such as the 100% gravel sediments at grab sample stations or the rippled sand dunes observed from submer- sibles, may harm suspension feeders through clogging of feeding apparatus, or the burial of lar- vae (Sanders 1956; Ross 1968; Rhoads and Young 1970; Rhoads 1974). However, Ceriantharia were generally cosmo- politan with respect to substrate (Fig. 4; Appendix Tables 1, 2). They are well adapted to withstand strong currents, sediment movement, and extreme deposition of fine material because their tubes pro- vide firm anchorage (Frey 1970) and protection against clogging or burial (Pearce 1972). Pearce et al. (1976) found Ceriantharia were dominant macro- fauna in fine carbon-rich sediments stressful to other benthic species, near New York Bight sewage sludge disposal sites. Just as 100% gravel substrate is unfavorable for burrowing, a gravel veneer might also be expected to limit space available for burrowing. However, on submersible dives, Ceriantharia were frequently seen in gravel-covered areas (less than about 50% gravel cover). These deposits, probably Pleistocene ice-rafted glacial debris, are exposed in areas which usually experience higher currents than adjacent areas (Valentine et al. 1980; Valentine in press), a favorable consideration for suspension feeders. Spatial Pattern Local conditions of food supply, substrate, or micro topography, may enhance Ceriantharia aggre- gation (Fig. 7). Local differences in food supply may allow Ceriantharia to survive in aggregations. Grassle et al. (1975) observed that strongly clumped suspension-feeders were able to maintain aggrega- tions because their food supply was continually renewed. Unusually high Ceriantharia abundances near a sewage sludge/dredge spoil disposal area may have occurred owing to the increased amounts of organic matter (Pearce et al. 1976). Grassle et al. (1975) found Ceriantharia, similar to Cerianthid A, more randomly distributed on the continental slope, south of Cape Cod (depth of 1,465 to 1,830 m, homogeneous sandy silt-clay substrate). In comparison, substrata in canyon heads where aggregations were observed from submersibles are heterogeneous (Hecker et al. 1980; Valentine et al. 1980). Our grab samples showed the same contrast between heterogeneous substrata shallower than 500 m and homogeneous silt-sands and clays down- slope (Shepard and Theroux fn. 4). Since inver- tebrates are capable of substrate selectivity (Thor- son 1966; Gray 1974), a variable substrate may be characterized by patchy inhabitant distributions (Hecker et al. 1980). The Cerianthid B aggregation in Lydonia Canyon (Fig. 7), located on a knoll, may benefit from elevated positioning and swifter currents (Hughes 1975; Sebens 1984), thus aggregations may also form in response to local changes in surface elevation. 639 FISHERY BULLETIN: VOL. 84, NO. 3 Functional Role An increase in structural complexity of the sub- strate vertically and/or horizontally increases the number of microhabitats, and if the appropriate colonizers and mortality sources are present, within- habitat diversity will likely be increased (Steimle and Stone 1973; Abele 1974; Hughes 1975; Woodin 1976, 1978; Connell 1978; Suchanek 1979; Hulbert et al. 1982). Ceriantharia tubes may increase species diversity and abundance on featureless soft bottom areas by 1) attracting motile megafauna seeking refuge near tubes and 2) serving as a favorable substrate for epifauna and infauna, particularly suspension-feeders and tubiculous species. By acting as a three-dimensional refuge, the tubes may ease predation pressure on smaller motile species (Ware 1972; Whoriskey 1983). Demersal fish and crustaceans similar to those we observed have been noted by others in association with Cerian- tharia (Uzmann et al. 1977; Hecker et al. 1980; Valentine et al. 1980). The species most commonly observed near tubes, Helicolenus dactylopterus, Sebastes sp., and Bathynectes sp., characteristical- ly exhibit thigmotactic behavior. Associations similar to the ones we found between suspension feeders and Ceriantharia tubes in Block Canyon (Figs. 3D, 8), and polychaetes and tubes from Oceanographer Canyon (Appendix Table 3), have been recorded for Ceriantharia and polychaetes (Kingsley 1904; O'Connor et al. 1977), phoronids (Ponder 1971; Emig et al. 1972; Hartog 1977), and bivalves (Ponder 1971). These associations have been alternately referred to as commensalism or in- quilinism; we prefer the latter definition as it high- lights the role of the ceriantharian tube. Emig et al. (1972) speculated that Cerianthus maua Carlgren tentacles may act as baffles, causing waterborne food particles to settle out, and become available to suspension feeders (Phoronis australis Haswell) in- habiting the C. maua tubes, in which case the term commensalism may be more appropriate. However, Emig et al. also stated that increased food supply is probably a secondary benefit to the phoronids and that the suitability of the tube as a settlement sur- face for larvae motivates the association. O'Connor et al. (1977) studied a Pachycerianthus multiplicatus Carlgren population inhabiting deposit substrates (85% silt-clay, 15% sand) off Ireland and suggested tubes were prime settlement surface for the larvae of inquiline filter-feeding polychaetes, Myxicola in- fundibulum (Renier). The associates (sponges, hydroids, and colonial anemones) of Ceriantharia tubes in Block Canyon are generally nonmotile so they probably had to arrive on the tubes as larvae. More unstable substrate surrounding the tubes may be less suitable as a settlement surface for larvae of suspension feeders (Rhoads and Young 1970, 1971; Rhoads 1974). The vertical aspect of Ceriantharia tubes may enhance diversity and abundance by 1) allowing ver- tical stratification of trophic types (MacArthur and Levins 1964; Hughes 1975; Schoener 1975; Ausich and Bottjer 1982), and 2) affording inhabitants, such as the filter feeder Potamilla neglecta, elevated feed- ing stations where clogging by resuspended sedi- ments is less likely, and current velocities tend to be greater (Dyer 1980), thus the food supply is more rapidly renewed (Hughes 1975; Sebens 1984). The stable nature of the tubes may serve species behaviorally inclined to attach themselves to firm substrate. The three species of polychaetes, Poly- cirrus eximius, Marphysa sp., and Potamilla neglec- ta, most abundant on ceriantharian tubes caught in Oceanographer Canyon, but rarely found in the ad- jacent sediments (Appendix Table 3), usually attach their tubes to solid surfaces such as stones, algae, or hydroids (Gosner 1971; Fauchald and Jumars 1979). Infaunal species may also gain relief from preda- tion pressure by inhabiting ceriantharian tubes. The feltlike tubes are generally more consolidated that the sediments surrounding them, thus more difficult to graze. Ponder (1971) viewed protection as the principal benefit to a leptonid bivalve, Montacutona ceriantha Ponder, inquiline with Cerianthus sp. in Japanese waters. Protection may be enhanced for tubiculous infauna since their retraction may be stimulated by a similar response to disturbance by the host ceriantharian (Emig et al. 1972). Ceriantharia tubes may serve as a preferential food source for some infauna. O'Connor et al. (1977) noted sipunculids, Golfingia elongata (Keferstein), inquiline with Pachycerianthus multiplicatus had tube remains in their guts. Scavengers, such as Mar- physa sanguinea may benefit from the inquilinism for this reason. Ceriantharia may also negatively affect the in- fauna in sediments adjacent to the tubes; large motile species, attracted to the tubes for shelter, might selectively graze near tubes. We hope to in- vestigate Ceriantharia "forest" communities more thoroughly on future submersible cruises: Substrate collections taken away from tubes will further define their functional role. We believe Ceriantharia influ- ence the ecology of the northwest Atlantic contin- ental shelf and slope more than has been revealed from data collected by conventional surface tech- 640 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA niques alone; methods inadequate for collecting deep-burrowing adults, and providing information on behavioral and spatial relationships between Ceriantharia and other community residents. ACKNOWLEDGMENTS Funding for submersible time was provided by NOAA's Office of Undersea Research, Washington, D.C. Ann Frame, NMFS Sandy Hook Laboratory, NJ, identified invertebrate specimens. We thank Kenneth Sebens and J. C. den Hartog for critically reviewing the manuscript. Alan Hulbert and Michael Pennington provided advice on data analysis. Han- nah Goodale, Jean Klemm, and Connie Fontaine typed the various drafts of the manuscript. Special thanks to the submersible and ship crews who made the data collection possible: R/V Johnson (JSL), R/V Atlantic Twin (Nekton), and R/V Lulu (Alvin). We dedicate this effort to the memory of John Lamont, whose talents, patience, and humor made our daily burdens easier to bear. LITERATURE CITED Abele, L. G. 1974. Species diversity of decapod crustaceans in marine habitats. Ecology 55:156-161. Agassiz, A. 1859. On some new Actinoid polyps of the coast of the United States. Proc. Boston Soc. Nat. Hist. 7:23-24. Andrewartha, H. G., and L. C. Birch. 1954. The distribution and abundance of animals. Univ. Chicago Press, Chicago, IL, 782 p. Arai, M. N. 1971. Paehycerianthus (Ceriantharia) from British Columbia and Washington. J. Fish. Res. Board Can. 28:1677- 1680. AUSICH, W. I., AND D. J. BOTTJER. 1982. Tiering in suspension-feeding communities on soft substrata throughout the Phanerozoic. Science 216:173- 174. Barham, E. G., N. J. Ayer, Jr., and R. E. Boyce. 1967. Macrobenthos of the San Diego Trough: photographic census and observations from bathyscaphe Trieste. Deep- Sea Res. 14:773-784. Bland, E. L., J. D. Donnelly, and L. A. Shumaker. 1976. Alvin user's manual. Woods Hole Oceanogr. Inst., Tech. Memo. No. 3-76, 36 p. Caracciolo, J. V. and F. W. Steimle. 1983. An atlas of the distribution and abundance of dominant benthic invertebrates in the New York Bight Apex with reviews of their life histories. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-766, 58 p. Carlgren, 0. 1912. Ceriantharia. Danish-Ingolf Exped. 5a(3):l-78. 1931. On some Ceriantharia. Ark. Zool. 23A(2):1-10. Cerfontaine, P. 1909. Contribution a l'etude des Cerianthides, nouvelles recherches sur le Cerianthus oligopodus. Arch. Biol. 24: 653-707. Connell, J. H. 1978. Diversity in tropical rain forests and coral reefs. Sci- ence 199:1302-1310. Cooper, R. A., P. C. Valentine, J. R. Uzmann, and B. Butman. In press. Georges Bank Submarine Canyons. In R. Backus (editor), Georges Bank, Chap. 10. MIT Press, Cambridge, MA. Crisp, D. J. 1965. Observations on the effects of climate and weather on marine communities. In C. G. Johnson and L. P. Smith (editors), The biological significance of climatic changes in Britain, p. 63-77. Acad. Press. N.Y. Dyer, K. R. 1980. Velocity profiles over a rippled bed and the threshold of movement of sand. Estuarine Coast. Mar. Sci. 10:181- 199. Emery, K. O., A. S. Merrill, and J. V. A. Trumbull. 1965. Geology and biology of the sea floor as deduced from simultaneous photographs and samples. Limnol. Oceanogr. 10:1-21. Emig, C. C, C. Herberts, and B. A. Thomassin. 1972. Sur l'association de Phoronis australis (Phoronida) avec Cerianthus maua (Ceriantharia) dans les zones recifales de Madagascar. Mar. Biol. (Berl.) 15:304-315. Fauchald, K. 1977. The polychaete worms: definitions and keys to the Orders, Families, and Genera. Nat. Hist. Mus. Los. Ang. Cty., Sci. Ser. No. 28, 188 p. Fauchald, K., and P. A. Jumars. 1979. The diet of worms: A study of polychaete feeding guilds. Oceanogr. Mar. Biol. Annu. Rev. 17:193-284. Field, L. R. 1949. Sea anemones and corals of Beaufort, North Carolina. Duke Univ. Mar. Stn. Bull. 5, 39 p. Frey, R. W. 1970. The lebensspuren of some common marine inverte- brates near Beaufort, N.C. II. Anemone burrows. J. Paleontol. 44:308-311. Gosner, K. L. 1971. Guide to identification of marine and estuarine inverte- brates. Wiley-Interscience, N.Y., 693 p. 1979. A field guide to the Atlantic seashore, from Bay of Fundy to Cape Hatteras. Houghton-Mifflin Co., Boston, MA, 329 p. Grassle, J. F., H. L. Sanders, R. R. Hessler, G. T. Rowe, and T. McLellan. 1975. Pattern and zonation: a study of the bathyal megafauna using the research submersible Alvin. Deep-Sea Res. 22: 457-481. Gray, J. S. 1974. Animal-sediment relationships. Oceanogr. Mar. Biol. Annu. Rev. 12:223-262. Haedrich, R. L., G. T. Rowe, and P. T. Polloni. 1975. Zonation and faunal composition of epibenthic popula- tions on the continental slope south of New England. J. Mar. Res. 33:191-211. 1980. The megabenthic fauna in the deep sea south of New England, USA. Mar. Biol. (Berl.) 57:165-179. Hartog, J. C. den. 1977. Description of two new Ceriantharia from the Carib- bean region, Paehycerianthus curacaoensis N. SP. and Arachnanthus nocturnus N. SP., with a discussion of the cnidom and of the classification of the Ceriantharia. Zool. Meded. (Leiden) 51(14):211-248. 641 FISHERY BULLETIN: VOL. 84, NO. 3 Hecker, B., G. Blechschmidt, and P. Gibson. 1980. Canyon assessment study in the Mid and North Atlan- tic areas of the U.S. Outer Continental Shelf: epifaunal zona- tion and community structure in three mid and north Atlantic canyons. Final Rep., U.S. Dep. Inter., Bur. Land Manage., Wash., DC, 139 p. Hughes, R. G. 1975. The distribution of epizoites on the hydroid Nerner- tesia antennina (L.). J. Mar. Biol. Assoc. U.K. 55:275- 294. Hulbert, A. W., K. J. Pecci, J. D. Witman, L. G. Harris, J. R. Sears, and R. A Cooper. 1982. Ecosystem definition and community structure of the macrobenthos of the NEMP monitoring station at Pigeon Hill in the Gulf of Maine. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NEC-14, 143 p. Hutchins, L. M. 1947. The bases for temperature zonation in geographical distribution. Ecol. Monogr. 17:325-335. Hyman, L. H. 1940. The invertebrates, Protozoa through Ctenophora. McGraw-Hill, N.Y., 726 p. Ingham, M. C. (editor). 1982. Summary of the physical oceanographic processes and features pertinent to pollution distribution in the coastal and offshore waters of the northeastern United States, Virginia to Maine. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NEC-17, 166 p. KlNGSLEY, J. S. 1904. A description of Cerianthus borealis. Tufts Coll. Stud. 8:345-361. Kinne, 0. 1970. Temperature. In O. Kinne (editor), Marine ecology, Vol. 1., p. 321-616. John Wiley and Sons, N.Y. MacArthur, J. W. 1975. Environmental fluctuations and species diversity. In M. L. Cody and J. M. Diamond (editors), Ecology and evolu- tion of communities, p. 74-80. Belknap Press of Harvard Univ. Press, Cambridge, MA. MacArthur, R., and R. Levins. 1964. Competition, habitat selection, and character displace- ment in a patchy environment. Proc. Natl. Acad. Sci. USA 51:1207-1210. MacIlvaine, J. C. 1973. Sedimentary processes on the continental slope off New England. Woods Hole Oceanogr. Inst. Tech. Rep. 73-58, 211 p. MARISCAL, R. N., E. J. CONKLIN, AND C. H. BlGGER. 1977. The ptychocyst: a major new category of cnida used in tube construction by a cerianthid anemone. Biol. Bull. 152:392-405. McMURRICH, J. P. 1890. Contributions on the morphology of the Actinozoa. I. The structure of Cerianthus americanus. J. Morphol. 4(2): 131-152. Menzies, R. J., R. Y. George, and G. T. Rowe. 1973. Abyssal environment and ecology of the world oceans. John Wiley and Sons, Inc., N.Y., 488 p. Meyer, D. L. 1980. Ecology and biogeography of living classes. In T. W. Broadhead and J. A. Waters (editors), Echinoderms: notes for a short course, p. 1-14. Stud. Geol. 3, Dep. Geol. Sci., Univ. Tenn. Miner, R. W. 1950. Field book of seashore life. G. P. Putnam's Sons, N.Y., 888 p. O'Connor, B., G. Konnecker, D. McGrath, and B. F. Keegan. 1977. Pachycerianthus multiplicatus Carlgren - biotope or biocoenosis? In B. F. Keegan, P.O. Ceidigh, and P. J. S. Boaden (editors), Biology of benthic organisms, p. 475-482. Pergamon Press, Lond. Paine, R. T. 1966. Food web complexity and species diversity. Am. Nat. 100:65-75. Parker, G. H. 1900. Synopses of N. American invertebrates. Am. Nat. 34: 747-758. Pearce, J. B. 1972. The effects of solid waste disposal on benthic com- munities in the New York Bight. In M. Ruivo (editor), Marine pollution and sea life, p. 404-411. Fishing News (Books) Ltd., Lond. Pearce, J. B., J. V. Caracciolo, M. B. Halsey, and L. H. Rogers. 1976. Temporal and spatial distributions of benthic macro- invertebrates in the New York Bight. Am. Soc. Limnol. Oceanogr. Spec. Symp. 2:394-403. Pearce, J. B., D. J. Radosh, J. V. Caracciolo, and F. W. Steimle. 1981. Benthic fauna. MESA N.Y. Bight Atlas Monogr. 14, 79 p. Ponder, W. F. 1971. Montacutona ceriantha n. sp., a commensal leptonid bivalve living with Cerianthus. J. Conchyliol. 109:15-25. Reid, R., D. Radosh, A. Frame, F. Steimle, J. Ward, and C. Mackenzie. 1981. Coastal ecosystems investigations, 1980. U.S. Dep. Commer., NOAA Northeast Monit. Program Rep. No. NEMP-111-80-G-001, 50 p. Rex, M. A. 1976. Biological accommodation in the deep-sea benthos: comparative evidence on the importance of predation and productivity. Deep-Sea Res. 23:975-987. Rhoads, D. C. 1974. Organism-sediment relations on the muddy sea floor. Oceanogr. Mar. Biol. Annu. Rev. 12:263-300. Rhoads, D. C, and D. K. Young. 1970. The influence of deposit-feeding organisms on sediment stability and community trophic structure. J. Mar. Res. 28:150-178. 1971. Animal-sediment relations in Cape Cod Bay, Massa- chusetts. II. Reworking by Molpadia oolitica (Holothu- roidea). Mar. Biol. (Berl.) 11:255-261. Richmond, S. B. 1964. Statistical analysis. Ronald Press Co., N.Y., 633 p. Robson, E. A. 1966. Swimming in Actiniaria. In W. J. Rees (editor), The Cnidaria and their evolution. Symp. of the Zool. Soc. Lond. 16:334-359. Rogers, C. A. 1979. Siphonophore ("Lipo") swarming in New England coastal waters— update 1976. In J. R. Goulet, Jr. and E. D. Haynes (editors), Ocean variability in the U.S. Fishery Conservation Zone, 1976, p. 349-352. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ. 472. Ross, D. A. 1968. Current action in a submarine canyon. Nature 218: 1242-1245. Rowe, G. T. 1971. Observations on bottom currents and epibenthic populations in Hatteras Submarine Canyon. Deep-Sea Res. 18:569-581. 642 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA 1972. The exploration of submarine canyons and their ben- thic faunal assemblages. Proc. R. Soc. Edinb. (B) 73(17): 159-169. Rowe, G. T., and R. J. Menzies. 1969. Zonation of large benthic invertebrates in the deep-sea off the Carolinas. Deep-Sea Res. 16:531-537. Rowe, G. T., P. T. Polloni, and R. L. Haedrich. 1975. Quantitative biological assessment of the benthic fauna in deep basins of the Gulf of Maine. J. Fish. Res. Board Can. 32:1805-1812. 1982. The deep-sea macrobenthos on the continental margin of the Northwest Atlantic Ocean. Deep-Sea Res. 29:257- 278. Sanders, H. L. 1956. The biology of marine bottom communities. In G. A. Riley et al. (editors), Oceanography of Long Island Sound, 1952-1954. Bull. Bingham Oceanogr. Collect., 15:345-414. Sanders, H. L., and R. R. Hessler. 1969. Ecology of the deep-sea benthos. Science 163:1419- 1424. Schoener, T. W. 1975. Resource partitioning in ecological communities. In W. E. Hazen (editor), Readings in population and commun- ity ecology, p. 370-390. W. B. Saunders Co., Phila., PA. Sebens, K. P. 1984. Water flow and coral colony size: interhabitat com- parisons of the octocoral Alcyonium siderium. Proc. Natl. Acad. Sci. 81:5473-5477. In press. Marine flora and fauna of the northeastern United States, Anthozoa: Actiniaria, Zoanthidea, Corallimorpharia, and Ceriantharia. U.S. Dep. Commer. NOAA Tech. Rep. NMFS. SlEBENALLER, J., AND G. N. SOMERO. 1978. Pressure-adaptive differences in lactate dehydroge- nases of congeneric fishes living at different depths. Science 201:255-257. Smith, W., and A. D. McIntyre. 1954. A spring-loaded bottom sampler. J. Mar. Biol. Assoc. U.K. 33:257-264. Steimle, F. W., Jr., and R. B. Stone. 1973. Abundance and distribution of inshore benthic fauna off southwestern Long Island, N.Y. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-673, 50 p. Suchanek, T. H. 1979. The Mytilus californianus community: studies on the composition, structure, organization, and dynamics of a mussel bed. Ph.D. Thesis, Univ. Washington, Seattle, 300 P- Sutherland, J. P. 1978. Functional roles of Schizoporella and Styela in the foul- ing community of Beaufort, North Carolina. Ecology 59: 257-264. Theroux, R. B. In press. Bottom animals. In R. Backus (editor), Georges Bank, Chap. 6.2. MIT Press, Cambridge, MA. Thorson, G. 1966. Some factors influencing the recruitment and establish- ment of marine benthic communities. Neth. J. Sea Res. 3:267-293. TlFFON, Y., AND M. DAIREAUX. 1974. Phagocytose et pinocytose par ectoderme et l'en- doderme de Cerianthus Iloydi Gosse. J. Exp. Mar. Biol. Ecol. 16:155-165. Trigom-Parc (The Research Institute of the Gulf of Maine- Public Affairs Research Center). 1974. A Socio-economic and environmental inventory of the North Atlantic region, Vol. I., Book Three. BLM Contract 08550-CT3-8. Res. Inst. Gulf of Maine, South Portland, ME, USA. UCHIDA, H. 1979. Cerianthid (Anthozoa, Coelenterata) from Kii Re- gion, Middle Japan. Mem. Nat. Sci. Mus., Tokyo 12:185- 199. Uzmann, J. R., R. A. Cooper, R. B. Theroux, and R. L. Wigley. 1977. Synoptic comparison of three sampling techniques for estimating abundance and distribution of selected mega- fauna: submersible vs. camera sled vs. otter trawl. Mar. Fish. Rev. 39(2): 11-19. Valentine, P. C. In press. Sediment texture and dynamics at outer shelf and upper slope depths on the southern flank of Georges Banks. In Undersea research and technology - scientific applications and future needs, Vol. 2. Symposia Series for Undersea Research, NOAA, Undersea Research Program, Univ. Con- necticut, Groton, CT 06340. Valentine, P. C, J. R. Uzmann, and R. A. Cooper. 1980. Geology and biology of Oceanographer Canyon. Mar. Geol. 38:283-312. Verrill, A. E. 1864. List of polyps and corals sent by the Museum of Com- parative Zoology to other institutions in exchange, with annotations. Mus. Comp. Zool 13:29-60. 1873. Dredging on the coast of New England. Am. J. Sci. 5:1-15. Ware, D. M. 1972. Predation by rainbow trout, Salmo gairdneri: the in- fluence of hunger, prey density, and prey size. J. Fish. Res. Board Can. 29:1193-1201. Whoriskey, F. G., Jr. 1983. Intertidal feeding and refuging by cunners, Tautogola- brus adspersus (Labridae). Fish. Bull., U.S. 81:426- 428. WlDERSTEN, B. 1976. Ceriantharia, zoanthidea, corallimorpharia, and acti- niaria from the continental shelf and slope off the eastern coast of the United States. Fish. Bull., U.S. 74:857-878. Wigley, R. L. 1968. Benthic invertebrates of the New England fishing banks. Underwater Nat. 5(1):8-13. Wigley, R. L., and K. O. Emery. 1967. Benthic animals, particularly Hyalinoecia (Annelida) and Ophiomusium (Echinodermata), in sea-bottom photo- graphs from the continental slope. In J. B. Hersey (editor), Deep-sea photography, p. 235-249. John Hopkins Oceanogr. Stud. 3. Wigley, R. L., and R. B. Theroux. 1981. Atlantic Continental Shelf and Slope of the United States - macrobenthic invertebrate fauna of the Middle Atlantic Bight Region - faunal composition and quantitative distribution. U.S. Dep. Inter., Geol. Survey Prof. Pap. 529-N, 198 p. WOBBER, D. R. 1970. A report on the feeding of Dendronotus iris on the an- thozoan Cerianthus sp. from Monterey Bay, California. Veliger 12:383-387. Woodin, S. A. 1976. The importance of structural heterogeneity in a marine infaunal system. In J. D. Costlow (editor), The ecology of fouling communities, p. 207-208. Duke Univ., Beaufort, NC. 1978. Refuges, disturbance, and community structure: a marine soft-bottom example. Ecology 59:274-284. 643 FISHERY BULLETIN: VOL. 84, NO. 3 X) CO *~ 2 • - To CO xi

, ii J3 o CO _L i- CD .. > CD CO CO OS ■o n 55 OT w >; „ « co « § I it; co c co 75 b* CO a> & To >. »3 CO CO o X) 3 II W CO -r^ S T3 CO Q) CO o -Q CD 3>s 8 " CD O CD ~0 5 co CO co « 5= 9> J2 .2 o | 2 75 » &? >> o c jr o co o 2 x: o> D> g £ n > to "O Q) m -Q - X3 O o CO CM - 9> £ o CO O o >, S ii CO O '■= O II CM 55 c !S co E £ CO co w- co r c co V- a> O T> a> co (0 — n cd > •D CD r *^ c o T> a> T) i- o o CD k- (0 CD i— "J cn T3 11 CD Q. )— h C m CD •o TO C 73 CO c CO CD CO CD CO CJ ^ CD CO Q. XI CO -3 CO 3 — X) m c CO a> \_ V cd CO c CO c CO CO 1— JZ It) O c CO 3 o r 1— CD O S II CO + > "D CO CD O a. CO C o CD CD CD ■o CO > (11 h_ c CD o 5 T3 01 CD T3 CO O o r CD CO L— CO CD CD o U CJ CD Q. CO c g II c a CD h_ 10 o o c O CD rn CO TJ < CD UJ c — 0) (0 CO a> X) >_ "J CO CO II CD o a. F co cn CD U u CD c n CO CO 1 CO c m .55 5 O CD o ~ Q. co To i_ to X) 3 CO EO CD o CO cd ^ k_ oio CO c o jc CO T- c co sz x k_ CD O-c CD t U T3 w x: o~ a CD° O)^ CO X CD CD > Q E, CO CD "D c o O c al CD ^ "cdE > .*: 1— o X CO "O Year/submersible Lat. Long. tra dive(s)1 Geographic name N W ( m + + + + +o + + + + + +m+mmmmmmmmmmm+coocoajajoo+ +o tj- i- i- CM ■* CO t^ soids 00 r«T-i-r O O) OOCM^Jo^^J^ ^ : ,J CDCMCOCOCJ5CMCOo6 ^- !•>■ t- ■>* 00 CD CM •^- CO ^ CN t-' d O) 1^ T- 1- 00 CD 00 CM CM i- O CM ' • O CM CM <* i— y— CJ> «— y— t- t- CO d co o cm co ^ co 1^ in T^ T^ CM T^ T^ CM CM O CM CO cm in d> co r^ in oo CO CO Ol CO o "* p ^ *~. CM CO t^ cvi t- co i co co oS CM r^ CM r^ t^ CD CT> h d cm h Olr-r-O) CMt^T^^^CDOTCMCOCJ5dinCMCMTJ-CMd'*o6 dddddindr^dooodi^cjJoJoJi^d'i-t^ o in •* co o in o *6oi r~- °?o ^o>qr^r*-o>h~ CM CMCO CMCMCOCMCMCMCM co co^i^^i^oqin4cocq'4'4coin ^^r-'cvicxiw^r-'cjCXir-VwCvi^ r^ co I °H °N ■■a- oo CM c\i r^ I I o co ai | TT CM in •** 1 ^1- CO CM t- co CM i- (oinosr-^Nw CM CMCOCMCMCMCMCM sinoindiscocDN t-'t^CMi^t-:'-:'JtJ'i-: ooqo^ooooo)oc«Nooon in CO COCOCOCMCOCOCOCOCMCOCMCM'cOCO'd CM (jSdioinT^cMinin^^r^incopincocqoqr^- ^ ^t ai m I °P ^ (Or-r-r-ION (0^600)4 CM E oq coooooo qnwNnqqNWioqcpqqqwiqwN^^qo^r-cococo cq t^ ^cm'cmcmcmcm r-V^o^riri^w^wwri^^o^'w^^N^cxicxiw^^^ co c^-c^-c^- CM ■* co r^ 00 CO S CO CO CO N in co co co co co co o o o o o o o O 01 01 U> G> O) o> CO CO CO CO CO CO CO r-.incMcomooina>r^ooin-<--*ococM--'>-'>-'-0000000000)CJ)0> •^■'^■^•^■■^^••^••^■•^•^••^•^TfTf^t^-Tf^-^-^^J-^-Tl-'Ja-rJ-COCOCO OIi-NWCOt- m m 'cf m in o Ol Ol Ol Ol oi o co co co co co ^~ CO CO CQ c o CO c Veatch Canyon Veatch Canyon Veatch Canyon Veatch Canyon Veatch Canyon Veatch Canyon CCCCCCC c ccccooooooo o ^^^^■COCOCOCOCOCOCO CO co co co co OO OO O O O . O o |- j- j- irx:x:x:x:x:x:x: S-S-S-^^^^^^^o o Jo 9PO ££ uvu 7i7i"ci7iQ.Q.7ic?c c c 5 Scccc >-?a.£££ a 0.0.0.0. S 2 2 2 S S 2 « 1 « « o « « oo o « ffl 2 « * « O 52««OTc3SraoSrao)oS0 000000000oo E3SOOO £0)010)010000000^^^^. CO «. CO. co ««o t3oooo5SEEE££cococococ:ccccc5;5;Et5t3t5 ^CC^^COCOCOCOCOCOCOSininioOOOOOoCDCDCOiiiii! C0T0-O"OT3CD(JJCDCDCDCDCD"l2v2l2-5-S-S-S-Q-SXIX)CDe0e0C0 CD>>>'>.>>CJCJOOOOCJOOOO>,>,>,>,>,>s=:r=O0JCDCD > X X IIOOOOOOOOOOOjjjjjjOOO> >> CM ^Ni-NCO'f o CO Ol O CM CO 4 '-'-»- cncONCOOlOCO'ClOlOSCOCDOT-WCO^lOlOSCO t-t-T-'r-^CMCMCMCMCMCMCMCMCOCOCOCOCOCOCOCOCO c c c c c c O O O O O O >. ^ >. >. ^ >< c c c c c c CO CO CO CO CO CO O O O OO o x: x: x: x: x: x: o o o o o o CO CO CO CO CO CO CD CD CD CD CD CD >>>>>> N- OO Ol O t- CM CO CO CO N S N CO CO CO CO CO CO -c ro 0) N cn 01 C) Q. r s 01 o S o> CO Ol 644 SHEPARD ET AL.: ECOLOGY OF CERIANTHARIA ID 3 CO CO h» Q.^ d °2 d EO V c~ c~ «? V CD o i- t>» co s cos (0(1)^ i_ CDO s CO c o o °> cd 1 °V T o .C CO T- —' I— CO CO i-c x LO 1 CO CO •* a> §" E O-a — £ o ' = x CD CO ^ CO oj "o § Is J 5 - ■ w 0 E CO c o !c Q. CO w D) O CD O 0) z~. F w n CD 3 > in ■D 01 + o + < + o CO CO CO CO "cc CM £)" CO CO CO CO CO CO t- CO CO t- t- CO cd s 9 cm * cd - CM l- CM r- CNJ CM O O O O O O C C C C C C O O O O O O c c c c c c CO co co CO CO CO oooooo CD CD CD CD CD CD .C .C .C .C .C .C aaaaaa co co co co co co k_ i_ &_ i_ k_ k_ CD CD O) O) O) CD OOOOOO C C C C C C CO CO CO CO CO CO CD CD CD CD CD CD OOOOOO OOOOOO incDScocno CO CO CO CO CO ^ GO CO CO CO CO CO CO cd m E E a CD c I cu 5 CD o I m o 0- n n a ffl £ CD E n £ 8 5 >. c ' O CD O CO o a o o" co" < CD w -Q CO 1 - CO 2 CO CD a. *P oS E 2 g LO " 1 ■I a » -i Q. £ - > CO CD II CO ■g CO o> o 7. II CD CD _ = co ■£> O CO CO «D .. | I V > to ~ 2 CO x> T- CD II «> co iS > g co S - co CD O CD Q. co T3 CO CO .E CO "5 -0 O CO I* » 5 -Q co c» ,5 •- C CO CD ■c CO 3 . CO ■O ^- J3 CO I >< CD O ■ ||« " ■C ^ ^j Q. O e = § -o E^ 3 = 7 ° w o ^ I O M yW « CD E CO" < CO ® s*s O CD Q LU c Q. C CO O CD M co CD £ | co t CO CD O II = CO CO -c C CO co :: ^ CD c O ■g II g E I * » m CD ffl > « ■o E o 2 . w c" co CD CD CD •.= CO (S CD C .2 CO" CO a> * 3 CO o - II to a co nj < O CM CD *^ o LO CO c o +| Q E CO CD O CD Q. CO CD cn to 3 CO EO CD o ca to ^ w 0)0 CO c o n co t- '*- O- c- CD CD E O-o — i= o cd SK.-"- CD CO * CD _ a> 111 o CD c CD CO CO i— CD CO n Q. c -i CO o w i— _j o. CO I— o> o CD (3 CO cu "co" E^ -Q > CO CO o CD ~ > CM CO O) ^> i- t- CO I CM O O O i- 00 8 | | d o • +1 +1 +1 +1 +1 LO N- CM CO O O co CO o •<* 5 CO o O O o * o o * o o i CO s o o +1 +1 CM h- O LO l- CO CM CM 00 CO y- O O O O odd +1 +1 +1 N- O O) h- CO CM CO O O CO CD CO m m o Q m" m CO OO CO CO CO cd" qq CD CD CD CD CD CD CM CO CM o co 1— 1— & CO CO CO -O JD CO CO -Q T- T~ r- m CO CO CO CO CO CO CO CO CO -Q T- CO CO CO CO CO CO co ca co r- CO CO co CO CO CO CO t- i- CO i- cd T— CO CO CO CO CO ^— t- CO CO CO CO CM k oS co 1 o CO V I 1 CM 9.5 11.9 11.9 10.8 9.9 10.3 10.4 O CO LO uo o 1 ■A 1 1 ■-0)(D01* •* LO LO LO 1- a> O) LO O) oioiot- CO CO G) O) O) O) O CO CO O CO CO O) ^■' CO CO CM A A CM r-LoN. cos4 mo CMCMCM r^rrrri CO CM pJcot-cm coco/f qq oj^cm T^-'*cpcr> cm't^cmcoco cot^-co ^-Aobr^- cor^-'i-LocD ■Aco-A I^CMt^CM t^i-^^t^t-: cmi-^cm •* O) CD CO LO LO '* co 1 °V V t- CM i- CO h~ i- 05 00 1-"- CM CJ) CO CO N ^ i- CM CO O O T- T- O O i- t- r- CM O CD CO rj CD 00 LO CM <5 d i-1 t1 N CO LO CO •A CM •A Tf CD CO CM 4 4 O OOLOO CMOOCOOOOO st iot- s i- inios t-_ oq co_ q io t s co co *f (O LO" LO" OOLO CM O Ol CO LOO»-LOOCO SCOO CDCOLOCO CO'-LOLOLOCD 0_ LO O CD CM_ S T- N r-^ LO CO T-_ LO_ r^ od d lo" in h- ,_ cn o CM in o h» CD 00 in CM CO 00 in CO CM CD CD CO CO CO CO T— O CM CM CM 1 — o O CM ■* ^r •* CM ■* •* *f O O o o o o o n o o o o o o o o 0 o o o o 0 o 0 01 m 09 CD cn an T— T— T— cn 00 00 r-- N N K N s S N CO CO CD CD ID ID CD CD N S N CD CD CD CD CD CD CD CD CD CD CD CD CD ■^ CD CO LO LO LO LO LO CD CD CD CD CO CO CO CO CD 00 CM O 00 CM CM O CO CO O O O CD CD ■* •* > >> CD CD .c s: a. o. co co k_ 1_ O) O) o o c c CO CO CD CD O O OO c c O c c >< ca co c OO c c c O O O c c c CO CO CO OOO CD CD CD £££ ca ca co c c o o co co -o n 3 3 1 1 O) oi o> OOO c c co co CD CD O O OOO c c ca co co co CO CO CD CD a> oi w- k_ O O CD CD OO c c o o c c ca ca OO eg ca c c o o CO N- CM CO O CO Tt CO CO CM CO CM oooooo OOOOOO ■* ^- -* ■* ■* •* c c o o >. >> c c ca ca OO c c co ca m co CO CO CD CD O) O) O O CD CD OO c c O O >. >. c c co ca OO CO CO c c o o ■a -a >< >. o CD r £- n O co CO k_ k- a) a o o r c CO 01 CI) CD o <) OO •^- in co CO CD ^ CO CO CO ssr-s "cm^ooo lO lO LO lO i- CM CO r- r- T-T-i- ■. .c to -Q Q. d) ■o 2 £ 2 O) „ o o to = CO O 03 O o o o . o m T> C CO inn = CD l- n t\i td «i o m •t: co g £ "O C CL C E 03 CO -o -o - g> co £ c « 03 rf o t 2 c o m >- ° .2, CO o m 2 a. "O CO p q 1 « Q. en CO en c 0 co ■9 E CD i: a 5 i: co „ co CO Q. 03 i « a > O O) ~" ~3 O C 2 T3 •^ £ 03 5 I C "- p CO £ c 3 CO 05 O °" o E CO o OJ I CO ai < o t- CT> §4 LU Q. a- E 3? m < en o 2 o o Q) CD 0) c CD E CD cn c CO CO CO CO CO CM CD CM Q. E CO CO ie •g.| 2 g.o iS CD s CD CO CO J£ o >- o o -2, £, o c c .j?CL <* CL Q_ o CO 0] a a m 10 o 3 3 cn c E OI cn ■x. D £ CD CO c E O u CD TO .C CO c O a. o £ Q. CO 5. co _ CD CO O C co CD O s>| °> o Q CD 15^ °>g> —" CO CO - 03 fc. 3 g§ 1| >< . 4) 10 C 2 CD S o £ o 00 CO 3 u CD b CD IS CO 3 E co "1 &| m ■£= o,° CO D) co II ■2 § g> Q. CD CD § £ Q uj .co co O) CO CO "> o P-~ 3 ^a?S CO o3 c c < cd cn g .q „ co co _, ° « 3 CO S el) * O q E m a. CD O) 3 CD CD c o CO CD -Q O *■ 03 CO .C c CD O CD CO E CO CD CO 3 ro •O O co CO CO o = « iS co £ci ~ CO CO ^ T3 CD co a o o ^ ■£; CD _q •^ t^ CO CM CM CM CM X3 CO C O CO n a> CO £i C/l j m < 5 CI) c o n >. _CD Q. E CO CO O ±= a5 2 S- co Oa. * s 3 ^ o .C CD gw - . S Q. D) CD LU Uj S O "5; ^ c^ m **- P. CO CO CD CD -C a p •t: o ■S ^ -c Q. CO CO S s "o coo Q. O 55 ,3 CD CB & O 03 g CO -2 1 .w 7= o ^ *ot S 2 o. a. a. CO CO t- CD If) CO o ■S2 d w CO CD CO 03 5 Q- -^ CO CO CD •a cj no" ■<*■ CO CD LO CM iu a ^w ~- 5T CO 03 Q. "O < CX3 CM t- t- 1^ CO 00 CO CD 03 O. >. I "O o 0) = o y- co 3 .E . S< Q. 03 CO CO CO 3 co o a 03 = "5 §O0- < a a O ,03 03 O C O w oL ^= CO E g o c CD CD C "S 03 co S: 03 £ « CO is O S ~ CD "O co 03 -Q 03 ■- -C Is ° o ■ .*< <5 m <" 5. _ to 03 O 2 5? ■ co Q. O CO 03 2 & g E co 3 CO 3 C C CO °s o 2 o ES 2 co CO 03 o CO CO CO 03 "O ■£ CO CO 03 £ to ^^fSPScoQ.'?: - 2 c < 646 CARTILAGE AND BONE DEVELOPMENT IN SCOMBROID FISHES Thomas Potthoff, Sharon Kelley, and Joaquin C. Javech1 ABSTRACT Early development of cartilage and bone was examined in representative species of the scombroid fish families Scombrolabracidae, Gempylidae, Trichiuridae, Scombridae, Istiophoridae, and Xiphiidae from cleared and stained larval size series. Development of the dorsal and anal fins and their pterygiophore supports, development of the neural and haemal spines and hypural complex, and ossification of the vertebrae were studied. The first appearance and location of these skeletal elements in cartilage were noted, and then the direction of new additions was observed. Direction of ossification of these elements was also noted. There were three major kinds of verebral column development: The first was shared by Scombrolabracidae, Scombridae in part - Scombrini, Scomberomorini, and Thunnini; the second was shared by Gempylidae, Sarda (Scombridae in part - Sardini), Istiophoridae, and Xiphiidae; the third kind was found in Trichiurus (Trichiuridae). Saddle-shaped ossifications of the vertebrae were found only in the Scombrolabracidae, and Gempylidae, and Scombridae. Four major kinds of fin and pterygiophore development were observed in the scombroid families: Scombrolabracidae and Scombridae in part - Scom- brini shared one kind; Gempylidae, Trichiuridae, and Scombridae in part - Scomberomorini, Sardini, and Thunnini shared another kind, which had some variations for different taxa; Istiophoridae had the third kind; and Xiphiidae had the fourth kind. Initial ossification of the vertebral column started in one place mScombrolabrax, Gempylidae, Trichiurus, and Xiphias, in two places in Scomber omorus, Sarda, Thun- nus, and Istiophorus, and in four places in Scomber and Acanthocybium. From our investigation, we are just beginning to learn about developmental characters and we cannot interpret their full meaning until more developmental work has been accomplished; we can only state that billfish (Istiophoridae, Xiphiidae) are very different from all other scombroids studied and that Scombrolabrax shows affinity with the scombroids. In this paper we describe development of selected osteological features of families in the suborder Scombroidei. We believe that this ontogenetic data will be useful in future taxonomic studies to aid in establishing familial relationships. Under current classification the scombroids comprise various num- bers of families. Greenwood et al. (1966) recognized six families in the suborder Scombroidei: Scom- bridae, Gempylidae, Trichiuridae, Istiophoridae, Xiphiidae, and Luvaridae. Gosline (1968), Potthoff et al. (1980), and Collette et al. (1984) included the family Scombrolabracidae in the Scombroidei, but Johnson (in press) removed it recently. Collette et al. (1984), Leis and Richards (1984), and Tyler et al.2 removed the Luvaridae from the Scombroidei. For this study we examined ontogenetic series of representative genera of the families Scombrola- bracidae, Gempylidae, Trichiuridae, Scombridae (four tribes), Istiophoridae, and Xiphiidae. Southeast Fisheries Center Miami Laboratory, National Marine Fisheries Service, NOAA, 75 Virginia Beach Drive, Miami, FL 33149. 2Tyler, J. C, G. D. Johnson, I. Nakamura, and B. B. Collette. Osteology and relationships of the oceanic fish Luvarus imperialis (Luvaridae): an acanthuroid not a scombroid. Unpubl. manuscr. National Museum of Natural History, Wash., DC 20560. Research on the larvae and young stages of scom- broids, particularly tunas (Richards and Klawe 1972) has been extensive. In general, most papers deal with the external description of the larvae and juveniles (Okiyama and Ueyanagi 1978); few exist that address the internal morphology and develop- ment of scombroids and those are mostly on scom- brids. Kramer (1960) described bone development in the mackerel (Pneumatophorus diego = Scomber japonicus). Potthoff and Richards (1970), Matsu- moto et al. (1972), and Richards and Potthoff (1974) published osteological characters for juvenile scom- brids. Cartilage and bone development were de- scribed in Thunnus atlanticus (Potthoff 1975), Scom- brolabrax heterolepis (Potthoff et al. 1980), and Xiphias gladius (Potthoff and Kelley 1982). Kohno et al. (1984) described fin and cartilaginous fin sup- port development in Scomber japonicus. To our knowledge no developmental studies of cartilage and bone have been made for the scombroid families Istiophoridae and Gempylidae, although a part of the research presented here was published in Col- lette et al. (1984). Since Collette et al. (1984), we have conducted additional research and have dis- covered several errors in our published observations. We have added developmental series of Scomber Manuscript accepted February 1986. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 647 FISHERY BULLETIN: VOL. 84, NO. 3 spp., Scomberomorus spp., Acanthocybium solanderi and Sarda sarda (Scombridae), Trichiurus lepturus (Trichiuridae), and Makaira nigricans (Istiophor- idae). We examined numerous juvenile and adult Trichiuridae; our findings are incorporated here. In table 161 of Collette et al. (1984), observations from the gempylid Diplospinus multistriatus were er- roneously listed under Trichiuridae. In this paper we have revised and corrected that table and incor- porated all our new findings (Tables 1, 2). Table 1 .—Developmental and osteological features and counts Gempylidae Trichiruidae Scombridae Scombrolabracidae (Scombrolabrax) without tail and pelvic fin, Trichiurus with tail and pelvic fin, Benthodesmus Evoxymetapon Lepidopus Scombrini (Scomber) Predorsal bones: present or absent absent present or or absent2 number 0 0 or 1 First anteriormost dorsal pterygiophore: supports number of 2 2 fin spines inserts in interneural space number 3 2 First anteriormost anal pterygiophore: supports number of 3 32or 3 spines or rays Middle radials: present or absent present present5 Dorsal and anal stay: present or absent present present ossifies to one or two one part one or parts two parts6 posteriorly bifurcated nonbifurcated bifurcated or nonbifurcated Pelvic fin: spine, ray count l,5 l,5;l,4;l,2; M;l Preural centrum 3: neural spine with or without cartilage tip with with haemal spine autogenous autogenous autogenous or nonautogenous Vertebrae inclusive of urostyle supporting caudal rays: number 3 3 Number of vertebrae: precaudal + caudal = 13 + 17 = 30 usually more total precaudal, fewer caudal total 31-67 Epurals: number 3 73 Anterior epural fused with neural arch of Pu2 No No absent present not determined 40 + 126 = 166 absent 0 43 present present one part nonbifurcated l,1;l,2 with ontogenetically fused fewer precaudal, more caudal, total 99-192 1 (ontogenetic fusion from 2) No absent 0 present present one part bifurcated l,5 with autogenous 13,14 + 17,18 = 31 No 1Data from Fritzsche and Johnson (1980) and G. D. Johnson (text footnote). 2Ruvettus, Thyrsitops and Tongaichthys have one predorsal bone. 3Rexea and Thyrsites (Leionura) have two spines, Nealotus ontogenetically has three spines but second spine fuses to basipterygium during devel- opment. 4Two of these spines are extreme vestiges. 648 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES METHODS Scombroid larvae were cleared and stained for cartilage and bone (Potthoff 1984) and subsequent- ly measured in millimeters with a calibrated ocular micrometer under a binocular microscope. Noto- chord length (NL) was measured on preflexion and flexion stage larvae from the anterior tip of the up- per jaw to the posterior tip of the notochord. Stan- dard length (SL) was measured from the anterior for the scombroid families and Morone, a primitive perciform fish. Scombridae— ■Continued Scomberomorini (Scomberomorus) Scomberomorini (Acanthocybium) Sardini (Sarda) Thunnini (Thunnus) Istiophoridae (Istiophorus) Xiphiidae (Xiphias) Percichthyidae (Moroney absent 0 absent 0 absent 0 absent 0 absent 0 absent 0 present 3 2 2 2 2 3 1 to 3, mostly 2 3 3 3 2 3 1 2 3 3 not known 3 3 2 1 to 3, 3 mostly 2 present present present present present absent present present one part nonbifurcated present one part slightly bifurcated present one part bifurcated present one part bifurcated present one part bifurcated, sometimes non- bifurcated present one part non- bifurcated present one part nonbifurcated l,5 l,5 l,5 l,5 l,2 l,5 with with with with with without with autogenous autogenous autogenous autogenous autogenous non- autogenous autogenous 4,5 (16-22) + (24-32) = (41-53) (30-32) + (31-33) = (62-64) 26 + 25 = 51 fewer precaudal, more caudal, total 39-41 12 11 + + 12 = 24 13 = 24 15 + 11 =26 16 + 10 = 26 12 + 13 = 25 11 + 14 = 25 2 2 2 2 3 3 3 No No No Yes No No No 5Neoepinnula lacks middle radials. 6Lepidocybium, Rexea, Diplospinus, Paradiplospinus, Tongaichthys, and Gempylus have a one-part stay, all other gempylids have a two-part stay. 7Diplospinus ontogenetically usually has three epurals, posterior two epurals are fused to one in adults, but some Diplospinus develop only two epurals. 649 FISHERY BULLETIN: VOL. 84, NO. 3 Table 1 .—Continued. Scombrolabracidae (Scombrolabrax) Gempylidae Trichiuridae Scombridae with tail and without pelvic fin, tail and Benthodesmus pelvic fin, Evoxymetapon Scombrini Trichiurus Lepidopus (Scomber) Uroneural: number 2 2 Hypural 5: present or absent present present fused or separate separate separate Ontogenetic hypural fusion: fusion of hypurals 1 & 2 to ventral plate is in cartilaginous or no fusion if present, ossified state ossified fusion of hypurals 3 & 4 to dorsal plate is in cartilaginous or no fusion If present, ossified state ossified Procurrent spur (Johnson 1975): present or absent present present, reduced or absent Stay on 4th pharyngo- branchial (G. D. Johnson, text footnote): present or absent absent absent absent present not known fused to uroneural proximally not known cartilaginous not known absent cartilaginous or ossified absent absent absent present Table 2.— Developmental features for the scombroid Neural and haemal arches and spines, parapophyses and hypural parts initially develop in the following places on the notochord by the following se- quence. Addition is in a given direction. Developing pterygiophores and fin spines and rays are added in a direction. Scombrolabracidae (Scombrolabrax) Gempylidae (Gempylus, Nesiarchus, Diplospinus) 1. 2. 3. Anterodorsad, posteriorly. Posteroventrad, posteriorly and anteriorly. Ventrad at center, posteri- orly and anteriorly. Dorsad at center, posterior- ly and anteriorly. Anterodorsad, posteriorly. Posteroventrad, posteriorly and anteriorly. Ventrad at center, posteri- orly and anteriorly. First dorsal: anteriorly and posteriorly. Second dorsal: an- teriorly and posteriorly. Anal: anteriorly and posteriorly. First dorsal: posteriorly. Sec- ond dorsal: anteriorly and posteriorly. Anal: anteriorly and posteriorly. Trichiuridae (Trichiurus) 1. Anterodorsad, posteriorly. 2. Ventrad at center, posteri- orly and anteriorly. Entire dorsal and anal: poste- riorly. 650 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES Scombridae — Continued Scomberomorini Scomberomorini Sardini Thunnini Istiophoridae Xiphiidae Percichthyidae (Scomberomorus) (Acanthocybium) (Sarda) (Thunnus) (Istiophorus) (Xiphias) (Moroney 1 111112 present present present present absent separate fused to uroneural proximally separate separate cartilaginous cartilaginous cartilaginous cartilaginous cartilaginous or ossified cartilaginous cartilaginous cartilaginous cartilaginous cartilaginous or ossified or ossified or ossified present separate present separate ossified ossifed no fusion no fusion absent absent absent absent absent absent present present present present present present present absent families and Morone, a primitive perciform fish. Sequence of fin and associ- ated pterygiophore develop- ment. First anteriormost dorsal and anal pterygiophore develop from one or two pieces of carti- lage. Number of initial places of ossification along vertebral column; centra develop from saddle-shaped ossifications at bases of neural and haemal arches. 1. Second dorsal and anal concurrently. 2. First dorsal. First dorsal separated from second dorsal during part of devel- opment. Dorsal from one piece, anal from two pieces. 1;Yes 1 . First dorsal. 2. Second dorsal and anal concurrently. First dorsal separated from second dorsal during part of devel- opment. 1 . All dorsal rays and pteryg- iophores dorsoanterior to anal fin. 2. All dorsal rays and pteryg- iophores opposite future anterior portion of anal fin. 3. All anal rays and pterygio- phores. Dorsal from one piece, anal from two pieces. Dorsal and anal from one piece. 1;Yes 1;No 651 FISHERY BULLETIN: VOL. 84, NO. 3 Table 2.— Continued. Neural and haemal arches and spines, parapophyses and hypural parts initially develop in the following places on the notochord by the following se- quence. Addition is in a given direction. Developing pterygiophores and fin spines and rays are added in a direction. Scombridae, Scombrini (Scomber) Scombridae, Scomberomorini (Scomberomorus) Scombridae, Scomberomorini (Acanthocybium) 1 Posteroventrad, posteriorly and anteriorly. Ventrad at center, posteri- orly and anteriorly. Dorsad at center, posterior- ly and anteriorly. Anterodorsad, posteriorly. Anterodorsad, posteriorly. Posteroventrad, posteriorly and anteriorly. Ventrad at center, posteri- orly and anteriorly. Dorsad at center, posterior- ly and anteriorly. Not entirely known. Smallest specimen available had al- ready two centers of initial development: anterodorsad and posteroventrad. First dorsal: pterygiophores anteriorly and posteriorly. Spines: one anteriorly, rest posteriorly. Second dorsal: an- teriorly and posteriorly. Anal: anteriorly and posteriorly. First dorsal: posteriorly. Sec- ond dorsal: anteriorly and pos- teriorly. Anal: anteriorly and posteriorly. First dorsal: probably posteri- orly. Second dorsal: anteriorly and posteriorly. Anal: anterior- ly and posteriorly. Scombridae, Sardini (Sarda) Scombridae, Thunnini (Thunnus) Istiophoridae (Istiophorus) Xiphiidae (Xiphias) 1. Anterodorsad, posteriorly. 2. Posteroventrad, posteriorly and anteriorly. 3. Ventrad at center, posteri- orly and anteriorly. 1. Anterodorsad, posteriorly. 2. Posteroventrad, posteriorly and anteriorly. 3. Ventrad at center, posteri- orly and anteriorly. 4. Dorsad at center, posterior- ly and anteriorly. 1. Anterodorsad, posteriorly. 2. Posteroventrad, posteriorly and anteriorly. 3. Ventrad at center, haemal spines posteriorly, para- pophyses anteriorly. 1 . Anterodorsad, posteriorly. 2. Posteroventrad, posteriorly and anteriorly. 3. Ventrad at center, posteri- orly and anteriorly. First dorsal: pterygiophores posteriorly. Spines: first one anteriorly, rest posteriorly. Second dorsal: probably ante- riorly and posteriorly. Anal: anteriorly and posteriorly. First dorsal: pterygiophores posteriorly. Spines: first one anteriorly, rest posteriorly. Second dorsal: anteriorly and posteriorly. Anal: some ante- riorly, most posteriorly. Entire dorsal: very few anteri- orly, most posteriorly. Anal: very few anteriorly, most pos- teriorly. Entire dorsal: anteriorly and posteriorly. Anal: very few an- teriorly, most posteriorly. Percichthyidae (Moroney Anterodorsad, posteriorly. Ventrad at center, posteriorly and anteriorly. Posteroven- trad, posteriorly and anterior- ly. Initial sequence not known, not known if neural arches and spines develop initially at center. First dorsal: anteriorly and posteriorly. Second dorsal: an- teriorly and posteriorly. Anal: anteriorly and posteriorly. 'Data from Fritzsche and Johnson (1980) and G. D. Johnson (text footnote 3). 652 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES Sequence of fin and associ- ated pterygiophore develop- ment. First anteriormost dorsal and anal pterygiophore develop from one or two pieces of carti- lage. Number of initial places of ossification along vertebral column; centra develop from saddle-shaped ossifications at bases of neural and haemal arches. 1. Second dorsal and anal Dorsal concurrently. piece. 2. First dorsal. and anal from one 4;Yes 1. First dorsal. 2. Second dorsal and anal concurrently. First dorsal separated from second dorsal during part of devel- opment. Dorsal from one piece, anal from two pieces. 2;Yes First dorsal. Second dorsal and anal concurrently. First dorsal separated from second dorsal during part of devel- opment. First dorsal. Second dorsal and anal concurrently. Not known if first dorsal is separated from second dorsal during part of development. First dorsal. Second dorsal and anal al- most concurrently. First dorsal separated from sec- ond dorsal during part of development. Dorsal probably from one 4;Not known piece, anal not known. Dorsal from one piece, anal 2?;Yes probably from two pieces. Dorsal from one piece, anal 2;Yes from two pieces. 1. First dorsal. 2. Second dorsal and anal concurrently. First dorsal not separated from second dorsal during development. Dorsal from one piece, anal from two pieces. 2;No 1. Second dorsal and anal concurrently. First dorsal. First dorsal and first anal nor separated from second dorsal and second anal during devel- opment. Second dorsal and anal concurrently. First dorsal. Separation or continuity of first and sec- ond dorsals not known. Variable, dorsal and anal may 1;No develop from one or two pieces. Dorsal and anal from two ? ; N o pieces. 653 FISHERY BULLETIN: VOL. 84, NO. 3 tip of the upper jaw to the posterior margin of the hypural bones. Xiphias larvae were measured from the anterior margin of the eye to the posterior tip of the notochord for eye notochord length (ENL) or from the anterior margin of the eye to the posterior margin of the hypural bones for eye standard length (ESL). FAMILY SCOMBROLABRACIDAE Figure 1 Thirty Scombrolabrax heterolepis larvae (2.9-10.4 mm NL or SL) were available. Development of the vertebral column initially started in four places on the notochord: 1) antero- dorsad (neural arches and spines of future centra 1-3), 2) posteroventrad (parhypural, hypurals), 3) ventrad at the center (haemal arches and spines on future centra 16-21), and 4) dorsad at the center (neural arches and spines on future centra 12-28). The anterior neural spines were added in a posterior direction whereas the neural and haemal spines at the center of the body were added anteriorly and posteriorly. The two areas of neural spine develop- ment coalesced around the eighth neural spine anteriorly and just anterior to the hypural complex posteriorly. The hypurals were added in a posterior direction, but the parhypural and the two autoge- nous haemal spines were added anteriorly (Table 2). Ossification of the vertebral column in Scombrola- brax initially started in one place with the ante- riormost neural arches and spines and proceeded in a posterior direction. The hypural complex was the last along the vertebral column to start ossifying. Vertebrae first ossified by forming saddles of bone dorsad and ventrad around the notochord. As ossi- fication proceeded the saddles merged laterally forming an hourglass-shaped vertebra in the lateral view. Cartilaginous second dorsal and anal fin pterygio- phores developed first simultaneously above inter- neural spaces 15-17 and below interhaemal spaces 16-19 before the anterior neural arches and spines had coalesced. The addition of cartilaginous second dorsal and anal fin pterygiophores was in an ante- rior and posterior direction. First dorsal fin pteryg- iophores appeared second above interneural spaces 4-7, to which pterygiophores were added anterior- ly and posteriorly, terminating anteriorly in the third interneural space and joining with the second dorsal fin pterygiophores posteriorly. Dorsal and anal fin rays and spines developed in the same se- quence as their corresponding pterygiophores, but a little later (Table 2). Scombrolabrax heterolepis does not develop pre- dorsal bones. The first dorsal pterygiophore orig- inated from one piece of cartilage and inserted in the third interneural space supporting two fin spines (one supernumerary spine). The first anal pterygio- phore developed from two pieces of cartilage and supported three spines (two supernumerary spines). The posteriormost five or six dorsal and anal pte- rygiophores had middle radials. The last dorsal and anal pterygiophore supported a double ray and had a nonbifurcated stay (Table 1). In S. heterolepis, first caudal development of the cartilaginous parhypural and hypurals 1 and 2 was concurrent with the anterior development of the neural spines and the central appearance of haemal spines. The hypural complex development was described by Potthoff et al. (1980). Scombrolabrax heterolepis had the basic perciform caudal skeleton (Gosline 1968), with no hypural fusion observed in adults. The neural and haemal elements of preural centra 2 and 3 supported the procurrent caudal rays. A procurrent spur was present on the posteriormost ventral secondary caudal ray with a basally fore- shortened ray anterior to it (Johnson 1975) (Table 1). FAMILY GEMPYLIDAE Figures 2-4 One hundred and ten gempylids in 11 genera were available: 33 Gempylus serpens (3.7-9.9, 160 mm NL or SL), 2SNesiarchus nasutus, (2.6-10.2, 55, 242 mm NL or SL), 7 Neoepinnula orientalis (3.3-7.1, 112 mm NL or SL), 11 Nealotus tripes (3.4-11.9, 24-140 mm NL or SL), 5 Lepidocybium flavobrunneum (5.5-35.3 mm NL or SL), 5 Promethichthys prome- theus (26.4-161 mm SL), 2 Rexea sp. (132, 155 mm SL), 2 Ruvettus pretiosus (209, 212 mm SL), 1 Thyrsitops lepidopoides (160 mm SL), 16 Diplospinus multistriatus (3.4-13.5 mm NL or SL), 5 Thyrsites atun (= Leionura, 83-254 mm SL). Of these, G. serpens, D. multistriatus, and TV. nasutus yielded complete developmental series. Development of the vertebral column initially started in three places on the notochord: 1) antero- dorsad (neural arches and spines on future centra 1-6); 2) posteroventrad (hypurals); and 3) ventrad at the center (anterior haemal arches and posterior parapophyses). The neural arches and spines were Figure 1.— Schematic representation of vertebral column, dorsa and anal fin, pterygiophore, and hypural development in Scorn brolabrax heterolepis, Scombrolabracidae. Cartilage, white; ossi fying, stippled. Scale represents interneural and interhaemal spac< number and vertebra number. 654 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 i — i — r i 1 1 1 r i 1 — i — i 1 — i — i — i — i — r "i — i 1 -U- 3 2 mm NL 3 7 mm NL 55° UUJUUUUUUUUUO-l TTTTTT-TTT-r^ 0i\0 OnO 35 mm NL UXXU UMUIIMIjUUU^l^ « ao fc & O 3-8 mm NL il^' ^/ n## 1° nO0S n° „00„ V^X^p* % % ^ * ^ * ^ 45 mm NL PffSS ^^<<^<<^Pf^f -o— -0- nvmxmxnr^ \w \%w\^*>^ 4 5mm NL 10 12 14 16 18 20 22 24 26 28 30 655 FISHERY BULLETIN: VOL. 84, NO. 3 12 14 16 18 20 22 24 26 28 30 32 i — i — i — i — i — i — n-! — r~\ — i — i — i — [— m — i — i — m — rn I i I r 34 36 38 40 42 44 46 48 50 52 i — i — i — i — i — i — i — r 54 ~ 1 t — i — i — r t— i — i — i — i — i — r 37 mm NL 3 7 mmNL fl ilUlUlUUUUUUJu^- 4 7 mm NL 47 mm NL fit ■/M/tltM/t/Hlirit ft/it nil • :■■:. ■ - -. ■ ^ ^ ^ ^ r-r 5 5 mm NL IJJJJJJJJJJJLdJJJU. ^^^v^vKV^^^w^^y^ 70 mmNL \0 0 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 Figure 2.— Schematic representation of vertebral column, dorsal and anal fin, pterygiophore, and hypural development in Gempylus serpens, Gempylidae. Cartilage, white; ossifying, stippled. Scale represents interneural and interhaemal space number and vertebra number. added in a posterior direction. Haemal arches and spines developed only when the neural spines reached the caudal area, and they were added in a posterior direction. Parapophyses were added ante- riorly. The hypurals were added posteriorly, the par- hypural and the autogenous haemal spines were added anteriorly (Table 2). Ossification of the verte- bral column in the gempylid genera examined by us initially started in one place and was similar to the ossification in Scombrolabrax, except in Diplospinus the vertebral column was ossified to preural centrum 6 when the urostyle and the hypurals initially started to ossify. Saddle-shaped vertebral ossifications were observed in all gempylids examined, similar to those described for Scombrolabrax. Gempylids developed first dorsal fin pterygio- phores and fin spines first, after only a few carti- laginous neural spines had developed. Development of first dorsal fin pterygiophores and spines was in a posterior direction. During early development the neural spines were anterior to the first dordal fin pterygiophores and fin spines, but later they developed faster and were posterior to the pterygio- phores. Pterygiophores of the second dorsal and anal fins developed before the developing first dorsal fin pterygiophores and had joined with the second dor- 656 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES 2 4 6 8 10 12 14 16 18 -i — i — i — r 20 ~~l — 22 — I- 24 26 28 30 32 34 36 ~l r ~r "i r T T i — r ~T T T" T T 3 6 mm NL ^Mmm^ 4 0 mm NL - *8 ^t SS <*=? i 658 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES t* O) c *- 3 * 5 _: * CD t- & > -" C CO 3 TO bo <- •C -° CO 3 C s •- & * p s bO o o % a O CS &H CS Thyrsites (Leionura), and Nealotus where only two spines were supported (one supernumerary). Lar- vae of Nealotus have three spines associated with the first anal pterygiophore, but in juveniles the sec- ond anal spine was fusing to the posterior process of the pterygiophore. No evidence of a similar fu- sion was observed in Rexea or Thyrsites (Leionura). Gempylids had middle radials in one to six posterior- most dorsal and anal pterygiophores (except Neo- epinnula lacked middle radials). A double ray, and a two-part posteriorly bifurcated stay was associated with the last dorsal and anal pterygiophore in ap- proximately one half of the genera. Lepidocybium, Gempylus, Diplospinus, Paradiplospinus, Tongaich- thys, and Rexea had a one-part posteriorly bifurcated stay (Table 1). First caudal development of the cartilaginous parhypural and hypurals 1 and 2 was concurrent with anterior development of a few neural spines and some first dorsal fin pterygiophores and fin spines. The gempylid genera studied by us developed all parts found in basic perciform caudal skeletons (Gosline 1968), even the smaller second uroneural. Caudal parts then fuse differently in the various genera of adults (Matsubara and Iwai 1958). The neural and haemal elements of preural centra 2 and 3 supported the procurrent caudal rays. In the gem- pylids the procurrent spur on the posteriormost ven- tral secondary caudal ray may be present, reduced, or absent. Johnson (1975) examined two species in which it was absent (Table 1). FAMILY TRICHIURIDAE Figures 5-8 Seventy-three trichiurids in four genera were available: 61 Trichiurus (4.5-26, 300, 303, 510 mm TL), 8 Benthodesmus (4.5, 12 mm NL, 65-120, 541, 545 mm SL), 3 Evoxymetapon (210-550 mm SL), 1 Lepidopus (280 mm SL). Only Trichiurus yielded a complete developmental series. Development of the vertebral column in Trichi- urus initially started in two places on the noto- chord: 1) anterodorsad (neural arch and spine on future centrum 1), and 2) ventrad at the center (an- terior haemal arches and posterior parapophyses). Cartilaginous neural arches and spines were added in a posterior direction. Haemal arches and spines developed when the neural spines reached the ante- rior future caudal vertebrae. Addition of haemal arches and spines was also in a posterior direction (Table 2). Trichiurus lacked a caudal complex. Ossi- fication of the vertebral column started initially in one place, with the anteriormost neural spines and 659 FISHERY BULLETIN: VOL. 84, NO. 3 2 4 6 8 10 12 14 36 38 40 42 44 i — (— — i — i — i — i — i — i — i — i — i — i — i — i — i i — i — i 1 — i 1 1 — i — i — i — 108 110 112 114 116 118 164 166 i i i i 1 1 1 1 1 1 1 i 1 1 ? YTffimr"" 4 5 mm SL 5 0 mm SL Z\ 6 0 mm SL , — — - — -— . ■ ■■ . ~\ 8 0 mm SL 'ffwmr^ h^s 9 8 mm SL nfTTTTTTTTTTTT' v N X. \ \ ^X\V\\\^ -i i t i i i i ''ii 6 8 10 12 14 16 0 mm SL _i i i i i i i i i i i i i i i i i i i i i i i p 36 38 40 42 44 108 110 112 114 116 118 164 166 Figure 5.— Schematic representation of vertebral column, dorsal and anal fin, pterygiophore, and hypural development in Trichiurus lepturus, Trichiuridae. Cartilage, white; ossifying, stippled. Scale represents interneural and interhaemal space number and vertebra number. 660 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES Figure 6.— Left lateral view of the anteriormost three dorsal pterygiophores inserting in the interneural spaces 2-4 from a juvenile Trichiurus lepturus 510 mm TL. D, distal radial; Ns, neural spine; P, proximal radial; R, ray or spine; X, a new pterygiophore element of unknown homology. Cartilage, white; bone, stippled. arches and proceeded in a posterior direction. Saddle-shaped ossifications of the vertebrae as seen in Scombrolabrocidae, Gempylidae, and Scombridae were not observed in Trichiurus, instead vertebral ossification started laterally on both sides of the notochord as a thin strip of bone. During further development the lateral strip elongated dorsad and ventrad joining the strip from the opposite side and forming a ring of bone around the noto- chord. Trichiurus first developed two of the three ante- rior dorsal fin spines. Next the first dorsal pteryg- iophore developed. Then dorsal pterygiophores, the third dorsal fin spine, and the dorsal fin rays were added in a posterior direction, with the pterygio- phore development being slightly posterior to the ray development and considerably posterior to the neural arch and spine development. The single large anal spine developed first after dorsal fin ray and pterygiophore development had dorsally passed the anterior portion of the anal fin fold. Next, the large first anal fin pterygiophore and some haemal arches and spines developed. Further development con- sisted of the addition of anal fin rays, pterygio- phores, and haemal arches and spines in a posterior direction. The haemal arches and spines and the anal fin rays developed slightly anterior to the anal pte- rygiophores. The anal pterygiophores were slightly anterior to the dorsal fin ray and pterygiophore development (Table 2). Trichiurus lacked predorsal bones. The first dor- sal pterygiophore supported two fin spines (one supernumerary) and originated from one piece of cartilage. In larvae the first dorsal pterygiophore inserted between the split neural arch and spine of the first centrum, thus inserting into the first and second interneural spaces. However, in adults the first dorsal pterygiophore inserted into the second interneural space. All following interneural and interhaemal spaces accommodated one pterygio- phore per space. The first anal pterygiophore was larger than the following pterygiophores, but it developed from one piece of cartilage and sup- ported one supernumerary spine and one ray (Table 1). The pterygiophores in Trichiurus and probably in most if not all species of the Trichiuridae are anatomically different from those of other scom- 661 FISHERY BULLETIN: VOL. 84, NO. 3 Figure 7.— Two dorsal fin pterygiophores from Trichiurus lepturus 510 mm TL, taken directly from opposite the anterior portion of the anal fin. A, left lateral view of the pterygiophores and rays; the left side of the posterior ray has been removed. Cartilage, white; ossifying, stippled. B, dorsal view of one of the two pterygiophores; unfused parts have been disarticulated. C, dorsal view of pterygiophore in B, unfused parts have been left articulated. For abbreviations see Figure 6. broids (G. D. Johnson3). The anteriormost two dor- sal pterygiophores supported three spines, which were the only dorsal fin spines and which had ser- rations in larvae and juveniles, but were smooth in adults. The anterior two pterygiophores had two parts each and supported fin spines. The 3d-127th pterygiophores had three parts and supported fin 3G. David Johnson, Curator (Fishes), Smithsonian Institution, National Museum of Natural History, Wash., DC 20560, pers. com- mun. 1985. rays, the distal parts being located between the bifurcate bases of the rays. These distal parts were not homologous with distal radials and are labeled "X" in Figures 6-8. The 128th-130th pterygiophores had four parts, and the last three pterygiophores (131st-133d) had become vestigial having a variable number of parts, usually from two to four. Anal fin pterygiophores were anatomically similar to the dor- sal fin pterygiophores. The first anal fin spine was large and serrated in larvae and juveniles but became small and smooth in adults. Trichiurus lar- 662 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES V 1 ft O CD 0) 0) c _o 1 ft t- 0> T3 0) C o> -a a> +-» o o ft ft s ■ tfl £ c o o JS X> O oj en I* a> si « ft a '5b £ ►» o R a ft C *0 13 to S-, o (-1 0) o •a >> O cS ■— a .2 ** sl & E co ,-< o cd co tt i2 s- cd a. J -2 bt &h vae and juveniles developed an anal fin in which the rays were of the same length as those in the dorsal fin, but the anal rays became very short and vestigial in adults. In adult Trichiurus the posterior end of the dorsal fin was anterior to the posterior end of the anal fin. Other trichiurids (Benthodesmus, Evox- ymetapon, Lepidopus) examined by us had pte- rygiophore arrangements similar to Trichiurus. FAMILY SCOMBRIDAE The family is a very speciose group which is divided into two subfamilies (Collette et al. 1984). For the monotypic Gasterochismatinae, larvae were not obtainable, but one or more species for each of the four tribes of the Scombrinae was studied. Tribe Scombrini Figure 9 Twenty-two Scomber japonicus (4.4 mm NL - 9.6 mm SL, 100, 103 mm SL) and 12 S. scombrus (5.7 mm NL - 8.2 mm SL) were used in this study. Many more Scomber smaller than 5.5 mm NL were avail- able but showed no cartilage development along the notochord. In addition, developmental studies on Scomber by Kramer (1960) and Kohno et al. (1984) were consulted. Development of the vertebral column in Scomber initially started in four places on the notochord: 1) posteroventrad (parhypural, hypurals 1 and 2), 2) ventrad at the center (anterior haemal arches and spines), 3) dorsad at the center (neural arches and spines above developing haemal arches and spines), and 4) anterodorsad (neural arches and spines of future centra 1-3). The anterior neural spines were added posteriorly, the neural spines at the center of the notochord were added anteriorly and poste- riorly, the haemal spines were added posteriorly, but the parapophyses were added anteriorly. The hypu- rals were added in a posterior direction, but the two autogenous haemal spines were added anteriorly. The dorsal and ventral areas of development co- alesced completing the cartilaginous ontogeny of the vertebral column. Ossification of the vertebral column (neural and haemal spines, vertebrae, and hypural complex) initially started in four places: 1) dorsoanteriorly (anteriormost neural arches and spines), 2) ventrad at the center (anterior haemal arches and spines and posterior parapophyses), 3) posteriorly (hypural complex), and 4) dorsad at the center (neural arches and spines). The four initial areas of ossification coalesced as ossification pro- gressed. Vertebrae in Scomber initially had saddle- 663 FISHERY BULLETIN: VOL. 84, NO. 3 I 3 5 7 9 II 13 15 17 19 21 23 25 27 29 31 I I I I I I I I I I I 1 I I 1 I I I I I I 1 1 1 1 1 — 1 1 1 — i 59 mm SL Tirnrmro- UM^J^W 68 mm SL VTTtTTTW^-v^ J — l — I — l — I — l — I — I — I I i I I I I i i i i i i i i 3 5 7 9 II 13 15 17 19 21 23 25 27 29 l I I I I Figure 9.— Schematic representation of vertebral column, dorsal and anal fin, pterygiophore, and hypural develop- ment in Scomber japonicus, Scombrini, Scombridae. Cartilage, white; ossifying, stippled. Scale represents interneural and interhaemal space number and vertebra number. 664 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES shaped ossifications similar to those described for Scombrolabrax (Table 2). Cartilaginous second dorsal and anal fin pteryg- iophores developed first simultaneously above inter- neural spaces and below interhaemal spaces 17-19. The addition of cartilaginous second dorsal and anal fin pterygiophores was in an anterior and posterior direction. Cartilaginous first dorsal fin pterygio- phores appeared second above interneural spaces 5-8 and were added anteriorly and posteriorly, ter- minating anteriorly in the third interneural space and joining with the second dorsal fin pterygio- phores posteriorly. Second dorsal and anal fin rays developed in the same sequence as their correspond- ing pterygiophore, but a little later. The first dor- sal fin spines developed from anterior in a posterior direction, but the anteriormost (supernumerary) spine first developed when seven first dorsal fin spines were already present (Table 2). Scomber lacked predorsal bones. The first dorsal pterygiophore originated from one piece of cartilage and inserted in the third interneural space support- ing two fin spines (one supernumerary spine). The first anal pterygiophore was considerably larger than all other pterygiophores, but it originated from only one piece of cartilage supporting two anal spines (one supernumerary spine). The posterior- most six dorsal and anal pterygiophores had mid- dle radials. The last dorsal and anal pterygiophore supported a double finlet and had a posteriorly bi- furcated stay (Table 1). In Scomber, caudal development of the cartilag- inous parhypural and hypurals 1 and 2 was first before any other development of cartilaginous haemal or neural arches and spines along the noto- chord. The development of the hypural complex from the first appearance of cartilaginous hypurals to ossification onset was described by Kohno et al. (1984) and our findings are in agreement with theirs. Kramer (1960) described the ossification sequence in the hypural complex of Scomber. In our speci- mens, hypurals 1 and 2 were fusing to a ventral hypural plate before ossification onset. Hypurals 3 and 4 were fusing in some larvae before and in others after ossification onset. The neural and haemal elements of preural centra 2 and 3 supported the procurrent caudal rays. A procurrent spur and a basally foreshortened ray were absent in Scomber (Johnson 1975) (Table 1). Tribe Scomberomorini Figures 10, 11 Thirty-nine specimens were available: 9 Scomber- omorus cavalla (4.1-6.2 mm NL), 17 S. maculatus (6.1 mm NL - 10.2 mm SL, 40.5-67.5 mm SL), 3 S. regalis (5.3, 6.5 mm NL, 85.0 mm SL), 4 S. tritor (6.0 mm NL - 8.0 mm SL), 6 Acanthocybium solan- deri (6.2 mm NL - 10.8 mm SL). None of the above five species yielded complete developmental series. However, S. cavalla specimens showed the cartilag- inous ontogeny of the vertebral column, of the dor- sal and anal fin pterygiophores and of the hypural complex. The S. maculatus specimens showed the latter phases of pterygiophore and hypural complex development, dorsal and anal fin development, and the ossification of the vertebral column and the hypural complex. Specimens of 5. regalis and S. tritor provided evidence that development for the Atlantic species of Scomberomorus is very similar. Specimens of A. solanderi gave incomplete infor- mation on cartilaginous vertebral column develop- ment, but adequate information on dorsal and anal pterygiophore, on dorsal and anal fin, on hypural complex development, and on the ossification se- quence of the vertebral column. Development of the vertebral column in Scomber- omorus initially started in four places on the noto- chord: 1) anterodorsad (neural arches and spines on future centra 1-3), 2) posteroventrad (parhypural, hypurals 1 and 2), 3) ventrad at the center (four haemal arches and spines), and 4) dorsad at the center (six neural arches and spines above initial haemal spine development). The anterior neural spines were added posteriad, the neural spines at the center of the notochord were added anteriorly and posteriorly, the haemal spines were added most- ly posteriorly but a few were added in an anterior direction. All parapophyses were added in an ante- rior direction. The hypurals were added in a poste- rior direction, but the two autogenous haemal spines were added in an anterior direction. The dorsal and ventral areas of development coalesced and thus car- tilaginous ontogeny of the vertebral column was complete. Ossification of the vertebral column ini- tially started in two places: 1) anteriorly (neural arches and spines, and centra) and 2) posteriorly (hypural complex). Ossification of the neural arches and spines and centra was in a posterior direction. In the hypural complex ossification started with the urostyle and proceeded anteriorly to preural cen- trum 3. Then the ventral hypural plate started to ossify followed by the dorsal plate, the parhypural, and the two autogenous haemal spines. Last to start ossification were the epurals, the uroneural, and the neural spines. Vertebrae in Scomberomorus had saddle-shaped ossifications similar to those de- scribed for Scombrolabrax (Table 2). 665 FISHERY BULLETIN: VOL. 84, NO. 3 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 i — i — i — ■ — rn — i — n — i — i — i — i i i i I — n — rn — i — n — I r— n i i i i i n i i i i i i I ll Scomberomorus cavalla 11 00 0 0 0 WW \ 4.1 mm NL 5 Omm NIL 5.0 mm NL 5.1 mm NL 5.5mm NL JfffffMM i i i i i i i i i i i i i i i i i i i i i i i i i i i i iii — i i i i i i i i i 1 I 3 5 7 9 II 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 Figure 10.— Schematic representation of vertebral column, dorsal and anal fin, pterygiophore, and hypural development in Scomberomorus cavalla and S. maculatus, Scomberomorini, Scombridae. Cartilage, white; ossifying, stippled. Scale represents inter- neural and interhaemal space number and vertebra number. 666 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES <£> no XI E 03 ft 03 -B N O CU T3 O c3 — . A? CS -a £ ^^ u «! W OS »- C 2 "ft § co 5b £ I £>| -1 ft § Hi-"" O 03 j. J- B <» 667 FISHERY BULLETIN: VOL. 84, NO. 3 Two to five cartilaginous first dorsal fin pteryg- iophores developed first above interneural spaces 3-5 at the time of ossification onset of the anterior- most neural arches and spines. The addition of car- tilaginous first dorsal fin pterygiophores was in a posterior direction. Five cartilaginous second dor- sal and anal fin pterygiophores developed second simultaneously in the anterior portions of the future second dorsal and anal fins. Some addition of carti- laginous second dorsal and anal fin pterygiophores occurred in an anterior direction, but most of the addition was posteriorly. Dorsal and anal fin rays and spines developed in the same sequence as their corresponding pterygiophores, but a little later (Table 2). Scomberomorus does not develop predorsal bones. The first dorsal pterygiophore originated from one piece of cartilage and inserted in the third inter- neural space supporting two fin spines (one super- numerary spine). The first anal pterygiophore developed from two pieces of cartilage and sup- ported three spines (two supernumerary spines). The posteriormost nine dorsal and anal pterygiophores had middle radials. The last dorsal and anal pteryg- iophore supported a double finlet and had a non- bifurcated stay (Table 1). In Scomberomorus, first caudal development of the cartilaginous parhypural and hypurals 1 and 2 was concurrent with the anterior development of the neural spines and the central appearance of haemal spines. Hypurals 3-5 were added posteriorly, the two autogenous haemal spines anteriorly. Hypurals 1 and 2 and hypurals 3 and 4 fused before ossifica- tion onset to a cartilaginous ventral and dorsal hypural plate. The dorsal and ventral plates fused after ossification to a single hypural plate with a cen- tral notch (Collette and Russo 1984). Hypural 5 gradually fused with the paired uroneural forming an autogenous bone resembling a third epural and mistaken as such by Leccia (1958). Two epurals developed anterior to the uroneural-hypural 5. These epurals remained autogenous. The neural and hae- mal elements of preural centra 2, 3, 4, and 5 sup- ported the procurrent caudal rays. A procurrent spur and basally foreshortened ray were absent in Scomberomorus (Johnson 1975) (Table 1). Only six Acanthocybium solanderi were available. We were therefore unable to ascertain a complete developmental sequence. Our smallest 6.2 mm NL specimen had two cartilaginous development centers along the notochord: some neural spines and arches anteriorly and the parhypural, hypural 1-3 poste- riorly. The next larger specimen 9.2 mm SL had all neural and haemal arches and spines developed, thus we were unable to tell if in Acanthocybium four ini- tial centers (as in Scomberomorus) or only three centers (as in Xiphias and Sarda) of cartilaginous development along the notochord were present. In all our Acanthocybium specimens, hypurals 1 and 2 gradually fused before ossification onset to a ven- tral cartilaginous hypural plate. In the 8.5 mm SL Acanthocybium, hypurals 3 and 4 were fusing before ossification onset; in the larger 9.5 and 10.4 mm SL specimens hypurals 3 and 4 were ossifying while still separate. The dorsal and ventral hypural plates were fused in adults to one plate with a notch (Conrad 1938; Collette and Russo 1984) (Table 1). Ossifica- tion of the vertebral column initially started in four places and was similar to the ossification in Scomber. The development of the dorsal and anal fins and their supporting pterygiophores in Acanthocybium was similar to that described in Scomberomorus. Tribe Sardini Figure 12 Ninety-nine Sarda sarda (2.4-9.0 mm NL or SL, 59-102 mm SL) were available. Of the larval speci- mens (2.4-9.0 mm NL or SL) only 32 were larger than 5 mm NL, and of these 10 were between 6.0 and 6.9 mm NL or SL, 6 were between 7.0 and 7.9 mm NL or SL, and 3 were larger than 8 mm SL. Thus, since development of the vertebral column in Sarda begins around 5 mm NL, only 32 specimens were useful to our study and they were too few to yield a complete developmental series. Our conclu- sions on Sardini development are not as well sup- ported as for most other scombroids. Development of the vertebral column in Sarda ini- tially started in three places on the notochord: 1) anterodorsad (neural arch and spine of future cen- trum 1), 2) posteroventrad (parhypural, hypurals 1 and 2), and 3) ventrad at center (haemal arches and spines, parapophyses). The anterior neural spines were added in a posterior direction and the haemal spines probably first appeared when the correspond- ing neural spines developed above them at the center of the notochord. Our evidence, however, is only indirect, because one 7.5 mm NL specimen had 21 neural spines and no haemal spines, but our 8.1 mm SL specimen had all neural and haemal spines developed. The cartilaginous hypurals were added posteriorly, but we could not observe the anterior addition of the autogenous haemal spines, although we assume that it happens in Sarda as in other scombroids with tails. Ossification of the vertebral column in Sarda initially started in two places: ante- riorly (neural arches and spines) and posteriorly 668 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 i I 1 I I I I I I I I I I i I I I — I i I I I I i i I i I I I I ii I I I I I 1 5.3 mm NL TJS UnUHHUlHM 6.1 mm NL I I I l i I l I l I I I l l i I l W^ 9.0 mm SL I I I I I I l I l I I i i i I i I i i l l i l I I l 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 Figure 12.— Schematic representation of vertebral column, dorsal and anal fin, pterygiophore, and hypural development in Sarda sarda, Sardini, Scombridae. Cartilage, white; ossifying, stippled. Scale represents interneural and interhaemal space number and vertebra number. (hypural complex). Our largest 9.0 mm SL specimen showed ossification to the 11th neural spine. We do not know if ossification in Sarda proceeds entirely posteriad or if in Sarda, as in Scomber and Acantho- cybium, there is some central ossification of neural and haemal spines before the anterior ossification has reached the center of the column. The hypural complex started to ossify early at the time ossifica- tion on the neural spines began anteriorly. Verte- brae in Sarda had saddle-shaped ossifications similar to those described for Scombrolabrax (Table 2). Cartilaginous first dorsal fin pterygiophores devel- oped first anteriorly above interneural spaces 2-10 in the 8.1 mm SL specimen. Addition of cartilag- inous first dorsal fin pterygiophores was in a poste- rior direction. The 8.4 mm specimen had all first dor- sal fin pterygiophores and some second dorsal and anal fin pterygiophores and they were continuous with each other. Therefore, we are unable to deter- mine if second dorsal and anal fin pterygiophores in Sarda developed before first dorsal fin pteryg- iophores were joined with the second dorsal fin pterygiophores. Three first dorsal fin spines were present in the 8.1 mm SL specimen, serially asso- ciated with the first three pterygiophores. Addition of first dorsal fin spines was in a posterior direc- tion, except for the anteriormost first spine (super- numerary), which developed later in the 9.0 mm SL Sarda. Second dorsal and anal fin rays were not developed in our 9.0 mm SL specimen. Our 59 mm SL specimen had the full adult compliment of fin rays (Table 2). 669 FISHERY BULLETIN: VOL. 84, NO. 3 Sarda did not develop predorsal bones. The first dorsal pterygiophore originated from one piece of cartilage and inserted in the second interneural space supporting two spines (one supernumerary spine). We do not know if the first anal pterygio- phore originated from one or two pieces of cartilage, but it is most likely that it originated from two pieces because it supported three fin elements (two super- numerary spines). The posteriormost seven to nine dorsal and anal pterygiophores had middle radials. The last dorsal and anal pterygiophore supported a double finlet and had a posteriorly bifurcated stay (Table 1). In Sarda first caudal development of the carti- laginous parhypural and hypurals 1 and 2 was con- current with the beginning development of the ante- riormost neural arches and spines. Hypurals 1 and 2 fused in the cartilaginous state to form the ven- tral hypural plate. In three specimens, hypurals 3 and 4 were separate after ossification onset. These hypurals were fused to the dorsal hypural plate in juveniles. Hypural 5, the uroneural and two epurals were separate in our juveniles. Collette and Chao (1975) found that in adults the dorsal and ventral plates fused to one hypural plate without a notch and that the uroneural fused with hypural 5, but the two epurals remained autogenous. The neural and haemal elements of preural centra 2, 3, 4, and 5 sup- ported the procurrent caudal rays. A procurrent spur and a basally foreshortened ray were absent in Sarda (Johnson 1975) (Table 1). Tribe Thunnini Figure 13 More than 86 specimens were available: 86 Thun- nus (mostly T. atlanticus and a few Thunnus spp., 3.7-9.7 mm NL or SL), and a small number of Auxis, Euthynnus, and Katsuwonus. We were unable to observe early cartilaginous development in all genera except Thunnus. Development of the vertebral column in Thunnus initially started in four places on the notochord: 1) anterodorsad (neural arches and spines of future vertebrae 1-3), 2) posteroventrad (hypurals 1 and 2), 3) ventrad at the center (anteriormost five haemal arches and spines and posteriormost two parapophyses), and 4) dorsad at the center (five neural arches and spines above initial haemal arch and spine development). The anterior neural arches and spines were added in a posterior direction, the central neural arches and spines were added ante- riorly (coalescing around the future 14th centrum) and posteriorly toward the epurals. The parapophy- ses were added in an anterior direction, whereas the haemal arches and spines were developing in a pos- terior direction. In the hypural complex hypurals were added posteriorly, but the parhypural and the two autogenous haemal spines were added in an anterior direction, coalescing with the central haemal arches and spines. Ossification of the verte- bral column in Thunnini initially started in two places similar to the ossification described for Scom- beromorus. Saddle-shaped vertebral ossification development was observed in all Thunnini examined, similar to the development described for Scombro- labra-x (Table 2). In Thunnini, cartilaginous first dorsal fin pteryg- iophores developed anteriorly in interneural spaces 3-6 when only few cartilaginous neural spines were present. Additional pterygiophores were added in a posterior direction. Later, small cartilaginous sec- ond dorsal fin pterygiophores appeared in the mid- dle of the vertebral column above interneural spaces 15-22. As the first dorsal fin pterygiophores devel- oped in a posterior direction, the second dorsal fin pterygiophores developed in an anterior and poste- rior direction until all the dorsal pterygiophores were continuous. Anal pterygiophores appeared below interhaemal spaces 20-25 and developed in an anterior and posterior direction. Addition of the first dorsal fin spines was in a posterior direction, except for the anteriormost spine (supernumerary), which developed when the second and third spine were already present. The second dorsal and anal fin rays developed in the same sequence as their correspond- ing pterygiophores but a little later (Table 2). All Thunnini species examined lacked predorsal bones. The first dorsal pterygiophore originated from one piece of cartilage and inserted in the third interneural space supporting two fin spines (one supernumerary spine). The first anal pterygiophore developed from two pieces of cartilage and sup- ported three fin spines (two supernumerary spines) (Potthoff 1975). Middle radials were present on the posterior eight or nine finlet supporting dorsal and anal pterygiophores. A one-part posteriorly bifur- cated stay developed with the posteriormost dorsal and anal fin pterygiophores (Table 1). In Thunnus, the caudal complex began to develop very early concurrently with the first anteriormost neural spines. Hypurals 1 and 2 and hypurals 3 and 4 developed separate cartilages and fused to a car- tilaginous dorsal and ventral hypural plate. Potthoff (1975) stated that hypurals 1 and 2 developed as one piece of cartilage from the start, but he examined only specimens larger than 5.0 mm NL not stained for cartilage. The dorsal and ventral hypural plates 670 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES I 3 5 7 9 II 13 15 17 19 21 23 25 27 29 31 33 35 37 39 i i I i i i i l i i l l i i i i i i l i i l l l — l — l — l — i — i — i — i — i — i — i — i — i — i — i 1 3 7 mm NL T^ \1 U VJ V U U u L- 40mm NL ^3^ 0 0 0 Do l/(/Ul/UUl/U u u v y d [7 (j^ \j \J V v « u ^V- 4 5 mm ^T />„„ „ I (I /! II II /I /I {/rl/fi^ffJ0.00.0O 00 00 51 mm ENL "VT^ WO *™W!tfmPtm£M tTTTTTTT^ u^A 19 5mm ESL Figure 15.— Schematic representation of vertebral column, dorsal and anal fin, pterygiophore, and hypural development in Xiphias gladius, Xiphiidae. Cartilage, white; ossifying, stippled. Scale represents interneural and interhaemal space number and vertebra number. 674 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES neural spaces 13-16. The anal pterygiophores first developed in a group above the future anterior part of the anal fin below interhaemal spaces 16-18. Fur- ther addition of cartilaginous dorsal and anal pte- rygiophores was in an anterior and posterior direc- tion. The posterior pterygiophore additions dorsad and ventrad were completed before the anterior additions ceased. The full complement of anal pte- rygiophores was reached before the full dorsal com- plement. Dorsal and anal fin rays first originated in the same areas as the pterygiophores, but at larger sizes with addition of rays in the same direc- tions (Table 2). Xiphias did not have predorsal bones. The first dorsal pterygiophore originated from one or two pieces of cartilage and inserted in the second inter- neural space, supporting from one to three fin spines. The first anal pterygiophore developed from one or two pieces of cartilage, supporting from one to three fin spines. Xiphias had no middle radials in the dorsal or anal pterygiophores, but a double ray and a nonbifurcated stay were associated with the posteriormost dorsal and anal pterygiophores (Potthoff and Kelley 1982) (Table 1). In Xiphias, cartilaginous hypurals were first seen before precaudal neural spine development was com- plete, but after dorsal and anal pterygiophore devel- opment had started. The hypural complex develop- ment was described by Potthoff and Kelley (1982). Hypurals 1-5 and the parhypural developed from separate cartilages, and there was no cartilage fu- sion. There were three epurals and one uroneural. Only one autogenous haemal spine was present on preural centrum 2. In adults the three epurals, the uroneural, hypural 5, and the parhypural remained autogenous, but hypurals 1-4 fused with each other and the urostyle forming a notched hypural plate (Gregory and Conrad 1937). The neural and haemal elements of only preural centrum 2 supported the procurrent caudal rays. A procurrent spur and basal- ly foreshortened ray were absent in Xiphias (John- son 1975) (Table 1). DISCUSSION AND CONCLUSION Developmental features observed in our study are illustrated in Figures 4-5 and 9-15. These features along with meristic and osteological characters are compared among the six scombroid families and the primitive percoid Morone in Tables 1 and 2. Al- though our conclusions are still preliminary because of lack of adequate developmental series for many genera, some comparisons, based largely on devel- opment, are worth noticing. There are three major kinds of early development and addition of the cartilaginous neural and haemal arches and spines along the notochord. Each kind may differ slightly between taxa. Scombrolabrax, Scomber (Scombrini), Scomberomorus (Scombero- morini), and Thunnini have one kind in which there are four initial developments on the notochord, but not necessarily in the given order, e.g., anteriorly dorsad, centrally dorsad, centrally ventrad, and pos- teriorly ventrad with a subsequent merger of the initial areas. Gempylidae, Sarda (Sardini), Istio- phoridae, and Xiphiidae have a second kind in which there are three initial developments, e.g., anterior- ly dorsad, centrally ventrad, and posteriorly ven- trad; then the addition is from anterior in a poste- rior direction with a merger in the posterior, near the hypural complex. Trichiurus, which lacks hypu- rals, has the third kind in which there are two ini- tial developments, e.g., anteriorly dorsad and cen- trally ventrad with addition in a posterior direction. We could not fully determine the cartilaginous development for Acanthocybium, because of an in- complete series, and for trichiurids with tails, because a series was lacking. In the Scombrolabracidae, Gempylidae, and Scom- bridae, the vertebrae first develop by coalescence of saddle-shaped ossifications positioned dorsad and ventrad. We were not able to observe saddle-shaped ossification in Acanthocybium because we lacked specimens. The other scombroid families, Trichiu- ridae (Trichiurus), Istiophoridae, and Xiphiidae, and the primitive percoid Morone did not have these saddle-shaped ossifications. Saddle-shaped ossifica- tions have been observed during ontogeny in other perciform fish such as Enchelyurus brunneolus (Blenniidae) by Watson5 and Lutjanus campechanus (Lutjanidae) by Potthoff and Kelley6. We are unable to comment at this time on the significance of these saddle-shaped ossifications until the ontogeny of many more taxa is studied. In the Scombrinae two species belonging to two different tribes share a peculiar ossification se- quence not observed by us in any other scombroids. Both in Scomber (Scombrini) and Acanthocybium (Sardini), initial ossification of the neural and haemal arches and spines and the hypural complex started at four locations on the vertebral column (Kramer 6Watson, W. Larval development of Enchelyurus brunneolus from Hawaiian waters (Pisces: Blennidae: Omobranchini). Un- publ. manuscr. Marine Ecological Consultants of Southern California, 533 Stevens Avenue, Soloma Beach, CA 92075. 6Research on the development of Lutjanus campechanus is in progress at the Southeast Fisheries Center Miami Laboratory, Na- tional Marine Fisheries Service, NOAA, 75 Virginia Beach Drive, Miami, FL 33149. 675 FISHERY BULLETIN: VOL. 84, NO. 3 1960). In other scombroids initial ossification was only anterior and posterior (S comber omorus, Sarda ?, Thunnus, Istiophoridae) or only anterior (Scom- brolabrax, Gempylidae, Trichiurus, Xiphias). We believe that the relationship of Acanthocybi- um to the Sardini should be re-examined in the future. The Scombrini and Scombrolabrax (Figs. 1, 9) share a primitive development in which the second dorsal fin, anal fin, and pterygiophores develop first from a center anteriorly and posteriorly, and the first dorsal fin and pterygiophores develop second, from a center anteriorly and posteriorly in Scom- brolabrax, but posteriorly only in Scomber except for the first dorsal fin spine, which was added later. The Gempylidae, Thunnini, and Scomberomorus (Figs. 2, 3, 4, 10, 13) share an advanced develop- ment in which the first dorsal fin and pterygiophores develop first from the anteriormost element in a posterior direction, and the second dorsal fin, anal fin, and pterygiophores develop second from a center anteriorly and posteriorly, the first dorsal fin being separate from the second dorsal fin during part of the ontogeny. In Acanthocybium, Sarda, and Thunnini, the development is similar to the advanced development of the Gempylidae and Scomberomorus except in Acanthocybium, Sarda, and Thunnini, the second dorsal fin spine developed first, the first dor- sal fin spine was added later. The first dorsal fin was separate for part of the ontogeny from the second dorsal in Acanthocybium, but we were unable to observe this in Sarda because of the lack of an ade- quate size series. In Trichiurus (Fig. 5), the dorsal fin and pterygiophores develop from the anterior- most element posteriorly. When dorsal fin develop- ment reaches above the anal fin, the anal fin develops from its anteriormost element in a poste- rior direction. Dorsal and anal fin development then proceed posteriorly at about the same pace. Tri- chiurus has a peculiar developmental feature, which was not observed in any other scombroid. It was that the anteriormost dorsal fin spines and anal spine and rays develop before their corresponding pterygio- phores. Pterygiophore development soon overtook fin ray development and during further development more pterygiophores are present than fin rays. In the Istiophoridae and Xiphiidae, dorsal and anal fin development differ from the previously described groups. In the Istiophoridae (Fig. 14) the first dor- sal fin and pterygiophores develop first from a center anteriorly and posteriorly. When the poste- rior portion of the first dorsal fin development reaches above the anterior portion of the anal fin, anal rays and pterygiophores are added mostly pos- teriorly, although a few elements develop in an anterior direction. The second dorsal fin develops only in a posterior direction consecutive to the first dorsal fin. In Xiphias (Fig. 15), the second dorsal and anal fins and pterygiophores develop first from a center anteriorly and posteriorly. Development of the first dorsal fin and pterygiophores then is con- tinuous with the second dorsal fin in an anterior direction only. The hypurals in all scombroids develop as separate cartilages. Only in Scombrolabrax is there no fusion of the hypurals in the adults. In the Gempylidae the extent of the hypural fusion varies for different genera and we did not observe fusion in the carti- laginous state. For the trichiurids with tails, not enough specimens were available to make observa- tions on hypural fusion. In the remaining scombroids (Scombridae, Istiophoridae, Xiphiidae) hypurals 1-4 are fused to one hypural plate in adults. Fusion to one hypural plate came about during ontogeny by fusion of hypurals 1 and 2 to a ventral and hypurals 3 and 4 to a dorsal hypural plate, with subsequent fusion of these into one plate. For the ventral plate, cartilaginous fusion occurs in all tribes of the Scom- bridae, but in the Istiophoridae fusion is either from cartilaginous or ossifying hypurals 1 and 2 and in Xiphias it is always from ossifying hypurals (Table 1). In Scomber, Acanthocybium, and Istiophoridae, the fusion of hypurals 3 and 4 to the dorsal hypural plate is variable and occurs either during the carti- laginous or ossifying state. In Sarda three speci- mens have fusion of hypurals 3 and 4 in the ossify- ing state. In Scomberomorus and Thunnus the fusion to the dorsal hypural plate occurs always in the car- tilaginous state, whereas in Xiphias it is always in the ossifying state (Table 1). The number of centra supporting the caudal rays varies in the scombroids. In Scombrolabrax, Gem- pylidae, Trichiuridae with tails, Scomber, and Istio- phoridae, three vertebrae (including the urostyle) support the caudal rays. In Xiphias only two verte- brae support the rays. In the Scombridae more vertebrae are involved with the support of the caudal rays, except in Scomber. In the Scombero- morus species examined by us, five centra support the rays, but in some species of Scomberomorus only four centra are involved (Collette and Russo 1984). In Acanthocybium (Collette and Russo 1984) and Sarda, five centra are involved with the support of the rays, whereas in Thunnus only four centra sup- port the caudal rays (Table 1). Johnson (fn. 3; in press) is of the opinion that Scombrolabrax does not belong in the Scombroidei because it lacks most defining specializations of this 676 POTTHOFF ET AL.: DEVELOPMENT OF SCOMBROID FISHES group. Bond and Uyeno (1981) removed Scombrola- brax from the Scombroidei on the basis of one spe- cialized character. We are of the opinion that Scom- brolabrax should be retained in the Scombroidei until we fully understand the significance of devel- opmental characters. Scombrolabrax shares many characters with other scombroids, in particular the absence of predorsal bones coupled with the ante- rior pterygiophore interneural insertion sequence, the saddle-shaped ossifications of the vertebrae, the sequence of neural and haemal arch and spine devel- opment and the striking resemblance of Scombro- labrax to Thunnini larvae. Gempylid and trichiurid relationships await fur- ther study when complete series of larvae of more species become available. We believe that Gempy- lus and Diplospinus are similar and very closely related. We also believe that the gempylids and tri- chiurids are very closely related, the trichiurids representing an advanced gempylid group. Johnson (in press) has discovered a specialization (a stay on the 4th pharyngobranchial) unique to the Scombridae, Istiophoridae, and Xiphiidae but absent in other Perciformes. From our study we believe that the billfish (Xiphias and Istiophoridae) do not belong in the Scombroidei because they differ in many developmental and meristic characters from other scombroid members (Tables 1, 2). However, until more developmental studies are done to deter- mine the meaning and significance of developmen- tal characters, it would be premature to suggest rearranging the Scombroidei. The full value of early developmental studies for systematic purposes will be realized when similar studies have been completed on a greater variety of fishes. Only then will we be able to interpret the meaning and significance of some developmental characters presented here. ACKNOWLEDGMENTS We thank G. L. Beardsley, B. B. Collette, A. C. Jones, G. D. Johnson, and W. J. Richards for critical- ly reading the manuscript and P. Fisher for typing many drafts of the manuscript. We thank B. B. Col- lette, R. H. Gibbs, M. F. Gomon, G. D. Johnson, W. J. Richards, and J. L. Russo for providing gempy- lid and trichiurid fishes for clearing and staining. The Scomberomorus and Acanthocybium material was loaned to us by M. Leiby and J. Gartner from the SEAMAP collections. M. P. Fahay, G. H. Moser, and B. Sumida MacCall generously provided Scomber and Sarda specimens. LITERATURE CITED Bond, C. E., and T. Uyeno. 1981. Remarkable changes in the vertebrae of perciform fish Scombrolabrax with notes on its anatomy and systematics. Jpn. J. Ichthyol. 28:259-269. Collette, B. B., and L. N. Chao. 1975. Systematics and morphology of the bonitos (Sarda) and their relatives (Scombridae, Sardini). Fish. Bull., U.S. 73: 516-625. Collette, B. B., and J. L. Russo. 1984. Morphology, systematics and biology of the Spanish mackerels (Scomberomorus, Scombridae). Fish. Bull., U.S. 82:545-692. Collette, B. B., T. Potthoff, W. J. Richards, S. Ueyanagi, J. L. RUSSO, AND Y. NlSHIKAWA. 1984. Scombroidei: development and relationships. In H. G. Moser, W. J. Richards, D. M. Cohen, M. P. Fahay, A. W. Kendall, Jr., and S. L. Richardson (editors), Ontogeny and systematics of fishes, p. 591-620. Am. Soc. Ichthyol. Herpetol., Spec. Publ. 1. Conrad, G. M. 1938. The osteology and relationships of the wahoo (Acan- thocybium solandri), a scombroid fish. Am. Mus. Nov. 1000, p. 1-32. Fritzsche, R. A., and G. D. Johnson. 1980. Early osteological development of white perch and striped bass with emphasis on identification of their larvae. Trans. Am. Fish. Soc. 109:387-406. Gosline, W. A. 1968. The suborders of perciform fishes. Proc. U.S. Natl. Mus. 124(3647):l-78. Greenwood, P. H., D. E. Rosen, S. H. Weitzman, and G. S. Myers. 1966. Phyletic studies of teleostean fishes, with a provisional classification of living forms. Bull. Am. Mus. Nat. Hist. 131:341-355. Gregory, W. K., and G. M. Conrad. 1937. The comparative osteology of the swordfish (Xiphias) and the sailfish (Istiophorus). Am. Mus. Novit. 952, p. 1- 25. Johnson, G. D. 1975. The procurrent spur: an undescribed perciform caudal character and its phylogenetic implications. Occas. Pap. Calif. Acad. Sci. 121, p. 1-23. In press. Scombroid phylogeny: an alternative hypothesis. Bull. Mar. Sci. 39. Kohno, H., M. Shimizu, and Y. Nose. 1984. Morphological aspects of the development of swimming and feeding functions in larval Scomber japonicus. Bull. Jpn. Soc. Sci. Fish. 50:1125-1137. Kramer, D. 1960. Development of eggs and larvae of Pacific mackerel and distribution and abundance of larvae 1952-56. U.S. Fish Wildl. Serv., Fish. Bull. 60:393-438. Leccia, F. M. 1958. The comparative osteology of the scombroid fishes of the genus Scomberomorus from Florida. Bull. Mar. Sci. Gulf. Caribb. 8:299-341. Leis, J. M., and W. J. Richards. 1984. Acanthuroidei: development and relationships. In H. G. Moser, W. J. Richards, D. M. Cohen, M. P. Fahay, A. W. Kendall, Jr., and S. L. Richardson (editors), Ontogeny and systematics of fishes, p. 547-551. Am. Soc. Ichthyol. Herpetol. Spec. Publ. 1. 677 FISHERY BULLETIN: VOL. 84, NO. 3 Matsubara, K., and T. Iwai. 1958. Anatomy and relationships of the Japanese fishes of the family Gempylidae. Mem. Coll. Agric. Kyoto Univ., Fish. Ser. Spec. No., p. 23-54. Matsumoto, W. M., E. H. Ahlstrom, S. Jones, W. L. Klawe, W. J. Richards, and S. Ueyanagi. 1972. On the clarification of larval tuna identification, par- ticularly in the genus Thunnus. Fish. Bull, U.S. 70:1-12. Merrett, N. R. 1971. Aspects of the biology of billfish (Istiophoridae) from the equatorial western Indian Ocean. J. Zool. 163:351-395. Nakamura, I., and E. Fujii. 1983. A new genus and species of Gempylidae (Pisces: Per- ciformes) from the Tonga Ridge. Publ. Seto Mar. Biol. Lab. 27(4/6, Art. 10):173-191. Okiyama, M., and S. Ueyanagi. 1978. Interrelationships of scombroid fishes: an aspect from larval morphology. Bull. Far Seas Res. Lab. 16, p. 103-113. Potthoff, T. 1975. Development and structure of the caudal complex, the vertebral column, and the pterygiophores in the blackfin tuna (Thunnus atlanticus, Pisces Scombridae). Bull. Mar. Sci. 25:205-231. 1984. Clearing and staining techniques. In H. G. Moser, W. J. Richards, D. M. Cohen, M. P. Fahay, A. W. Kendall, Jr., and S. L. Richardson (editors), Ontogeny and systematics of fishes, p. 35-37. Am. Soc. Ichthyol. Herpetol., Spec. Publ. 1. Potthoff, T., and S. Kelley. 1982. Development of the vertebral column, fins and fin sup- ports, branchiostegal rays, and squamation in the swordfish, Xiphias gladius. Fish. Bull., U.S. 80:161-186. Potthoff, T., and W. J. Richards. 1970. Juvenile bluefin tuna, Thunnus thynnus (Linnaeus), and other scombrids taken by terns in the Dry Tortugas, Florida. Bull. Mar. Sci. 20:389-413. Potthoff, T., W. J. Richards, and S. Ueyanagi. 1980. Development of Scombrolabrax heterolepis (Pisces, Scombrolabracidae) and comments on familial relation- ships. Bull. Mar. Sci. 30:329-357. Richards, W. J. 1974. Evaluation of identification methods for young bill- fishes. In R. S. Shomura and F. Williams (editors), Proceedings of the International Billfish Symposium, Kailua- Kona, Hawaii, 9-12 August 1972, Part 2. Review and con- tributed papers, p. 62-72. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-675. Richards, W. J., and W. L. Klawe. 1972. Indexed bibliography of the eggs and young of tunas and other scombrids (Pisces, Scombridae), 1880-1970. NOAA Tech. Rep. NMFS SSRF-652, 107 p. Richards, W. J., and T. Potthoff. 1974. Analysis of the taxonomic characters of young scom- brid fishes, genus Thunnus. In J. H. S. Blaxter (editor), The early life history of fish, p. 623-648. Springer-Verlag, Berlin, Heidelberg, New York. Sato, S. 1983. Identificacao, Distribuicao e desenvolvimento larval de "Lanceta" Thyrsitops lepidopoides (Cuvier, 1931) (Pisces: Gempylidae) da regiao compreendida entre cabo frio (23 °S) e cabo de sta. Marta Grande (29°S). M.S. Thesis, Univ. Sao Paulo, Brasil, 64 p. 678 AGE AND GROWTH OF THE MARINE CATFISH, NETUMA BARBA (SILURIFORMES, ARIIDAE), IN THE ESTUARY OF THE PATOS LAGOON (BRASIL)1 Enir Girondi Reis2 ABSTRACT Otolith cross sections from Netuma barba were used for age and growth determinations. There is close agreement between average back-calculated lengths and average observed lengths determined from otoliths at capture for each year class. One opaque and one hyaline zone is formed annually. The hyaline zone appears to be formed during the breeding season when the estuarine mature population is scarcely feeding. Von Bertalanffy growth parameters were estimated through Beverton's method which showed the smallest residual variance between observed and calculated lengths for year class. The growth equa- tion (mm) is Lt = 638 [1 - exp (-0.1287(f + 0.195))]. The largest specimen observed was a 980 mm female, 36 years old. The life span of N. barba was estimated to be 23.1 years and the natural mortality rate 0.13. The sea catfish, Netuma barba (Lacepede 1803), ranges in the western Atlantic from Bahia (lat. 17°00'S) in Brasil (Gunther 1864) to San Bias (lat. 40°32'S) Argentina (Lopez and Bellisio 1965). It is the second most important estuarine fishery re- source in the Patos Lagoon and is caught with gill nets (Reis 1982a). The species accounts for about 29% of the total fish landings in the estuary from October to December, a period when it migrates from the sea to spawn. During the remaining months the species is dispersed in low abundance in the ocean (Reis in press). Observations on Netuma barba in Brasil have been restricted to taxonomy (Higuchi et al. 1982) and to feeding and reproduc- tion (Ihering 1888, 1896; Nomura and Menezes 1964; Reis in press). Age determinations in catfishes are usually based on reading vertebrae and pectoral or dorsal spines (Pantulu 1962; Tweddle 1975). Pectoral spines of Netuma barba were not used in the present study because they showed inconsistencies in age deter- mination. However, a preliminary investigation revealed the presence of clear and readable zones in otoliths. This paper deals with the interpretation of these zones, the possible causes of zone forma- tion, and the determination of growth of Netuma barba in the estuary of the Patos Lagoon. 'Based on a thesis in partial fulfillment of the requirements for the MS degree, Fundacao Universidade do Rio Grande - Rio Grande (Brasil). 2Departamento de Oceanografia, Fundacao Universidade do Rio Grande, Caixa Postal 474, 96200 - Rio Grande - RS, Brasil. MATERIALS AND METHODS Study Area The Patos Lagoon, the largest lagoon system in southern Brasil (10,360 km2), is connected to the Atlantic Ocean by a narrow access canal (Fig. 1). The estuary of the lagoon serves as a breeding, nursery, and feeding ground for most of the coastal fish which migrate through the canal and represent a significant percentage of the national fishery resource. Collections of adult Netuma barba were made from fish-processing plants located in the estuarine zone of the lagoon, off the coast of Rio Grande to Sao Lourenco do Sul, a town located 94 km inland (Fig. 1). Juveniles were collected by special research surveys carried out in the estuary. Data were col- lected from September 1977 to December 1980 on 4,120 specimens. No samples were available from January to March because of a closed fishing season of Ariidae in the area, and few samples were col- lected from April to July due to the absence of the species in the estuary. Sampling Procedure Specimens were measured (total length, mm), weighed (g), and sexed. Lapillus otoliths were re- moved, sectioned transversally next to the nucleus, polished, and were examined under a 10 x binocu- lar microscope. The dorsal, polished half of the otoliths was observed with transmitted light. The Manuscript accepted January 1986. FISHERY BULLETIN: VOL. 84, NO. 3, 1986 679 FISHERY BULLETIN: VOL. 84, NO. 3 ESTUARY OF THE PATOS LAGOON 32?- S. LourtnfO do Su PATOS LAGOON RIO GRANDE do SUL »* >v 53< 52° 5I( Figure 1.— Coastal lagoon system of southern Brasil (A) and the study area (B). type of deposit (opaque or hyaline) on the otolith margin and the number of hyaline zones were re- corded for each otolith. Back-calculation was done over the surface, the total length of the otolith (Co) and the length between the nucleus and each hya- line zone (ci) (Fig. 2) were measured with an ocular micrometer. The term nucleus used here refers to the central area of the otolith limited by the first zone (Jearld 1983). Growth curves for males and females were cal- culated using the mean lengths for year class. The parameters of the von Bertalanffy growth equation were determined: Lt = Lm [1 - e-^'-W] (1) where Lt is the total length at time t, LM is the maximum attainable size, K is the growth coeffi- 680 REIS: AGE AND GROWTH OF MARINE CATFISH VS. a.e. s.e. d.s. Figure 2.— A lapillus otolith oiNetuma barba showing opaque ( + ) and hyaline (b) zones, the nucleus (N), the axes where back- calculation was made (Co = distance between the nucleus and the otolith's edge; ci = distance from the nucleus to "i" hyaline zone) and the position of otolith on fish head (a.e. = antisulcal end; s.e. = sulcal end; d.s. = dorsal surface; v.s. = ventral surface; and hyaline zones = I-IX). cient, and t0 a correction on the time axis. The parameters of Equation (1) were estimated by deter- mining the predictive regression of ln(Loo - Lt) against t (Beverton 1954): lnCZ^ - Lt) = \nL„ + K(t0 - t) (2) where K is the slope of the regression line and the ^/-intercept of Equation (2) can be equated to In Lm + Kt0 providing the value of t0 (Ricker 1975). Trial plots, including values of Lm first derived by the methods of Walford (1946) and Gulland (1964), yielded the L^ which gives the straightest line. The agreement between observed and calculated lengths for year class was determined by residual variance (S2y) expressed by „,„ Z. (observed Lt - calculated Lt)2 o y = (o) N - 1 where N is the number of age classes. Length-weight relationship was determined for males and females Wt = t*Ltv (4) where Wt is the weight at time t, and \x and v the coefficients of the functional regression between Wt and Lt (Ricker 1973). The condition factor was calculated for each sex as follows: K = Wt Ltv (5) Wt = W*, [1 - e-K«-y]" (6) expressed growth in weight, where W„ is the max- imum attainable weight obtained by solving for L^ in Equation (4). The life span was estimated: ^•0.95 - ^0 ln(l - P) K (Taylor 1960) (7) where A0 95 is the time required to attain 95% of Loo, P = 0.95 and t0 and K are derived from the growth equation. The natural mortality coefficient (M) was estimated according to Taylor (1960) M = ln(l - P) A0. (8) 95 and Statistical analyses were done when necessary (Snedecor and Cochran 1970; Sokal and Rohlf 1981). RESULTS AND DISCUSSION Age Determination The lapillus otolith used for the determination of age oiNetuma barba is the most developed ear bone in the Ariidae (Stinton 1975), its length attaining 3% of fish fork length (Reis 1982b). Growth zones can be observed on a sectioned otolith from the sulcal to the antisulcal end and from the nucleus on the dorsal face to the ventral one (Fig. 2). The hya- line and opaque zones are clearly evident even in otoliths of old specimens. Under transmitted light the opaque zones, or fast-growth zones, are white (broad) and hyaline zones, or slow-growth zones, are dark (narrow) (Fig. 2). Warburton (1978) counted growth checks on whole otoliths of Galeichthys caerulescens (Giinther), and Dmitrenko (1975) studied Arius thalassinus (Riippel) by viewing the otoliths the same way as in the present paper. The number of hyaline zones on sectioned otoliths and of growth checks observed on whole otoliths (War- burton 1978) were compared. A smaller number of growth checks was encountered in all cases when using whole otoliths. In the present study only 2.4% of the otoliths were considered illegible. About 60% agreement was ob- tained when otoliths were read on two different oc- casions separated by a month. Disagreement was due to the inability to distinguish the first hyaline zone and those near the otolith's edge. When the same otoliths were analyzed for the third time, the 681 FISHERY BULLETIN: VOL. 84, NO. 3 agreement between observations increased to 79.9%. Time of Zone Formation The percentage of hyaline and opaque edged oto- liths was plotted for each month (Fig. 3). Otoliths showing hyaline edge are more abundant in Decem- ber when they comprise 63.7% of the total; opaque edged otoliths are fewer in this month (33.9%). Stu- dent's £-test (Snedecor and Cochran 1970) showed that proportions between hyaline and opaque edged otoliths differ significantly (P < 0.05) for most months (Fig. 3). Also, the mean width of the opaque zone on the otolith edge decreased towards the end of the year (Fig. 4), indicating a recent hyaline zone formation. The period of zone formation is not the same for all individuals, the result of individual growth differences. It is evident, however, that only one hyaline and one opaque zone is formed each year. Formation of slow-growth zones during warm months in Netuma barba coincides with the spawn- ing period and the cessation of feeding activity (Reis in press). Both events suggest a decrease or a pause in growth when hyaline zones are formed. Menon (1953) observed that decreased feeding and gonad maturation may cause a periodic formation of the growth marks in skeletal parts of fish. During the too 90 80- j 70 Ul 0 60 ui ec "■ 50 ui «* 40 § 30 e Ul 01 20 10 / \ / \ / \ — HYALINE EDGE -x OPAQUE EOGE 155 2W 179 OCT. NOV. DEC. 1977 48 74 22 12 75 42 73 44 APP..MAY.JUN. JUL. AUG. SEPT. OCT. NOV.DEC. 1978 35 23 76 82 67 AUG.SEPTQCT. NOV DEC. 1979 65 175 251 200 n SEPT.OCT.NOV. DEC. MONTH 1980 Figure 3.— Percentage of hyaline and opaque edge on otoliths of Netuma barba related to the months of four years (* P < 0.05; n = number of specimens). z o N 4- o o ui X X \- o MALES t SEPT. 55 — i — OCT. 101 ft 80 ft FEMALES fl 26 71 NOV. DEC. T SEPT. OCT. 141 — I — NOV. 94 n — i DEC. MONTH Figure 4.— Mean width and confidence limits at P < 0.05 level of the last opaque zone on otolith's edge for males and females of Netuma barba (n = number of specimens). 682 REIS: AGE AND GROWTH OF MARINE CATFISH remaining months, when Netuma barba is at sea actively feeding and the gonads are resting (Reis in press), the opaque zones appear to be laid down due to fast somatic growth. According to Pannella (1974), fishes of temperate environments tend to form opaque zones or fast-growth zones during warm months but the synonymity of the terms sum- mer and opaque, winter and hyaline has to be demonstrated in each instance rather than accepted as a general fact. For Netuma barba slow-growth zones are formed during warm months and may be related to the maturation of the gonads and a pause in feeding activity. Gonad maturation may be one of the causative factors of hyaline formation in adults; however, a plausible cause still needs to be established for immature specimens. GROWTH Growth in Length Sectioned otolith lengths (measured as shown in Figure 2), and fish lengths were best fitted to the power curve: Lt = 1.89 Co1047 r = 0.960; n = 689, and the equation for back-calculation was , 1.047 Lt i = Lt ~Co where Lt % is the length of fish when zone "i" was formed. Observed and back-calculated mean lengths for year class for each sex increase as one opaque and one hyaline zones are formed in the otolith each year (Table 1). Up to age 11, the mean lengths are similar; older females had mean lengths greater than males. The same was true for mean weight although a small number of specimens were analyzed from age 11 onward. Observed lengths are usually higher than back-calculated lengths except in ages that few specimens were analyzed. Lengths corresponding to ages 8 to 12 are most frequent in the samples since they are most affected by the mesh size of the fishing gear used in the estuary. Mean observed lengths at these ages agree closely with mean back-calculated lengths (Fig. 5) for both sexes. Gill nets are highly size selective and retain fish at lengths of 370-520 mm (Reis 1982a). The analysis of variance (Sokal and Rohlf 1981) showed that observed lengths at ages 5, 6, and 7 are significantly higher than back-calculated lengths (P < 0.05) which could be due to the capture of the largest specimens of these ages since the minimum size of fish held by gill nets depends on the maxi- mum body girth (opercle). Mean back-calculated lengths showed no definitive tendencies for any age class (Fig. 6) indicating no growth changes. Fur- Table 1 .—Mean observed and back-calculated lengths of males and females of Netuma barba for each age class (sample size in parentheses). Estimated age Mean observed length Male Female Mean back-calculated length Male Female 1 84 96 65 63 ( 10) | 4) (310) (370) 2 145 152 140 137 ( 31) | 21) (303) (371) 3 203 197 193 192 ( 40) 50) (295) (362) 4 228 261 244 245 ( 7) 2) (291) (359) 5 348 146 300 299 ( 16) 1) (286) (357) 6 378 365 347 347 ( 28) 11) (281) (349) 7 403 394 386 385 ( 59) 56) (263) (336) 8 415 416 413 412 (144) 111) (241) (313) 9 433 431 430 433 (291) 281) (140) (193) 10 452 463 444 446 ( 94) 120) ( 79) (129) 11 476 493 455 462 ( 73) 138) ( 31) ( 54) 12 490 526 460 507 ( 57) 51) ( 8) ( 8) 13 464 602 508 612 ( 15) 23) ( 4) ( 3) 14 551 667 480 637 ( 10) 21) ( 3) ( 2) 15 522 622 520 578 ( 13) 5) ( 2) ( 1) 16 533 620 528 608 ( 10) 3) ( 1) ( 1) 17 494 647 546 637 ( 2) 2) ( 1) ( 1) 18 620 714 564 657 ( 1) ; 1) ( 1) ( 1) 19 588 554 — 666 ( 4) 3) — ( 1) 20 550 860 — 696 ( 1) 1) — ( 1) 21 520 520 — 706 ( 2) 2) — ( 1) 22 490 649 — — ( 1) 1) — — 23 — 736 ■n — — 24 680 ( 1\ '/ — — 36 930 980 — — ( 1) 1) — — 683 FISHERY BULLETIN: VOL. 84, NO. 3 Figure 5.— Mean observed and back-calculated lengths for year class of Netuma barba (* P < 0.05). | 500 h 400 o y 300 < 200 O •- 100 .-^1*-* -*-~ .-3P* S* S -+■ BACK-CALCULATION -OBSERVED DATA I 'I I I T I" I 2 3 4 5 6 7 8 9 10 II 12 ESTIMATED AGE 40 20 40 20 40 20 40 20 1st HYALINE ZONE 2nd HYALINE ZONE ESTIMATED AGE '\. / / \ f I * f I "^^^T" / n- II n-19 T ■ T » ■ V r-£ n»20 40- v o 2 UJ 2 20 cc b. UJ ° 40 Z^y\ v n=34 ■i r- ■ i"1 / /• S r \ ./ . — ■- n-206 2 UJ cc 20 40 20 40 20 / /' / \ n*l36 :/v r / \ V n-126 VL \. \ n«94 40' 20 vx I * I n«!8 10 12 30 70 90 110 130 90 110 130 150 170 190 210 TOTAL LENGTH (mm) Figure 6.— Back-calculated lengths frequencies at first and second hyaline zones for year class of Netuma barba (n = number of specimens). thermore, as the modes for each year class are similar, age determination can be considered consistent. Validation of Age Validation of the otolith method for aging Netuma 684 REIS: AGE AND GROWTH OF MARINE CATFISH barba is supported by the following: 1) one opaque and one hyaline zone is formed annually (Figs. 3, 4); 2) a gradual decrease of length increments with age (Table 1); 3) observed lengths generally agree with back-calculated lengths (Fig. 5); and 4) distri- bution of back-calculated lengths for previous ages shows similar modes for each year class (Fig. 6). Length-weight Relationship and Condition Factor A total of 685 specimens captured during 1980 was used to compute the length-weight relationship for each sex: Male Wt = 4.70 x 10"6 Lt3u Female Wt = 2.19 x 10"6 Lt326 Total Wt = 4.41 x 10"6 Lt315 r = 0.992 n = 332 r = 0.952 n = 363 r = 0.987 n = 685 The analysis of covariance (Snedecor and Cochran 1970) at P < 0.05 level showed significant difference only for the \i value, and for that reason condition factor (K) was determined for each sex. There is a decrease of mean K values towards the end of the year (Fig. 7). The condition factor for males is always higher probably due to a more intense feed- ing prior to reproduction. Low K values reveal the stress the fish suffers when it is scarcely feeding and fat reserves are being diverted to gonad maturation (Reis in press), thereby causing a cessation of growth. I proposed that K values for males will sharply decrease after spawning due to an oral in- cubation period that lasts 1 to 2 mo and prevents males from feeding (Reis in press). Calculation of Growth Parameters cr o t- z o Q Z o o cc o b. O o fl 8 fl ft MALES & m hlO o m z H O I I> 10 34 35 54 66 56 44 FEMALES m 3J O o i > 10? 26 H SEPT. 36 48 I TC OCT. 72 85 X JL NOV. 32 X 68 n DEC. MONTH Figure 7.—K condition factor and percent change for males and females of Netuma barba related to time (n = number of specimens; I = first half of the month; II = second half of the month) Von Bertalanffy growth parameters were esti- mated by Beverton's (1954) method which presented the smallest residual variance between observed and calculated lengths for year class on the ages that are most affected by gear selectivity (8-12 yr old). For fish populations captured from a certain age on- ward, the smallest residual variance should be sought for all year classes from age at first capture. For Netuma barba the smallest residual variance could not be ascertained by this method because the true length distribution is unknown due to the use of gillnets as fishing gear. Growth equation for age 1 to 12 for both sexes in represented by Lt = 638 [1 - e-o.i287(t+o.i95)]_ Figure 8 shows both calculated and observed lengths for each year class. Growth in weight for each sex resulted in Male Female Wt = 2981.89 [1 - e-°-1287200 2100- S .^" Loo=638nrn * * Colculoted Ltngtht K =0,1287 . _ Observed Length* I 2 3 4 5 6 7 8 9 10 II 12 ESTIMATED AGE Figure 8.— Growth curve of Netuma barba. 685 FISHERY BULLETIN: VOL. 84, NO. 3 Maximum Size and Age, Life Span, and Mortality Rate Netuma barba is a long lived, slow growing species with a low mortality rate. Specimens as long as the theoretical mean length (638 mm) are frequently captured. The largest catfish observed was a 980 mm female 36 yr old. Netuma barba life span was estimated to be 23.1 yr and its mortality rate was 0.13. I assumed that the estimate of M (natural mortality) is accurate, since Netuma barba reveals a long life span, a capacity to avoid predation through the defense represented by its hard dorsal and pectoral spines and a parent-juvenile care behavior (Reis in press). Pauly (1980) suggested that species with low mortality rates are related to high Loo values and to low growth coefficients. These characteristics combined with the fact that Netuma barba has a low fecundity (Reis in press) define the species as /f- strategists (Gunderson 1980). ACKNOWLEDGMENTS I am greatly indebted to J. P. Castello for his assistance and valuable suggestions and to the staff of Fisheries Biology Laboratory, Department of Oceanography, Fundacao Universidade do Rio Grande. LITERATURE CITED Beverton, R. J. H. 1954. Notes on the use of theoretical models in the study of the dynamics of exploited fish populations. U.S. FisKLab., Beaufort, N.C., Misc. Contrib. 2, 159 p. Dmitrenko, E. M. 1975. Size-age composition of the giant catfish, Arius thalassinus in the vicinity of Kathiawar Peninsula (India). Vopr. Ikhtiol. 15:695-702. GULLAND, J. A. 1964. Manual of methods for fish population analysis. FAO Fish. Tech. Pap. 40, 61 p. Gunderson, D. R. 1980. Using r-K selection theory to predict natural mortality. Can. J. Fish. Aquat. Sci. 37:2266-2271. GUNTHER, A. 1864. Catalogue of the fishes in the British Museum. 5. Catalogue of the Physostomi; Ariina. Br. Mus., p. 138- 182. Higuchi, H., E. G. Reis, and F. G. Araujo. 1982. A new species of marine catfish from the coast of Rio Grande do Sul, with comments on the nominal genus Netuma Bleeker, 1858 of the Southwest Atlantic (Siluriformes, Ariidae). Atlantica 5(1):1-15. Ihering, H., Von. 1888. Ueber brutpluge und Entwicklung des bagre (Arius commersoni). Biol. Cent. 8:268-271. 1896. Os peixes da costa do mar do Estado do Rio Grande do Sul. Rev. Mus. Paul. 2:25-63. Jearld, A., Jr. 1983. Age determination. In L. A. Nielsen and D. L. John- son (editors), Fisheries techniques, p. 301-324. Am. Fish. Soc, Southern Print. Co., VA. LOpez, R. B., and N. B. Bellisio. 1965. Contribucion al conocimiento del Tachysurus barbus (Lacepede), bagre del mar argentino (Pisces, Ariidae). In Anais II Congreso Latino-Americano Zoologia, p. 145-153. Menon, M. D. 1953. The determination of age and growth of fishes of tropical and subtropical waters. J. Bombay Elist. Sec. 51: 623-635. Nomura, H., and N. A. Menezes. 1964. Peixes marinhos. In P. E. Vanzolini (editor), Hist6ria natural dos organismos aquaticos do Brasil, p. 343-386. Sao Paulo. Pannella, G. 1974'. Otolith growth patterns: An aid in age determination in temperate and tropical fishes. In T. B. Bagenal (editor), The ageing of fish, p. 13-27. Proc. Int. Symp. Ageing Fish, Univ. Reading, Engl.; Unwin Brothers, Ltd., Engl. Pantulu, V. R. 1962. On the use of pectoral spines for the determination of age and growth of Pangasius pangasius (Hamilton Buch). J. Cons. Int. Explor. Mer 27:192-216. Pauly, D. 1980. On the interrelationship between natural mortality, growth parameters, and mean environmental temperature in 175 fish stocks. J. Cons. Int. Explor. Mer 39:175-192. Reis, E. G. 1982a. Idade, crescimento e reproducao de Netuma barba (Siluriformes, Ariidae) no estuario da Lagoa dos Patos (RS). M.S. Thesis, Fundacao Universidade do Rio Grande, Brasil, 114 p. 1982b. Anatomy of the inner ear of Netuma barba (Lacepede, 1803), Siluriformes, Ariidae. Atlantica 5(1): 16-22. In press. Reproduction and feeding habits of the marine cat- fish Netuma barba (Siluriformes, Ariidae) in the estuary of the Lagoa dos Patos, Brasil. Atlantica. Ricker, W. E. 1973. Linear regressions in fishery research. J. Fish. Res. Board Can. 30:409-434. 1975. Computation and interpretation of biological statistics of fish populations. Fish. Res. Board Can. Bull. 191:1-382. Snedecor, G. W., and W. G. Cochran. 1970. Metodos estadisticos aplicados a la investigacion agri- cola y biologica. 3d ed. Ed. Cont., Mex., 626 p. SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry. 2ded. W. H. Freeman & Co., N.Y., 859 p. Stinton, F. C. 1975. Fish otoliths from the English Eocene. Palaentogr. Soc. Monog. (Lond.), p. 1-56. Taylor, C. C. 1960. Temperature, growth and mortality - the Pacific cockle. J. Cons. Int. Explor. Mer 26:117-124. TWEDDLE, D. 1975. Age and growth of the catfish Bagrus meridionalis Giinther in southern Lake Malawi. J. Fish. Biol. 7:677-685. Walford, L. 1946. A new graphic method of describing the growth of ani- mals. Biol. Bull. (Woods Hole) 90:141-147. Warburton, K. 1978. Age and growth determination in a marine catfish using an otolith check technique. J. Fish. Biol. 13:429-434. 686 MONITORING THE SEA SURFACE CHLOROPHYLL CONCENTRATION IN THE TROPICAL PACIFIC: CONSEQUENCES OF THE 1982-83 EL NINO Yves Dandonneau1 ABSTRACT The sea surface chlorophyll concentration (SSCC) is routinely measured in the tropical Pacific using filtra- tions made aboard merchant ships that sail from New Caledonia to Japan, North America, Panama, New Zealand, and Australia. About 4,000 measurements are collected every year, allowing a tentative monitor- ing of SSCC in the Pacific. Heavy smoothing made it possible to map quarterly charts of SSCC which cover the 1982-83 El Nino episode. The usually enriched belt which corresponds to the equatorial upwell- ing vanished after September 1982, except for a reduced zone east of long. 120° W, where a moderate enrichment persisted throughout the warm event. It recovered after July 1983, spreading westwards to long. 170°E. During the mature phase of El Nino (October 1982-June 1983), an enriched zone ap- peared in the western Pacific, centered at about lat. 7°N, consistent with a rise of the thermocline in this region. An examination of oceanographic data collected in this region since 1970 shows that nutrients from below the thermocline are consumed by the phytoplankton during each El Nino. This western Pacific enrichment was weakened with time, and the period from April to June 1983 was characterized by low SSCC values over most of the tropical Pacific. Unusually high SSCC values are reported in subtropical zones, during the austral winters of 1982 and 1983 in the southwestern Pacific and during the 1982 autumn in the northeastern Pacific, which may be due to advection of rich water from higher latitudes and to intensified vertical mixing by strong westerly winds, respectively. El Nino was first observed and experienced in Peru, where it was given its name and became a familiar part of Peruvian life. Although the southern oscilla- tion was identified more than 60 yr ago (Walker 1924), the relation between the El Nino phenomenon and ocean-scale features was only established after the 1957-58 event by Bjerknes (1966). It is now well established that El Nino is simply the most obvious consequence of important oceanographic and meteorological changes in the Pacific Ocean (Donguy and Henin 1976; Quinn et al. 1978; Cane 1983). One would expect biological changes at the same scale. These, however, have only been studied in the eastern Pacific (Walsh 1981; Chelton et al. 1982; Barber and Chavez 1983) where a pronounced decrease in phytoplankton biomass and primary pro- duction is observed. Farther west in the equatorial zone, the decrease in primary productivity has been shown only by indirect observations on marine birds (Schreiber and Schreiber 1984) and on abnormal distributions of some fishing grounds in relation to changes of water mass (Donguy et al. 1978; Yama- naka 1984). The difficult problem of monitoring the intensity of primary production on a large scale is troupe SURTROPAC, Centre ORSTOM, B.P. A5, Noumea, New Caledonia. 4S7 -US' Manuscript accepted January 1986. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. usually reserved for satellite-borne sensors. A modest attempt, however, is in progress, as a part of the SURTROPAC program (ORSTOM, Noumea) based upon chlorophyll samples taken by voluntary observers on ships of opportunity. Each year about 4,000 sea surface chlorophyll concentrations (SSCC) are collected in this way, distributed along maritime lanes from the Tasman Sea to Panama, North America, or Japan. These data cover the tropical Pacific from lat. 30 °S to 30 °N, and from long. 140°E to 80°W. There are large gaps both in space, between the main lanes, and in time, between con- secutive crossings. But, on a quarterly basis, the SSCC data are numerous enough to allow a crude view of the whole tropical Pacific Ocean, with the advantage of using a single methodology. The con- sequences of the 1982-83 El Nino can thus be ex- amined, and most of the attention will be directed towards the central and western Pacific, where pres- ent knowledge is very incomplete. METHODS Chlorophyll Measurements SSCC measurements are made according to a non- extractive method (Dandonneau 1982). Twenty milli- 687 FISHERY BULLETIN: VOL. 84, NO. 3 liters of seawater are filtered on 13 mm HAWP Millipore filters, using a syringe and Swinnex type filtering cartridges. The filters are then stored in a dark place at ambient temperature. When the observing ship reaches Noumea, the filters are taken to the laboratory for fluorescence measurements. A 3-wk minimum time lag is needed between filtra- tion and measurement, after which degradation pro- cesses lead to stable fluorescent chlorophyll by- products on the filters. The fluorescence (Ff) of the filters is then measured without extraction, using a specially adapted sample holder. The measurement error e is proportional to the chlorophyll concentration C and can be expressed as e = |SSCC-C|/C where SSCC is measured by the non-extractive method while C is obtained by a more conventional technique (Holm-Hansen et al. 1965). Ninety-five percent of e values are <0.6 (Dandon- neau 1982, and confirmed by later tests). This value is probably an overestimate of e since it results both from the error on SSCC and from the unknown error on C. Different phytoplankton populations can also result in different fluorescence to chlorophyll ratios for the dry filters. This ratio has shown no signifi- cant change between winter and summer conditions around New Caledonia where a mixed regime alter- nates with a stratified one (Dandonneau and Gohin 1984). The risk of a variation of the ratio in other environments has not been examined, and must be kept in mind. The few SSCC data points at latitudes higher than 30° were not taken into account for this reason. Calibrations SSCC is estimated using SSCC = k Ff where k is a calibration coefficient that must be corrected from time to time. Twenty milliliters from a sea- water sample are filtered giving a fluorescence Ff0 after 21 d of storage. A larger volume V from the same sample is filtered on a glass fiber filter, ground, and extracted by a volume v of 90% acetone. The fluorescence of the extract is Fe0. Knowing the fluorescence to chlorophyll ratio of the fluorometer, R0, determined from a known solution of pure chlorophyll a, we can estimate the following chloro- phyll concentration of the seawater sample: C0 = (Fe0 x v)l(R0 x V); we obtain then k0 = Ff0IC0. k0 is sensitive to detrital material in turbid coastal waters, so these main calibrations are made during offshore oceanographic cruises. As such op- portunities are infrequent, secondary calibrations are made more frequently with known solutions of pure chlorophyll a, giving Rt instead of R0. We then assume that kt = k0 x RJRq. This procedure does not consider correction for chlorophylls b and c, nor does it consider correction for phaeopigments, which has recently proven to be uncertain when the fluor- ometer is fitted with a commonly supplied blue ex- citation lamp (Baker et al. 1983). Although the SSCC data presented in this work are expressed in milli- grams of chlorophyll a, they should be considered only as indices of phytoplankton abundance. Data Rejection The crew members who take the seawater samples and make the filtrations are voluntary observers. Errors may occur which are difficult to detect because, unlike temperature or salinity, 1) any SSCC value in the interval 0-1 mg-m-3, which covers almost the whole data set, is a possible one anywhere in the tropical Pacific, and 2) the auto- correlation of SSCC decreases very quickly with time or space, so that surrounding data cannot help in error detection. Therefore, all the data are ac- cepted, unless the filter exhibits an obvious fault (i.e., breaking, stain, extraneous material). Occa- sionally, all the data from a ship's voyage were evidently too high, by a factor 3 or 5. Contamina- tion by a polluted sampling bucket was the cause, and the data from the entire voyage were rejected. Other possible errors are more insidious, such as insufficient care in keeping the filters out of light, or using an oxidized sampling bucket. These errors result in slightly lowered values, but there is no way to correct them and, in most cases, no way to even detect these biases. Such data are entered in the data bank. As a resulting constraint, any estimate from this SSCC data set must be developed from many data, in order to minimize the effect of a few possibly biased values. Mapping Techniques In a previous work (Dandonneau and Gohin 1984) the principles of objective analysis were applied to compute best estimates of SSCC at a given place and time in the southwestern tropical Pacific. The studied area in the current study is much larger and more complex, and the density of data is not high enough to allow good estimates of the statistics of the field. Hence, the use of an objective analysis of the SSCC data has been excluded. The SSCC mapped here on Figure 1 have been estimated using 688 DANDONNEAU: MONITORING SEA SURFACE CHLOROPHYLL CONCENTRATION 20*N M'S 20*N 20*S . 20*N M'S 20* N 20*S UO'E 100" W 20* N 20*S 100'W Figure 1. -Quarterly charts of SSCC (sea surface chlorophyll concentrations) in the tropical Pacific from January 1, 1982 to December 31, 1983. Areas where SSCC is >0.10 mg nT3 are shaded with large dots. Smaller dots represent the data points. 1=1 1=1 where i3 is the SSCC estimate at longitude Xj and latitude y^, and ptj is the weight given to observa- tion t{ for the estimation tj. p^ is given by Vij = [R2 + (Xi - xf + a2 (yt - t//]"1 where a accounts for anisotropy of the SSCC varia- tions in space. We used a = 2, so that observations at a distance Ay in latitude are given the same weight as observations at a distance kx = 2hy in longitude, p^ was set to zero when (xt - Xjf + a2 (Vi ~ Vj)2 was >160, so that the observations were considered as "non useful" when outside an ellipse centered at (Xj, y^) with a principal axis equal to about 25° longitude, and a small axis equal to about 13° latitude. In order to avoid hazardous estimates at the margin of the contoured area, tj has not been estimated when rij (the number of useful observa- tions) was <12. 689 FISHERY BULLETIN: VOL. 84, NO. 3 R = 0 would give an infinite weight to an obser- vation k available at xk = Xj and yk = jjj. We would then obtain tj = tk regardless of the other observa- tions. This is acceptable only if the instrumental and sampling errors on tk were null, which is not the case. Thus, R accounts for the errors on the obser- vations. We choose R2 = 25, which, together with a = 2 and ptj > (25 + 160)" \ performed an effi- cient smoothing and preserved the large-scale infor- mation. RESULTS The sequence of quarterly mean SSCC for 1982 and 1983 is presented in Figure 1, together with the positions of the data. The western part, north of lat. 20°N, is poorly sampled. The data range between 0.05 and 0.20 mg-m-3. The highest values are found during the northern spring of 1982, and the northern winter of 1983. The 1982 winter, and the spring and fall of 1983 exhibit a few values >0.10 mg-m"3. The 1982 winter and fall show low SSCC, like the summer of both years, below 0.10 mg-m-3. The eastern part, north of lat. 10°N, has gener- ally low SSCC values, often below 0.05 mg-m-3. Exceptions are the spring of 1982 at the extreme north, and, mainly, the fall of 1982 during which the mean values exceeded 0.20 mg-m-3 off California. Low SSCC values are observed in the western part between the Equator and lat. 20°N until the summer of 1982. They are abruptly replaced at the end of 1982 by high values which persist until March 1983. Later, low values, generally below 0.05 mg-m"3, dominate again between lat. 5°N and 20°N, while SSCC >0.10 mg-m"3 shift back south- ward to the Equator. The equatorial zone shows high SSCC in January- March 1982, between America and long. 160° E. Values higher than 0.10 mg-m-3 spread from lat. 10°N and 10°S in the central Pacific, and to 15°S at 120°W. From April to June 1982, the enriched zone shifts eastwards and southwards. The east- wards shift continues between July and September and is accompanied by a decrease of SSCC in the eastern Pacific, with mean values <0.15 mg- nr3. From October 1982 to June 1983, a narrow band with SSCC between 0.10 and 0.15 mg-m"3 in the eastern Pacific is the only remnant of the equator- ial enrichment. A normal situation returned after the El Nino, in July-September 1983, with SSCC values >0.15 mg-m-3 spreading westwards to long. 170°E. In October-December 1983, SSCC >0.10 mg-m-3 are seen all along the Equator. South of lat. 20° S, an SSCC increase is observed during the austral winter. The increase started in April-June in 1982, the maximum was reached in July-September, with SSCC >0.20 mg-m-3 spread- ing northward to 22 °S, and low values were seen again in October-December. The increase during the austral winter of 1983 was of a lesser extent, being well developed only during July-September, with SSCC >0.20 mg-m-3 limited to the south of 28°S. The intermediate zone, from lat. 10°S to 20°S, between the equatorial upwelling and higher lati- tudes where a winter increase is observed, gener- ally has low chlorophyll concentrations, below 0.10 mg-m-3. The lowest concentrations are seen in austral summer, from October 1982 to June 1983, and in October-December 1983. The highest concen- trations are associated with a strengthening of the equatorial upwelling (around long. 140 °W in April- June 1982; westwards spreading of richer waters from the eastern Pacific in July-September 1982 and 1983). When looking at the whole series of maps, the most striking feature is the reduction of the equa- torial upwelling enriched area after the onset of El Nino. The most pronounced stage was in April-June 1983, with poor waters over most of the tropical Pacific. On the contrary, a zone centered at lat. 10°N, west of the dateline, which is usually occupied by chlorophyll-poor waters, had higher SSCC dur- ing the 1982-83 El Nino. DISCUSSION Equatorial Upwelling The collapse of the equatorial upwelling after the onset of El Nino, when westerlies have replaced the trade winds at the Equator, consistently results in a decrease in SSCC. This decrease has already been documented for the eastern Pacific in the Galapagos Islands region by Feldman et al. (1984) using sea color satellite images. It corresponds to a decrease in primary production of the whole photic layer (Barber and Chavez 1983). The data presented here show that the equatorial zone was impoverished westwards to nearly 180°. This is in agreement with the reproductive failure and disappearance of sea- bird communities at Christmas Atoll (lat. 2°N, long. 157°W) in November 1982; Schreiber and Schrei- ber (1984) attributed these events to the establish- ment of an oligotrophic oceanic ecosystem instead of a productive one. Successful reproduction started again for some birds species in June 1983, and hatch- ing occurred in July-September 1983, when SSCC 690 DANDONNEAU: MONITORING SEA SURFACE CHLOROPHYLL CONCENTRATION higher than 0.15 mg-m~3 reappeared at the Equator (Fig.l). Western Pacific Around Lat. 7°N. Under normal conditions (see Figure 1: January to March 1982, July to December 1983) the equa- torial upwelling also drives a chlorophyll-rich zone west of 180°. This does not appear on the map of Koblentz-Mishke et al. (1970) on the primary pro- duction in the world ocean, but is described as an episodic feature by Oudot and Wauthy (1976). The area with SSCC >0.15 mg-m~3 which appears north of the Equator, centered at about 7°N from October 1982 to March 1983 (Fig. 1) has nothing to do with the equatorial upwelling. Based on approx- imately 100 SSCC data points obtained by three dif- ferent merchant ships, this chlorophyll-rich area can hardly be thought to result from measurement errors. It rather may be related to the eastward draining of warm water from the western tropical Pacific and consequent thinning of the surface mixed layer and drop of the sea level (Wyrtki 1985). A simultaneous cooling of the sea surface by 1°C oc- curred in this region during El Nino, which can be explained by advection of cooler water, and also by other potentially important processes which are more difficult to quantify (Meyers and Donguy 1984). The observed SSCC increase supports the hypothesis that vertical mixing of cooler nutrient- rich deep water might be one of these processes. Even if vertical mixing is unlikely, the 50 m rise of the thermocline which has been observed at lat. 7°N between January 1982 and January 1983 (Meyers and Donguy 1984) allows more light to penetrate to the deep chlorophyll maximum. This hypothesis is supported by the shift which occurred between January 1982 and January 1983 in the nitrate-tem- perature relationship (Fig. 2; data collected by the Japan Meteorological Agency along long. 137°E aboard RV Ryofu Maru; Anonymous 1972-84). The nitrate concentration at a given temperature (which we assume to represent a given water mass) dropped by about 2 ^moles-L-1. Shifts in the nitrate-temperature relationship provide informa- tion on the consumption of nitrate by the phyto- 25 20- TCC) N03 (yumole.1"1 ) (♦) January 1982 (♦) January 1972 January 1973 •ITCC) Figure 2.— Nutrient-temperature relationships between lat. 6°N and 9°N. Crosses: observations before an El Nino; open circles: observations after an El Nino. (Data from the RV Ryofu Maru cruises at long. 137°E, Anonymous 1972 to 1984). 691 FISHERY BULLETIN: VOL. 84, NO. 3 plankton (Voituriez and Herbland 1984). We can then suggest that new nitrates have been assim- ilated during El Nino in the western Pacific at lat. 6-9°N. The 2 /imoles-L-1 drop in nitrate concentra- tion is observed in the interval 17°-22°C, corre- sponding to a 35 m thick water layer (Anonymous 1972-84), so that the amount of new nitrates used by photosynthesis is 70 fimoles-m"2, or 980 mg-m-2. If CIN = 9.01 and C/Chl = 114 in surface waters of the oligotrophic central North Pacific (Sharp et al. 1980), this amount of nitrogen corre- sponds to 77 mg Chi a-m-2. It represents an im- portant supply in an ecosystem where the chloro- phyll concentration is usually low. Figure 3 shows the variations of integrated chlorophyll (0-200 m) between lat. 6°N and 9°N at long. 137°E, obtained from the Ryofu Maru data (Anonymous 1972-84). Values during the 1982-83 El Nino are similar to those since July 1981, i.e., below 50 mg-m-2. SSCC from the same data set also shows low values during the 1982-83 El Nino, con- flicting with the results mapped on Figure 1. Re- cent El Nino events in 1972 and 1976 resulted in a drop of the sea level in the western Pacific (Meyers 1982). Low sea level was also recorded during an El Nino like event in the western Pacific in 1979-80 (Donguy and Dessier 1983). These low sea level episodes during which the thermocline is shallow (Wyrtki 1978), yet do not correspond to high SSCC or high integrated chlorophyll values in the Ryofu Maru results (Fig. 3). It seems however that the nutricline depth is shallower during these four epi- sodes (Fig. 3). All of them are moreover charac- terized by a shift in the nutrient-temperature rela- tionship (Fig. 2) indicating a consumption of new nutrients. We are dealing with an SSCC enrichment in the northwestern tropical Pacific which persists for several months (October 1982-March 1983) and is consistent with an input of new nutrients from below, but which does not appear in the chlorophyll concentrations measured every 6 mo on the Ryofu Maru. Both data sources have weaknesses. The SSCC monitoring does not measure what occurs below the surface. A significant correlation exists between SSCC and integrated chlorophyll on broad data sets (Lorenzen 1970; Piatt and Herman 1983), but oligotrophic ecosystems often show no relation- ship or, sometimes, inverse relationships (Hayward and Venrick 1982). The Ryofu Maru data at 137 °E between 6°N and 9°N allow a look at this problem (Fig. 4): the correlation between SSCC and in- tegrated chlorophyll is significant at the 1% level. The value r = 0.52 obtained with individual stations increases to r = 0.70 when enlarging the spatial scale (i.e., taking mean values between 6°N and 9°N instead of individual stations); a further improve- ment would probably be obtained by enlarging the time scale, but appropriate time series do not exist SSCC(mg m-3) A- .2- 0 — i »— ^ 1 1^ r- — -t ^ Integrated Chlor. (0-200m , mg.m-2) 100-. 50- 0- —I r-^ 1 1 1 — ■ Depth of nutricline ( m ) 100- 50- 0- 70 75 80 Figure 3.— Long-term evolution of lat. 6°N-9°N averaged parameters related to the primary production (data from the RV Ryofu Maru cruises at long. 137°E, Anonymous 1972 to 1984). Upper and middle panels: the chlorophyll concentrations primarily expressed in active chlorophyll a and pheophytin have been converted into chlorophyll a equivalents (Dandonneau 1979). Lower panel: the continuous line joins the depth of P04 = 0.35 /^moleL"1; open circles represent the depths of N03 = 1 ^mole-L"1. Thickened marks on the horizon- tal axis indicate the low sea level episodes in the western tropical Pacific. 692 DANDONNEAU: MONITORING SEA SURFACE CHLOROPHYLL CONCENTRATION Integrated Chlorophyll (0- 200m) : (mg.m~z) ^S • ♦ 100- • ♦*•* ~.i*r n = 128 r = .52 50- .•V n=26 r =.70 0- 1 1 1 — i 1 .10 .20 .30 .40 .50 5SCC(mg.m-3) Figure 4.— Integrated chlorophyll (0-200 m)/SSCC relationship between lat. 6°N and 9°N (data from the RV Ryofu Maru cruises at long. 137°E, Anonymous 1972 to 1984). Points and continuous line: individual stations. Crosses and dashed line: averaged values for each cruise. in this region. We can thus conclude that SSCC is a reasonable index of the chlorophyll content in the photic layer. The weakness of the Ryofu Maru data series is that only 4-6 stations within 3 d are avail- able for each El Nino episode. This sampling pat- tern can describe the vertical structure of the ocean, but it is not helpful in large-scale studies based on chlorophyll, in which the signal to noise ratio is very low (Dandonneau and Gohin 1984). Subtropical Zones At the start of the 1982-83 El Nino (and a possi- ble cause of it?) strong southerly winds were re- corded east of Australia in June and July 1982 (Harrison and Cane 1984). In the Coral and Tasman Seas, a chlorophyll enrichment occurs in austral winter between 22 °S and higher latitudes (Dandon- neau and Gohin 1984). This chlorophyll enrichment can be seen in austral autumn and winter of 1982 (Fig. 1), while it only appears in winter in 1983. Moreover, SSCC higher than 0.15 mg-m-3 spread northward to lat. 20 °S in July-September 1982 around long. 160°E, but only to 24°S in July-Sep- tember 1983 at the same longitude. The long and intense SSCC winter increase in this area in 1982 may be the result of advection of richer water from the south after the strong wind anomaly. In the Northern Hemisphere, a zone with high SSCC values is observed off North America during the fall of 1982 (Fig. 1); this feature is especially noteworthy since most regions of the Pacific (even those from the same merchant ship voyage) show low SSCC values. Like other El Ninos, the 1982-83 one re- sulted in temperatures and sea levels higher than normal along the California coast, and strong westerly winds at about 30°N. One would not ex- pect increased chlorophyll concentrations with higher temperatures, and according to Chelton et al. (1982), El Nino episodes are likely to diminish advection of water from the north which generates a higher biomass. However, our data points cor- responding to the enriched zone were far offshore (Fig. 1) and the thermal anomaly there did not great- ly differ from zero. The high SSCC values off North America during the fall of 1982 might then be related to the severe wind conditions which pre- vailed during this time, and probably induced ver- tical mixing of deep nutrients. A few more features which appear on Figure 1 would be worthy of discussion, but conclusion is hindered by the lack of accordance with a poorly known field of oceanic properties and by the risk of sampling or instrumental errors in SSCC measurements. For instance the shape of the area with SSCC >0.15 rng-m-3 centered slightly south of the Equator at 165°W in July- September 1982 (Fig. 1), while the upwelling was collapsing, is sur- 693 FISHERY BULLETIN: VOL. 84, NO. 3 prisingly similar to the shape of the maximum of cloudiness in September 1982, derived from satellite measurements of outgoing long wave radiation (Gill and Rasmusson 1983). Similarity might be causal, i.e., high SSCC values might result from a response of the phytoplankton to attenuation of light by the clouds, or from enhanced phytoplankton growth caused by precipitation of dust and aerosols by the rain (Menzel and Spaeth 1962). It may also result, at least partly, from sampling artifacts. The major features shown by this SSCC monitor- ing experiment are in agreement with the large- scale processes that affect the tropical Pacific dur- ing El Nino episodes. The collapse of the equatorial upwelling in October 1982 resulted in a nearly com- plete disappearance of the chlorophyll-rich area which is usually located across the Equator. A moderate enrichment persisted, however, east of long. 120°W. In the northwestern tropical Pacific, the eastward drift of the warmwater pool was followed by conditions which stimulated photosyn- thesis: a shallower thermocline, and more light penetrating to the nutrients gave rise to unusually high chlorophyll concentrations west of 180° from October 1982 to March 1983. In April-June 1983, the equatorial upwelling in the eastern Pacific was still reduced by the El Nino conditions, and the enrichment in the northwestern tropical Pacific was less intense; during this period, low chlorophyll con- centrations prevailed over most of the tropical Pacific. ACKNOWLEDGMENTS I would like to thank Henri Walico for the thousands of chlorophyll measurements which are the basis of the present work. I am indebted to the captains and crews of the merchant ships who call at Noumea for kindly and carefully sampling and filtering at sea. LITERATURE CITED Anonymous. 1972-84 The results of marine meteorological and ocean- ographical observations. Jpn. Meteorol. Agency, Tokyo, Nos. 47-72. Baker, K. S., R. C. Smith, and J. R. Nelson. 1983. Chlorophyll determinations with filter fluorometer: lamp-filter combination can minimize error. Limnol. Ocean- ogr. 28:1037-1040. Barber, R. T., and F. P. Chavez. 1983. Biological consequences of El-Nino. Science 222:1203- 1210. Bjerknes, J. 1966. Survey of El-Nino 1957-58 in its relation to tropical Pacific meteorology. Inter-Am. Trop. Tuna Comm. Bull. 12:25-86. Cane, M. A. 1983. Oceanographic events during El-Nino. Science 222: 1189-1195. Chelton, D. B., P. A. Bernal, and J. A. McGowan. 1982. Large scale interannual physical and biological inter- action in the California Current. J. Mar. Res. 40:1095- 1125. Dandonneau, Y. 1979. Concentrations en chlorophylle dans le Pacifique tropical sud-ouest: comparaison avec d' autres aires ocean- iques tropicales. Oceanol. Acta 2:133-142. 1982. A method for the rapid determination of chlorophyll plus phaeopigments in samples collected by merchant ships. Deep-Sea Res. 29:647-654. Dandonneau, Y., and F. Gohin. 1984. Meridional and seasonal variations of the sea surface chlorophyll concentration in the southwestern tropical Pacific (14 to 32°S, 160 to 175°E). Deep-Sea Res. 31:1377- 1393. DONGUY, J. R., W. BOUR, P. GALENON, AND J. A. GUEREDRAT. 1978. Les conditions oceanographiques et la peche de la bonite (Katsuwonus pelamis) dans le Pacifique occidental. Cah. ORSTOM, Ser. Oceanogr. 16:309-317. DONGUY, J. R., AND A. DESSIER. 1983. El-Nino like events observed in the tropical Pacific. Mon. Weather Rev. 111:2136-2139. DONGUY, J. R., AND C. HENIN. 1976. Anomalous navifacial salinities in the tropical Pacific Ocean. J. Mar. Res. 34:355-364. Feldman, G., D. Clark, and D. Halpern. 1984. Satellite color observations of the phytoplankton distribution in the eastern equatorial Pacific during the 1982-83 El-Nino. Science 226:1069-1070. Gill, A. E., and E. M. Rasmusson. 1983. The 1982-83 climate anomaly in the equatorial Pacific. Nature 306:229-234. Harrison, D. E., and M. A. Cane. 1984. Changes in the Pacific during the 1982-83 event. Oceanus 27(2):21-28. Hayward, T. L., and E. L. Venrick. 1982. Relation between surface chlorophyll, integrated chlorophyll and integrated primary production. Mar. Biol. (Berl.) 69:247-252. Holm-Hansen, O., C. J. Lorenzen, R. W. Holmes, and J. D. Strickland. 1965. Fluorometric determination of chlorophyll. J. Cons. 30:3-15. Koblentz-Mishke, O. J., V. V. Volkovinsky, and J. G. Kabanova. 1970. Plankton primary production of the world ocean. In W. S. Wooster (editor), Scientific exploration of the southern Pacific. Nat. Acad. Sci., Wash., D.C., p. 183-193. Lorenzen, C. J. 1970. Surface chlorophyll as an index of the depth, chloro- phyll content, and primary productivity of the euphotic layer. Limnol. Oceanogr. 15:479-480. Menzel, D. W., and J. P. Spaeth. 1962. Occurrence of ammonia in Sargasso sea water and in rain water at Bermuda. Limnol. Oceanogr. 7:159-162. Meyers, G. 1982. Interannual variation in sea level near Truk Island - A bimodal seasonal cycle. J. Phys. Oceanogr. 12:1161-1168. Meyers, G., and J. R. Donguy. 1984. The north equatorial countercurrent and heat storage 694 DANDONNEAU: MONITORING SEA SURFACE CHLOROPHYLL CONCENTRATION in the western Pacific Ocean during 1982-83. Nature 312: 258-260. OUDOT, C, AND B. WAUTHY. 1976. Upwelling et dome dans le Pacifique tropical occiden- tal distributions physico-chimiques et biomasse vegetale. Cah. ORSTOM, Ser. Oceanogr. 14:27-48. Platt, T., and A. W. Herman. 1983. Remote sensing of phytoplankton in the sea: surface layer chlorophyll as an estimate of water column chlorophyll and primary production. Int. J. Remote Sensing 4:343- 351. Quinn, W. H., D. 0. Zopf, K. S. Short, and R. T. W. Kuo Yang. 1978. Historical trends and statistics of the Southern Oscil- lation, El-Nino, and Indonesian droughts. Fish. Bull., U.S. 76:663-678. SCHREIBER, R. W., AND E. A. SCHREIBER. 1984. Central Pacific seabirds and the El-Nino southern oscillation: 1982 to 1983 perspectives. Science 225:713- 716. Sharp, J. H., M. J. Perry, E. H. Renger, and R. W. Eppley. 1980. Phytoplankton rate processes in the oligotrophic waters of the central north Pacific Ocean. J. Plankton Res. 2:335- 353. VOITURIEZ, B., AND A. HERBLAND. 1984. Signification de la relation nitrate/temperature dans l'upwelling equatorial du Golfe de Guinee. Oceanol. Acta 7:169-174. Walker, G. T. 1924. World weather II. Mem. India Meteorol. Dep. 24: 275-332. Walsh, J. J. 1981. A carbon budget for overfishing of Peru. Nature 290: 300-304. Wyrtki, K. 1978. Monitoring the strength of equatorial currents from XBT sections and sea level. J. Geophys. Res. 83C:1935- 1940. 1985. Sea level fluctuations in the Pacific during the 1982-83 El-Nino. Geophys. Res. Lett. 12:125-128. Yamanaka, H. 1984. The relationship between El-Nino episodes and fish migration and yields in the western Pacific. Trop. Ocean- Atmos. Newsl. 25:2-4. 695 ANATOMICAL TRAUMA TO SPONGE-CORAL REEF FISHES CAPTURED BY TRAWLING AND ANGLING S. Gordon Rogers,1 Hiram T. Langston,2 and Timothy E. Targett3 ABSTRACT External signs of trauma were examined in 15 sponge-coral reef fish species captured while trawling and angling at 37 m depth. Internal evidence of trauma was noted for all species and quantified for a sample of angling-caught black sea bass, Centropristis striata. Distinct differences were noted in the types and frequencies of trauma experienced among species, and between gear types within species. Black sea bass; red snappers, Lutjanus campechanus; short bigeyes, Pristigenys alta; and Mycteroperea groupers exhibited high frequencies of oral protrusions. Planehead filefish, Monacanthus hispidus; orange filefish, Aleuterus schoepfi; and blue angelfish, Holacanthus bermudensis, were particularly prone to cloacal protrusions. External signs of trauma were few in vermilion snappers, Rhomboplites aurorubens; porgies (Stemotomus chrysops, Calamus leucosteus, and Pagrus pagrus); tomtates, Haemulon aurolineatum; and two trawl-caught serranids (Centropristis ocyurus and Diplectrum formosum). Angling produced oral protrusions in black sea bass more frequently than trawling. Trawl-caught red snappers had a higher stomach eversion frequency when brought to the surface more quickly. Angling-caught black sea bass experienced high frequencies of tissue emphysema and swim-bladder rupture. These results should be considered in studies of feeding biology, released-fish survivorship, and fishery management. Anatomical trauma experienced by fishes during capture is interesting from several standpoints. Mor- tality of individuals caused by stress, tissue damage, organ displacement, and resulting aberrant behavior has been recognized primarily for its effects on the survival of released fish in mark-and-recapture studies (Ricker 1949; Parker et al. 1959, 1963; Got- shall 1964; Beamish 1966; Moe 1966; Laird and Stott 1978; Pawson and Lockwood 1980; Fable 1980; Grimes et al. 1983). Mortality of fishes released by fishermen is an important consideration for stock assessment and management (Black 1958; Pawson and Lockwood 1980; Matheson and Huntsman 1984). Recent management plans for the U.S. Gulf and South Atlantic snapper-grouper fisheries (GOMFMC 1981; SAFMC 1983a, b) recommended implementation of minimum sizes for several species. The sizes in the South Atlantic were deter- mined from yield-per-recruit (YPR) models incor- porating assumed survival rates for undersized, ^kidaway Institute of Oceanography, University System of Georgia, POB 13687, Savannah, GA 31416; present address: Coastal Resources Division, Georgia Department of Natural Resources, 1200 Glynn Avenue, Brunswick, GA 31523. 2Skidaway Institute of Oceanography, University System of Georgia, POB 13687, Savannah, GA 31416. 3Skidaway Institute of Oceanography, University System of Georgia, POB 13687, Savannah, GA 31416; present address: University of Delaware, College of Marine Studies, Lewes, DE 19958. released fishes (SAFMC 1983a). Size regulations were predicted on survivorship of ^60%. Gulf YPR models did not incorporate survival rates, effectively assuming 100% survival. Other workers have indicated difficulty in obtain- ing specimens of snapper-grouper species for quan- titative analyses of feeding biology from depths which caused stomach eversion and loss of gut con- tents (Stearns 1884; Adams and Kendall 1891; Camber 1955; Mosely 1966; Moe 1969; Bradley and Bryan 1975; Link 1980; Ross 1982). This is of par- ticular concern for studies comparing food habits across depth zones (Moseley 1966). Differences be- tween fish species captured by identical gear at similar depths and differences within species be- tween gear types introduce additional variation. This study addresses the types and frequencies of anatomical trauma experienced by sponge-coral reef fishes captured by angling and trawling at a single depth. These data are discussed in relation to trophic studies, future studies of trauma during capture, survival following release, and management of snapper-grouper fisheries. METHODS Fishes were caught by angling and trawling at a low-relief (<1 m) sponge-coral reef 37 m deep on the continental shelf 84 km east of Sapelo Island, GA Manuscript accepted January 1986. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 697 FISHERY BULLETIN: VOL. 84, NO. 3 (lat. 31°26'N, long. 80°20'W; central South Atlan- tic Bight). Angling gear was standard hand-operated boat rods rigged with double-hooked terminal tackle and baited with squid. Hook sizes were 3/0 to 5/0. Fishes were brought to the surface as quick- ly as possible (about 1 m/s; somewhat slower for large snappers and groupers). Trawling was con- ducted from two vessels, each rigged for stern trawling but with some differences in gear and handling. The trawl gear on the RV Georgia Bulldog was a 25 m, 4-seam high-rise roller trawl with tongue. Meshes were (stretched) 20 cm in the wings and tongue, 10 cm in the belly and bag (2 cm liner), and 7.5 cm in an extension. Cables connecting the trawl and doors produced a sweep of 31.1 m; the rise on the tongue was 6.1 m (J. B. Rivers4). The rig had a vertical haulback rate of 0.12-0.15 m/s. The trawl gear on the RV Blue Fin was a modified No. 36 Yankee flat roller trawl. Meshes were (stretched) 5 cm in the wings and belly and 3.5 cm in the bag (2 cm liner). The total sweep was 22.1 m and the rise at the center of the headrope was 3.7 m (Rivers fn. 4). The rig had a vertical haulback rate of 0.1 m/s. Gear handling was otherwise identical. Tows were 20 min long. The fish catch was sorted to species and the alimentary tracts samples re- moved; or samples were placed in 20 L buckets with ice-seawater mixture, frozen on board, and pro- cessed in the laboratory. Data on anatomical trauma were recorded during dissections. An angling catch of 34 black sea bass, Centropristis striata, was put on ice and dissected 2 days later for examination of internal trauma. No samples were subjected to the bin-type icing procedures common on commer- cial snapper-grouper vessels. Fishes were collected from July through December in 1983 and in Sep- tember 1984. External evidence of trauma consisted of several types of protrusion of the gastrointestinal tract. These were classified as 1) Oral eversion - stomach everted into the pharynx and often present in the mouth, pull- ing the pyloric area and the intestine with it. 2) Cloacal protrusion - intestine protruded from the cloacal area. Initially such protrusions were not classified further; however, detailed dissec- tions showed that they were either 4J. B. Rivers, Marine Fisheries Specialist, University of Georgia Fisheries Extension Station, POB Z, Brunswick, GA 31523, pers. commun. October 1984. a) Herniations - disruptions of the body wall in the pericloacal area through which the gut protruded or b) Intussusceptions - actual eversion of the terminal portion of the intestine through its own lumen. 3) Branchial protrusions - portions of the gut pro- truded through the branchial opening. Results are expressed as occurrences and percent- age frequencies. Frequencies of herniations and in- tussusceptions were calculated by dividing the observed number in a class by the total number of classified cloacal protrusions, then multiplying the result by the total proportion of cloacal protrusions. Example (from Table 1): planehead filefish herni- ations, (99/(99 + 22)) (160/440) = 0.30. Internal evidence of trauma included 1) the pres- ence of gas in the tissues (tissue emphysema) and 2) rupture of the swim bladder. Although notes on both phenomena were kept for all fish species, their frequencies were enumerated only for the 34 care- fully examined, angling-caught black sea bass. Among-species and between-gear comparisons of trauma were performed by using Pearson's test for goodness of fit (yielding a x2 value). The null hypotheses were specified as homogenous (equal) proportions of specimens exhibiting a particular symptom, based on the overall proportion of fish with the symptom across species or gears (signifi- cant departures were P < 0.05). RESULTS Dissection records of 1928 trawl-caught and 235 angling-caught fishes of 15 species were collated for external evidence of trauma (Table 1). Seven species were not caught with angling gear. Scamp, Myctero- perca phenax, and gag, M. microlepis, were com- bined to form a Mycteroperca grouper category due to low numbers collected. Trawl-caught red snappers, Lutjanus campecha- nus; Mycteroperca groupers; short bigeyes, Pristi- genys alta; planehead filefish, Monacathus hispidus; orange filefish, Aleuterus schoepfi; and blue angel- fish, Holacanthus bermudensis, experienced fre- quent gut displacements (Table 1). These were oral eversions in red snappers, short bigeyes, and Myc- teroperca groupers; cloacal protrusions in orange filefish and blue angelfish; and all three categories (including branchial protrusion) in planehead filefish. Alimentary tract displacements were minimal in trawl-caught black sea bass; bank sea bass, Centro- pristis ocyurus; sand perch, Diplectrumformosum; 698 ROGERS ET AL.: ANATOMICAL TRAUMA TO SPONGE-CORAL REEF FISHES Table 1 .—Numbers and percentage frequencies (in parentheses; a = 1 %) of alimentary tract displacements in sponge-coral reef fishes collected by trawling (T) and angling (A) in 37 m depth. Dashes (— ) indicate no data. Within cloacal protrusions, H = herniations, I = intussusceptions, U = unclassified, and TC = total cloacal. N = number of specimens examined. Oral eversions Cloacal protrusions Branchial Total Species H I u TC L^ 1 Ul Iwl IIUI protrusions displacements N black sea bass T 4(2) 0 0 0 0 0 4(2) 200 Centropristis striata A 45(27) 0 0 0 0 0 45(27) 169 red snapper T 26(55) 0 0 0 0 0 26(55) 47 Lutjanus campechanus A 1(50) 0 0 0 0 0 1(50) 2 bank sea bass T 0 0 0 0 0 0 0 39 Centropristis ocyurus A 1(33) 0 0 0 0 0 1(33) 3 short bigeye T 8(22) 0 0 0 0 0 8(22) 37 Pristigenys alta A — — — — — — — 0 sand perch T 0 0 0 0 0 0 0 19 Diplectrum formosum A 2(18) 0 0 0 0 0 0 11 Mycteroperca groupers T 5(29) 0 0 0 0 0 5(29) 17 A 0 0 0 0 0 0 0 1 planehead filefish T 3(1) 99(30) 22(7) 39 160(36) 14(3) 177(40) 440 Monacanthus hispidus A — — — — — — — 0 orange filefish T 0 1(4) 4(17) 7 12(21) 0 12(21) 58 Aleuterus schoepfi A — — — — — — — 0 blue angelfish T 0 4(30) 1(8) 4 9(38) 0 9(38) 24 Holacanthus bermudensis A — — — — — — — 0 vermilion snapper T 0 0 0 0 0 0 0 339 Rhomoboplites aurorubens A 0 0 0 1 1(4) 0 1(4) 28 whitebone porgy T 0 1(3) 1(3) 0 2(6) 0 2(6) 33 Calamus leucosteus A — — — — — — — 0 scup T 0 1(1) 0 2 3(1) 0 3(1) 286 Stenotomus chrysops A — — — — — — — 0 tomtate T 0 0 0 2 2(a) 0 2(a) 372 Haemulon aurolineatum A — — — — — — — 0 red porgy T 0 0 1(6) 0 1(6) 1(6) 2(12) 17 Pagrus pagrus A 0 0 0 0 0 0 0 21 tomtate, Haemulon aurolineatum; scup5, Stenoto- mus chrysops; whitebone porgies, Calamus leucos- teur; red porgies, Pagrus pagrus; and vermilion snappers, Rhomboplites aurorubens. Angling-caught black sea bass had high frequen- cies of oral eversion. Angling-caught red porgies and vermilion snappers exhibited few or no protrusions. Angling data for all other species are too sparse to estimate protrusion frequencies. There was a significant lack of homogeneity in the frequencies of oral eversions between species within trawl (x2 = 695, df = 13, P « 0.01) and angling- caught (x2 = 14.2, df = 6, P < 0.05) samples. The trawling value resulted from high frequencies for red snapper, Mycteroperca groupers, and short bigeye; these three categories accounted for 95% 5The taxonomic status of this species is unclear (B. Roumillat, South Carolina Marine Resources Research Institute, POB 12559, Charleston, SC, 29412 pers. commun.) and is properly listed as scup (Stenotomus chrysops (Robins et al. 1980; SAFMC 1983a, b)) although several authors have recently used the nomen southern porgy (S. aculeatus (Miller and Richards 1980; Wenner 1983; Sedberry and Van Dolah 1984)). Still others have classified South Atlantic-caught Stenotomus as longspine porgy (S. caprinus (Chester et al. 1984)). of the x2 statistic. Among angling-caught fishes, a high value for black sea bass and low values for red porgy and vermilion snapper accounted for 91% of the x2 statistic. The high frequencies of cloacal protrusions in trawl-caught planehead filefish, orange filefish, and blue angelfish (21-38%) and low values in all other species (<7%) produced a highly significant depar- ture from homogeneity (x2 = 470, df = 13, P « 0.001). Seven of the 15 fish species did not display the symptom (Table 1). Only one of the angling- caught specimens (a vermilion snapper) experienced cloacal protrusion. Of those cloacal protrusions classified for blue angelfish and the two filefish species, all herniations (Table 1) had fecal material in the protruded gut portion. Only planehead filefish experienced branchial pro- trusions. Tomtate, vermilion snapper, scup, red porgy, and whitebone porgy were notably free of all forms of alimentary tract displacement. Swim-bladder rupture was noted for all fish species. Tissue emphysema was detected only in black sea bass. Of the 34 black sea bass exam- ined in detail for internal trauma, 33 (97%) 699 FISHERY BULLETIN: VOL. 84, NO. 3 exhibited swim-bladder rupture (1 specimen had 2 points of rupture), and 27 (79%) had tissue emphy- sema. Significantly more angling-caught black sea bass had oral protrusions than those caught by trawling (X2 = 138, df = 1, P « 0.001). For trawl-caught red snappers, significantly more fish caught aboard the Georgia Bulldog (26 of 39) had oral eversions than those caught aboard the Blue Fin (0 of 8) (x2 = 5.34, df = 1, P < 0.025). No other comparisons for combinations of symptoms, species, and gear types yielded significant results. However, all oral eversions noted for Mycteroperca groupers were produced by Georgia Bulldog trawling gear, and those noted for sand perch were produced by angling gear. DISCUSSION Differences Due to Species and Gear Differences between fish species (captured by identical gear) in the type and frequency of gut displacement are likely due to differences in bone structure and relative swim-bladder volume. Except for planehead filefish, which exhibited all forms of external evidence, those species which experienced frequent oral eversions did not present cloacal ever- sions and vice versa (Table 1; refer also to the anal- yses of categorized data). In this study the leather- jackets (Balistidae) and angelfishes (Holacanthidae) experienced high frequencies of gut displacements toward the cloacal area. These taxa have a relatively restricted pharyngeal area and the leatherjackets have a bony sternum which further defines a "path of least resistance" toward the cloaca. Other fishes which may be similarly susceptible to cloacal pro- trusions include other balistids, acanthurids, chae- todontids, and scarids. Larger mouthed species such as lutjanids (Stearns 1884; Adams and Kendall 1891; Camber 1955; Moseley 1966; Bradley and Bryan 1975; this study), serranids (Moe 1969; Link 1980; Matheson and Huntsman 1984; this study), priacanthids (this study), and scorpaenids (Gotshall 1964) experience oral eversion more frequently than cloacal protru- sion. Fishes with medium-sized mouths and "non- directing" body morphologies (e.g., vermilion snap- per, tomtate, and sparids in this study) exhibit neither type of gut protrusion, instead having a general swelling of the body cavity. The relative volume of the swim bladder varies from 0 to 6% of total body volume in marine fishes (Jones 1957). Although measurements were not made, the patterns of protrusion in this and other studies (above) suggest that species-specific differ- ences in swim-bladder volume result in varying degrees of internal pressure on ascent. This may contribute to differences in gut protrusion and the extent of body cavity swelling. It is not clear why varying rates of ascent would induce varying frequencies of gut protrusion within a fish species. Differences between fish species in the rates at which gases can be resorbed from the swim bladder likely had little effect on patterns of protrusion. Achievement of equilibrium through resorption requires time on the order of hours (Brown 1939; Jones 1951). This is a longer time-scale than the normal vertical movements of most fishes (Steen 1970) and vertical displacements while trawl- ing and angling. Also, the absolute magnitude of swim-bladder expansion is independent of the rate of ascent and should not be considered a factor. Yet, a pattern is apparent in the higher values for angling versus trawl-caught black sea bass and also for red snappers caught with Georgia Bulldog versus Blue Fin trawling gear. Mosely (1966) reported higher oral eversion frequencies for red snappers taken by angling versus those taken while trawling at inter- mediate shelf depths (42-60 m). Bradley and Bryan (1975) also noted for red snappers that angling pro- duced more stomach eversions than trawling, but stated that their data were confounded by differ- ences in the average depths of fishing efforts. Addi- tionally, stomach eversion frequencies for our trawl- caught red snappers (83% taken with "rapid ascent" Georgia Bulldog gear) were 7.5-9.5 times higher than those reported in the literature from similar depths (Fig. 1A; Moseley 1966; Bradley and Bryan 1975). It is tempting to attribute these results to dif- ferences in vertical haulback rates. The rate of swim- bladder expansion, linked directly to changes in hydrostasis (Steen 1970) and therefore qualitative- ly more or less "violent", may govern the nature and extent of injuries. An additional factor contributing potentially to the types and frequencies of gut protrusion is the con- sistency, amount, and position of prey material in the alimentary tract. Firm material may function as a bonelike directing structure or be what an ex- panding swim bladder acts upon. It is interesting that all of the herniated intestines in planehead file- fish, blue angel fish, and orange filefish contain fecal material. If hydrostatic forces within a fish's body cavity are influenced by gut contents, unequal and variable allocation of sampling effort and catch over a diel feeding cycle could alter estimates of protru- sion frequency for a given fish species. The major- 700 ROGERS ET AL.: ANATOMICAL TRAUMA TO SPONGE-CORAL REEF FISHES A 60 50 a. 40 c 30- oj O 20^ 10 0 Trawl Caught i i i i i i 1 1 1 1 0 10 20 30 40 50 60 70 80 90 100 e. Ul S 100 90 80 70 60 C 50 £ 40 30 20 10- o- Angling Caught i i i i i i i 1 1 1 0 10 20 30 40 50 60 70 80 90 100 Bottom Depth (m) Figure 1.— Plots of the proportions of red snappers with everted stomachs (PE) captured by (A) trawling and (B) angling as a func- tion of bottom depth (data from Camber 1955; Moseley 1966; Bradley and Bryan 1975; except this study). Ordinates were arc- sine transformed (Snedecor and Cochran 1980). Abscissas are plotted as actual depths or midpoints of ranges. The dashed line (plot ^4) is the least-squares line including data from this study. The only significant relationship was for trawl-caught fishes from the literature (r = 0.90, df = 5, P < 0.01). ity of orange filefish were collected during periods of the day when there was very little material in the alimentary tract, which is likely responsible for her- niation/intussesception rates at variance with plane- head filefish and blue angelfish values. Sampling of other species was more equitably distributed over the 24-h period. Considerations for Feeding Studies Negligible biases in stomach and intestinal con- tents are expected among trawl-caught black sea bass, bank sea bass, tomtate, the three porgy species, sand perch, and vermilion snapper at depths of 37 m. Angling-caught red porgies and vermilion snappers should be equally free of bias-producing gut displacements at these depths. However, cau- tion is necessary in analyses of stomach contents for trawl-caught red snappers, Mycteroperca groupers, short bigeyes, and angling-caught black sea bass from 37 m. Stomach content data for angling-caught red snappers, groupers, bank sea bass, and sand perch should also be interpreted with attention to the likelihood of bias. These considerations have been previously acknowledged for angling-caught black sea bass and bank sea bass (Link 1980), trawl and angling-caught red snappers (Stearns 1884; Adams and Kendall 1891; Camber 1955; Moseley 1966; Bradley and Bryan 1975), angling-caught red groupers, Epinephelus morio (Moe 1969), and angling and longline-caught blueline tilefish, Caulo- latilus microps (Ross 1982), from southeastern U.S. shelf and slope waters. Moseley (1966) and Link (1980) both stated that partial or full stomach ever- sion renders quantification of consumed prey suspect, particularly with respect to across-depth comparisons (e.g., Godfriaux 1974). Studies of food habits of fishes in the South Atlantic and Gulf of Mexico shelf snapper-grouper complex have either not discussed depth as a diet-determining variable (Camber 1955; Moseley 1966; Moe 1969; Bradley and Bryan 1975; Dixon 1975; Henwood et al. 1978; Ross 1982; Steimle and Ogren 1982) or if depth was considered, dealt with fishes not prone to stomach- eversion bias (Manooch 1977; Grimes 1979; Sed- berry 1985). Species and gear-specific considerations should also be made for analyses of daily feeding chron- ologies and rations based on stomach content weights. Fishes with partially or completely everted stomachs should be eliminated from the data set. It is clear that trawl-caught specimens of most species are more suited to such analyses than those caught with angling gear. However, some species cannot be efficiently collected with trawling gear at certain times of day, over certain types of bottom, or indeed at all. Extra angling effort (offsetting eversion rates) and well-designed multigear ap- proaches (including traps and longlines) can be used to complete data sets for such fishes. Displacements of the posterior portion of the alimentary tract can also have significant effects on studies of feeding biology. Trawl-caught planehead filefish, blue angelfish, and orange filefish are sub- ject to such bias. Prey position data used to examine the rate of movement and evacuation of material through the gut (e.g., Klumpp and Nichols 1983) will be affected by both herniations and intussusceptions. During herniation, fecal material is either shifted into the protruded portion of the intestine or the 701 FISHERY BULLETIN: VOL. 84, NO. 3 material already present in that segment is isolated from what might otherwise be a continuous column of material. Both potentially produce gaps or "clumping" of intestinal contents. Collection of data affected by intestinal displacements should also in- corporate increased sampling so that specimens with herniations or intussusceptions can be eliminated from the data set without a significant loss of information. Survivorship: Experimental Design and Fishery Management Our data show that experimental studies of sur- vivorship and the physiological responses of sponge- coral reef fishes following capture and release should stratify their designs by gear. Traps and longlines should be considered in future studies because of the gear-specific vertical haulback rates and other stress factors. Additional considerations are capture depth (Gotshall 1964; Moe 1966, 1969; Moseley 1966; Bradley and Bryan 1975; Grimes et al. 1983), preda- tion on injured and disoriented fishes (Parker et al. 1959, 1963; Randall 1960; Topp 1963; Gotshall 1964; Fable 1980), crowding and abrasion in the gear (Pawson and Lockwood 1980), degree of gut full- ness (related to stress from diverted blood supply; Beamish 1966), physiological state related to long- term feeding/activity cycles (Parker et al. 1959), water column temperature structure, currents, and turbidity. Many of the factors covary with depth and fluctuate seasonally. The anatomical derangements investigated in the present study are severe trauma. Oral and cloacal protrusions would very likely cause high rates of mortality in subsequently released fishes. Obstruc- tion of the gastrointestinal tract would normally be serious and interference with the blood supply to the gastric and intestinal walls would lead to severe cir- culatory impairment. Gotshall (1964) has shown that returns from tagged blue rockfish, Sebastes mys- tinus requiring stomach replacement and swim- bladder deflation were less than half those from fish requiring only swim-bladder deflation. These fish en- dured everted stomachs for only a few minutes. Topp (1963) has noted that the everted stomachs of Lutjanus snappers are frequently perforated by the fish's teeth. The effects of such injuries on survival require further study. Expansion of the swim bladder in specimens which do not experience gut protrusions likely induces in- ternal damage undetected by external examination. Aquarists commonly use swim-bladder deflation techniques to increase survivorship of specimens suf- fering from decompression symptoms (D. Miller6). Gotshall (1964) increased tag returns of blue rock- fish by deflating expanded swim bladders of speci- mens collected as deep as 90 m. The technique also reduces the effects of exopthalmia (protruding eyes produced by expansion of gas into the cranial region) on blue rockfish (Gotshall 1964), vermilion snapper, big eye (Priacanthus arenatus), and short bigeye (D. Miller fn. 6). Although tissue emphysema per se may not be lethal, swim-bladder rupture probably is for some species. Jones (1949) reported 90% mortality of 600 perches, Percafluviatilis, with swim bladders rup- tured while being raised rapidly from 13.7 m. Topp (1963) speculated that survivorship of sponge-coral reef fishes with ruptured swim bladders is very low. However, R. O. Parker7 has observed healing of rup- tured swim bladders in black sea bass. Further ex- perimentation is needed to determine the effects of swim-bladder rupture on a species-specific basis. It is likely that survivorship following release varies with depth due to hydrostatic factors alone. Regression of trawl-caught red snapper stomach eversion proportions on capture depth (values from the literature) explains 80% of the variance in the observed data (Fig. L4; r = 0.90, df = 5, P < 0.01). Inclusion of our trawl-caught red snapper data rendered the relationship nonsignificant (Fig. L4; r = 0.70, df = 6, P < 0.05). A similar plot of angling- caught red snapper data from the literature was not significant (Fig. IB; r = 0.58; df = 5; 0.10 < P < 0.05), possibly because of the differences in the sizes of red snappers hooked with respect to depth (see Figure 1 citations) and resultant differences in the rates of ascent, or ontogenetic differences in relative swim-bladder volume. Note that increased depth eventually outweighs any real effect of the size of the fish and tenacity of its struggle against the angling gear, or anatomical variation, rendering the overall relationship positive albeit nonlinear. The above data (Fig. LA, B) also indicate that red snap- pers caught with any gear over bottoms <30 m deep do not suffer significant trauma. Similarly, depths < 20 m introduced no difficulties to a food habits study of this species (Moseley 1966). Clearly, regulations which diminish removal of fishes (e.g., gear/method restrictions, area/time closures) will be more effective over a larger depth 6D. M. Miller, Curator, University of Georgia Marine Education Center, POB 13687, Savannah, GA 31416, pers. commun. November 1984. 7R. O. Parker, National Marine Fisheries Service, Southeast Fisheries Center, Beaufort Laboratory, POB 500, Beaufort, NC 28516, pers. commun. October 1984. 702 ROGERS ET AL.: ANATOMICAL TRAUMA TO SPONGE-CORAL REEF FISHES range than release measures. Current management of the southeastern U.S. snapper-grouper fisheries guarantees subjection of protected size classes to stress and trauma. However, it is conceivable that swim-bladder deflation techniques could improve the effectiveness of current regulations. The data we have presented show that the effects of capture on sponge-coral reef fishes vary between species and gears. These are important considera- tions for studies of feeding biology. Additional data on fish species survivorship following releases, stratified by gear and depth, will allow fine-tuning of present snapper-grouper management policies. ACKNOWLEDGMENTS Special thanks go to the captains and crews of the RV Blue Fin, RV Georgia Bulldog, and FV Sanc- tuary. Trawl doors for the RV Blue Fin were loaned by S. Drummond, Southeast Fisheries Center Pas- cagoula Laboratory, National Marine Fisheries Ser- vice. D. Wyanski, B. Wells, C. Haney, T. Chestnut, D. Miller, P. Schlein, B. Goggins, J. Rivers, M. Har- ris, J. Hightower, and a host of volunteers assisted in the field. D. Wyanski, B. Wells, A. Acevedo, and L. Creasman assisted in the laboratory. The figure was drafted by A. Boyette. The manuscript was im- proved with comments from G. Helfman, S. Larson, and two anonymous reviewers; it was typed by L. Wainright and W. Roberts. The research benefited from discussions with R. Parker, C. Haney, D. Miller, P. Van Veld, and L. Wall. This work was sup- ported by NOAA Sea Grants to T. E. Targett and M. V. Rawson (NA80AA-D-00091 collectively) and by operating funds at Skidaway Institute of Ocean- ography. LITERATURE CITED Adams, A. C, and W. C. Kendall. 1891. Report upon an investigation of the fishing grounds off the west coast of Florida. Bull. U.S. Fish. Comm. 9:289- 312. Beamish, F. W. H. 1966. Muscular fatigue and mortality in haddock, Melano- grammus aeglefinus, caught by otter trawl. J. Fish. Res. Board Can. 23:1507-1521. Black, E. C. 1958. Hyperactivity as a lethal factor in fish. J. Fish. Res. Board Can. 15:573-586. Bradley, E., and C. E. Bryan. 1975. Life history and fishery of the red snapper (Lutjanus campechanus) in the northwestern Gulf of Mexico: 1970- 1974. Proc. Gulf Caribb. Fish. Inst. 27:77-106. Brown, F. A., Jr. 1939. Responses of the swimbladder of the guppy, Lebistes reticulatus, to sudden pressure decreases. Biol. Bull. 76:48-58. Camber, C. I. 1955. A survey of the red snapper fishery of the Gulf of Mex- ico, with special reference to the Campeche Banks. Fla. Board Conserv. Mar. Res. Lab. Tech. Ser. 12, 64 p. Chester, A. J., G. R. Huntsman, P. A. Tester, and C. S. Manooch, III. 1974. South Atlantic Bight reef fish communities as repre- sented in hook-and-line catches. Bull. Mar. Sci. 34:267-279. Dixon, R. L. 1975. Evidence for mesopelagic feeding by the vermilion snapper, Rhomboplites aurorubens. J. Elisha Mitchell Sci. Soc. 91:240-242. Fable, W. A. Jr. 1980. Tagging studies of the red snapper (Lutjanus cam- pechanus) and vermilion snapper (Rhomboplites aurorubens) off the south Texas coast. Contrib. Mar. Sci. 23:115-121. Godfriaux, B. L. 1974. Food of snapper in western Bay of Plenty, New Zea- land. N. Z. J. .Mar. Freshw. Res. 8:473-504. GOMFMC (Gulf of Mexico Fishery Management Council). 1981. Environmental impact statement and fishery manage- ment plan for the reef fish resources of the Gulf of Mexico. GOMFMC, St. Petersburg, FL. GOTSHALL, D. W. 1964. Increasing tagged rockfish (genus Sebastodes) survival by deflating the swim bladder. Calif. Fish Game 50:253- 260. Grimes, C. B. 1979. Diet and feeding ecology of the vermilion snapper, Rhomboplites aurorubens (Cuvier) from North Carolina and South Carolina waters. Bull. Mar. Sci. 29:53-61. Grimes, C. B., S. C. Turner, and K. W. Able. 1983. A technique for tagging deepwater fish. Fish. Bull., U.S. 81:663-666. Henwood, T., P. Johnson, and R. Heard. 1978. Feeding habits and food of the longspined porgy, Steno- tomus caprinus Bean. Northeast Gulf Sci. 2:133-137. Jones, F. R. H. 1949. The teleostean swimbladder and vertical migration. Nature 164:847. 1951. The swimbladder and the vertical movements of teleo- stean fishes. I. Physical factors. J. Exp. Biol. 28:553-566. 1957. The swimbladder. In M. E. Brown (editor), The physiology of fishes. II. Behavior, p. 305-322. Acad. Press, N.Y. Klumpp, D. W., and P. D. Nichols. 1983. Nutrition of the southern sea garfish Hyporhamphus melanochir: gut passage rate and daily consumption of two food types and assimilation of seagrass components. Mar. Ecol. Prog. Ser. 12:207-216. Laird, L. M., and B. Stott. 1978. Marking and tagging. In T. Bagenal (editor), Methods for assessment of fish production in fresh waters, p. 84-100. Blackwell Sci. Publ., Oxf. Link, G. W., Jr. 1980. Age, growth, reproduction, feeding, and ecological observations on the three species of Centropristis (Pisces: Serranidae) in North Carolina waters. Ph.D. Thesis, Univ. North Carolina, Chapel Hill, NC. Manooch, C. S., III. 1977. Foods of the red porgy, Pagrus pagrus, Linnaeus (Pisces: Sparidae), from North Carolina and South Carolina. Bull. Mar. Sci. 27:776-787. Matheson, R. H., Ill, and G. R. Huntsman. 1984. Growth, mortality, and yield-per-recruit models for 703 FISHERY BULLETIN: VOL. 84, NO. 3 speckled hind and snowy grouper from the United States South Atlantic Bight. Trans. Am. Fish. Soc. 113:607-616. Miller, G. C, and W. J. Richards. 1980. Reef fish habitat, faunal assemblages, and factors determining distributions in the South Atlantic Bight. Proc. Ann. Gulf Caribb. Fish. Inst. 32:114-130. Moe, M. A., Jr. 1966. Tagging fishes in Florida offshore waters. Fla. Board Conserv., Div. Southwest Fish. Tech. Ser. 49, 40 p. 1969. Biology of the red grouper Epinephelus morio (Valen- ciennes) from the eastern Gulf of Mexico. Fla. State Board Conserv., Dep. Nat. Resour. Mar. Lab., Prof. Pap. Ser. 10, 95 p. Moseley, F. N. 1966. Biology of the red snapper, Lutjanus aya Bloch, of the northwestern Gulf of Mexico. Publ. Inst. Mar. Sci., Univ. Tex. 11:90-101. Parker, R. R., E. C. Black, and P. A. Larkin. 1959. Fatigue and mortality in troll-caught Pacific salmon (Oncorhynchus). J. Fish. Res. Board Can. 16:429-448. 1963. Some aspects of fish-marking mortality. In Northwest Atlantic Fish Marking Symposium, 1961, p. 117-122. Int. Comm. Northwest Atl. Fish. Spec. Publ. 4. Pawson, M. G., and S. J. Lockwood. 1980. Mortality of mackeral following physical stress, and its probable cause. Rapp. P. -v. Reun. Cons. int. Explor. Mer 177:439-443. Randall, J. E. 1960. The case of the free-loading barracuda. Sea Front. 6:174-179. RlCKER, W. E. 1949. Effects of removal of fins upon the growth and sur- vival of spiny-rayed fishes. J. Wild]. Manage. 13:29-40. Robins, C. R., R. M. Bailey, C. E. Bond, J. R. Brooker, E. A. Lachner, R. N. Lea, and W. B. Scott. 1980. A list of common and scientific names of fishes from the United States and Canada. Am. Fish. Soc. Spec. Pub. 12, 174 p. Ross, J. L. 1982. Feeding habits of the gray tilefish, Caulolatilus microps (Goode and Bean, 1878), from North Carolina and South Carolina waters. Bull. Mar. Sci. 32:448-454. SAFMC (South Atlantic Fishery Management Council). 1983a. Source document for the snapper-grouper fishery of the South Atlantic region. SAFMC, Charleston, SC. 1983b. Fishery management plan, regulatory impact review, and final environmental impact statement for the snapper- grouper fishery of the South Atlantic region. SAFMC, Charleston, SC. Sedberry, G. R. 1985. Food and feeding of the tomtate, Haemulon aur'o- lineatum (Pisces, Haemulidae), in the South Atlantic Bight. Fish. Bull., U.S. 73:461-466. Sedberry, G. R., and R. F. Van Dolah. 1984. Demersal fish assemblages associated with hard bot- tom habitat in the South Atlantic Bight of the USA. En- viron. Biol. Fishes 11:241-258. Stearns, S. 1884. On the position and character of the fishing grounds of the Gulf of Mexico. Bull. U.S. Fish. Comm. 4:289-290. Steen, J. B. 1970. The swim bladder as a hydrostatic organ. In W. S. Hoar and D. J. Randall (editors), Fish physiology. Vol. IV. The nervous system, circulation, and respiration, p. 413-443. Acad. Press, N.Y. Steimle, F. W., Jr., and L. Ogren. 1982. Food of fish collected on artificial reefs in the New York Bight and off Charleston, South Carolina. Mar. Fish. Rev. 44(6-7):49-52. Topp, R. 1963. The tagging of fishes in Florida 1962 program. Fla. Board Conserv., Mar. Lab. Prof. Pap. Ser. 5, 76 p. Wenner, C. A. 1983. Species associations and day-night variability of trawl- caught fishes from the inshore sponge-coral habitat, South Atlantic Bight. Fish. Bull, U.S. 81:537-552. 704 ANNUAL PRODUCTION OF EVISCERATED BODY WEIGHT, FAT, AND GONADS BY PACIFIC HERRING, CLUPEA HARENGUS PALLASI, NEAR AUKE BAY, SOUTHEASTERN ALASKA Jay C. Quast1 ABSTRACT Pacific herring, Clupea harengus pallasi, grow according to the constant-proportion growth model, which requires that yearly growth in body length be a constant proportion of growth during the previous year. Herring have one or two growth stanzas (periods of constant-proportional growth) in the eastern Pacific Ocean and eastern Bering Sea, and grow faster in the eastern Bering Sea than in the northeastern Pacific Ocean. With growth, total and eviscerated body weights of fresh Auke Bay herring bear an exponential relationship to body length (BL) that is slightly greater than cubic, and evisceration does not lower variabili- ty in length-weight relationships. With growth, an increasing part of the annual product (growth plus gonads) is partitioned into gonads so that in the largest fish most of the annual product is gonads. The annual product is constantly proportional to BL through ages 2-6 and also through ages 9-12, but the proportion is considerably smaller in the 9- to 12-yr-old fish. The two differing proportions may indicate that young and old Auke Bay herring occupy slightly different feeding niches and that the trophic en- vironment in the Auke Bay vicinity may not support the older fish as well as the younger. Pacific herring spawn in April or May in the Auke Bay vicinity, as zooplankton density rapidly in- creases to its peak in June. The time of spawning seems optimal for rapid building of fat reserves and feeding of newly hatched larvae. Pacific herring, Clupea harengus pallasi, range off western North America, from the Chukchi Sea to San Diego, CA, and have been commercially ex- ploited over the entire range (Rounsefell 1930; McLean and Delaney 1978; Spratt 1981). Pacific herring usually occupy extensive reaches of coast, from tens to hundreds of miles, and populations are particularly dense around the Alexander Archi- pelago of southeastern Alaska and the archipelago off British Columbia (from charts or fisheries maps in Rounsefell 1930, McLean and Delaney 1978, and Spratt 1981). Yet, even where dense, populations can be locally distinctive in vertebral number and spawning time (Rounsefell and Dahlgren 1935; Hourston 1980). Pacific herring have been commercially harvested in Alaska since the late 1800's (Rounsefell 1930), principally for reduction to meal and oil. Herring were also pickled, starting in 1900, but the industry never became large and declined in the 1920's. A fishery for Pacific halibut, Hippoglossus stenolepis, bait had a similar rise and decline. The reduction Northwest and Alaska Fisheries Center Auke Bay Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 210155, Auke Bay, AK 99821; present address: 1565 Jamestown Street S.E., Salem, OR 97302. Manuscript accepted December 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. fishery ended in the 1960's, and the principal fishery for Pacific herring in Alaska now is sac roe, which is exported to Japan. The biology of Pacific herring in Alaska has not been thoroughly described. The study by Rounse- fell (1930) is the most comprehensive work, and Rounsefell and Dahlgren (1935) separated stocks in southeastern Alaska on the basis of vertebral counts. Skud (1963) analyzed tag returns, and Carlson (1980) described the ecology of Auke Bay herring. Reid (1971) summarized some biological character- istics of herring taken for the reduction fishery from 1929 to 1966. Because Pacific herring are economically and ecologically important in southeastern Alaska and there is little information on the growth, produc- tivity, and life history of this species in this region, I undertook a 1-yr study of a population in the Auke Bay vicinity (Auke Bay is about 16 km northwest of Juneau). Goals of the study were to compare growth of Pacific herring in the Auke Bay vicinity with growth of Pacific herring from other locales in the eastern Pacific Ocean and relate annual pro- duction of fat, gonads, and eviscerated weight in the Auke Bay herring to the annual cycle of food supply. Pacific herring of the Auke Bay vicinity are one of the innermost and northernmost populations in 705 FISHERY BULLETIN: VOL. 84, NO. 3 the Alexander Archipelago (Auke Bay is about 80 nmi [148 km] by water from the open coast). Although this population may contain more than one spawning stock, it will be identified with Auke Bay in the present study (local populations spawn with- in weeks of each other and within a few nautical miles). METHODS Auke Bay herring were sampled several times monthly from April 1973 to March 1974; however, no fish were taken in February 1974. The fish were captured principally by jigging with bright hooks or hooks wrapped with colored yarn. Samples were also taken during spring 1973 from nearby locales in southeastern Alaska, including Hood Bay (off Chatham Strait, southwest of Juneau), Carroll In- let (near Ketchikan), and Katlian Bay (near Sitka), and also from the eastern Bering Sea west of Nunivak Island. Auke Bay herring were usually examined fresh but sometimes were frozen and examined within 1 wk. Lengths were originally measured as standard lengths (SL, tip of upper jaw to end of hypural bones) but were later converted to body length (BL, tip of lower jaw to end of hypural bones) by multi- plying SL by 1.0132, the average ratio in 126 specimens from Auke Bay. Body lengths were back-calculated from scales taken from above the pectoral fins and posterior to the opercular flap. The calculations followed the pro- portional method of Whitney and Carlander (1956), which should reduce the variation in BL-scale size relationships because the method adjusts for possi- ble differences in scale length in the same-sized fish. This method requires that the regression between BL and scale length be linear, which was satisfied (Fig. 1). The intercept of the regression (55 mm) was somewhat higher than the median BL (36.5 mm) for first squamation of 16 preserved specimens; however, the differences between estimates for BL at first squamation are probably important only for fish younger than 1 yr. The regression fit the data well for herring >1 yr old (Fig. 1). I also attempted to reduce variability in the back-calculations for the Auke Bay fish by averaging focus-to-annulus distances from left and right sides of the scales (an- nuli were as well defined at the sides as in the centerline of the scale), but only a centerline measurement was used in samples from other geographic regions. After the growth data were analyzed by Walford graphs (Walford 1946), linear regressions (Walford regressions) were fit by least squares to adult sec- tions of constant parameters (stanzas) that were indicated on the graphs. Both the Walford regres- sion and von Bertalanffy formulation are variants of the constant-proportion growth model, which requires that growth in one year be a constant pro- portion of growth the preceding year (Ricker 1975). (The slope of a Walford regression equals the von Bertalanffy e~K, and the intercept equals L (1 - e~K).) Annual changes in development of fat and gonads were evaluated by indices that were derived from total body weights, eviscerated body weights, and gonad weights. I estimated unbound water in the eviscerated body tissues and gonads as the percent- age weight lost by drying 1 cm wide transverse body sections and entire gonads in a drying oven for more than 4 d at 27° C, a period that yielded weight stabil- ity. Visual estimates of visceral fat used a four-point scale (from none to heavy), and visual estimates of maturity used a seven-point scale, as follows (Roman numerals in brackets refer to a similar scale developed by Hay and Outram (1981) for Pacific herring): 1) Newly regenerating [VIII], 2) regen- erating [III], 3) nearly mature [IV], 4) ripe [V], 5) ripe and running [VI], 6) partially spawned [VII], and 7) spawned out [VII]. Fresh body, eviscerated, and gonad weights were regressed on body lengths by least squares after logarithmic transformation of variates. Statistical tests were significant when P < 0.05. Scales of Pacific herring from southeastern Alaska and the eastern Bering Sea probably have two an- nuli in the first growth year. When the annulus nearest to the scale focus of Auke Bay herring was used for back-calculations, the BL's were much smaller (average of 65 mm) for the first winter than the BL's of juvenile herring (at least 80 mm) cap- tured at the end of their first year in Auke Bay by Jones (1978). Pacific herring in British Columbia at- tain a length of at least 80 mm by their first Sep- tember (Hourston 1958). Furthermore, when the first annulus was used as the first year mark, Wal- ford graphs of the growth data were erratic and differed markedly from graphs of the same type of data in the literature. When the second annulus was used as the first year mark, the graphs were sim- ple and corresponded to graphs of similar data from the literature. There was no indication of Lee's phenomenon (slower growth in longer lived individuals) in the back-calculated BL's, but there was evidence of a changing relation between growth back-calculated for ages 1 and 2 and the span of years that was used 706 QUAST: BODY WEIGHT, FAT, AND GONADS OF PACIFIC HERRING 250 200 — x r- O z UJ _l > a o 150 100 — BL N 54.855 + 0.6309 S = 234, R = 0.84 OL^J I I I I I I I I I I I I I I I I I I I I L 100 200 300 400 PROJECTED SCALE SIZE (S) IN MM 500 Figure 1.— Relationship between body length (BL) and projected scale size (S) of Pacific herring from Auke Bay, AK. in the back-calculations. When the estimates of growth to ages 1 and 2 were compared for all specimens, those from herring aged 2-4 at time of capture (back-calculated over a span of 0-2 yr) had slower-than-average growth, and those herring aged 4-7 (back-calculated over a span of 3-5 yr) had faster- than-average growth (Fig. 2). Estimates for the oldest herring (back-calculated over a span of >6 yr), however, gave mixed results. The trends in fish of 5 yr and younger may have been caused by en- vironmental influences because the trends occur in sets of years (fish aged 2-4, when captured, spent their first or second growth years in 1970-72, and those aged 4-7 spent their first or second growth years principally in 1966-69). GROWTH The average size-at-age data in my samples of Pacific herring from the eastern Pacific and east- ern Bering Sea and data from the literature for those regions usually formed two stanzas on Wal- ford graphs and inflected at ages 2 or 3 (see Figure 3 for examples). The data for Norwegian and Mur- man stocks of Atlantic herring, Clupea harengus harengus, (Svetovidov 1952) also formed two stan- zas and intersected at age 2. Although the stanzas for all of my back-calculated data from the eastern Pacific Ocean intersected at age 2, stanzas for two populations from California (data from Spratt 1981) intersected at age 3, and a plot of Naumenko's 707 FISHERY BULLETIN: VOL. 84, NO. 3 10 < a: < LU > OS O LL z < LU o OH Z o < > G LU (J < H Z LU u LU a: c (+)5- 0 - (-)5- 10 - - • t * . A A A A A • A • • • • A ~ A A BACK CALCULATED TO 1ST YR. MARK • • BACK CALCULATED TO 2ND YR. MARK i i 1 1 1 1 1 1 1 i _L 2 3 4 5 6 7 NO. ANNULI BACK-CALCULATED 10 Figure 2.— Relationship between extent of back-calculation from scales and the body length estimated at 1 and 2 yr in Pacific herring from Auke Bay, AK. Points indicate the deviation of size estimates for age-1 and age-2 Pacific herring from the average for all annuli (the second annulus was taken as the first year's mark). (1979) data from the eastern Bering Sea had only one stanza. Regardless of the data source, linear regressions (Walford regressions) closely fit the data in the growth stanzas (Tables 1, 2). The method of aging Pacific herring can influence estimates of growth. In the data that I examined for this study, adult stanzas based on back-calculated lengths usually had lower slopes than adult stanzas based on terminal-lengths-at age (Table 1). Further- more, the plots of back-calculated data inflected either at 2 yr or not at all, in contrast to plots of lengths-at-terminal-age, which inflected at 3 yr in three of six examples (Table 1). Important factors, however, remain uncontrolled in this comparison. For instance, the lengths-at-terminal-age from the literature were based on summer sampling; hence, they include additional growth after annulus formation. The lengths-at-terminal-age were from populations near or on the open coast, which may grow faster than populations from protected and possibly less productive waters within the Alex- ander Archipelago. Furthermore, it is not clear that the Alaskan data for lengths-at-terminal-age used the second scale annulus as the first year's mark. Walford graphs for Pacific herring from Tomales Bay, CA (data from Spratt 1981), and the eastern Bering Sea (this study) indicated that juvenile growth success and age at inflection (intersection of juvenile and adult stanzas) are more important determinants of adult size at age than either length at year 1 or the slope of the adult stanza, the adult growth proportion (Table 1; Fig. 3). The data in- dicate that herring from the Bering Sea quickly outgrow those from Tomales Bay although the BL's of the two groups were almost identical at ages 1 708 QUAST: BODY WEIGHT, FAT, AND GONADS OF PACIFIC HERRING Table 1 .—Growth characteristics and growth parameters of Pacific herring from the northeastern Pacific Ocean and eastern Bering Sea, based on data from the present study and from the literature. Growth is portrayed by the Walford version of the constant-proportion growth model (see text). Because Reid's (1971) data were gathered from a summer fishery, body lengths are longer than they were at the time of annulus formation and may not be comparable to back-calculated data or to lengths-at-terminal-age collected on or near the time of annulus formation. The inflections column refers to the junction of juvenile stanzas with stanzas for adults. Juvenile stanzas on the Walford graphs were fit by eye to sizes at ages 1 and 2, or ages 1-3; adult stanzas were fit by least squares. Size at age 1 Inflec- tion Adult stanza Aging Capture location (mm) at year Intercept Slope R2 method Source Back-calculated lengths: Auke Bay vicinity 93.3 2 64.89 0.709 0.998 scales this study Hood Bay, Chatham Strait 90.8 2 66.31 0.664 0.995 scales this study Katlian Bay 101.5 2 81.50 0.659 0.991 scales this study Carroll Inlet 102.7 2 83.61 0.632 0.999 scales this study Eastern Bering Sea 112.8 2 79.26 0.727 0.993 scales this study Eastern Bering Sea 90.3 ?1 88.79 0.722 0.999 scales Naumenko 1979 Lengths-at-terminal-age: Auke Bay vicinity — 2 64.94 0.716 0.983 scales Blankenbeckler 19792 Prince William Sound, AK 131.4 3 40.60 0.859 0.985 scales Reid 1971 Kodiak vicinity, AK 132.1 2 + 55.99 0.792 0.990 scales Reid 1971 Southeastern Alaska 145.1 2 + 52.14 0.788 0.967 scales Reid 1971 San Francisco, CA 113 3 44.87 0.816 0.989 otoliths Spratt 1981 Tomales Bay, CA 113 3 36.95 0.871 0.996 otoliths Spratt 1981 No inflection apparent. 2Blankenbeckler, D. 1978. Age, growth, maturation, and parasite occurrence of Pacific herring (Clupea pallasi) from southeastern Alaska, 1974 through 1976. Alaska Dep. Fish Game, Tech. Data Rep. 39, 88 p. and 2, and the adult stanza was steeper for herring from Tomales Bay. The Bering Sea herring, how- ever, inflected to a steeper slope at age 2 rather than age 3. Environment may not determine the time of inflection in Pacific herring because juveniles both from the Bering Sea and Tomales Bay had similar BL's during the first 2 yr (Fig. 3) although the environments of the locales probably differ greatly. Weight-Length Relationships Total weight (W, grams) relates to BL (milli- meters) in fresh Pacific herring from the Auke Bay vicinity as W = (4.4467 x 10"6)BL3-2232 (N = 491; R2 = 0.97). The lower confidence limit for the ex- ponent exceeds 3.0, and the exponent exceeds 3.0 in reports for herring in most locales; e.g., Pacific herring from Tomales Bay, 2.93 (Spratt 1981); San Francisco Bay, 3.23 (Spratt 1981); the east coast of Vancouver Island, 3.26 (Hart et al. 1940), and Barkley Sound, British Columbia, 3.46 (Hart et al. 1940); and in Atlantic herring, 3.15 and 3.5 (Hart et al. 1940). Many differences between exponents, as cited, may not be biologically significant because weight-length relationships vary seasonally and be- tween sexes, even in eviscerated fish. The exponent for the relationship between BL and total weight probably exceeds 3.0 in healthy herring populations because, as noted in later paragraphs, both eviscer- ated and gonad fresh weights also have exponents >3.0 when related to BL. Eviscerated weight of Auke Bay herring also had an exponential relationship to BL that significantly exceeded 3.0 [(W = 5.0894 x 10"6)BL3 16640; Fig. 4]. In theory, evisceration avoids large potential weight variations caused by seasonal changes in gonads and fat deposits about the viscera, and vari- able food content; yet, eviscerated weight (Sy ■ x = 0.1030) was at least as variable a function of BL as total weight (Sy -x = 0.0953) in the same specimens, and both total weight and eviscerated weight had the same coefficient of determination (0.97). The lack of decreased variability in the weight of evis- cerated herring, as a function of BL, compared with whole fish is evidence that building of visceral fat and gonads does not simply add weight, but rather that some compensatory mechanism may act be- tween these apparent weight sources and the evis- cerated body. In contrast to the results of Hart et al. (1940), Hickling (1940) found markedly low exponents, 2.13 and 2.37, for the relationship between eviscerated weight and BL for Atlantic herring from the North Sea, values that are strikingly lower than those ex- pected for fishes in general. For example, Quast (1968) gave exponents of 2.7-4.5 for 32 species of marine fishes in southeastern California, including 3.9 for the northern anchovy, Engraulis mordax. Hickling' s exponents may be too low because the 709 FISHERY BULLETIN: VOL. 84, NO. 3 *-,- CO CD >- N 3 » CO "5 CD Q. .Q c "O c 5 c © s * >• CO C W I £ S E-.9 ffl o -o > 3 ■— S o <" 2 !|l to £ « » o § BQ « -a a. a, E ■ 53 J 0) B 0 £ .Q E O) CD i s I £? » © p E § J= o o ^B "- g « < $ ™ -T 0> "D S - i O) o _. ■= EI 0) ^ CO CD -g T3 £ 6 g a di- ss c © S .2 £ Q) C c £ O Q) ■•- Q. CD sa-i o c Olffl » « w S Q) C c f 8 s < £ -6 **& « a> 5 rf "D < CD S^ ** CD "P E co | O) — en o S £ CO to 0- "J "S CD co C/> c . o CD | o ^ 3 Q_ 10 .2 CO °> V >< c _* CD a> ° ^ — co o >. ja ^ •o r- CO O S ts 0) "O E -o ^ te E m S & < s m hTJ V CD I "D (0 C J5 X3 CD •^ •o c IS m o o. 0) h- =3 CO- CO o CO -C O) c * c o T3 n (O CD o c CD k_ CD >*— TO CD O) CO 1— CD > CO >. E 'cj > >> CO CD CD j*: 3 < CD -C ■*-- c a> c ^ o l_ CO CD O) -C O **— (0 o CO CO 5 CL ^_ ^ '— lO o LL O _0^ CO CD co to c CD O C W CD 3 I— CO O a. CO o c jk CO CD 3 O CO -Q CO C > o i o- CO I— CD CO (1) (11 to r 01 ID CD CO CO m c co TO ra >, k_ COCO m h u CO a r u X co O ^. (0 ^ aj r a) jj: o 3 > < E E_ i:l c >, Q) CO co a, .E ^: CD ^ CD CD D) < co in NOi-nincooou) "-towot^'-toonintD CDOlW'tiniDSNCOCOCOCO ■^••■-CMOUCVJCNJOJOJCNJCvJCNJCM oqcO'^^LnooojcM'^^-o OJCOOOT^CD^T^Tt'^OOCO T-cocncMcO'a-cot^cooocj) ■"-■■-T-CNJOJCVJCViCMCMCNICN inmconcDcocMOCMOi^s air^uSi^^cTicviTtuSiricDccJ ■^•I^OlOi-^-CvJCNJCVJCAjeMCM i-i-i-CvJC\JCNJCMOJCNJCVJC\JC\J siflsoioi^wuioicoow cvicDcccDoo's-ai^ccicociicjD O^h-CDOi-i-CVJOJCvJCJ-^ i-i-i-i-CNJC\JCMOJCMC\JCNJCvJ coa>^tcNji-~.cvjo>coc3)or^T- inCOOi-OJCMCOCOCOCOCOCO t-t-WCvJCvJCvJCvJCMCvJCAJCMOJ inncoS'ttos^ ^cocjtDcocooocr) OU00003i-C\JC\JCM f-^-i-i-CMOJCMCN II I I I cO'*c\jcqcoc\jcocoi^coe\ico cbocbcbco'oo^cri'^iricbcb cMmcoi^cocoOTOcnoajcn COO)0)(Dt-S*-t- di/iViriiricor-ic oiwiniosoococo I I ocqcocvicoiflqN'-cooo) ninsojoiO'-'-T-T-cMCM i-T-^-t-^CMCNJCNJCNJCMCMCNJ noqc\iqinqcoc\iwcoc\i-NCMW i-t-i-»-i-C>JC\JCMCVJC\JCMCNJ c\jcqtqif)i^cO'*oo>cooj I |Qp>cocooioo>c\icoincp I Icor^OlOT-T-^CVJCvJCMCM i-T-T-CMCMCMCMWOJWW CO CNJ CNJ CO CO CO CD CD CD CO q-co 5 '-'-'-i-CvJCJCMCAi T-pjcoM-iniDScooiOr-ojn CD a> CD o^ > w < <1) E a C3 a. O a < s CA S < Q. co CD o o CO 0. a. c (0 CD O) < CO O ffi oo j: CO O . CO en ^3 CO c 9 ci co 2 - ra Q 710 QUAST: BODY WEIGHT, FAT, AND GONADS OF PACIFIC HERRING • AUKE BAY, BACK-CALCULATED (THIS STUDY) *" * TOM ALES BAY, CALIF., TERMINAL LENGTHS (SPRATT 1981) 0 ° BERING SEA, BACK-CALCULATED (THIS STUDY) A * BERING SEA, BACK-CALCULATED (NAUMENKO 1979) 50 100 BL. 150 (MM) 200 250 Figure 3.— Examples of Walford trends in body length of Pacific herring from widely separated locales in the eastern Pacific Ocean and Bering Sea. Heavy solid line through origin is the line of zero growth, and numbered points indicate ages in years. Adult stanzas were fit by least squares, and the juvenile stanzas were fit by eye. Data from Naumenko (1979) represent 25 yr of collections. effective range of BL's was limited (near 50 mm) in his data sets and his data were grouped in 10 mm size classes. (In contrast, BL's extended over about 130 mm in the Auke Bay herring, and lengths were taken to 1 mm.) In Pacific herring, the relationship between evis- cerated weight and BL varies with season and sex (Fig. 5), and the relationship for Atlantic herring should vary similarly. Although Hickling (1940) con- cluded that regressions of eviscerated weight on BL differed by sex in Atlantic herring (W = 0.0661 BL2 312 in males, and W = 1.1471 BL1-456 in females), his samples probably were too restricted seasonally to estimate reliably the relationship be- tween eviscerated weight and BL for all seasons. Because of seasonal variation in fat content of the musculature (discussed in the next section), data for a single season cannot represent an average over all seasons in Pacific or Atlantic herring of either sex. 711 FISHERY BULLETIN: VOL. 84, NO. 3 5. 175 — 4.425 r- X u Q LU H 3.675 < LU U in > LU 5 2.925 2.175 Y = -12. 18836 + 3. 16640 X N = 491 _L l J L J L 4.600 4.800 5.000 5.200 LN BODY LENGTH (X) 5.400 Figure 4.— Relationship between body length and eviscerated body weight in fresh Pacific herring in the vicinity of Auke Bay, AK, sexes combined. Variates transformed to their natural logarithms (LN). Points represent 1-9 specimens. Seasonal Cycles in Fat and Gonads Adult Pacific herring feed chiefly on zooplankton and small fishes (Hart 1973). In the Auke Bay vicin- ity, zooplankton peak in abundance in June or July and are virtually absent from November to March (Fig. 6; fig. 3 in Carlson 1980). In an unpublished study of Auke Bay herring, stomachs were mostly empty during late fall and winter (R. E. Haight, cited in Carlson 1980). Pacific herring spawn in Auke Bay in late April or May but may spawn as late as 4 June (Wing2). Eggs hatch 14-20 d after spawning, based on incu- bation temperatures for herring in British Colum- bia (Outram 19653) and temperatures for mid- April and May in Auke Bay, which are similar to 2B. L. Wing, Northwest and Alaska Fisheries Center Auke Bay Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 210155, Auke Bay, AK 99821, pers. commun. November 1981. 3Outram, D. N. 1965. Canada's Pacific herring. Dep. Fish. Can., Ottawa, Fish. Res. Board Can., Biol. Stn., Nanaimo, B.C., 23 p. those for British Columbia (Wing4). The time of spawning seems optimal to allow spawned fish and their newly hatched larvae to feed during the heaviest zooplankton concentrations of the year (Fig. 6). Because the peak in zooplankton abundance is relatively brief, the period immediately after spawn- ing is critical for fattening of adults and for growth and survival of newly hatched larvae. Feeding and fattening of all life stages of Auke Bay herring may also be aided by the submarine illumination afforded by the longest days and highest levels of light, early in the summer. Fat accumulated about the viscera during the period of maximum zooplankton abundance and reached highest indices shortly afterward, about mid-July (Fig. 6). It then declined rapidly but slightly differently in each sex. There is evidence, also, based 4B. L. Wing, Northwest and Alaska Fisheries Center Auke Bay Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 210155, Auke Bay, AK 99821, pers. commun. July 1983. 712 QUAST: BODY WEIGHT, FAT, AND GONADS OF PACIFIC HERRING JAN I FEB I MAR ■ APR H MAY I JUN ' JUL ' AUG ' SEP ' OCT1 NOV ' DEC Figure 5.— Seasonal variation in eviscerated weight as shown by monthly samples of fresh Pacific herring near Auke Bay, AK, given as percentage departure from the weight predicted by the general eviscerated weight/BL regression for these fish (see Figure 4). The percentage departure is given relative to its yearly average to highlight seasonal changes. Data fit by eye. on the water content of the musculature, that intra- muscular fat varied seasonally and paralleled the development of visceral fat— water content of evis- cerated body sections for the sexes behaved in an opposite fashion to visceral fat, being highest in April-May and at low levels between June and Octo- ber (Table 3). In contrast to the water content of the musculature, eviscerated weight increased relative to BL after May (Fig. 5). If the increase in eviscerated weight were caused by increased somatic hydration, variation in hydration would have paralleled variation in eviscerated weight, but instead, the values for hydration decreased after May. Some other factor must be responsible for the increased eviscerated weights after May, and a like- ly candidate is fat, because eviscerated weight in- creased over the same period that visceral fat was building. Hart et al. (1940) also described an ap- parent reciprocal relationship between water and oil content in Pacific herring from British Colum- bia, and Love (1970) discussed the same relation- Table 3.— Average hydration of musculature as a percentage of wet weight, by month in Pacific herring from Auke Bay, AK. Males Females Month N Percent N Percent January 15 69.2 6 68.4 March 17 71.1 19 71.0 April 26 75.3 25 76.0 May 29 77.0 29 76.0 June 8 61.7 12 66.5 July 8 60.8 12 61.5 August 10 60.7 10 61.1 September 23 61.7 30 62.7 October 12 61.9 18 61.5 November 1 62.6 6 60.6 December 3 65.0 25 63.1 ship in Atlantic herring and other fish species with fatty tissues. The timing of gonad development, as indicated by seasonal development of gonads, differed in the sexes in Pacific herring from Auke Bay. Males were 713 FISHERY BULLETIN: VOL. 84. NO. 3 ' JAN ' FEB ' MAR ' APR ' MAY ' JUN ' JUL ' AUG1 SEP ' OCT ' NOV1 DEC Figure 6.— Three annual cycles that relate to the condition of Pacific herring in the vicinity of Auke Bay, AK: A visual index of visceral fat (see text); gonad indices based on (wet) gonad weights as a percentage of the eviscerated (wet) body weights that would be expected at various BL's (see Figure 4); and an annual cycle of zooplankton density, from displacement volumes for 1962-64 given in Wing and Reid (1972). Points based on less than five specimens are enclosed in parentheses. Curves fit by eye. 714 QUAST: BODY WEIGHT, FAT, AND GONADS OF PACIFIC HERRING nearly ready to spawn in November but females delayed readiness until perhaps 4 mo later (Fig. 6), a delay that was confirmed by visual judgments of maturity, see table below (sample size in paren- theses): Percentage of herring judged ripe Sept. Oct. Nov.- Jan. Mar. Males Females 4(23) 0(31) 92(12) 11(18) 95(19) 79(38) 94(18) 90(20) These data differ in some important respects from those of Hay and Outram (1981) for Pacific herring in British Columbia. Their gonadosomatic index has sharper peaks in maturity of gonads and different timing of the peaks than the Pacific herring from Auke Bay. For example, in their data, testes were only developing (a low gonadosomatic index) in Oc- tober (the fish spawned in late February and early March), but testes were near maximum fullness (high index values) in October in herring from Auke Bay (Fig. 6). However, Hay and Outram used total weight in their index. If total weight is used for the index, the divisor will include a considerable weight of fat about the viscera in the fall and negligible weight in the spring, with the result that even if gonad weights remain the same from November to February, the decline in the amount of fat would cause the index to increase. In my study of the Pacific herring in Auke Bay, I divided gonad weight by eviscerated body weight, which should avoid an appreciable error in the gonadosomatic index that would be caused by variation in visceral fat. Within each sex, seasonal profiles for gonad in- dices are nearly opposite the profiles for indices of visceral fat (Fig. 6). The annual cycles in fat and gonad indices (Fig. 6) in Pacific herring from Auke Bay resemble those noted by Blaxter and Holliday (1963) for spring spawning in Atlantic herring: "In winter-spring herring the good feeding conditions in late spring and early summer (after spawning) build up the fat reserves. With development of the gonads in late autumn feeding stops and spawning in December-March means that the fish overwinter and spawn with fat reserves considerably lower than the autumn spawners." Visceral fat in male Auke Bay herring is lowest in winter (perhaps as early as November), but in females does not reach lowest values until April. Correspondingly, the testes build rapidly in late summer and fall and appear to be heaviest by October or shortly after, but the ovaries are not at their heaviest until shortly before spawn- ing, in April or May. Hydration is not responsible for sexual differences in development of gonad weight from January to March because, as the following table indicates, hydration remains virtual- ly constant from November to March in both sexes (Table 4). Table 4. — Average hydration of gonads, as a percentage of wet weight, by month in Pacific herring from Auke Bay. AK. Males Females Month N Percent N Percent January 14 76.2 5 73.6 March 16 76.1 17 71.3 April 39 82.6 33 84.5 May 24 83.7 24 77.2 June 6 75.5 9 77.7 July 18 73.6 19 76.2 August 25 77.9 21 80.7 September 19 77.6 25 78.6 October 12 76.9 18 74.5 November 1 76.1 6 72.7 December 3 74.2 25 71.2 This seasonal, mirror imagery between develop- ment of fat and gonads, with the images differing for sexes, is evidence for a strong physiological coupling between fat depots and gonads. Fat depots enable Pacific herring to accommodate two critical cycles in their life history that are badly out of phase: The zooplankton cycle, with its brief, summer peak that builds fat depots rapidly and is followed by low levels of food abundance from October to March; and the gonad cycle that slowly removes fat from the depots with the slow building of testes from July through October and the slower building of ovaries from July through March. Are the seasonal cycles of gonad maturity in Pacific herring from Auke Bay determined by gene- tics or are the gonads responding principally to cyclical changes in the immediate environment? lies (1984) felt that Atlantic herring are remarkably in- dependent of their environment. Genetic control of gonad maturity seems likely except for spawning, which appears to respond to local temperatures (Outram 1965, see footnote 3). Gonads must build well in advance of spawning, and spawning dates vary from November in the southern limits of the eastern Pacific range (Spratt 1981) to June in Auke Bay. Female Auke Bay herring mature sexually and use fat deposits later in the fall than do males and thus anticipate a later spawning date. Male herring in the eastern Pacific Ocean, in contrast, appear to build testes early enough to spawn at any date be- 715 FISHERY BULLETIN: VOL. 84, NO. 3 tween November and June. Only the ovarian cycle seems to correspond closely to the local environmen- tal conditions that seem optimal for larval growth and survival. Possibly, the genes that are respon- sible for local adaptation of spawning stocks are sex linked for females and are selected through larval survival. Annual Production of Eviscerated Weight and Reproductive Tissues Although Pacific herring usually have only one major spawning per site in the Auke Bay vicinity, there may be a succession of lesser spawnings each spring. Unspawned fish are rarely seen as late as July (author's observations and comments by salmon fishermen who jig herring for bait). Although Wing (see footnote 2) recorded spawnings in Auke Bay between 24 and 29 April 1973, herring must spawn for at least 2 mo in Auke Bay because some fish sampled in 1973 were partially spawned or ripe and running in May and June (Fig. 7). Presumably, local conditions influence the number of eggs deposited on any date. The relationship between fecundity, as indicated by mature ovarian weight, and BL was greater than cubic, in agreement with data on other clupeiod species (Blaxter and Hunter 1982). Samples of Auke Bay herring had an exponent of 3.94 (Fig. 8), within the range (3.07-4.50) for Atlantic herring as given by Paulson and Smith (1977), from the literature. These authors gave an exponent of 3.32 for Pacific herring they sampled in Prince William Sound. Perhaps, the exponent for fecundity would have been higher for the herring Paulson and Smith sampled in Prince William Sound had their collec- tions included smaller fish (their smallest were near 180 mm long, but fish as small as 130 mm were available in samples from Auke Bay). The exponent for testicular weight was considerably higher than that for ovarian weight in Auke Bay herring (Fig. 8); however, the difference may not be real because the confidence limits for the two exponents over- lapped considerably. The scatter in the plots of gonad and testes weights on BL for Auke Bay herring (Fig. 8) and for Pacific herring from Prince William Sound (fig. 1 in Paulson and Smith 1977) indicate that some of the herring may have been partially spawned when they were collected (fully spawned fish were not used in my data). If samples for fecundity are taken in the spawning season, there is the risk that some PERCENTAGE IN STAGE: 1-10 11-50 51-100 0 0 0 1 10 8 1 8 10 :15> :49:: ::20: 0 0 0 0 85 60 0 0 0 i/i 2 Q 28:: :13x < o 0 0 0 0 0 0 0 0 o3 25 19 :41: u LU ::5:: 0 0 0 0 0 57 2« ::5::: 24 :29> 14: > ts 86 92 :25> 10 0 0 0 2 :;43';': 71 86 CL 3 r- < S6 0 0 35: 64 :5:: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7 39 : 1 7: .Ian Fph Mar Anr Ma v ini Can Or-t Mnu r\ar- MATURITY STAGE: 1. IMMATURE 2. SPAWNED OUT AND REGENERATING 3. REGENERATING U. MATURING 5. RIPE 6. PARTIALLY SPAWNED OR RIPE AND RUNNING 7. SPAWNED OUT N = 21 0 38 80 59 20 55 80 54 30 7 29 Figure 7.— Maturity of Pacific herring near Auke Bay, AK, by month (sexes combined). Numbers in boxes are percent- ages of herring that were visually classified into maturity stages on examination. Total fish by month are given in the bottom line. Data for February were extrapolated from January and March. 716 QUAST: BODY WEIGHT, FAT, AND GONADS OF PACIFIC HERRING fish will have spawned partially and that fecundity estimates will be too low. When the relationships between BL, weight, and fecundity in Pacific herring from Auke Bay were used in a model of annual changes in gonad weight and eviscerated weight, production of eviscerated weight decreased rapidly with age or size (Table 5, col. 3). Gonad production (Table 5, col. 4), in con- trast, increased yearly but appeared to approach an asymptote at about 31-34 g in the oldest fish. With age, more of the annual product (annual increment in eviscerated weight plus gonad weight) was par- 3.525 2.775 2.025 1.275 h- x 2 0.525 LU a < 4.125 z o u 3.375 LN CONW =~ 20.44468+ 4.41971 (LN BL) MALES N = 63 i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — r 2.625 1.875 1 .125 LN CONW =- 17.78385 + 3.93569 (LN BL) FEMALES N = 61 ' I ' ' ' I 1 ' I J L _L I J L 4.875 5.025 5.175 LN BL (MM) 5.325 5.475 Figure 8.— Relationships between (wet) gonad weight (GONW) and BL in fresh Pacific her- ring collected from March to May 1973 near Auke Bay, AK (variates transformed by natural logarithms (LN). Data (not shown) that formed a separate cluster of points near the abscissa for each sex probably represented spawned fish and were not used in the regressions. Points represent 1-2 specimens. 717 FISHERY BULLETIN: VOL. 84, NO. 3 oo co cd — S m O) 5 „ io2 s o ■o c -= = CO CD 2 "° C o ^ O U Q- CO CO .C ,_ « •- -C >, -~. O O) ~ 3 !£, q) '55 °> p cj CO £. Q.53 5 tt CO co CD _ CO 'CD g> co ^_ >> _ « O O) ^ — ' 0)T3 £ «- CO 0) C o — o < 3 f, CD T3 CD k- o y q. Q..C (A ■a ■D 2 , k CO c o CO .c co CD D) 3 O) o 0) § C en g C < > CD "O 10 o o 3 -° Q. Q. E.2>co D (1)5 X T3 O CO CO £ co .2 Is co CD ^ "S S co Is CD co CD .-^ COS CO C\J co£ «2 o.S>|3 »5o > c UJ Si- CO k_ rn^^ c (d y'a) .E O -Ot-C0NCDCDCD1O C\JCMOJCMi-t-t-i-t-i- I N CAI t- Bl rri- i- .- i- O5C0i-mc\JOiir>T-cqcp cbcoiridi/i^^^cjco ajcoco^-cocococococo m^ococor-cqwni- i^T-'coo6cocoo>ic\JcAJc\i CMCOCOCOCOCOCOCOCOCO ncgs^^N^scpcq cbcoKT^'^-cooooSci'r^ i-i-OJNCMWCOncO wncDiflinoiSt- cD^mo5C\j^-cDooa>a) ■^t-i-C\JC\ICMC\JCM(M CDi-^i-OOCNJi-^CNCO oorv!oioo6cb'inr^oic3>o-r-T-T-7-cM i-i-^t-i-WWCMCMCMW wco^incDScooiOi-cy titioned into gonads, which by age 12 composed nearly all of the annual production (Table 5, col. 8). As the herring grew, the annual product was more closely related to BL than to eviscerated body weight (Table 5, cols. 6, 7), evidence that individual herring in the Alike Bay vicinity may use food more in proportion to their BL than their weight. Further- more, although the relationship between annual product and BL was nearly constant within age groups 3-6 and 8-12 (Table 5, col. 6), productivity was much lower in the 8-12 group. Young Auke Bay herring may be more successful for their size in find- ing food than are older individuals, because impor- tant foods needed by older herring may be scarce in the Auke Bay vicinity, which is about 80 nmi (148 km) by water from the open ocean. There is indirect evidence from the characteristics of growth and production in Auke Bay herring and growth characteristics of other Pacific herring in the northeastern Pacific Ocean and Bering Sea that not only growth but the annual product relative to her- ring size and the partitioning of the annual product may vary with the population. If the relationship between eviscerated weight and BL in Auke Bay herring is used with growth data of Pacific herring from other locales in the eastern Pacific Ocean and Bering Sea (Table 1), striking differences are visi- ble in the annual product of eviscerated weight (Fig. 9). For example, if 190 mm herring (3-yr-olds) in the Bering Sea produce gonads in the same proportion to eviscerated weight as 190 mm herring in Auke Bay (5-yr-olds; Table 5, col. 8), gonads and eviscerated weight would each form about one-half of the annual product of Bering Sea herring. How- ever, this proportion would be much too high for herring in Bering Sea during their first year of gonad production, according to the model based on Auke Bay herring, if Bering Sea herring mature as 2- or 3-yr-olds. If production of eviscerated weight and gonads is scheduled according to age, rather than proportion of annual product, similar conflicts result; thus, Pacific herring from different regions probably differ in characteristics for production of eviscerated weight and gonads. CONCLUSIONS Pacific herring grow in the eastern Pacific Ocean according to the constant-proportion model; i.e., growth in one year is a constant proportion of the amount grown the previous year. Growth stanzas (regions of constant growth parameters) for juve- niles and adults usually inflect near age 2, and the change in growth is probably related to sexual 718 QUAST: BODY WEIGHT, FAT, AND GONADS OF PACIFIC HERRING 50 40 3 |(a8.0) BERING SEA (NAUMENKO) 4 30 G LU H < a: LU O CO > 1 20 < 10 SAN FRANC J L_L TOMALES BAY 100 150 BL (MM) AT 200 ANNULUS 250 Figure 9.— Hypothetical growth in (fresh) eviscerated body weight by Pacific herring at locales in the eastern Pacific Ocean and eastern Bering Sea. Data were based on relationships between eviscerated body weight and BL in samples from the Auke Bay vicinity and data on growth reported in the literature (see Table 1). Numbered points are ages at the beginning of annual growth increments. The dashed vertical line is for comparative purposes and intersects the graphs at 190 mm BL. The second annulus was taken as the first year mark in specimens from Auke Bay, Katlian Bay, and Bering Sea (this study). maturity. Size of adults is influenced more by growth rate of juveniles and the size at inflection of growth stanzas than by the constant of propor- tional growth after inflection. In the Auke Bay vicinity, a sharp increase of zoo- plankton abundance in June is the determinent for the annual cycles of fattening and spawning in Pacific herring, and spawning in April or May seems optimally timed for growth of newly hatched fry. In summer, fat builds rapidly about the viscera and in the musculature of adults, as a reserve for gonad development and metabolism in fall and winter when food is scarce and herring do not feed. lies (1984) found that in Atlantic herring fat is assimilated and deposited almost unchanged during the feeding cycle and is not utilized for metabolism until the metabolic pool of protein is exhausted. He also hypothesizes that annual somatic growth declines 719 FISHERY BULLETIN: VOL. 84, NO. 3 with gonad growth and ceases with depletion of the protein pool. Male Pacific herring from Auke Bay build gonads and use their fat reserves more rapidly than do females. Testes may be near spawning condition in November, but ovaries are not full sized until April. Males may be ready or nearly ready to spawn in November over the entire eastern Pacific Ocean, but females delay spawning until local conditions of temperature and food abundance are optimal for larval growth. The exponents for total and eviscerated body weights, as functions of BL, exceed 3.0 in Pacific herring from Auke Bay, and probably in Atlantic herring as well because of their similar morphology. Weight of mature gonads also have a greater-than- cubic relationship to BL in Auke Bay herring (the exponent was 4.4 for testes; the exponent of 3.9 for ovaries was within the range for ovaries in Atlan- tic herring). The annual product (eviscerated body weight and gonad weight) is constantly proportional to BL through ages 2-6 and also through ages 8-12 in Pacific herring from Auke Bay, but the proportion is considerably lower in the 8-12 group. However, despite the two levels of production relative to BL, annual production corresponds more closely to BL than to eviscerated weight. Annual production may be lower relative to BL in the older group because suitable foods for adults may not be abundant in the Auke Bay vicinity. Most annual production in young Auke Bay herring goes into growth of eviscerated body weight. After age 6, production of sex prod- ucts predominates, and by age 12, sex products com- pose over 90% of annual production. Pacific herring probably develop genetic stocks that are distinguished by locale, spawning time, and cycles of gonad maturity and fat utilization in the females. The stocks probably are distinguished also by growth rate, age, or size at growth inflection and by partitioning of annual product between eviscer- ated body weight and gonads. ACKNOWLEDGMENTS I especially thank Elizabeth L. Hall, NMFS Auke Bay Laboratory, for her exacting scale measure- ments and painstaking preparation of specimens, and H. Richard Carlson and Richard E. Haight, also of the Auke Bay Laboratory, who obtained the her- ring samples from the Auke Bay vicinity, sometimes under severe weather conditions. My thanks to the Alaska Department of Fish and Game for samples from Carroll Inlet, Katlian Bay, and the eastern Ber- ing Sea, and to Petersburg Fisheries, Inc., for the opportunity to collect specimens from the herring fishery at Hood Bay. Helpful reviews of the manu- script were provided by H. Richard Carlson, Robert R. Simpson, and Bruce L. Wing of the Auke Bay Laboratory. LITERATURE CITED Blaxter, J. H. S., and F. G. T. Holliday. 1963. The behavior and physiology of herring and other clupeids. In F. S. Russell (editor), Advances in marine biology, Vol. 1, p. 261-393. Acad. Press, N.Y. Blaxter, J. H. S., and J. R. Hunter. 1982. The biology of the clupeoid fishes. In J. H. S. Blax- ter, F. S. Russell, and M. Young (editors), Advances in marine biology, Vol. 20, p. 1-223. Acad. Press. N.Y. Carlson, H. R. 1980. Seasonal distribution and environment of Pacific her- ring near Auke Bay, Lynn Canal, southeastern Alaska. Trans. Am. Fish. Soc. 109:71-78. Hart, J. L. 1973. Pacific fishes of Canada. Fish. Res. Board Can. Bull. 180, 740 p. Hart, J. L., A. L. Tester, D. Beall, and J. P. Tully. 1940. Proximate analysis of British Columbia herring in rela- tion to season and condition factor. J. Fish. Res. Board Can. 4:478-490. Hay, D. E., and D. N. Outram. 1981. Assessing and monitoring maturity and gonad develop- ment in Pacific herring. Can. Tech. Rep. Fish. Aquat. Sci. 998, 31 p. Hickling, C. F. 1940. The fecundity of the herring of the southern North Sea. J. Mar. Biol. Assoc. U.K. 24:619-632. HOURSTON, A. S. 1958. Population studies on juvenile herring in Barkley Sound, British Columbia. J. Fish. Res. Board Can. 15:909- 960. 1980. Timing of herring spawnings in British Columbia, 1942- 1979. Can. Ind. Rep. Fish. Aquat. Sci. 118, 101 p. Iles, T. D. 1984. Allocation of resources to gonad and soma in Atlantic herring, Clupea harengus L. In G. W. Potts and R. J. Wootton (editors), Fish Reproduction: Strategies and tac- tics, p. 331-347. Acad. Press, N.Y. Jones, J. D. 1978. Growth of larval Pacific herring in Auke Bay, Alaska, in 1975 and 1976. M.S. Thesis, Univ. Alaska, Juneau, 23 p. Love, R. M. 1970. The chemical biology of fishes. Acad. Press, NY., 547 P- McLean, R. F., and K. J. Delaney. 1978. Alaska's fisheries atlas. Alaska, Dep. Fish Game, 2 vol., 81 p. Naumenko, N. I. 1979. Features of growth of young eastern Bering Sea her- ring, Clupea harengus pallasi. J. Ichthyol. 19(6):152-156. Paulson, A. C, and R. L. Smith. 1977. Latitudinal variation of Pacific herring fecundity. Trans. Am. Fish. Soc. 106:244-247. Quast, J. C. 1968. Estimates of the populations and the standing crop of 720 QUAST: BODY WEIGHT, FAT, AND GONADS OF PACIFIC HERRING fishes, pp. 57-79. In W. J. North and C. L. Hubbs (editors), Utilization of kelp-bed resources in southern California, 264 p. Calif. Dep. Fish Game Fish Bull. 139. Reid, G. M. 1971. Age composition, weight, length, and sex of herring, Clupea pallasi, used for reduction in Alaska, 1929-66. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSR-F 634, 25 p. RiCKER, W. E. 1975. Computation and interpretation of biological statistics of fish populations. Fish. Res. Board Can. Bull. 191, 382 p. ROUNSEFELL, G. A. 1930. Contribution to the biology of the Pacific herring, Clupea pallasi, and the condition of the fishery in Alaska. Bull. U.S. Bur. Fish. 45:227-320. ROUNSEFELL, G. A., AND E. N. DAHLGREN. 1935. Races of herring, Clupea pallasi, in southeastern Alaska. Bull. U.S. Bur. Fish. 48:119-141. Skud, B. E. 1963. Herring tagging experiments in southeastern Alaska. U.S. Fish and Wild]. Serv. Fish. Bull. 63:19-32. Spratt, J. E. 1981 . Status of the Pacific herring, Clupea harengus pallasi, resource in California, 1972 to 1980. Calif. Dep. Fish Game Bull. 171, 107 p. Svetovidov, A. N. 1952. Clupeidae. InE. N..Pavlovskii and A. A. Shtakel'berg (editors), Fauna of U.S.S.R., Fishes, Vol. 11, No. 1. Israel Program for Scientific Translations (translated from Rus- sian by Z. Krauthamer and E. Roifer), 428 p. Walford, L. A. 1946. A new graphic method of describing the growth of animals. Biol. Bull. 90(2):141-147. Whitney, R. R., and K. D. Carlander. 1956. Interpretation of body-scale regression for computing body length of fish. J. Wildl. Manage. 20:21-27. Wing, B. L., and G. M. Reid. 1972. Surface zooplankton from Auke Bay and vicinity, south- eastern Alaska, August 1962 to January 1964. U.S. Dep. Commer., NOAA, NMFS Data Rep. 72, 764 p. 721 CONTRIBUTIONS TO THE LIFE HISTORY OF BLACK SEA BASS, CENTROPRISTIS STRIATA, OFF THE SOUTHEASTERN UNITED STATES1 Charles A. Wenner, William A. Roumillat, and C. Wayne Waltz2 ABSTRACT Ages of black sea bass, Centropristis striata, from the South Atlantic Bight were determined from otoliths. Analysis of marginal increments showed that annulus formation occurred in April and May. The von Bertalanffy growth equation derived from back-calculated mean standard lengths at age was It = 341 (1 - e-0-2309(f+o.30io)^ wjjere £ js age jn years an(j \f _ standard length at age. The oldest fish was age 10. Centropristis striata is a protogynous hermaphrodite that undergoes sex succession at ages 1 through 8. The process of sex succession is described from histological examination of the gonads. The major spawning period is from March to May, and a minor spawn occurs in September-October. Mature males and females were encountered at age 1. Fecundity estimates ranged from 17,000 in a 108 mm SL female to 1,050,000 in a 438 mm SL fish, and were significantly related to length, weight, and age. The instantaneous rate of total mortality of C. striata from catch curve analysis, ranged from 0.721 in 1978 to 1.320 in 1981 for commercial fish traps and 0.726 in 1979 to 1.430 in 1981 for hook-and-line gear. Petersen mark-recapture techniques were used to determine the population size of C. striata on two shallow-water patch reefs. Conversions of these estimates to densities gave 14-125 individuals per hectare. The black sea bass, Centropristis striata (Linnaeus), is an important recreational and commercial ser- ranid (Huntsman 1976; Musick and Mercer 1977; Low 1981) that occurs along the east coast of the United States from Massachusetts to Florida, with occasional individuals as far south as the Florida Keys (Fischer 1978). Within this range, C. striata is thought to form two populations separated at Cape Hatteras (Mercer 1978). The northern popula- tion migrates seasonally from shallow waters along the Middle Atlantic and southern New England coasts during summer to deeper water in the south- ern part of the Middle Atlantic Bight during the winter (Musick and Mercer 1977). Black sea bass in the Middle Atlantic Bight are harvested commer- cially with traps in shallow water during summer and with otter trawl gear when aggregated in deeper water in winter (Frame and Pearce 1973). Commercial catches are almost exclusively from traps in that part of the South Atlantic Bight from Cape Fear, NC to Cape Canaveral, FL where fish- ing is largely confined to patch reefs (live bottom habitat of Struhsaker 1969 or inshore sponge-coral Contribution No. 205 from the South Carolina Marine Resources Center, Marine Resources Research Institute. 2Marine Resources Research Institute, South Carolina Wildlife and Marine Resources Department, P.O. Box 12559, Charleston, SC 29412. habitat of Powles and Barans 1980) at depths from 20 to 46 m. South Carolina commercial landings were as high as 350.7 metric tons (t) in 1970, but show large annual fluctuations (Fig. 1). Both the northern and southern populations have been aged by analyzing otoliths (Mercer 1978), with nine age groups identified north of Cape Hatteras and eight along the southeastern U.S. coast. How- ever, sampling techniques could have biased the findings on southern C. striata since fishes came from commercial catches which are frequently culled at sea (Mercer 1978). Black sea bass are protogynous hermaphrodites (Lavenda 1949), wherein most individuals function first as a female and later as a male. Most females mature by age 2; older age classes are composed predominately of male fish although sexually active males are in all age groups. Sexual succession oc- curs at ages 1 through 5 (Mercer 1978). The north- ern population spawns from June to October with peak reproduction in July and August off Virginia (Mercer 1978). There is insufficient published information to describe the life history of this valuable commercial and recreational species in the South Atlantic Bight in detail. This report describes aspects of the life history of C. striata from the South Atlantic Bight, including age and growth, sex ratios, size and age Manuscript accepted October 1985. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 723 FISHERY BULLETIN: VOL. 84, NO. 3 300 c o|2 c — 9 9 o e w Z il < i UJ 0.15 mm in diameter were counted since histological examination of the gonads of maturing and spent females showed only oocytes >0.15 mm developed during the spawning season. Total fecundity was estimated by expanding the mean of the subsamples to the total sample volume. Total fecundity was related to length and weight by standard least squares linear regression (Sokol and Rohlf 1981) and GM functional regression (Ricker 1973). Mortality Estimates Plots of logp frequency on age indicated that black sea bass are fully recruited to commercial traps and hook-and-line gear at age 4, so mortality analysis applies to age 4 and older. The instanta- neous rate of total mortality (Z) was estimated by standard least squares regression (Sokol and Rohlf 1981) from the slope of the right descending limb of the catch curve (Ricker 1975). Values of Z were also obtained by converting (appendix I in Ricker 1975) rates of survival (S) derived by Heinke, and Chapman and Robson estimates (Everhart and 727 FISHERY BULLETIN: VOL. 84, NO. 3 Youngs 1981). Not all fish collected were aged, so fish of known age were grouped into 1 cm length intervals by gear type for each survey to calculate percentages of each age in each size interval. These percentages were applied to the number of C. striata in each length interval to estimate age composition for the unaged fish (Ricker 1975). Population Estimates at Specific Reef Sites Petersen mark-recapture experiments were con- ducted at site 1 (lat. 32°30.3'N, long. 79°41.9'W; depth = 20 m; area =160 ha) during the summer of 1981 to estimate the population size of C. striata on this reef. In the summers of 1982 and 1983, we studied a second reef also (site 2: lat. 32°28.3'N, long. 78°14.5'W; depth = 23 m; area = 120 ha). These reef areas were defined by the presence of attached algae and invertebrate growth (porifera, corals, echinoderms, bryozoans, anthozoans, and ascidians) as observed with a HYDRO products TC-125-5DA low-light-level underwater television camera during transects across the sites (for more details, see Wenner 1983). Study areas were mapped with an EPSCO C-Plot II LORAN-C plotter inter- phased with a SITEX 707 LORAN-C receiver. Black sea bass were captured and recaptured at each site with commercial traps (Rivers 1966) and Florida snapper traps (0.9 m wide x 1.2 m long x 0.6 m high) fished for 45-90 min with cut clupeid bait (Brevoortia tyr annus and Alosa aestivalis). Black sea bass >20 cm TL, the approximate size of full retention in the traps, were measured to the nearest mm TL and tagged with 13 mm diameter plastic disc tags attached under the first dorsal fin with a nickel pin trimmed to the proper length and held in place with a 13 mm diameter plastic backing disc. Expan- sion of the swim bladder, due to reduced hydrostatic pressure, caused captured fish to float, so gas was released from the swim bladder with a 20-gauge hypodermic needle to enable fish to return to the bottom. For each experiment, 50-75 tagged fish, handled in the same fashion as those released, were held on the bottom in wire cages for about 24 h to determine tag-related mortality. Tagged fish were released over the reef, and sampling for recaptures started 24 h after tagging began. Experiments were completed in 48 h except at site 2 during the sum- mer of 1982 when tagging was interrupted for 48 h by weather. Preliminary estimates of population size are needed to determine sample sizes required for precise Petersen estimates (Everhart and Youngs 1981). Powles and Barans (1980) estimated the mean density of C. striata on reefs was 51 fish/ha from underwater television transects; expansion to the areas of our study sites gave preliminary estimates of 8,160 C. striata on site 1 and 6,120 on site 2. At site 1 we needed to tag 1,000 fish and examine 550 for tags to have an error no greater than 25% for 19 times out of 20 (1 - a = 0.95 and P = 0.25). At site 2, we needed to tag 1,000 fish and examine 500. We used the adjusted Petersen estimate (Ricker 1975, p. 78): N* = (M + 1) (C + 1) (R + 1) where N* M C R estimated population size number of fish tagged sample taken for census number of tags returned in the sam- ple taken for census. The biomass of C. striata for each year and site was estimated as Biomass = 21— x PE x g\ where n1 = number of tagged fish in each 1 cm TL interval n = total number of tagged fish PE = population size from the Petersen estimate g = weight in grams for the midpoint of each 1 cm TL size interval derived from the total length-weight rela- tionship a = number of 1 cm TL intervals of tagged fish. In addition, the upper and lower confidence limits were substituted for PE in the above expression for estimates of the biomass at those population sizes. RESULTS Age and Growth We believe that the central opaque zone of the sagitta may represent the first 1-4 mo of life in C. striata. This zone varied in length from 1.16 to 3.60 mm in the anteroposterior plane and from 0.56 to 1.54 mm in the dorsoventral plane (Fig. 4). We were not always able to make counts along a continuous 728 WENNER ET AL.: LIFE HISTORY OF BLACK SEA BASS DORSAL ANTERIOR i POSTERIOR w NUCLEUS CENTRAL FIELD VENTRAL — length- Figure 4.— Schematic representation of the left sagitta in young of the year Centropristis striata showing the orientation and direction of growth ring-counts in the central opaque zone, a = anterior, d = dorsal, p = posterior, v = ventral. line because grinding did not expose all rings equally in the central zone. Also, in some instances, count- ing was halted at a distinct mark, such as a ring more distinctive than others, and we followed this mark around the sagitta to a site where rings were again visible and resumed counting. The number of rings in the central zone varied because of the oto- lith asymmetry and with the direction of the count (Table 2). For example, we obtained the following counts from the central primordium in one specimen (number 9 of Table 2): 90 rings to the dorsal edge of the central field (d of Fig. 4); 95 rings to the ven- tral edge (v of Fig. 4); 129 rings to the posterior edge (p of Fig. 4). Since marginal increments on the otoliths should approach zero during the time of annulus formation, we calculated their monthly means to determine if one opaque band was laid down during each year on the sagittae of C. striata. Generally, a single an- nulus was formed during April and May in all age groups (Fig. 5). We found that the ring was deposited unevenly around the sagitta, with the dor- sal margin of the annulus being the last to be completed. We identified 10 age groups in the South Atlan- tic Bight population of C. striata, which exceeded the previous reports of 7 (Cupka et al. 19735) and 8 (Mercer 1978) groups. Observed mean lengths and weights increased with age; however, small sample sizes in ages 8 through 10 masked this trend (Table 3). Regressions of weight on length (TL and SL) and length on length are in Table 4. 5Cupka, D. M., R. K. Dias, and J. Tucker. 1973. Biology of black sea bass, Centropristis striata from South Carolina waters. Unpubl. manuscr. South Carolina Wildlife and Marine Resources Department, P.O. Box 12559, Charleston, SC 29412. Table 2.— Data from Centropristis striata examined for daily growth rings. Refer to Figure 2 for otolith morphology and terms (d, v, p, and a) used in the counts. Numbers in parentheses are ranges of several counts; dashes indicate no counts made. Fish Otolith Central field Daily ring counts TL SL WT Height Length Height Length No. (mm) (mm) (g) (mm) (mm) (mm) (mm) d V P a 1 66 54 4 1.56 broken 0.92 1.76 — (84-89) — 2 60 48 2 1.52 3.08 0.56 1.16 28 (25-26) (50-51) — 3 95 74 11 2.24 4.44 1.54 3.25 — (109-120) — — 4 125 98 30 2.92 4.88 1.54 3.60 — 135 — — 5 78 61 4 1.82 broken 0.98 broken 105 — — — 6 76 58 5 1.84 3.44 1.08 2.68 106 — 121 — 7 78 60 4 1.78 3.40 1.12 2.88 (98-106) — — — 8 81 63 5 1.92 3.60 1.52 3.40 51 — — 81 9 — 64 6 1.84 3.52 1.16 3.08 90 95 — 129 729 6-i 5- cr f> o h 4. -I < K < 3 s 2 < UJ 2 3- 2- (28) (246) FISHERY BULLETIN: VOL. 84, NO. 3 (167) (2310 M ~i 1 r M J J MONTHS Figure 5.— Mean marginal increment by month for otoliths of Centropristis striata. Number in parentheses represent monthly number of otoliths examined. Table 3.— Means (x), standard deviations and sample sizes for observed lengths (mm) and weights (g) by age for Centropristis striata. Total length Standard length n (x) SD Weight Age n (x) SD n (*) SD 0 185 94 25 186 73 20 186 17 16 1 818 163 27 830 127 21 822 70 38 2 2,712 215 28 2,714 167 21 2,688 152 65 3 4,271 249 34 4,263 192 24 4,246 228 102 4 2,376 291 40 2,371 222 28 2,350 348 142 5 951 337 46 950 256 32 904 520 206 6 497 375 48 497 284 35 460 711 266 7 138 395 50 139 299 38 121 823 280 8 48 394 50 48 301 38 43 838 289 9 10 406 58 10 305 46 7 816 383 10 4 404 45 4 303 35 3 685 85 Table 4. — Least square linear and geometric mean functional regression equations of weight (WT) on total length (TL) and standard length (SL), and length-length for Centropristis striata. Weight units are grams and lengths are millimeters. All least squares regressions were significant at o = 0.01. Least squares equation n r2 GM functional equation log10 WT = -4.375 + 2.800 log10 TL log10 WT = -4.328 + 2.978 log10 SL TL = -9 + 1.4 SL SL = 12 + 0.7 TL 12,281 12,284 12,473 12,473 0.97 0.98 0.97 0.97 log10 WT = -4.478 + 2.844 log10 TL log10 WT = -4.398 + 2.949 log10 SL TL = -12 + 1.4 SL SL = 9 + 0.7 TL Least squares regressions of SL (mm) on otolith radius (OR in ocular units) are log10 SL = 0.668 + 1.056 log10 OR; n = 12,011; r2 = 0.89. The intercept of the SL-OR relationship was used to obtain the mean back-calculated SL's at age which were lower than observed SL's in all cases (Table 5). Weighted least square estimates of von Ber- talanffy parameters, asymptotic 95% confidence Table 5.— Observed and back-calculated mean standard length in mm and von Bertalanffy standard length at age for Centropristis striata. Observed Back-calculated von Bertalanffy Age n SL SL SL 1 830 127 88 88 2 2,714 167 142 141 3 4,263 192 180 182 4 2,371 222 212 215 5 950 256 244 241 6 496 284 271 261 7 139 299 283 278 8 48 301 289 291 9 10 305 296 301 10 4 303 303 309 730 WENNER ET AL.: LIFE HISTORY OF BLACK SEA BASS limits and asymptotic standard errors were also derived from these data (Table 6). Estimates of an average asymptotic size (LJ depended not only on the number of age groups present and the distribu- tion of individuals within each group, but also on the curvature of the age-size relationship. An average asymptotic size of 341 mm SL appeared conser- vative. The largest fish aged was 390 mm SL and only 0.6% of all C. striata sampled were larger than 341 mm SL. The largest specimen caught off the South Carolina coast was estimated to be about 490 mm SL (S.C. Wildlife and Marine Resources Depart- ment6). Comparisons of von Bertalanffy back- calculated and observed SL at age are in Table 5. Table 6. — Estimated von Bertalanffy parameters describing the growth of Centropristis striata. The weighted residual sums of squares = 238.46. SE = standard error; C.L. = confidence limits. Param- eter Estimate Asymptotic SE Asymptotic lower 95% C.L upper L 341 0.2309 -0.3010 17.818 0.0221 0.0560 298 0.1787 -0.4335 383 0.2831 -0.1685 Reproduction The generalized ovarian structure of C. striata is similar to that of Epinephelus fulva (Smith 1965), E. morio (Moe 1969), and Hemanthias vivanus (Hastings 1981). The bilobed organ is suspended by mesenteries from the swim bladder in the posterior region of the body cavity. The lobes fuse posteriad, and their lumina form a common oviduct. Blood vessels and nerves enter the ovary at the anterior point of each lobe's suspension and continue pos- teriad medial to the supportive mesenteries along the dorsomedial surface of each lobe. The lumina are lined with folded germinal epithelium (ovarian lamellae), within which oocytes develop and mature. The lamellae are first seen protruding from the dor- sal region of the lumen at the boundary of the ovary and the alamellar oviduct. They continue along both sides of the lumen in the area of gonadal confluence until only the ventralmost region of the ovary is alamellar. This alamellar area is confluent with the oviduct and extends anteriad to half of the lengths of each ovarian lobe (Fig. 6 A). The alamellar regions of female gonads were bordered throughout their extent by testicular precursor cells (Figs. 6A, B; 7A). Although these bands of cells were found in vary- ing stages of development in all ovarian tissues, the most active proliferation of identifiable sperma- togenic tissue (as manifested by transitional gonads; Table 7) occurred after the spring and fall spawn- ing seasons (described later). Both increased ovarian inactivity and degeneration coincided with the pro- liferation of testicular tissue during sexual succes- sion (simultaneous hermaphroditic development is treated below). No instance of active ovarian devel- opment concurrent with testicular degeneration was observed. Sexual transition commenced in the posterior region of the ovary with the expansion of testicular lobes into the ovarian lumen. This proliferation pro- ceeds anteriad, with sperm sinus forming in the ovarian tunic adjacent to the testes. Testicular growth appears to be the result of mitotic sperma- togonial development, although limited spermato- genic processes, including spermatozoa formation, are not uncommon (Fig. 7B). The sperm sinuses, as well as the vas deferens (which form within the ovi- ductal wall) apparently result from ruptures in their respective surrounding tissues, as suggested by Hastings (1981), because there was no cell lining associated with these structures (Fig. 7B). Simultaneously developed hermaphroditic gonads were found in all maturity stages. However, only 3% of the fishes exhibited this phenomenon, and we were unable to determine if the vas deferens had an external opening; therefore, we lack definitive proof that these fish were functional simultaneous hermaphrodites. Histological sections of immature ovaries con- tained only oogonia and small basophilic, previtel- logenic oocytes about 8-100 /^m in diameter. Matur- ing ovaries had oocytes 100-500 ^m in diameter, in Table 7.— Monthly sex composition data for Centropristis striata along with x2 values for tests of a 1 :1 sex ratio. * * = P < 0.01 , 1 df; * = P< 0.05, 1 df. 6Office of Conservation, Management, Marketing and Recrea- tional Fisheries, S.C. Wildlife and Marine Resources Department. 1982. South Carolina Saltwater Sport Fishing Tournaments and State Record Fish. S.C. Wildlife and Marine Resources Depart- ment, P.O. Box 12559, Charleston, SC 29412. Transi- Transi- tional Month Males Females tional (%) o-:9 x2 January 13 13 1 3.7 1 1 — February 111 107 8 3.6 1 0.96 0.07 March 15 7 0 0 1 0.47 2.91 April 928 1,685 122 4.5 1 1.82 219.30** May 404 497 145 13.9 1 1.23 9.60** June 509 1,039 383 19.8 1 2.04 181.46** July 132 368 84 14.4 1 2.79 111.39** August 112 246 42 10.5 1 2.20 50.16** September 668 1,109 262 12.8 1 1.66 109.44** October 35 17 1 1.9 1 2.05 5.12* November 64 150 27 11.2 1 2.34 34.56** December 34 19 13 19.7 1 0.56 4.45 731 FISHERY BULLETIN: VOL. 84, NO. 3 Figure 6.— Schematic representation of a functionally female ovary from Centropristis striata. A) Ventral view of ovary showing the alamellar region. Cross sections of the ovary were made in planes I-I and II-II, and show the positioning of the primordial testicular tissue at the boundary of the alamellar regions. B) An enlargement of the area indicated, showing the cellular relationships of the alamellar area, testicular primordia and ovarian lamellae. AL = alamellar region, 0 = oocytes, OL = ovarian lamellae, OT = ovarian tunic, SP = chords of spermatogonia, TP = testicular primordia. Figure 7.— Photomicrographs of histological sections of gonads from Centropristis striata. A) Cross section taken from the posterior region of a functional ovary showing the alamellar region and testicular primordia, 250 x magnification. B) Cross section taken from the posterior region of a gonad undergoing transition, 250 x magnification. C) Cross section of immature testicular tissue from a 133 mm SL fish, 100 x magnification. D) Cross section of a simultaneous gonad showing active testicular and ovarian development, 100 x magnification. AL = alamellar region, DO = developing oocyte, OL = ovarian lamellae, OT = ovarian tunic, S = spermatozoa, SP = spermatogonia, T = testicular tissue. 732 WENNER ET AL.: LIFE HISTORY OF BLACK SEA BASS AL / N>t' SP A fc,. * - '^ ■ ;* ■ 339 55 0 2 3.6 Total 2,105 3,595 1:1.70 Table 9.— Sex composition and x2 values for tests of a 1:1 sex ratio of Centropristis striata by age. * * = P < 0.01 , 1 df . Transi- Transi- tional Age Males Females tional (0/0) ct:9 x2 0 10 55 0 0 1:5.50 31.2** 1 42 315 20 5.3 1 7.50 208.8** 2 251 1,066 185 12.3 1 4.20 504.3** 3 561 1,449 447 18.2 1 2.60 392.3** 4 635 479 223 16.7 1 0.75 21.8** 5 274 84 59 5.0 1 0.31 100.8** 6 189 17 8 3.7 1 0.09 143.6** 7 54 2 2 3.4 1 0.04 48.3** 8 13 0 0 0 9 2 0 0 0 10 2 0 0 0 734 WENNER ET AL.: LIFE HISTORY OF BLACK SEA BASS . "V.. \ to < 20- \ O X o < < LjJ 100- UJ ° o tr UJ 0- 60- \ »-*-> — i — i 1 1 i — 100 150 200 250 300 SIZE CLASS(cm) • — — • f»mol«8 x x transitional r yf * \ \ '■V S 60 ^ -I o 40 X o < UJ X V) < z loo o h- 80 | cr — i 1 1 1 1 1 1 — 12 3 4 5 6 7 AGE Figure 8.— Percent female and transitional Centro- pristis striata by size and age. Fall spawning probably extended into October because many fishes obtained in November had recently spent ovaries. Fecundity was significantly related to SL, TL, weight, and age with the former three equations having by far the highest r2 values (Table 10). The Female • = Developing + Ripe Maturity ■ = Spent + Resting Stage A M J J MONTHS Figure 9.— Maturity stages of female Centropristis striata by month to illustrate bimodal spawning. least squares linear regression model of fecundity on age explained only 33% of the variation in fecun- dity. Observed mean fecundity and its standard error increased with age (Table 11). The lowest observed fecundity (17,000) was in a 2-yr-old fish (SL = 108 mm; TL = 140 mm; weight = 45 g) and the largest (1,050,000) was in a 438 mm SL fish (TL = 454; weight = 1,371 g) of undetermined age. Mortality Instantaneous rates of total mortality, as derived from catch curves, for C. striata ranged from 0.721 Table 10. — Least squares linear and geometric mean functional regression equations of fecundity (fee) on total length (TL), standard length (SL), weight (WT), and age for Centropristis striata. Weight units are grams and lengths are millimeters. All least squares regressions were significant at o = 0.01. Least squares equation GM functional equation log10 fee = -0.605 + 2.335 (log10 TL) log 10 fee = -0.309 + 2.318 (log10 SL) log10 fee = 3.057 + 0.822 (log10 WT) log 10 fee = 4.529 + 0.913 (log10 Age) 115 0.62 log10 fee = -2.098 + 2.959 (log10 TL) 115 0.65 log10 fee = -1.589 + 2.879 (log10 SL) 115 0.65 log10 fee = 2.587 + 1.022 (log10 WT) 110 0.33 log10 fee = 4.196 + 1.580 (log10 Age) Table 1 1 . — Observed mean fecundity at age and its standard error (S^) for Centropristis striata, in the South Atlantic Bight. Age Mean fecundity Sx n 2 61,846 8,089 13 3 94,801 4,406 55 4 115,411 8,900 27 5 160,000 50,720 7 6 226,040 46,706 5 7 287,350 80,650 2 8 137,400 — 1 to 1.430, and actual mortality rates were from 0.513 to 0.761. Values increased from 1978 to 1981. For example, values of A rose from 51.3 to 73.3% for trap-caught fish and from 51.6 to 76.1% for hook- and-line caught fish older than age 4 (Table 12). Mor- tality values of trap-caught and hook-and-line caught C. striata were similar within years for each estima- tion procedure. Mortality values, moreover were similar between estimation procedures. We found a significant correlation between the instantaneous 735 FISHERY BULLETIN: VOL. 84, NO. 3 Table 12. — Instantaneous (Z) and actual (>A) rates of total mortality for Centropristis striata in the South Atlantic Bight. Gear types: T = trap; H&L = hook and line. Gear Catch Z curve A Heinke Z A Chapmar Z -Robson A Means Year Z A 1978 T 0.721 0.513 0.841 0.568 0.991 0.628 0.851 0.569 1979 T 0.906 0.595 0.819 0.559 0.872 0.582 0.866 0.579 1979 H&L 0.726 0.516 0.759 0.532 0.650 0.478 0.712 0.509 1980 T 1.030 0.643 1.020 0.639 1.181 0.693 1.077 0.658 1980 H&L 0.905 0.595 0.944 0.611 1.016 0.638 0.955 0.615 1981 T 1.320 0.733 1.347 0.740 1.328 0.735 1.332 0.736 1981 H&L 1.430 0.761 1.347 0.740 1.492 0.775 1.423 0.759 1982 T 1.279 0.722 1.382 0.749 1.443 0.764 1.368 0.745 1982 H&L 1.246 0.712 1.277 0.721 1.309 0.730 1.277 0.721 rate of total mortality from trap data and the South Carolina commercial landings from 1978 to 1982 (Fig. 10). Population Estimates at Specific Sites Mortality of C. striata attributable to tagging occurred only once, during the 1983 experiment of site 2 when 6% of the fishes (3 of 50) died during the holding period. Therefore, we reduced the number of tagged fish-at-large (M) by 6% to account for this tagging related mortality. Between 1981 and 1983, a decline in the order of magnitude from 20,070 to 2,236 individuals (88.9%) occurred in the estimated abundance of C. striata at site 1. On this reef, the abundance declined 60.6% from 1981 to 1982 and another 75.5% from 1982 to 1983 (Table 13). Abundance at site 2 declined 52.9% between 1982 and 1983. Biomass of C. striata declined by an order of magnitude on site 1 from 4,836 kg in 1981 to 491 kg in 1983 (Table 13). This was an overall decrease of 89.9%. Site 2 had a 62% decline in biomass from 2,150 kg in 1982 to 810 kg in 1983. Our estimates are for fish >20 cm TL, the only ones vulnerable to the traps. Therefore, density and biomass estimates are minimum val- ues. In addition to the declines in population size and biomass of C. striata, there were decreases in mean size and age. Mean TL was 3 cm less in 1983 than in 1981 at site 1, whereas C. striata were on average 2 cm smaller in 1983 than 1982 at site 2. Not only were the means reduced, but also the frequency distribution became more skewed towards the smaller size intervals and the contributions of larger Figure 10.— Plot of the instantaneous rate of total mortality (Z) as determined from resource survey data (1978-81) and the South Carolina commercial landings of Centropristis striata for that year. 736 Cfl 1 3 >> O o o CO OS .5- r-0.970 100 200 300 South Carolina Commercial Landings(metric tons) WENNER ET AL.: LIFE HISTORY OF BLACK SEA BASS Table 13.— Summary of Petersen mark-recapture population estimates, biomass, and density (number and kg/ha) estimates for black sea bass, Centropristis striata, on two sponge-coral habitat sites. 95% confidence limits (C.L.) of p(= RIC) were determined by the methods of Cochran (1977). Site 1 Site 2 1981 1982 1983 1982 1983 c 634 529 438 446 679 M 1,042 1,169 1,084 901 854 R 32 67 212 50 155 95% C.L of R 21.9-44.2 53.4-83.1 193.2-230.8 33.9-57.5 135.1-175.2 P 0.50 0.127 0.484 0.112 0.228 95% C.L. of p 0.035-0.070 0.101-0.157 0.441-0.527 0.076-0.129 0.199-0.258 AT1 20,070 9,119 2,236 7,906 3,727 95% C.L. of AT 14,653-28,921 7,347-1 1 ,399 2,055-2,453 6,892-11,553 3,300-4,272 M 0.032 0.058 0.196 0.056 0.182 95% C.L of h 0.022-0.043 0.046-0.072 0.179-0.213 0.039-0.065 0.159-0.206 Biomass (kg) 4,836 2,077 491 2,150 810 95% of biomass (kg) 3,530-6,959 1 ,673-2,595 451-539 1,874-3,142 717-928 Number/ha 125 57 14 66 31 95% C.L. of number/ha 92-181 46-71 13-15 57-96 28-36 kg/ha 30.2 13.0 3.1 17.9 6.7 95% C.L. of kg/ha 22.1-43.5 10.5-16.2 2.8-3.4 15.6-26.2 6.0-7.7 'Adjusted Petersen estimate (Ricker 1975). fishes to the populations was greatly reduced (Fig. (Fig. 12). Fishes age 4 and older went from 42% of 11). Mean age declined 0.5 years at site 1, and the the population in 1981 to 25% in 1982 and 9% in age composition shifted towards younger age classes 1983. 20 15- 10- 5- 20-i JU- 18- Sit* 2. 1S82 5 TL"26cm 10- 5- ' "l i i i i i i i i Site 2, 1883 x TL-=24cm 1 I I I I I I I I I I I I i I I i I i r 20 25 30 35 40 TOTAL LENGTH(cm) Figure 11.— Size-frequency distribution of Centropristis striata from five discrete mark- recapture tagging studies. 737 FISHERY BULLETIN: VOL. 84, NO. 3 Figure 12.— Age composition of Centropristis striata at the two experimental mark-recapture sites. 50 30 x Age 3.4 Site 1 1981 n Site 1 1982 50-, 30 10 x Age3.3 Site 2 1982 a >i _ 50 30 10 x Age3.1 Site 2 1983 n 3 AGE DISCUSSION Age and Growth The cause of the variation in size and shape of the sagitta's central field in C. striata is unknown, how- ever, differing size of the nuclei of Atlantic herring, Clupea harengus, can be related to spawning season (Postuma 1974). Further studies are needed to determine if these differences can be related to spawning time of C. striata in the South Atlantic Bight. Inadequate validation in many studies that esti- mate age have been noted by Beamish and McFar- lane (1983), and they have reemphasized the need for verification of aging technique. Our attempts at validation have shown that one annulus is formed each year during April-May. Also, our counts of presumed daily growth rings have provided circum- stantial support for the formation of the first an- nulus. A similar approach was used by Radtke et al. (1985) in their study of the oyster toadfish, Op- sanus tau. Our mean back-calculated lengths agree well with Mercer's (1978) data for C. striata from the South Atlantic Bight up to age 5; however, ours are much smaller than Cupka et al. (fn. 5). Our lengths at age are consistently smaller than C. striata from the Middle Atlantic Bight (Mercer 1978). Mercer (1978) attributed size at age differences between the two areas to gear selectivity, yet our results suggest that C. striata from the South Atlantic Bight are smaller than those of the Middle Atlantic Bight. The larger size at age found by Cupka et al. (fn. 5) may reflect the population of C. striata in the South Atlantic Bight prior to heavy exploitation that began in 1969. Since estimates of Lm, K, and t0 are affected by several nonbiological, methodical factors, direct comparisons of these growth parameters between different studies are of limited value. However, when viewed in relative terms, they can indicate general differences or similarities between studies, species, or areas. Our estimate of L^ (341 mm SL) was much closer to Mercer's (1978) value (L^ = 352 mm SL) than that of Cupka et al. (fn. 5) (°°625 mm SL). Our growth coefficient (K) was higher, in- dicating that C. striata achieves maximum attain- able size more rapidly than previously reported. These differences could have been caused by sam- pling methodologies and/or conditions of the popula- 738 WENNER ET AL.: LIFE HISTORY OF BLACK SEA BASS tion of South Atlantic Bight C. striata at the time the studies were conducted. Reproduction Smith (1965) established a phylogeny of serranid fishes based on three types of hermaphroditism. Most primitive is the Serranus-type gonad found in Serranus and Hypoplectus, genera which are simul- taneously hermaphroditic with male and female ger- minal tissues well separated by connective tissues. The middle type of this trio is the protogynously her- maphroditic Rypticus-Anthias-type gonad where testicular takeover commences with proliferation of preexisting spermatogonia located in crypts along the alamellar regions of the ovary and gametogenic tissues remain separated by connective tissue throughout sexual transition. Most advanced is the protogynous hermaphroditic Epinephelus-type gonad where testicular tissue cannot be found before sexual transition commences. During this process, crypts of spermatogonia differentiate and prolifer- ate within the ovarian lamellae where they are inter- mixed with oogonia and oocytes. Citing Lavenda (1949), Smith (1965) classified C. striata within the Epinephelus-type, an error cor- rected by Mercer's (1978) demonstration that mor- phological events during sexual transition in C. striata most resemble those of the Rypticus- Anthias-type gonad. Sexual succession in C. striata results from hypertrophy of bands of testicular primordia that lie along borders of the alamellar region of the ovary, not the proliferation of crypts of tissue that Mercer (1978; see also Smith 1965) reported. The arrangement of the primordial tes- ticular ridges in C. striata is the same as in the pro- togynous Hemanthias vivanus (Hastings 1981). The testicular primordia in C. striata is located in a similar region of the gonad as is the testicular portion of the simultaneously functioning gonad of Serranus tigrinus (Smith 1965). Though not stated by Smith (1965), the testes of S. tigrinus might border the alamellar region of the ovarian section as does the testicular primordial cells in C. striata, a gonadal similarity also noted between H. vivanus and S. tigrinus (Hastings 1981). No phylogenetic in- ferences should be drawn from these data, because gonadal development varies even among the close- ly related simultaneous hermaphrodites of the genera Serranus and Diplectrum. Centropristis striata, H. vivanus, and probably R. maculatus (see Smith 1965) have similar gonadal morphologies and strategies of sex succession, but these species are usually not considered closely related. Gonadal mor- phologies may one day be important in determin- ing serranid phylogenetic relationships, but more observations of all serranids are necessary. The simultaneously functioning gonad of C. striata has morphology similar to that of Serranus (Smith 1965) in which discrete areas of testicular tissue empty into peripherally located sinuses, and oocytes discharge centrally. Sperm sinuses within the wall of the simultaneous gonads are well developed in C. striata, but it is not known if they are functional, i.e., permit sperm to exit the body along with the oocytes. We found sizes and ages of C. striata undergoing sex succession which were similar to those Mercer (1978) reported in the South Atlantic Bight; how- ever, we found a much higher incidence of transi- tional fish. Since Mercer (1978) found only 4% of C. striata from this area were undergoing sex suc- cession, she offered two mechanisms for her abun- dance (38%) of mature males: 1) development of mature males from both immature males and juvenile hermaphrodites was very important, or 2) the rate of sexual transition was very rapid in this species. We feel that both of Mercer's arguments were at best incomplete because of her small sample sizes from the South Atlantic Bight. Since we found few immature males and juvenile hermaphrodites in our samples, the probability is low that mature males develop solely from these. Also, we acknowledge the presence of serranids which show rapid sex succes- sion (Fishelson 1970; Fricke and Fricke 1977) and believe the low frequency of individuals undergoing sex succession seen in most Epinepheline groupers probably reflects a similarly short-lived process. However, the presence of C. striata undergoing sex succession throughout the year, and their occur- rence at sizes where the frequency of females declines, leads us to conclude that the primary source of mature males is through sex succession from active females. We found secondary testes (sensu Harrington 1971) in all male C. striata including immature specimens. This morphology is not unique to C. striata. Hastings (1981) observed no primary male H. vivanus and suggested they all passed through an initial female phase. This same secondary gonadal morphology occurs in the secondarily gonochoristic serranid Paralabrax clathratus (Smith and Young 1966), and Reinboth (1970) indicated all male ser- ranids are derived from females. Overall, sex ratios of C. striata were significantly different from an hypothesized la:l9 in favor of females. Females significantly outnumbered males 739 FISHERY BULLETIN: VOL. 84, NO. 3 up to an intermediate size and age, at which time the significantly different ratios favored males. Fishelson (1975) stated that sex ratios should ap- proximate la: 19 at some stage if all protogynous females undergo sex succession. Given the alter- nating ratios of sexual abundance with size and age, and considering that no female older than age 7 and few larger than 330 mm SL were found in our samples, leads us to conclude that all C. striata have the potential to undergo sex succession. Population Estimates The underlying assumptions of the Petersen method for population estimates were met in this study. We found tag-related mortality in only one experiment and adjusted the number of fish marked for it. We feel all tags were accounted for and tag loss was minimal, because tags were firmly anchored to the fish and were bright orange. Tagged fish were not randomly distributed over the study site, but they were released during vessel drifts governed by wind and surface currents and may be effectively random. We assumed minimal immigration and emigration because our experiments covered a brief time period. Powles and Barans (1980) estimated density of C. striata in the sponge-coral habitat of the South Atlantic Bight. The estimates of 51 fish/ha and 7.6 kg/ha derived from the data of Powles and Barans (1980) were 37-66% and 23-44% of our mark-recap- ture values. Powles and Barans (1980) indicated that possible sources of error in their study were distance determinations from loran-A, which are much less precise than distances derived from loran-C read- ings, and variable visibility. ACKNOWLEDGMENTS This work was funded by the National Marine Fisheries Service under contract NA-84-WCC-06101 to the South Carolina Wildlife and Marine Resources Department. We appreciate the Assistance of A. J. Kemmerer and W. Nelson of NMFS. We thank Captain John Causby and First Mate Julian Mikell of the RV Oregon for the exceptional navigational and vessel handling skills enabling us to sample open ocean patch reefs not much larger than a few football fields with a 90-ft vessel. The difficulties involved can be appreciated only by one who has been there. It would not have been possi- ble to process the numerous histological samples without the help of D. Stubbs. We are grateful to the many individuals who participated in the field effort, several of whom suffered punctures by fish spines and lacerations by sea bass preopercles dur- ing the tagging study. A. G. Gash provided assis- tance with the computer analysis, K. Swanson drew the figures and N. Beaumont and M. Lentz typed the manuscript. Helpful critical reviews of the manuscript were made by C. A. Barans, E. L. Wen- ner, R. Warner, P. Hastings, G. Huntsman, P. Eldridge, and two anonymous reviewers. LITERATURE CITED Bagenal, T. B. 1978. Aspects of fish fecundity. In S.D. Gerking (editor), Ecology of freshwater fish production, p. 75-101. John Wiley & Sons, Inc., N.Y. Beamish, R. J., and G. A. McFarlane. 1983. The forgotten requirement for age validation in fish- eries biology. Trans. Am. Fish. Soc. 112:735-743. Bertalanffy, L. von. 1938. A quantitative theory of organic growth. II. Inquiries on growth laws. Hum. Biol. 10:181-213. Carlander, K. D. 1982. Standard intercepts for calculating lengths from scale measurements for some centrarchid and percid fishes. Trans. Am. Fish. Soc. 111:332-336. Cochran, W. C. 1977. Sampling techniques. 31st ed. John Wiley & Sons, Inc., N.Y., 428 p. Combs, R. M. 1969. Embryogenesis, histology and organology of the ovary of Brevoortia patronus. Gulf. Res. Rep. 2:333-436. Everhart, W. H., and W. D. Youngs. 1981. Principles of fishery science. Cornell Univ. Press, Ithaca, N.Y., 349 p. Fischer, W. (editor). 1978. FAO species identification sheets for fishery purposes. Western Central Atlantic. FAO., Rome, Vol. 4, pag. var. Fishelson, L. 1970. Protogynous sex reversal in the fish Anthias squami- pianis (Peters), (Teleostei:Anthiidae). In R. Reinboth (editor), Intersexuality in the animal kingdom, p. 284-294. Spring- Verlag, N.Y. 1975. Ecology and physiology of sex reversal in Anthias squamipianis (Peters), (TeleosteiAnthiidae). In R. Rein- both (editor), Intersexuality in the animal kingdom, p. 284-294. Spring- Verlag, N.Y. Fricke, H., and S. Fricke. 1977. Monogamy and sex change by aggressive dominance in coral reef fish. Nature (Lond.) 266:830-832. Frame, D. W., and S. A. Pearce. 1973. A survey of the sea bass fishery. Mar. Fish. Rev. 35 (l-2):19-26. Harrington, R. S., Jr. 1971. How ecological and genetic factors interact to deter- mine when self-fertilizing hermaphrodites of Rivulus mar- moratus change into functional secondary males, with a reappraisal of the modes of intersexuality among fishes. Copeia 1971:389-432. Hastings, P. A. 1981. Gonad morphology and sex succession in the protogy- nous hermaphrodite Hemanthias vivanus (Jordan and Swain). J. Fish. Biol. 18:443-454. 740 WENNER ET AL.: LIFE HISTORY OF BLACK SEA BASS Helwig, J. T., and K. A. Council (editors). 1979. SAS users guide. 1979 ed. SAS Instit., Inc., N.C., 294 p. Hilge, V. 1977. On the determination of the stages of gonad ripeness in female bony fishes. Meeresforchung 25:149-155. Humason, G. L. 1972. Animal tissue techniques. W. H. Freeman and Co., San Francisco, CA, 641 p. Huntsman, G. R. 1976. Offshore headboat fishing in North Carolina and South Carolina. Mar. Fish. Rev. 38(3): 13-23. Hyder, M. 1969. Histological studies on the testes of Tilapia leucosticta and other species of the genus Tilapia (Pisces:Teleostei). Trans. Am. Microsc. Soc. 88:211-231. Lavenda, N. 1949. Sexual differences and normal protogynous hermaphro- ditism in the Atlantic sea bass Centropristis striatus. Copeia 1949:185-194. Low, R. A., Jr. 1981. Mortality rates and management strategies for black sea bass off the southeast coast of the United States. No. Am. J. Fish. Manage. 1:93-103. Mercer, L. P. 1978. The reproductive biology and population dynamics of black sea bass, Centropristis striata. Ph.D. Thesis, College of William and Mary, Williamsburg, VA, 196 p. Moe, M. H., Jr. 1969. Biology of the red grouper Epinephelus morio (Valen- ciennes) from the eastern Gulf of Mexico. Fla. Dep. Nat. Resour. Mar. Res. Lab., Prof. Pap. 10, 95 p. Musick, J. A., and L. P. Mercer. 1977. Seasonal distribution of black sea bass, Centropristis striata, in the Mid-Atlantic Bight with comments on the ecology and fisheries of the species. Trans. Am. Fish. Soc. 106:12-25. Poole, J. C. 1961. Age and growth of the fluke in Great South Bay and their significance in the sport fishery. N.Y. Fish Game J. 8:1-11. Postuma, K. H. 1974. The nucleus of the herring otolith as a racial character. J. Cons. Int. Explor. Mer 35:121-129. Powles, H., and C. A. Barans. 1980. Groundfish monitoring in sponge-coral areas off the southeastern United States. Mar. Fish. Rev. 42(5):21- 35. Radtke, R. L., M. L. Fine, and J. Bell. 1985. Somatic and otolith growth in the oyster toadfish (Op- sanus tau L.). J. Exp. Mar. Biol. Ecol. 90:259-275. Reinboth, R. 1970. Intersexuality in fishes. Mem. Soc. Endocr. 18:515- 544. RlCKER, W. E. 1973. Linear regressions in fishery research. J. Fish. Res. Board Can. 30:409-434. 1975. Computation and interpretation of biological statistics of fish populations. Fish. Res. Board Can. Bull. 191, 382 p. Rivers, J. B. 1966. Gear and technique of the sea bass trap fishery in the Carolinas. Commer. Fish. Rev. 28(4):15-20. Smith, C. L. 1965. The patterns of sexuality and the classification of ser- ranid fishes. Am. Mus. Novit. 2207, 20 p. Smith, C. L., and P. H. Young. 1966. Gonad structure and the reproductive cycle of the kelp bass, Paralabrax clathratus (Girard) with comments on the relationships of the serranid genus Paralabrax. Calif. Fish Game 52:283-292. SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry. 2d ed. W. H. Freeman & Co., N.Y., 859 P- Struhsaker, P. 1969. Demersal fish resources: composition, distribution, and commercial potential of the continental shelf stocks off south- eastern United States. U.S. Fish. Wildl. Serv., Fish. Ind. Res. 4:261-300. Wallace, R. A., and K. Selman. 1981. Cellular and dynamic aspects of oocyte growth in teleosts. Am. Zool. 21:325-343. Wenner, C. A. 1983. Species associations and day-night variability of trawl caught fishes from the inshore sponge-coral habitat, South Atlantic Bight. Fish. Bull, U.S. 81:537-552. 741 NOTES COMPARISON OF VISCERAL FAT AND GONADAL FAT VOLUMES OF YELLOWTAIL ROCKFISH, SEBASTES FLAVIDUS, DURING A NORMAL YEAR AND A YEAR OF EL NINO CONDITIONS One of the severest El Nino events of the century occurred off California during late 1982 and most of 1983 (Rasmusson 1984). Associated with the warm water and lack of upwelling were impressions by many fishermen and biologists that macroplank- tonic organisms were at low densities and that fish were thinner than normal. A semiquantitative sam- pling program off of San Francisco indicated that euphausiids, a major component of the macroplank- ton, were considerably less common in 1983 than in either 1982 or 1984 (Smith1). Yellowtail rockfish are abundant off northern California and are an important component of recreational and commercial catches in some areas. The species feeds mostly on macroplanktonic organisms such as euphausiids, salps, and small fish (Phillips 1964; Pereyra et al. 1969; Lorz et al. 1983). Annual cycles of visceral fat volume and gonad volume are documented in Guillemot (1982) and Guillemot et al. (1985). The studies showed that visceral fat volume in both sexes of yellowtail rock- fish is at a maximum during fall. The viviparous species (Boehlert and Yoklavich 1984) mates in early fall (September) and releases larvae during winter (January-March) (Wyllie Echeverria2). Guillemot (1982) and Guillemot et al. (1985) showed that male gonad volumes peak in fall and female gonad volumes peak in winter. The purpose of this study is to determine possi- ble effects of El Nino conditions by comparing visceral fat and gonad volumes during 1983, a year of El Nino conditions, with data collected during 1980, a normal year (Guillemot 1982). Methods and Materials Guillemot (1982) and Guillemot et al. (1985) util- ized data collected throughout the year. The 1983 data were collected only on 21 September, the ap- proximate sexual activity peak for males, and 20 December, which slightly precedes the peak time of larval release for females. Only 1980 data collected within 20 d of the two 1983 collection dates and samples collected from central California, between Bodega Bay and Half Moon Bay, were used in this study. In 1983 all specimens were collected from landings made at Bodega Bay. Specimens were sexed, measured to the nearest millimeter for total length, and viscera were re- moved and preserved in 10% buffered Formalin3 in the field following the procedures of Guillemot et al. (1985). After about 90 d of storage, visceral fat and gonad volumes were measured to the nearest milliliter by water displacement. Visceral fat of some fish had dissolved to form a floating liquid. The volume of this liquid was measured and added to the total fat volume. Data from males larger than 379 mm, when 90% are mature, and from females larger than 380 mm, when 85% are mature, were used (Wyllie Echeverria fn. 2). As in Guillemot (1982) and Guillemot et al. (1985) we used the following power equation to describe the relationship between fat or gonad volume and length: Y = aXP where Y = fat or gonad volume, and X = total length. The parameters were estimated by first trans- forming the variables to natural logarithms and then using standard least squares linear regression techniques. Analysis of covariance was used to determine if separate lines for the two years significantly reduced the variance from a common line (Kleinbaum and Kupper 1978). This is a fairly robust test in that if there is not a significant linear relationship between the two variables for one or both time periods, the test is nearly as powerful for comparing the two means as an analysis of variance. 'Smith, S. Unpublished data. Tiburon Laboratory, Southwest Fisheries Center, National Marine Fisheries Service, NOAA, 3150 Paradise Drive, Tiburon, CA 94920. 2Wyllie Echeverria, T. 1983. Reproductive seasonality and maturity of the rockfishes (Scorpaenidae; Sebastes) off central California. Unpubl. manuscr., 66 p. Southwest Fisheries Center, Tiburon Laboratory, National Marine Fisheries Service, NOAA, Tiburon, CA 94920. 3Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 743 Results The regression lines for the male fat volume for the two years intersect and are not significantly dif- ferent (Table 1). The results of the analysis of covariance for fat volume of females are highly significant (Table 1). Females had significantly higher fat volumes in 1980 for both months (Fig. 1). The comparisons of gonad volumes produced highly significant results in December for both sexes, and for females in September (Table 2). Female gonad volumes were higher in 1980 during December and lines intersected in September (Fig. 2). Male gonad volumes were significantly higher in December 1983 than in December 1980. The seasonality of gonad development was similar in the two years, but appeared to be delayed in 1983. Table 1.— Results of analysis of covariance of fat volumes of yellowtail rockfish regressed on length. Observations were transformed to natural logarithms for the analysis. Month 1980 1983 Sex Sample size Intercept Slope Sample size Intercept Slope F Male Male Female Female September December September December 20 17 25 19 -9.884 7.198 -9.327 21.813 2.011 -0.978 2.003 -3.262 38 35 46 50 22.272 35.402 5.443 1.609 -3.355 -5.693 -0.449 - 0.058 2.753 2.571 11.917** 5.889** * "Significant at 99% level of confidence. 3.5 3.0 E 2.5 i 0 E •2 2.0 o > 5 1.5 1.0 0.5 a a Sep. 1980 a- -* Sep. 1983 • • Dec. 1980 •---• Dec. 1983 _L _L 5.9 6.0 6.1 6.2 6.3 In (total length-mm) Figure 1.— Relationships between In (visceral fat volume) and In (total length) for female yellowtail rockfish in 1980 and 1983. Males were 50% maturing and 50% resting in September 1980, and 100% resting during Decem- ber. In 1983 males were 100% maturing during September, and 8% maturing and 92% resting dur- ing December. Females were 35% maturing and 65% resting in September 1980, and 83% maturing and 17% resting in December. In 1983 females were 100% maturing during September, and 97% matur- ing and 2% resting during December. Data on season or parturition for 1981-84 (Table 3) indicate that parturition was delayed in 1983 and 1984 com- pared with 1981 and 1982. Discussion The results tend to agree with expectations. Female fat volumes were lower in 1983 than in 1980, which is in agreement with the impressions of fisher- men and the expectation that El Nino would pro- duce relatively poor feeding conditions and conse- quently result in thin fish. Table 2.— Results of analysis of covariance of gonad volumes of yellowtail rockfish regressed on length. Observations were transformed to natural logarithms for the analysis. Month 1980 1983 Sex Sample size Intercept Slope Sample size Intercept Slope F Male Male Female Female September December September December 20 17 25 19 -35.589 -50.003 - 54.09 - 56.674 6.161 8.290 9.171 9.855 38 35 46 50 -25.205 - 1 1 .037 -31.019 -45.723 4.450 1.962 5.406 7.908 0.274 8.170** 3.404* 12.224** '•Significant at 99% level of confidence. 'Significant at 95% level of confidence. 744 co <1> E o > <0 C o 2.5 2.0 1.5 -^ Sep. 1980 *---* Sep. 1983 • • Dec. 1980 • • Dec. 1983 6.0 6.1 In (total length-mm) 6.2 1980 were not expected. December is later than the normal period of sexual activity for males, but the unexpected gonad volume results may be caused by delayed mating. The gonad stage data indicated that male sexual activity was later in 1983 than in 1980. While fish condition and reproduction were dif- ferent in 1983 than in the preceding non-El Nino years, the documentation of such differences for marine fish is uncommon. The season of parturition of yellowtail rockfish is more variable than we realized when the study was designed and the data on fish condition and gonad volume should have been collected over a wider period of time. The results of our study indicate that the assumption of constant adult fish condition and reproductive effort that is usually made in models of the population dynamics of fish is questionable. 5.0 ■= 4.0 ■D CO c o 3.0 2.0 1.0 * * Sep. 1980 *- — ■* Sep. 1983 • «Dec. 1980 • ^Dec. 1983 5.9 6.0 6.1 6.2 6.3 In (total length-mm) Figure 2.— Relationships between In (gonad volume) and In (total length) for yellowtail rockfish in 1980 and 1983. (Top) males; (bottom) females. Table 3.— Percent of yellowtail rockfish females with eyed-larvae observed in samples collected in central and northern California, 1981-1984. Year January February March April May June 1981 5 0 0 0 0 0 1982 0 15 0 0 0 0 1983 0 18 16 6 0 4 1984 3 10 15 0 0 0 The lower ovary volumes in 1983 than 1980 could have been related to either delayed parturition and/or lower reproductive effort. Wootton (1979) described relationships between feeding conditions and fish fecundity. The significantly higher gonad volumes for males in December 1983 compared with Literature Cited BOEHLERT, G. W., AND M. M. YOKLAVICH. 1984. Reproduction, embryonic energetics, and the maternal- fetal relationship in the viviparous genus Sebastes (Pisces: Scorpaenidae). Biol. Bull. (Woods Hole) 167:354-370. Guillemot, P. J. 1982. Seasonal cycles of fat content and gonad volume in species of northern California rockfish (Scorpaenidae; Sebastes). M.A. Thesis, San Francisco State Univ., San Francisco, CA, 167 p. Guillemot, P. J., R. J. Larson, and W. H. Lenarz. 1985. Seasonal cycles of fat and gonad volume in five species of northern California rockfish (Scorpaenidae; Sebastes). Fish. Bull., U.S. 83:299-311. Kleinbaum, D. G., and L. L. Kupper. 1978. Applied regression analysis and other multivariable methods. Duxbury Press, North Scituate, MA, 556 p. Lorz, H. V., W. G. Pearcy, and M. Fraidenburg. 1983. Notes on the feeding habits of the yellowtail rockfish, Sebastes flavidus, off Washington and in Queen Charlotte Sound. Calif. Fish Game 69:33-38. Pereyra, W. T., W. G. Pearcy, and F. E. Carvey, Jr. 1969. Sebastodes flavidus, a shelf rockfish feeding on meso- pelagic fauna, with consideration of the ecological implica- tions. J. Fish. Res. Board Can. 26:2211-2215. Phillips, J. B. 1964. Life history studies on ten species of rockfish (genus Sebastodes). Calif. Dep. Fish Game, Fish Bull. 126, 70 p. Rasmusson, E. M. 1984. El Nino: the ocean/atmospheric connection. Oceanus 27 (2):5-12. Wootton, R. J. 1979. Energy costs of egg production and environmental determinants of fecundity in teleost fishes. Symp. Zool. Soc. Lond. No. 44, p. 133-159. William H. Lenarz Tina Wyllie Echeverria Southwest Fisheries Center Tiburon Laboratory National Marine Fisheries Service, NOAA 3150 Paradise Drive Tiburon, CA 94920 745 DIEL FORAGING ACTIVITY OF AMERICAN EELS, ANGUILLA ROSTRATA (LESUEUR), IN A RHODE ISLAND ESTUARY Although the American eel, Anguilla rostrata (LeSueur), is abundant and commercially exploited along the entire Atlantic coast of North America, its basic biology is not well understood (Tesch 1977; Fahay 1978; Helfman et al. 1984). Foraging activity has not been studied. Helfman et al. (1983) examined daily movement patterns in an estuary and found, as had laboratory studies (Bohun and Winn 1966; Edel 1976; van Veen et al. 1976; Westin and Nyman 1979), that American eel locomotor activity is noc- turnal and suggested that American eel foraging activity is also nocturnal. This study sought to describe the diel foraging patterns of wild estuarine American eels by monitoring capture rates in baited eel traps on a 24-h basis. Eight eel traps were set 10 m apart along a transect in a tidal portion of the Pettaquamscutt River estuary, R.I. The water was turbid (the bot- tom could not be seen at midday in areas <1 m deep) and the salinity ranged from 20 to 30%u, depend- ing on the tide. Cylindrical traps are commercially constructed of 0.64 cm2 wire mesh and are 78 cm long and 20 cm in diameter with two single funnel openings of 5 cm in diameter. The traps were baited with 500-700 g pieces of freshly killed horseshoe crab, Limulus polyphemus, an effective eel bait (Bianchini et al. 1981). Capture rates probably reflected contempor- aneous foraging because the traps were thought to have a high escape rate. A high escape rate was suspected for two reasons: 1) When we changed from checking the traps once every afternoon to once every 3 h, the daily capture rate increased near- ly 50 fold; and 2) when 40 eels were placed into 4 unbaited traps in the river, only 1 eel remained 24 h later. Feeding activity in the traps was evidenced by several factors: an examination of the gut con- tent of 10 captured eels found 6 to contain horse- shoe crab and the rest to be empty, anesthetized animals frequently regurgitated bait, eels were often found burrowing in the bait, and unbaited traps rare- ly caught anything. Starting at 1200 e.d.t., traps were checked and rebaited at 3-h intervals for six 24-h periods evenly spaced over a 15-d span in early September 1982. This design removed any possible tidal influence because the lunar tidal period is 14.8 d. Within 10 min of their capture, eels were released 10 m to one side of the transect' s center point. Traps were rebaited every 6 h or whenever the bait was found to have been consumed (which rarely occurred). Baiting schedules were designed so that every other trap was rebaited at each 3-h check, and all portions of the crabs (heads and tails of both males and females) were equally distributed with respect to time and location. A total of 322 American eels were captured (some were probably recaptures): 178 (55% of the total) were caught just after sunset at 2000 e.d.t., 140 (44%) were caught during the remainder of the night, and 4 (1%) were caught during daylight (Fig. 1). Although daily capture rates were variable and ranged from 113 to 22, all exhibited this pattern. To determine when foraging activity commenced, the traps were checked and rebaited at 30-min inter- vals between 1715 e.d.t. (40 min before sunset) and 2015 e.d.t. for 6 evenings during a 15-d period in early October. Eels were consistently first captured just after sunset, with captures peaking 1 h after sunset and declining thereafter (Fig. 2). Daily cap- ture totals varied considerably but all exhibited this 60- 50- X o I- < 40 O _J < h- 30H I- z w 20 o UJ a. 10- N = 322 11 14 17 20 23 2 TIME OF DAY -1 11 Figure 1.— Percentage of total catch of American eels by time of day for the 24-h experiment. The histograms cover the time between checks; i.e., their right boundaries mark the times when traps were checked. The bold section of the x-axis denotes the period between sunset and sunrise. 746 FISHERY BULLETIN: VOL. 84, NO. 3, 1986. pattern. A total of 588 American eels were captured, 83% more than in the 24-h experiment, possibly reflecting the intensity of foraging activity just after sunset. To characterize the population, eels caught on the third evening were measured. They had an average total length of 30.7 cm (SD = 5.4, n = 121), and 10 of the 121 animals caught had the silvered pigmentation pattern which characterizes maturing individuals (Tesch 1977). These data show that the foraging activity of estuarine American eels in late summer through autumn is nocturnal and peaks sharply at nightfall. Whether the subsequent decline in captures was caused by a decrease in foraging because of satia- tion or by an unrelated decline in locomotor activity cannot be determined. The swimming activity of unfed eels in the laboratory often exhibits a dramatic peak at lights-off (Bohun and Winn 1966; Edel 1976; van Veen et al. 1976). Spring and autumn captures of wild short-finned New Zealand eels, Anguilla australis schmidtii, in baited traps displayed the nocturnal activity pattern described here (Ryan 1984). However, capture patterns in the latter study changed with the season, as did the locomotor pat- terns of the yellow European eel, Anguilla anguilla, studied by Westin and Nyman (1979). Further research is required to understand the relationship between foraging and locomotor activity patterns 40-| X o < 30- O < o 20- UJ O 10- K U N = 588 17=45 18:45 TIME OF DAY 19 = 45 Figure 2.— Percentage of total catch of American eels by time of day for the evening experiment. The histograms cover the time between checks; i.e., right boundaries mark the times when traps were checked. The bold section of the x-axis denotes the period after sunset. and how environmental and physiological factors might influence them. Literature Cited BlANCHINI, M., P. W. SORENSEN, AND H. E. WlNN. 1981. Horseshoe crabs as bait for estuarine American eels, Anguilla rostrata. J. World Maricul. Soc. 12:127-129. Bohun, S., and H. E. Winn. 1966. Locomotor activity of the American eel (Anguilla rostrata). Chesapeake Sci. 7:137-147. Edel, R. K. 1976. Activity rhythms of maturing American eels (Anguilla rostrata). Mar. Biol. 36:283-289. Fahay, M. P. 1978. Biological fisheries data on American eel, Anguilla rostrata (LeSueur). Natl. Mar. Fish. Serv. Sandy Hook Lab., Highlands, NJ, Rep. 17, 87 p. Helfman, G. S., D. L. Stoneburner, E. L. Bozeman, P. A. Christian, and R. Whalen. 1983. Ultrasonic telemetry of American eel movements in a tidal creek. Trans. Am. Fish. Soc. 112:105-110. Helfman, G. S., E. L. Bozeman, and E. B. Brothers. 1984. Size, age and sex of American eels in a Georgia river. Trans. Am. Fish. Soc. 113:132-141. Ryan, P. A. 1984. Diet and seasonal feeding activity of the short-finned eel, Anguilla australis schmidtii, in Lake Ellesmere, Canter- bury, New Zealand. Environ. Biol. Fishes 11:229-234. Tesch, F.-W. 1977. The eel. Chapman and Hall, Ltd., Lond./J. Wiley & Sons, N.Y., 434 p. J. Greenwood, translator. Van Veen, T., H. G. Hartwig, and K. Muller. 1976. Light-dependent motor activity and photonegative behavior in the eel (Anguilla anguilla L.). J. Comp. Physiol. 111:209-219. Westin, L., and L. Nyman. 1979. Activity, orientation, and migration of Baltic eel (An- guilla anguilla L.). Rapp. P. -v. Reun. Cons. int. Explor. Mer 174:115-123. Peter W. Sorensen Graduate School of Oceanography University of Rhode Island Narragansett, RI 02882 Present address: Zoology Department University of Alberta Edmonton, Alberta T6G 2E9, Canada Marco L. Bianchini Graduate School of Oceanography University of Rhode Island Narragansett, RI 02882 Present address: I.P.R.A., Consiglio Nazionale Delle Ricerche Via Nizza 128 00188 Roma, Italy Howard E. Winn Graduate School of Oceanography University of Rhode Island Narragansett, RI 02882 747 FIRST RECORD OF THE LONGFIN MAKO, ISURUS PAUCUS, IN THE GULF OF MEXICO The longfin mako, Isurus paucus, (Guitart-Manday 1966) is a large, pelagic shark that has been reported from the western Indian, central Pacific, eastern North Atlantic, and the western North Atlantic Oceans (Compagno 1984). Guitart-Manday (1975, cited by Dodrill and Gilmore 1979) described the longfin mako as a relatively common catch of pelagic longliners off northwest Cuba. They are usually cap- tured off the continental shelf at depths of 60-120 fathoms and infrequently at 10-50 fathoms. Dodrill and Gilmore (1979) reported the first North Ameri- can continental longfin mako, found beached in the surf at Melbourne Beach, FL. This paper reports the first recorded occurrence of the longfin mako in the Gulf of Mexico. A large female /. paucus was collected 1 April 1985 by longline fisherman, 80 mi south of Panama City, FL Gat. 28°55'N, long. 85°35'W) near the sur- face, over 300 fathoms of water. Standard length (precaudal length) measured 313.0 cm and fork length measured 342.0 cm. Total length could not be measured directly because of the sharks position on the boat deck and was estimated using a ratio of total length to fork length (TL/FL = 1.152) cal- culated from 7 large /. paucus (Harold Pratt1). Using this ratio, total length was estimated to be ca. 390 cm. Although no embryos were present in the ovi- duct, this fish appeared reproductively mature. The oviducts were 3-4 cm in diameter and ovarian eggs measured 2-3 mm in diameter. Gilmore (1983) pro- posed the reproductive strategy of /. paucus to be oviphagous, as remnants of yolk were found in the digestive tract and mouth of an examined embryo. The ventral surface of the snout and gill areas of our shark exhibited a dark grey coloration. Garrick (1967) reported this coloration as an important distinguishing characteristic between /. paucus and the shortfin mako, /. oxyrinchus, which exhibits a creamy white coloration in that area. Gilmore (1983) reported the dusky coloration to be more extensive in larger /. paucus. Pectoral fin length of our shark measured 80.6 cm. Gilmore (1983) compared an adult and embryo /. paucus and found that the pectoral fin length represented a greater percentage of SL in the em- bryo (31% of SL) than in the adult (28% of SL). Our Gulf of Mexico specimen was slightly larger than the specimen reported by Gilmore (1983) (313.0 cm vs 303.5 cm SL), and the pectoral fin represented 26% of SL. Guitart-Manday (1966) examined smaller /. paucus— 195, 203, and 226 cm TL— and found pec- toral fin length as percent total length to be 30.4%, 30.0%, and 29.2%, respectively. For this specimen, pectoral fin length as percent TL was about 21%. It appears that as /. paucus increase in length, the pectoral fins do not increase proportionately, result- ing in reduced pectoral length to total length ratios in larger sharks. This record suggests that the longfin mako at least occurs infrequently in the northern Gulf of Mexico. Three male /. paucus (191, 193, and 220 cm SL) cap- tured 16 April 1985 off the Mississippi River (lat. 27°35'N, long. 89°55'W) further supports this sug- gestion (Stephen Branstetter2). These captures ex- tend the known range of this species well into the northern Gulf of Mexico. Acknowledgments We would like to extend a most sincere thanks to Lew Bullock of the Florida Department of Natural Resources for his help in examining this shark. We are grateful to Stephen Branstetter and Wes Pratt for reviewing the manuscript and providing unpub- lished data. Literature Cited Compagno, L. J. V. 1984. FAO species catalogue. Vol. 4. Sharks of the world. An annotated and illustrated catalogue of sharks species known to date. Part 1. Hexanchiformes to Lamniformes. FAO Fish. Synop. 125, Vol. 4(Pt. 1), 249 p. Dodrill, J. W., and R. G. Gilmore. 1979. First North American continental record of the longfin mako (Isurus paucus Guitart-Manday). Fla. Sci. 42:52-58. Garrick, J. A. F. 1967. Revision of sharks of genus Isurus with description of a new species (Galeoidea, Lamnidae). Proc. U.S. Natl. Mus. 118:663-690. Gilmore, R. G. 1983. Observations on the embryos of the longfin mako, Isurus paucus and the bigeye thresher, Alopias super cili- osus. Copeia 1983:375-382. Guitart-Manday, D. 1966. Nuevo nombre para una especie de tibur6n del genero Isurus (Elasmobranchii:Isuridae) de aguas Cubanas. Poe- yana Ser. A, No. 15, 9 p. 1975. Las pesquerias pelagico-oceanicas de corto radio de ac- tion en la region noroccidental de Cuba. Oceanogr. Inst., 'Harold Pratt, Northeast Fishery Center Narragansett Labora- tory, National Marine Fisheries Service, NOAA, South Ferry Road, Narragansett, RI 02882-1199, pers. commun. June 1985. 2Stephen Branstetter, Department of Wildlife and Fisheries Science, Texas A&M University, College Station, TX 77843-2258, pers. commun. August 1985. 748 FISHERY BULLETIN: VOL. 84, NO. 3, 1986. Acad. Sci., Havana, Cuba. Ser. Oceanologica, p. 1-41. Kristie Killam Glenn Parsons Department of Marine Science University of South Florida at St. Petersburg HO 7th Avenue South St. Petersburg, FL 33701 MOVEMENT OF SEA-RUN SEA LAMPREYS, PETROMYZON MARINUS, DURING THE SPAWNING MIGRATION IN THE CONNECTICUT RIVER1 Adult sea lampreys, Petromyzon marinus, first enter New England rivers in late March and early April (Bigelow and Schroeder 1953). The only infor- mation on river water temperatures during the migration were collected in 1974 from the St. John River, New Brunswick, where Beamish and Potter (1975) captured the first prespawning adults in a fish lift at Mactaquac Dam (river km 140) at 13° C in mid- June and the run peaked at 17°-19°C. Because thousands of sea lampreys are annually passed up- stream of Holyoke Dam (river km 140) on the Con- necticut River, the passage records provide an ideal opportunity to characterize the run relative to tem- perature. River flow was partially or totally con- trolled by the hydroelectric facilities at the dam, so we did not examine the effects of flow on the run. The behavior and rate of movement of landlocked sea lampreys in the Great Lakes was determined using mark and recapture of adults at stream weirs (Applegate 1950; Applegate and Smith 1950; Smith and Elliot 1952; Moore et al. 1974). The only esti- mate of the rate of movement of sea-run sea lam- preys was done by Beamish (1979) who used the energy expended during an upstream movement to estimate the distance traveled and the rate of move- ment of adults in the St. John River. Because this estimate of the rate of movement was not verified by direct observations on fish in the field, we be- lieved that additional study was necessary. We selected radio telemetry to determine the rate of movement and diel behavior of sea lampreys. The Contribution No. 101 of the Massachusetts Cooperative Fishery Research Unit, which is supported by the U.S. Fish and Wildlife Service, Massachusetts Division of Fisheries and Wildlife, Mass- achusetts Division of Marine Fisheries, and the University of Massachusetts. abundance, size, and sex ratio of the Connecticut River population were reported by Stier and Kynard (1986). Methods Radio-tagged sea lampreys were observed in the 46 km stretch of the Connecticut River from Brun- elle's Marina to Cabot Station, a hydroelectric facil- ity located 4.5 km below Turners Falls Dam (Fig. 1). The downstream half of this stretch flows slow- ly, creating a deep channel and shoals; the upstream half flows swiftly with pools and riffles. Major spawning areas are in the upper main-stem near Cabot Station, Russelville Brook, and the Fort, Mill, Sawmill, and Deerfield Rivers (Fig. 1). The number of sea lampreys passed daily by the fish lifts from 1980 to 1983 were counted by per- sonnel of the Massachusetts Cooperative Fishery Research Unit. Daily maximum river temperature was recorded at Holyoke Dam. Sea lampreys were captured in the trap at the fish lifts during May and June 1982, measured for total length, and held for <24 h in a 1,325 L circular tank supplied with river water. We anesthetized fish with MS-222 (1:20,000) and tagged them first with a Floy tag inserted through the posterior dorsal fin, and second with a transmitter placed on the left side of the body along the first dorsal fin. Sex could not be accurately determined visually. Cylindrical radio transmitters were constructed from the design of Knight (1975) and operated at a frequency of 30.05-30.25 MHz. Tags measured 34 x 10 mm, weighed 3.5-4.5 g in air, and transmitted for about 20 d. Each fish was identified by frequency and pulse rate. We located fish to within about 10 m, using receivers equipped with an omnidirectional, 1/8- wave antenna and a directional, tuned-loop antenna. We released two to six^sea lampreys at a time and observed them continuously for >6 h or until dark- ness. Subsequently, sea lampreys were located each day until they reached Cabot Station or entered a tributary. During all surveys, we noted the locations of fish to the nearest river kilometer. Diel movement was monitored for five 24-h periods. Additional fish were released during the day for this study. Results and Discussion The water temperatures, and the year in paren- theses, when sea lampreys first entered the fish lifts were 12.5°C (1980), 10.5°C (1981), 12.5°C (1982), and 15.5°C (1983) (Fig. 2). The lifts sampled the en- FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 749 Turners Falls Dam km 198 Figure 1.— Section of the Connecticut River from river km 140 to 198, showing the loca- tions of the Holyoke and Turners Falls Dams, the release site for radio-tagged sea lampreys at Brunelle's Marina, and the major spawning tributaries between the two dams. tire run each year except in 1981 when sea lampreys were present in the first lifts of the year (the lifts began operating on 29 or 30 April of each year). Dur- ing the peak 7 d, the temperature ranges, and year in parentheses, were 16°-19°C (1980), 17°-19°C (1981), 16°-17°C (1982), and 17°-21°C (1983). Movement into the fish lift ceased at 24°C in 1983 and at 21°-22°C in the other years (Fig. 2). Information on the maximum daily temperature during the migration of landlocked sea lampreys in a large river comes from the Ocqueoc River (Lake Huron drainage) which for some years supported an annual run of 25,000-40,000 (Applegate 1950; Apple- gate and Smith 1950). The temperatures, and date in parentheses, when the first sea lampreys entered a weir near the mouth of the river were 10°C (27 April 1949) and 6°C (11 May 1950); and the run peaked at 14°-17°C (first week of May 1949) and 18° -20°C (third week of May 1950). Most movement at the weir ceased at 21 °C (about 11 July), but dur- ing both years one or two sea lampreys per day con- tinued to enter the weir throughout the summer at 22°-26°C. The temperature regimes in the Ocqueoc and Con- necticut Rivers during the peak and at the end of the principal migration were in general agreement. Runs peaked at 14° -20°C in the Ocqueoc River and 16° -21 °C in the Connecticut River; most of the run ceased at 21 °C in the Ocqueoc River and at 21°-24°C in the Connecticut River. The migrations differed because a few adults in the Ocqueoc River continued to migrate throughout the summer, whereas none were captured after 25 June during 3 yr in the Connecticut River. Therefore, even 750 15 h- 10 z Ld 5 O cr 0 Ld Ql 35 i 30 5 APRIL 15 20 25 30 MAY 10 15 20 25 30 5 10 JUNE JULY Figure 2.— Daily percent of total sea lampreys lifted at the Holyoke fish lifts each year, 1980-83. Temperatures are the daily maximum river temperatures. The lifts began operating about 1 May in all years and ceased about 15 July. Wavy line near the base of each panel identifies days on which the lifts were not operated. 751 though the data from the two runs differed greatly in time and space, the general migration pattern in relation to river temperature was remarkably similar. The behavior of the sea lampreys in the St. Johns and Connecticut Rivers also appeared similar. In 1974, the first migrants were collected at 13°C at the Mactaquac fish lift (Beamish and Potter 1975), and from 1980 to 1983 the first migrants were passed in the Holyoke fish lift at 10.5° -15.5°C. The peak of the run was also similar— 17°-19°C in the St. Johns River and 16°-21°C in the Connecticut River. Mean length of the 45 sea lampreys tagged was 73.2 cm (range, 63.0-80.0 cm). Five were not re- located either because the tag failed or the fish moved downstream over the dam. No tagged sea lamprey died during the study. The remaining 40 fish were followed for a total of 224 h during 24 d (12 May-4 June; Fig. 3). Since sea lampreys mi- grated upstream at Holyoke until 30 June 1982 (Fig. 2), for the most part we observed the movement of early migrants. During the study, water tempera- ture increased from 13° to 22 °C; river discharge gradually decreased from 60.4 m3/s on 12 May to 50.9 m3/s on 31 May. Twenty sea lampreys moved >23 km and 4 reached Cabot Station. Nineteen were last located near the mouths of the Fort or Mill Rivers or Russelville Brook (Fig. 1). Spawning of tagged fish was verified in the tributaries— an in- dication that normal behavior resumed after the sea lampreys were tagged. Sea lampreys moved upstream at ground speeds of 0.1-3.5 km/h. The daily mean rate of movement including rest periods was 1.01 km/h ± 0.75 (mean ±SD; range, 0.1-2.7 km/h; N = 40) or 0.4 body length/s. The mean rate, excluding rest periods, was 1.51 km/h ± 0.53 (range, 0.1-3.5 km/h; N = 39) or 0.6 body length/s. Early migrants moved a mean of 0.1-1.2 km/h; and three migrants that were observed during the peak passage at the fish lift on 2 June had the fastest mean daily rate of 2 km/h (Fig. 3). Among landlocked sea lampreys, early migrants have a slower rate of movement than peak migrants because they rest more (Applegate 1950; Skidmore 1959; Larsen 1980). Our observations during the peak period after 30 June did not indicate a sus- tained increase in the rate of movement (Fig. 3). Because we only observed a few peak migrants, additional study is necessary to compare the rates of movement between early and peak migrants. The movement rates of sea lampreys in the Con- necticut River were the highest reported for the species. Landlocked sea lampreys moved at much lower rates of 0.02-0.21 km/h (Applegate and Smith 1950; Skidmore 1959; Wigley 1959). Beamish (1974) found a maximum swimming speed of 1.08 km/h (30 cm/s) for landlocked adults in the laboratory. Using the energetics of adult sea-run sea lampreys dur- ing a 35-d upstream move into the fish lift at Mac- taquac Dam on the St. John River, Beamish (1979) estimated the rate to be 0.23 km/h for males and 0.26 km/h for females, or 0.1 body length/s for both. This rate was similar to that of the landlocked form. Because the sea-run adults are much larger than landlocked adults, they should swim faster. Our results suggest that the 0.2 km/h rate which was estimated for the St. John River adults may be in- correct, possibly because the fish were delayed ii 2 £ 2 E Ld cr o 23446474845222431 634312 t \\\ -X l_l l_J_ X J„ J i I i L_. L 12 14 16 18 20 22 24 26 28 30 I 3 MAY JUNE Figure 3.— Daily mean rates of movement of radio-tagged sea lampreys (open circles). (Vertical lines show standard errors; numbers of lampreys monitored are shown above each mean.) 752 several days before finding the entrance to the fish lift at Mactaquac Dam. Diel movement rates were monitored on 13 and 17 May (early migrants) and 26 and 30 May and 1 June (peak migrants). Movement was slowest from 1200 to 1700 h (Fig. 4). Nocturnal behavior was strongest among the early migrants; peak migrants had a higher rate of movement because they also moved during the day (mornings only). A similar pat- tern for landlocked adults was found by Kleerekoper et al. (1961). In summary, except for the longer summer migra- tion and the slower rate of upstream movement, the behavior of sea-run sea lampreys in the Connecticut and St. Johns Rivers was similar to that of the land- locked sea lampreys in the Ocqueoc River. The timing of the runs in relation to temperature and the diel movement patterns appears very stable, probably with important survival or reproductive advantages. 2400 0500 1200 1700 2000 to to to to to 0500 1200 1700 2000 2400 HOUR Figure 4.— Mean movement rates of early migrants (solid circles) monitored 13 and 17 May (N = 13), and peak migrants (open circles) monitored 26 and 30 May and 1 June 1982 (N = 7). (Ver- tical lines show standard errors.) Acknowledgments We thank D. Stier, A. Richmond, J. Nicholson, T. Clifford, C. Hall, J. Burnett, J. Bain, and J. Idoine for assistance with field work. The project was funded by Federal Aid Project AFS-4-R-21 and Dingell-Johnson Project 5-29328 to the Massachu- setts Cooperative Fishery Research Unit. We thank Holyoke Water Power Company for providing space for the holding tanks. Literature Cited Applegate, V. C. 1950. Natural history of the sea lamprey, Petromyzon mari- nus, in Michigan. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 55, 237 p. Applegate, V. C, and B. R. Smith. 1950. Sea lamprey spawning runs in the Great Lakes in 1950. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 61, 49 p. Beamish, F. W. H. 1974. Swimming performance of adult sea lamprey, Petro- myzon marinus, in relation to weight and temperature. Trans. Am. Fish. Soc. 103:355-358. 1979. Migration and spawning energetics of the anadromous sea lamprey, Petromyzon marinus. Environ. Biol. Fishes 4:3-7. Beamish, F. W. H., and I. C. Potter. 1975. The biology of the anadromous sea lamprey (Petro- myzon marinus) in New Brunswick. J. Zool. (Lond.) 177: 57-72. Bigelow, H. B., and W. C. Schroeder. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53:1-577. Kleerekoper, H., G. Taylor, and R. Wilton. 1961. Diurnal periodicity in the activity of Petromyzon mari- nus and the effects of chemical stimulation. Trans. Am. Fish. Soc. 90:73-78. Knight, A. E. 1975. A tuned-antenna radio telemetry tag for fish. Under- water Telem. Newsl. 5:13-16. Larsen, L. 0. 1980. Physiology for adult lampreys, with special regard to natural starvation, reproduction, and death after spawning. Can. J. Fish. Aquat. Sci. 37:1762-1779. Moore, H. H., F. H. Dahl, and A. K. Lamsa. 1974. Movement and recapture of parasitic phase sea lam- preys (Petromyzon marinus) tagged in the St. Marys River and Lakes Huron and Michigan, 1963-67. Great Lakes Fish. Comm. Tech. Rep. 27, 19 p. Skidmore, J. F. 1959. Biology of spawning-run sea lampreys (Petromyzon marinus) in the Pancake River, Ontario. M.S. Thesis, Univ. Western Ontario, London, Ont., 87 p. Smith, B. R., and 0. R. Elliott. 1952. Movement of parasitic-phase sea lampreys in Lakes Huron and Michigan. Trans. Am. Fish. Soc. 82:123-128. Stier, K., and B. Kynard. 1986. Abundance, size, and sex ratio of adult sea-run sea lam- prey, Petromyzon marinus, in the Connecticut River. Fish. Bull., U.S. 84:476-480. Wigley, R. L. 1959. Life history of the sea lamprey of Cayuga Lake, New York. U.S. Fish Wildl. Serv., Fish. Bull. 59:559-617. Kathleen Stier Boyd Kynard Massachusetts Cooperative Fishery Research Unit University of Massachusetts 204 Holdsworth Hall Amherst, MA 01003 753 VARIATIONS IN THE MORPHOLOGY OF FISTULICOLA PLICATUS RUDOLPHI (1802) (CESTODA:PSEUDOPHYLLIDEA) FROM THE SWORDFISH, XIPHIAS GLADIUS L., IN THE NORTHWEST ATLANTIC OCEAN During the course of a survey of the helminth parasites of the swordfish, Xiphias glasius L., from the Northwest Atlantic Ocean, several morpholo- gical variations were observed in specimens of the pseudophyllidean tapeworm, Fistulicola plicatus. The most notable of these variations were pseudo- scolex form and proglottid shape and size. Methods of scolex attachment to the organ wall, descriptions of pseudoscolex structures, and organ specific varia- tions in the morphology of F. plicatus are given. Materials and Methods A sample of 303 gills and gastrointestinal tracts of swordfish was collected from four geographical areas in the Northwest Atlantic Ocean in the late summer and early fall of 1980. The areas sampled and the number of swordfish collected from each geographical area are as follows: Cape Hatteras (74), Georges Bank (90), Scotian Shelf (69), and Grand Bank (70); all collected by longline gear and frozen at sea. The swordfish were later dissected and ex- amined for helminth parasites in the laboratory. Pseudophyllidean cestodes were removed from the infected organ and fixed whole in 70% alcohol or 10% Formalin1. Several infected organs were fixed whole in Bouin's fluid or 10% Formalin. Speci- mens used for taxonomic examinations were stained in Erlich's hematoxylin, Blachin's lactic acid car- mine, or Semichon's aceto-carmine. Camera lucida drawings were made from fixed, unstained speci- mens. Results Fistulicola plicatus has been reported from the swordfish by Linton (1901), Cooper (1918), Nigrelli (1938), and lies (1970). In this study F. plicatus was found in the intestines and recturns of swordfish from all four sampling areas. Considerable morpho- logical variation was found between individuals of this species. Variations were in scolex form, overall parasite length, and proglottid shape and size. About 50% of specimens recovered exhibited a scolex and proglottid structure characteristic of specimens 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. described by Yamaguti (1959). Scolices from these were arrow-shaped and possessed two simple, leaf- shaped bothridia (Fig. 1). Any variation from this scolex form were considered to be pseudoscolices. Proglottids from specimens described by Yamaguti (1959) were short and broad with foliate lateral edges. Internal proglottid morphology was not easily seen in any of the specimens examined during the present study, although nerve trunk location (near lateral margins), cirrus-sac and vagina location (on opposite lateral margins), and egg shell structure (thick-shelled and operculate) were occasionally observable. A total of 29 specimens recovered had penetrated the wall of the infected organ. Occasionally the tapeworms penetrated the organ wall and retained typical scolex form, i.e., arrow-shaped with simple, well-developed bothridia but, in the majority of cases, complete perforation of the organ wall re- sulted in the formation of a pseudoscolex. Attach- ment to the organ wall (rectum and intestine) was achieved in the following four ways: 1) By complete perforation of the organ wall, the scolex and a portion of the neck encapsulated in a rounded, host-produced cyst attached to the organ serosa. Scolices recovered from these cysts were usually arrow-shaped with typical bothridia, or occasionally found as a round, transparent, fluid-filled bag, which pos- sessed rudimentary or no apparent bothridia (Fig. 2). 2) By complete perforation of the organ wall, the scolex and a portion of the neck encased in a tubular, host-produced sheath, attached along its entire length to the organ serosa. Occa- sionally this sheath was entwined with the mesenteries associated with the infected organ. Pseudoscolices found within these sheaths were long, rounded, and slender, and exhibited no bothridia (Fig. 3). 3) By complete penetration of the organ wall, the scolex markedly enlarged (up to 6 cm in length), lying free, and unencapsulated in the peritoneal cavity. Pseudoscolices of this type were long, broad, pseudosegmented, and pos- sessed well-developed bothridia (Fig. 4). 4) In this case the scolex did not fully penetrate the organ wall, but perforated the wall to a slight depth, and remained in that position. Often specimens were found to exhibit this slight organ wall penetration and re-emerge in- to the lumen of the organ. In these cases the 754 FISHERY BULLETIN: VOL. 84, NO. 3, 1986. 4mm 0.5cm L 1.0cm Figure 1.— Typical scolex from Fistulicola plicatus, as described and figured by Yamaguti (1959). Figure 2.— Pseudoscolex (1st type). Figure 3.— Pseudoscolex (2d type). Figure 4.— Pseudoscolex (3d type). 755 1) scolex type described by Yamaguti (1959) was retained. Fistulicola plicatus specimens recovered from the lumen of the intestines were morphologically dif- ferent from those collected from the rectum. They were long, up to 1 m in length, and exhibited longer, less-broad strobila than those characteristic of the rectal forms. All specimens of F. plicatus recovered from the anterior portion of the intestine exhibited the previously described first type of scolex attach- ment to the organ wall, i.e., the scolex perforated the organ wall and was encapsulated in a rounded, host-produced cyst attached to the intestinal serosa. The scolex penetrated the anterior portion of the intestine, with the strobila projecting posterior through the length of the organ. Very small F. plicatus were found in the posterior portion of the intestine. These exhibited shallow penetration by an unmodified scolex. Fistulicola plicatus specimens found in the rec- tum of swordfish were usually <20 cm in length and possessed very broad strobila. These rectal forms exhibited all of the previously described types of scolex attachment and structure, penetrating the organ wall near the rectal sphincter (Fig. 5). Occa- sionally, several tapeworms were found with their necks passing through a single perforation of the rectal wall, their scolices jointly encapsulated in a rounded serosal cyst. Discussion Apex type predators such as the swordfish eat and digest large amounts of prey species and, conse- quently, the intestines and rectums of these fish exhibit high levels of muscular activity. Without per- foration of the organ wall (by the scolex and neck), many tapeworms would probably be voided with the faeces. The development of the pseudoscolex is an adaptation for anchoring the simple, unarmed scolex to the organ wall. It is clear that F. plicatus secretes a powerful digestive enzyme which enables the scolex to penetrate the very muscular walls of the intestine and rectum of swordfish. lies (1970) found many examples of pseudoscolex variation in 24 specimens from swordfish in the Northwest Atlan- tic Ocean. Several of these variations are similar to those found in this study. It is obvious from this study, and studies such as lies (1970), that F. plicatus is a very adaptable tapeworm and will develop any pseudoscolex structure which is neces- Figure 5. —Fistulicola plicatus (in situ) from rectum of Xiphias gladius. 756 sary to anchor itself to the organ wall. Large samples of swordfish intestines and rectums will in- variably show many variations in the pseudoscolex structure of F. plicatus. Acknowledgments We thank L. S. Uhazy for initiating the study of these parasites, C. lies for comments on the manu- script, P. W. G. McMullon for advice on figures, and B. Garnett for preparing the typescript. Literature Cited Cooper, A. R. 1918. North American pseudophyllidean cestodes from fishes. III. Biol. Monogr. 4:1-243. Iles, C. 1970. A preliminary investigation of the parasites of the gills and gastrointestinal tract of the swordfish Xiphias gladius L. from the Northwest Atlantic. J. Fish. Res. Board Can. MS Rep. Ser. 1092, 11 p. Linton, E. 1901. Parasites of fishes of the Woods Hole region. Bull. U.S. Fish. Comm. 19:405-492. Nigrelli, R. F. 1938. Parasites of the swordfish, Xiphias gladius Linnaeus. Am. Mus. Novit. 996:1-16. Yamaguti, S. 1959. Systema helminthum. Vol. 2. The cestodes of verte- brates. Intersci. Publ., N.Y., 860 p. W. E. Hogans Identification Centre Department of Fisheries and Oceans Biological Station St. Andrews, Nova Scotia E0G 2X0, Canada Marine Fish Division Bedford Institute of Oceanograph Department of Fisheries and Oceans P.O. Box 1006 Dartmouth, Nova Scotia B2Y JfA2, Canada P. C. F. Hurley 757 INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN Manuscripts submitted to the Fishery Bulletin will reach print faster if they conform to the following instructions These are not absolute requirements, of course, but desiderata. CONTENT OF MANUSCRIPT The title page should give only the title of the paper, the author's name, his affiliation, and mailing address, in- cluding ZIP code The abstract should not exceed one double-spaced page In the text, Fishery Bulletin style, for the most part, follows that of the U.S. Government Printing Office Style Manual. Fish names follow the style of the American Fisheries Society Special Publication No. 12, A List of Com- mon and Scientific Names of Fishes from the United States and Canada, Fourth Edition, 1980. lext footnotes should be typed separately from the text. Figures and tables, with their legends and headings, should be self-explanatory, not requiring reference to the text. Their placement should be indicated in the right-hand margin of the manuscript. Preferably figures should be reduced by photography to 53A inches (for single-column figures, allowing for 50% reduction in printing), or to 12 inches (for double-column figures). The maximum height, for either width, is 14 inches. Photographs should be printed on high quality or glossy paper. Do not send original drawings to the Scientific Editor; if they, rather than the photographic reductions, are needed by the printer, the Scientific Publications Office will request them. Each table should start on a separate page. Consistency in headings and format is desirable Vertical rules should be avoided, as they make the tables more expensive to print Footnotes in tables should be numbered sequentially in arabic numerals. To avoid confusion with powers, they should be placed to the left of numerals. Acknowledgments, if included, are placed at the end of the text. Literature is cited in the text as: Lynn and Reid (1968) or (Lynn and Reid 1968). All papers referred to in the text should be listed alphabetically by the senior author's sur- name under the heading "Literature Cited." Only the author's surname and initials are required in the literature cited. The accuracy of the literature cited is the respon- sibility of the author. Abbreviations of names of periodicals and serials should conform to Serial Sources for the BIOSIS DataBase. (Chemical Abstracts aiso uses this system, which was developed by the American Standards Associa? i Common abbreviations and symbols, such as mm, m, g, L, mL, mg, °C (for Celsius), %, °/00, and so forth, should be used. Abbreviate units of measure only when used with numerals. Periods are only rarely used with abbrevia- tions. We prefer that measurements be given in metric units; other equivalent units may be given in parentheses. FORM OF THE MANUSCRIPT The original of the manuscript should be typed, double- spaced, on white bond paper. Please triple space above headings. We would rather receive good duplicated copies of manuscripts than carbon copies. The sequence of the material should be: TITLE PAGE ABSTRACT TEXT LITERATURE CITED TEXT FOOTNOTES APPENDIX TABLES (Each table should be numbered with an arabic numeral and heading provided). LIST OF FIGURES (Entire figure legends) FIGURES (Each figure should be numbered with an arabic numeral; legends are desired) ADDITIONAL INFORMATION Send the ribbon copy and two duplicated or carbon copies of the manuscript to: Dr. William J. Richards, Scientific Editor Fishery Bulletin Southeast Fisheries Center Miami Laboratory National Marine Fisheries Service, NOAA 75 Virginia Beach Drive Miami, FL 33149-1099 Fifty separates will be supplied to an author free of charge and 50 supplied to his organization. No covers will be supplied. Contents— Continued ROGERS, S. GORDON, HIRAM T. LANGSTON, and TIMOTHY E. TARGETT. Ana- tomical trauma to sponge-coral reef fishes captured by trawling and angling .... 697 QUAST, JAY C. Annual production of eviscerated body weight, fat, and gonads by Pacific herring, Clwpea harengus pallasi, near Auke Bay, southeastern Alaska. . . 705 WENNER, CHARLES A., WILLIAM A. ROUMILLAT, and C. WAYNE WALTZ. Con- tributions to the life history of Black sea bass, Centropristis striata, off the south- eastern United States 723 Notes LENARZ, WILLIAM H., and TINA WYLLIE ECHE VERRIA. Comparison of visceral fat and gonadal fat volumes of yellowtail rockfish, Sebastesjlavidus, during a normal year and a year of El Nino conditions 743 SORENSEN, PETER W, MARCO L. BIANCHINI, and HOWARD E. WINN. Diel foraging activity of American eels, Anguilla rostrata (Lesueur), in a Rhode Island estuary 746 KILLAM, KRISTIE, and GLENN PARSONS. First record of the longfin mako, Isurus paucus, in the Gulf of Mexico 748 STIER, KATHLEEN, and BOYD KYNARD. Movement of sea-run sea lampreys, Petro- myzon marinus, during the spawning migration in the Connecticut River 749 HOGANS, W E., and P. C. F. HURLEY. Variations in the morphology of Fistulicola plicatus Rudolphi (1802) (Cestoda:Pseudophyllidea) from the swordfish, Xiphias gladius L., in the Northwest Atlantic Ocean 754 • GPO 593-096 ^O'Cq, S*TES O* * Fishery Bulletin LIBRA* Vol. 84, No. 4 'i«L* !-; . October 1986 POLOVINA, JEFFERY J., and STEPHEN RALSTON. An approach to yield assessment for unexploited resources with application to the deep slope fishes of the Marianas. . 759 YANG, W. T, R. F. HIXON, P. E. TURK, M. E. KREJCI, W. H. HULET, and R T. HANLON. Growth, behavior, and sexual maturation of the market squid, Loligo opalescens, cultured through the life cycle 771 BODKIN, JAMES LEE. Fish assemblages in Macrocystis and Nereocystis kelp forests of central California 799 STEPIEN, CAROL A. Life history and larval development of the giant kelpfish, Hetero- stiehus rostratus Girard, 1854 809 PAULY, DANIEL. A simple method for estimating the food consumption of fish popu- lations from growth data and food conversion experiments 827 FINUCANE, JOHN H., L. ALAN COLLINS, HAROLD A. BRUSHER, and CARL H. SALOMAN. Reproductive biology of king mackerel, Scomberomorus cavalla, from the southeastern United States 841 FARLEY, C. A., S. V. OTTO, and C. L. REINISCH. New occurrence of epizootic sarcoma in Chesapeake Bay soft shell clams, Mya arenaria 851 FOLKVORD, ARILD, and JOHN R. HUNTER. Size-specific vulnerability of northern anchovy, Engraulis mordax, larvae to predation by fishes 859 SMITH, JOSEPH W, and JOHN V. MERRINER Observations on the reproductive biology of the cownose ray, Rhinoptera bonasus, in Chesapeake Bay 871 McGOWAN, MICHAEL F Northern anchovy, Engraulis mordax, spawning in San Fran- cisco Bay, California, 1978-79, relative to hydrography and zooplankton prey of adults and larvae 879 HUNTER, J. ROE, BEVERLY J. MACEWICZ, and JOHN R. SIBERT. The spawning frequency of skipjack tuna, Katsuwonus pelamis, from the South Pacific 895 HOUDE, EDWARD D, and LAWRENCE LUBBERS III. Survival and growth of striped bass, Morone saxatilis, and Morone hybrid larvae: laboratory and pond enclosure experiments 905 DAILEY, MURRAY D, and STEPHEN RALSTON. Aspects of the reproductive biology, spatial distribution, growth, and mortality of the deepwater caridean shrimp, Heterocarpus laevigatas, in Hawaii 915 RALSTON, STEPHEN. An intensive fishing experiment for the caridean shrimp, Hetero- carpus laevigatus, at Alamagan Island in the Mariana Archipelago 927 DITTY, JAMES G. Ichthyoplankton in neritic waters of the northern Gulf of Mexico off Louisiana: composition, relative abundance, and seasonality 935 (Continued on back cover) Seattle, Washington U.S. DEPARTMENT OF COMMERCE Malcolm Baldrige, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION Anthony J. Calio, Administrator NATIONAL MARINE FISHERIES SERVICE William E. Evans, Assistant Administrator Fishery Bulletin The Fishery Bulletin carries original research reports and technical notes on investigations in fishery science, engineering, and economics. The Bulletin of the United States Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates were issued as documents through volume 46; the last document was No 1 103. Beginning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate appeared as a numbered bulletin. A new system began in 1963 with volume 63 in which papers are bound together in a single issue of the bulletin instead of being issued individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued quarterly. In this form, it is available by subscription from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. It is also available free in limited numbers to libraries, research institutions, State and Federal agencies, and in exchange for other scientific publications. SCIENTIFIC EDITOR, Fishery Bulletin Dr. William J. Richards Southeast Fisheries Center Miami Laboratory National Marine Fisheries Service, NOAA Miami, FL 33149-1099 Editorial Committee Dr. Bruce B. Collette Dr. Jay C. Quast National Marine Fisheries Service National Marine Fisheries Service Dr. Reuben Lasker Dr. Carl J. Sindermann National Marine Fisheries Service National Marine Fisheries Service Mary S. Fukuyama, Managing Editor The Fishery Bulletin (ISSN 0090-0656) is published quarterly by the Scientific Publications Office, National Marine Fisheries Service, NOAA, 7600 Sand Point Way NE, BIN C15700, Seattle, WA 98115. Second class postage is paid in Seattle. Wash., and additional offices. POSTMASTER send address changes for subscriptions to Fishery Bulletin, Superintendent of Documents, U.S. Government Printing . Washington, DC 20402. Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated. Fishery Bulletin CONTENTS Vol. 84, No. 4 October 1986 POLOVINA, JEFFERY J., and STEPHEN RALSTON. An approach to yield assessment for unexploited resources with application to the deep slope fishes of the Marianas. . 759 YANG, W. T, R. F. HIXON, P. E. TURK, M. E. KREJCI, W. H. HULET, and R. T. HANLON. Growth, behavior, and sexual maturation of the market squid, Loligo opalescens, cultured through the life cycle 771 BODKIN, JAMES LEE. Fish assemblages in Macrocystis and Nereocystis kelp forests of central California 799 STEPIEN, CAROL A. Life history and larval development of the giant kelpfish, Hetero- stichus rostratus Girard, 1854 809 PAULY, DANIEL. A simple method for estimating the food consumption of fish popu- lations from growth data and food conversion experiments 827 FINUCANE, JOHN H., L. ALAN COLLINS, HAROLD A. BRUSHER, and CARL H. SALOMAN. Reproductive biology of king mackerel, Scomberomorus cavalla, from the southeastern United States 841 FARLEY, C. A., S. V. OTTO, and C. L. REINISCH. New occurrence of epizootic sarcoma in Chesapeake Bay soft shell clams, Mya armaria 851 FOLKVORD, ARILD, and JOHN R. HUNTER. Size-specific vulnerability of northern anchovy, Engraulis mordax, larvae to predation by fishes 859 SMITH, JOSEPH W., and JOHN V. MERRrNER. Observations on the reproductive biology of the cownose ray, Rhinoptera bonasus, in Chesapeake Bay 871 McGOWAN, MICHAEL F Northern anchovy, Engraulis mordax, spawning in San Fran- cisco Bay, California, 1978-79, relative to hydrography and zooplankton prey of adults and larvae 879 HUNTER, J. ROE, BEVERLY J. MACEWICZ, and JOHN R. SIBERT The spawning frequency of skipjack tuna, Katsuwonus pelamis, from the South Pacific 895 HOUDE, EDWARD D, and LAWRENCE LUBBERS III. Survival and growth of striped bass, Morone saxatilis, and Morone hybrid larvae: laboratory and pond enclosure experiments 905 DAILEY, MURRAY D, and STEPHEN RALSTON. Aspects of the reproductive biology, spatial distribution, growth, and mortality of the deepwater caridean shrimp, Heterocarpus laevigatus, in Hawaii 915 RALSTON, STEPHEN. An intensive fishing experiment for the caridean shrimp, Hetero- carpus laevigatus, at Alamagan Island in the Mariana Archipelago 927 DITTY, JAMES G Ichthyoplankton in neritic waters of the northern Gulf of Mexico off Louisiana: composition, relative abundance, and seasonality 935 (Continued on next page) Seattle, Washington 1986 ^■riaa B^W^jmJ I -*- *-- M9KN DM^ni UUUI41W LIBRARY JAN 29 1987 -oas Hole, Mass. For sale by the Superintendent of Documents, U.S. Government Printing Office,, Washington DC 20402— Subscription price per year: $21.00 domestic and $26.25 foreign. Cost per single issue: $6.50 domestic and $8.15 foreign. Contents— Continued REXSTAD, ERIC A., and ELLEN K. PIKITCH. Stomach contents and food consump- tion estimates of Pacific hake, Merluccius productus 947 PEREZ, MICHAEL A., and MICHAEL A. BIGG. Diet of northern fur seals, Callorhmus ursinus, off western North America 957 TESTER, PATRICIA A., and ANDREW G. CAREY, JR. Instar identification and life history aspects of juvenile deepwater spider crabs, Chionoecetes tanneri Rathbun . . . 973 Notes CODY, TERRY J., and BILLY E. FULS. Comparison of catches in 4.3 m and 12.2 m shrimp trawls in the Gulf of Mexico 981 POWELL, ALLYN B., and GERMANO PHONLOR. Early life history of Atlantic men- haden, Brevoortia tyrannus, and gulf menhaden, B. patronus 991 PODNIESINSKI, GREG S., and BERNARD J. McALICE. Seasonality of blue mussel, Mytilus edulis L., larvae in the Damariscotta River estuary, Maine, 1969-77 995 Index 1003 Notices 1017 The National Marine Fisheries Service (NMFS) does not approve, recommend or en- dorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to NMFS, or to this publication furnished by NMFS, in any advertising or sales promotion which would indicate or imply that NMFS ap- proves, recommends or endorses any proprietary product or proprietary material mentioned herein, or which has as its purpose an intent to cause directly or indirect- ly the advertised product to be used or purchased because of this NMFS publication. AN APPROACH TO YIELD ASSESSMENT FOR UNEXPLOITED RESOURCES WITH APPLICATION TO THE DEEP SLOPE FISHES OF THE MARIANAS Jeffrey J. Polovina and Stephen Ralston1 ABSTRACT A comprehensive approach to estimate the maximum sustainable yield (MSY) for a tropical multispecies resource which lacks catch and effort data is presented. This yield assessment approach was used to design a fishery resource assessment survey of the Mariana Archipelago. An application of the method is presented to estimate the MSY for a multispecies bottom fish resource, based on data collected during the survey. The annual MSY for the deep slope fishes (primarily snappers and groupers) of the Mariana Archipelago is estimated to be 109 t, which for comparative purposes is equivalent to 222 kg/nmi of 200 m isobath or 0.3 t/km . Assessment of tropical resources has always created major problems in fisheries research (Saila and Roedel 1979; Pauly and Murphy 1982). This has been largely due to three factors: technical dif- ficulties in aging, a high species diversity in tropical communities, and what is typically a multiplicity of artisanal gears used in these fisheries. The latter problem has been especially difficult to surmount, making it difficult to determine not only the level of fishing effort but sometimes even the total catch. Without these data many standard fish- eries techniques such as stock-production methods are inapplicable (but see Csirke and Caddy 1983). In recent years, however, new methods and modifications of existing methods have been pro- posed to estimate growth and mortality parameters, standing crop, and yield for fish stocks in the absence of a time series of commercial catch and ef- fort data (Beddington and Cooke 1983; Pauly 1983; Polovina 1986a; Wetherall et al. in press). We will show that several of these techniques can be com- bined, producing an integrated approach to yield assessment designed specifically for tropical fisheries resources in situations where catch and ef- fort data are lacking. The approach is then applied to data gathered in a fishery survey of the Mariana Archipelago to estimate maximum sustainable yield (MSY) for a multispecies resource of deep slope snappers and groupers. 'Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI 96822-2396. YIELD ASSESSMENT The equilibrium yield assessment is presented schematically in Figure 1. This approach assumes that growth follows the deterministic von Berta- lanffy curve with parameters K and Lx, that the mortality of fish above the smallest length fully represented in the catch (Lc) occurs at a constant instantaneous rate (Z), and that recruitment is con- stant with R recruits entering the first vulnerable age class annually. It is also assumed that the resource is essentially pristine, such that an estimate of the biomass recruited to the fishery in the absence of exploitation (B^) can be obtained from a catch- per-unit-effort (CPUE) survey and an estimate of catchability. In the discussion section, the effect of relaxing some of these assumptions will be con- sidered. For each species under consideration, the data re- quired for this program, at a minimum, consist of a large length-frequency sample, otolith data and/or a time series of length-frequency data, a systematic CPUE survey, and an estimate of catchability, such as that obtained from an intensive fishing experi- ment. The large length-frequency sample is used to jointly estimate the asymptotic length (Lm) and the ratio of total instantaneous mortality (Z) to the von Bertalanffy growth parameter (K) based on the following relationship: 0 = ZIK = (Lm - 1)1 (T - Lc), where Lc is a parameter defined above and I is the mean length of all fish greater than Lr (Beverton Manuscript accepted February 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 759 FISHERY BULLETIN: VOL. 84, NO. 4 Intensive Fishing Length Frequency © Systematic Survey of Relative Abundance Otoliths Time Series Length Frequency I Yield and Relative Spawning Stock Equations f MSY j Figure 1.— Schematic of the yield assessment approach. A more general approach to fishery assessment which includes a treatment of catch and effort data as well is given in Munro (1983); our Figure 1 represents a detailed subset of Munro's figure 1 (1983). and Holt 1956). For a series of Lc values at inter- vals beginning with the smallest Lc and going up to L^, there will be a corresponding set of I values. By solving the ZIK equation above for I as a func- tion of Lc, the following relationship is obtained: T = LJ(Q + 1) +Lc(0/(0 + 1)). Thus, regressing a sequence of I values on the cor- responding L( values will produce estimates for the slope and intercept which can be solved for esti- mates of Lx and ZIK (Wetherall et al. in press). Once an estimate of L has been obtained by this method, otolith data and/or a time series of length- frequency data can be fit to the von Bertalanffy growth curve to estimate the growth coefficient K. Estimation of L^ from length-frequency data was used for the Marianas bottom fish data because a large length-frequency sample was available and otolith readings were difficult to interpret for old stages of growth. With an estimate for K, the total mortality rate, Z, can then be estimated as the pro- duct of K and the ratio of ZIK obtained in the previous step. Alternatively, one can estimate Z from a catch curve constructed from a length- frequency sample which has been corrected for nonlinear growth and converted to an age-frequency sample (Pauly 1983). If these techniques are applied to unexploited or lightly exploited resources, the estimate of Z pro- vides an estimate of the instantaneous rate of natural mortality (Af ). However, if fishing mortal- ity is believed significant, an equation to estimate M as a function of K, L^, and mean annual water temperature (T) (in °C) has been developed as follows (Pauly 1983): log10 M = -0.0066 - 0.279 log10 L^ + 0.6543 log10 K + 0.4634 log10 T. Given estimates of K, M, and age of entry to the fishery (tc), the Beverton and Holt (1957) yield per recruit (Y/R) equation can be used to compute the ratio of equilibrium yield to unexploited recruited biomass as a function of fishing mortality (F). The 760 POLOVINA and RALSTON: MSY FOR DEEP SLOPE FISHES equilibrium yield (Y) can be expressed as oo Y = R F J exp {-tM - (t - tc) F) w(t) dt, where w(t) = W^ (1 - exp(-Kt))b, and where W^ is the asymptotic weight and b is the exponent of the length-weight relationship. The unexploited recruited biomass (B ) can be expressed as B = R J w(t) exp ( -Mi) dt. The ratio of equilibrium yield to unexploited re- cruited biomass (Y/BJ) is then independent of Wx and R, depending only on K, M, tc, F, and b. Tables and computational formulae are readily available to evaluate these integrals for Y and 5m as functions of tc and F (Beverton and Holt 1966; Beddington and Cooke 1983). Upon estimation of B^, the equi- librium yield is estimated for a given level of F as the product of YIB and B . * oo oo If a stock is unfished, Bm can be estimated by mapping the relative abundance of the stock in terms of CPUE from a systematic survey and then converting estimates of relative abundance into biomass with an estimate of catchability. There are a number of methods which have been used to esti- mate catchability (Ricker 1975). For work on Pacific island fishery resources, an intensive fishing ap- proach, which fishes a small isolated location heavily and regresses CPUE on cumulative catch (Leslie model), has been used successfully to estimate catch- ability for bottom fishes and shrimp (Polovina 1986a; Ralston 1986). If only one estimate of catchability is obtained, then the standing stock per unit of area is determined as the ratio of CPUE to catchability in the appropriate units of weight or numbers. If several estimates of catchability are available corre- sponding to different levels of CPUE, then it might be appropriate to fit a more general power function relationship between CPUE and standing stock (Bannerot and Austin 1983). The product of YIB^ and B^ as a function of F is the equilibrium yield based on the assumption of constant recruitment. While this assumption will be valid for low levels of exploitation, there will come a point as F increases that recruitment will begin to decline and sustainable yield may thus be less than the yield predicted under the assumption of constant recruitment. Estimating MSY yield as the maximum equilibrium yield obtained over all F from the prod- uct of Y/B^ and B^ may, therefore, overestimate the actual MSY. There are two adjustments which have been proposed to estimate MSY in the absence of detailed knowledge of the spawner-recruit rela- tionship. One approach is to estimate MSY from the constant recruitment yield curve as that yield cor- responding to that level of F where the addition of one unit of mortality increases the yield by 10% of the amount caught by the first unit of F (Gulland 1983, 1984). This level of mortality and correspond- ing yield have been denoted as F01 and Y0A, respectively. A second approach to estimating MSY from the constant recruitment yield curve is to use the Beverton and Holt equation to calculate the ratio of the spawning stock biomass under exploitation (S) to the spawning stock biomass in the absence of exploitation (S0) and to use this ratio as an in- dicator of the sustainability of a yield for a given combination of F and tc . For simplicity, we assume that the age of sexual maturity (tm) is identical for both sexes. Then the unexploited spawning stock biomass (50) is S0 = R J exp(-Aft) w(t) dt, and S = R J exp(-M - (t-tc) F) w(t) dt. Thus, the ratio of S/S0 depends only on M, K, tc, tm, andF. It has been suggested that the spawning stock biomass of a species should not be reduced below 20% of its unexploited level if a substantial reduc- tion in the recruitment is to be avoided (Beddington and Cooke 1983). Thus, the estimate of MSY is determined as the maximum yield from the constant recruitment curve subject to the constraint that F does not exceed the level which reduces the relative spawning stock biomass below 0.20 of S0. ASSESSMENT OF SNAPPERS AND GROUPERS IN THE MARIANAS The Mariana Archipelago consists of a chain of islands and banks on a north-south axis beginning with Galvez Banks and Santa Rosa Reef at the southern end and extending northward to Farallon de Pajaros (30 nmi north of Maug Island). A chain of seamounts also runs on a north-south axis 761 FISHERY BULLETIN: VOL. 84, NO. 4 about 120 nmi west of the high island chain (Fig. 2). Six resource assessment cruises of 40 d each were conducted in the Marianas during the period from May 1982 through June 1984. During these cruises, the deepwater snapper and grouper community along the outer slope was sampled at all 22 islands and banks labeled in Figure 2. Thirteen of these 22 sampling sites were visited at least once during the first three cruises and, again, during the second set ■Maug a Asuncion -20°- -19° o Bank C OAgrihan P Pagan I. -18°- BankO ; uj o co O c 3 •e CD O Uj £2 CD O Maug 0.016 0.000 0.347 0.016 0.425 0.000 0.102 0.094 Asuncion 0.036 0.036 0.089 0.000 0.589 0.018 0.036 0.196 Agrihan 0.016 0.041 0.103 0.064 0.602 0.016 0.110 0.048 Pagan 0.007 0.002 0.089 0.013 0.699 0.023 0.126 0.042 Alamagan 0.010 0.013 0.232 0.011 0.495 0.143 0.059 0.037 Guguan 0.020 0.004 0.182 0.004 0.613 0.047 0.083 0.047 Sarigan 0.016 0.010 0.141 0.010 0.646 0.042 0.057 0.078 Anatahan 0.015 0.035 0.119 0.148 0.540 0.040 0.045 0.059 38-Fathom 0.064 0.028 0.228 0.047 0.434 0.019 0.045 0.136 Esmeralda 0.017 0.051 0.040 0.366 0.397 0.029 0.026 0.074 Farallon de Medimlla 0.052 0.021 0.093 0.166 0.477 0.021 0.093 0.078 Saipan 0.013 0.138 0.087 0.338 0.225 0.000 0.075 0.125 Tinian 0.000 0.000 0.083 0.694 0.000 0.056 0.139 0.028 Aguijan 0.021 0.188 0.063 0.417 0.271 0.000 0.000 0.042 Rota 0.019 0.143 0.162 0.114 0.362 0.029 0.067 0.105 Guam 0.064 0.161 0.258 0.129 0.161 0.000 0.129 0.097 Galvez- Santa Rosa 0.085 0.017 0.364 0.051 0.322 0.009 0.059 0.093 Bank C 0.000 0.017 0.390 0.000 0.356 0.017 0.212 0.009 Bank D 0.015 0.010 0.091 0.005 0.480 0.045 0.349 0.005 Pathfinder 0.059 0.011 0.172 0.004 0.506 0.000 0.215 0.032 Arakane 0.116 0.057 0.188 0.003 0.412 0.000 0.169 0.055 Bank A 0.008 0.004 0.184 0.008 0.607 0.000 0.159 0.029 Table 4. — Mean weight (kg) of the fish caught by bank and species group. Banks and islands Caranx lugubris eo CD 6 CO E S S c .Q. CD •2 c a CO a: .CO C c .§. s» CO a: co a CO c o N a; CO C CO o eo co 11 Uj o 1 3 O c CO o Uj CD -C O Maug 3.784 1.930 0.815 1.585 0.977 6.113 0.893 2.559 Asuncion 3.784 1.930 0.848 1.265 1.344 6.113 0.670 4.595 Agrihan 3.784 1.930 0.784 1.235 1.169 6.113 0.741 7.787 Pagan 3.784 1.930 0.651 1.169 1.094 6.113 0.652 5.068 Alamagan 3.784 1.930 0.834 1.354 1.326 6.113 1.010 2.992 Guguan 3.784 1.930 0.773 1.780 1.216 6.113 0.815 2.400 Sarigan 3.784 1.930 0.642 1.025 1.204 6.113 0.811 5.279 Anatahan 3.784 1.930 0.556 1.169 0.874 6.113 0.586 2.631 38-Fathom 3.784 1.930 0.532 1.193 0.874 6.113 0.798 2.523 Esmeralda 3.784 1.930 0.567 1.014 0.782 6.113 0.702 8.409 Farallon de Medinilla 3.784 1.930 0.439 1.265 0.891 6.113 0.575 1.963 Saipan 3.784 1.930 0.577 0.992 0.837 6.113 0.773 1.353 Tinian 3.784 1.930 0.653 1.003 1.017 6.113 0.422 0.520 Aguijan 3.784 1.930 0.480 0.927 0.760 6.113 0.753 0.770 Rota 3.784 1.930 0.542 1.222 0.667 6.113 0.506 3.168 Guam 3.784 1.930 0.606 1.112 0.780 6.113 0.673 0.600 Galvez- Santa Rosa 3.784 1.930 0.522 1.206 0.979 6.113 0.801 2.085 Bank C 3.784 1.930 0.761 1.265 1.267 6.113 0.923 0.920 Bank D 3.784 1.930 0.961 1.710 1.169 6.113 0.983 1.070 Pathfinder 3.784 1.930 1.953 1.381 1.218 6.113 0.875 7.348 Arakane 3.784 1.930 0.860 1.350 0.949 6.113 0.791 2.338 Bank A 3.784 1.930 0.636 1.605 0.984 6.113 0.811 5.504 765 FISHERY BULLETIN: VOL. 84, NO. 4 Table 5.— The unexploited recruited biomass by bank for each species groups in metric tons. Banks and islands Caranx lugubris Pristipomoides filamentosus -5 co .co c c S. CD 1 CO c o N CL co c CO o CO w 11 CO 3 O c ■s CO o uj i2 0) O Maug 0.5 0 2.2 0.2 3.3 0 0.7 1.9 Asuncion 0.5 0.2 0.3 0 2.9 0.4 0.1 3.3 Agrihan 0.7 0.9 0.9 0.9 8.2 1.1 0.9 4.4 Pagan 0.6 0.1 1.2 0.3 15.9 2.9 1.7 4.4 Alamagan 0.1 0.1 0.8 0.1 2.7 3.5 0.2 0.5 Guguan 0.3 <0.1 0.6 <0.1 3.2 1.2 0.3 0.5 Sarigan 0.2 0.1 0.3 <0.1 2.8 0.9 0.2 1.5 Anatahan 0.3 0.4 0.4 1.0 2.8 1.5 0.2 0.9 38-Fathom 0.3 0.1 0.2 0.1 0.5 0.2 <0.1 0.5 Esmeralda 0.3 0.4 0.1 1.6 1.3 0.7 0.1 2.7 Farallon de Medinilla 7.5 1.5 1.6 8.0 16.3 4.9 2.1 5.8 Saipan 0.6 3.6 0.7 4.6 2.6 0 0.8 2.3 Tinian 0 0 0.5 6.0 0 2.9 0.3 0.1 Aguijan 0.7 3.3 0.3 3.6 1.9 0 0 0.3 Rota 0.7 2.6 0.8 1.3 2.2 1.6 0.5 3.1 Guam 4.8 6.1 3.1 2.8 2.5 0 1.7 1.1 Galvez- Santa Rosa 7.5 0.8 4.5 1.4 7.4 1.2 1.1 4.6 Bank C 0 0.1 0.8 0 1.2 0.3 0.5 <0.1 Bank D 0.2 0.1 0.2 <0.1 1.5 0.7 0.9 <0.1 Pathfinder 0.5 <0.1 0.4 <0.1 0.3 0 0.4 0.5 Arakane 0.6 0.2 0.2 <0.1 0.6 0 0.2 0.2 Bank A 0.1 <0.1 0.2 <0.1 1.2 0 0.3 0.3 Total 32.5 32.0 24.5 43.1 85.4 26.2 15.9 42.8 Table 6.— The total unexploited recruited biomass (BJ in metric tons (t) and the total unexploited recruited biomass per nautical mile (nmi) of 200-m con- tour in kilograms (kg) by bank. Total unexploited Biomass per nmi of Banks and islands recruited biomass (t) 200 i m contour (kg) Northern banks and islands Maug 8.8 850.3 Asuncion 7.7 689.7 Agrihan 18.1 991.1 Pagan 27.0 900.7 Alamagan 8.0 706.8 Guguan 6.1 659.7 Sarigan 6.1 714.0 Anatahan 7.6 440.6 38-Fathom 1.8 637.1 Esmeralda 7.2 584.1 Total 98.4 Mean 717.4 Southern banks and islands Farallon de Medinilla 47.7 620.2 Saipan 15.3 290.9 Tinian 10.0 346.0 Aguijan 10.1 637.0 Rota 12.4 391.2 Guam 22.2 260.6 Galvez-Santa Rosa 28.5 542.4 Total 146.2 Mean 441.2 Western seamounts Bank C 2.9 973.0 Bank D 3.6 1,207.0 Pathfinder 3.1 1,024.0 Arakane 2.0 695.9 Bank A 2.1 594.7 Total 13.7 Mean 898.9 766 POLOVINA and RALSTON: MSY FOR DEEP SLOPE FISHES ploited biomass per nautical mile of 200 m contour in the subsequent yield estimation, in place of the values computed from the bank CPUE values for the inhabited southern islands (Saipan, Tinian, Rota, and Guam). For each species group with values of K, M, tc, and F, the ratio of fishery yield to unexploited recruited biomass {YIBJ can be computed from the Beverton and Holt yield equations (Beddington and Cooke 1983). The product of YIB^ with the species group unexploited recruited biomass estimates (Table 5) results in estimates of equilibrium yield for the seven species for which estimates of K and M are available. For the eighth group, which consists of all other species, the ratio of yield to B^ is taken as the ratio of total yield for the seven species divided by their total Bm. For a fixed F, the sum of the equilibrium yield of the eight species groups at a bank is the bank equilibrium yield, and the sum of the equilibrium yields for a species group over all the banks is the species group equilibrium yield. The equilibrium yield for the multispecies bottom fish complex fished with handline gear in the 125- 275 m depth range for the 22 islands and banks of the Mariana Archipelago increases rapidly as a func- tion of F to a level of about 90 t and beyond that exhibits a gradual increase with increased fishing mortality (Table 7). The MSY estimation approach estimates MSY as the yield from the constant re- cruitment yield curve corresponding to that level of mortality where a marginal increase in one unit of Table 7. — Total annual sustainable handline yield in metric tons (t) for a range of fishing mortalities. Fishing mortality (F) Total yield (t) 0.1 23 0.5 64 11.0 182 1.5 89 2.0 92 2.5 94 mortality increases the catch by 0.1 of the amount caught by the first unit of F. The value of F0A for the bottom fish resource in the Marianas is esti- mated to be F = 1.0 and the corresponding annual equilibrium yield is 82 t (Table 7). The equilibrium yield value of 82 t, which cor- responds to a fishing mortality of 1.0, is based on the current estimated age of entry to the fishery and not necessarily the age of entry which maximizes the YIR. For a fishery mortality of 1.0, the estimated age of entry which maximizes YIR is computed from the Beverton and Holt equation and compared with the current age of entry for each species (Table 8). With the exception of the jack, Caranx lugubris, the age of entry which maximized YIR is less than the current age of entry (Table 8). Based on the age of entry which maximized the YIR, new levels of sus- tainable yield for each species group as a function of F can be computed as the product of the yield for the current age of entry with the ratio of YIR maximized over age of entry to the YIR for the cur- rent age of entry. The values of F01 and Y01 for the ages of entry which maximize the YIR are 1.0 and 109 t, respectively (Table 9). An approximate confidence interval (C.I.) for this yield estimate can be obtained from a Taylor series expansion which incorporates the variance estimate for catchability (Polovina 1986a) and a sampling variance of the bank CPUE values (Table 2). The standard error of the yield estimate is 14 t, and thus a 95% C.I. for the yield at F01 for the archipelago is 81-137 t annually. The estimation of MSY based on the relative spawning stock approach requires estimates of the age of sexual maturity (tm). A relationship express- ing the length at sexual maturity (Lm) as a fraction of the length of the upper one percentile (Lmax) for tropical bottom fishes is as follows (Anonymous 1977, from Brouard and Grandperrin 1984): Lm = 0.576 Lmax. 1F01 and /01 as defined by Gulland (1983). Table 8.— Current age at entry and age at entry which maximizes the yield per recruit (YIR) at F = 1.0. Current age at entry Age at entry which Species tc (yr) maximizes YIR (yr) Caranx lugubris 1.3 1.75 Pristipomoides filamentosus 4.3 2.75 P. auricilla 3.6 2.25 P. flavipinnis 3.7 2.00 P. zonatus 4.65 3.00 Etelis coruscans 6.2 4.50 E. carbunculus 3.45 2.50 Table 9.— Annual sustainable handline yield in metric tons (t) for the age at entry which maximizes the yield per recruit for each species. Fishing mortality (F) Total yield (t) 0.1 35 0.5 91 11.0 1109 1.5 114 2.0 116 2.5 116 1F01 and y0. as defined by Gulland (1983). 767 FISHERY BULLETIN: VOL. 84, NO. 4 The tm can then be computed from Lm with the von Bertalanffy growth equation. The tm for the seven species, which is assumed to be the same for both sexes of a species, is given in Table 1, and the ratio of spawning stock biomass under exploitation to the unexploited spawning stock biomass is presented for three levels of F (Table 10). As expected, the ratio decreases as F increases. However, without the spawner-recruit relationship, it is difficult to deter- mine the extent that the spawning stock biomass can be reduced before recruitment is substantially affected. It has been suggested that as a lower bound, the spawning stock biomass should not be reduced below 20% of its unexploited level before there is a deleterious reduction in recruitment (Beddington and Cooke 1983). The level of F = 1.0 is the largest level of F which insures that the relative spawning stock biomass for all the species does not fall below 20% and hence the spawning stock approach also estimates the MSY for the bot- tom fish in the Marianas at 109 t/year. Table 10. — The ratio of spawning stock biomass to unexploited spawning stock biomass for three levels of fishing mortality (F) at the age of entry which maximizes the yield per recruit. Species F = 0.5 F = 1.0 F = 2.0 Caranx lugubris 0.44 0.26 0.12 Pristipomoides filamentosus 0.46 0.33 0.25 P. auricilla 0.45 0.29 0.19 P. flavipinnis 0.45 0.26 0.12 P. zonatus 0.39 0.24 0.14 Etelis coruscans 0.31 0.20 0.13 E. carbunculus 0.58 0.42 0.30 DISCUSSION The assessment proposed here is a multispecies approach which is most suitable for resources where prey-predator interactions are negligible. Two assumptions initially required to implement this program, i.e., constant recruitment and that the resource be essentially unexploited, can in some instances be relaxed. Simulation results suggest that if recruitment is seasonal and a pooled length fre- quency is constructed from individual length- frequency samples collected over the year, the length-frequency based method used here to esti- mate mortality produces an essentially unbiased estimate (Ralston4). Furthermore, the assumption that stocks be unexploited can be relaxed if an estimate of the average of F for the archipelago can be obtained. Then M can be estimated by the dif- ference between F and total mortality, and instead of estimating unexploited recruited biomass from the CPUE survey, the biomass under F will be estimated, and yields calculated as the product of exploited biomass with the ratio of yield/biomass resulting from F computed from the Beverton and Holt yield equation. The estimate of maximum equilibrium yield from the Beverton and Holt (1957) equation for the deep slope snappers and groupers from 22 banks in the Mariana Archipelago is 109 1 annually with a fishing mortality of 1.0. About 70% of this yield would be expected to come from the southern islands of the chain, including Guam and Saipan. Another 27% would come from the northern islands and only 3% from the seamounts (Table 11). The mean of the annual sustainable yield levels per nautical mile of 200 m contour for the northern banks, southern banks, and western seamounts are 212.9, 228.5, and 264.4 kg, respectively, with a ratio of total yield for the archipelago to the total length of the 200 m contour of 222.4 kg/nmi (95%) C.I. of 165.3-279.6) (Table 11). Detailed bathymetry data to establish a correspondence between contour length and area are available from Guguan Island in the northern Marianas, and it is estimated that 1 nmi of 200 m isobath corresponds to 0.23 nmi2 of habitat in the 125-275 m depth range (Polovina and Roush5). Based on this correspondence the unit MSY of 222.4 kg/nmi of 200 m contour for the Marianas is equivalent to about 1.0 t/nmi2 or 0.3 t/km2. These values suggest that the Marianas may be slightly less productive for bottom fishes than the Hawaiian Archipelago where a lower bound esti- mate for MSY of 272 kg/nmi of 200 m contour was obtained from a stock production model applied to commercial catch and effort data that did not include the recreational fishing component of snappers and groupers. Also, an estimate of 286 kg/nmi of 200 m contour was derived from an ecosystem model ap- plied to an island system in the Northwestern Hawaiian Islands (Ralston and Polovina 1982; Polovina 1984). The species composition of the catch should depend to some extent on levels of F and tc. As F increases and tc decreases, the contribution of 4Ralston, S. The effect of pooling length-frequency distributions on mortality estimation in seasonally breeding fish populations: A Monte Carlo simulation. Manuscr. in prep. Southwest Fish- eries Center Honolulu Laboratory, National Marine Fisheries Ser- vice, NOAA, Honolulu, HI 96822-2396. 6Polovina, J. J., and R. C. Roush. 1982. Chartlets of selected fishing banks and pinnacles in the Mariana Archipelago. South- west Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv., NOAA, Admin. Rep. H-82-19, 7 p. 768 POLOVINA and RALSTON: MSY FOR DEEP SLOPE FISHES Table 11.— Annual sustainable yield in metric tons (t) and yield in kilograms (kg) per nautical mile (nmi) of 200 m contour for the age at entry which max- imizes the yield per recruit at a level of fishery mortality of F = 1.0. Total yield Yield (kg per nmi of Banks and islands (t/yr) 200 m contour/yr Northern banks and islands Maug 2.7 262 Asuncion 2.1 188 Agrihan 5.6 304 Pagan 7.7 255 Alamagan 2.0 178 Guguan 1.7 179 Sarigan 1.6 194 Anatahan 2.5 144 38-Fathom 0.5 187 Esmeralda 2.9 237 Total 29.3 Mean 213 Southern banks and islands Farallon de Medinilla 16.7 217 Saipan 13.4 254 Tinian 8.8 304 Aguijan 4.2 267 Rota 6.1 192 Guam 17.2 202 Galvez-Santa Rosa 8.6 164 Total 76.0 Mean 229 Western seamounts Bank C 0.9 288 Bank D 1.1 351 Pathfinder 0.9 304 Arakane 0.6 200 Bank A 0.6 180 Total 4.1 Mean 264 Total yield from all banks s: 109 t/yr. Total yield/length of 200 m contour = 222.3 kg/nmi. those species to the catch with the high MIK values, particularly P. flavipinnis andE1. carbunculus tends to increase (Table 12). A form of succession is, therefore, predicted as exploitation proceeds. There are two approximations which have been used to determine MSY which express it as a frac- tion of the unexploited biomass. Gulland's formula estimates MSY as 0.5 MB, where M is the instan- taneous rate of natural mortality and B is the unex- ploited biomass. An approach proposed by Pauly estimates MSY as B 2.3w~02S, where w is the mean of the weight (in grams) at sexual maturity and the asymptotic weight (Gulland 1983; Pauly 1983). A comparison of these two estimators with the values obtained here shows that for four out of seven species the YIB values estimated with the Bever- ton and Holt equation lie between the values ob- tained from the Pauly and Gulland approximations. Table 12.— The percentage of annual sustainable yield by species groups for two ages at entry with two levels of fishing mortality. Percentage of total catch by i weight Age at entry which maximizes yield Current F = 0.10 age of entry F = 1.50 per recruit Species groups F = 0.10 F = 1.0 Caranx lugubris 10.0 5.5 7.3 8.3 Pristipomoides filamentosus 10.3 8.5 9.4 8.2 P. auricilla 8.5 9.4 7.4 7.3 P. flavipinnis 15.6 21.7 24.3 28.7 P. zonatus 28.1 26.7 26.1 23.1 Etelis coruscans 7.6 5.1 7.1 5.0 E. carbunculus 5.9 9.2 4.6 5.9 Others 14.0 13.8 13.8 13.5 769 FISHERY BULLETIN: VOL. 84, NO. 4 For the other three species, the YIB values fall slightly below the Pauly and Gulland approximations for two species and substantially above for the third species. The mean YIB values obtained by the Pauly and Gulland approximations are, moreover, in substantial agreement with the mean value of YIB obtained with the approach proposed here (Table 13). Table 13. — Annual maximum sustainable yield as a fraction of unexploited recruited biomass (YIB J at F = 1 .0 together with 0.5 M and 2.3 w"026. Species groups YIB 0.5 M 2.3 w-026 Caranx lugubris 0.261 0.335 0.252 Pristipomoides filamentosus 0.262 0.270 0.296 P. auricilla 0.306 0.325 0.403 P. flavipinnis 0.680 0.475 0.348 P. zonatus 0.280 0.270 0.363 Etelis coruscans 0.201 0.175 0.226 E. carbunculus 0.375 0.515 0.289 Mean 0.338 0.338 0.311 ACKNOWLEDGMENT This paper is the result of the Resource Assess- ment Investigation of the Mariana Archipelago at the Southwest Fisheries Center Honolulu Labora- tory, NOAA. LITERATURE CITED Anonymous. 1977. Inventaire critique des donnees utilisees pour l'etude de la reproduction. In Rapport du groupe de travail sur la reproduction des especes exploiters du Golfe de Guinee, p. 8-20. ORSTOM-ISRA, Dakar. Bannerot, S. P., and C. B. Austin. 1983. Using frequency distributions of catch per unit effort to measure fish-stock abundance. Trans. Am. Fish. Soc. 112:608-617. Beddington, J. R., and J. G. Cooke. 1983. The potential yield of fish stocks. FAO Fish. Tech. Pap. 242, 47 p. Beverton, R. J. H., and S. J. Holt. 1956. A review of methods for estimating mortality rates in exploited fish populations, with special reference to sources of bias in catch sampling. Rapp. P.-v. Reun. Cons. int. Ex- plor. Mer 140:67-83. 1957. On the dynamics of exploited fish populations. Fish. Invest. Ser. II, vol. 19, Mar. Fish., G. B. Minist. Agric. Fish. Food, 533 p. 1966. Manual of methods for fish stock assessment. Part II. Tables of yield functions. FAO Fish. Tech. Pap. 38 (Rev. 1), 67 p. Brouard, F., and R. Grandperrin. 1984. Les poissons profonds de la pente recificale externe a Vanuatu. Mission ORSTOM, Vanuatu, Notes et Documents D'Oceanographie 11, 131 p. Csirke, J., and J. F. Caddy. 1983. Production modeling using mortality estimates. Can. J. Fish. Aquat. Sci. 40:43-51. Gulland, J. A. 1969. Manual of methods for fish stock assessment. Part 1. Fish population analysis. FAO Fish. Ser. 3, 154 p. 1983. Fish stock assessment. A manual of basic methods. John Wiley & Sons, 223 p. 1984. Advice on target fishery rates. ICLARM Fishbyte 2(1):8-11. Munro, J. L. 1983. A cost-effective data acquisition system for assessment and management of typical multispecies, multi-gear fish- eries. ICLARM Fishbyte 1(1):7-12. Pauly, D. 1983. Some simple methods for the assessment of tropical fish stocks. FAO Fish. Tech. Pap. (234), 52 p. Pauly, D., and G. I. Murphy (editors). 1982. Theory and management of tropical fisheries. ICLARM Conf. Proc 9, 360 p. ICLARM, Manila, Philip- pines, and Division of Fisheries Research, CSIRO, Cronulla, Australia. POLOVINA, J. J. 1984. Model of a coral reef ecosystem. I. The ECOPATH model and its application to French Frigate Shoals. Coral Reefs 3:1-11. 1986a. A variable catchability version of the Leslie model with application to an intensive fishing experiment on a multispecies stock. Fish. Bull, U.S. 84:423-428. 1986b. Variation in catch rates and species composition in handline catches of deepwater snappers and groupers in the Mariana archipelago. Proc. 5th Int. Coral Reef Congr. 1985, Tahiti. Ralston, S. 1986. An intensive fishing experiment for the caridean shrimp, Heterocarpus laevigatas, at Alamagan Island in the Mariana Archipelago Fish. Bull., U.S. 84:927-934. In press. Length-weight regressions and condition indices of lutjanids and other deep slope fishes from the Mariana Archipelago. Micronesica. Ralston, S., and J. J. Polovina. 1982. A multispecies analysis of the commercial deep-sea handline fishery in Hawaii. Fish. Bull., U.S. 80:435-448. RlCKER, W. E. 1975. Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Board Can. 191, 382 P- Saila, S. B., and P. M. Roedel (editors). 1979. Stock assessment for tropical small-scale fisheries. Proceedings of an International Workshop Held September 19-21, 1979, at the University of Rhode Island, Kingston, RI, 198 p. Wetherall, J. A., J. J. Polovina, and S. Ralston. In press. Estimating growth and mortality in steady state fish stocks from length-frequency data. In D. Pauly and G. R. Morgan (editors), Length-based methods in fishery research. The ICLARM/KISR Conference on the Theory and Applica- tion of Length-Based Methods for Stock Assessments, Sicily, Italy, February 1985. Vol. 1. 770 GROWTH, BEHAVIOR, AND SEXUAL MATURATION OF THE MARKET SQUID, LOLIGO OPALESCENS, CULTURED THROUGH THE LIFE CYCLE W. T. Yang, R. F. Hixon, P. E. Turk, M. E. Krejci, W. H. Hulet, and R. T. Hanlon1 ABSTRACT Loligo opalescens, a commercially important species of the eastern Pacific, is the first pelagic cephalopod to be cultured through the entire life cycle. Squid were cultured twice to viable second generation progeny in closed seawater systems using artificial and natural seawater. The reasons for success compared with previous attempts were 1) increased depth in the culture tank, 2) improvements in water conditioning methods, and 3) an increase in availability, density, and species diversity of food organisms. The diet consisted of live zooplankton (predominantly copepods), mysid and palaemonid shrimp, and estuarine fishes. Mean daily group feeding rates of subadults and adults were 14.9% and 18.0% of body weight. Growth was fast, increasing exponentially the first 2 months of the life cycle (8.35% increase in body weight per day) then slowing to a logarithmic rate thereafter (5.6-1.6% increase per day). Growth rings in statoliths corresponded to one per day for the first 65 days. Maximum life span was 235 and 248 days in the two experiments, with a maximum size of 116 mm dorsal mantle length. Viable eggs were pro- duced within 172 and 196 days, respectively. Eggs developed in 30 days at 15°C. Survival through the life cycle was low, with the highest mortality occurring in the first few weeks when squid made the transi- tion from feeding on yolk to active predation on fast-moving plankton. Fin or skin damage and senescence after reproduction accounted for late mortality. The laboratory life cycle of less than a year is compati- ble with existing field data that propose either a 1- or 2-year life cycle, depending upon the season of hatching. Since 1975 we have been studying loliginid squid to develop methods of providing a consistent supply for neuroscience research. These studies include aspects of fishery biology (Rathjen et al. 1979; Hix- on 1980a, b, 1983; Hixon et al. 1980), capture and maintenance methods (Hanlon et al. 1978, 1983; Hulet et al. 1979; Hanlon and Hixon 1983), behavior (Hanlon 1978, 1982), and mass-culture methods (Hanlon et al. 1979; Yang et al. 1980a, b, 1983a, b). Much of the baseline information acquired through these controlled culture experiments will also be im- portant to the fisheries biology of commercially ex- ploited loliginid squids (cf., Roper et al. 1983). About 20 major attempts have been made to culture loliginid squids through the life cycle, but none have been successful (see review in Yang et al. 1980b), even though wild-caught mature females of Loligo and Doryteuthis spawn readily in captiv- ity (Hamabe 1960; Fields 1965; Takeuchi 1969, 1976; Hurley 1977; Arnold et al. 1974; Hanlon et al. 1983). Fields (1965) attempted unsuccessfully to culture Loligo opalescens as early as 1947. Hurley (1976) 'The Marine Biomedical Institute, The University of Texas Medical Branch, 200 University Boulevard, Galveston, TX 77550- 2772. reared L. opalescens for 100 d to a mantle length (ML) of 13 mm. Hanlon et al. (1979) reared this species to 17 mm ML in 79 d and, based upon that work, reared L. opalescens from hatching to sub- adults (Yang et al. 1980b, 1983a). We have now im- proved previous culture methods by increasing the rearing population density and by improving the space requirements for young and adult squid. With a more consistent supply of foods and improvement of water management, we have now successfully cultured this squid twice from egg to second genera- tion, thus closing the life cycle. MATERIALS AND METHODS Two culture experiments are reported herein: L.0. 1981 (full life cycle partly published in Japanese by Yang et al., 1983b); and L.O. 1982 (full life cy- cle). A third experiment, L.O. 1980, was published by Yang et al. (1980b, 1983a) and is referenced for comparison in the Discussion and figures. For L.O. 1981, freshly laid eggs were obtained from wild-caught squid kept in holding tanks at Sea Life Supply (Sand City, CA 93955). Eggs were col- lected from spawning grounds in Monterey Bay, CA for experiment L.O. 1982. Eggs were air-shipped Manuscript accepted February 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 771 FISHERY BULLETIN: VOL. 84, NO. 4 to Galveston (Yang et al., 1980b; 1983a, b). Only early stage eggs were shipped and cultured (never beyond stage 19, Arnold 1965). The eggs were ac- climated gradually to the temperature and salinity of the culture tank water; incubation temperature was maintained around 15°C while salinity ranged between 34 and 36%o. Bundles of a few capsules each were suspended from a rack at the water sur- face to ensure oxygenation and uniform develop- ment of eggs. Styrofoam panels covered the rear- ing tank and the illumination level was kept below 1 lux to prevent the growth of benthic diatoms on egg capsules. A circular tank (CT) system consisting of two cir- cular tanks (each 1,300 L) was used for incubation and early rearing of hatchlings and juvenile squid. Water circulation was modified in L.O. 1982 when compared with earlier culture experiments (Yang et al. 1980b: fig. 1, 1983a: fig 1). Prior to L.O. 1982, a laboratory-constructed particle/carbon filter was used with circulation first passing through an ultra- violet (UV) sterilizer. L.O. 1982 used modular type particle and carbon filters, with the UV sterilizer in the last position in the water conditioning pro- cess. The raceway (RW) system (RW culture tank volume-10,970 L in L.O. 1981, and 13,180 L in L.O. 1982) was used for final grow-out after transferring the squid from the CT culture tanks. The transfer was necessary to give the squid greater horizontal swimming space. The initial RW system in experi- ment L.O. 1981 had been modified from previous experiments (Yang et al. 1980b, 1983a) to improve water quality by 1) adding a rectangular, 960 L capacity water conditioning tank (0.46 x 1.22 x 1.83 m, water depth 0.43 m) with water circulation of 54 L/minute, 2) adding another cooling unit, 3) adding three protein skimmers, 4) adding three UV light sterilizers (each 30 W, total 90 W), 5) modi- fying the water uptake system in the RW with a float near the center to remove near-surface water without sucking up squid or food organisms and to increase the lateral swimming space for the squid, 6) painting an irregular mottled pattern on the sides of the RW to make the walls more visible to the squid, and 7) most importantly, by increasing RW water depth gradually from 24 cm initially to 40 cm (average depth 38.8 cm) to provide swimming space for the squid and to increase the average culture water volume in the RW from 5,990 to 8,610 L. A further improved RW system (Fig. 1) was used in experiment L.O. 1982. It consisted of two bio- logical filter tanks (A, C) with oyster shell subgravel filters and airlifts for water circulation, a tank for growing macroalgae (B), the RW where the squid were cultured (D), and a separate tank where pro- tein skimmers were operated continuously (E). The surface water was taken from the RW through pipes suspended in a screened floating core. Water within the system was recirculated by three routes. First, water was pumped to filter tank A that contained approximately 0.15 m3 of oyster shell over a false bottom. Water passed through the filter bed, then flowed through a constant-level siphon to tank B where algae were illuminated by two 400-W metal halide lamps. Water flowed by gravity into the sec- ond filter tank C that contained 0.18 m3 of oyster shell substrate and two 1-hp cooling units, and final- ly returned by gravity to the RW proper. Second, water was pumped through two sets of six modular filters: four modules containing pleated 20 ^m fiber particle filters and two containing activated carbon. From the modular filters, water either flowed direct- ly into the RW or through a 60 W UV sterilizer before returning to the RW. Third, water was pumped at 36 L/minute to a tank that contained five protein skimmers and then flowed back into the RW. The outflow of the three recirculating routes created a clockwise water flow in the RW proper. This mo- tion accumulated dead squid and food organisms in one place on the bottom. The bottom was painted solid black with nontoxic Thixochlor2 paint and the sides were painted with an irregular mottled pat- tern. Three 11 x 28 cm windows were mounted in one side of the RW for observing the squid's feeding and behavior. The tanks were insulated with poly- styrene sheeting and 2.3 cm thick polystyrene covers. To ensure activation of the biological filter for both CT and RW systems, filter beds were inoculated 2 to 3 wk beforehand with nitrifying bacteria on oyster shell from other systems. Fish and shrimp were placed in the water conditioning tank to build up the bacterial population. Thus the filter beds were estab- lished by organic conditioning methods (Moe 1982) instead of by directly adding ammonia source chemicals. A set of black silk nets was used to transfer squid from the CT system into the RW system. A tri- angular lift net was laid on the bottom of the tank while two rectangular net curtains were slowly drawn from the left and right sides of the tank to concentrate the small squid above the lift net. The lift net was gradually raised, a wash tub placed underneath, and both were moved to the RW tank where the squid were gently released into the tank. 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 772 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS Lamps 1 meter Cooling units 'w///;;////////////////;/////;////////////////////////////?/ Length x Width x Depth (m) A = 1.8x1.2x0.5 B = 1.2 x 0.5 x0.3 C = 1.8 (diameter) x 0.8 D = 6.1 x2.4 xO 9 E = 1.8x0.6x0.6 Protein skimmers Filter module (4 particle + 2 carbon) Ultraviolet sterilizer Figure 1.— Raceway (RW) system (L.O. 1982) with recirculating culture seawater (17,000 L total) for grow-out of juvenile and adult squid. In L.O. 1981, 129 squid were not transferred from the CT tank and they continued to grow in the CT, thereby allowing comparisons of temperature toler- ance and survival in small versus large culture systems. Natural seawater and artificial sea salts (Instant Ocean) dissolved in deionized water were used in CT systems for L.O. 1982 and L.O. 1981, respectively, and artificial seawater was used exclusively in the RW system in both experiments. Salinity was main- tained between 34 and 37°/. Trace elements were supplemented regularly with Wimex Trace Ele- ments. Temperature was maintained at 15°C unless otherwise noted. The pH was maintained between 7.8 and 8.2, and low pH was corrected by the gradual addition of sodium bicarbonate. Temperature and salinity were measured daily, pH every other day, and metabolic waste products (ammonia, nitrite, and nitrate) were measured week- ly. Ammonia-nitrogen levels were determined by the Solorzano method (Strickland and Parsons 1972), and nitrite-nitrogen was determined by the Shinn method (applied to seawater by Bendschneider and Robinson in Strickland and Parsons 1972). Nitrate- nitrogen levels were determined using a prepacked Hach reagent kit. Various live food organisms were fed to the squid several times daily throughout the experiments. Live planktonic organisms such as zooplankton (mainly copepods) and small mysidacean shrimp (Mysidopsis almyra) were the primary foods dur- ing the first 60 d in the CT system. Brine shrimp, Artemia salina; larval red drum, Sciaenops ocellatus; and mysis stage penaeid shrimp were fed as supplemental foods. Food organisms were added to the CT system four or five times daily. Thereafter 773 in the RW system, adult mysids; palaemonid shrimp, Palaemonetes pugio; and a variety of marine or estuarine fishes were fed to the squid at least twice daily. Zooplankton were washed carefully in clean sea- water. Mysids and palaemonid shrimp were treated overnight with quinacrine, while erythromycin and/or tetracycline were used to treat fish (Yang et al. 1980b, 1983a, b). Before feeding, all foods were counted or weighed and slowly acclimated to the temperature and salinity of the cultured water. Dead squid and dead food organisms from previous feedings were removed by siphoning once or twice daily from the CT or RW systems. Daily food consumption in the RW was derived by sub- tracting the weight of uneaten food remains siph- oned each day from the weight of food organisms added daily to each culture system. Daily feeding rate (wet weight) is expressed as the percentage of food consumed by the total estimated biomass of the squid. Daily biomass of squid was estimated by multiplying the number of live squid on a given day by the average weight of an individual squid on that day. Daily squid weight estimates were projected from linear regression of the weights of freshly dead squid against time. All measurements and wet weights (WW) were usually made with freshly dead squid although live squid were occasionally used. Badly damaged or partially cannibalized squid were not measured or weighed for this analysis. The ini- tial squid population was derived from the number of dead or sacrificed specimens removed from the culture systems. Overhead fluorescent lights provided illumination. In the CT systems for both experiments there was constant light that measured 11 to 15 lux in the mid- dle of the water column. In the RW systems there was also constant light although light only filtered in through plastic-covered holes in the polystyrene tops. In L.O. 1981 it measured 17 lux in the center of the RW and 0.5 to 0.7 lux at each end. In L.O. 1982 it measured 4 to 7 lux near the ends under the opaque top and 11 lux near the center where light passed through the clear plastic. Statoliths from hatchlings of known age in L.O. 1982 were dissected from the squid and decalcified in a 1:1 mixture of 4% EDTA in distilled water and 0.2 H sodium cacodylate buffer (pH 7.4). Decalcifica- tion facilitated the counting of rings in statoliths from squid age 65 d or younger, but older statoliths were distorted by the process. The rings were counted from photographs taken with a Leitz Com- biphot II and Kodak copy film #4125. FISHERY BULLETIN: VOL. 84, NO. 4 RESULTS Water Quality There were no obvious differences in growth or survival between squid cultured in artificial sea- water (L.O. 1981) and filtered natural seawater (L.O. 1982). Water quality in the CT systems was maintained in very good condition due to the short culture period, while water quality in the RW system was more difficult to maintain because of the long grow-out period and the greater biomass of squid and food organisms. In L.O. 1981 (Fig. 2) from days 180 to 190 the estimated total biomass reached the maximum peak of 1,706 g (cf., Fig. 7), which is equivalent to 155 g/m3 of rearing water volume. After the 160th day, food organism biomass in- creased to between 300 and 400 g/day. As a result, the amount of nitrate-nitrogen gradually accu- mulated to over 23.0 mg/L during the period from day 180 to day 193 (Fig. 2). On day 164, 1,900 L (17% of total volume) of fresh Instant Ocean was replaced in this system. However, the nitrate- nitrogen level did not drop in proportion to the per- cent water change. Concurrently, pH dropped to 7.75 by day 169 and dissolved sodium bicarbonate (Atz 1964; Bower et al. 1981) was introduced to the system to adjust the pH above 7.9. The sodium bicar- bonate solution required very strong aeration to be effective when it was put into the culture water. A similar trend of slightly increased nitrate-nitrogen and decreased pH occurred (about day 200) in L.O. 1982 (Fig. 2). This was corrected in the same manner. The vegetative macroalgae, Gracilaria tikvahiae, was cultured in the water conditioning tank of the RW system in L.O. 1982 to remove ammonia and prevent the accumulation of nitrate-nitrogen, but its effectiveness was not clear. Incubation and Hatching of Eggs Average hatchling size in both experiments was 2.7 mm ML (range 2.3-2.8 mm ML) with a hatching success of over 90%. In L.O. 1981, hatching began on 14 October and lasted until 17 October. Embry- onic development required 27 to 30 d at 15 °C. The hatching period lasted 4 d, compared with L.O. 1982 that took 5 to 6 d. The period of embryonic develop- ment in L.O. 1982 was not precisely known because the eggs were collected in nature. Development of eggs within the same egg cluster was different de- pending upon the capsule position within the cluster. Moreover, hatching time within the same capsule 774 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS differed, since distal embryos usually hatched first. Since we used early stage eggs removed from their habitat in California, no polychaete worms (Capitella ovincola) were observed in the egg capsules (cf., McGowan 1954), although we had observed worms in other late stage California egg capsules. During embryonic development, granules or crystals appeared in the perivitelline fluid of some eggs, but no significance to survival or development could be associated with this condition. The outer tunics of the egg capsules incubated in Instant Ocean were more elastic until the later stages (around stage 27) than those incubated in natural seawater. More bacteria and other benthic or- ganisms grow on the capsules incubated in natural seawater. These differences did not influence devel- opment or hatching success. Embryos near hatch- ing (stage 29) generally moved little or were nearly static, but in most individuals the external yolk sac was already broken off within the egg. External yolk sacs were observed on a few hatchlings. In L.O. 1981, bright illumination stimulated hatching in very late stage eggs and therefore light levels were in- creased during later stages of egg development in L.O. 1982. Foods and Feeding The species and size of food organisms were similar in the two experiments. The general progres- sion of food types began with zooplankton, then mysidean shrimps, then palaemonid shrimp larvae and adults, and finally fishes (Fig. 3). The use of brine shrimp has been curtailed since they were found to be unattractive to the hatchlings. The size range of food organisms fed in the first 30 d is large, especially when compared with the size of 1-d-old hatchlings (2.3-2.8 mm ML, Fig. 4). How- ever, as shown in Figure 4, the hatchlings have only small fins and are not strong swimmers; therefore, feeding on active prey at this stage is not excellent. A summary of the types and quantities of food of- fered in the experiments (L.O. 1981 is used as an example) is given in Figure 5. Large amounts of food were available to the squid; this was important dur- ing the first weeks when hatchlings could only cap- ture food organisms drifting very close to them. The relationship of hatchling to food organism density during the first 59-d period in each experiment is summarized in Table 1. Unfortunately there was no clear relationship between densities and survival. For example, in L.O. 1982, there were twice as many food organisms per squid as in L.O. 1981, but survival (cf., Fig. 13) was not better. Figure 5 shows more specifically the number of food organisms fed daily in L.O. 1981. The early rearing period in L.O. 1981 and 1982 coincided with the spawning of mysid shrimp in the Galveston estuaries. Therefore, small mysids with a total length of about 2.0 mm (Fig. 4B) were abun- dantly supplied. This was particularly important since small mysids swim more frequently in the water column than do adults. Young mysid hatch- lings were given as food by day 12 in L.O. 1981 and immediately in L.O. 1982 (Fig. 3). Small mysids distribute themselves more evenly in the culture tanks and are easier for hatchlings to capture. Palaemonetes spp. were fed to juvenile and adult squid (Fig. 3). Shrimp ranged in size from 2.0 to 25.0 mm. They were graded by size and fed based on size and availability. Daily siphoned remains indicated that only the abdominal flesh was consumed, with the thorax and carapace discarded. Fish were generally used for juvenile or older squid. However, fertilized red drum eggs were available in L.O. 1981, and larvae up to 13-d old (Fig. 4E) were given to the hatchlings. In the two ex- periments, a total of over 14 fish species of 10 families were fed (Table 2). To determine the diet preference for fish species, the actual consumption of fish (i.e., total weight of fish put in tank minus total weight of fish remains) was compared for a total of 5 kg fish fed in L.O. 1982 (Fig. 6). The cyprinodont fish were most preferred (consumption of 83%). Only small Fundulus spp., smaller than 31 mm (Cyprinidontidae), were fed because large Fun- Table 1.— The mean density of squid and food organisms per liter of culture water from days 0-30 and 30-59. Initial hatchling population Day 0-30 Day 30-59 Exp. No. Squid No./L Food organisms No./L Ratio of food organisms to squid Squid No./L Food organisms No./L Ratio of food organisms to squid L.O. 1980 L.O. 1981 LO. 1982 864 2,061 1,704 0.46 0.93 0.27 14.2 24.0 14.6 30:1 25:1 54:1 0.35 0.54 0.14 9.4 12.4 5.6 26:1 23:1 40:1 775 FISHERY BULLETIN: VOL. 84, NO. 4 LU DC LU CL X LU 00 i ( I I L _L ■■ J_ O) 00 P^ CO LO *t oo r» co m o YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS LU or LU Q_ X LU CN 00 CC 5 :i cj I I I I Lf_l L o o) co r* co in ^- CM »-»-«-»- f- «- o I I I ' '■ o a I I L_ O 00 CO ^f CN OCN«- t-OOOO OOO J I I L o ■in CM a D o 0 o o a a o o o CN a a a D o o a o o ■m D 0 o a a c G a o° o a a >- -8 m ' 1 I L mom CN CN i- o m CN r- p O) oo oo oo oo r** p* \ i EX \ i EO \ i O) co EO o PS o> s CD — _C o _CD ■s "to" Oi s CD ft S £ El 0) ™ +-> *- CO 0) >> ft £ E Eh Oi 5 ~ oi E -C o |.S to cd £>•* .-^ 0) c? g) •S > 8 -H TO 25 — T. s — ' ^H 1° to .5 I* =S a" *^ CO u a o> 3 S u o .fa 03 CN DO co "O co CD V Sd 1 I ^S cs a> CM o i? FISHERY BULLETIN: VOL. 84, NO. 4 >- > _l _l 1- CO z Z> LU o t= z > t- CL 7 LU o o z Q Q LU LU o I- z < _l CL O O ISI DC X CO LU z CD X CO CO < > DC < CO o Q_ LU < _l < Q_ O r LO CM CO LU X CO o o CM o LO CO >- < O O O LO c 0) o> u 3 0) fcuo §■ o "o -J -o -1> c ho t- o T3 O to 3 0) > IS s o o I Ed OS O O *- CM O «~ CN CO GO 00 CO CO CO CD CD CO CD CD CD O i- CM O «- CM O «- a. a; s- 0) c s-. > 00 o J c 01 c C3 c s- o T3 O 03 0> c T3 e o> I Eh I 63 D O SIAISINVOdO QOOd dO S3dAl 69 01 0 AVa lAIOdd U31I1 U3d Q3d SIAISINVOHO QOOd dO UdaiAIDN 780 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS Table 2. — Fish species and size range given as food in all three experiments. Size (TL in mm) L.O. experiment Family and species 1980 1981 1982 Family: Clupeidae Brevoortia spp. 15.0-31.0 — — X Family: Engraulididae Anchoa mitchilli (Valenciennes) 20.0-25.0 — — X Family: Cyprinodontidae Adinia xenica — — — X Cyprinodon variegatus Lacepede 10.0-28.0 X X X Fundulus spp. 15.0-31.0 X X X Family: Poeciliidae Gambusia affinis (Baird and Girard) 12.0-28.0 X X X Poecilia latipinna (Lesueur) 22.0-41.0 X X X Family: Atherinidae Menidia beryllina (Cope) 18.0-52.0 X X X Family: Carangidae Hemicaranx amblyrhynchus (Cuvier) — — — — Family: Gerreidae Eucinostomus gula (Quoy and — — X X Gaimard) Family: Sparidae Lagodon rhomboides (Linnaeus) — — X X Family: Sciaenidae1 Sciaenops ocellatus (Linnaeus) 1.5-14.5 — X X Pogonias cromis (Linnaeus) 10.0-15.0 — X X Family: Mugilidae Mugil spp. 18.0-38.0 X X X 'There were about six more species of Sciaenidae: minority species were not identified. -™ FISH MEAT ^^ CONSUMED □ UNEATEN REMAINS < o z j CYPRINODONTIDAE^';< ^?^,^^->^w;r :v iCLUPEIDAE==c 0 10 20 30 40 50 60 70 80 90 100 PERCENT Figure 6.— Food preference for fishes by squid in experiment L.O. 1982. Total fish weight fed to the squid was 5.0 kg. initiation of spawning; biomass then decreased because of the mortality accompanying spawning. Squid in L.O. 1982 were fed ad libitum and daily group feeding rates could not be determined. How- ever, the average group feeding rate calculated weekly for L.O. 1982 allowed an estimate of 18.0% for the daily feeding rate. 781 FISHERY BULLETIN: VOL. 84, NO. 4 3 5? "o o 3 t; en Q. E O W (/) C (/) o co o E -o o o >■ >• CD CO Q Q DAILY GROUP FEEDING RATE A FOOD WEIGHT x 1 00 m SQUID WEIGHT o CO o CM 2 o O CM 1 r „*- O — J I I I I L _I_SL o o CM CO < O ID 00 O lO (z0 1*) SIAIVH9 Nl 1H9I3AA 13AA 00 O J c £ ■c oo 00 o I— I T3 o to I J3 -2 e o o 1 -a c c3 1 ■i he a $ ! 6b •a m to C < I £ D O 782 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS Growth Figure 8 illustrates growth data through the life cycle for both experiments. At hatching, Loligo opalescens has a mean mantle length of 2.7 mm, a wet weight of 0.001 g, and has approximately 100 chromatophores on its body. In L.O. 1981, the largest reared squid was a male of 113 mm ML and 58 g. In L.O. 1982, the largest reared squid was a female of 116 mm ML and 63 g. Mean sizes for adults from the two experiments were 87 mm ML (Sx = 2.7) and 23.8 g (Sx = 1.9) for 35 males, and 83 mm ML (Sx = 1.9) and 21.2 g (Sx = 1.5) for 58 females. Growth equations for the squid in L.O. 1981 clear- ly describe two separate phases of growth. The man- tle length of squid cultured in the CT system (days 1-56) increased at an exponential rate (ML = 2.121 eo.o2398«. r2 _ o 92) or 2.4% increase per day, while those cultured in the RW system (days 56-248) grew logarithmically (ML = 0.2884 t1A95; r2 = 0.97). Weights were only measured on squid from the RW (23) 60 50 CT X 40 O LU LU 30 20 10 0 L O 1981 c )- n - Range *T~' Standard Deviation Mean (3 (1 7) 5) (22) (7) L.O 1982 (8) (20) 12) T i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — | 50 100 150 200 250 DAYS "i i i i i i i i i i — i — i — r — i — i — i — i — i — i — | — i — i — n — | 50 100 150 200 250 " L.O. 1981 (51) C )— n E E, 100 - p — Range _ -I^> Standard Deviation Mean (8 5) 1 | (20) \ ' 1 o (42) J 11 z i LU s _l t LU 50 — . 1 _l (99) k r- I I 1 z (129) 1 I < ^ T (90) i 0 -(258)1 I 1TTFI[1IT1ITTTTTTT t — r t i i i i i L.O. 1982 (62) (18) J :18) (30) (30) (13) (20) ;io> i i — i — i — i — | — n — i — i — | — i — i — i — i — | — n — i — i — | — n — i — i — | 50 100 150 200 250 50 100 150 200 250 DAYS Figure 8.— Comparison of growth (wet body weight and mantle length) in experiments L.O. 1981 and L.O. 1982. 783 FISHERY BULLETIN: VOL. 84, NO. 4 system (days 108-248) and the growth curve showed a logarithmic increase (W = 6.283 x 10"7 tSM0; r2 = 0.92). Hence, younger squid grew at an exponen- tial rate and growth slowed to a logarithmic rate in older squid. Squid exhibited fast exponential growth for the first 2 mo in L.O. 1982 and slower logarithmic growth thereafter (Fig. 9). Wet weight data from live animals in L.O. 1982 indicated a mean growth rate of 8.35% increase in body weight per day for the first 2 mo. Mantle length increased 3.19%/day or the equivalent of 8 mm/month. The squid were doubling their weight every 8 d and doubling their length every 21 d. Growth rates declined from 5.6%/day WW at day 60 (and 2.2%/day mm ML) to 1.6%/day WW (and 0.63%/day mm ML) at day 240. B .25 3 20 X a .15 UJ uJ .10 .05 L.O. 1 982 Live Loligo opalescens W = 0.0023e° °835 ' r2 = 0 98 V/-I— 20 30 40 DAYS 50 60 50 40 H 30 X CD UJ UJ 20 10 L.O. 1982 Dead Loligo opalescens W = 0.3894 x 10"7 t3827; r2 = 0.98 E £ x h- o z UJ _J UJ < 2 15 - L= 2.73 10 L.O. 1 982 Live Loligo opalescens ,00319 1 V 20 — i — 30 40 DAYS — i — 50 60 E E X \- o z UJ < 2 100 50 10 L0. 1982 Dead Loligo opalescens *# i — 50 100 150 DAYS 200 235 Figure 9.— Early exponential growth of Loligo opalescens in experiment L.O. 1982: A, Live wet weight, illustrating exponential growth through day 60. B, Dead wet weight, illustrating logarithmic growth from day 60 to maturity. C, Live mantle length measurements, showing exponential growth as in A. D, Dead mantle length measurements, showing logarithmic growth to maturity as in B. Numbers above rectangles indicate actual number of squid measured for that mean. 784 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS Mean growth was 16 mm/month for this period. Doubling times for weight increased from 12 d at day 60 to 42 d at day 240, and for length from 31 d at day 60 to 109 d at day 240. The length-weight relationships of squid in L.O. 1981 and 1982 are illustrated in Figure 10 and are compared with data on wild squid (Fields 1965). The slopes of the curves are slightly higher in laboratory- reared animals, indicating that these squid are heavier per unit length than wild squid. Table 3 illus- trates differences in predicted weights for repre- sentative mantle lengths from L.O. 1982 data versus Fields' (1965) data. The length-weight relationship for males vs. females in L.O. 1981 is shown in Figure 11; no significant differences between sexes were detected (P > 0.05). Statoliths from 55 early hatchlings (L.O. 1982) aged 21 to 79 d (± 5 d) were examined to correlate statolith ring numbers with the age of individual 70 •3 LU LU 60 50 40 30 20 10 •••••• L.O. 1981 W = 0.0002 L2-56 r2 = 0.96 n= 104 — L.O. 1982 W = 0.0002 L2-60 r2 = 0.98 n = 81 1 — Wild Loligo opalescens (Fields, 1965) W= 0.0013 L2-15 0 — i — i — i — i — i — i — i — i — i — i — i — i 0 20 40 60 80 100 120 MANTLE LENGTH (mm) Figure 10.— Comparison of length-weight relationship of squid cultured in L.O. 1981 and 1982, and squid collected in the field at Monterey Bay, CA by Fields (1965). Table 3.— Examples of length-weight differences between L.O. 1982 and the data of Fields (1965). Reference Figure 10. ML = mantle length; WW = wet weight. L.O. 1982: Fields (1965) ML (mm) WW (g) WW(g) 25 50 75 100 125 0.86 1.31 5.22 5.80 15.00 13.90 31.70 25.90 56.60 41.90 squid (Fig. 12). The linear relationship between the number of rings (R) and the age in days (D) for 43 statoliths aged 21 to 65 d was R = -7.24 + 1.13 D, with an r2 value of 0.90. Counts of rings dif- fered from the actual age by an average of ±4.2 d (range -12 to +8 d). Survival Figure 13 compares survival in the two experi- ments. The longest lived squid were 248 d in L.O. 1981 and 235 d in L.O. 1982. Survival dropped below 50% on day 15 in L.O. 1981 and on day 2 in L.O. 1982. In L.O. 1982, the early rapid population reduc- tion was due to the removal of newly hatched squid for a different experiment. Mortality rates slowed after the early heavy population reduction; 10% sur- vival occurred on day 120 in L.O. 1981 and on day 49 in L.O. 1982. In all cases, mortality gradually slowed after 60- to 70-d posthatching. Survival reduction after day 180 in both experiments was considered to be related to spawning (Figs. 13A, B). In L.O. 1981 experiment (Fig. 13A), 50% survival of 391 squid transferred to the large RW system oc- curred at day 114, but at day 84 for the 129 squid left in the same small CT system. For example, 10 d after transfer the squid in the CT system had 30% mortality whereas those in the RW system experi- enced only 20% mortality. Thus, transferring squid at about 60 d gave better results by reducing the mortality from fin and skin damage that accrues in the smaller CT system. In the middle of L.O. 1981 (day 108) cannibalism was observed. The fins and/or posterior mantle were clearly eaten in some squid; these squid differed from those that died from fin damage or from scraping on the bottom of the tank since the latter developed lesions near the tip of the mantle (Fig. 14). From days 108 to 206 there were 16 partly eaten squid in the RW system (7% of the popula- tion on day 108), compared with two squid (of 10 total) in the CT system between days 157 and 172. Slightly higher levels of cannibalism (19% between days 97 and 191) were observed in L.O. 1982. 785 FISHERY BULLETIN: VOL. 84, NO. 4 Figure 11.— Length-weight relationship of males versus females in L.O. 1981, compared with the data of Fields (1965). 35 30 25 20 s 15 h- X CD LLI LU 10 / Wild Loligo opalescens (Fields. 1965) / W=.0013L215 1 Hi ob. h 1- 0 33 40 L.O. 1981 Laboratory reared female Loligo opalescens W = .0007 L2 31 r2 = 0.88 n = 40 Laboratory reared male Loligo opalescens W = .0002 L2 59 r2 = 0.96 n = 18 "60~ 80 1 00 1 20 MANTLE LENGTH (mm) Figure 12.— Increase in statolith rings with age (L.O. 1982). Closed circles represent ring counts of 55 statoliths from L. opalescens of known-age (21-79 d). Each statolith was counted twice from different photographs. Unclear exposures (16) were not counted. The solid line represents a linear relationship of age and ring numbers. The space between the solid and dashed lines reflects the 5 d of hatching. 80 r 70 (/) CD -z. 60 cc O 50 DC III CQ 40 30- 20 20 30 40 50 60 70 80 ACTUAL AGE (DAYS) 786 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS NUMBER OF EGG CAPSULES SPAWNED PER DAY o o o o o o O 00 (O 4 N o o o o o o O 00 (O ■ < Q T3 C- 01 CO -i-' o ~° -a 3 * 2 s s « o> £ -a . '3 co ^ i- — 3 >> o o ft o> s- a> 3 « £ he t-i a, » bo c '2 c3 en Q, 5?^ a> 4d a -C o eg oi c Ol o aj *- ti o> as ft "O o ^ cu en j3 'E a s,§ O 0) » B" a) -d 3.2 ^ 3 >. ,£ u 3 -O § us *> s „ P « .2 .5 6 £ ° lO JO >> to tu i^q c S o> 2 tn Oi £ ii _e -n en _, ft ft 3 X ni eS 01 So § "S IK 2 tS 01 CI O) ft o faS bo 03 EJ C tn ■g ° o ft o — £ c o> o> 3 3 o S> 01 .5 u ft.S c .01 CQ 01 .H 3 01 a -° 3 2 w 3 0) c 53 o> '3 ° ~ 01 .-I Ol C "3 — i «? h CO e» 3 o ^1 CO C 01 u 00 133 -o =4-1 O C o ft. 2 5 ® 3 -C £ tn cij _cS to2 q ca £ o O I. co _c X. 0J ■- D co O * r-* 3 — CO T3 S 01 ^. oi m <; co C3 £ oi 01 w S-, C3 01 3 ^ 3 i> a C en Oj cS § > Ol 01 ft -*-> Ol ol Ol *- tn .S Oi > « 5 +j o 2 S tn J 5? jo f_, 787 FISHERY BULLETIN: VOL. 84, NO. 4 Other causes of injury and death in the later part of RW culture were 1) swimming into the water in- take pipes, 2) jetting out of the water and hitting the bottom of the polystyrene tank covers, and 3) colliding occasionally with the walls and slowly accruing fin damage. The resulting abrasions on the body and fins (Fig. 14) were probably the main fac- tor influencing mortality after about 60 d of cul- ture. 1 or 2 d. They usually had some obvious skin damage and were probably unable to maintain disciplined swimming with the school. Sexual Maturation, Mating, and Egg Laying In L.0. 1981, the first signs of sexual maturation were when chromatophore patterns associated with Figure 14.— Fin and skin damage that resulted in mortality of cultured squid. A, Epidermis and dermis missing on periphery of fins, with fin margin thickened from scar tissue. B, More extreme case with damage extended to mantle. C, Excessive skin damage on ventral mantle caused by scrapping the tank bottom. A hole (arrow) was produced in the mantle wall and prevented jet-propulsed swimming. Schooling Behavior The squid were able to hold a swimming position in the tank between days 41 and 44 in both L.0. 1981 and 1982, corresponding to a mantle length of about 10 mm. In the early phase of RW culture in L.O. 1981 and 1982, squid swam in two or three loose groups throughout the RW. Later, they schooled together at both ends. The reasons for this behavior are unknown, but it may have been related to lower illumination levels at the RW ends or to the well- aerated seawater entering the RW at these points. Individuals not schooling were often found dead in courting were observed in males. On day 174 two males showed the "Shaded testis" component of pat- terning similar to that described in Loligo plei (Hanlon 1982). Later, other chromatic components of patterns seen in mature males of Loligo plei were observed: faint, lateral stripes on the mantle ("Lateral flame"); a discontinuous suture line along the fin margin ("Stitch work fins"); a clear area in the dorsal portion of the mantle above the testis ("Accentuated testis"). Maturation and spawning occurred earlier in L.O. 1982 than L.O. 1981 (Fig. 13). The penis was first recognizable in a 100-d-old male (25 mm ML) and 788 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS the nidamental gland was observed in a 101-d-old female (23 mm ML) in L.0. 1981. The penis was first recognizable in a 93-d-old male (29 mm ML, 1.15 g WW), and the nidamental gland was observed in a 92-d-old female (33 mm ML, 1.71 g WW) in L.0. 1982. Figure 15 shows that females become mature at approximately 60 mm ML. This maturation in- dex is based upon reports by Hixon (1980a) and Macy (1982) in which the ratio of nidamental gland length to mantle length is >0.20. Squid this size could produce fully formed egg capsules. The smallest male with spermatophores was 71 mm ML in L.O. 1982. In L.O. 1981, first mating activity was observed on day 193. A pair was swimming together, a sec- ond male interrupted, and a third male grasped the female in the midmantle area but she jetted away. On day 197 another pair was swimming together at the end of the RW and a brief head-to-head mating was observed. They separated for about 1 min, then 40 r ®® ® E E 30 I H (J -z. LU _l Q < _l o _l < r- LU < Q 20 10 ® ® ® ® ® ® © ® ®® ® ® © ® ® ® ® ® ® *-» I I I I 1 1 1 1 1 1 1 1 0 20 40 60 80 100 120 MANTLE LENGTH (mm) Figure 15.— Maturation index for females from pooled data of L.O. 1981 and 1982. Dots with circles indicate sexually mature females in which the ratio of nidamental gland length to mantle length is >0.20. See text. the male grabbed the female by the arms for a sec- ond time. Drew (1911) illustrated this copulating position in Loligo pealei. A second mating position was observed on day 226. A male grasped a female's mantle one-third of the way from the posterior tip of the mantle, then he gradually moved to the female's head near the mantle opening and then let go. This was done very swiftly and it was impossible to see if a spermatophore was passed. Freshly dead females had spermatophores attached around the sperm receptacle below the mouth after head-to- head matings. Mating activity was not as closely monitored in L.O. 1982, but first observations were several weeks earlier (before day 175). Spawning started on day 196 and lasted till day 239 in L.O. 1981. Of 151 egg capsules, 24 (16%) were unfertilized (Table 4). Squid kept in the CT system (3°C higher temperature from day 125) spawned first on day 185, 11 d earlier than the RW system, but none were fertile. Spawning occurred earlier in L.O. 1982, beginning day 175 and ending day 222. All of the 199 spawned capsules were infertile. The maximum number of spawned capsules in a single day was 27 on day 203 (Fig. 13B). Most eggs were collected in the morning indicating that spawning Table 4. —Spawning date and n umber of egg capsules spawned in the raceway system (L.O. 1981). Capsules Month/ Age/ Capsules without day day spawned eggs 05/01 196 3 0 05/03 198 2 0 05/04 199 5 0 05/05 200 6 0 05/06 201 1 0 05/08 203 1 0 05/09 204 6 0 05/10 205 5 0 05/16 211 5 1 05/17 212 8 0 05/19 214 5 2 05/20 215 3 3 05/21 216 17 8 05/22 217 17 2 05/23 218 14 0 05/24 219 7 0 05/25 220 3 0 05/26 221 6 0 05/27 222 3 0 05/28 223 7 2 05/29 224 6 1 05/30 225 1 0 05/31 226 2 1 06/02 228 5 2 06/03 229 4 2 06/11 237 5 0 06/12 238 4 0 Total 151 24(16%) 789 FISHERY BULLETIN: VOL. 84, NO. 4 took place mainly at night, but some individuals spawned during the day. Egg capsules in the early portion of the spawning period were small, with a length of 2.2 to 4.7 cm when laid. Superficially there were no differences with normal capsules, but usual- ly the early ones contained only a few eggs while a few had none. Typical newly laid egg capsules were between 6.0 and 9.0 cm and contained an average of 156 eggs (range 107-199). These egg cap- sules were normal in length and egg number com- pared with L. opalescens in nature (Hixon 1983). A large number of typical egg capsules were in- cubated and a normal second generation hatched. The average mantle length of second generation hatchlings was 2.3 mm ML (range 1.9-2.7 mm ML, n = 13). This was smaller compared with first generation hatchlings (average 2.7 mm ML) but there was no difficulty in rearing them on copepods for 10 d. Since initial survival was confirmed, fur- ther rearing ceased. In L.O. 1982, three patches of artificial egg cap- sules made of silicon glue were placed on the bot- tom of the RW tank to stimulate spawning. The squid spawned 15 fertilized egg capsules around the artificial capsules (Fig. 16). DISCUSSION Water Quality and System Design Water quality was consistently good throughout both experiments and was probably a major con- tributor to culture success. The CT systems were particularly clean (Fig. 2) because the water volume was relatively large for the small biomass of animals. In the large RW system, water quality changed only slightly when the biomass of squid and food or- ganisms reached its maximum from approximately days 150 to 220 (Figs. 2, 5, 7). The highest total biomass level was 1,706 g between days 180 and 190 in L.O. 1981, which is equivalent to approximately 155 g/m3 of water. At this point, the nitrate- nitrogen level reached 23 mg/L, which is still low [Spotte (1979a) gave a conservative safe level of 20 mg/L for most marine organisms]. Ammonia-nitro- gen and nitrite-nitrogen levels always stayed below the recommended safe level of 0.1 mg/L (Spotte 1979a) in both experiments. We know from our re- cent unpublished data that a drop in pH (which ac- companies nitrogen level increase; Hirayama 1966) is more dangerous to squid; therefore, addition of sodium bicarbonate was necessary to keep the pH near 8.0. Several improvements in system design helped improve water quality over our L. opalescens experiment in 1980 (Yang et al. 1983a), when nitrite- nitrogen reached 1.22 mg/L and nitrate-nitrogen reached 39.20 mg/L. These included increased culture water depth and volume in the RW (5,990 to 8,610 L), increased number of protein skimmers from 2 to 5 and generally more oyster shell substrate area for increased biological filtration. Furthermore, regular addition of trace metals assured high levels since losses occur through foam fractionation in pro- tein skimmers (Spotte 1979b) and metabolism of filter bed bacteria, squid, and food organisms. Growth and Survival Growth in L. opalescens is very fast (Figs. 8, 9) and conforms to a general trend among cephalopods in which the early life cycle is characterized by rapid exponential growth, followed by slower logarithmic growth until reproduction and death (Boyle 1983; Forsythe and Van Heukelem in press). Egg development is temperature-dependent and takes 19 to 25 d at 16.5°C (Fields 1965), 27 to 30 d at 15°C (L.O. 1981, this report) and 30 to 35 d at 13.6°C (McGowan 1954). Hatching success was high, and young squid survived several days on in- ternal yolk. Many squid will feed before internal yolk is absorbed (Boletzky 1975). The young will feed on a variety and wide size range of crustaceans and fishes (Fig. 4). Zooplankton, but especially copepods, are readily attacked and eaten by very young squid. It is noteworthy that relatively large mysids could be fed successfully to hatchlings within the first week (Fig. 3: L.O. 1982) and for 3 to 4 mo there- after as a primary food. Mysids are easier to col- lect and acclimate to laboratory conditions and are thus attractive to the culturist for pragmatic reasons. Loligo opalescens hatchlings (2.3-2.8 mm ML) are much larger than those of L. pealei (1.7 mm ML) or L. plei (1.6 mm ML) (McConathy et al. 1980) and are consequently easier to rear because larger food organisms can be used immediately. Larval fish were attractive to young squid but are difficult to provide. Major mortality occurred within 10 d posthatch- ing. Although high food densities and variety were provided (Tables 1,2; Figs. 3, 4, 5), many squid ap- peared to have difficulty making the transition from passive yolk absorption to active feeding on live organisms. A learning process may be involved, because capturing copepods was initially difficult (squid have been observed to miss 40 times con- secutively) and improved when squid attacked from behind. Past experience (cf., Yang et al., 1983a) sug- gested that increasing food abundance relative to 790 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS Figure 16.— Fertilized egg capsules laid at the base of artificial silicon egg capsules (erected). 791 FISHERY BULLETIN: VOL. 84, NO. 4 squid abundance would enhance survival, but no change has been observed. Further experimentation is required, but a central question is whether many squid are genetically unfit to survive or whether we have not yet provided the proper foods and environ- ment for good survival. Although the former pros- pect seems unlikely from the evolutionary viewpoint, our experimental design has certainly promoted outstanding growth in surviving squid. With the growth data from live squid in L.0. 1982, we confirmed that squid grow exponentially both by weight and length during the first 2 months (Fig. 9). Weight increases at a rate of 8.35% body weight/day (doubling their weight every 8 d) and this compares very favorably with octopods (4-7%), other squid (5-7%), and cuttlefishes (5-12%) (Forsythe and Van Heukelem in press). Logarithmic growth dur- ing the rest of the life cycle also conforms general- ly to other cephalopods, except that some cepha- lopods have a longer exponential growth period up to one-half their life cycle (Forsythe and Van Heukelem in press). The length-weight relationship (Figs. 10, 11) generally conforms to those of wild- caught squid, but indicates that laboratory-reared squid weigh more per unit length (Table 3), possibly, as a result of reduced swimming. The slopes of the lines (all <3.0) indicate allometric growth (Forsythe and Van Heukelem in press). The estimated feeding rates of 18.0% body weight/day (days 121-176) in L.O. 1982 and 14.9% (days 108-230) in L.O. 1981 compare well with the estimate of 14.4% (on a dry weight basis) for L. opalescens of a similar size in the natural population (Karpov and Cailliet 1978). Younger L. opalescens (48-56 d) fed on Artemia were estimated by Hurley (1976) to feed at rates of 36 to 80%/day (dry weight). Another loliginid squid, Sepioteuthis sepioidea, had feeding rates of 20 to 25% (wet weight) between days 70 and 105 (La Roe 1971). Other squids of similar size show comparable rates: Loligo plei, 10 to 18% (Hanlon et al. 1983); L. pealei, ca. 11% (Macy 1980); Illex illecebrosits, ca. 10% (Hirtle et al. 1981); and Todarodes pacificus, ca. 24% (Soichi 1976). Maximal survival and size in our three major ex- periments were L.O. 1980 - 233 d, 77 mm ML (Yang et al. 1983a); L.O. 1981 - 248 d, 113 mm ML; L.O. 1982 - 235 d, 116 mm ML. Figure 13 illustrates sur- vival throughout these experiments and shows that there was a long, steady mortality after the initial high mortality of the first 2 wk. Once in the RW systems (i.e., after 2 mo) most mortality was attrib- uted to fin and skin damage (Hulet et al. 1979; Fig. 14) that accrued slowly from colliding with the sides of the tank. The painted designs on the walls were clearly helpful in reducing wall collisions but damage over time was lethal in many squid. Cannibalism ac- counted for a minor number of deaths (ca. 7-10%) Most mortality after day 170 in L.O. 1981 and L.O. 1982 was due to 1) sexual maturation and spawn- ing and 2) an unusual situation where fully mature females scraped the bottom of the tank often enough to wear a large lesion through the ventral mantle (Fig. 14C). It should be noted that survival rate was greater where large tanks such as the RW were used. In L.O. 1981 (Fig. 13A), 50% survival of squid left in the smaller CT system occurred only on day 84 compared with day 114 for those transferred to the RW. In summary, growth was excellent, indicating that estuarine foods were sufficient and that system design and water quality were conducive to growth, especially in the first 2 mo. Survival was good from the historical perspective (cf., Arnold et al. 1974; Yang et al. 1983b) but rather poor from the produc- tion standpoint. A recent hypothesis concerning temperature effects on growth (O'Dor and Wells in press) indicates that higher temperature in the first half of the life cycle and lower temperature in the latter half may enhance growth and survival of laboratory-reared squid. In future work it would be desirable to enhance growth during the latter half of the life cycle and to provide an environment in which somatic growth continues for a longer period before sexual maturation occurs. Behavior Squid are generally sensitive laboratory animals, responding very quickly with their sophisticated sen- sory systems to any fast environmental change. They habituate to many daily disturbances in the tank system (e.g., tank cleaning, etc.) provided everything is done slowly. Later in the life cycle they become slightly less sensitive. Hatchlings were positively phototaxic and often swam at the water surface. In nature, young squid have been caught mainly by plankton nets mounted on a sled and towed along the bottom (Recksiek and Kashiwada 1979). It is not possible at this time to explain the movements of hatchlings in nature based upon laboratory observations of positive phototaxis. A key component in feeding behavior was move- ment by the prey, regardless of the size or age of the squid or food organisms. Young squid preferred copepods but ate a variety and a very wide size range of organisms (Fig. 4). In general, the squid preferred crustaceans over fish, but the relatively 792 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS restricted diet offered to them may have influenced that. Fields (1965) and Karpov and Cailliet (1978) agreed that L. opalescens adults prefer fish over crustaceans but there was no clear-cut preference in younger squid. It is clear from laboratory obser- vations that squid learned to associate certain events with feeding (e.g., opening the tank top), and the general level of activity increased markedly during these periods. We were also able to stimulate feeding in the CT systems by dimming and bright- ening the lights to attract the planktonic food organisms into the water column near the squid. Schooling behavior was correlated with size. Larger body size and growth of the fins were re- quired before squid could swim in place against a current; this occurred at about 10 mm ML (41-44 d in L.O. 1981 and 1982). Hurley (1976) reported that L. opalescens 4 to 5 mm ML could briefly form loose schools when disturbed, but this may have been in static water. At 15 mm ML, L. opalescens were powerful enough to form distinct schools (Yang et al. 1983a), indicating the size at which one could expect schooling to appear in nature. How and why squid begin schooling in nature has not been investigated. Cannibalism was not seen in L.O. 1980 (Yang et al. 1980b, 1983a) and accounted for 7 to 19% of mor- talities in experiments L.O. 1981 and 1982. Lack of food did not precipitate this behavior. On the spawn- ing grounds in Monterey, CA, mature squid often have cephalopod remains in their stomachs (Lou- kashkin 1977; Karpov and Cailliet 1978); in one case as many as 75% of males had squid remains in their stomachs (Fields 1965). This could be a behavioral response to overcrowding (Fields 1965) or to restrict prey organisms on the spawning grounds. We anti- cipate that cannibalism in tanks would be a signi- ficant problem only during prolonged food shortage or if squid of a very wide size range were in the same system (cf., Hanlon et al. 1983). Body patterning was not studied in great detail but several observations are noteworthy. Young animals are capable only of simple chromatic expres- sion such as "All dark" or "Clear". When excited, L. opalescens of all sizes show some degree of dark- ening; this is similar to other loliginid squids (cf., Hanlon 1982; Hanlon et al. 1983). By the time the squid are approximately 80 to 100 mm ML they can show a repertoire that includes about a dozen chromatic components of patterning (e.g., Dark arm tips, Ring on the mantle, etc.). This places L. opalescens in a category of rather simple pattern- ing, making it comparable to L. pealei and L. vulgaris, slightly more complex than Lolliguncula brevis (Dubas et al. 1986), but simpler than Loligo plei (Hanlon 1982; Hanlon et al. 1983). Further analysis is warranted because much behavior is ex- pressed through patterning and may yield impor- tant behavioral clues. Social behavior was first manifest in schooling (see above) then much later in mild intraspecific aggres- sion. Occasionally two squid would fight over one fish, but the first firm observations came at the time of sexual maturation when mating was seen. As Hurley (1977) noted, there were no obvious inter- actions among males to form a dominance hierarchy for mate selection. Mating was initiated by males, and both typical forms of mating were observed: "head-to-head" matings in which spermatophores were stored in the bursa copulatrix; and male- underneath matings in which spermatophores were deposited in the mantle near the oviduct (cf., Drew 1911; McGowan 1954; Hurley 1977). Females mated promiscuously as they do in nature, and females were also stimulated visually to lay eggs around ar- tificial facsimiles of egg mops (Fig. 16). Males were not observed to guard or defend egg capsules as described by Hurley (1977), but this may have been because relatively few egg capsules were left in the tank each day. Reproduction In L.O. 1980 only the subadult stage was reached in 233 d (Yang et al. 1980b, 1983a). Full sexual maturity was achieved in L.O. 1981 and 1982 and spawning of viable eggs occurred from days 196 to 239 and 175 to 226, respectively (Fig. 13). Relatively few egg capsules were laid per female, and these capsules were generally shorter and contained slightly fewer eggs per capsule than those reported from natural populations, but this was probably due to the smaller size of these spawning females (Hixon 1983). Laboratory cultured Loligo opalescens matured precociously and since they are terminal spawners this prevented attainment of full adult size. In the laboratory, males as small as 71 mm ML had fully formed spermatophores and females became sexual- ly mature beginning at about 60 mm ML (Fig. 15). In nature, the average adult size is 150 mm ML for males and 140 mm ML for females, although size at onset of maturity is variable and can be as low as 72 mm ML for males and 81 mm ML for females (Fields 1965; Hixon 1983). Precocious maturation has also been reported in other squid maintained in the laboratory (cf., Durward et al. 1980; Hanlon et 793 FISHERY BULLETIN: VOL. 84, NO. 4 al. 1983). The stimuli (or stressors) that cause this are unknown. Van Heukelem (1979) reviewed environmental fac- tors that influence maturation in cephalopods and reported that light, temperature, and nutrition are the key stimuli. In our experiments, light was con- stant (24 h on), temperature was consistent (ca. 15°C) and food was relatively constant and highly available compared with natural populations. How- ever, all three conditions are different from nature. The most interesting result concerns light, which is thought to have a major effect on maturation through the light-optic gland-gonad pathway (cf., Mangold and Froesch 1977; Wells and Wells 1977). Long daylength of high intensity is thought to delay maturation; in our experiments daylength was 24 h but intensity (ca. 4-17 lux) was low compared with full sunlight. However, we do not know what light intensity subadult L. opalescens are subject to in nature. Clearly, long daylength alone does not delay maturation in L. opalescens. Future experimenta- tion will be necessary to identify the combinations of environmental factors that affect maturation in the laboratory. Life Cycle Comparisons: Laboratory vs. Fishery Data In general, five major rearing attempts have been successful in varying degrees: 1) Hurley (1976), to 100 d; 2) Hanlon et al. (1979), to 79 d; 3) L.0. 1980, to 233 d and subadult stage (Yang et al, 1980b, 1983a); 4) and 5) L.O. 1981 and 1982, to sexual maturity and egg laying within 8 mo (this report). From this it is clear that the life cycle can be <1 yr under laboratory conditions. Fields (1965) stated, based upon fishery data, that "Almost all females spawn at the age of 3 years...." However, more recent field (cf., Recksiek and Frey 1978) and laboratory studies of L. opalescens (above) indicate that life span estimates beyond 2 years are excessive. Further- more, recent books on cephalopod life cycles (Boyle 1983, in press) indicate that few squid live beyond 2 years. Growth information on laboratory populations is now quite good. The present data allow an accurate assessment by weight from hatching onwards (Fig. 9) and firmly verify that young squid are capable of dramatically fast, exponential growth when food is not limiting. This indicates that in nature squid are capable of exploiting plankton blooms and other instances of greater food availability; the highest feeding rates we estimated (29%) also confirm field observations that squid will eat large quantities of food when available and when necessary. Field estimates of growth by Fields (1965) and Spratt (1978) are compared with laboratory data in Figure 17. Field's data are very conservative (averaging 4 mm/month) and based only upon monthly modal length-frequency diagrams from squid on or near spawning grounds. Spratt (1978) estimated growth from statolith rings and hypothesized that growth is rapid during the first few months then decreases with age. Laboratory growth was much faster, but animals were not subject to environmental fluctua- tions. We estimate that growth in nature approx- imates something between the laboratory data and Spratt' s data, and that date of hatching, seasonal temperature fluctuations, and food availability result in life cycle variations between 1 and 2 years. One would expect to observe exponential growth of young squid during spring and summer when tem- peratures and food availability are high, slower logarithmic growth in fall and winter, and spawn- ing the following spring. Field evidence (McGowan 1954; Fields 1965) and reproductive physiology studies (Grieb and Beeman 1978; Knipe and Beeman 1978) indicate that L. opalescens is a terminal spawner (Hixon 1983), and our laboratory observations verify this since all animals died shortly after spawning (Fig. 13). Rings in statoliths may eventually be used as a reliable age marker to determine growth rate and life span. Our preliminary results in this paper from 43 statoliths of known age support Spratt's (1978) conclusion that ring deposition occurs roughly on a daily basis during the first 65 d. However, our laboratory data indicate that the relationship does not hold well beyond that age, although Spratt sug- gested that daily ring deposition occurs up to 150 d. Thereafter, Spratt (1978) hypothesized lunar (monthly) rings on statoliths but there are no lab- oratory data for comparison. Daily, fortnightly, or monthly growth rings have been hypothesized in the squid Gonatus fabricii (Kristensen 1980), Todarodes sagittatus (Rosenberg et al. 1981), Illex illecebrosus (Hurley and Beck 1980), and Loligoforbesi (Martins 1982), but there are no hard data to confirm these estimates. The mechanism of ring formation is unclear but may be related to feeding, since in this part of our laboratory study the squid received food during 12 h and none for the next 12, while concur- rently there was constant light and no temperature fluctuation (Hixon and Villoch 1984). Hurley et al. (1985) and Dawe et al. (1985) found evidence of daily rings in statoliths by inoculating squid with tetra- cycline or strontium. Further work is required to 794 YANG ET AL.: CULTURE EXPERIMENTS OF LOLIGO OPALESCENS 175 -i 150- 125- | 100- 2 75- 50- 25 Yang, unpublished Yang eta/. (1980) Spratt (1978) Fields (1965) —r- 9 —i — 12 — i — 15 18 21 — i — 24 Age (mths) Figure 17.— From estimates of growth rate in mantle length oiLoligo opalescens. Fields (1965) used population data. Spratt (1978) combined age (statolith ring counts) and ML data and calculated a mean (horizontal line), range (vertical line) and standard deviation (vertical bar) values for 3-month intervals throughout the life cycle. Yang et al. data are from laboratory rearing studies (1980b, 1983a, this report) (modified from figure 7.1, Hixon 1983). determine if and how statolith rings are correlated with age. A major gap in fisheries studies concerns where the hatchlings go from the spawning grounds. Very few young squid have been captured (Okutani and McGowan 1969; Recksiek and Kashiwada 1979) even in the vicinity of spawning grounds. Hatchlings are positively phototaxic and this may serve to disperse them immediately from the spawning grounds. Thereafter their movements are unknown, although rarely young squid 3.5 to 7.0 mm ML have been caught in neritic plankton samples, usually at depths of 25 to 40 m nearshore in water between 12.5° and 21.0°C (Okutani and McGowan 1969). Detailed knowledge of water currents between spawning grounds and nearshore, combined with monitoring of plankton abundance (especially copepods and lar- val fish) by surface, bottom and oblique tows may provide important clues about movements and feed- ing patterns of young-of-the-year squid. Laboratory studies indicate that squid can swim well enough to hold their position against a current by 10 mm ML, or about 40 to 45 d posthatching. By 15 mm ML (ca. 60-80 d) they can form and maintain well-formed schools. The functions of schooling in nature prob- ably relate to defense, feeding and migratory behavior. The California squid fishery has nearly collapsed since El Nino of 1983, and the squid population has been generally displaced northward as far as south- ern Canada. Some small spawning populations are still present in southern and central California. It may be rewarding to investigate feeding and migra- tory patterns of young and adult squid to better understand population recruitment into this ecologi- cally and economically important fishery resource. Biomedical Research Applications Loligo opalescens has proved to be a suitable model for giant axon preparations (e.g., Llano and Bezanilla 1980). However, for most axon experi- ments the largest axons (>400 fim diameter) are needed; this requires the largest squid taken in the fishery, usually 150 mm ML and larger. Our largest squid, 116 mm ML, had an axon about 240 ^m in 795 FISHERY BULLETIN: VOL. 84, NO. 4 diameter. Unknown factors in our laboratory en- vironment resulted in precocious sexual maturation and thus smaller animals. Therefore, we are now evaluating the culture potential of Loligo forbesi, a much larger squid from the eastern Atlantic, since precocious maturation in that species would still result in axons >500 /urn. Preliminary experiments bear out this proposition as we have recently cultured L. forbesi to 140 mm ML and 400 \xm diameter axons. However, L. opalescens would be an excellent model for the giant synapse preparation in which smaller squid are most suitable. Therefore, L. opalescens, with a now substantial amount of culture information, may be a highly suitable species in the United States for providing squid on a con- sistent basis for neuroscience research. Moreover, the recent disappearance of L. opalescens (1983-85) from traditional fishing grounds in California make laboratory culture an attractive alternative for animal supply. ACKNOWLEDGMENTS We acknowledge funding from DHHS grant RR01024, Division of Research Resources, National Institutes of Health, and from the Marine Medicine General Budget 7-11500-765111 of The Marine Bio- medical Institute, The University of Texas Medical Branch. We especially appreciate the assistance of John W. Forsythe on the rearing experiments and the growth data analyses. We also thank Joseph P. Hendrix Jr. for assistance in rearing and Lea A. Bradford for water analyses and data gathering. Connie Arnold and Joan Holt of the Port Aransas Laboratory, University of Texas, kindly supplied the red drum eggs. We are grateful to Aquabiology (Seibutsu Kenkyusha Publishing Co., Tokyo) for per- mission to reprint several figures. Academic Press Inc. kindly gave us permission to use a modification on Figure 17. Phillip G. Lee kindly read and im- proved the final draft. Note: We dedicate this paper to our coauthor and dear friend Dr. Raymond F. Hixon, who passed away on 19 March 1984 as he valiantly fought to recover from chronic myelogenous leukemia. LITERATURE CITED Arnold, J. M. 1965. Normal embryonic stages of the squid, Loligo pealei (Lesueur). Biol. Bull. (Woods Hole) 128:24-32. Arnold, J. M., W. C. Summers, D. L. Gilbert, R. S. Manalis, N. W. Daw, and R. J. Lasek. 1974. A guide to laboratory use of the squid Loligo pealei. Marine Biological Laboratory, Woods Hole, MA, 75 p. Atz, J. W. 1964. Some principles and practices of water management for marine aquariums. In J. R. Clark and R. L. Clark (editors), Sea-water systems for experimental aquariums, p. 3-16. U.S. Fish Wil'dl. Serv., Res. Rep. 63. BOLETZKY, S. V. 1975. A contribution to the study of yolk absorption in the Cephalopoda. Z. Morphol. Tiere 80:229-246. Bower, C. E., D. T. Turner, and S. Spotte. 1981. pH maintenance in closed seawater culture systems: limitations of calcareous filtrants. Aquaculture 23:211-217. Boyle, P. R. (editor). 1983. Cephalopod life cycles, Vol. I: Species accounts. Acad. Press, Lond., 475 p. In press. Cephalopod life cycles, Vol. II: Comparative re- views. Acad. Press, Lond. Dawe, E. C, R. K. O'Dor, P. H. Odense, and G. V. Hurley. 1985. Validation and application of an ageing technique for short-finned squid (Illex illecebrosus). J. Northwest Atl. Fish. Sci. 6:107-116. Drew, G. A. 1911. Sexual activities of the squid, Loligo pealei (Les.). J. Morphol. Wistar Inst. Ana. Biol. Phila. 22(2):327-359. Dubas, F., R. T. Hanlon, G. P. Ferguson, and H. M. Pinsker. 1986. Localization and stimulation of chromatophore moto- neurons in the brain of the squid Lolliguncula brevis. J. Exp. Biol. 121:1-25. Durward, R. D., E. Vessey, R. K. O'Dor, and T. Amaratunga. 1980. Reproduction in the squid, Illex illecebrosus: first obser- vations in captivity and implications for the life cycle. ICNAF, Sel. Pap. 6, p. 7-13. Fields, W. G. 1965. The structure, development, food relations, reproduc- tion, and life history of the squid Loligo opalescens Berry. Calif. Dep. Fish Game, Fish Bull. 131, 108 p. Forsythe, J. W., and W. F. Van Heukelem. In press. Cephalopod growth. In P. R. Boyle (editor), Cephalopod life cycles, Vol. II: Comparative reviews. Acad. Press, Lond. Grieb, T. M., and R. D. Beeman. 1978. A study of spermatogenesis in the spawning popula- tion of the squid, Loligo opalescens. In C. W. Recksiek and H. W. Frey (editors), Biological, oceanographic, and acoustic aspects of the market squid, Loligo opalescens Berry, p. 11-21. Calif. Dep. Fish Game, Fish. Bull. 169. Hamabe, M. 1960. Observations of early development of a squid, Loligo bleekeri Keferstein. [In Jpn. with Engl, abstr.] Ann. Rep. Jpn. Sea Reg. Fish. Res. Lab. 6:149-155. Hanlon, R. T. 1978. The biology of the squid Loligo (Doryteuthis) plei in cap- tivity. Ph.D. Thesis, Univ. Miami, Coral Gables, 120 p. 1982. The functional organization of chromatophores and iridescent cells in the body patterning of Loligo plei (Cephalopoda: Myopsida). Malacologia 23:89-119. Hanlon, R. T., and R. F. Hixon. 1983. Laboratory maintenance and culture of octopuses and loliginid squids. In C. J. Berg, Jr. (editor), Culture of marine invertebrates: Selected readings, p. 44-61. Hutchinson Ross Publishing Co., Stroudsburg, PA, 386 p. Hanlon, R. T., R. F. Hixon, and W. H. Hulet. 1978. Laboratory maintenance of wild-caught loliginid squids. In N. Balch, T. Amaratunga, and R. K. O'Dor (editors), Fisheries and marine service technical report 833, p. 20.1- 20.14. Dalhousie University, Halifax, Nova Scotia. 1983. Survival, growth, and behavior of the loliginid squids 796 YANG ET AL.: CULTURE EXPERIMENTS OF LOL1GO OPALESCENS Loligo plei, Loligo pealei, and Lolliguncula brevis (Mollusca: Cephalopoda) in closed sea water systems. Biol. Bull. (Woods Hole) 165:637-685. Hanlon, R. T., R. F. Hixon, W. H. Hulet, and W. T. Yang. 1979. Rearing experiments on the California market squid Loligo opalescens Berry, 1911. Veliger 21:428-431. Hirayama, K. 1966. Influence of nitrate accumulated in culturing water on Octopus vulgaris. [In Jpn. with Engl, abstr.] Bull. Jpn. Soc. Sci. Fish. 32(2):105-111. Hirtle, R. W. M., M. E. DeMont, and R. K. O'Dor. 1981. Feeding, growth, and metabolic rates in captive short- finned squid, Illex illecebrosus, in relation to the natural population. J. Shellfish Res. 1:187-192. Hixon, R. F. 1980a. Growth, reproductive biology, distribution and abun- dance of three species of loliginid squids (Myopsida, Cepha- lopoda) in the northwest Gulf of Mexico. Ph.D. Thesis, Miami, Coral Gables, 233 p. 1980b. Potential commercial squid resources of the Gulf of Mexico - an updated review. Sea Grant Rep. MASGP-80- 012 (Mississippi-Alabama Sea Grant Consortium), p. 54-73. 1983. Loligo opalescens. In P. R. Boyle (editor), Cephalopod life cycles, Vol. I: Species accounts, p. 95-114. Acad. Press, N.Y., 475 p. Hixon, R. F., R. T. Hanlon, S. M. Gillespie, and W. L. Griffin. 1980. Squid fishery in Texas: biological, economic, and market considerations. Mar. Fish. Rev. 42(7-8):44-50. Hixon, R. T., and M. R. Villoch. 1984. Growth rings in the statoliths of young laboratory cultured squids (Loligo opalescens). Am. Malacol. Bull. 2:93. Hulet, W. H., M. R. Villoch, R. F. Hixon, and R. T. Hanlon. 1979. Fin damage in captured and reared squids. Lab. Anim. Sci. 29:528-533. Hurley, A. C. 1976. Feeding behavior, food consumption, growth, and respiration of the squid Loligo opalescens raised in the laboratory. Fish. Bull, U.S. 74:176-182. 1977. Mating behavior of the squid Loligo opalescens. Mar. Behav. Physiol. 4:195-203. Hurley, G. V., and P. Beck. 1980. The observation of growth rings in statoliths from the Ommastrephid squid, Illex illecebrosus. NAFO SCR. Doc. 80/II/1, 17 p. Hurley, G. V., P. H. O'Dense, R. K. O'Dor, and E. G. Dawe. 1985. Strontium labelling for verifying daily growth incre- ments in the statolith of the short-finned squid {Illex il- lecebrosus). Can. J. Fish. Aquat. Sci. 42:380-383. Karpov, K. A., and G. M. Cailliet. 1978. Feeding dynamics of Loligo opalescens. In C. W. Recksiek and H. W. Frey (editors), Biological, ocean- ographic, and acoustic aspects of the market squid, Loligo opalescens Berry, p. 45-65. Calif. Dep. Fish Game, Fish. Bull. 169. Knipe, J. H., and R. D. Beeman. 1978. Histological observations on oogenesis in Loligo opalescens. In C. W. Recksiek and H. W. Frey (editors), Biological, oceanographic, and acoustic aspects of the market squid, Loligo opalescens Berry, p. 23-33. Calif. Dep. Fish Game, Fish. Bull. 169. Kristensen, T. K. 1980. Periodical growth rings in cephalopod statoliths. Dana 1:39-51. La Roe, E. T. 1971. The culture and maintenance of the loliginid squids Sepioteuthis sepioidea and Doryteuthis plei. Mar. Biol. (Berl.) 9:9-25. Llano, I., and F. Bezanilla. 1980. Current recorded from a cut-open giant axon under voltage clamp. Proc. Natl. Acad. Sci. USA 77(12):7484- 7486. LOUKASHKIN, A. S. 1977. On biology of the market squid, Loligo opalescens, a contribution toward the knowledge of its food habits and feeding behavior. Calif. Coop. Ocean. Fish. Invest. Rep. 18:108-111. Macy, W. K., III. 1980. The ecology of the common squid, Loligo pealei LeSueur, 1821, in Rhode Island waters. Ph.D. Thesis, Univ. Rhode Island, Kingston, 249 p. 1982. Development and application of an objective method for classifying long-finned squid, Loligo pealei, into sexual maturity stages. Fish. Bull., U.S. 80:449-459. Mangold, K., and D. Froesch. 1977. A reconsideration of factors associated with sexual maturation. Symp. Zool. Soc, Lond. 38:541-555. Martins, H. R. 1982. Biological studies of the exploited stock of Loligo forbesi (Mollusca: Cephalopoda) in the Azores. J. Mar. Biol. Assoc. U.K. 62:799-808. McConathy, D. A., R. T. Hanlon, and R. F. Hixon. 1980. Chromatophore arrangements of hatchling loliginid squids (Cephalopoda, Myopsida). Malacologia 19:279-288. McGowan, J. A. 1954. Observations on the sexual behavior and spawning of the squid, Loligo opalescens, at La Jolla, California. Calif. Fish Game 40:47-54. Moe, Jr., M. A. 1982. The marine aquarium handbook: beginner to breeder. Norns Publishing Co., Marathon, FL, 169 p. O'Dor, R. K., and M. J. Wells. In press. Energy and nutrient flow in cephalopods. In P. R. Boyle (editor), Cephalopod life cycles, Vol. II: Comparative reviews. Acad. Press, Lond. Okutani, T., and J. A. McGowan. 1969. Systematics, distribution, and abundance of the epi- planktonic squid (Cephalopoda, Decapoda) larvae of the California current April, 1954 - March, 1957. Bull. Scripps Inst. Oceanogr. 14, Univ. Calif. Press, Berkeley, 89 p. Rathjen, W. F., R. F. Hixon, and R. T. Hanlon. 1979. Squid fishery resources and development in the North- west Atlantic and Gulf of Mexico. Gulf Caribb. Fish. Inst. 31:145-157. Recksiek, C. W., and H. W. Frey (editors). 1978. Biological, oceanographic, and acoustic aspects of the market squid, Loligo opalescens Berry. Calif. Dep. Fish Game, Fish. Bull. 169, 185 p. Recksiek, C. W., and J. Kashiwada. 1979. Distribution of larval squid, Loligo opalescens, in various nearshore locations. Calif. Coop. Ocean. Fish. Invest. Rep. 20:31-34. Roper, C. F. E., C. C. Lu, and F. G. Hochberg (editors). 1983. The biology and resource potential of cephalopods. Memoirs of the National Museum of Victoria, Melbourne, Australia, No. 44, 311 p. Brown-Prior-Anderson Pty. Ltd., Burnwood, Victoria. Rosenberg, A. A., K. F. Wiborg, and I. M. Bech. 1981. Growth of Todarodes sagittatus (Lamarck) (Cephalo- poda, Ommastrephidae) from the Northeast Atlantic, based on counts of statolith growth rings. Sarsia 66(l):53-57. Soichi, M. 1976. On the growth and food quantity of young common 797 squid, Todarodes pacificus, in captivity. [In Jpn.] Dosuishi 13(4):79-82. Spotte, S. 1979a. Fish and invertebrate culture. Water management in closed systems. 2d ed. Wiley-Interscience, N.Y., 179 p. 1979b. Seawater aquariums: The captive environment. John Wiley and Sons, N.Y., 413 p. Spratt, J. D. 1978. Age and growth of the market squid, Loligo opalescens Berry, in Monterey Bay. In C. W. Recksiek and H. W. Frey (editors), Biological, oceanographic, and acoustic aspects of the market squid, Loligo opalescens, Berry, p. 35-44. Calif. Dep. Fish Game, Fish. Bull. 169. Strickland, J. D. H., and R. T. Parsons. 1972. A practical handbook of seawater analysis. 2d ed. Fish. Res. Board Can., Bull. 167, 310 p. Takeuchi, T. 1969. Strange spawning of the YARI-IKA (Doryteuthis bleekeri): letting incubate the eggs by a crab. [In Jpn.] Fish Mag. 1969:86-89. 1976. Spawning behavior of YARI-IKA (Doryteuthis bleekeri). [In Jpn.] Fish Mag. 1976:1-4. Van Heukelem, W. F. 1979. Environmental control of reproduction and life span FISHERY BULLETIN: VOL. 84, NO. 4 in Octopus: an hypothesis. In S. E. Stancyk (editor), Reproductive ecology of marine invertebrates, p. 123-133. The Belle W. Baruch Library in Marine Science, No. 9, Univ. So. Carolina Press, Columbia. Wells, M. J., and J. Wells. 1977. Cephalopoda: Octopoda. In A. C. Giese and J. S. Pearse (editors), Reproduction of marine invertebrates 4, p. 291-336. Acad. Press, Lond. Yang, W. T., R. T. Hanlon, R. F. Hixon, and W. H. Hulet. 1980a. First success in rearing hatchlings of the squid Loligo pealei Lesueur 1821. Malacol. Rev. 13:79-80. Yang, W. T., R. T. Hanlon, M. E. Krejci, R. F. Hixon, and W. H. Hulet. 1980b. Culture of California market squid from hatching - first rearing of Loligo to sub-adult stage. [In Jpn. with Engl, abstr.] Aquabiology 2:412-418. 1983a. Laboratory rearing of Loligo opalescens, the market squid of California. Aquaculture 31:77-88. Yang, W. T., R. F. Hixon, P. E. Turk, M. E. Krejci, R. T. Hanlon, and W. H. Hulet. 1983b. Culture of California market squid from hatching - completion of the rearing cycle to second generation hatchlings. [In Jpn. with Engl, abstr.] Aquabiology 5:328-339. 798 FISH ASSEMBLAGES IN MACROCYSTIS AND NEREOCYSTIS KELP FORESTS OFF CENTRAL CALIFORNIA James Lee Bodkin1 ABSTRACT The abundance and species composition of conspicuous fishes were compared within two canopy forming kelp forests (giant kelp, Macrocystis pyrifera, and bull kelp, Nereocystis luetkeana) in Central Califor- nia. The primary investigative method was a subtidal belt transect, in which visual observation was used. The species composition of fish assemblages in the two canopy types was similar. Densities of fish were generally greater in Macrocystis than in Nereocystis forests. The major difference was the density of midwater species of the genus Sebastes. The blue rockfish, Sebastes mystinus, was the numerically domi- nant species in both canopy types. Estimates of the biomass of fish were about 2.4 times greater in Macrocystis beds than in Nereocystis beds. Many species of fish exhibit an affinity for substrate and cover within their habitat, such as rock or coral reefs or kelp beds, as well as man-made objects such as piers, jetties, and offshore oil platforms. This structure may provide shelter, a base for foraging activity, or nursery habitat for young fish. Within the temperate nearshore marine environment, macroalgae may provide a large portion of this substrate and cover. Kelp forests are one of the major features of the nearshore environment along the west coast of North America. The two most con- spicuous canopy-forming kelps are the giant kelp, Macrocystis pyrifera, a perennial, and the bull kelp, Nereocystis luetkeana, an annual (Abbott and Hollenberg 1976). Besides the difference in peren- nial versus annual growth pattern, Macrocystis and Nereocystis differ markedly in physical structure (Fig. 1) and seasonal patterns of abundance. Macro- cystis plants typically have many stipes originating from a single large holdfast, and large fronds at- tached to each stipe throughout its length. Nereo- cystis plants consist of a single stipe, with large fronds only at the distal end. During periods of full development (typically late summer), Macrocystis can develop a completely closed canopy, whereas Nereocystis typically has a broken canopy. Winter storms usually remove large portions of the Macro- cystis canopy, but many plants remain secured to the substrate and provide structure within the water column to varying depths throughout the year. Nereocystis canopies are also typically removed dur- iU.S. Fish and Wildlife Service, P.O. Box 70, San Simeon, CA 93452. ing these storms, and, because Nereocystis is an annual, it provides little or no structure from mid- winter through late spring. Nereocystis may be more abundant than Macro- cystis in the presence of severe and persistent disturbances such as continued exposure to large swells or heavy grazing pressure (Dayton et al. 1980). In the absence of this pressure, Macrocystis may be competitively dominant, in that it forms a dense and often complete surface canopy earlier in the year, and thus may exclude or limit Nereocystis which has light-sensitive germination requirements (Dayton et al. 1980, 1984). This study was designed to test the hypothesis that the fish component of the Macrocystis pyrifera community differs from that of the Nereocystis luet- keana community in Central California. METHODS Studies were conducted from 6 km south to 15 km north of Point Piedras Blancas, San Luis Obispo County, CA (lat. 35°40'N, long. 121°17'W) (Fig. 2). Additional studies were also done near Big Creek, Monterey County, CA Gat. 36°04'N, long. 121°36'W). The surface canopies of kelp beds consist almost ex- clusively of Nereocystis from Point Piedras Blancas north to Ragged Point, an area about 13 km long, but are dominated by Macrocystis south of Piedras Blancas. I searched 74 transects in the Piedras Blancas study area and 4 in the Big Creek area: 26 transects in Macrocystis forests and 14 in Nereo- cystis in 1982 and 17 in Macrocystis and 21 in Nereo- cystis in 1983. Field studies extended from June Manuscript accepted March 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 799 FISHERY BULLETIN: VOL. 84, NO. 4 01 he -a c 08 a o o> o ft c o 'E es ft o o < I « a 800 BODKIN: MACROCYSTIS AND NEREOCYSTIS KELP FORESTS N Ragged Pt. Pt. Sierra Nevada Pt. Piedras Blancas Figure 2 . — Location of areas sampled . Piedras Blancas Pt., south, to San Simeon Pt.: kelp canopies are dominated by Macrocystis pyrifera. Piedras Blancas Pt. north, to Ragged Pt.: kelp canopies are dominated by Nereocystis luetkeana. Big Creek (not shown) is about 38 km north of Ragged Pt. '■:■■ *&K-- • San Simeon 10 km 1982 to October 1983. Transects were apportioned evenly throughout early summer to late fall in each of the forest types. A belt transect, as described by Brock (1954) and modified by Quast (1968), was used with the aid of scuba to conduct subtidal fish surveys. Each survey consisted of two components, benthic and midwater. A 50 m fiberglass tape was extended across the ocean floor in differing compass courses, extending from eye bolts permanently embedded in the sub- strate, or from the anchor of a dive boat on hap- hazardly located transect sites. The width of the midwater transect was determined by measuring the horizontal water visibility 2 m above the sub- strate. This was done by sighting down the transect line (fiberglass tape) toward the zero end, where a small bicolored float (13.5 x 5.5 cm) was suspended 2 m off the bottom. The observer moved away from the float along the line. When the float could not be readily discerned, the position on the tape was recorded. This value was doubled (to include obser- vations on either side of the transect line) to obtain the width of the midwater transect. This survey technique may lead to a slight underestimation of fish densities due to decreasing searching efficiency with increasing distances from observer to observed (Caughley 1977). Surveys were conducted only when visibility exceeded 4.4 m. Horizontal water visibil- ity ranged from 4.4 to 12.1 m (Macrocystis x = 6.6 m, SE = 2.6; Nereocystis x = 7.4 m, SE = 0.49). The width of the benthic survey was 4 m (2 m on each side of the transect tape). All sampling 801 FISHERY BULLETIN: VOL. 84, NO. 4 wasconducted beneath and within either of the two forest types, in water 6 to 22 m deep. Underwater observations were recorded on formated data sheets using plastic paper. In conducting the benthic survey, I slowly swam from one end of the transect to the other and iden- tified and enumerated the fish that were observed. A fish was included in the benthic survey if it was observed within 0.5 m of the bottom and was not a member of a school of typically midwater fish located momentarily near the bottom. A fish ob- served swimming through the transect in front of the diver was included. An effort was made to in- spect all crevices, caves, and ledges, and to move aside algae to locate fish. A description of unfamiliar fish was made in the field and its identity later deter- mined in field guides if possible. Small, relatively cryptic species were probably underestimated in the process of these visual surveys (Brock 1982). The midwater transect was searched about 3 m above the tape. Repetitive ascents and descents were made at 5 m intervals to detect fish occurring throughout the water column. The sizes of very large schools were estimated. All fish observed within the length of the 50 m tape were recorded. Unidentified species were treated as they were dur- ing the benthic survey. After the survey was completed, an index of the bottom profile was recorded by measuring the water depth at each meter mark along the tape. Two methods of determining bottom profile were used: first, an objective, and later, a subjective measure. The objective relief index was the sum of the dif- ferences between each of the 50 consecutive depth measurements along the 50 m transect. During the second half of this study (1983) a subjective relief index was assigned to the general vicinity of each transect; this was determined by the greatest ver- tical relief observed along the transect line: 0 = flat, no relief; 1 = low relief <1 m); 2 = moderate relief (1 to 2 m); 3 = high relief (2 to 4 m); and 4 = ex- treme relief (more than 4 m). Two measures of species diversity were used to compare the fish assemblages in Macrocystis and Nereocystis forests: 1) total number of species found on all transects within one canopy type and 2) the Shannon-Weaver index of diversity, H' (Pielou 1966). Because of heterogeneity between sample vari- ances, fish density distributions were compared with the nonparametric Mann- Whitney test. A minimum acceptable level of significance of 0.05 was assigned. RESULTS Twenty-seven species of fish were identified within the spatial limits of the transects (Tables 1, 2, 3). An additional 8 species were identified within the kelp forest, but outside the transect limits. Juvenile rockfish were considered a single group, and occasionally an unidentified fish was observed. In Macrocystis forests, 26 species were identified within the transects and 10 species outside the transects; in Nereocystis forests, the respective totals were 23 and 4 species. Three additional types of fish were observed that could be identified only to the family level (Table 3). Four species observed only in Macrocystis forests were white seaperch, Phanerodon furcatus; rainbow seaperch, Hypsurus caryi; China rockfish, Sebastes nebulosus; and black- eye goby, Coryphopterus nicholsi. One species was observed only in Nereocystis beds, the jacksmelt, Atherinopsis calif orniensis. Species not observed within both transect types were relatively uncom- mon, but were observed in and around both forest types during this study. Fishes that could not be identified to species or family level were rare, occurring on only 6 (8%) of the transects (Table 3). Table 1.— Summary of presence/absence of fish species encountered [midwater (M) and benthic (B), years pooled] throughout study. Macro- Nereo- Principal Species cystis cystis habitat Sebastes mystinus X X M Sebastes serranoides X X M Sebastes atrovirens X X M Sebastes melanops X X M Sebastes chrysomelas X X B Sebastes carnatus X X B Sebastes miniatus X X B Sebastes rastrelliger X X B Sebastes caurinus X X B Sebastes nebulosus X B Sebastes sp. (juveniles) X X M/B Oxyjulis califomica X X M Aulorhynchus flavidus X X M Atherinopsis californiensis X M Phanerodon furcatus X M Oxylebius pictus X X B Hexagrammos decagrammus X X B Embiotoca lateralis X X B Embiotoca jacksoni X X B Orthonopias triads X X B Scorpaenichthys marmoratus X X B Ophiodon elongatus X X B Rhachochilus vacca X X B Coryphopterus nicholsi X B Anarrhichthys ocellatus X X B Jordania zonope X X B Hypsurus caryi X B 802 BODKIN: MACROCYSTIS AND NEREOCYSTIS KELP FORESTS Midwater Transects Differences in abundance of fish in the Macro- cystis and Nereocystis forests were most apparent among the midwater species, primarily within the genus Sebastes. Of the nine species of midwater fish (juvenile Sebastes treated as a single "species"), three were significantly more abundant in Macro- cystis than in Nereocystis forests: blue rockfish, S. mystinus; kelp rockfish, S. atrovirens; and olive rockfish, 5. serranoides (Tables 1, 2). A fourth species, the black rockfish, S. melanops, was not ob- served on Nereocystis transects, though it was only occasionally seen in Macrocystis. Although there were no general changes in fish abundance between 1982 and 1983 among the mid- water species, some individual species differences were noted. Densities of blue rockfish were signifi- cantly lower in 1983 than in 1982 (Table 2). During this same period there was an insignificant increase in the density of juvenile rockfish. Densities of the senorita, Oxyjulis californica, appeared to increase within both forest types in 1983, but the increase was significant only when canopy types were com- bined for each year. This annual variation should be considered in light of the extremely anomolous El Nino event which occurred during this period (Cane 1983), and may be atypical. Benthic Transects Among the 19 principally benthic species found in both the Macrocystis and Nereocystis benthic transects, three (16%) were significantly more abun- dant in Macrocystis forests: Striped seaperch, Embiotoca lateralis, painted greenling, Oxylebius pictus, and the gopher rockfish, Sebastes carnatus (Tables 1, 3). One other species, the kelp rockfish, which occurred on benthic transects, was considered as primarily a midwater species. Gopher rockfish are bathymetrically segregated from the sibling species, S. chrysomelas (black-and-yellow rockfish). Gopher rockfish are relatively more abundant at depths >12 to 14 m (Larson 1980). In my study, the densities of black-and-yellow rockfish increased significant- ly in the second year while during the same period, densities of gopher rockfish decreased. Due to sampling methodology and the occurrence Table 2.— Mean densities (no. fish/100 m2) and frequency of occurrence of fishes on midwater transects through kelp (standard error of mean in parenthesis). Mean densities (fish/100 m2) Freque sncv of i Macrocystis Nereocystis occurrence Species 1982 1983 1982-83 1982 1983 1982-83 Macrocystis Nereocystis Sebastes mystinus"12 19.4 8.25 15.0 6.68 2.09 3.9 1.00 0.82 Blue rockfish (1.8) (1.0) Sebastes serranoides1 0.51 0.36 0.45 0.17 0.07 0.11 0.74 0.34 Olive rockfish (0.09) (0.03) Sebastes atrovirens1 0.19 0.16 0.18 0.007 0.005 0.006 0.44 0.06 Kelp rockfish (0.05) (0.004) Sebastes melanops1 0.03 0.01 0.02 0 0 0 0.16 0 Black rockfish (0.009) Sebastes sp. 3.4 7.7 5.1 0.06 0.95 0.59 0.19 0.11 Juvenile rockfish (3.1) (0.5) Oxyjulis californica2 3.1 26.6 12.4 1.6 18.7 11.9 0.40 0.40 Senorita (6.6) (6.2) Aulorhynchus flavidus 0.43 0.014 0.07 0.06 Tube-snout (0.4) (0.01) Atherinopsis californiensis 0 6.0 0 0.20 Jacksmelt (6.5) Phanerodon furcatus 1.37 0 0.05 0 White seaperch (1.4) Species observed incidental to transects Scomber japonicus 0 0.09 Chub mackerel Myliobatis californica 0 0.03 Bat ray Sphyraena argentea 0.02 0 Pacific barracuda Torpedo californica 0.02 0 Pacific electric ray 1 Difference significant between Macrocystis and Nereocystis, years combined. 2 Difference significant between years, kelp canopies combined. 803 FISHERY BULLETIN: VOL. 84, NO. 4 Table 3.— Mean densities (no. fish/100 m2) and frequency of occurrence of fishes on benthic transects through kelp forests (standard error of mean in parenthesis). Mean densities (fish/100 m2) Frequency of Macrocystis Nereocystis occurrence Species 1982 1983 1982-83 1982 1983 1982-83 Macrocystis Nereocystis Sebastes chrysomelas1 1.52 1.91 1.67 1.11 2.21 1.77 0.74 0.91 Black-and-yellow rockfish (0.25) (0.26) Oxylebius pictus2,3 1.13 1.35 1.2 0.21 0.79 0.56 0.86 0.51 Painted greenling (0.1) (0.1) Hexagrammos decagrammus 0.33 0.35 0.34 0.36 0.43 0.40 0.44 0.57 Kelp greenling (0.07) (0.07) Sebastes carnatus^2 1.29 0.76 1.04 0.75 0.22 0.43 0.61 0.31 Gopher rockfish (0.2) (0.15) Embiotoca lateralis2 0.63 1.1 0.84 0.25 0.12 0.17 0.58 0.20 Striped seaperch (0.2) (0.08) Sebastes atrovirens2 0.52 0.97 0.70 0.04 0.15 0.11 0.58 0.14 Kelp rockfish (0.1) (0.05) Sebastes sp. 0.87 0.21 0.62 0.23 0.14 0.17 0.42 0.26 Juvenile rockfish (0.2) (0.07) Embiotoca jacksoni 0.39 0.44 0.41 0 0.27 0.16 0.42 0.17 Black perch (0.1) (0.06) Orthonopias triads 0.20 0.23 0.21 0.04 0.13 0.09 0.33 0.14 Snubnose sculpin (0.06) (0.04) Sebastes mystinus 0.08 0.26 0.15 0.04 0.17 0.15 0.23 0.17 Blue rockfish (0.05) (0.05) Scorpaenichthys marmoratus 0.107 0.11 0.16 0.20 Cabezon (0.04) (0.04) Ophiodon elongatus 0.13 0.09 0.21 0.09 Ling cod (0.04) (0.04) Sebastes melanops2 0.209 0.029 0.23 0.06 Black rockfish (0.06) (0.02) Rhachochilus vacca 0.135 0.0149 0.21 0.06 Pile perch (0.04) (0.01) Sebastes miniatus 0.042 0.094 0.07 0.14 Vermilion rockfish (0.03) (0.04) Coryphopterus nicholsi 0.198 0 0.21 0 Blackeye goby (0.09) Sebastes rastrelliger 0.0116 0.0143 0.05 0.11 Grass rockfish (0.01) (0.01) Sebastes caurinus 0.035 0.0143 0.07 0.03 Copper rockfish (0.02) (0.01) Anarrhichthys ocellatus 0.023 0.0143 0.05 0.03 Wolf-eel (0.02) (0.01) Jordania zonope 0.014 0.0143 0.02 0.03 Longfin sculpin (0.01) (0.01) Hypsurus caryi 0.034 0 0.05 0 Rainbow seaperch (0.01) Sebastes nebulosus 0.019 0 0.02 0 China rockfish (0.02) Unidentified fish 0.128 (0.09) 0.29 (0.3) 0.05 0.11 Species observed incidental to transects Sebastes serriceps 0.02 0.03 Treefish Cephaloscyllium ventriosum 0.05 0 Swellshark Sebastes auriculatus 0.02 0 Brown rockfish Sebastes pinniger 0.02 0 Canary rockfish Clinidae 0.12 0 Clinids Cottidae 0.07 0 Sculpins Gobiesocidae 0.02 0 Cling fishes Unidentified fish 0.12 0.06 'Difference significant between years, kelp canopies combined. 2 Difference significant between Macrocystis and Nereocystis years combined. 3 Difference significant between years, Nereocystis. 804 BODKIN: MACROCYSTIS AND NEREOCYSTIS KELP FORESTS of Macrocystis in water up to 4 m deeper than that occupied by Nereocystis within the study area, the mean water depth at which surveys were made dif- fered between sites (Macrocystis mean depth =12.2 m; Nereocystis mean depth = 10. 5m, t = 2.73, P = 0.008 (two sample £-test)). When the five transects in Macrocystis which occurred at depths beyond the maximum depth of Nereocystis transects (16 m) were excluded from analysis, the difference in water depths between sites became insignificant. Follow- ing the removal of these deep transects, all species of fish, both midwater and benthic, were reevalu- ated. There were no changes in the results presented above following this treatment. There was little correlation between densities of fish and either of the bottom relief indices (r values, 0.025 to 0.482). Throughout the study, bottom relief typically ranged from 1 to 4 m and relief <1 m was not encountered. Mean values of the objective relief index were 44.1 (SE = 2.8) for Macrocystis tran- sects and 37.2 (SE = 2.2) for Nereocystis transects. This difference resulted in a P value of 0.061 (two sample £-test), which I considered significant. How- ever, when all species of fish which demonstrated significantly different densities between canopy types were reevaluated, after excluding the six Macrocystis transects with relief values more than one standard deviation above the mean, no change in results was observed for any species tested. The total number of species encountered on the transects was 26 in Macrocystis and 23 in Nereo- cystis. The two kelp forests had 22 species in com- mon. Five species were found in only one of the two canopy types, although none of these were present in more than 21% of the transects within the canopy in which it was found. The H' values calculated were 1.76 for Macrocystis transects and 1.58 for the Nereocystis transects. Although the value of diver- sity indices has been questioned (Goodman 1975), such indices are widely used in ecological literature. Neither measure of diversity used in the present study indicated differences in the diversity of fish assemblages between the two kelp forest types investigated. DISCUSSION Several measures of comparison were considered in the analysis of these two kelp communities: species composition, species diversity, and abun- dance of fishes. The data presented here demon- strate very little difference in either composition or diversity of fish assemblages (Table 1), while esti- mates of biomass were markedly higher in giant kelp compared with bull kelp (Table 4). The single most obvious difference between the two kelp communities was in the abundance of the blue rockfish: mean density of fish (no./lOO m2) was Table 4.— Estimates of biomass of fish of Macrocystis and Nereocystis kelp forests. Species that were uncommon, (<20% of transects), or small are not included. Macrocystis Nereocystis Mean Mean Density weight1 Biomass Density weight1 Biomass Species (#/100 m2) (kg) (kg/100 m2) (#/100 m2) (kg) (kg/100 m2) Midwater transects Sebastes mystinus 15.0 0.44 6.6 3.92 0.50 1.96 Sebastes serranoides 0.45 0.63 0.28 0.11 0.72 0.08 Sebastes atrovirens 0.18 0.54 0.09 0.006 0.57 0.003 Sebastes melanops 0.02 0.44 0.009 0 0 0 Oxyjulis californica 12.4 0.024 0.30 11.9 0.024 0.29 Benthic transects Sebastes chrysomelas 1.7 0.36 0.61 1.8 0.36 0.65 Sebastes carnatus 1.0 0.36 0.36 0.43 0.36 0.15 Sebastes atrovirens 0.70 0.38 0.27 0.11 0.38 0.04 Sebastes mystinus 0.15 0.44 0.07 0.15 0.50 0.07 Sebastes melanops 0.21 0.44 0.09 0.03 0.44 0.01 Sebastes miniatus 0.04 2.0 0.08 0.09 2.0 0.18 Hexagrammos decagrammus 0.34 0.5 0.17 0.40 0.5 0.2 Embiotoca lateralis 0.84 0.47 0.39 0.17 0.47 0.08 Embiotoca jacksoni 0.41 0.47 0.19 0.16 0.47 0.08 Scorpaenichthys marmoratus 0.11 0.7 0.08 0.11 0.7 0.08 Ophiodon elongatus 0.13 2.6 0.34 0.09 2.6 0.23 Rhachochilus vacca 0.13 0.47 0.06 0.01 0.47 0.005 Total 9.99 kg/100 m2 = 0.0999 kg/m2 4.11 kg/100 m2 = 0.0411 kg/m2 'Mean weights from collections at Piedras Blancas Field Station, U.S. Fish and Wildlife Service, or estimated from mean total lengths. 805 FISHERY BULLETIN: VOL. 84, NO. 4 15.0 in Macrocystis and 3.9 in Nereocystis. Blue rockfish probably are the largest contributor to the total biomass of kelp forest fish communities in Cen- tral California. Miller and Geibel (1973) estimated blue rockfish densities at 6.66 fish/100 m2 in 1969 and 8.35 in 1970 in Macrocystis beds at Hopkins Marine Life Refuge, Monterey County, CA. They suggested that this represents about 50% of the ac- tual biomass because their survey method under- represented midwater species. Considering this ad- justment, my data for blue rockfish in Macrocystis forests agree well with theirs. Near Pt. Piedras Blancas, blue rockfish made up 33% and 18% of the mean number of fish within the Macrocystis and Nereocystis forests, respectively. Assuming an aver- age weight of 440 g (Table 4), blue rockfish con- tributed about 70% of the total biomass of the Macrocystis fish assemblage and about 50% of Nereocystis (species weighing a few ounces or less were not included in this analysis). The importance of juvenile blue rockfish as forage for large car- nivorous kelp forest fishes (primarily Sebastes sp.) has been well documented (Miller and Geibel 1973; Burge and Schultz 1973; Hallacher and Roberts 1985). Tagging studies have suggested that the home range of blue rockfish is relatively small (Miller and Geibel 1973). The evidence given here illustrates the important role that blue rockfish play in the kelp forest communities of central California. My estimate of the biomass of fish within each of the two canopy types (Table 4) included only species that were relatively common and of sufficient size to contribute significantly to the total. For exam- ple, although the estimated mean weight of Oxyjulis californica was only 24 g, its abundance made its total contribution rather large. My data showed that in this study area off Cen- tral California Macrocystis supported a larger stand- ing crop of fish, primarily midwater species of the genus Sebastes, than did forests of Nereocystis (Table 4). The following explanations are offered for the observed differences. These explanations are not mutually exclusive; several or all of the proposed ex- planations may have contributed to the observed patterns. 1) The amount of algae consumed by blue rock- fish fluctuates seasonally. Hallacher and Roberts (1985) showed that blue rockfish may use algae as a major source of energy during the non-upwelling period (September through March), which partly coincides with the period of minimum development in Nereocystis forests. During this period blue rock- fish may rely on Macrocystis directly as a food source, or indirectly as a substrate from which in- vertebrates are taken. The resulting increased biomass of blue rockfish in Macrocystis may help support larger numbers of other carnivorous fish. Four of the seven species that were densest in Macrocystis (Table 5) forests are known to rely heavily on juvenile rockfish for food (Hallacher and Roberts 1985). Although juvenile rockfish densities were not statistically greater in the Macrocystis forest (Table 2) because of large variations in den- sities (occurring on transects in either very large or very small schools), they were generally more avail- able in Macrocystis forests. Subsequent field obser- vations of juvenile rockfish in central California kelp forests have indicated that kelp forest rockfish recruitment may have been poor during the course of this study. Table 5. — Summary of species for which densities in the two kelp types differed significantly. Species Canopy type which presented significantly higher density Midwater Sebastes mystinus Blue rockfish Sebastes serranoides Olive rockfish Sebastes atrovirens Kelp rockfish Sebastes melanops Black rockfish Benthic Sebastes carnatus Gopher rockfish Embiotoca lateralis Striped seaperch Oxylebius pictus Painted greenling Sebastes atrovirens Kelp rockfish Macrocystis Macrocystis Macrocystis Observed on Macrocystis mid- water transects only Macrocystis Macrocystis Macrocystis Macrocystis (considered primarily as a midwater species) 2) The perennial nature of Macrocystis forests compared with the annual nature of Nereocystis forests may contribute to increased fish densities in Macrocystis forests. Macrocystis forests provide some structure throughout the year with new growth providing both vertical and canopy structure 1 to 3 mo earlier than Nereocystis. This temporal stability may afford necessary habitat structure within the water column permiting relatively higher densities of fish. 3) Differences in abiotic factors such as the physical orientation of the reef systems to oceanic swells and the resultant surge and scour effects may play a role in determining habitat suitability for some species of fish. The effects of sediment trans- port and scouring, caused by water movement, 806 BODKIN: MACROCYSTIS AND NEREOCYSTIS KELP FORESTS would be most evident at the sea floor and may in fact have contributed to the observed differences in densities in the bottom dwelling surf perch (Table 5). My data indicated that the major differences in densities of fish were in midwater species, suggest- ing that exposure to bottom disturbance per se was not a primary influence on observed patterns. 4) The differing physical characteristics of the Macrocystis and Nereocystis plants themselves may play a role in determining their suitability as habitat for kelp bed fishes. During periods of full develop- ment, within this study area, Macrocystis typically has widely spaced, thick bundles of stipes with large fronds throughout the water column, leading to a canopy that is frequently closed. Nereocystis, in con- trast, has single, frondless stipes with large terminal fronds that generally form a broken surface canopy (Fig. 1). Due to the distinct physical structure of these two plants, both within the water column and at the canopy, the foliage biomass is usually con- siderably greater within the Macrocystis forest. This abundance of structure, combined with its persis- tance over time, may enhance the carrying capacity of giant kelp forests compared with those of bull kelp (Leaman 1980). A comparison of the standing crop estimates pre- sented in this study is made with those from other marine reef systems in Table 6. While values for both Macrocystis and Nereocystis forests are below those representing fringing coral reefs (Brock 1954; Randall 1963), my estimates for Macrocystis forests compare favorably with the upper values obtained in Monterey, CA (Miller and Geibel 1973) and north- east New Zealand (Russell 1977), while the Nereo- cystis estimate corresponds to the estimates from Southern California Macrocystis forests (Quast 1968; Larson and DeMartini 1984). In conclusion, Macrocystis forests supported a biomass of fish about 2.4 times greater than that supported by Nereocystis forests (Table 4) where perennial, water column foliage provided a more persistant, structurally diverse habitat. Larger numbers of midwater fish, primarily 5. mystinus, found in the Macrocystis forest can account for this difference. ACKNOWLEDGMENTS This work was supported by the U.S. Fish and Wildlife Service, Denver Wildlife Research Center, Marine Mammal Section. I thank R. Brownell, R. Curnow, J. Estes, R. Jameson, C. Jones, M. Layman, L. Rathbun, P. Vohs, and S. Wright for their support. D. Hilger, D. Martin, F. Scott, M. Shawver, and G. VanBlaricom contributed their time as dive partners to this work. I would like to thank the members of my graduate committee— A. Roest (advisor), F. Clogston, R. Gambs, and R. Nakamura— and staff— R. Bowker and L. Maksou- dian, and the Biological Science Department, California Polytechnic University, San Luis Obispo, CA. Valuable comments on earlier drafts of this manuscript were offered by P. Eschmeyer, R. Table 6.— Comparison of biomass estimates of fish from marine communities (after Russell 1977). Location and Standing crop reference Bottom type (kg.m2) Hawaii (Brock 1954) Fringing coral reef: open sand, broken rock, coral reef, reef flat 0.001-0.0184 Virgin Islands Fringing coral reef: boulders, (Randall 1963) coral 0.160 Southern California Kelp bed: broken rocky bottom, (Quast 1968) dense algal cover 0.0351 Southern California Cobble, low relief Macrocystis (Larson and DeMartini 1984) forest Cobble, low relief kelp- 0.039-0.065 depauperate 0.024 Monterey Bay, CA Kelp bed: broken rocky bottom (Miller and Geibel (1973) dense algal cover, rocky reef 0.001 ->0.1 12 N.E. New Zealand Rocky reef: open low relief, (Russell 1977) sparse algal cover. Rocky reef: high bottom relief, <0.001 extensive algal cover 0.103 Central California Rocky reef: high bottom relief; (Present study) Macrocystis canopy 0.0999 Nereocystis canopy 0.041 1 'Average estimate. 807 FISHERY BULLETIN: VOL. 84, NO. 4 Jameson, R. Nakamura, G. Rathbun, A. Roest, and G. VanBlaricom and three exceptional anonymous reviewers. A special thanks to D. Bodkin and G. VanBlaricom for their support and encouragement. LITERATURE CITED Abbott, I. A., and G. J. Hollenberg. 1976. Marine algae of California. Stanford University Press, Stanford, CA. Brock, R. E. 1982. A critique of the visual census method for assessing coral reef fish populations. Bull. Mar. Sci. 32:269-276. Brock, V. E. 1954. A preliminary report on a method of estimating reef fish populations. J. Wildl. Manage. 18:297-308. BURGE, R. T., AND S. A. SCHULTZ. 1973. The marine environment in the vicinity of Diablo Cove with special reference to abalones and bony fishes. Calif. Dep. Fish Game, Mar. Res. Tech. Rep. 19, 433 p. Cane, M. A. 1983. Oceanographic events during El Nino. Science 222: 1189-1195. Caughley, G. 1977. Analysis of vertebrate populations. John Wiley and Sons, Lond. Dayton, P. K., V. Currie, T. Gerrodette, B. D. Keller, R. Rosenthal, and D. Ven Tresca. 1984. Patch dynamics and stability of some California kelp communities. Ecol. Monogr. 54:253-289. Dayton, P. K., B. D. Keller, and D. A. Ven Tresca. 1980. Studies of a nearshore community inhabited by sea otters. Final Report MMC-78/14. Mar. Mammal Comm., Wash., D.C., 91 p. (Available U.S. Dep. Commer., Natl. Tech. Inf. Serv., as PB81-109860.) Goodman, D. 1975. The theory of diversity-stability relationships in ecology. Q. Rev. Biol. 50:237-266. Hallacher, L. E., and D. Roberts. 1985. Differential utilization of space and food by the inshore rockfishes (Scorpaenidae: Sebastes) of Carmel Bay, Califor- nia. Environ. Biol. Fish. 12(2):91-110. Larson, R. J. 1980. Competition, habitat selection, and the bathymetric segregation of two rockfish (Sebastes) species. Ecol. Monogr. 50:221-239. Larson, R. J., and E. E. DeMartini. 1984. Abundance and vertical distribution of fishes in a cobble-bottom kelp forest off San Onofre, California. Fish. Bull, U.S. 82:37-53. Leamon, B. M. 1980. The ecology of fishes in British Columbia kelp beds. I. Barkley Sound Nereocystis beds. Fish. Dev. Rep. 22. Ministry of Environment, British Columbia, 100 p. Miller, D. J., and J. J. Geibel. 1973. Summary of blue rock fish and ling cod life histories; a reef ecology study and a giant kelp, Macrocystis pyrifera, experiments in Monterey Bay, California. Calif. Dep. Fish Game, Fish Bull. 158, 137 p. Pielou, E. C. 1966. Species-diversity and pattern-diversity in the study of ecological succession. J. Theoret. Biol. 10:370-383. Quast, J. C. 1968. Estimates of the population and standing crop of fishes. In W. J. North and C. L. Hubbs (editors), Utilization of kelp bed resources in southern California, p. 57-79. Calif. Dep. Fish Game, Fish Bull. 139. Randall, J. E. 1963. An analysis of the fish populations of artificial and natural reefs in the Virgin Islands. Caribb. J. Sci. 3:31- 47. Russell, B. C. 1977. Population and standing crop estimates or rocky reef fishes of northeastern New Zealand. N.Z. J. Mar. Freshw. Res. 11:23-36. 808 LIFE HISTORY AND LARVAL DEVELOPMENT OF THE GIANT KELPFISH, HETEROSTICHUS ROSTRATUS GIRARD, 1854 Carol A. Stepien1 ABSTRACT Life history data from about 1,200 giant kelpfish, including age, length, and weight relationships, are described and analyzed. Additionally, differences in habitats and behavior between larvae, juveniles, and adults are reported. Female giant kelpfish were found to be larger than males at given ages past sexual maturity. Age data indicate that females live longer and all individuals larger than 28 cm TL collected in this study were females. Males guard the algal nests until hatching, about 2 weeks after spawning. Giant kelpfish from nests collected in the field were reared in the laboratory, surviving for up to 9 months. Feeding and development of laboratory-reared larvae were compared with field-collected specimens. In situ, they school in the kelp canopy until 2 months old, gradually developing juvenile coloration and becom- ing increasingly thigmotactic and solitary. Giant kelpfish reach sexual maturity at 1-1.5 years, at which time they commence to defend territories in given plant habitats. The cryptically colored giant kelpfish, Heterostichus 7,ostratus, is abundant in southern California kelp forests and surrounding subtidal plant habitats. Heterostichus is one of the largest members of the clinid family, reaching a length of 41.2 cm and an age of 5 yr (J. E. Fitch in Feder et al. 1974). Al- though ranging from British Columbia, Canada, to Cape San Lucas, Baja California, Mexico, it is most commonly found from Point Conception to central Baja in depths of 35 m (Roedel 1953). Giant kelpfish occur in three different colormorphs— red, brown, and green— which closely match the color of their surrounding plant habitats (Hubbs 1952; Stepien 1985, 1986). They additionally exhibit four different dark melanin patterns, which appear superimposed on the basic color of the fish and, unlike color- morphs, can change rapidly (Stepien 1985, 1986). Giant kelpfish spawn year-round, but most fre- quently during spring months (Limbaugh 1955; Feder et al. 1974). The eggs are attached to algal nests with entangling threads that extend from the egg membranes (Holder 1907; Feder et al. 1974). The males alone guard the nests from predators until hatching, averaging 2 wk after spawning (Coyer 1982). Giant kelpfish are relatively well- developed at hatching and are planktonic for several weeks. They school in the kelp canopy until they are about 6 cm long, then develop juvenile coloration department of Biological Sciences, University of Southern California, Los Angeles, CA 9008S; present address: Marine Biology Research Division A-002, Scripps Institution of Ocean- ography, University of California at San Diego, La Jolla, CA 92093. and become solitary, living close to nearshore algae (Limbaugh 1955). Although Heterostichus larvae are not uncommon in the nearshore ichthyoplankton, their development has not been previously described. Heterostichus egg morphology was described by Barnhart (1932), and the egg-laying process was described by Holder (1907). Matarese et al. (1984) published two draw- ings of kelpfish larvae. Although diet and some aspects of general life history have been described qualitatively by several investigators (Hubbs 1920, 1952; Roedell 1953; Limbaugh 1955; Quast 1968; Hobson 1971; Feder et al. 1974; Hobson et al. 1981; Coyer 1982) and one quantitative study was con- ducted on feeding and distribution of juveniles and adults in giant kelp (Coyer 1979), specific morpho- metric data for larval, juvenile, and adult stages have not previously been reported. This paper presents life history data, including the following: 1) Differences in larval, juvenile, and adult habitats and behavior; 2) size, weight, and age relationships, in- cluding differences between males and females; and 3) the sequence of larval development and meta- morphosis. MATERIALS AND METHODS Collection and In Situ Observations In situ observations were made during approx- imately 280 scuba dives from 1978 to 1983, the majority in the vicinity of the University of South- ern California's Catalina Marine Science Center Manuscript accepted March 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 809 FISHERY BULLETIN: VOL. 84, NO. 4 (CMSC) on Santa Catalina Island (Fig. 1). Most observations and collections were made in protected cove areas having well-developed kelpbeds of the giant kelp, Macrocystis pyrifera, and associated plant habitats, including surfgrass, Phyllospadix tor- reyi, and red and brown algae. Approximately 1,200 giant kelpfish were observed during the course of the study. The aging and sexing study material from Catalina was also supplemented by 42 specimens col- lected from subtidal sites off the southern Califor- nia mainland, including Ventura, Lunada Bay on the Palos Verdes Peninsula, Huntington Beach, and La Jolla (Fig. 1). Kelpfish were collected using a 0.5 x 0.8 m net, mounted on a 1 m long handle and constructed of 0.25 cm mesh dyed either brown or red to match the kelpfish algal habitats (it was found that white netting alarmed the fish, making them difficult to —i — 119° — T" 118° —l — 117° t N , Ventura 34° -33' Los • Angeles Santa Monica\ Bay \ * Long Beach LUoadaBBV ^ v. Huntington ^\4 Beach ^Q^U Newport Bay y. C$ v — J Catalina Island. kfOceanside A COLLECTION SITES La Jolla .San 'Diego u 20 _i_ 40 MILES " — i — i — i — i — i — r- 0 20 40 60 KILOMETERS Figure 1.— Giant kelpfish collection sites (open triangles) off the southern California coast. 810 STEPIEN: LIFE HISTORY AND DEVELOPMENT OF KELPFISH catch). Kelpfish were collected by sliding the net for- ward and downward over the fish. Collection of kelp- fish was facilitated by their habit of hiding in algae when pursued rather than escaping by rapid swim- ming. Those that were actively swimming (usually through the kelp canopy) were less frequently cap- tured. They were placed in a collecting bucket having mesh sides, a snap-on lid, and a funnel entry- way, preventing escapes when the lid was opened for other fish. Care was taken to avoid putting the larger kelpfish in the same bucket as the smaller ones, because the smaller ones were occasionally eaten by the larger ones. Life History Data From Juveniles and Adults In the present study, 140 juveniles and adults of representative sizes (ranging from 10 to 42 cm TL) were measured live to the nearest 0.1 cm. Total length (TL) was found to be more quickly measur- able than standard length (SL). Both SL and TL were measured, in order to allow comparisons with other studies. Kelpfish were weighed to the nearest 0.1 g on a triple-beam balance while briefly con- tained in plastic bags, in which they were quiescent and unabraded. These data were graphed, and regression and F-test analyses were performed (Sokal and Rohlf 1981; Zimmerman and Kremer 1983). The fish were sexed and aged. Females had clear or pink, rounded ovaries and most individuals over 14 cm TL had clearly visible developing eggs. Male gonads were cream-colored and had a characteristic ventral groove. In cases when sex of juveniles was questionable, the gonads were examined under a dissection microscope. Otoliths (sagitta) were removed and stored dry in labeled glass. They were briefly submerged in water and examined against a black background with a dissecting microscope (25-50 x magnification) for ring counting (Fig. 2). Ages were determined by counting alternating white (opaque) and translucent (hyaline) bands, each representing 6 mo of growth, KELPFISH OTOLITH B NUCLEUS 1 YEAR OLD 2 YEARS OLD 3 YEARS OLD 4 YEARS OLD O I .1 _L- .2 _l .7 CM. Figure 2.— (A) Photograph of otolith (actual length = 6.5 mm) of a 4 yr-old female giant kelpfish, 33 cm TL. (B) Drawing of otolith (sagitta) showing ring counts. 811 FISHERY BULLETIN: VOL. 84, NO. 4 using standard methods outlined by Fitch (1951), Jensen (1965), and Collins and Spratt (1969). Each pair of otoliths was read independently by me and another reader, neither knowing the identity of the fish. Our age estimates were in agreement in 80% of the examinations. When differences in ring count occurred, a joint reevaluation was made. Total length versus age comparisons were graphed, and regression analysis and F-tests were performed on the log-log transformations. Mean sizes of male and female kelpfish in age classes where differences appeared to occur were tested for significance using i-tests and 2-way ANOVA. Sep- arate regression equations were also calculated for males and females, and ANCOVA was performed to determine whether the distributions were signifi- cantly different (Sokal and Rohlf 1981). Seasonal population structure was estimated from collection data taken from February 1981 through January 1983. Kelpfish were grouped in six size classes. Distribution of kelpfish in size classes was analyzed for significant seasonal variations using contingency tables and G-tests (Sokal and Rohlf 1981). Larval Rearing Nine giant kelpfish nests were collected, four in spring 1980 and five in spring 1982, off Santa Cata- lina Island. Both parents of the eggs were collected in three cases when spawning was observed. In six cases, only the male parents, which were guarding the nests, were collected. Eggs were also laid in the laboratory on five separate occasions, but did not hatch normally, apparently because of inadequate dispersion in the nests. Algal nests containing eggs were suspended from a glass rod connected to an electric stirring device, simulating wave motion in shallow subtidal habitats (Fig. 3A). This method substantially decreased bacterial and fungal attacks. Parents were not kept with the eggs, as both males and females were sometimes found to eat eggs in the laboratory. Nests were placed in aerated 190 L plastic containers cooled in 1 m deep aquaria of running seawater. Filtered seawater in the containers was replaced every few days. Several eggs were removed daily for examination of development. Newly hatched larvae were isolated in lightly aerated 76 L brown plastic containers bathed in large aquaria. Kelpfish larvae were fed laboratory- raised Brachionus plicatilis (marine rotifers) within 24 h after hatching. Brachionus plicatilis were cul- tured in high densities of the green flagellate, Tetraselmis tetrahele, which was grown in a nutrient-rich medium under constant light, follow- ing methods developed for feeding northern anchovy larvae (Theilacker and McMaster 1971). Brachionus plicatilis, ranging from 0.01 x 0.02 mm to 0.07 x 0.20 mm in size, were maintained in the larval kelp- fish containers at concentrations of 10-40/mL. At age 1 wk, kelpfish larvae were changed from closed to open containers of filtered and aerated running seawater, having two 20 x 30 cm panels of 100 ^m mesh. After age 2 wk, kelpfish larvae were also fed wild plankton, which primarily contained various devel- opmental stages of the copepod Acartia sp. (92% wet weight) and some barnacle nauplii and cyprid larvae (7% wet weight). Wild plankton were col- lected using a submersible pump attached to a float off the laboratory pier. A light was suspended over the pump and the system connected to an electrical timer. Plankton were filtered through a 335 pm mesh bag into a 190 L plastic container. The con- tainer had a removable inner 100 pm mesh lining and a spillover pipe, retaining only appropriate-sized plankton between the two filter bags (Fig. 4). Best copepod catches were obtained from dusk to 2 h after sunset. Running filtered seawater and an aerator were used to maintain temperature and oxygen levels in the collecting container until the fish larvae were fed the following morning. Den- sities averaged 1-3/mL, which have been shown to support high survival rates in laboratory rearing of other fish larvae (Houde 1973; Hunter 1981). When plankton catches were low, giant kelpfish diet was supplemented with cultured Artemia salina (brine shrimp) nauplii. Brachionus plicatilis were discontinued after age 3 wk and plankton continued until age 3 mo. After age 2 mo, diet was supple- mented with frozen adult brine shrimp, Tetramin2 commercial flake food, and live mysids captured from net tows in kelpbeds. Ten larvae were removed every 2 d during the first 2 wk of development for measurement and description. After this period, 10 larvae were ex- amined weekly until 2 mo had elapsed. All measure- ments were made on fresh material. Drawings of several stages of larval development were made using a camera lucida and a dissecting microscope. Gut contents of three specimens from each weekly sample through age 4 wk were analyzed. While viewing with a dissecting microscope, guts were dissected away from the body and food particles 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service. NOAA. 812 STEPIEN: LIFE HISTORY AND DEVELOPMENT OF KELPFISH Figure 3.— (A) Giant kelpfish nest in aquarium, attached to an electric stirrer, which simulated wave motion. (B) Photograph of nest with eggs in brown algae, taken with 70 mm macrolens. (C) Photograph under compound scope of 24-h kelpfish egg show- ing blastodisc, egg diameter = 1.4 mm. (D-F) Developing kelpfish eggs photographed under dissection microscope (diameters = 1.4 mm). (D) 72 h after spawning. (E) 10 d after spawning. Note adhesive threads attaching egg to red alga. (F) 12 d after spawning. 813 FISHERY BULLETIN: VOL. 84, NO. 4 PLANKTON COLLECTOR DESIGN LID WITH HOLE INNER MESH BAG OUTER MESH BAG TIMER LIGHT FLOAT ^SUBMERSIBLE PUMP Y£.s« Figure 4.— "Automatic" plankton collector design for feeding giant kelpfish larvae. Plankton were attracted to light on timer after dark. Submersible pump, suspended beneath the float, pumped plankton into large plastic container on dock. Plankton ranging from 100 to 325 ^m were filtered between two mesh bags. Aeration and running seawater kept the plankton alive. teased out using either a single human hair or a modified paint brush from which only a few long strands protruded. Gut contents were viewed under a compound microscope and identified, and average lengths and widths of prey items were recorded. At age 2 mo, the kelpfish larvae were moved to containers having 0.3 cm mesh panels and contain- ing artificial plant habitats (see Stepien 1985 and 1986). They were subsequently measured bimonth- ly and their development described. Development and feeding of laboratory-reared kelpfish larvae were also compared with 20 field-collected in- dividuals. Kelpfish larvae of various ages and sizes were collected in hand nets while night-lighting from a dock and while scuba diving in kelpbed canopies using a 1 mm mesh handnet. Other kelpfish larvae were examined from bongo net collections made in Santa Monica Bay in 1982. Their development was compared with similar-sized laboratory-reared lar- vae. Gut contents of four early-stage larvae (esti- mated 0-9 d old) were analyzed for food types and sizes, in comparison with laboratory-reared kelp- fish. RESULTS Spawning Giant kelpfish nests were guarded by the male parent, the eggs being interspersed and held by adhesive threads in either red or brown algae (Fig. 3B). Seven of the nine nests collected were located in isolated clumps of algae, and all were found be- tween 6 and 12 m deep. Kelpfish nests were most common in the red alga Gelidium nudifrons (6 of 9 nests collected) in areas where clumps of taller brown algae covered patches of red algae. Three of the nests were located in brown algae, two in Cys- toseira neglecta, and one in Sargassum muticum. The male parent hid in the overlying clump of brown algae, emerging to chase away intruding fishes. Male kelpfish were observed to defend their nests against other kelpfish, sheephead, and rock wrasse. Female kelpfish may spawn several times a year since a female kept in the laboratory laid eggs twice within 3 mo. Gonads of all females examined after spawning were almost entirely spent. Since 814 STEPIEN: LIFE HISTORY AND DEVELOPMENT OF KELPFISH all eggs in the nests examined were in similar stages of development, it is likely that each nest contains the eggs of a single female. After spawning (the behavior sequence of which is described in Coyer 1982), the male kelpfish chases away the female parent, as was observed in the laboratory on three separate occasions. In one case, the male's repeated pursuits resulted in the female jumping out of the aquarium. Eggs occurred in two different colors, red and brown, which microscopic examination showed was due to color of the yolk. All eggs in a given nest were either red or brown and remained that color throughout development. Nest and egg color did not always match. Brown eggs were found in four nests of red algae and two nests of brown algae, while red eggs were found in two nests of red algae and one nest of brown algae. Fertilized eggs laid in the laboratory developed poorly and few of them hatched, apparently due to abnormal dispersal in the algal nests by the females. In all three cases of laboratory spawnings, eggs were laid in clumps rather than being well-spaced throughout the algae, as observed in field-collected nests. Freshly laid nests were collected in the field on three occasions from pairs that had just com- pleted spawning. Two of the three spawning females were brown colormorphs and one was a red morph, but all three showed the barred melanin pattern. All nine field-collected male parents were brown color- morphs exhibiting the characteristic male nuptial striped melanin pattern (Coyer 1982; Stepien 1985, 1986). Egg Development and Hatching Eggs from freshly laid nests hatched in 12-17 d at 18°C, the largest number hatching in 13 d. Eggs averaged 1.4 mm in diameter and nests contained an average of 700 eggs, ranging from 400 to 1,200 eggs. An estimated 800 of the 1,200 eggs hatched from the most successful laboratory incubation. Nests that were rotated vigorously and kept well- aerated produced the most successful hatchings. The sequence of egg development is summarized in Table 1 and photographs of the developing eggs are shown in Figure 3. Hatching occurred from day 14 through day 15. Hatching took about 20 min, the larvae emerging head-first from the egg membrane. Early Larval Development (Prenotochord Flexion) Giant kelpfish larvae can be distinguished from other southern California clinid larvae by their large numbers of myomeres, averaging 55-59. Newly hatched larvae had large yolk sacs and well- developed mouths, guts, melanophores, and fin folds and averaged 6.2 mm TL (Fig. 5A). Larvae floated upside-down, yolk up, for the first 24-36 h after hatching. They swam with wriggling movements, lasting about 30 s, interspersed with longer periods of inactivity, lasting up to several minutes. Yolk sacs were present 36-48 h after hatching. Two- day-old larvae averaged 7.0 mm TL and swam strongly upright, showing positive attraction to light and concentrating near the white mesh areas of the containers. After 4 d, the larvae were less positive- ly photo tactic, concentrating towards the bottom of the containers. Mean sizes and a summary of the sequence of larval development are listed in Table 2. Illustrations of larvae are found in Figure 5. Later Larval Development (Postnotochord Flexion) Flexion of the notochord had begun by 7-9 d and an average size of 8.5 mm (ranging from 7.6 to 8.9 mm, N = 12). Field-collected giant kelpfish larvae also showed the beginnings of notochord flexion at a similar size (7.4-9.3 mm, N = 5). Size at notochord flexion is smaller than that reported by Matarese et al. (1984) for other clinid larvae. Two-week-old giant kelpfish larvae began swim- ming in organized schools, which also were observed in situ in giant kelp canopies. Other researchers have also noted this phenomenon (Feder et al. 1974), which was not observed in giant kelpfish past the age of 2 mo in both the laboratory and the field. By 3 wk, the schooling larvae became progressively more difficult to catch with dip nets, exhibiting well- Table 1.— Summary of kelpfish egg developmental stages. Developmental features well-developed blastodisc, beginnings of epiboly head fold apparent, neural tube forming embryo wrapped 180° around egg's circumfer- ence; notochord, somites, eyes, and lenses visible embryo wrapped 240° around egg's circumfer- ence, myomeres well-developed, lenses of eyes pigmented, heart beating 95 times per minute yolk shrunk to 1/2 size of egg; embryo curled 1.5 times around egg; mouth differentiated; gut, liver, and inner ear developing otoliths and pectoral and dorsal fin folds visible, vigorous tail movements, heart beats 90 to 100 times per minute hatching at 18°C, larva exits head-first, hatching takes about 20 min Time after spawning 24 h 36 h 72 h 6d 10 d 12d 14 d 815 FISHERY BULLETIN: VOL. 84, NO. 4 816 STEPIEN: LIFE HISTORY AND DEVELOPMENT OF KELPFISH Table 2.— Mean sizes (TL, mm) and developmental stages of laboratory-reared giant kelpfish larvae 0-60 d. Age1 (d) Mean length Range (TL, mm) No. Developmental features 0 6.2 6.0-6.5 10 2 7.0 6.7-7.5 10 4 7.7 7.0-8.0 5 7.9 7.4-8.4 10 7 8.3 7.5-9.0 5 9 8.8 7.9-9.4 10 11 9.7 7.9-10.7 10 13 10.3 8.2-11.2 5 15 10.9 9.5-11.7 10 17 11.4 10.0-12.3 10 19 11.5 10.1-12.5 5 21 11.7 10.6-13.6 10 23 12.0 11.4-14.3 10 25 12.2 10.3-16.8 10 30 16.8 13.0-19.0 10 39 23.8 18.0-28.0 5 46 25.6 22.0-27.0 10 53 25.7 18.0-35.0 10 60 30.6 25.0-37.0 10 'Age (d) after hatching. well-developed mouth, gut, and fin folds; 12 postanal serial melanophores 12-19 postanal melanophores, first feeding, yolk sac 1/3 original size 20 postanal melanophores, 2 melanophore spots on liver, melanophores dorsal to anus, yolk sac disap- peared some ventral caudal fin rays visible, gill rakers formed, operculum visible notochord flexion begun in some notochord flexion completed, swim bladder formed caudal fin rays well-developed schooling behavior is pronounced fin rays in rear of dorsal and anal fin folds scattered melanophores on top of head and lower jaw, melanophores over gut well-developed schooling and avoidance behavior pectoral, dorsal, and anal fin rays formed continuous line of stellate melanophores above the gut pelvic fins beginning to develop, melanin pigmentation in pelvic region orange xanthophore pigmentation on top of the head, over the gut and at the base of the caudal fin; teeth visible pelvic fins formed, 32 postanal ventral melanophores larvae are pale gold in color schooling no longer pronounced most have settled onto algae developed avoidance patterns and fright responses. By 5 wk, schooling was no longer as pronounced and the larvae were observed to stalk their copepod prey very efficiently. Larval Feeding Unless giant kelpfish larvae were given food with- in the first 48 h, a point of no-return was reached, after which they starved to death even if given food. Best results were obtained if larvae were fed within 24 h of hatching. Brachionus (rotifers) and Tetra- selmis (algae) were found in the guts of 2-d-old lar- vae in the laboratory. Three-day-old larvae, even those still having yolk sacs, contained an average of 5.6 Brachionus and 2.9 Tetraselmis (Table 3). High mortality (nearly 60% of those hatched) oc- curred after hatching and through day 5. Dead lar- vae examined had apparently never eaten, despite relatively high levels of appropriately sized food items. Gut contents of field-collected kelpfish larvae (estimated to range from 0 to 9 d old) showed that they fed on a wide variety of food items, including single-celled algae, rotifers, mollusk larvae, and bar- nacle and copepod larvae (Table 4). Similar sizes and quantities of food items were consumed by both the laboratory-reared and field-collected larvae (Tables 3,4). Significantly larger food items were consumed by 2-wk-old laboratory- reared larvae, the largest widths being 52% of the mouth size (Fig. 6). Larger copepods were eaten more frequently than rotifers, although both food items were present in guts (Table 3). High mortality (ranging from 20 to 40%) also occurred at about 2.5 wk of age in both the 1980 and 1982 rearing experiments. At this age, gut ex- aminations indicated that the larvae were switch- ing from the smaller prey (rotifers and algae) to the larger copepods. Older larvae progressively con- sumed larger copepods whose size reached 70% of the mouth width by week 3 (Fig. 6, Table 3). Figure 5.— Drawings of laboratory-reared giant kelpfish larvae, made with camera lucida and dissection microscope. (A) Day 0 (after hatching), 6.1 mm TL. (B) Day 4 after hatching, 7.0 mm TL. (C) Day 7 after hatching, 8.4 mm TL. (D) 2 wk, 10.9 mm TL. (E) 3 wk, 11.6 mm TL. (F) 5 wk, 22.2 mm TL. Settlement and Metamorphosis After 8 wk and at a mean length of 30.6 mm, giant kelpfish larvae had well-developed, pale gold-brown pigmentation. They became increasingly thigmotac- 817 FISHERY BULLETIN: VOL. 84, NO. 4 Table 3.— Gut contents of laboratory-reared giant kelpfish larvae, 3 d to 5 wk, indicating mean numbers and sizes of prey items. N = 18 (N = 3/sample). Laboratory diets 0-3 wk consisted of Tetraselmis and Brachionus. Acartia copepods were added to the diet at 2 wk. Sizes of kelpfish (TL, mm) and mean sizes of prey items (width x length) given. larvae Size (mm) 6.8 7.1 7.4 Prey items Kelpfish Tetraselmis algae Brachionus rotifers Acartia copepods Age Mean No. 2.9 Size (mm) Mean No. Size (mm) Mean No. Size (mm) 3d 0.039 x 0.120 5.6 0.10 x 0.149 — 1 wk 8.0 8.2 8.2 3.3 0.050 x 0.120 14.7 0.103 x 0.157 2 wk 9.4 9.7 10.8 10.0 0.078 x 0.130 10.2 0.160 x 0.220 1.2 0.100 x 0.390 3 wk 10.7 11.5 13.6 " 6.8 0.130 x 0.195 2.4 0.221 x 0.520 4 wk 11.9 13.3 16.4 3.3 0.221 x 0.520 5 wk 19.7 22.0 23.0 7.9 0.220 x 0.850 Table 4.— Gut contents of field-collected giant kelpfish larvae 6.24-8.2 mm TL. N = 4. Mean TL of kelpfish = 6.93 mm (range 6.24-8.82 mm). Mean mouth width = 0.42 mm (0.40-0.44 mm). Mean no./ Mean width Mean length Food iterr I larva and range (mm) and range (mm) Diatoms 3.00 0.03 (0.01-0.07) 0.06 (0.04-0.08) Dinoflagellates 2.00 0.03 (0.01-0.07) 0.04 (0.02-0.20) Tintinnid protozoans 0.75 0.04 (0.03-0.07) 0.13 (0.10-0.16) Rotifers 0.75 0.08 (0.03-0.13) 0.19 (0.08-0.35) Barnacle nauplii and cyprids 0.75 0.10 (0.07-0.13) 0.16 (0.12-0.23) Copepod nauplii and copepodites 3.50 0.12 (0.07-0.21) 0.40 (0.14-0.46) Mollusk larvae 1.00 0.11 (0.09-0.12) 0.25 (0.22-0.29) Nemertean worms 0.25 0.10 0.34 Siphonophores 0.25 0.29 0.30 tic during the next few weeks, darting amongst the artificial plants placed in their containers. Similar- ly, kelpfish individuals observed in situ had "settled" onto juvenile habitats by 30-50 mm TL. Juvenile habitats included the fronds of giant kelp; the brown alga, Sargassum muticum; and green surfgrass. Juveniles were usually in loose aggregations of three to seven similar-sized individuals until reaching a size of 7-9 cm TL. At 5-7 cm (between 2 and 4 mo), laboratory-reared and field-collected giant kelpfish lost their trans- parent light gold-colored appearance, developing either green, gold, or brown pigmentation depend- ing on their juvenile habitat, whether surfgrass, kelp, or Sargassum. The majority of juveniles found in surfgrass were green with striped or mottled melanin patterns and had silvery horizontal patches. Those in kelp were usually plain or mottled gold- brown with gold bellies while those in Sargassum developed brown pigmentation and barred or mot- tled melanin patterns (see Stepien 1985 and 1986 for detailed descriptions of color patterns). Morphometries of Larvae, Juveniles, and Adults The SL and TL of giant kelpfish larvae were linearly related (Fig. 7). Early growth (to 40 d) of laboratory-reared larvae was logarithmic (Fig. 8A) while length and age were linearly related between 1 and 9 mo of age (Fig. 8B). Otoliths of laboratory- reared kelpfish showed abnormal ring patterns, 818 STEPIEN: LIFE HISTORY AND DEVELOPMENT OF KELPFISH MEAN MOUTH WIDTH AND MAXIMUM PREY WIDTH 10 r E E 5 N* 15 N (ea. sample) ■ 3 0 12 3 4 Age (Weeks) Figure 6.— Mean mouth width and maximum prey width consumed by laboratory-reared giant kelpfish larvae 0-4 wk old. N = 15 (N each sample = 3). having several "checks" (false rings). Maximum age reached by laboratory-reared kelpfish in these ex- periments was 9 mo, at which time they succumbed to a bacterial infection. Weight versus length of juvenile and adult kelp- fish was exponentially related (Fig. 9), and SL and TL were directly linearly related (Fig. 10). Length versus age determinations also followed an exponen- tial curve (Fig. 11). Sexual maturity occurred at a mean size of 18.6 cm TL and an age of 1-1.5 yr. Regressions of sizes of adult males and adult females on age class were found to be significantly different using ANCOVA (see Fig. 11 legend). When sizes at given ages were compared using £-tests, females were found to be significantly larger than males at given ages past 2 yr (Fig. 12). The largest males sampled in this study were not older than 3 yr or larger than 28 cm TL. In contrast, large females, reaching ages of 4.5 yr and sizes of 42 cm (TL) were collected. Larger individuals collected throughout the 5-yr sampling regime were consis- tently females. Population Structure Seasonal size class structure of the giant kelpfish population was consistent over 2 yr of regular sam- STANDARD LENGTH (SL) VS. TOTAL LENGTH (TL) OF KELPFISH LARVAE 30r- 10 14 18 22 Total Length (mm) 26 34 Figure 7.— SL (mm) versus TL (mm) of laboratory-reared giant kelpfish larvae 0-30 d old. * = one fish. N = 108. Regression equation: SL = 0.598 + 0.819 (TL). F = 11,588.62, P < 0.00001. 819 FISHERY BULLETIN: VOL. 84, NO. 4 LOG LARVAL LENGTH VS. AGE (0 - 40 DAYS) E c TO o 2 4 6 7 9 T1 13 15 17 21 23 25 Age (days) 30 B LENGTH VS AGE (0 - 9 MONTHS) 3 4 5 Age (months) Figure 8.— (A) Log length (TL, mm) versus age (0-40 d) of laboratory-reared giant kelpfish larvae. * = one fish. N = 130. Regression equation: Log TL = 0.814 + 0.013 (days). F = 1,211.9, P < 0.0001. (B) Growth of laboratory-reared giant kelpfish (0-9 mo), length (cm) versus age (months). * = one fish. N = 100. Regression equation: TL = 0.379 + 1.482 (months). F = 2,230.8, P < 0.0001. pling (Fig. 13). Contingency tests of independence showed that numbers of individuals in various size classes differed significantly with season in 1981-82 and 1982-83. Juveniles appeared in significant num- bers during the spring and summer months. These data agreed with observations on spawning and appearance of larvae in the water column, indicating that most Catalina Island kelpfish in these years spawned from January through May. During spring and summer, a large portion of the population was estimated to be 1 and 2 yr old, composed of in- dividuals of reproductive age. During the fall 820 STEPIEN: LIFE HISTORY AND DEVELOPMENT OF KELPFISH LOG WEIGHT VS. LOG LENGTH OF JUVENILES AND ADULTS 2.6 1.0 1.1 1.2 1.3 Log Total Length (cm) 1.4 1.5 1.6 Figure 9.— Log weight (g) versus log TL (cm) of juvenile and adult giant kelpfish. * = one fish. N = 140. Regression equation: Log weight = -2.508 + 3.243 flog TL). F = 3,622.7, P < 0.0001. 45 r- STANDARD LENGTH (SL) VS. TOTAL LENGTH (TL) OF JUVENILES AND ADULTS Total Length (cm) Figure 10.— SL (cm) versus TL (cm) of juvenile and adult giant kelpfish. * = one fish. A = 11 fish. N = 140. Regression equation: SL = -0.580 + 0.906 (TL). F = 15,993.0, P < 0.0001. 821 FISHERY BULLETIN: VOL. 84, NO. 4 LOG LENGTH VS. LOG AGE .7 -.6 .5 -.4 -.3 .2 -.1 0 .1 .2 .3 .4 LOG AGE CYEARS) 8 .9 1.0 Figure 11.— Log TL (cm) versus log age class (years) of juvenile and adult giant kelpfish (males and females). * = one fish. N = 137. Regression equation: Log TL = 1.234 + G.528 (log age). F = 1,589.28, P < 0.0001. Regression equation for females only (N = 77): Log TL = 1.234 + 0.561 Gog age); F = 1,460.7, P < 0.0001. Regression equation for males only (N = 60): Log TL = 1.235 + 0.453 (log age); F = 535.0, P < 0.001. ANCOVA regression analysis of log TL for males and females (two different groups versus log age class (years): F = 5.82 (P < 0.05). MEAN LENGTH VS. AGE OF FEMALES AND MALES Figure 12.— Mean TL (cm) versus age class (years) of female and male giant kelpfish. Significant differences between male and female mean sizes indicated. * = Significant difference in t-test results (0.05 level). Standard error bars shown. N = 137. Two-way ANOVA with replication for mean lengths of male and female kelpfish at three ages (2.0, 2.5, and 3.0 yr) showed signifi- cant differences between the sexes (F = 38.52, P < 0.001) and the age classes (F = 78.01, P < 0.001), but no inter- action (sex x ages; F = 3.37). 1.0 1.5 2.0 2.5 Age (.5 years) 3.5 822 STEPIEN: LIFE HISTORY AND DEVELOPMENT OF KELPFISH SIZE FREQUENCIES OF KELPFISH 50 40 30 - >20 o a» 10 3 a> £ 0 0> re § 50 S. 40 30 H 20 10 Feb- Apr 81 N=63 May-Jul 81 N =111 Aug -Oct 81 N = 128 Nov 81 -Jan 82 N = 61 ■F3*- 5 15 25 35 5 15 25 35 5 15 25 35 5 15 25 35 Feb- Apr 82 N = 74 May-Jul 82 N = 153 AugOct82 Nov82-Jan83 N--130 N--24 15 25 35 5 15 25 35 5 15 25 35 Total Length (cm) 15 25 35 Figure 13.— Percentage frequencies of giant kelpfish size classes collected seasonally from February 1981 to January 1983. N = 744. Contingency table R x C G tests of independence showed significant seasonal variations in frequencies of kelpfish size classes in 1981-82 (N = 363; *2 = 167.73, 15 df, P < 0.001) and 1982-83 (N = 381; x2 = 86.07, 15 df; P < 0.001). (Sokal and Rohlf 1981.) months, the most abundant size classes were esti- mated as 0.5 and 1.5 yr of age. These size frequen- cies also indicate that a relatively low percentage of the population is composed of individuals 3 yr and older. DISCUSSION Reproduction and Development Unlike Heterostichus, whose nests contain eggs in similar stages of development, those of the fringe- head Neoclinus bryope (family Clinidae; subfamily Chaenopsidae) contain various developmental stages, apparently from several spawnings (Shiogaki and Dotsu 1972). Heterostichus eggs have a single large oil globule (see Barnhart 1932 and Figure 3C), while other described clinid eggs have several (Spar- ta 1948; Shiogaki and Dotsu 1972; Matarese et al. 1984). Unfertilized eggs of Gibbonsia elegans con- tain a mass of 6-16 small oil globules (Stepien3). Like Heterostichus (see Figure 3D), Clinus argentatus eggs develop large black melanophores over the sur- face (Sparta 1948). Early larval development in other clinids resem- bles that of Heterostichus, although few species have been studied and none have been reared past the yolk-sac stage. Other clinids are reported to hatch at similar sizes and at comparable development (Sparta 1948; Shiogaki and Dotsu 1972; Matarese et al. 1984). As in Heterostichus, the yolk-sac stage persists for 2-3 d (Shiogaki and Dotsu 1972), caudal fin rays develop first (Matarese et al. 1984), and dor- sal and anal fin rays form posteriorly to anteriorly (Risso 1948; Shiogaki and Dotsu 1972; Matarese et al. 1984). Flexion of the notochord appears to occur at a smaller size in Heterostichus (mean 8.5 mm TL) than in some other clinids (by 11.1 mm TL in Neo- clinus and 11.52 mm TL in Clinus argentatus) (Spar- ta 1948; Shiogaki and Dotsu 1972). 3Stepien, C. A. 1986b. Life history of the spotted kelpfish, Gib- bonsia elegans Cooper. Unpubl. manuscr. Marine Biology Research Division A-002, Scripps Institution of Oceanography, University of California at San Diego, La Jolla, CA 92093. 823 FISHERY BULLETIN: VOL. 84, NO. 4 Swimming behavior of newly hatched kelpfish lar- vae, characterized by short periods of swimming interspersed with longer periods of inactivity, is common in many small marine yolk-sac larvae (Hunter 1972; Ellertsen et al. 1980; Weihs 1980). Like kelpfish, some other newly hatched larvae in- cluding cod, Gadus morhua, (Ellertsen et al. 1980) and white seabass, Atractoscion (Cynoscion) nobilis, (Orhun4) swim upside-down for the first 24 h after hatching. This behavior is due to positive buoyancy of the yolk (Hunter5). Kelpfish larvae, in situ as well as in the laboratory, schooled between 2 wk and 2 mo of age. Larval schooling is common in species of nearshore fishes which also school as adults (Smith 1981; Hunter 1981) and may serve to in- crease the probability of locating patches of food and/or may help them avoid predation. No reference to larval schooling in fishes that do not school as adults was found in the literature. Larval Feeding A point of no-return at which starvation occurs even if larvae are fed appears to occur earlier in giant kelpfish (after 36 h) than in fish larvae hatch- ing from pelagic eggs (Hunter 1981) and is probably due to their greater degree of development at hatch- ing (i.e., smaller yolks and well-developed mouths and digestive tracts). Only a small number of species are sufficiently developed to consume exogenous food shortly after hatching (Balon 1984a, b). Early feeding during the yolk-sac stage may be critical for the larvae to develop a "search" image and capture skills (Hunter 1981). In this study, high mortality following the yolk- sac stage was apparently due to starvation, despite relatively high levels of appropriate-sized food items. In many marine fishes, relatively low feeding suc- cess is apparently common in field-collected, as well as laboratory-reared, larvae (Hunter 1981). During the first week, field-collected, as well as the labora- tory-reared, larvae consume a wide variety of food items, primarily smaller ones such as unicellular algae. Like Heterostichus, most species of larval fishes have been found to eat many more small prey items than larger ones (Hunter and Kimbrell 1980; Hunter 1981). "Orhun, R. M. 1986. Culture and growth of larval and early juvenile white seabass, Atractoscion (Cynoscion) nobilis. M.S. Thesis in preparation, Center for Marine Studies, Department of Biology, San Diego State University, San Diego, CA 92182. 6 John Hunter, Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, 8604 La Jolla Shores Drive, La Jolla, CA 92038, pers. commun. January 1986. High mortality also occurred in the laboratory at about 2.5 wk, when larvae were apparently switch- ing from smaller to larger prey. This may be a critical period when the larvae have to learn to cap- ture larger, faster swimming crustaceans as the primary dietary component in order to obtain suffi- cient caloric intake. Studies on other fish larvae have demonstrated the necessity of increasing prey size with growth (Hunter 1977; Hunter and Kimbrell 1980). Juvenile and Adult Life History Ages of juveniles and adults calculated in the pres- ent study agree with estimates for giant kelpfish determined by J. E. Fitch (in Feder et al. 1974) and by R. Collins6. Ages by Coyer (1982), based on 42 kelpfish samples, do not agree with those in the pres- ent study. Coyer appeared to have overestimated the oldest kelpfish by 3 yr. This may have been due to the prevalence of "checks" or partially completed false rings on the otoliths which are commonly formed during spawning (Collins and Spratt 1969) and were frequently observed in the present study. Estimated size at sexual maturity (mean 18.6 cm TL) agrees with that reported by Coyer (1982). Past the age of sexual maturity, female giant kelp- fish are significantly larger than males and also live several years longer. Size discrepancy between adult males and females may have evolved from the females' behavior of venturing away from their ter- ritories during the spring spawning season into those occupied by males (Stepien 1985, 1986). They are often readily visible at this time while away from plants of matching colors. Large size may help females to avoid predation or, alternatively, may be the result of selection for increased fecundity. ACKNOWLEDGMENTS Grants and funds supporting this research included Sigma Xi, the Lerner Fund for Marine Re- search, the Theodore Roosevelt Memorial Scholar- ship Fund of the American Museum of Natural History, the University of Southern California Department of Biological Sciences and Graduate School, and a Sea Grant Traineeship. Laboratory facilities were provided by the Catalina Marine Science Center, U.S.C.'s Fish Harbor Research Laboratory, and Southern California Edison (Redon- do Beach). 6Robson Collins, California State Department of Fish and Game, Long Beach, CA 90813, pers. commun. March 1982. 824 STEPIEN: LIFE HISTORY AND DEVELOPMENT OF KELPFISH I thank the following people for giving technical assistance and information: Robson Collins, John Hunter, Kenneth Rich, C. Robert Bostick, Charles Winkler, Peter McGroddy, Steve Edwards, Eric Lynn, Robert Lavenberg, Gerald McGowen, Gary Brewer, and Laura Alderson. Steven Naffziger, Donald Wilkie, Robert Moore, Neale Jones, Mark Carr, Brandon Kulik, Richard Wright, John Sudick, Neale Jones, and Larry Allen assisted in collecting kelpfish. Robert Provin helped draw the figures and Stanley Azen provided statistical advice. This manu- script benefited substantially from critical reviews by Basil Nafpaktitis, Richard Brusca, Gerald Bakus, Gerald McGowen, and Bernard Abbott. LITERATURE CITED Balon, E. K. 1984a. Reflections on some decisive events in early life of fishes. Trans. Am. Fish. Soc. 113:178-185. 1984b. Patterns in the evolution of reproductive styles in fishes. In G. W. Potts and R. J. Wooton (editors), Fish reproduction: strategies and tactics, p. 35-53. Acad. Press, Lond. Barnhart, P. S. 1932. Notes on the habits, eggs, and young of some fishes of southern California. Bull. Scripps Inst. Oceanogr., Univ. Calif. 3:87-99. Collins, R. A., and J. D. Spratt. 1969. Age determination of northern anchovies, Engraulis mordax, from otoliths. Calif. Fish Bull. 147:39-55. Coyer, J. A. 1979. The invertebrate assemblage associated with Macro- cystis pyrifera and its utilization as a food resource by kelp- forest fishes. Unpubl. Ph.D. Thesis, Univ. Southern Califor- nia, Los Angeles, 314 p. 1982. Observations on the reproductive behavior of the giant kelpfish, Heterostichus rostratus (Pisces: Clinidae). Copeia. 1982:344-350. Ellertsen, B., P. Solemdal, T. Stromme, S. Tilseth, T. Westgard, E. Moksness, and V. Oiestad. 1980. Some biological aspects of cod larvae (Gadus morhua L.). Fiskeridir. Skr. Ser. Havunders. 17:29-47. Feder, H. M., C. H. Turner, and C. Limbaugh. 1974. Observations on fishes associated with kelp beds in southern California. Calif. Fish Bull. 160:1-144. FlTCH, J. E. 1951. Age composition of the southern California catch of Pacific mackerel 1939-40 through 1950-51. Calif. Fish Bull. 83:1-75. Hobson, E. S. 1971. Cleaning symbiosis among California inshore fishes. Fish. Bull., U.S. 69:491-523. Hobson, E. S., W. N. McFarlane, and J. R. Chess. 1981. Crepuscular and nocturnal activities of Californian nearshore fishes, with consideration of their scotopic visual pigments and the photic environment. Fish. Bull., U.S. 79:1-30. Holder, C. F. 1907. The nest of the kelpfish. Am. Nat. 41:587-588. Houde, E. D. 1972. Some recent advances and unsolved problems in the culture of marine fish larvae. Proc. World Maricult. Soc. 3:83-112. Hubbs, C. 1952. A contribution to the classification of the Blennioid fishes of the family Clinidae, with a partial revision of the eastern Pacific forms. Stanford Ichthyol. Bull. 4:41-165. Hubbs, C. L. 1920. Protective coloration and habits in the kelpfish, Hetero- stichus rostratus. Copeia 1920:19-20. Hunter, J. R. 1972. Swimming and feeding behavior of larval anchovy Engraulis mordax. Fish. Bull., U.S. 70:821-838. 1977. Behavior and survival of northern anchovy Engraulis mordax larvae. Calif. Coop. Oceanic Fish. Invest. Rep. 19: 138-146. 1981. Feeding ecology and predation of marine fish larvae. In R. Lasker (editor), Marine fish larvae: morphology, ecology, and relation to fisheries, p. 33-77. Wash. Sea Grant Program, Seattle. Hunter, J. R., and C. A. Kimbrell. 1980. E arly life history of Pacific mackerel , Scomber japoni- cus. Fish. Bull., U.S. 78:89-101. Jensen, A. C. 1965. A standard terminology and notation of otolith readers. Int. Comm. Northwest Atl. Fish. Res. Bull. 2:5-7. Limbaugh, C. 1955. Fish life in the kelp beds and the effects of kelp har- vesting. Inst. Mar. Res. Ref. 55-9, 158 p. Matarese, A. C, W. Watson, and E. G. Stevens. 1984. Blennioidea: Development and relationships. In H. G. Moser et al. (editors), Ontogeny and systematics of fishes, p. 565-577. Am. Soc. Ichthyol. Herpetol. Spec. Pub. 1. Quast, J. C. 1968. Observations on the food of the kelp-bed fishes. In W. J. North and C. L. Hubbs (editors), Utilization of kelp-bed resources in southern California, p. 109-142. Bull. Calif. Dep. Fish Game 139. Roedel, P. M. 1953. Common ocean fishes of the California coast. Calif. Fish Bull. 91:1-184. Shiogaki, M., and Y. Dotsu. 1972. The life history of the blenniid fish, Neoclinus bryope. [In Jpn., Engl, abstr.] Bull. Fac. Fish. Nagasaki Univ. 34, p. 1-8. Smith, P. E. 1981. Fisheries on coastal pelagic schooling fish. In R. Lasker (editor), Marine fish larvae: morphology, ecology, and relation to fisheries, p. 1-31. Wash. Sea Grant Program, Seattle. SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry: The principle and practice of statistics in biological research. 2d ed. W. H. Freeman and Co., San Franc, 859 p. Sparta, A. 1948. Uova ovariche, uova fecondate tenute in colture larva alia schiusa, stadi larvali e giovanile di Cristiceps argenatus. Risso. [In Ital.] Arch. Oceanogr. Limnol., Mem. 315 (1-3): 79-84. Stepien, C. A. 1985. Life history, ecology, and regulation of the colormor- phic patterns of the giant kelpfish Heterostichus rostratus Girard (Family Clinidae). Ph.D. Thesis, Univ. Southern California, Los Angeles, 318 p. 825 FISHERY BULLETIN: VOL. 84, NO. 4 1986. Regulation of the colormorphic patterns in the giant Weihs, D. kelpfish, Heterostichus rostratus Girard: Genetic versus en- 1980. Energetic significance of changes in swimming modes vironmental factors. J. Exp. Mar. Biol. Ecol. 100:181-208. during growth of larval anchovy, Engraulis mordax. Fish. Theilacker, G. H., and M. F. McMaster. Bull., U.S. 77:597-604. 1971. Mass culture of the rotifer Brachionus plieatilis and Zimmerman, R. C., and J. L. Kremer. its evaluation as a food for larval anchovies. Mar. Biol. 1983. Crunch: The friendly data analysis program. Zimmer- (Berl.) 10:183-188. man and Kremer, Los Ang., 17 p. 826 A SIMPLE METHOD FOR ESTIMATING THE FOOD CONSUMPTION OF FISH POPULATIONS FROM GROWTH DATA AND FOOD CONVERSION EXPERIMENTS1 Daniel Pauly2 ABSTRACT Experimental data on the gross food conversion efficiency of fishes (Kl = growth increment/food in- gested) are usually reduced to a model of the form Kx = aWb ; it is shown that the model K-, = 1 - (WIWao)1' has a number of advantages over the traditional model. The new model can be used to compute the food consumption per unit biomass of an age-structured fish population, by relying on the first derivative of the von Bertalanffy growth formula (VBGF) to ex- press growth increments, and the identity of W^, in the VBGF and in the model expressing Kx as a func- tion of weight. Computed examples, using published growth and mortality parameters, and the results of food con- version experiments were used to obtain consumption estimates in a carnivorous grouper (Epinephelus guttatus) and an herbivorous angelfish (Holacanthus bermudensis). Results were shown to be most sen- sitive to the parameter /J. Various applications of this simple model are discussed, particularly as a method to estimate key inputs in J. J. Polovina's ECOPATH model. A multiple-regression extension of the basic model is presented which accounts for the impact of factors other than body weight on values of Kl and /?. This method is illustrated with an analysis of data on dab (Limanda limanda). Estimating the quantity of food eaten during a cer- tain period by a fish population from field data is usually a difficult task and various sophisticated methods developed for this purpose have data re- quirements which can make their routine applica- tion impossible (Beverton and Holt 1957; Ursin 1967; Daan 1973, 1983; Andersen 1982; Armstrong et al. 1983; Rice et al. 1983; Stewart et al. 1983; Pennington 1984; Majkowski and Hearns 1984). Polovina (1984) recently presented a technique for construction of ecosystem models which is structured around a well-documented computer program called ECOPATH (Polovina and Ow3). In situations where classical fishery data are sparse this technique has the potential of becoming a standard method for consolidating and examining the data available on aquatic ecosystems. ECOPATH esti- mates equilibrium biomass (B), annual production 'Based on Chapter 3 of a "Habilitationschrift" presented in December 1984 to the Dean of the Mathematics and Science Facul- ty, Kiel University (Federal Republic of Germany) and titled "Zur Biologie tropischer Nutztiere: eine Bestandsaufnahme von Konzepten und Methoden." ICLARM Contribution No. 281. international Center for Living Aquatic Resources Manage- ment, MCC P.O. Box 1501, Makati, Metro Manila, Philippines. 3Polovina, J. J., and M. D. Ow. 1983. ECOPATH: a user's manual and program listings. Southwest Fish. Cent. Admin. Rep. H 82-83. Southwest Fisheries Center Honolulu Laboratory, Na- tional Marine Fisheries Service, NOAA, 2570 Dole Street, Hono- lulu, HI 96822-2396. (P), and annual consumption (Q) for each group in the model. ECOPATH requires a number of data inputs for each group treated in the model and usual- ly the most difficult to obtain is the average food consumption per unit biomass (Q/B) of each group. The present study derives a method to estimate Q/B through a combination of experimental and field data that are easily obtained. In the process, a model is derived which will allow for more information to be extracted from feeding experiments than has hitherto been the case. MODEL FOR REDUCING EXPERIMENTAL DATA ON THE CONVERSION EFFICIENCY OF FISHES Usually laboratory or pond feeding experiments lead to estimates of Kx, the gross conversion ef- ficiency, which are obtained, for short intervals, from Kx = growth increment/food ingested (1) (Ivlev 1939, 1966). Usually, Kx declines with body size (other factors affecting Kx are discussed below) and it has become Manuscript accepted April 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 827 FISHERY BULLETIN: VOL. 84, NO. 4 a standard procedure to plot empirical values of Kx obtained against the corresponding body weights, i.e., the mean weights (W) corresponding to each growth increment, or logio Ki = log10 a + b log10 W which leads to the model Kx = aWb. (2) (3) (See Sprugel 1983 for a method to correct the bias due to log transformation in this and the other models below.) A discussion of this model may be found in Jones (1976) (see Figure la for an example). This model has three liabilities, the first of which is the most serious: 1) The parameters "a" and "6" have no biological meaning, i.e., cannot be predicted from one's knowledge of the biology of a given fish. Converse- 0.4 o o 0.5 0.6 0.7 0.8 0.9 0.25 *r 0.20 0.15 o 1 0.10 0.05 0.00 Traditional model (r2 = 0.821) New model (r2 = 0.888) ,oqiow°> 2 3 Body weight (g,log|0units) ooi o .s> i.o c o o 0.6 0.4 0.2 Traditional model Upper limit for new model (K=l,when W=0) — Traditional model — New model Lower limit for new model Traditional (KpO when W» Woo) model \ 50 100 _i ty\/\ s---ii- , 150 1,500 00 Body weight (g) Figure 1.— Relationship of gross food conversion efficiency (K{) and body weight (W) in Channa striata, a) Plot of \ogwK1 on log10W^, as needed to estimate parameters "a" and "b" of traditional model for prediction of Kx from body weight, b) Plot of -log10(l -Kx) on log10W, as needed to estimate parameters Wm and p of new model, c) Comparison of the two models. Note that both fit the data well over the range for which data points are available, but that the traditional model provides nonsensical results beyond this range (see text). Based on the data in Pandian (1967). 828 PAULY: ESTIMATING FOOD CONSUMPTION OF FISH POPULATIONS ly, these parameters do not provide information which can be interpreted via another model. 2) The model implies values of K1 > 1 when a~llb > W > 0, which is nonsensical. 3) The model implies that, except when W = 0, Kx is always > 0, even in very large fish, although it is known that fish cannot grow beyond certain species-specific and environment-specific sizes, whatever their food intake. The new model proposed here has the form Kx = 1 - (W/Wy (4) with ft as a constant and Wm as the weight at which Kx = 0. The model implies that Kx = 1 when W = 0, whatever the values of p and W^ (see Discussion for comments on using values other than 1 as up- per bound for Kx in Equation (4)). The new model can, as the traditional model, be fitted by means of a double logarithmic plot: C = p log10 W„- p log10 W (5) where C = -log10 (1 - Kx), the sign being changed here to allow the values of C to have the same posi- tive sign as the original values of Kx. Interesting- ly, it also appears that negative values of Kx (based on fish which lost weight), which must be ignored in the traditional model, can also be used in this model (as long as they do not drag the mean of all available Kx values below zero, see Table 1), although their interpretation seems difficult. The new model requires no more data, nor markedly more computations than the old one. It produces "possible" values of Kx over the whole range of weights which a given fish can take. The values of W^, which represent the upper bound of this range can be estimated from W^ = antilog10 (C intercept/ 1 slope |). (6) Thus, while p has no obvious biological meaning, the values of Wx obtained by this model do have a biological interpretation, which is, moreover, anal- ogous to the definition of Wx in the von Bertalanffy growth function (VBGF) of the form Table 1 .—Data on the food conversion efficiency of Channa striata (= Ophiocephalus striatus) (after Pandian 1967), Epinephelus striatus (after Menzel 1960), and Hola- canthus bermudensis (after Menzel 1958). Body weight Food conv. Transformed data C = Species and (g)1 (KiP iog10 w log^K, -log10(1-K ,) remarks 1.86 0.391 0.270 -0.408 0.215 v 9.92 0.274 0.998 -0.562 0.139 ' 13.09 0.320 1.117 -0.495 0.167 19.65 0.284 1.293 -0.547 0.147 24.63 0.278 1.391 -0.556 0.141 35.09 0.234 1.545 -0.631 0.116 | 45.15 0.199 1.655 -0.701 0.096 ' Channa striata 50.70 0.227 1.705 -0.644 0.112 (see Figure 1) 51.30 0.235 1.710 -0.629 0.116 57.00 0.208 1.756 -0.682 0.101 79.80 0.177 1.897 - 0.752 0.085 93.80 0.232 1.972 -0.635 0.115 107.50 0.157 2.031 - 0.804 0.074 , 0.079 / 123.80 0.166 2.093 - 0.780 216 0.247 2.334 -0.607 0.123 \ 285 0.219 2.455 - 0.600 0.107 Epinephelus 319 0.160 2.504 -0.796 0.076 guttatus; 392 0.153 2.593 -0.815 0.072 ( log10 W = 2.617; C = 0.0894 424 0.179 2.627 -0.747 0.086 [ 628 0.161 2.798 -0.793 0.076 I (see Figure 2) 647 0.177 2.811 - 0.752 0.085 649 0.187 2.812 -0.728 0.090 ' 66 0.222 1.820 - 0.654 0.109 ) Holacanthus 139 0.178 2.143 -0.750 0.085 > bermudensis 256 -0.258 2.408 not de- fined -0.100 J (28°C only)3 log10 W = 2.124 C = 0.031 'Mean of starting and end weights. 2Growth increment/food intake. 3Note that the experiment considered here was conducted with a food which led to deposition of fat, but not of protein (see also Table 2), a consideration that is ignored for the sake of this example. 829 FISHERY BULLETIN: VOL. 84, NO. 4 Wt = Wm (1 - e-^-fo))3 (7) (von Bertalanffy 1938; Beverton and Holt 1957), and where Wt, the weight at time t, is predicted via the constants K, t0, and W^, all three of which are usually estimated from size-at-age data obtained in the field (see Gulland 1983 or Pauly 1984a). That Wx values obtained via Equations (2) and (6) are realistic can be illustrated by means of that part of the data in Table 1 pertaining to Channa striata (= Ophiocephalus striatus), the "snakehead" or "mudfish" of south and southeast Asia. These data give, when fitted to the traditional model KY = 0.482 If"0-205. (8) The same data, when fitted to the new model give Kx = 1 - (Ml,580)0073. (9) (See Figure 1 for both models.) The value oiW00 = 1,580 g is low for a fish which can reach up to 90 cm in the field (Bardach et al. 1972). However, its growth may have been reduced in laboratory growth experiments conducted by Pandian (1967). Equation (6) used here to predict W^ is extreme- ly sensitive to variability in the data set investigated, and two approaches are discussed to deal with this problem. The first approach is the appropriate choice of the regression model used. In the example above (Equa- tion (9)), the model used was a Type I (predictive) regression, which is actually inappropriate, given that 1) the log10 W values are not controlled by the experimentator and 2) regression parameters are required, rather than prediction of C values (see Ricker 1973). The use of a Type II ("functional", or "Geometric Mean") regression appears more appropriate; con- version of a Type I to Type II regression (with parameters a, b') can be performed straight- forwardly through b' = bl\r\ and a = C - b' log10 W (10) (11) where r is the correlation coefficient between the C and the log10 W values (Ricker 1973). In the case of the example here, one obtains with r = 0.942 a new model: Kx = 1 - (W71,290)0077 (12) close to that obtained using a Type I regression, due to the high value of r of this example. However, in cases where the fit to the model is poor, the use of a Type II regression can make all the difference between realistic and improbable values of W^. Another approach toward optimal utilization of the properties of the new model (4) is the use of "ex- ternal" values of asymptotic weight, which will here be coded W^ to differentiate them from values of Wx estimated through the model. In such case, (i can be estimated from P = C/(log10 WM - log10 W) (13) in which W{ao) is an asymptotic size estimated from other than food conversion and weight data, e.g., from growth data or via the often observed close- ness between estimates of asymptotic size and the maximum sizes observed in a given stock (see Pauly 1984a, chapter 4). These two approaches are illustrated in the exam- ple below, which is based on the data in Table 1 per- taining to the grouper Epinephelus guttatus. When Equation (6) is interpreted as a Type I regression, these data yield a value of W^ > 12 kg, which is far too high for a fish known to reach 55 cm at most (Randall 1968). Interpreting Equation (5) as a Type II regression leads to a value of Wx = 3.5 kg which is realistic, although still not close to the asymptotic weight of 1,880 g estimated by Thompson and Munro (1977). Finally, using the latter figure as an estimate of W{oo) yields the model Kx = 1 - (W/1,880)0136 (14) as a description of the relationship between Kx and weight in Epinephelus guttatus (Fig. 2). The value of p in Equation (14) lies within the 95% confidence interval of the value of p = 0.060 which generated the first unrealistically high estimate of Wm. MODEL FOR ESTIMATING THE FOOD CONSUMPTION OF FISH POPULATIONS When feeding experiments have been or can be conducted under conditions similar to those prevail- ing in the sea (food type, temperature, etc.), the 830 PAULY: ESTIMATING FOOD CONSUMPTION OF FISH POPULATIONS 0.14 Wo, estimated through Type I regression Wm estimated through Type H regression ■)£ Mean of y,x values W(a,) input from outside loginAB(TypeI)\. logotype I) 10 0.00 K/ i X 7 2.0 2.5 3.0 3.5 4.0 Body weight (g, log units) Figure 2.— Relationship between gross food conversion efficiency (KJ and body weight in Epinephelus guttatus. Note that a Type I "predictive" regression leads to an overestimation of W„ while a Type II "functional" regression leads to a value of W^ close to an estimate of W^ based on growth data (see text). Based on data in Menzel (1960). model presented above can be made a part of a model for estimation of food consumption per unit biomass (Q/B), provided a set of growth parameters is also used in which the value of Wm or W{oo) is iden- tical to that estimated from or used to interpret the feeding experiments. In this case, inserting Equation (8) into Equation (5) leads to Km = 1 - (1 - e-*«- K> Z » tmax > to > tr (24) These results suggest that, when using this model, most attention should be given to an accurate esti- mation of p (see below). It should be also noted that P and K have opposite effects on the estimation of Q/B (see Figure 3). Thus, a biased (e.g., high) estimate of Wx will be associated with too low values of p and K which partially compensate each other. 832 PAULY: ESTIMATING FOOD CONSUMPTION OF FISH POPULATIONS 90 t w w 'o ' r » i max -50 -10 o 10 Input change (%) 50 Figure 3.— Sensitivity analysis of Equation (22), based on parameter estimates in Table 4 for Epinephelus guttatus. Note strong effects of changes in /?, intermediate effects of K and Z, and negligible effects of Wmax, Wr , and t0 . Table 2.— Properties and parameter values of Epinephelus gut- tatus and Holacanthus bermudensis relevant to the computation of their food consumption (based on data in Menzel 1958, 1960; See Table 1 and text). Property/ Epinephelus Holacanthus parameter guttatus bermudensis Asymptotic weight (g) 1 1,880 2800 K(1/yr) 10.24 30.25 to (yo 4 -0.2 -0.2 t, (yr) 50.35 50.45 (3 60.136 20.040 Z (1/yr) 10.64 70.72 'max (y) 812 812 food (in fish (Anchoa, Algae (Monostroma experiments) Sardinella oxysperma and Haren- and Enteromorpha gula) satina) 1From Thompson and Munro (1977); Z = 0.64 refers to an unfished stock and is thus an estimate of M. 2From data in Table 1 and Equation (13). 3Based on method in Pauly and Munro (1984) and on growth parameter estimates pertaining to members of the related family Acanthuridae, in Pauly (1978). 4Assumed; has little influence on results (see text and Figure 3). Corresponding to a fish of 1 g with growth parameters W^, K, and f0 as given. 6See text and Figure 2. 7Based on equation (11) in Pauly (1980), with T = 28°, L„ = 30 cm, K = 0.25, and M = Z. eAssumed; has little influence on results (see text and Figure 3). QUANTITIES OTHER THAN QIB ALSO ESTIMATED BY THE MODEL In addition to estimating QIB, the model presented above can be used to obtain other useful quantities; namely, 1) maintenance ration and related informa- tion, and 2) trophic efficiency. Although there are differences between authors, maintenance ration is usually defined as the food used by fish to just maintain their weight at some "routine" level of activity. Usually, maintenance ra- tion is estimated by feeding fish over a wide range of rations and determining by interpolation the ra- tion generating neither weight gains nor losses (Jones 1976). The model presented here allows the estimation of maintenance ration (even if fish have been fed constant rations) through extrapolation of weight- specific estimates of QIB, such as presented in Figure 4 to the size W^, i.e., to the size at which, by definition, all food consumed by a fish is used for maintenance. In the case of the feeding data on E. guttatus analyzed here, an estimate of daily main- 833 FISHERY BULLETIN: VOL. 84, NO. 4 (/> W > M » S. (33) See Li (1975) for further inferences based on path coefficients. In the southern North Sea in late summer-early autumn, Limanda limanda experiences tempera- tures usually ranging between 10° and 20°C (Lee 1972). Solving Equation (31) for T = 18°C, the highest temperature in Pandian's experiments (i.e., assuming the higher late summer-early autumn temperatures limit WJ leads to estimates of W = 500 g for the females and 298 g for the males, com- pared with the values of 756 and 149 g obtained by Lee (1972) on the basis of growth studies. Estimating values of /? that are wholly compatible with the latter estimates of W^ is straightforward, however, since it consists of solving Equation (31) forT=18°C,M=0, and the appropriate value of S, based on the equation P = 1/log WM (a + VVi + b2'V2 . . . bn'Vn) (34) In the present case, this leads to /3 values of 0.073 and 0.089 for females and male dab, respectively. The "average" relationship (if such exists) between food conversion efficiency and body weight in female dab fed herring meat is thus Kx = 1 - (H7756)0073 while for males it is Kx = \ - (IF/149)0089 (35) (36) with both values of fi within the 95% confidence interval of the first estimate of /3 (in Equation (27), see Table 3). DISCUSSION The model presented here for the computation of Q/B is not meant to compete against the more sophisticated models whose authors were cited above. Rather, it was presented as a mean of link- ing up the results of feeding experiments with elements of the theory of fishing such that infer- ences can be made on the food consumption of fish populations which 1) do not invoke untenable assumptions, 2) make maximum use of available data, and 3) do not require extensive field sampling. A distinct feature of the method is that it does not require sequential slaughtering of fish for the esti- mation of their stomach evacuation rate, nor field sampling of fish stomachs, which may be of rele- vance when certain valuable fishes are considered (e.g., coral reef fishes in underwater natural parks). Several colleagues who reviewed a draft version of this paper suggested that Equation (4) should in- corporate an upper limit for Kx smaller than unity. This model would have the form Kx = Klmax - (W/WJP™ (37) with parameters W^ and (im identical and analogous respectively to those in Equation (4) and a value of 836 PAULY: ESTIMATING FOOD CONSUMPTION OF FISH POPULATIONS Klmax to be estimated independently prior to fitting Equation (37) to data. Data do exist which justify setting the upper limit of Kx at or near unity. They pertain to fish em- bryos, whose gross conversion efficiency can be defined by Kx = Wh we - wy (38) where Wh is the larval weight at hatching, We the egg weight, and Wy is the weight of the yolk sac at hatching. Values of Kx as high as 0.93 have been reported using this approach (From and Rasmussen 1984), extending further toward unity the range of Kx values reported by earlier authors, e.g., 0.85 in Solea solea (Fliichter and Pandian 1968), 0.79 in Sar- dinops caerulea (Lasker 1962), and 0.74 in Clupea harengus (Blaxter and Hempel 1966). Thus, for a wet weight of 0.5 mg corresponding to a spherical egg of 1 mm diameter, one obtains, using Equation (14) for E. guttatus, a value of K1 = 0.87 which is within the range of Kx values given above. This example is not meant to suggest that Kx values pertaining to large fish should be used in combination with the model presented here to "estimate" K1 in eggs or larvae. Rather, it is meant to illustrate the contention that, of the possi- ble choices of an upper bound for Kx in Equation (4), the one selected here has the feature of making the model robust, particularly with respect to high values of Kx and extrapolations toward low values of W. Apart from (1, the key elements of the model (isometric von Bertalanffy growth, constant ex- ponential decay, steady-state population) are all parts of other, widely used models. Thus, whether estimates of QIB obtained by this model are con- sidered "realistic" or not will depend almost entirely on the value of (i used for the computation. There are several ways of reducing the uncertain- ty associated with p. The following may need special consideration: 1) Feeding experiments used to estimate p could be run so as to mimic as closely as possible the crucial properties of the habitat in which the popula- tion occurs whose QIB value is estimated, inclusive of seasonally oscillating factors. 2) Further research and study should lead to the identification of anatomical, physiological, and ecological properties of fish correlating with their most common value of ft. 3) An additional parameter could be added to account for fish reproduction, which is not explicit- ly considered in Equation (22). Little needs to be said about item 1 which should be obvious since (except in the context of aquacul- ture) feeding and growth experiments are conducted in order to draw inferences on wild populations. With regards to item 2, it suffices to mention that relative gill area ( = gill surface area/body weight), which appears to a large extent to control food con- version efficiency (Pauly 1981, 1984b), should be a prime candidate for correlational studies. Item 3 could cause QIB values obtained by the model pre- sented here to substantially underestimate actual food consumption, were it not for three circum- stances which produce opposite tendencies: a) The assumption that the energy needed by fish to develop gonads is taken from the energy other- wise available for growth may not apply (lies 1974; Pauly 1984b). Rather, the reduction of activity occurring in some maturing fish may more than compensate for the energy cost of gonad develop- ment (Koch and Wieser 1983). b) Growth parameters are usually computed using size data from fish whose gonads have not been removed, thus accounting for at least a fraction of the food converted into gonad tissue. When the value of Z used in the model is high, this fraction will be large because the contribution of the older fish to the overall estimate of QIB will be small. c) Experimental fish are usually stressed and therefore have lower conversion efficiencies than fish in nature, even though they may spend little energy on food capture (see Edwards et al. 1971). This effect leads to low values of ft and hence high estimates of QIB. Because of these factors, the values of QIB obtained by the method proposed here may lack a downward bias. ACKNOWLEDGMENTS I wish to thank R. Jones (Aberdeen), as well as E. Ursin (Charlottenlund), A. McCall (La Jolla), J. J. Polovina (Honolulu), P. Muck (Lima), and the two anonymous reviewers for their helpful comments on the draft of this paper. LITERATURE CITED Allen, K. R. 1971. Relation between production and biomass. J. Fish. 837 FISHERY BULLETIN: VOL. 84, NO. 4 Res. Board Can. 28:1573-1581. Andersen, K. P. 1982. An interpretation of the stomach contents of fish in relation to prey abundance. Dana 2:1-50. Armstrong, D. W., J. R. G. Hislop, A. P. Robb, and M. A. Brown. 1983. A preliminary report on the analysis of whiting stomachs collected during the 1981 North Sea Stomach Sampling Project. ICES. Doc. CM. 1983/G:59. Bardach, J. E., J. H. Ryther, and W. 0. McLarney. 1972. Aquaculture: the farming and husbandry of freshwater and marine organisms. Wiley-Inter-Science, N.Y., 868 p. Bertalanffy, L. von. 1938. A quantitative theory of organic growth (Inquiries on growth laws. II). Hum. Biol. 10:181-213. Beverton, R. J. H., and S. J. Holt. 1957. On the dynamics of exploited fish populations. Fish. Invest. Minist. Agric. Fish Food (G.B.) Ser. II, Vol. 19, 533 p. Blaxter, J. H. S., and G. Hempel. 1966. Utilization of yolk by herring larvae. J. Mar. Biol. Assoc. U.K. 46:219-234. Condrey, R. E. 1982. Ingestion-limited growth of aquatic animals: the case for Blackman kinetics. Can. J. Fish. Aquat. Sci. 39:1585- 1595. Daan, N. 1973. A quantitative analysis of the food intake of North Sea cod, (Gadus morhua). Neth. J. Sea Res. 6:479-517. 1983. Analysis of the cod data collected during the 1981 stomach sampling project. ICES Doc. CM. 1983/G:61. Edwards, R. R. C, J. H. S. Blaxter, U. K. Gopalan, C. V. Mathew, and D. M. Finlayson. 1971. Feeding, metabolism, and growth of tropical flatfish. J. Exp. Mar. Biol. Ecol. 6:279-300. Fluchter, J., and T. J. Pandian. 1968. Rate and efficiency of yolk utilization in developing eggs of the sole Solea solea. Helgol. Wiss. Meeresunters. 18:53-60. From, J., and G. Rasmussen. 1984. A growth model, gastric evacuation and body composi- tion in rainbow trout, Salmo gairdneri Richardson 1836. Dana 3:61-139. Gulland, J. A. 1983. Fish stock assessment: a manual of basic methods. John Wiley and Sons, N.Y., 223 p. Iles, T. D. 1974. The tactics and strategy of growth in fishes. In F. R. Harden-Jones (editor), Sea fisheries research, p. 331-345. Elek Sci. Books, Ltd., Lond., 511 p. Ivlev, V. S. 1939. Balance of energy in carps. [In Russ.] Zool. Zh. 18: 303-318. 1966. The biological productivity of waters. J. Fish. Res. Board Can. 23(11): 1727-1759. W. E. Ricker (transl.). Jones, R. 1976. Growth of fishes. In D. H. Cushing and J. J. Walsh (editors), The ecology of the seas, p. 251-279. Blackwell Scientific Publications, Oxford. Kinne, O. 1960. Growth, food intake, and food conversion in a eury- plastic fish exposed to different temperatures and salinities. Physiol. Zool. 33:288-317. Koch, F., and W. Wieser. 1983. Partitioning of energy in fish: can reduction of swim- ming activity compensate for the cost of production? J. Exp. Biol. 107:141-146. Lasker, R. 1962. Efficiency and rate of yolk utilization by developing em- bryos and larvae of the Pacific sardine Sardinops caerulea (Girard). J. Fish. Res. Board Can. 19:867-875. Lee, C K. G. 1972. The biology and population dynamics of the common dab Limanda limanda (L.) in the North Sea. Ph.D. Thesis, Univ. East Anglia. Li, C C. 1975. Path analysis: a primer. The Boxwood Press, Pacific Grove, CA, 347 p. Majkowski, J. 1982. Usefulness and applicability of sensitivity analysis in a multispecies approach to fisheries management. In D. Pauly and G. I. Murphy (editors), Theory and management of tropical fisheries, p. 149-165. ICLARM Conf. Proc. 9, 360 p. Majkowski, J., and W. S. Hearns. 1984. Comparison of three methods for estimating the food intake of a fish. Can. J. Fish. Aquat. Sci. 41:212-215. Menzel, D. W. 1958. Utilization of algae for growth by the Angelfish Hola- canthus bermudensis. J. Cons. Int. Explor. Mer 24:308- 313. 1960. Utilization of food by a Bermuda reef fish, Epinephe- lus guttatus. J. Cons. Int. Explor. Mer 25:216-222. Paloheimo, J. E., and L. M. Dickie. 1966. Food and growth of fishes. III. Relations among food, body size, and growth efficiency. J. Fish. Res. Board Can. 23:1209-1248. Pandian, T. J. 1967. Intake, digestion, absorption and conversion of food in the fishes Megalops cyprinoides and Ophiocephalus striatus. Mar. Biol. (Berl.) 1:16-32. 1970. Intake and conversion of food in the fish Limanda limanda exposed to different temperatures. Mar. Biol. (Berl.) 5:1-17. Pauly, D. 1978. A preliminary compilation of fish length growth param- eters. Ber. Inst. f. Meeresk., Univ. Kiel No. 55, 200 p. 1980. On the interrelationships between natural mortality, growth parameters, and mean environmental temperature in 175 fish stocks. J. Cons. Int. Explor. Mer 39:175-192. 1 98 1 . The relationship between gill surface area and growth performance in fish: a generalization of von Bertalanffy's theory of growth. Meeresforschung 28:251-282. 1982. Studying single-species dynamics in a tropical multi- species context. In D. Pauly and G. I. Murphy (editors), Theory and management of tropical fisheries, p. 33-70. ICLARM Conf. Proc. 9, Manila, 360 p. 1984a. Fish population dynamics in tropical waters: a manual for use with programmable calculators. ICLARM Studies and Reviews 8, Manila, 325 p. 1984b. A mechanism for the juvenile-to-adult transition in fishes. J. Cons. Int. Explor. Mer 41:280-284. Pauly, D., and J. L. Munro. 1984. Once more on the comparison of growth in fish and invertebrates. Fishbyte 2(1):21. Pennington, M. 1984. Estimating the average food consumption by fish in the field from the stomach contents data. I.C.E.S., Doc. CM. 1984/H:28. POLOVINA, J. J. 1984. Model of a coral reef ecosystem. I. The ECOPATH model and its application to French Frigate Shoals. Coral Reefs 3:1-11. 838 PAULY: ESTIMATING FOOD CONSUMPTION OF FISH POPULATIONS Randall, J. E. 1968. Caribbean Reef Fishes. T.F.H. Publications, Neptune City, NJ, 318 p. Rasch, D. (editor) 1983. Biometrie: Einfuhrung in die Biostatistik. VEB Dtsch. Landwirtschaftsverlag, Berl., 276 p. Rice, J. A., J. E. Breck, S. M. Bartell, and J. F. Kitchell. 1983. Evaluating the constraints of temperature, activity and consumption on growth of largemouth bass. Environ. Biol. Fishes 9:263-275. RiCKER, W. E. 1973. Linear regressions in fishery research. J. Fish. Res. Board Can. 30:409-434. 1975. Computation and interpretation of biological statistics offish populations. Bull. Fish. Res. Board Can. 191, 382 p. Sprugel, D. G. 1983. Correcting for bias in log- transformed allometric equa- tions. Ecology 64:209-210. Stewart, D. J., D. Weininger, D. V. Rottiers, and T. A. Edsall. 1983. An energetics model for lake trout, Salvelinus namay- cush: application to the Lake Michigan population. Can. J. Fish. Aquat. Sci. 40:681-698. Taylor, C. C. 1958. Cod growth and temperature. J. Cons. Int. Explor. Mer 23:366-370. Thompson, R., and J. L. Munro. 1977. Aspects of the biology and ecology of Caribbean reef fishes: Serranidae (hinds and groupers). J. Fish. Biol. 12: 115-146. Ursin, E. 1967. A mathematical model of some aspects of fish growth, respiration, and mortality. J. Fish. Res. Board Can. 24: 2355-2453. 839 FISHERY BULLETIN: VOL. 84, NO. 4 APPENDIX List of symbols used in model development and illustration b' B AX-AA terms used in computation of biomass per recruit (Equation (21)) a - multiplicative term in equation linking Kx and body weight (Equation (3)) - intercept of a Type I (multiple) linear re- gression a - intercept of a Type II (multiple) linear re- gression b - slope of a Type I linear regression - exponent in equation linking Kx and body weight bc - slope of a Type I multiple linear regression - slope of a Type II linear regression - slope of a Type II multiple linear regression - biomass (under equilibrium condition) - exponent in model linking Kx and body weight (Equation (4)) - similar to ft, but estimated jointly with #imax (Equation (37)) - (-log10(l - KJ) - same as C, but expressed in standard deviation units - rate of food consumption - rate of growth in weight - trophic efficiency, i.e., production by popu- lation/food consumption by population i - counter for number of variables in a multi- ple regression j - counter for number of multiple regres- sions K - constant in VBGF Kx - gross conversion efficienty (Equation (1)) ■^lmax ■ hypothetical upper limit for Kx (with #lmax < 1) (Equation (37)) M - instantaneous rate of natural mortality - a dummy variable expressing food type (Equation (27)) M' - a dummy variable expressing food type in standard deviation units n - number of partial regression coefficient used in computing a given value of 6/ N - number of fish in population (Equation (19)) 840 C C dq/dt dwldt Q - food consumption of a population (per unit time) Q/B - food consumption per unit biomass of an age-structured animal population Qc - cumulative food consumed by a single fish between ages tr and tmax (Equation (22)) R - number of recruits (Equation (19)) r - product moment correlation coefficient 5 - a dummy variable expressing sex S' - a dummy variable expressing sex in stan- dard deviation units t - age tc - mean age at first capture (in an exploited stock) £0 - a parameter of the VBGF expressing the theoretical age at size zero ^max ■ maximum age considered (= longevity) tr - mean age at recruitment to the part of the population considered when computing Q/B temperature in °C temperature in °C, expressed in standard deviation units (Equation (32)) any variable beyond W which affects Kx the von Bertalanffy growth function body weight (in log units in some cases) body weight (in log10 units), expressed in standard deviation units weight of a fish egg weight of a fish at hatching (yolk sac ex- cluded) body weight corresponding to tmax body weight corresponding to tr mean weight at age t yolk sac weight in a newly hatched fish asymptotic weight in the VBGF or in new model (Equation (4)) an estimate of asymptotic weight obtained indirectly (i.e., from data of a type differ- ent than those in model using value of W Y{ - any variable included in a multiple regres- sion Z - instantaneous rate of mortality (= PIB ratio) T T Vi VBGF W w wc wh w " max wr wt wy w, (oo) REPRODUCTIVE BIOLOGY OF KING MACKEREL, SCOMBEROMORUS CAVALLA, FROM THE SOUTHEASTERN UNITED STATES John H. Finucane, L. Alan Collins, Harold A. Brusher, and Carl H. Saloman1 ABSTRACT The reproductive biology of king mackerel, Scomberomorus cavalla, was studied from specimens collected off Texas, Louisiana, and northwest Florida in the Gulf of Mexico and off North and South Carolina in the Atlantic Ocean. Gonads were examined from 1,163 females and 595 males obtained in 1977-78. Spawning was prolonged. Most king mackerel were reproductively active from May through September. A few fish were in spawning condition as early as April and as late as October. All females were mature at 850-899 mm fork length (FL). Estimates of fecundity ranged from about 69,000 to 12,207,000 eggs for fish from 446 to 1,489 mm FL, 618 to 25,610 g total weight (TW), and 1 to 13 years of age. Fecundity (F) was usually significantly correlated with FL, TW, and age in each area but TW was the best predic- tor of fecundity in all areas combined (F = 1.854 x 101 (TW)1'361) with r2 = 0.856. King mackerel, Scomberomorus cavalla, is one of the most valuable commercial and recreational fish in the Gulf of Mexico and south Atlantic. It is an epipelagic, neritic species that occurs in the western Atlantic Ocean from Massachusetts to Rio de Janeiro, Brazil (Collette and Russo 1979, 1984). Most of the king mackerel caught off the southeastern United States are landed in Florida (Manooch 1979) where it is an important component of charter boat catches (Moe 1963; Brusher et al. 1978). Commer- cial landings in Florida during 1983 totaled 2,017 t and the estimated recreational catch from the Gulf of Mexico was 1,090,000 fish in 1984 (U.S. Depart- ment of Commerce 1985a, b). Although much has been written on king mack- erel, little is known of its reproductive biology (Manooch et al. 1978). Ovarian histology and size- at-maturity has been described by Alves and Tome (1967) for fish from Brazil and by Beaumariage (1973) for fish from Florida. Maturation based on blood hormone levels from fish off northwest Florida was reported by MacGregor et al. (1981). Spawn- ing times and areas have been inferred from ichthyo- plankton collections of king mackerel larvae (Dwinell and Futch 1973; Finucane and Collins 1977; Houde et al. 19782; McEachran et al. 1980). The only fecun- 'Southeast Fisheries Center Panama City Laboratory, National Marine Fisheries Service, NOAA, 3500 Delwood Beach Rd., Panama City, FL 32407-7499. 2Houde, E. D., J. C. Leak, C. E. Dowd, S. A. Berkely, and W. J. Richards. 1979. Ichthyoplankton abundance and diversity in the eastern Gulf of Mexico. Part I: Executive summary, abstract, text reference. Unpubl. manuscr., 119 p. Draft Final Report to dity estimates in the literature were made by Ivo (1974) for fish from Brazil. The purpose of our study was to provide additional information on king mackerel reproductive biology by determining spawning season, length-at- maturity, and fecundity from four areas off the southeastern coast of the United States. This infor- mation will be useful in the management of king mackerel since the measure of reproductive poten- tial is a basic element of productivity and stock dynamics (Baglin 1982). METHODS King mackerel were sampled from commercial and recreational catches in four separate areas along the coast of the southeastern United States during 1977 and 1978 (Fig. 1). These areas were I, the northwestern Gulf of Mexico off the central and south coasts of Texas; II, the northcentral Gulf off Louisiana and Mississippi; III, the northeastern Gulf off northwest Florida; and IV, the western Atlan- tic Ocean off South and North Carolina. Procedures for processing gonads, weighing, and measuring fish followed the methods of Finucane and Collins (1984). If no total weight had been recorded for a fish, we estimated TW by using the formula TW = 1.4959 x 10 "5 (FL)2-89284 (TW = total weight in grams; FL = fork length in milli- meters). This formula (Ricker 1975) was derived Manuscript accepted April 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. Bureau of Land Management, Contract AA550-CT7-28. Rosen- stiel School of Marine and Atmospheric Science, University of Miami, Miami, FL 33149. 841 FISHERY BULLETIN: VOL. 84, NO. 4 GULF OF MEXICO Figure 1.— Sampling areas for king mackerel, Scomberomorus cavalla, in the Gulf of Mexico and Atlantic Ocean during 1977-78. from a length-weight regression (r = 0.996; n = 186) of king mackerel data from all areas. Egg size distributions within the ovary were sta- tistically compared to ensure that subsamples taken for studies of maturation and fecundity were repre- sentative (Yuen 1955; Otsu and Uchida 1959). Both ovarian lobes were divided into three sections (anterior, middle, and posterior) of about equal length. At a selected point along each of these sec- tions, a 2-4 mm thick cross section was cut and removed. A wedged-shaped portion was then taken from each of the three cross sections and divided into three zones: inner, middle, and outer. A sam- ple of 150 yolked eggs from each of the zones was examined with a microscope and all eggs were mea- sured to the nearest 0.02 mm at 500 x on whatever axis the egg happened to be located in respect to an ocular micrometer scale (Clark 1934). A chi- square test of independence (Steel and Torrie 1960) was used to test for significant differences in mean egg diameters (EDs) among the sections, zones, and zones within a section in each lobe. Each wedge-shaped sample of eggs was placed in a dish with 10% Formalin3 and the eggs were then teased apart. Samples containing only unyolked eggs (<0.20 mm ED) were considered to be from immature fish and only 100 eggs from these samples were measured. Samples with yolked eggs (^0.20 mm ED) were considered to be from mature fish and 300 eggs were measured. Seasonal maturation was determined by plotting monthly mean EDs of the most advanced eggs found in each ovary and by gonadosomatic indices (GSI = the percentage of TW represented by gonad weight). The range and 95% confidence interval of the monthly mean GSIs were also plotted. To com- pare the variation of GSIs, we calculated the coef- ficient of variation for each month. We estimated the length at which the fish first matured by com- puting mean GSIs for fish in each 50 mm interval and used the length at which the greatest increase in mean GSIs between consecutive FL intervals oc- curred. For this analysis we only used data that were collected during the fish's most sexually active months as indicated by the highest values of mean EDs and GSIs. An additional estimate was made for females by assigning immature or mature status to each fish according to egg stage and then calculating the percentage of mature fish by FL intervals. 3Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 842 FINUCANE ET AL.: REPRODUCTIVE BIOLOGY OF KING MACKEREL Fecundity estimates were based on the number of yolked eggs ^0.20 mm in diameter in the most mature ovaries. Similar methods were discussed by Hunter and Goldberg (1980) and used by Morse (1980). A diameter of 0.20 mm was used to separate immature and mature eggs, because it was at this size that yolk first appeared. A gravimetric method was used for fecundity and followed the procedures of Finucane and Collins (1984). Ages of fish were determined from otoliths (Johnson et al. 1983). Analysis of covariance was used to test for differ- ences in fecundity by year and area. Regression and correlation were used to examine the linear and curvilinear relationships between fecundity and fork length, total weight, and age. RESULTS Gonads from 1,165 female and 593 male king mackerel were examined. Fish ranged in FL from 351 to 1,554 mm, in TW from 658 to 31,780 g, and in age from 1 to 13 yr. Temporal coverage varied from 3 mo in area I to 12 mo in area II. Number and percentage composition of fish by area were area I, 85 and 4.8%; area II, 646 and 36.7%; area III, 768 and 43.7%; and area IV, 259 and 14.7%. Analysis of the egg size distribution indicated that there were significant differences (a = 0.05) in ED between the inner, middle, and outer zones within ovarian sections; there were no differences between sections. Therefore, we took a wedge-shaped sam- ple (representing the three cross-sectional zones) from the middle of the right or left ovary as repre- sentative of the entire ovary for ED analysis. King mackerel ovaries were grouped into five reproduc- tive stages based on ED. Stage I (immature ovaries) contained eggs <0.06 mm. Eggs in stage II (resting ovaries) ranged from 0.07 to 0.20 mm. Stage III (maturing) and stage IV (mature) ovaries contained eggs 0.21-0.50 mm and 0.51-0.71 mm, respective- ly. Stage V eggs measured 0.71-1.20 mm and in- dicated ripe ovaries. The seasonal progression of mean GSIs and EDs indicated that king mackerel have a prolonged spawning season that varied between areas (Figs. 2-5). Peak spawning months occurred from May through September as observed in 14 ripe females from areas I, II, and IV. A few fish were in spawn- ing condition as early as April and as late as Octo- ber. In area I, GSIs and EDs peaked in July and August for both sexes. Area II fish had the highest GSIs and EDs for both sexes during May. In area III, GSIs for both sexes were greatest during June O 1 O 2 fir t VI -im- MALES NO. FISH 2 23 36 IX 0.50 0.40 0.30 0.20 0.10 0 FEMALES NO. FISH 3 11 12 J J A MONTHS Figure 2.— Seasonal maturation cycle of male and female king mackerel from area I (Texas) shown by monthly gonadosomatic index (GSI) and mean egg diameters (EDs) in mm. while EDs peaked in August. Area IV fish had the highest female GSIs and EDs during July. Serial spawning was suggested by several lines of evidence. Distribution of EDs was multimodal during spawning months. The highest coefficient of variation for GSIs occurred during the spawning months, suggesting that eggs were maturing and released serially throughout the spawning season (Table 1). The size at maturation of king mackerel also varied between areas. Maturity was based on the number and percentage of fish with stage Ill-stage V ova for each 50 mm FL interval. Length inter- vals in which at least 50% of the females were mature for areas I-IV, respectively, were 450-499 mm, 600-649 mm, 600-649 mm, and 650-699 mm 843 FISHERY BULLETIN: VOL. 84, NO. 4 5 4 to O 2 MALES 1 0 NO. FISH -B- 2 3 ■§• T-'-fx 1 8 10 6 8 14 0 7 6 5 4 3 2 1 ■FEMALES 95%C.I.-) RANGE JJj.MEAN MO. FISH 19 31 49 39 4 58 A -*HHi- ■£*-*- J FMAMJ JASOND MONTHS Figure 3.— Seasonal maturation cycle of male and female king mackerel from area II (Louisiana and Mississippi) shown by month- ly gonadosomatic index (GSI) and mean egg diameters (EDs) in mm. O 2 6 5 _ 4 O 3 MALES NO. FISH 1 5 59 78 74 163 r^GEj 95%C.l.n 11- MEAN 2 - 1 - 0 0.3 £0.2 IX 0.1 FEMALES NO. FISH 7 41 107 76 85 72 i i I J- M J J A S O MONTHS Figure 4.— Seasonal maturation cycle of male and female king mackerel from area III (northwest Florida) shown by monthly gonadosomatic index (GSI) and mean egg diameters (EDs) in mm. Table 1 .—Coefficient of variation for monthly GSIs of female (F) and male (M) king mackerel in each area. Area I Area II Area III Area IV Month F M F M F M F M January — — 12.7 — — — — February — — 12.5 — — — — — March — — 18.0 7.1 — — — — April — — 43.6 15.3 — — — — May — — 20.4 i 28.2 — 56.3 52.9 June 51.7 16.0 55.0 57.1 96.9 56.4 39.1 — July 33.6 35.4 77.3 58.2 95.5 104.8 36.2 1 August 54.2 38.9 66.2 43.4 95.5 75.7 44.4 2.5 September — — 56.7 38.7 85.9 51.9 64.9 61.5 October — — 36.8 48.4 62.7 51.6 41.7 54.5 November — — 32.8 — — — — — December — — 20.0 — — — — — 'n = 1 844 FINUCANE ET AL.: REPRODUCTIVE BIOLOGY OF KING MACKEREL 5 2 O 1 0 ■RANGE IX _ 3 «/» ° 2 1 0 0.50 0.40 0.30 0.20 0.10 0 MALES k NO.FISH 2 1 2 8 89 NO FISH 3 12 10 8 20 104 MJ J A S O MONTHS FIGURE 5.— Seasonal maturation cycle of male and female king mackerel from area IV (North and South Carolina) shown by monthly gonadosomatic index (GSI) and mean egg diameters (EDs) in mm. (Table 2). All females were mature at 850-899 mm. Another maturation pattern was noted when the midpoints of fork length intervals were plotted against mean GSIs for each area (Fig. 6). The size interval where greatest increases in GSIs occurred were 650-699 mm (area I), 700-749 mm (area II), 450-499 mm (area III), and 650-699 (area IV). Fecundity ranged from 69,000 to 12,207,000 eggs in 65 king mackerel from all areas. Fish ranged in FL from 446 to 1,489 mm, in TW from 681 to 25,610 g, and in age from 1 to 13 yr (Table 3). Analysis of covariance with TW as the covariate showed no significant differences (a = 0.05) in fecundity between years or among areas. The best predictor of fecundity based on regression and cor- relation analysis was TW for areas II, IV, and all areas combined and FL for areas I and III (Table 4). Log transformed linear models were better pre- dictors of fecundity than nontransformed models in all areas but area IV. DISCUSSION Our results on the seasonal maturation and pro- tracted spawning season of king mackerel agree closely with other studies. In waters off Florida, Beaumariage (1973) found late-maturing (stages III and IV) eggs in king mackerel from May through October. In the northeastern Gulf of Mexico (area III), Dwinell and Futch (1973) caught king mackerel larvae during the same time interval and MacGregor et al. (1981) reported early- or late-maturing ovaries from August through October. In the northwestern Gulf of Mexico off Texas (area I), Finucane and Col- lins (1977) and McEachran et al. (1980) noted catches of larvae from May through August, and April through October, respectively. In the area off Cape Fear, NC, to Cape Canaveral, FL, Powles4 col- lected king mackerel larvae from May through September. Length at maturation was difficult to determine because the sample size of small fish (<600 mm) was limited in all areas except area III (northwest Florida). Using only fish from this area, maturity first occurred about 450-499 mm and 50% of the fish were mature at about 550-599 mm. These estimates of maturity agreed with some of the other studies. Female king mackerel first reached sexual matur- ity at 630 mm and 4 yr of age (Gesteira and Mes- quita 1976) or at 586 mm (Alves and Tome 1967) off Brazil. Another study on Brazilian fish, however, noted that females were first mature at 770 mm and 5-6 yr of age (Ivo 1972). In Florida waters, Beau- mariage (1973) estimated that females 3 yr or younger were immature and probably had not spawned. He believed that the first major spawn- ing by females and males occurred at 880 and 770 mm SL, respectively. Some of his 1-yr-old females contained stage IV eggs that had been aborted or reabsorbed since he did not find ripe (stage V) eggs until the fish were 4 yr old. His standard length for king mackerel from Florida at age 1 was 610 mm (651 mm FL), which was higher than our estimate of length at first maturity. 4Powles, H. W. Abundance and distribution of king mackerel, (Scomberomorus cavalla) and Spanish mackerel (S. maculatus) lar- vae of the southeast United States. Unpubl. manuscr. Gouvern- ement du Canada, Peches et Oceans, Division des Sciences halieu- tiques, C. P. 15500, Quebec, Canada GlK 7Y7. 845 FISHERY BULLETIN: VOL. 84, NO. 4 Table 2.— Total sample number and percentage of mature (Stages lll-V) king mackerel females collected during the peak maturation season in each area.1 Area I Area II Area III Area IV (Louisiana, (Northwest (North and South I Texas) Mississippi) Mature Florida) Carolina) Mature Mature Mature Fork length No. (0/0) No. (%) No. (%) No. (%) 300-349 0 0 — 0 — 0 — 350-399 1 0.0 0 — 1 0.0 0 — 400-449 0 — 0 — 2 0.0 0 — 450-499 1 100.0 0 — 3 33.3 0 — 500-549 0 — 0 — 16 6.3 0 — 550-599 0 — 0 — 28 46.4 0 — 600-649 2 100.0 2 100.0 31 71.0 1 0.0 650-699 0 — 0 — 31 71.0 4 75.0 700-749 4 100.0 1 100.0 35 80.0 2 100.0 750-799 8 100.0 0 — 29 62.1 5 100.0 800-849 6 100.0 0 — 41 75.6 4 100.0 850-899 2 100.0 5 100.0 29 100.0 11 100.0 900-949 2 100.0 6 100.0 21 100.0 8 100.0 950-999 0 — 22 100.0 19 100.0 7 100.0 1,000-1,049 0 — 19 100.0 13 100.0 3 100.0 1,050-1,099 0 — 18 100.0 4 100.0 3 100.0 1,100-1,149 0 — 18 100.0 6 100.0 1 100.0 1,150-1,199 0 — 13 100.0 4 100.0 1 100.0 1 ,200-1 ,249 0 — 18 100.0 2 100.0 0 — 1,250-1,299 0 — 14 100.0 1 100.0 0 — 1 ,300-1 ,349 0 — 17 100.0 0 — 0 — 1,350-1,399 0 — 11 100.0 0 — 0 — 1,400-1,449 0 — 3 100.0 0 — 0 — 1,450-1,499 0 — 2 100.0 0 — 0 — Total 26 169 316 50 'Area I, June-August; Area II, May-August; Area III, May-September; and Area IV, June-September. Factors influencing the maturation cycle of king mackerel are not well known. Presumably, photo- period and water temperature are important for spawning, egg, and larval development. Beau- mariage (1973) indicated that seasonal changes in photoperiod influenced the spawning of king mackerel while McEachran et al. (1980) noted that larvae were more abundant at temperatures from 20.2° to 29.8°C and salinities from 28.2 to 34.47oo. A study by MacGregor et al. (1981) also showed that the levels of serum androgens and estrogens may be indicators of maturation in king mackerel. Our inferences on spawning peaks and activity of king mackerel, as determined by largest mean EDs, usually coincided with those of other studies. Our largest mean ED of 0.61 mm agrees with the 0.60 mm reported by Alves and Tome (1967). In contrast, the largest mean ED of 0.33 mm shown by Beau- mariage (1973) suggests that most of his fish were not ready to spawn. Our largest mean egg sizes from northwest Florida fish were similar to those re- ported by Beaumariage (1973) and probably in- dicates that spawning activity off the west coast of Florida is not extensive. Peak spawning months by area in this study were area I, August; area II, May; area III, August; and area IV, July. In the north- western and northeastern gulf, (our areas I and III) the highest catches of larval king mackerel occurred during September (Dwinell and Futch 1973; McEachran et al. 1980). Houde et al. (fn. 2) stated that because of their rare catches of larvae, king mackerel does not appear to spawn frequently in the eastern gulf. The reproductive cycle of king mackerel off the coast of Brazil is probably similar to that of this species from American waters. Ivo (1972) noted that spawning occurred throughout the year off the state of Ceara which is south of the Equator. Other studies indicate that they begin to spawn from Octo- ber through December (Menezes 1969) with peaks in November and March (Gesteria and Mesquita 1976). Since the seasons are reversed in this area, they would correspond to our spring and late sum- mer spawning peaks for king mackerel. We were unable to determine the number of times individual king mackerel spawn during the year from the data. Beaumariage (1973) concluded that king mackerel were multiple spawners, based on their extended spawning season and presence of several modal groups of yolked eggs. Morse (1980) reported that individual Atlantic mackerel, Scomber 846 FINUCANE ET AL.: REPRODUCTIVE BIOLOGY OF KING MACKEREL 3.0 1- 2.5 2.0 1.5 1.0 0.5 0 2.5 2.0 1.5 1.0 0.5 0 NORTH CAROLINA SOUTH CAROLINA AREA IV N = 50 -J I L J I L J I I I ' ' NORTHWEST FLORIDA woo O 1.5 z 1.0 2 0.5 * 0 2.5 2.0 1.5 1.0 0.5 0 _l L AREA III N=316 j i_j i LOUISIANA, MISSISSIPPI . TEXAS AREA I N=26 3 2 5 4 2 5 5 2 5 l i ' ' 6 2 5 7 2 5 8 2 5 _ 9 2 5 1 0 2 5 I I I l 1 1 2 5 1 2 2 5 ■ ' ' i 1 3 2 5 1 4 2 5 I L MIDPOINTS OF FORK LENGTH INTERVALS (mm) Figure 6.— Mean GSI plotted by midpoint of fork length interval for female king mackerel in each area. scombrus, are capable of spawning six batches of eggs during the spawning season. Documentation of spawning frequency and numbers of eggs pro- duced will require that king mackerel be held in captivity. Major spawning areas for king mackerel could not be determined during this study because of the scar- city of ripe fish. Gonad maturation data suggest that spawning occurs throughout the sampling areas but the magnitude of spawning and extent of spawning areas are unknown. Ichthyoplankton surveys con- ducted by Wollam (1970), Houde et al. (fn. 2), and McEachran et al. (1980) have revealed general spawning locations of king mackerel by the occur- rence of small larvae (<3 mm SL). These studies in- dicate that spawning probably occurs over the con- tinental shelf of the northwestern and northeastern Gulf of Mexico. Most small larvae collected by McEachran et al. (1980) were captured over the mid- dle and outer continental shelf in water depths of 35-130 m off the Texas coast. No comparative fecundity data were available from the southeastern U.S.; however, Ivo (1974) determined fecundity for 39 fish from Brazilian waters. He found great variation in fecundity for fish with the same fork length. The fact that disjunct spawning appears to occur off the Carolinas and in the northcentral and west- ern Gulf of Mexico from spring through fall may sug- gest separate stocks of king mackerel in these areas. 847 FISHERY BULLETIN: VOL. 84, NO. 4 Table 3.— Summary of data on king mackerel for which fecundity was estimated, 1977-78. Fork Gonad Total Gonad Fork Gonad Total Gonad length weight weight index Fecundity length weight weight index Fecundity Date (mm) (g) (g) (x100) Age (estimated) Date (mm) (g) (g) (x100) Age (estimated) Area I (Texas) Area III (Northwest Florida) 8/26/78 500 18.16 900 2.02 2 185,608 8/8/77 508 9.24 944 0.98 1 196,938 7/8/78 650 55.66 2,270 2,45 — 985,340 8/8/77 568 11.95 1,318 0.91 1 160,722 7/26/78 750 76.09 3,042 2.50 4 1,082,301 8/7/77 608 24.70 1,950 1.27 1 404,982 8/26/78 760 35.23 3,166 1.11 — 466,252 7/2/78 652 39.53 2,497 1.58 1 688,354 7/8/78 770 92.87 3,405 2.73 2 1,194,283 8/14/77 727 105.96 3,180 3.33 2 1 ,640,497 7/8/78 800 142.53 4,086 3.49 4 2,009,870 7/14/77 780 139.64 3,424 4.08 — 2,102,579 7/8/78 810 135.50 4,313 3.14 — 1,435,752 6/27/78 816 301.26 4,450 6.64 — 5,049,856 7/8/78 835 82.96 4,540 1.83 4 1 ,380,342 8/7/78 826 167.31 4,903 3.41 2 2,912,649 7/8/78 860 176.00 4,994 3.52 5 2,753,638 6/19/77 862 186.02 4,680 3.98 4 2,509,948 7/8/78 870 130.19 4,994 2.61 5 2,236,664 7/4/78 906 210.33 5,630 3.74 6 3,005,716 8/7/78 895 212.86 5,448 3.91 6 2,309,622 6/27/78 929 96.05 6,492 1.48 — 1,891,588 — — 239.41 4,183 — 3 4,183,921 7/13/77 980 205.49 8,170 2.52 — 3,346,332 Area II (Louisiana) 7/20/77 6/19/77 1,018 1,087 268.22 602.45 12,700 1 1 ,350 2.11 5.31 7 4,960,702 5,744,230 6/24/78 446 8.43 681 1.24 — 69,264 8/24/77 1,108 476.06 9,768 4.87 — 5,836,910 6/23/77 635 13.45 1,930 0.70 1 182,863 9/5/78 1,142 538.60 12,031 4.48 8 8,070,585 6/20/77 710 26.92 2,500 1.08 2 2,570,133 8/14/78 1,220 575.39 14,437 3.99 7 7,489,089 9/13/77 852 96.36 4,380 2.20 4 1,179,625 7/13/78 895 158.51 5,130 3.09 4 2,079,204 Area IV (North Carolina) 8/15/77 951 239.09 6,221 3.84 — 4,448,492 7/13/77 617 171.17 5,765 2.97 — 2,625,338 5/20/78 972 451.68 7,310 6.18 6 6,319,134 9/9/78 780 131.11 3,632 3.61 3 1,667,418 5/20/78 994 577.18 11,120 5.19 6 5,890,631 7/28/78 841 207.22 5,766 3.59 4 2,330,248 7/7/78 1,025 325.56 9,000 3.62 6 4,686,248 9/21/78 844 100.07 4,722 2.12 4 969,206 8/7/78 1,037 417.00 8,325 5.01 6 6,437,542 7/26/78 865 150.50 4,631 3.25 — 1,639,189 6/23/78 1,055 314.33 8,626 3.64 11 4,686,598 7/15/78 869 227.67 4,767 4.78 — 2,795,451 9/3/77 1,086 303.52 9,750 3.11 — 5,401,961 9/9/78 880 119.88 4,858 2.47 5 1 ,236,055 6/25/78 1,109 247.66 9,534 2.60 — 2,771,744 7/1/78 900 170.57 6,628 2.57 4 3,321,377 6/25/78 1,149 401.59 10,896 3.69 9 4,268,537 8/27/78 972 214.00 7,173 2.98 6 3,204,055 6/16/78 1,178 478.74 13,286 3.60 6 8,899,756 8/27/78 996 282.01 7,718 3.65 6 2,652,453 7/10/78 1,194 447.88 9,045 4.95 10 6,010,133 9/9/78 1,000 267.35 6,992 3.82 8 2,797,301 4/29/78 1,220 498.52 15,150 3.29 9 7,315,781 8/30/78 1,050 416.04 9,988 4.17 8 6,102,347 5/20/78 1,229 698.36 14,070 4.96 6 10,116,890 6/17/78 1,265 611.04 15,095 4.05 9 9,209,082 6/17/78 1,291 468.64 15,890 2.95 10 7,487,826 8/10/77 1,312 583.79 17,120 3.41 — 6,689,189 5/20/78 1,316 840.08 17,800 4.72 10 10,711,026 6/17/78 1,370 570.66 19,885 2.87 11 7,650,064 8/15/78 1,489 815.00 25,610 3.18 13 12,206,888 Table 4.— Regressions of fecundity (F) on total weight (TW), fork length (FL), and age (A) of king mackerel by areas. Area Predictor Equation r2 I TW F = 8.554 X 101(TW)1465 0.745 (TX) FL F = 8.816 X 10-7(FL)4 206 0.781 A F = 2.487 X 105(A)1.390 0.373 II TW F = 1.475 X 101(TW)1381 0.847 (LA-MS) FL F = 9.973 X 10_7(FL)4175 0.840 A F = 4.207 X 105(A)1313 0.721 III TW F = 1.327 X 101(TW)1408 0.877 (NWF) FL F = 1.918 X 10_7(FL)4455 0.884 A F = 4.684 X 105 + 9.494 x 105(A) 0.870 IV TW F = 1.419 X 106 + (6.658 x 102)TW 0.760 (NC-SC) FL F = -2.554 x 106 + (5.840 x 103)FL 0.257 A F = -2.778 x 105 + (5.579 x 105)A 0.436 l-IV TW F = 1.854 X 101(TW)1361 0.856 (All areas) FL F = 4.391 X 10-6(FL)3 974 0.820 A F = 3.399 X 105(A)1356 0.730 848 FINUCANE ET AL.: REPRODUCTIVE BIOLOGY OF KING MACKEREL Williams and Godcharles5 have postulated on the basis of mark-recapture data that two migratory groups occur: one in the South Atlantic and the other in the Gulf of Mexico. Both of their ranges overlap in south Florida. ACKNOWLEDGMENTS We thank Dale S. Beaumariage, Churchill B. Grimes, and Steven A. Bortone for their critical review of this manuscript. LITERATURE CITED Alves, M. I. M., and G. S. Tome. 1967. Alguns aspectos do desenvolvimento maturativo das gonadas da cavala, Scomberomorus cavalla (Cuvier, 1829). Arq. Estac. Biol. Mar. Univ. Fed. Ceara 7(l):l-9. Baglin, R. E., Jr. 1982. Reproductive biology of western Atlantic bluefin tuna. Fish. Bull., U.S. 80:121-134. Beaumariage, D. S. 1973. Age, growth, and reproduction of king mackerel, Scom- beromorus cavalla, in Florida. Fla. Mar. Res. Publ., No. 1, 45 p. Brusher, H. A., L. Trent, and M. L. Williams. 1978. Recreational fishing for king mackerel in Bay County, Florida during 1975. In C. B. Austin et al. (editors), Mackerel Workshop Report, p. 120-142. Univ. Miami Sea Grant, Spec. Rep., No. 14. Clark, F. N. 1934. Maturity of the California sardine (Sardina caerulea), determined by ova diameter measurements. Calif. Dep. Fish Game, Fish Bull. 42, 49 p. Collette, B. B., and J. L. Russo. 1979. An introduction to the Spanish mackerels, genus Scom- beromorus. In E. L. Nakamura and H. R. Bullis, Jr. (editors), Proceedings of Colloquium on the Spanish and King Mackerel Resources of the Gulf of Mexico, p. 3-16. Gulf States Marine Fisheries Commission. No. 4. 1984. Morphology, systematics, and biology of the Spanish mackerels (Scomberomorus, Scombridae). Fish. Bull., U.S. 82:545-692. DWINELL, S. E., AND C. R. FUTCH. 1973. Spanish and king mackerel larvae and juveniles in the northeastern Gulf of Mexico, June through October 1969. Fla. Dep. Nat. Resourc, Mar. Res. Lab., Leafl. Ser. 4 (pt. 1, no. 24), 14 p. FINUCANE, J. H., AND L. A. COLLINS. 1977. Environmental assessment of an active oil field in the northwestern Gulf of Mexico, 1976-1977. Ichthyoplankton, NOAA Final Report to EPA, 150 p. 1984. Reproductive biology of cero, Scomberomorus regalis, from the coastal waters of south Florida. Northeast Gulf Sci. 7(1):101-107. Gesteira, T. C. V., and A. L. L. Mesquita. 1976. Epoca de reproducao tamanho e idade na primeira 5Williams, R. 0., and M. F. Godcharles. Fisheries stock assess- ment, king mackerel tagging and stock assessment. Annual report for 1981-82. Florida Department of Natural Resources, Marine Research Laboratory, 100 Eighth Ave., S.E., St. Peters- burg, FL 33701. desova da cavala e da serra, na costa do Estado do Ceara (Brasil). Arq. Cienc. Mar. 16(2):83-86. Hunter, J. R., and S. R. Goldberg. 1980. Spawning incidence and batch fecundity in northern anchovy, Engraulis mordax. Fish. Bull, U.S. 77:641-652. Ivo, C. T. C. 1972. Epoca de desova e idade na primeira maturacao sex- ual da cavala, Scomberomorus cavalla (Cuvier), no Estado do Ceara. Arq. Cienc. Mar. 12(l):27-29. 1974. Sobre a fecundidade da cavala, Scomberomorus cavalla (Cuvier), em aguas costeiras do Estado do Ceara (Brasil). Arq. Cienc. Mar. 14(2):87-89. Johnson, A. G., W. A. Fable, Jr., M. L. Williams, and L. E. Barger. 1983. Age, growth, and mortality of king mackerel, Scom- beromorus cavalla, from the southeastern United States. Fish. Bull., U.S. 81:97-106. MACGREGOR, R. M., Ill, J. J. DlNDO, AND J. H. FINUCANE. 1981. Changes in serum androgens and estrogens during spawning in bluefish, Pomatomus saltator, and king macker- el, Scomberomorus cavalla. Can. J. Zool. 59:1749-1754. Manooch, C. S., III. 1979. Recreational and commercial fisheries for king mack- erel, Scomberomorus cavalla, in the South Atlantic Bight and Gulf of Mexico, U.S.A. In E. L. Nakamura and H. R. Bullis, Jr. (editors), Proceedings of Colloquium on the Spanish and King Mackerel Resources of the Gulf of Mexico, p. 33-41. Gulf States Marine Fisheries Commission No. 4. Manooch, C. S., Ill, E. L. Nakamura, and A. B. Hall. 1978. Annotated bibliography of four Atlantic scombrids: Scomberomorus brasiliensis, S. cavalla, S. maculatus, and S. regalis. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ. 418, 166 p. McEachran, J. D., J. H. FINUCANE, and L. S. Hall. 1980. Distribution, seasonality and abundance of king and Spanish mackerel larvae in the northwestern Gulf of Mex- ico (Pisces: Scombridae). Northeast Gulf Sci. 4(1):1-16. Menezes, M. F. 1969. Alimenta?ao da cavala, Scomberomorus cavalla (Cuvier), em aguas costeiras do Estado do Ceara. Arq. Cienc. Mar. 9(l):15-20. Moe, M. A., Jr. 1963. A survey of offshore fishing in Florida. Fla. State Board Conserv. Mar. Lab., Prof. Pap. Ser. 4, 115 p. Morse, W. W. 1980. Spawning and fecundity of Atlantic mackerel, Scomber scombrus, in the Middle Atlantic Bight. Fish. Bull., U.S. 78:103-108. Otsu, T., and R. N. Uchida. 1959. Sexual maturity and spawning of albacore in the Pacific Ocean. Fish. Bull., U.S. 59:287-305. Ricker, W. E. 1975. Computation and interpretation of biological statistics offish populations. Fish. Res. Board Can., Bull. 191, 382 p. Steel, R. G. D., and J. H. Torrie. 1960. Principles and procedures of statistics (with special references to biological sciences). McGraw-Hill Book Co., Inc., N.Y., 481 p. U.S. Department of Commerce, National Marine Fisheries Service. 1985a. Fisheries of the United States, 1984. Current Fish- ery Statistics 8360, 121 p. 1985b. Marine recreational fishery statistics survey, Atlan- tic and Gulf Coasts, 1983-1984. Current Fisheries Statis- tics 8326, 222 p. 849 FISHERY BULLETIN: VOL. 84, NO. 4 Wollam, M. B. Dep. Nat. Resourc. Mar. Res. Lab. Tech. Ser. 61, 35 p. 1970. Description and distribution of larvae and early juven- Yuen, H. S. H. iles of king mackerel, Scomberomorus cavalla (Cuvier), and 1955. Maturity and fecundity of bigeye tuna in the Pacific. Spanish mackerel, Scomberomorus maculatus (Mitchill), Spec. Sci. Rep. Fish. 150, 30 p. (Pisces: Scombridae) in the western north Atlantic. Fla. 850 NEW OCCURRENCE OF EPIZOOTIC SARCOMA IN CHESAPEAKE BAY SOFT SHELL CLAMS, MYA ARENARIA C. A Farley,1 S. V. Otto,2 and C. L. Reinisch3 ABSTRACT Maryland soft shell clams, My a arenaria, from Chesapeake Bay were sampled from 1969 through January 1983. Four cases of sarcomatous neoplasia were diagnosed histologically [1979 (1), 1982 (2), January 1983 (1)] in 3,584 animals. Hemocytologic sampling between December 1983 and May 1984 revealed peak prevalences of 42-65% in clams from five sites. Sarcomas in laboratory-held clams progressed from early to advanced stages and death. This is the first time epizootic neoplastic disease has been observed in a wild molluscan population which was previously documented to be sarcoma-free. An infectious etiology is implied and data indicate the potential for mass mortality of bay clams. Neoplastic diseases in soft shell clams, Mya arenar- ia, have been reported from New England popu- lations in both polluted and nonpolluted areas (Barry and Yevich 1975; Farley 1976a; Yevich and Barszcz 1977; Brown et al. 1977, 1979; Brown 1980; Cooper et al. 1982a; Reinisch et al. 1984). Generally, the types of neoplasia noted have been considered as having hemocyte (blood cell) (Yevich and Barszcz 1977; Brown et al. 1977, 1979; Brown 1980; Cooper et al. 1982a; Reinisch et al. 1984) and gonadal (Barry and Yevich 1975; Yevich and Barszcz 1977; Brown et al. 1977, 1979; Brown 1980) origins or have been designated as sarcomatous (Farley 1976a). A single neoplastic clam was reported from Chesapeake Bay with an apparent teratoma composed of nerve and muscle tissue and digestive epithelium (Harshbarger et al. 1977). Chesapeake Bay soft clams collected and examined by several authors between 1971 and 1978 were free of the neoplastic disease (Barry and Yevich 1975; Brown 1980) with the exception of 1 case found in a collection of 3,000 clams used as ex- perimental controls (Brown 1980). Evidence for a viral etiology for hematopoietic neoplasia in clams was reported in a Rhode Island study (Oprandy et al. 1981). Improved techniques such as examination of hemolymph using a combination of histologic and cytologic procedures (Cooper et al. 1982b) and the development of a monoclonal antibody test specific for neoplastic clam cells (Reinisch et al. 1983) have facilitated the identification and diagnosis of the disease. High prevalences of sarcomas have been ■Northeast Fisheries Center Oxford Laboratory, National Mar- ine Fisheries Service, NOAA, Oxford, MD 21654. 2Aspen Cove, Bozman, MD 21612. 3Tufts University School of Veterinary Medicine, Boston, MA 02111. found repeatedly in populations of Chesapeake Bay clams. This paper documents the first occurrence of epi- zootic sarcoma in soft shell clams in Chesapeake Bay, and the first time neoplastic disease has ap- peared in a wild molluscan population that was previously shown to be free of the disease. Epizootic prevalences of this condition may have a potential- ly devastating impact on the clam industry of the region. MATERIALS AND METHODS Sixty samples of 25 or more soft shell clams (total- ing over 3,500 clams) have been collected periodi- cally by the Maryland Department of Natural Resources (DNR) or purchased from seafood outlets from 51 sites in Chesapeake Bay since 1969. Each animal was necropsied and tissues were fixed, pro- cessed, and diagnosed histologically via standard methods (Howard and Smith 1983) for diseases and parasites. Recent samples (Table 1) were examined by cytologic methods to determine the percent prevalence and number of abnormal cells. Late spring samples (YCLP, YSWP, YAGH, and YPIS, Table 1) were diagnosed by both histology and histo- cytology (technique described below). Hemolymph was drawn via hypodermic syringe into sterile, ambient (15%o), artificial seawater to produce a 1:9 dilution of cells to seawater. One milli- liter of this sample was placed on a 25 mm, cham- bered, poly-L-lysine coated microscope slide and cells were allowed to settle for 1 h (the poly-L-lysine coating improves the adhesiveness of neoplastic cells which in vitro are rounded and do not usually stick to glass [Cooper 1982a]). Fluid and chambers were Manuscript accepted May 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 851 FISHERY BULLETIN: VOL. 84, NO. 4 C CO CO Q. E CO CO ^- c co o « ° 5 co 5 CO « Q. °r P E 1— CL o o U a o E o o M— CO CO T3 *~ C •- CO 0> .y o O) O 1° ■5. ■= o §> ■a o £ J= iS en CO CO c ? CD ^ °> E 3 O CD i o o ■D o CD o CO O CD co 5 T3 CO co E £ « to 4> O co CO "5 !c -a 'I co >> CD "2 O *- CD O (11 L. m CO F c/> n u ci (1) »_ > CO CO 0 »*— i_ 0 co 0 E r u CD ^ co > CO I— CL III .J CO < T3 CD , C CO CD o >> o — CO -zr ~P CO O ° o- a. .5 CD .Q O) 0-^.0 C T3 — OJ CD O ^ 5 o g^ ti:«.!2 T3 CO CD CD CO o c E LO CO CD °^ C «2 — CO to CD CO o c CO > < E*~ o co ~P CO CO <° to ofe£ "CO co Z Q. CD Q.0) ET3 CO o CO CD 1^ CM II I I I I I I 01 C\J CD O ^ I I o cm cm t-- CO LO CM I I CO ■«* CO CO O CO O CM ■* CM ■* LO *t oocoNinoimin CO CM CO CO t- ICIO^OOOO* 1- CM CM OOONtNOllfl T- T- CO CO T- 00 O CO O ** CO CM t- O ■>- CM CO LO OOO CO o CO CO j» to f- 0 0 0 Q. E CO n ■* ■* co 0 CO T3 CD > !c 0 CM O CO CD LO CO CM 00 co ■* o o o o t- 1- n m w lo in X0DZOW- >- > > OOOOO00CMCM o LOLO^-LOLOCOLOLO -~ CO CM CO LO t- CM CO 0- CL CL Q. Q. CL COCLCOCLC0COCOW >>>>>>>> CO CO ■* •*•*•* ■* * •* •* ■* ■* LO LO 10 >>CD CO CO CO CO 00 00 CO CO CO CO CO CD CO 00 CD 00 CO N- CD a a 0 0> 0> o> CTi O) CF> Oi Gh 05 a> O) ^ ci> £ CO Q CD CD CD QQLL Q_ o_ CO CO <<52 CD c 3 —3 q> o)-j3 0 <> CO 0 CD Q CO CZ co ■7 CO CD t- o> ■<* ^ CO CM -3 ~5 r^ CTi CM r^ en CZ C C CO TO T3 T3 »_ TD 5 5 5 . . ** IB C ^— c _ c — .^ *_ ^j CD CD OOO Point Poin Sisti CO C co c a) c c c c > -O £ c c c _cn O co 0 co 0 0 0 O ofl LX E CO _^£ -2£ Jul c c c j_ Cl — CL — CL CL CL CL >. co 3 3 3 ong wan hree CO c CO C CO c c c C CO if) CD O £ £ £ a. 0 CO 5 Q. CO Q. CO 0 5 0 5 CO S co 5 CO 5 CO (J5 V) CO _J C/3 H CL s C/>CLC/50_COC/)C/)cy3 a> o3 L. CD CD CD ■0 to CO CD > a. 0. 0. E E E _CD LU DC CO CO CO O. > c 0 0 0 _i CO CO CO CD CD > > ^ CO CO CO CD ^ CD ^ CD CO 2 CO 2 CD > > CD > if LT mmerci mmerci mmerci er R er R Rive m c 0 ncDLicDiccira: 5£cd£cdcdcdcd c eg £ V> (D ^Z O O x: O UJ GO *; CD CD £ 2 tr5 1- to CD 2 «5 2 tf5 (0 CD CO CD CO CD CO CD CO CD Q. 0 ffl 0 O O -C ^ ^ 00000 5 CO JZ co -c co ^: JZ ^Z .CZ .c S3 LU OUJOLUOOOO 0 _l 852 FARLEY ET AL.: EPIZOOTIC SARCOMA IN SOFT SHELL CLAMS removed, while slides were wet-fixed in an aldehyde fixative (1% glutaraldehyde/4% formaldehyde) (McDowell and Trump 1976) in half ambient sea- water and stained with Feulgen picromethyl blue (Farley 1969), dehydrated, and mounted with a coverslip using a synthetic mounting medium. We are designating the term "histocytology" to describe this technique. The significance of this method is that the monolayer preparations, which result from treating living cells with histologic procedures, are permanent. Cytologic artifacts are minimal and cases can be accurately staged using cell counting procedures. Since histocytologic preparations con- tain between 100,000 and 500,000 cells in a mono- layer, very early stages of the proliferative process can be diagnosed. Staging is arbitrarily determined by estimating the number and determining the ratio of both normal and neoplastic cells (Table 2). A similar diagnostic and staging method using cyto- logic techniques was reported by Cooper et al. (1982b); however, our method appears to have bet- ter accuracy and increased sensitivity to light cases. Diagnosis of histologic sections is reliable for stages 3-5 (Fig. 1A). As an example, comparison of late spring samples shows that histocytology is the more sensitive method while histology alone clearly demonstrates a massive increase in prevalence from zero in 1969-78 to 29.5% in 1984 (Table 1). Monoclonal antibody was developed against neo- plastic clam cells from Massachusetts clams by tech- niques described elsewhere (Reinisch et al. 1983). Periodic histocytologic diagnosis and mortality ob- servations were made on clams held in 55 L aquaria with 15°/oo, 10 °C artificial seawater, circulated through floss and charcoal filtering systems. RESULTS Sarcomas in clams were diagnosed histologically in 1/25 in November 1979 from Eastern Bay; 1/25 in May 1981 from West River; 1/50 in November 1981 from Little Choptank River; and 1/75 in Janu- ary 1983 from Chester River. In December 1983, histocytologic diagnoses of clams obtained from a local seafood restaurant showed 8/18 with sarcomas. An intensive survey and study of local populations was initiated in December 1983 to evaluate the ex- tent of this apparently new epizootic in Chesapeake Bay soft shell clams. Table 1 presents epizootiology of field collections while Table 2 shows comparable information on laboratory-held clams. Field preva- lences were found to be high in most samples from December 1983 through April 1985. At the same time, disease intensities which were light in Decem- 03 co ■5. 22 oo o E a> .c o co S 03 O .y c &< 03 ' O *- CO II I— C CO 03 — ' g 03 M II 03 03 E (5 03 03 03 £9 5 ii "~ C\J 03 ^ 3 _ 2 « CO o 03 fj, "~ O y 9 7 Q ° 2 »d'a P CO 5 II co I- s 03 N~' Q. 03 — O O LU O 03 .. o 2 g S. 03 rjj o "? « T3 rT-co J> o . .£ « S S O ffl f -O Q3 C JO 03 03 _- T3 i? CO -- 3 3 O 03 > 03 c c M g c o 03 CO "D 03 O c $ S c? Q. 03 CO (0 # £ Q. O O c ° «5 52 2 CO 03 ^ I -§ II c\i fc LU LO CO TJ T3 < C C I— CO CO 03 <5 a a. E CO CO CO 03 E E 5* O 03 Q. O 03 -z. 03 >>• o — TO . o => c S- II E o -- 03 lis-1- = r S~ ii E o ^ aE s 03 T CO C-- o CO T3 t= CD E o 03 a. co CO Q ooinin Jiominooii) ■^■■-CVJCAJCAJtJ-lOLn ooooooooooo OJCAJCAICAJCMCM^-I^O T-T-T-T-t\l(OIO(CO) i 2 2, 2- 2- 2- 2- 2 2- "S I in cm o t-~ ?T io lo of ~ (OiocoinoicncDQ o I o~ o~ c\T c\T c\T co" ^" | ^ ^ ^, ^_- -~-- -^ ^ ^ ^ ciddocM i ~ S I SSSS222 I ^ C0O3 i- 00 "O CJ3 Q ss"S 0!D~ r^- i^- Q COCOCO'<*'^-^-'*' < Artemia TRIALS B Artemia AND LARVAL TRIALS 0 12345678 9 10 1112 13 14 15 TRIAL NUMBER Figure 1.— Variation in feeding performance of northern anchovy predators fed live Artemia as a function of trial number (equivalent to elapsed time of experiment); shaded area indicates trials in which northern anchovy were fed only Artemia; un- shaded areas, Artemia trials alternated with larval trials. A, percent predator error in capturing adult Artemia (percentage of attacks in which a northern anchovy missed the prey); dashed line indicates mean. B, mean time required for predator group to capture 3 adult Artemia; bars are 2 x SE of the mean. (N = 21.) Success of Avoidance Movements Larval vulnerability depended not only on the responsiveness but also on the success of avoidance movements. The proportion of larvae escaping northern anchovy predators increased from 8% for 6 mm larvae to 92% for 33 mm larvae with an esti- mated 50% of the 17 mm larvae escaping. The percentage of larvae escaping the attacks of chub mackerel was lower than for adult northern an- chovy, but the curves given in Figures 2 and 3 had a similar form. Weibull curves were fit to the data to provide trend lines (equations and parameters given in Figure legends). The fraction of larvae that escaped increased from 6% of 6.7 mm larvae to an estimated 50% of the 30 mm larvae. Of the 50 mm juvenile northern anchovy used as prey only 64% escaped the attacks of the chub mackerel. The ability to successfully avoid predator attacks was strongly affected by species-specific differences in predator behavior since the fraction of larvae escaping the attacks of northern anchovy increased much more rapidly with larval length than did the fraction escaping the attacks of chub mackerel. In contrast, the fraction of smaller larvae (SL <20 mm) responding to the attacks of these two predators was similar (Fig. 4). This indicates that probability of a larva responding to an attack is less affected 862 FOLKVORD and HUNTER: VULNERABILITY OF NORTHERN ANCHOVY LARVAE PREDATOR - Engraulis mordax Escaping attack Taken in 5 min. i i i i B / Responding to attack i i i i i i i 2 6 10 14 18 22 26 30 34 38 42 46 50 LARVAL STANDARD LENGTH (mm) Figure 2.— Vulnerability of northern anchovy larvae to adult northern anchovy predators as a function of larval length. A, Percentage of northern anchovy larvae escaping attack; bars are 2 x SE; line is Weibull curve fit to six points using Marquardt's least squares method (Pielou 1981); equation is N = K(\ - exp (1 - (Lib) -A)) where K = 0.93, 6 = 17.85, A = 2.85, N is the percentage of larvae and L = larval length; and predation rate of the northern anchovy predators (percentage eaten in 5 min) where dashed line is data when ex- periments with biased predator feeding motivation are omitted. B, Percentage of northern anchovy larvae that responded to the attack of an adult northern anchovy; bars are 2 x SE; and Weibull parameters for curve are K = 1.00, b = 13.58, and A = 1.94. by differences in predator behavior than is its suc- cess in avoiding the attack. The success of avoidance movements can be separated from larval responsiveness by calculating the avoidance success of responding larvae (numbers escaping/numbers responding). Webb (1981) found no change in this fraction over the larval size range he examined (3-12 mm SL), indicating that changes in responsiveness alone were responsible for the decline in the vulnerability of northern anchovy lar- vae to Amphiprion with increasing larval length. In the present study, no significant trend existed in the success of avoidance movements over the size range of larvae studied by Webb (1981) but success of avoidance movements greatly increased in larger larvae (Fig. 5). The figure also indicates that north- ern anchovy larvae were much more successful in avoiding Amphiprion than in avoiding adult north- ern anchovy and that the larvae had the least suc- cess in avoiding chub mackerel. 863 FISHERY BULLETIN: VOL. 84, NO. 4 PREDATOR - Scomber Japonicus tu < > < -I > o z o z < I- z UJ o £C UJ 100 80 60 40 20 100(0) 80 60 40 20 TAKEN IN 5 MINUTES ESCAPING ATTACK ■ -ri > i i i i i i i i i i i i ■ i ' i i_j i i ' /, RESPONDING TO ATTACK ''''■■'■ ' ' ' ' ' ' J_L I I I 2 6 10 14 18 22 26 30 34 38 42 46 50 LARVAL STANDARD LENGTH (mm) Figure 3.— Vulnerability of northern anchovy larvae and juveniles to juvenile chub mackerel predators as a function of anchovy length. A, percentage of northern anchovy larvae escaping attack; bars are 2 x SE; line is Weibull curve fit to six points using Marquardt's least squares method (Pielou 1981); equation is N = K(\ - exp (1 - (Lib) - A)) where K = 0.66, b = 27.41, N = percentage of larvae, L = larval length, and A = 2.12; and predation rate of chub mackerel predators (percentage eaten in 5 min) where dashed line is data when experiments with biased predator feeding motivation are omitted. B, percentage of northern anchovy lar- vae that responded to the attack of a chub mackerel, bars are 2 x SE; and Weibull parameters for curve are K = 0.93, 6 = 12.61, and A = 1.24. Predation Rates The predation rate of northern anchovy (propor- tion of larvae consumed by northern anchovy predators in 5 min) reached a maximum somewhere between larval lengths of 8.5 and 15 mm when all data were used, but it occurred between larval lengths of 8.5 and 11 mm when we deleted the ex- periment where northern anchovy predator perfor- mance was lower than average (dashed line in Figure 2A). Statistical comparisons of the fraction of larvae consumed in the various size classes in- dicated that 6.8 mm larvae were taken less often than larvae in 8.5, 11, and 15 mm size classes despite the fact that these larvae had a low escape ability (P < 0.05; normal approximation to the binomial mean; n = 35, 48, 40, and 60, for 5.9, 8.5, 11, and 15 mm size classes). Owing to their small size and 864 FOLKVORD and HUNTER: VULNERABILITY OF NORTHERN ANCHOVY LARVAE St -J o U. H O o H Z Z Q UJ Z O O OC a Q. uj OC 100 80 60 40 20 Amphiprion percula Qr 2 Gf. 7 Engraulis mordax -Gr-3 -L«7/aDon*icus Scomber »aP J L 6 10 14 18 22 26 30 34 38 42 46 50 STANDARD LENGTH (mm) Figure 4.— Percentage of northern anchovy larvae that responded to attacks by adult northern anchovy (lines for the three different predator groups shown separately), chub mackerel and the aquarium fish Amphiprion percula (from Webb 1981). o z a. < O s < 2 > oc < o J Q. o Oz £1 o OL (/> UJ OC 1.00 0.8 0.6 0.4 0.2 0.0 Amphiprion percula (Webb, 1981) X _L 10 14 18 22 26 30 34 38 LARVAL STANDARD LENGTH (mm) 42 46 50 Figure 5.— The percentage of responding northern anchovy larvae in various length classes that escaped the predator. Species names identify the predator species. lack of pigmentation, 6 mm larvae may have been less visible to the predators than larger larvae and consequently were detected less frequently. The decline in predation rates in larvae longer than 15 mm was the result of their greater escape ability. The number of larvae consumed in 5 min was an insensitive measure of predation rates of chub mackerel, because they usually ate all larvae in the tank within 5 min regardless of their size. Only in the two smallest larval size groups (6-10 mm SL) were some larvae left after a 5-min elapsed time; the deletion of one experiment because of low chub mackerel predator performance changed the preda- tion rate on 10 mm larvae from 87 to 95% (dashed line in Figure 3A). These adjusted data indicate that the feeding rate of chub mackerel was lowest when 865 FISHERY BULLETIN: VOL. 84, NO. 4 the smallest larval size group (6.7 mm) were the prey. Predator Behavior Sighting distances, persistence of the attack, at- tack speed, and other characteristics of predator behavior were not well documented in our experi- ments because our focus was on the larvae. Such information could be quite useful if one were to develop a predation model for northern anchovy lar- vae, using northern anchovy or chub mackerel predators. We provide some general observations on the behavior of the predators. Chub mackerel attacked 6.7 mm larvae from a shorter distance than larger larvae (£-test, P = 0.05), but no statistically significant trend was evident when northern anchovy were predators. Mean at- tack distances were a poor measure of sighting range as they included repeated, short-range attacks on the larger larvae. We observed both predator species swimming within 2-3 dm of the smallest lar- vae without attacking them, whereas larger larvae were always attacked from this distance indicating that sighting distances may be shorter for small larvae. Adult northern anchovy usually attacked a larva only once during a feeding sequence, and if the larva escaped, it was rarely attacked again or pursued. On the other hand, if the chub mackerel did not cap- ture the larva on the first attack, it usually turned and attacked again. Chub mackerel usually chased an escaping larva until it was captured. The attack speeds of adult northern anchovy, although not measured, seemed to be similar over a wide range of larval prey sizes, whereas the attack speeds of chub mackerel clearly were faster when attacking larvae greater than about 10 mm SL than when at- tacking smaller larvae. DISCUSSION Factors Affecting Larval Vulnerability A low level of responsiveness seems to be the dominant feature of the vulnerability of northern anchovy larvae to fish predators over the smallest larval size classes we tested (6-10 mm SL). Presum- ably northern anchovy larvae <6 mm would respond even less frequently, as Webb (1981) found that only 9% of 2.9 mm northern anchovy larvae responded to the aquarium fish Amphiprion percula, whereas about 30% of 6 mm larvae did so. During this period vulnerability of northern anchovy larvae to fish predators seems to be primarily a function of visual detection by the predator, because when the larvae are detected they have a low probability of escap- ing. Our data on predation rates and maximum at- tack distances indicate that predation in the sea on the small, young larval stages might be lower than expected because of the short range at which such larvae may be detected. Thus, factors that affect the distance at which larvae are detected by predators, such as larval size, visual contrast, and water clar- ity (Vinyard and O'Brien 1976), may be the most im- portant variables during the first 3 wk of life. As larvae grow they more often respond to the attacks of predators and escape them more frequently. Maturation of visual and lateral line systems (O'Con- nell 1981) may be the principal cause of this general increase in responsiveness with larval length. Al- though older larvae are more responsive, they are also more readily detected by predators because they are larger and have more pigmentation. Im- proved avoidance behavior may not completely com- pensate for the greater visibility of larvae in the 8-12 mm range, as our data on predation rates by north- ern anchovy indicated that the rates of consump- tion were highest for larvae in this range. Larvae longer than 20 mm responded more fre- quently to northern anchovy than to chub mackerel predators, possibly because chub mackerel attacked such large larvae at much higher speeds. At higher attack speeds, less time is available for the larvae to respond; consequently, predators with the most rapid attack speeds evoke the lowest proportion of prey responses (Webb 1982). Thus one might expect a larva to respond to small fish predators more fre- quently than to larger ones, since attack speed would be expected to increase with predator size. This may explain why northern anchovy larvae (2.9-12 mm SL) responded more frequently to the small Amphiprion (44 mm) (Webb 1981) than they did to either northern anchovy or chub mackerel predators (Fig. 4). The pectoral swimming of Amphiprion might also provide more cues of an im- pending attack than did the swimming movements of either northern anchovy or chub mackerel. In addition to size-specific avoidance capabilities and visibility, many other larval characteristics affect their vulnerability to predators. We briefly consider here three of these: effects of starvation, effects of the onset of schooling, and effects of varia- tions in larval growth rates. Clupeoid larvae undergo degradation of muscle and other tissues during star- vation, and a reduced predator avoidance behavior might be anticipated (Ehrlich 1974; O'Connell 1980). In a preliminary experiment Folkvord (1985) 866 FOLKVORD and HUNTER: VULNERABILITY OF NORTHERN ANCHOVY LARVAE reported that only 50% of starved, 33 mm northern anchovy larvae responded to the attacks of adult northern anchovy as compared with 100% for fed larvae. No starved 10 mm larvae escaped attack whereas 15-20% of the fed 10 mm larvae did so. The numbers of observations were insufficient for a statistical comparison, but recent work by Booman (unpubl. data, Southwest Fisheries Center, La Jolla, CA) indicates starvation can have a statistically significant effect on responsiveness of 10 mm north- ern anchovy larvae to adult northern anchovy predators. The effect of the onset of larval schooling was not considered in these experiments; however, escape and response probabilities of individual larvae may not be altered greatly by the onset of schooling. The work of Major (1978) indicates that the most impor- tant effect of schooling may be to reduce the rate of attack by predators. He also found that the ma- jority of Hawaiian anchovy captured by predators were isolated individuals that had moved away from the school, and predator success on schooled prey was similar to that on isolated prey. The onset of schooling in larval northern anchovy occurs between 11 and 15 mm SL, but the time spent in organized, cohesive schools increases throughout the northern anchovy's larval and juvenile periods (Hunter and Coyne 1982). Thus attack rates of predators might be expected to decline throughout later larval and juvenile life as the northern anchovy spends more time in cohesive schools. The onset of schooling oc- curs over the size range in which we observed the maximum predation rate (numbers consumed in 5 min) on individual northern anchovy larvae by north- ern anchovy predators. Thus predation pressure may be an important evolutionary factor in the tim- ing of the onset of schooling during the larval stage. The interaction between larval growth rate and size-specific vulnerability to predation may be an im- portant source of interannual variation in larval mortality (Shepherd and Cushing 1980; Smith 1985). A simple calculation illustrates this point using the size-specific vulnerability of northern anchovy lar- vae (10-20 mm SL) to adult northern anchovy predators. We assumed larval escape ability to be an inverse measure of predator vulnerability and normalized it to the average mortality rate over this size interval (Table 1). Thus in our calculation, the rate larval mortality decreased with increasing lar- val size was inversely proportional to the rate escape ability increased with size (larval escape ability in- creased linearly with larval length over the 10-20 mm length range). Our calculation indicated that a 50% increase in growth rate from the average rate of growth in the sea resulted in a 58% increase in survival in 30 d compared with average conditions. Decreasing the growth rate by 50% gave a 37% decrease in survival over the same interval. A longer period of reduced or enhanced growth rates will, of course, give a larger deviation from average survival values. Table 1 .—Calculation of the effect of growth rate on survival of 10-20 mm northern anchovy larvae when mortality is inversely pro- portional to length specific escape probabilities. Terms Parameter values z = mortality rate Z = 0.05 at 16 mma s = larval length (mm) 10 mm < S< 20 mm G = growth rate G = 0.325 mm/db T = time T = 30 d N = relative numbers of larvae A/(0) = 1 Initial equations Z = 0.15 - (0.00625 x S) a S = 10 + (G x 7") dNIdt = - (Z x N) Final equations after substitution and integration N = 0.0724 x exp (2.808 x G) Estimates of survival after 30 days Growing conditions Growth rate (mm/d) Relative numbers of larvae Relative survival (%) Average 50% increase 50% decrease 0.325 0.488 0.162 0.1803 0.2850 0.1141 + 58 -37 aMortality function generated from larval anchovy escape data with adult northern anchovy as predators. Values are normalized to Z 0.05 at 16 mm (Smith 1985). "From Smith (1985), 0.325 ± 50% also used in calculation. Effect of Predator Size We examined the existing literature on predators of larval northern anchovies to determine how the ability to escape a predator varied among different predator species. Regardless of the predator species, larval escape ability always increases with larval size, but the rates vary greatly with predator size. In general the smaller the predator, the faster lar- val escape abilities improve with increasing larval length (Fig. 6). The results of our work on E. mor- dax were similar to those of Brownell (1985) on E. capensis. However, capture success of the 85 mm E. mordax predators used in our study was about 20% higher than the 34 mm E. capensis predators used by Brownell. The extent of the predator field for a given size and species of predator can be defined as the larval size range in which larval escape success is <100%. For adult northern anchovy predators (85 mm and 867 FISHERY BULLETIN: VOL. 84, NO. 4 100 - juv. Euphausia (6- 10mm) Engraulis capensis (15mm) / UJ < > < -I u. o UJ O CC UJ a. < H I- < O z a < o (/> UJ Engraulis capensis (34mm) Engraulis mordax (85mm) Scomber japonicus (190mm) ^~~" JL 2 6 10 14 18 22 26 30 34 38 42 46 50 60 70 80 90 STANDARD LENGTH (mm) Figure 6.— Percentage of larval and juvenile anchovies escaping attacks of various predators as a function of length. Data for Engraulis capensis feeding on larval E. capensis are from Brownell (1984); juvenile Euphausia fed E. mordax from Theilacker and Lasker (1974); Amphiprion percula fed E. mordax from Webb (1981); and others are from this study. Numbers indicate length (mm) of the various predators. larger), the field extends from the egg (Hunter and Kimbrell 1980) to about 40 mm. The field for juvenile chub mackerel is much wider than that for northern anchovy extending from northern anchovy eggs to adults (120 mm), whereas the predation field for Euphausia is restricted to the yolk-sac period (Theilacker and Lasker 1974). The limited data available (Table 2) provide a crude index for the up- per limit of the predator field for northern anchovy larvae. When the larval length exceeds about 50% of the predator length little or no predation occurs. CONCLUSIONS Much of the past research on recruitment has focused on early larval stages where mortality rates are the highest (May 1974). Our work supports a growing contention that later larval stages and early juvenile stages may be as important in determining year-class strength (Smith 1985) and that such ef- fects might be mediated through an interaction between larval growth and size-specific vulnerability to predators. Our results and those of others indicate that the ability of northern anchovy larvae to escape pelagic predators increases throughout the larval stage. On the other hand, the susceptibility of lar- vae to predation may not decrease strictly accord- ing to size because large larvae may be more easily Table 2.— Upper limit of some predator fields for larval anchovies, Engraulis mordax and E. capensis. Upper limit of predator Predator field Predator length Larval length Larval length Species (mm) (mm) Predator length Euphausia juveniles 8 4.5 0.6 (Theilacker and Lasker 1974) Engraulis capensis 15 8.2 0.6 (Brownell 1984) Engraulis capensis 34 20 0.6 (Brownell 1984) Amphiprion percula 44 18 0.4 (Webb 1981) Engraulis mordax 85 2~40 0.5 (this study) Scomber japonicus 190 2~120 0.6 (this study) 'Upper limit = larval size at which all larvae escape predator. 2Extrapolated values. detected by visual feeding planktivorous fishes than smaller ones. ACKNOWLEDGMENTS We wish to thank Roderick Leong for providing the northern anchovy eggs on demand, Carol Kim- brell for editorial assistance, and Clelia Booman and 868 FOLKVORD and HUNTER: VULNERABILITY OF NORTHERN ANCHOVY LARVAE Paul Smith for reviewing the manuscript and offer- ing helpful suggestions. LITERATURE CITED Baxter, J. L. 1967. Summary of biological information on the northern an- chovy Engraulis mordax Girard. CalCOFI Rep. 11:110- 116. Brownell, C. L. 1985. Laboratory analysis of cannibalism by larvae of the Cape anchovy Engraulis capensis. Trans. Am. Fish. Soc. 114:512-518. Curio, E. 1976. The ethology of predation. Springer- Verlag, N.Y., 249 P- Ehrlich, K. F. 1974. Chemical changes during growth and starvation of herring larvae. In J. H. S. Blaxter (editor), The early life history of fish, p. 310-323. Springer- Verlag, N.Y. Frank, K. T., and W. C. Leggett. 1984. Selective exploitation of capelin (Mallotus villosus) eggs by winter flounder (Pseudopleuronectes americanus): capelin egg mortality rates, and contribution of egg energy to the annual growth of flounder. Can. J. Fish. Aquat. Sci. 41:1294-1302. FOLKVORD, A. 1985. Size specific vulnerability of northern anchovy Engraulis mordax larvae to predation by fishes. MS Thesis, University of California at San Diego, 96 p. Hewitt, R. P., G. H. Theilacker, and N. C. H. Lo. 1985. Causes of mortality in young jack mackerel. Mar. Ecol. Ser. 26:1-10. Hunter, J. R. 1976. Culture and growth of northern anchovy, Engraulis mordax, larvae. Fish. Bull., U.S. 74:81-88. 1984. Inferences regarding predation on the early life stages of cod and other fishes. F10devigen rapportser., 1. ISSN 0333-2594 The propagation of cod Gadus morhua L, p. 533-562. Hunter, J. R., and K. M. Coyne. 1982. The onset of schooling in northern anchovy larvae, Engraulis mordax. CalCOFI Rep. 23:246-251. Hunter, J. R., and C. A. Kimbrell. 1980. Egg cannibalism in the northern anchovy, Engraulis mordax. Fish. Bull., U.S. 78:811-816. Leong, R. 1971. Induced spawning of the northern anchovy, Engraulis mordax Girard. Fish. Bull., U.S. 69:357-360. Major, P. F. 1978. Predator-prey interactions in two schooling fishes, Caranx ignobilis and Stolephorus purpureus. Anim. Behav. 26:760-777. May, R. C. 1974. Larval mortality in marine fishes and the critical period concept. In J. H. S. Blaxter (editor), The early life history of fish, p. 3-19. Springer- Verlag, N.Y. Moller, H. 1984. The reduction of a larval herring population by jellyfish predator. Science 224:621-622. O'Connell, C. P. 1980. Percentage of starving northern anchovy, Engraulis mordax, larvae in the sea as estimated by histological methods. Fish. Bull., U.S. 78:475-484. 1981. Development of organ systems in the northern an- chovy, Engraulis mordax, and other teleosts. Am. Zool. 21:429-446. Pielou, E. C. 1981. The usefulness of ecological models: a stock-taking. Q. Rev. Biol. 56:17-31. Purcell, J. E. 1985. Predation on fish eggs and larvae by pelagic cnidarians and ctenophores. Bull. Mar. Sci. 37:739-755. Schaefer, K. M. 1980. Synopsis of data on the chub mackerel, Scomber japonicus, 1782, in the Pacific Ocean. In W. H. Bayliff (editor), Synopses of biological data on eight species of scom- brids. Inter-Am. Trop. Tuna Comm., Spec. Rep. 2, p. 395-445. Shepherd, J., and D. H. Cushing. 1980. A mechanism for density dependent survival of larval fish as the basis of stock recruitment relationship. J. Cons, int. Mer 39:160-167. Smith, P. E. 1985. Year-class strength and survival of 0-group clupeoids. Can. J. Fish. Aquat. Sci. 42 (Suppl. l):69-82. Theilacker, G. H. 1986. Starvation-induced mortality of young sea-caught jack mackerel, Trachurus symmetricus, determined with histo- logical and morphological methods. Fish. Bull., U.S. 84: 1-15. Theilacker, G. H., and R. Lasker. 1974. Laboratory studies of predation by euphausiid shrimps on fish larvae. In J. H. S. Blaxter (editor), The early life history of fish, p. 287-299. Springer Verlag, N.Y. Van Der Veer, H. W. 1985. Impact of coelenterate predation on larval plaice Pleuronectes platessa and flounder Platichthys flesus in the western Wadden Sea. Mar. Ecol. Prog. Ser. 25:229- 238. Vinyand, G. L., and W. J. O'Brian. 1976. Effects of light and turbidity on the reactive distance of bluegill (Lepomis macrochirus). J. Fish. Res. Board Can. 33:2845-2849. Webb, P. W. 1981. Responses of northern anchovy, Engraulis mordax, lar- vae to predation by a biting planktivore, Amphiprion per- cula. Fish. Bull., U.S. 79:727-735. 1982. Avoidance responses of fathead minnow to strikes by four teleost predators. J. Comp. Physiol. 147:371- 378. 869 OBSERVATIONS ON THE REPRODUCTIVE BIOLOGY OF THE COWNOSE RAY, RHINOPTERA BONASUS, IN CHESAPEAKE BAY12 Joseph W. Smith and John V. Merriner3 ABSTRACT Cownose rays, Rhinopterabonasus, are abundant in Chesapeake Bay during summer. We made obser- vations on the reproductive biology of specimens collected primarily from commercial pound nets and haul seines from May through October 1976-78. Clasper development suggested that males began to mature at disc widths (DW) of 75-85 cm. Males judged as mature averaged about 90 cm DW. Macroscopic inspection of the oviducts suggested that females began to mature at 85-92 cm DW. Females judged as mature averaged 96 cm DW. Only the left reproductive tract in female cownose rays appeared func- tional and only one embryo per gravid female was observed. A total of 67 embryos ranging 18-440 mm DW were collected and the sex ratio of the embryos was 1:1. Gravid females carried three-quarter term embryos in May and parturition occurred in late June and July. Full-term embryos averaged about 40 cm DW. Gestation of another group of embryos began by August. Growth of these embryos was rapid and they were relatively large when cownose rays left the Chesapeake Bay in October. Cownose rays exhibited aplacental viviparity. Yolk reserves supplied the initial energy demands of the embryos (up to about 20 cm DW), but histotrophic secretions of uterine villi provided nutrition for the young through the remainder of gestation. The cownose ray, Rhinoptera bonasus, a large myliobatoid ray, which attains a maximum weight of 23 kg, is abundant in Chesapeake Bay during summer (Schwartz 1965; Musick 1972) where it preys heavily on commercially important shellfish (Merriner and Smith 1979). Because of the severe damage to shellfish beds and the paucity of infor- mation on the biology of the cownose ray, the Virginia Institute of Marine Science began a study on the life history of the cownose ray in 1976. Prior to our work, information on the cownose ray's reproductive biology was primarily limited to obser- vations of single gravid females, usually included in more general literature (Gudger 1910; Bigelow and Schroeder 1953; Joseph 1961; Hoese 1962; Bearden 1965; Orth 1975), and size at maturity was unknown (Bigelow and Schroeder 1953). Schwartz's (1967) brief abstract represented the most complete state- ment on the species' reproductive cycle. Here, we report on the reproductive biology of the cownose ray, specifically on 1) the estimated size at matur- 1 Based on part of a thesis submitted by the senior author in par- tial fulfillment of the degree of Masters of Arts at the College of William and Mary, Williamsburg, VA 23185. Contribution No. 1305 from the Virginia Institute of Marine Science and the College of William and Mary School of Marine Science, Gloucester Point, VA 23062. 3The College of William and Mary, School of Marine Science, Gloucester Point, VA 23062; present address: Southeast Fisheries Center Beaufort Laboratory, National Marine Fisheries Service, NOAA, Beaufort, NC 28516-9722. ity for both sexes, 2) the definition of the gestation period, and 3) the description of the embryonic development. MATERIALS AND METHODS Most cownose rays were taken from pound nets in the lower Chesapeake Bay during three summers, 1976-78, but some rays came from haul seines used in spring along the Virginia-North Carolina coast, and from gill nets and rod and reel catches. Disc width (DW = distance between tips of the pectoral fins) was measured in mm on a measuring board. References to specimen size (including embryos) hereafter are disc width measurements. We judged male cownose rays sexually mature if 1) the clasper rhipidion was fully developed and easily spread and 2) clasper cartilages were well calcified (rigid). We measured clasper length as the distance from the junction of the clasper and pelvic fin to the distal end of the clasper. Criteria modi- fied from Smith (1975) were used to determine the following stages of sexual maturity for females: 1) immature - ovaries thin and flaccid; uterus thin and elongate, lining appears rugous. 2) maturing - ovary slightly developed, yellowish ova visible, ova <1 cm diameter; uterus somewhat dilated, trophonemata (uterine villi) small, general- ly <0.5 cm long. Manuscript accepted May 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 871 FISHERY BULLETIN: VOL. 84, NO. 4 3) mature - ovary with large yellowish ova >1 cm diameter; uterus well-developed and rich in tropho- nemata, generally >1 cm long. Uteri and oviducts were opened and inspected for ova or embryos. Embryos were weighed and mea- sured for disc width (mm). Yolk-sac volume (mL) was measured by volumetric displacement in a graduated cylinder. RESULTS AND DISCUSSION Like many other elasmobranch populations which occur along the east coast of the United States, cow- nose rays are highly migratory and exhibit a north- ward coastal migration in spring and a southward movement in fall (Schwartz 1965; Smith 1980). Our earliest spring collection of adult rays occurred dur- ing 2-5 May 1977 on the North Carolina Outer Banks. Our latest fall collection of adult males was on 7 September 1978 in the lower York River, while the latest fall collection of adult females occurred on 12 October 1977 near Cape Henry, VA, at the mouth of the Chesapeake Bay. Adult rays were absent from pound net catches in the lower bay after mid-October; furthermore, they were unavailable to us until the following spring when they migrated back into Chesapeake Bay. Size at Maturity At the onset of sexual maturity, terminal cartilage elements develop distally on the claspers of male elasmobranchs (Bigelow and Schroeder 1953), and the allometric growth of these appendages has been used to determine the attainment of sexual matur- ity in various elasmobranchs (e.g., Clark and Von Schmidt 1965; Struhsaker 1969; Gilbert and Heath 1972). In male cownose rays the ratio of clasper length to disc width increases slightly at 75-85 cm DW suggesting the onset of sexual maturity (Fig. 1). Males <75 cm (n = 68) appear immature; their testes are thin, white and ribbonlike and their claspers are narrow and flexible. Males ranging 80-98 cm (x = 89.8 cm; n = 115) appear mature; their testes are pinkish white in color and greatly swollen, and their claspers are rigid and well-cal- cified. Based on clasper length to disc width ratio and cursory observations of the testes, we estimated that male cownose rays begin sexual maturation at about 80 cm and most are probably mature at disc widths >84 cm. Considerable discrepancies exist in the literature concerning size of female cownose rays at sexual maturity. Gudger (1910) claimed a female about 60 cm wide gave birth to a pair of young. Bearden (1965) reported four premature young from a female measuring 712 mm (disc width?) taken in South Carolina. Joseph (1961) and Orth (1975) collected gravid females in Chesapeake Bay of 89 and 90 cm, respectively. We classified females <84 cm (n = 86) as immature (immature ovaries are thin and flac- cid, and immature uteri are thin and elongate). Females that we judged as mature ranged 84.5-100 cm (x = 96 cm; n = 117). Mature ovaries possess yellowish ova >1 cm in diameter; the left uterus of mature females is well-developed and rich in tropho- nemata (uterine villi), which are generally >1 cm long, red in color, and spatulate distally. We clas- sified eight specimens (range: 84-92 cm) as matur- ing females. Although ova <1 cm in diameter are visible in the ovary, the left uterus is not well- developed and the trophonemata are generally <0.5 cm long. The smallest gravid female measured 87 cm. Based on these observations we estimated that female cownose rays begin sexual maturation at 85-90 cm and are mature at disc widths >90 cm. Only the left reproductive tract appears functional in female cownose rays. There is no macroscopic evidence of follicular development in the right ovary. The right uterus in mature specimens shows some distension (ca. 3 cm wide), but does not exceed the breadth of the left uterus. Embryos and ova occur only in the left uterus, although we found an empty shell capsule in the right uterus of several gravid females. Nonfunctional right reproductive tracts have been reported in the roughtail stingray, Dasya- tis centroura, (Struhsaker 1969) and the bluntnose stingray, D. sayi (Gudger 1912; Hamilton and Smith 1941; Hess 1959). Reproductive Cycle Numerous literature accounts reported on the cap- ture of singular gravid cownose rays (Smith 1907; Gudger 1910; Bigelow and Schroeder 1953; Joesph 1961; Hoese 1962; Bearden 1965; Orth 1975) and these provided fragmentary information on the ray's gestation cycle. Schwartz's (1967) abstract defined June through October as the breeding cycle and closely parallels our results, although we disagreed on size at parturition. We collected 67 embryos (range: 18-440 mm; sex undetermined for 3 speci- mens) from the lower Chesapeake Bay and vicinity. Data for 19 embryos (all specimens sexed, length undetermined for 8 specimens) taken in April 1978 near Cape Lookout, NC, were provided to us by W. 872 SMITH and MERRINER: REPRODUCTIVE BIOLOGY OF COWNOSE RAY S. Otwell4 (Fig. 2). Only one embryo per gravid 4W. S. Otwell, formerly of North Carolina State University Food Science Laboratory, Morehead City, NC; presently at University of Florida, Food Science Department, Gainesville, FL 32611, pers. commun. April and May 1978. female was observed. The overall sex ratio of em- bryos (40o":439) did not differ significantly from 1:1. Gravid female rays migrate into Chesapeake Bay in spring with well-developed embryos that we designated as approximately three-quarter term. 44 -i Figure 1.— Relationship of clasper length (mm) to disc width (cm) for 188 male Rhi- noptera bonasus. 12 - 8 - 4 0 40 - x = 2 Values n = 188 38 - 32- 28 - X o z UJ _l 24 - 20 - or UJ CL w < 16- 30 • •• x •• • •• • • • xxm a • • • XXX* x« • •• • •• • •• mx* • X •• • X» XX • • • • • •• • x» XXX AM • ••• X«5t •• • •• ••«XX XX X • XX X «H«I. • i ' i ■ i ■ i ' r 40 50 60 70 80 DISC WIDTH (cm) 90 100 450 400 350 £300 >- cc m u. o x t- Q ? O CO a 250 200 150 100 50- N = 78 •1 iM I— Mean > Range • One Value + Two Volues SEPT ' OCT "jAN ' FEB ' MAR APR ' MAY ' JUNE ' JULY AUG NOV DEC Figure 2.— Relationship of disc width (mm) to date of capture for Rhinoptera bonasus embryos collected 1976 through 1978. 873 FISHERY BULLETIN: VOL. 84, NO. 4 Embryos collected in early May on the Outer Banks and in the lower York River average 259 mm (range: 221-276 mm; n = 7), and those collected from Cape Lookout, NC, in mid- April (Otwell fn. 4) average 264 mm (range: 222-281 mm; n = 11) (Fig. 2). By late June and early July the embryos are full term (x = 413 mm; n = 4). Parturition occurs at this time and the first free-swimming young appear in pound net catches. Embryo weight gain in spring is note- worthy; three-quarter term embryos in April and May average 310 g (range: 192-392 g; n = 16), while the weight of full-term individuals in late June in- creases fourfold averaging 1,291 g (range: 1,134- 1,409 g; n = 3). Schwartz (1967) reported that term individuals average 305 mm DW, however, embryos we considered full term are considerably larger (ca. 400 mm) and the smallest free-swimming ray we col- lected was 323 mm. Perhaps, the embryos Schwartz (1967) considered full term were taken in early June and were not yet ready for parturition. Female rays ovulate following parturition. We found encapsulated uterine eggs in specimens taken on 28 June and 21 July. In early August the embryos are 20-30 mm wide and have lost the shell capsule. By late August they average 125 mm (Figs. 2, 3). When adult rays leave the Chesapeake Bay in late September and early October, the embryos are relatively large, up to 220 mm. Reproductive cycles of large elasmobranchs are often difficult to describe because during certain stages of pregnancy, individuals may be inaccessible as a result of schooling and migratory behavior (Holden 1974). Since cownose rays leave Chesapeake Bay by November and do not return until May, we could not determine precisely the length of gesta- tion. Nevertheless, an 11-12 mo gestion period seems most probable. Within this context, the rapid embryonic growth observed in summer would slow during winter. A slowdown or cessation of intra- uterine growth would be expected if gravid females experience high energy demands during an exten- sive migration to distant wintering grounds, possibly northern South America as suggested by Schwartz (1965). Thus, the embryos from late summer and fall Figure 3.— Series of Rhinoptera bonasus embryos ranging from 18 to 140 mm disc width collected in late summer and fall. 874 SMITH and MERRINER: REPRODUCTIVE BIOLOGY OF COWNOSE RAY would be born the following summer when the adults return to Chesapeake Bay, a gestation period of 11-12 mo beginning in July or August and ending in June or July. The relatively large size of cownose ray embryos in late September and early October suggests the possibility of two 5-6 mo gestation periods. Partu- rition might occur on the cownose rays' wintering grounds followed by the gestation of another brood of embryos destined for birth the following summer. This hypothesis is not unprecedented, since the presence of well-developed young in the spiny butterfly ray, Gymnura altavela, during May in Delaware Bay and during February off the coast of North Carolina (27 fathoms) led Daiber and Booth (1960) to propose two 5-6 mo gestation periods per year for this species. Precise definition of the cow- nose ray gestation cycle will require collecting gravid female rays on their wintering grounds. Embryonic Development and Nutrition The shell capsule of the cownose ray, which we observed twice in utero, is of a greenish amber, thin diaphanous material, and is about 10 cm long. One capsule held a single ovum, while the capsule from a second female contained three ova. Ova are yellow, extremely flaccid, and about 3-4 cm in diameter. The embryos in late summer and fall possess yolk stalks and yolk sacs, although these often become detached during collection (Fig. 3). The smallest em- bryos we collected are about 20 mm wide, batoid in appearance, and unencapsulated. Numerous exter- nal branchial filaments (ca. 15-30 mm long), which emerge from the gill slits, are highly conspicuous on small embryos (18-75 mm). These filaments are absent in embryos larger than 89 mm. Three-quarter term embryos are upright in the uterus (ventral surface of the embryo on the ven- tral wall of the uterus) with the rostrum pointed for- ward. The pectoral fins are folded dorsally. The tail and heavily sheathed spine are bent forward along the dorsum of the disc. The yolk sac and stalk are almost completely absorbed; only about 3 mm of the umbilicus protrudes from the abdomen. Full-term embryos are similarly oriented. How- ever, the umbilicus is completely absorbed, leaving only a small scar that is evident on free-swimming young. Pigmentation is that of the adults, i.e., chocolate-brown dorsally, white ventrally, and black caudally. Several tooth plates were discovered in the left uterus from which a full-term young was re- moved, confirming Bigelow and Schroeders' (1953) report that tooth replacement begins in utero. During the early stages of gestation the uterus is rigid and thick-walled, but it gradually expands to accommodate the developing young. Just prior to parturition, it is extremely distended (ca. 15 cm at its greatest breadth), thin-walled, and flaccid. Myliobatoids overcome spatial restrictions in utero by rolling the pectoral fins dorsally or ventrally, along the anterioposterior axis (Gudger 1951), and some studies report that larger than average females carry more and larger offspring (e.g., Babel 1967). Although we observed multiple encapsulated ova in cownose rays, and others have cited the oc- currence of multiple embryos in utero (Smith 1907; Gudger 1910; Bearden 1965), we never found more than one embryo per gravid female. Setna and Sarangdhar (1949) and James (1962, 1970) made similar observations for the Javanese cownose ray, R. javanica, from the Indian Ocean. Our data for term embryos (n = 4) are insufficient to corrolate embryo size with parent's size; however, we suspect that in general only one cownose ray embryo is car- ried to term. Embryonic nutrition is from yolk and histotrophe. Yolk of the late summer and fall embryos (n = 33) gradually diminishes between August and October (Fig. 4), and most yolk reserves are probably util- ized when embryos are about 20 cm. Histotrophe, a viscid, yellowish secretion of the uterus (as also cited by Schwartz 1967), also nourishes the embryos. The amount of histotrophe, although not quantified, increases considerably as gestation progresses. Tro- phonemata, the uterine villi that produce histo- trophe, are deep red, flattened in cross section and spatulate distally. They attain their greatest length (ca. 2-3 cm) in females with near full-term embryos. The trophonemata occasionally invade the gill slits. In summarizing chondrichthyan, fetal-maternal relationships, Wourms (1977) noted that the effi- ciency of placental analogues, the villiform tropho- nemata, far surpasses that of the yolk-sac placenta exhibited by some carcharhinids. In cownose ray em- bryos, yolk apparently provides initial nutritional re- quirements. Embryos may augment yolk supplies during the first month of gestation by absorbing histotrophe via the external branchial filaments, as was suggested for Urolophus halleri by Babel (1967). After October, histotrophe supplies nourish- ment for the remainder of the gestation period, probably engulfed via the mouth, spiracles, and gill slits. Viviparity and the use of nursery areas that are relatively free of predators, e.g., Chesapeake Bay, no doubt protect young cownose rays. Large car- charhinids, of which batoids are purported to be a 875 FISHERY BULLETIN: VOL. 84, NO. 4 I6-1 14 - 1 12" 10- 8- Ld 2 Z3 _J o > < V) 5 4 o 2- » — I 1 1 1 1 1 1 1 1 1 1 1 0 20 40 60 80 100 120 140 160 180 200 220 240 EMBRYO DISC WIDTH (mm) Figure 4.— Relationship of yolk-sac volume (mL) to disc width (mm) for Rhinoptera bonasus embryos collected in late summer and fall. favorite prey (Darnell 1958; Budker 1971) are abun- dant seaward of the Virginia capes during summer (Lawler 1976), but generally only the sandbar shark, Carcharhinus plumbeus, and the bull shark, C. leucas, frequent the Chesapeake Bay proper (Schwartz 1960; Musick 1972). Although gravid female sandbar sharks utilize the eastern shore of the Chesapeake Bay (Lawler 1976), they may not pose a threat to cownose rays, since the female sand- bar sharks generally abstain from feeding while on their pupping grounds and males tend to avoid such areas (Springer 1960). Bull sharks (Schwartz 1959) may represent the only major predators of rays in Chesapeake Bay during summer. ACKNOWLEDGMENTS We thank J. A. Musick, G. R. Huntsman, A. B. Powell, W. R. Nicholson, J. Colvocoresses, and two anonymous reviewers for their comments and critical review of the various drafts of this manu- script. Numerous students and staff at Virginia In- stitute of Marine Science lent valuable assistance during various phases of this study, especially J. Gourley, R. Lambert, R. K. Dias, E. F. Lawler, R. J. Orth, and C. E. Richards. Captains Buddy Pon- ton, George Ross, Benny Belvin, and Herman Greene kindly provided "stingers" from their com- mercial catches. W. S. Otwell and F. J. Schwartz generously shared their life history notes on the cownose ray. Support for the study was provided by the Sea Grant program of the Virginia Institute of Marine Science (Grant Nos. 04-6-158-44-047 and 04-7-158- 44-109). Additional financial assistance was provided by the Gulf and South Atlantic Fisheries Develop- ment Foundation, Inc., Tampa, FL. LITERATURE CITED Babel, J. S. 1967. Reproduction, life history, and ecology of the round stingray, Urolvphus halleri Cooper. Calif. Dep. Fish Game, Fish Bull. 137:1-104. Bearden, C. M. 1965. Elasmobranch fishes of South Carolina. Contrib. Bears Bluff Lab. 42:1-19. BlGELOW, H. B., AND W. C. SCHROEDER. 1953. Fishes of the Western North Atlantic. Pt. 2. Sawfishes, guitarfishes, skates, rays, and chimaeroids. Mem. Sears Found. Mar. Res., Yale Univ. 1:1-588. Budker, P. 1971. The life of sharks. Columbia Univ. Press, N.Y., 222 p. Clark, E., and K. von Schmidt. 1965. Sharks of the central Gulf Coast of Florida. Bull. Mar. Sci. 15:13-83. Daiber, F. C, and R. A. Booth. 1960. Notes on the biology of the butterfly rays, Gymnura altavela and Gymnura micrura. Copeia 1960:137-139. Darnell, R. M. 1958. Food habits of fishes and larger invertebrates of Lake Pontchartrain, Louisiana, an estuarine community. Publ. Inst. Mar. Sci. Univ. Tex. 5:353-416. Gilbert, P. W., and G. W. Heath. 1972. The clasper-siphon sac mechanism in Squalus acanthias and Mustelus canis. Comp. Biochem. Physiol. A Comp. Physiol. 42:97-119. Gudger, E. W. 1910. Notes on some Beaufort fishes-1909. Am. Nat. 44: 395-403. 1912. Natural history notes on some Beaufort, N.C., fishes, 1910-11. No. I. Elasmobranchii— with special reference to utero-gestation. Proc. Biol. Soc. Wash. 25:141-156. 876 SMITH and MERRINER: REPRODUCTIVE BIOLOGY OF COWNOSE RAY 1951. How difficult parturition in certain viviparous sharks and rays is overcome. J. Elisha Mitchell Sci. Soc. 67:56-86. Hamilton, W. J., Jr., and R. A. Smith. 1941. Notes on the sting-ray, Dasyatis say (LeSueur). Copeia 1941:175. Hess, P. W. 1959. The biology of two sting rays, Dasyatis centroura Mitchill 1815 and Dasyatis say Lesueur 1871, in Delaware Bay. M.S. Thesis, Univ. Delaware, Newark, 42. p. Hoese, H. D. 1962. Sharks and rays of Virginia's seaside bays. Chesa- peake Sci. 3:166-172. Holden, M. J. 1974. Problems in the rational exploitation of elasmobranch populations and some suggested solutions. In F. R. Harden- Jones (editor), Sea fisheries research, p. 117-137. Wiley, N.Y. James, P. S. B. R. 1962. Observations on shoals of the Javanese cownose ray Rhinoptera javanica Muller and Henle from the Gulf of Man- nar with additional notes on the species. J. Mar. Biol. Assoc. India 4:217-223. 1970. Further observations on shoals of the Javanese cownose ray Rhinoptera javanica Muller and Henle from the Gulf of Mannar with a note on the teeth structure in the species. J. Mar. Biol. Assoc. India 12:151-157. Joseph, E. B. 1961. An albino cownose ray, Rhinoptera bonasus (Mitchill) from Chesapeake Bay. Copeia 1961:482-483. Lawler, E. F. 1976. The biology of the sandbar shark Carcharhinus plum- beus (Nardo, 1827) in the lower Chesapeake Bay and adja- cent waters. M.A. Thesis, College of William and Mary, Williamsburg, 48 p. Merriner, J. V., and J. W. Smith. 1979. A report to the oyster industry of Virginia on the biology and management of the cownose ray (Rhinoptera bonasus, Mitchill) in lower Chesapeake Bay. Spec. Rep. Appl. Mar. Sci. Ocean Eng. 216, 33 p. Va. Inst. Mar. Sci. Musick, J. A. 1972. Fishes of Chesapeake Bay and the adjacent coastal plain. In M. L. Wass (editor), A check list of the biota of lower Chesapeake Bay, p. 175-212. Va. Inst. Mar. Sci., Spec. Rep. No. 65. Orth, R. J. 1975. Destruction of eelgrass, Zostera marina, by the cow- nose ray, Rhinoptera bonasus, in the Chesapeake Bay. Chesapeake Sci. 16:205-208. Schwartz, F. J. 1959. Two eight-foot cub sharks, Carcharhinus leucas (Muller and Henle), captured in Chesapeake Bay, Maryland. Copeia 1959:251-252. 1960. Additional comments on adult bull sharks, Carcharhi- nus leucas (Muller and Henle), from Chesapeake Bay, Maryland. Chesapeake Sci. 1:68-71. 1965. Inter-American migrations and systematics of the western Atlantic cownose ray, Rhinoptera bonasus. [Abstr.] Meet. Assoc. Isl. Mar. Lab. Caribb., 1 p. 6th Meet., Isla Margarita, Venez. 20-22 Jan. 1967. Embryology and feeding behavior of the Atlantic cow- nose ray Rhinoptera bonasus. [Abstr.] Meet. Assoc. Isl. Mar. Labs. Carib. 1 p. 7th Meet., Barbados, W.I., 24-26 Aug.. Setna, S. B., and P. N. Sarangdhar. 1949. Breeding habits of Bombay elasmobranchs. Rec. In- dian Mus. (Calcutta) 47:107-124. Smith, H. M. 1907. The fishes of North Carolina. N.C. Geol. Econ. Surv. 2:47. Smith, J. W. 1980. The life history of the cownose ray, Rhinoptera bonasus (Mitchill 1815), in lower Chesapeake Bay, with notes on the management of the species. M.A. Thesis, College of William and Mary, Williamsburg, 151 p. Smith, M. S. 1975. A. P. Knight Groundfish Cruise No. 74-2, October 15-17, 1974 (Data Record). Fish. Res. Board Can. Manuscr. Rep. Ser. No. 1336, 13 p. Springer, S. 1960. Natural history of the sandbar shark, Eulamia milber- ti. U.S. Fish Wildl. Serv., Fish. Bull. 61:1-38. Struhsaker, P. 1969. Observations on the biology and distribution of the thorny stingray, Dasyatis centroura (Pisces:Dasyatidae). Bull. Mar. Sci. 19:456-481. Wourms, J. P. 1977. Reproduction and development in chondrichthyan fishes. Am. Zool. 17:379-410. 877 NORTHERN ANCHOVY, ENGRAULIS MORDAX, SPAWNING IN SAN FRANCISCO BAY, CALIFORNIA, 1978-79, RELATIVE TO HYDROGRAPHY AND ZOOPLANKTON PREY OF ADULTS AND LARVAE Michael F. McGowan1 ABSTRACT Eggs and larvae of Engraulis mordax were sampled by nets monthly for one year. Either eggs or larvae were caught every month. Both were most abundant when water temperature was high. Mean egg abun- dance did not differ among stations but larvae were more abundant within the San Francisco Bay at high and low salinity than near the ocean entrance to the Bay. Larvae longer than 15 mm were collected over the shoals in spring and autumn but were in the channel during winter. Zooplankton and microzooplankton were abundant relative to mean California Current densities. Adult spawning biomass in the Bay was 767 tons in July 1978, based on egg abundance and fecundity parameters of oceanic animals. San Francisco Bay was a good spawning area for northern anchovy because food for adults and larvae was abundant and because advective losses of larvae would have been lower in the Bay than in coastal waters at the same latitude. The northern anchovy, Engraulis mordax, is the most abundant fish in San Francisco Bay (Aplin 1967), but little is known about the seasonal dura- tion or areal extent of northern anchovy spawning there (Eldridge 1977; Sitts and Knight 1979; Wang 1981). In the California Current, spawning is thought to be related to abundance of food for adults (Brewer 1978) or to seasonal patterns of abundance of food for larvae (Lasker 1978). Dense patches of appropriate food for larvae are believed to be neces- sary for survival of larvae (Lasker 1975; Scura and Jerde 1977). Zooplankton are generally more abun- dant in estuaries than in coastal and oceanic waters. Therefore, San Francisco Bay, the largest estuary on the west coast of North America, could be a favorable habitat for spawning northern anchovy and their developing larvae. The northern anchovy could affect plankton dynamics in the San Francisco Bay (the Bay) by preying on zooplankton and by excreting concen- trated nutrients for phytoplankton. It is the target of a seasonal bait fishery (Smith and Kato 1979), and it is an important forage fish for many other species (Recksiek and Frey 1978). Quantitative estimates of the adult stock size and numbers of eggs and lar- vae are needed to understand the ecology of this anchovy in the Bay. This paper reports the results of a 1-yr survey of Cooperative Institute for Marine and Atmospheric Studies, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149. the northern anchovy eggs and larvae, zooplankton, and microzooplankton in San Francisco Bay. Dis- tribution and abundance of eggs and larvae were related to water temperature, salinity, turbidity, stratification, abundances of potential adult prey, and potential larvae prey. The suitability of the Bay for spawning and development of larvae was as- sessed. An estimate of spawning stock abundance within the Bay was calculated from egg abundance, and the impact of this biomass of anchovies on the zooplankton was estimated. MATERIALS AND METHODS Study Site San Francisco Bay consists of three major parts (Fig. 1): 1) Central Bay opens to the Pacific Ocean through the Golden Gate at lat. 37°49'N, long. 112°29'W; 2) North Bay receives the drainage from the Sacramento and San Joaquin Rivers and in- cludes Suisun, San Pablo, and Richardson Bays; 3) South Bay is the largest single embayment, ex- tending some 27 nmi from Coyote Creek in the south to the Oakland-San Francisco Bay Bridge in the north. The following description of San Francisco Bay was taken from Conomos and Peterson (1977). Mean depth is 6 m at mean lower low water, or 2 m if the large expanses of mudflats are included. There is a 10 m deep dredged ship channel in the northern part. Tides are mixed semidiurnal ranging from 1.7 m at the Golden Gate to 2.7 m at the south- Manuscript accepted July 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 879 FISHERY BULLETIN: VOL. 84, NO. 4 122°30" 122c i 38°- 37°30- SAN FRANCI GULF OF THE FARALLONES Miles 0 L 5 _L_ Kilometers 10 10 _L_ 1 r SAN FRANCISCO BAY Figure 1.— Locations of stations and the areas represented by each station sampled monthly May 1978-April 1979. ern end of South Bay. The tidal prism is 27% of the Bay volume. Maximum tidal currents occur in the channels and may be 225 cm/s (4.5 kn) at the Golden Gate. More than 90% of the freshwater entering San Francisco Bay enters North Bay from the Sacra- mento and San Joaquin Rivers. Less than 10% enters South Bay from small tributary streams and sewage. Because of the difference in freshwater in- flow the northern and southern reaches are very dif- ferent types of estuary. North Bay is partially-to- well-mixed with true two-layer estuarine circulation. South Bay, dependent for water exchange on tidal circulation and occasional incursions of freshwater from the north during wet winters, resembles a coastal lagoon. The heterogeneous nature of San Francisco Bay requires that stations be representative of the diverse areas of the Bay. The stations (Fig. 1) were located in the channel adjacent to the shoals in the South Bay in 5-6 m of water (stations 1 and 2); just north of San Bruno Shoal in 3 m of water (station 3); east of Treasure Island over a dredge borrow pit in 10 m of water (station 4); in midchannel just south of the Richmond-San Rafael Bridge in 10-13 m of water (station 5); and north of Lime Point just in- side the Golden Gate Bridge in 25-35 m of water (Station 6). These sites were near those of a previous trawl study (Aplin 1967) and they represented loca- tions from South Bay, Central Bay, the outflow from North Bay, and the Pacific Ocean entrance to the Bay. All South Bay stations were sampled in one day, all Central Bay stations were sampled on another day, usually the day following the South Bay sam- pling. This schedule and the pattern of tidal flow in the Bay (Tidal Current Chart, San Francisco Bay 1973) enabled all stations to be sampled before noon at approximately slack tide, low water. This sched- ule controlled for the effects of time of day, tide, and currents which can affect catches of ichthyo- plankton (Eldridge 1977). Additional samples were taken in October 1978 and April 1979 at station 3 and over the shoals adjacent to this station. Duplicate oblique ichthyoplankton tows and dupli- cate surface microzooplankton tows were made monthly at six stations for one year, May 1978-April 880 McGOWAN: SPAWNING OF NORTHERN ANCHOVY 1979. Ichthyoplankton and zooplankton samples were collected from a 5 m Boston Whaler with a 1 m diameter, cylinder-cone net of 0.308 mm mesh nylon with a 0.200 mm mesh cod end. The net was attached to a sled which kept the lower rim of the net 10 cm above the bottom and which had a tow- bridle that did not obstruct the mouth of the net. A frame attached to the transom permitted the sled to be launched and retrieved over the stern while underway. The sled was lowered to the bottom while underway at 1-2 kn, towed at the bottom for 1 min, and then retrieved at a constant rate and constant wire angle. Tow time, excluding that spent lower- ing the net to the bottom, was approximately 6 min. The gear was effective because it often caught an- chovies and herring longer than 30 mm, a size not usually captured in towed gear (Clarke 1983) or in a plankton purse seine (Murphy and Clutter 1972). A calibrated flowmeter suspended off-center in the mouth of the net measured the amount of water filtered during the tow. Volumes calculated from the flowmeter readings were similar to a hypothetical volume calculated from net mouth area and tow distance: approximately 300 m3/tow. Microzooplankton was collected with a 0.5 m diam- eter net with 0.080 mm mesh, which was towed just submerged at the surface for 2 min during the ich- thyoplankton tow. Because the flowmeter in this net frequently malfunctioned, hypothetical volumes calculated from mouth diameter and tow length (ap- proximately 25 m3/tow) were used to standardize catches of microzooplankton. The net probably did not filter as much water as calculated so microzooplankton were underestimated. All samples were preserved with 2% formaldehyde in seawater buffered with sodium borate. Water turbidity was measured with a Secchi disk (Tyler 1968). Water samples for salinity and tem- perature measurements were taken with a Van Dorn water sampler from 1 m below the surface and from 1 m above the bottom. The temperature was measured to 0.1°C with a laboratory thermometer, and salinity was measured to 0.5°/oo with a temper- ature-compensated refractometer. Laboratory Procedures Northern anchovy eggs were easily recognized and distinguished from other regional pelagic fish eggs by their oval shape and their size, approxi- mately 0.75 mm x 1.25 mm. Eggs were not as- signed to stages, but some of the embryos were developed enough to be identified as those of north- ern anchovies. Northern anchovy eggs were counted under a dissecting microscope; at the same time, fish larvae were picked from the samples. The northern anchovy can be separated from other similar look- ing larvae by its myomere count (43-47), its gut length, and its median fin positions (Miller and Lea 1972; McGowan and Berry 1984). All northern anchovy larvae <10 mm long were measured to the nearest 0.1 mm using an ocular micrometer. Longer larvae were measured to 1 mm using vernier calipers or a plastic ruler graduated in millimeters. The distance from the tip of the snout to the tip of the notochord was measured in pre- flexion larvae, standard length in larger specimens. Zooplankton were subsampled from a 500 mL pharmaceutical beaker by stirring and taking an ali- quot with a 1 mL or 2 mL Stempel pipet. Zooplank- ton were identified to major taxonomic group under the dissecting microscope using standard references such as Smith (1977). All holoplanktonic, meroplank- tonic, and nektonic invertebrates were considered to be zooplankton if they were suitably sized prey for adult anchovies. Isopods were included; adult shrimp and gelatinous invertebrates were not. Plankton was allowed to settle in water in a grad- uated cylinder to estimate zooplankton volume. Microzooplankton were subsampled from a stirred beaker with a pipet. A settling chamber and inverted compound microscope with movable stage were used to count microzooplankton (0.050-0.200 mm diameter) at 100 x magnification. Dinoflagellates known to be eaten by anchovy larvae were counted as microzooplankton. Precision Estimates The precision of the microzooplankton counts was estimated by the method of Lund et al. (1958). If the counts are treated as a Poisson variable then the 95% confidence limits for a single count are Upper limit = X + 2.42 + 1.96(Z + 1.5)1/2 Lower limit = X + 1.42 - 1.96(X + 0.5)1/2. The limits are approximately ±20% if 100 organ- isms are counted. Confidence intervals for micro- zooplankton counts in this study range from + 50% at the lowest count (5) to + 9% at the highest count (659). The precision of the zooplankton subsampling esti- mates was evaluated by taking triplicate subsam- ples, with replacement, from 10 randomly selected samples. The mean coefficient of variation (standard deviation divided by the mean) of the triplicates was 0.29. 881 FISHERY BULLETIN: VOL. 84, NO. 4 The precision of the duplicate tows was evaluated by comparing numbers of eggs, larvae, and zoo- plankton settled volumes from the May, June, and July tows. No statistical difference was detected between first and second tows (2-tailed P = 0.407, Wilcoxon Matched Pairs test, Hull and Nie 1981: 228). The mean coefficient of variation for these paired tows was 0.22. Because there were no statis- tical differences between these duplicates, only one of each pair of the remaining samples was sorted. Data Analysis Eggs, larvae, zooplankton, microzooplankton, and plankton volume per 1,000 m3 were calculated based on flowmeter readings. Temperature and salinity stratification variables were created by taking the difference between surface and bottom values. Salinity stratification represented the inten- sity of estuarine circulation or freshwater runoff; temperature stratification represented water column stability and revealed atmospheric tempera- ture extremes. Distributions of the variables were examined for skewness, kurtosis, and unreasonable range limits indicative of keypunch errors. Normality of the original variables and of \og(X + 1) transformations was tested (Kolmogorov-Smirnov test; Hull and Nie 1981:224). Variances of the transformed variables were not heteroscedastic. Biological and environ- mental variables were plotted against month, sta- tion, and each other to look for spatial patterns, seasonal trends, and nonliner relationships (espe- cially nonmonotonicity) between pairs of variables. Analysis of variance (ANOVA) was used to assess the effects of month of the year and station loca- tion on numbers of eggs and numbers of larvae. Stepwise multiple linear regression was used to ex- amine which of the other variables could account statistically for the variability in numbers of eggs and larvae. Logarithmic transformations of stan- dardized numbers of eggs, larvae, zooplankton, and microzooplankton were used in the regressions and in the ANOVA' s. Ichthyoplankton abundance is often expressed as numbers of ichthyoplankton under an area of sea surface by multiplying density per cubic meter times water depth (Smith and Richardson 1977). In deep water tows are made below the depth range of most eggs and larvae, so the tow depth is used as the ef- fective water depth. Standardizing a unit of sea sur- face area allows comparisons of total numbers of eggs and larvae in the water column from areas with different water depths. Abundance standardized to area of sea surface was used to estimate total egg production. However, larvae that were relatively uncommon in deep water could be as abundant as more concentrated larvae in shallow water, but ex- posed to different concentrations of predators and prey; therefore, densities of larvae and plankton were used to examine relationships between ich- thyoplankton, other plankton, and environmental variables. The method used to estimate spawning stock bio- mass was a direct estimate because it incorporated batch fecundity from histological data (Hunter and Goldberg 1980) and daily egg production estimates from ichthyoplankton surveys (Parker 1980). Parker's equation for the direct estimate of biomass from egg abundance is 5 = P{ab'c)-ld where 5 P a = spawning biomass in tons = egg production in eggs/day = 3.96 x 108 egg/ton b' = 0.159 the observed daily spawning fraction c = 0.550 the proportional biomass of females d = 1.080 a correction for potential mis- classification of daily spawning frac- tion. Parker (1980) estimated the coefficient of varia- tion of the estimate of spawning stock to be 0.614. Most of this statistical error was due to error in the estimate of egg production. Daily egg production was estimated in my study by dividing the egg abun- dance by the number of days needed to hatch at the ambient temperature (interpolated from Zweifel and Lasker 1976, fig. 7). Numbers per square meter of Bay surface were calculated by multiplying density per cubic meter times water depth at the station. The areas repre- sented by the stations were estimated from the chart of the Bay in Conomos and Peterson (1977). Total numbers of eggs and larvae were calculated from estimates per square meter times the area repre- sented by the sample. RESULTS Eggs and Larvae Either eggs, larvae, or both were present every month of the year. Eggs were present every month except December and January. Only one egg was 882 McGOWAN: SPAWNING OF NORTHERN ANCHOVY collected in February and very few were collected in November. Larvae were present every month and at every station each month with four exceptions: during June, no larvae were collected at station 1, the southernmost station; during July and August no larvae were collected at station 6, the Golden Gate Bridge station; during March no larvae were collected at station 3 in South Bay. Eggs were pres- ent on each of the occasions when larvae were absent from the samples. Egg density varied from 0 to 55,000 per 1,000 m3 (mean = 3,000). The greatest number of eggs in a single sample was 14,640 at station 2 in July. Occur- rence of eggs was seasonal: they were abundant in summer and absent in winter (Fig. 2). Larvae varied from 0 to 4,400 per 1,000 m3 (mean = 259). The greatest number of larvae in a single sample was 1,420 in September at station 2. Larval abundance was also seasonal with peak den- sity in late summer and fall (Fig. 2). Two-way ANOVA of log-transformed standard- ized densities of eggs and larvae were performed with month and station as fixed factors in separate analyses. The interaction mean square (not signifi- cant) was used as the denominator in the F-tests because there was just one observation per cell of the design (Montgomery 1976:156). Densities of eggs differed significantly among months (P < 0.001, Table 1) but not among stations (P = 0.104). Densities of larvae were significantly different among months (P = 0.010) and among stations (P = 0.014) (Table 2). Seasonal patterns of abundance of eggs and lar- Table 1 . — Analysis of variance of northern anchovy eggs: month by station. Source of Sum of Mean Signif- variation squares df square F icance Residual 42.23 54 0.78 Constant 286.22 1 286.22 366.02 0.000 Month 128.93 11 11.72 14.99 0.000 Station 7.56 5 1.51 1.93 0.104 Month x station 0.48 1 0.48 0.61 0.437 Table 2.— Analysis of variance of northern anchovy larvae: month by station. Source of Sum of Mean Signif- variation squares df square F icance Residual 25.77 54 0.48 Constant 220.65 1 220.65 462.43 0.000 Month 13.72 11 1.25 2.61 0.010 Station 7.59 5 1.52 3.18 0.014 Month x station 1.16 1 1.16 2.43 0.125 vae were unmistakeable, but differences among sta- tions were not as clear so three hypotheses were tested: 1) stations 1, 2, and 3, South Bay stations, differed from stations 4, 5, and 6; 2) stations 4 and 6, Golden Gate and Central Bay stations, differed from stations 1, 2, 3, and 5, South Bay stations plus the station at the outflow of San Pablo Bay; 3) sta- tions 3, 4, and 6, the stations most influenced by ocean water, differed from stations 1, 2, and 5, the Bay stations. These hypotheses were tested using linear contrasts (Nie et al. 1975:425), a procedure that compared the geometric means of the groups of stations. None of the three contrasts was significant for eggs but all three were significant (P < 0.05) for lar- vae. The difference between the mean of stations 4 and 6 and the mean of stations 1, 2, 3, and 5 was highly significant (P = 0.001). Further comparisons of mean densities of larvae were done using Duncan's Multiple Range test. This a posteriori procedure identified groups of means which did not differ significantly from each other at a specified level (Nie et al. 1975:427). The rank order of the stations in increasing mean density of larvae was 4,6, 1,3,5,2. Three groupings were pro- duced by the Duncan procedure at the 0.05 level. The mean of stations 4 and 6 was smaller than the mean of the other four. The mean of stations 5 and 2 was greater than that of the other four. Station 4 was significantly lower and station 2 significant- ly higher than the mean of the other four stations. A summary of the analyses of variance follows. Eggs and larvae were seasonal in abundance, eggs more strongly than larvae. Numbers of eggs, which would be subject to passive drift and dispersal, were not significantly different among locations in the Bay. Larvae did differ in abundance among the six stations. Based on a priori and a posteriori tests, station 4 and station 6, the stations most influenced by oceanic water, had low densities of larvae while the other stations within the Bay had high mean den- sities of larvae. This pattern was true for station 5, near the Richmond-San Rafael Bridge, as well as for stations 1, 2, and 3 in the South Bay. Among the within-bay stations, station 1, the southernmost, ranked lowest in both egg density and larval den- sity although it was not statistically different from the other inner stations— 2, 3, and 5. The stations also differed in the proportion of eggs to larvae. While the ratio of eggs to larvae was generally greater than 10:1, at station 3 the ratio of the mean number of eggs to mean number of lar- vae was <10:1 (Fig. 3). The proportions were sta- tistically different among stations (Chi-square P < 883 FISHERY BULLETIN: VOL. 84, NO. 4 >- z •- ° •< Z h- is CO u. \B h 2 CO 1 1 1 CO CVJ SniS"130 S33d93Q CO z til Q Z o CL o o N < LU (ldd) NOLLVOIdLLVUiS A1INI1VS O l- z < _i O o N o cc o 2 < s - u_ -a Q Z o CO - < -2 en CO o> o • ■ a • • ♦ CVJ CJ *" *- o o (W) Hld3Q IH003S — r- O — I- o o >- CO z LU Q < 1X1 o n O O "O < o> u. r>* o> -3 *- - Q Z o 00 < O) r- -T" T~ T" -T- T T CO o CO CM o CVJ in o m (ew/N) N0±>lNV~ld00Z (l/N) (ldd) A1INI1VS 30VddnS NOlMNVldOOZOyQIlN >- CO z 111 Q (3 (D UJ < UJ •f -r o CO z UJ Q _l ? < < UJ O O -f- r< o o m o in (euu OOCH/N) S993 (EUJ OOCH/N) 3VAUV1 I — T in o evi cvj ■»- ■<- (0J 'diVvJ31 30VddnS 2 Q Z o oo < o> a. 3 3 .H > t- •g 0) c a £-s -o s c c rt B a j«s o o 4s § m c 'S "a "3 o w o . p § 3 oi £ £ .2 S3 03 o> CO bD 3 bC cS o> — >> a § S •§-8 P 'a ^ "3 r* TO 2 «J C n3 CD CO O C c8 -a a 3 ■a _>> 2 c o § I D 884 McGOWAN: SPAWNING OF NORTHERN ANCHOVY EGGS 10000- 9000 8000 7000 6000-- 5000 4000 3000 2000 Figure 3.— Relative abundances of northern -|000 anchovy eggs and larvae at each station show- ing the difference between station 3 and the q other stations. LARVAE 1000 900 MEAN CATCH EGGS AND LARVAE AT EACH STATION NUMBERS ARE PER 1000 METERS CUBED ♦ LARVAE -I 1 1 i 1 2 3 4 5 6 STATION 0.01 with 5 degrees of freedom). Station 3 deviated most from the expected ratio. Station 1 also differed by having relatively fewer larvae than expected. Zooplankton Zooplankton catch varied from 13.6-9,560 indivi- duals/m3. Mean catch was 1,170/m3. No seasonal pattern was apparent (Fig. 2). There was a gradual increase in zooplankton abundance over the course of the study. This linear trend was significant (P < 0.01). Copepods, especially A cartia spp., dominated Table 3.— Zooplar ikton: relative density, May 1978- April 1979 mean Taxon ±1 SE nm 3 % Copepoda Acartia 1,120 + 192 96.05 harpacticoida 4.67 + 1.55 0.40 other 3.48 ± 0.75 0.30 shrimp zoeae 3.82 + 1.28 0.33 crab zoeae 12.27 ± 4.27 1.04 mysids 1.31 + 1.16 0.10 amphipods 0.39 + 0.16 0.03 pelecypods 1.10 + 0.35 0.09 chaetognaths 0.59 + 0.23 0.05 polychaetes 1.26 + 0.50 0.11 isopods 0.23 + 0.12 0.02 barnacle nauplii 9.18 + 2.04 0.78 barnacle cyprids 6.18 + 1.49 0.52 gastropods 0.74 + 0.37 0.06 cumaceans 0.08 + 0.05 0.01 cladocerans 0.81 + 0.31 0.07 the catches (Table 3). Brachyuran (crab) zooeae and cirrepedian (barnacle) nauplii and cyprids were occa- sionally abundant. Potential predators on northern anchovy larvae, such as chaetognaths and pontellid copepods, were often present but in relatively low numbers. Counts of zooplankton for each sample are reported in McGowan (unpublished M. A. Thesis, San Francisco State University, San Francisco, CA). Zooplankton catch was significantly correlated with all variables except surface salinity and salin- ity stratification. Negative correlations were ob- served with egg density, surface temperature, temperature stratification, and Secchi depth. Posi- tive correlations were found with larvae and microzooplankton . Microzooplankton Microzooplankton catch at the surface (0.080 mm mesh net) varied from 1 to 300 per liter (mean = 28.8). No clear seasonal trend was apparent (Fig. 2). Copepod nauplii were the most abundant micro- zooplankton followed by tintinnids and rotifers (Table 4). Dinoflagellates such as Ceratium and Peridinium were occasionally more abundant than copepod nauplii. The spiny, armored Ceratium species were not included in the density estimates because northern anchovy larvae prefer unarmored forms (Scura and Jerde 1977). Microzooplankton density was negatively correlated with Secchi disk depth (r = -0.34, P = 0.004) and positively corre- 885 FISHERY BULLETIN: VOL. 84, NO. 4 Table 4.— Microzooplankton: relative density, May 1978-April 1979. Taxon mean + 1 SE n-1 copepod nauplii 15.14 ± 1.82 54.97 barnacle nauplii 0.56 + 0.09 2.03 polychaete larvae 0.36 ± 0.08 1.31 tintinnids 6.56 + 2.68 23.82 rotifers 1.24 + 0.45 4.50 harpacticoid copepods 0.03 + 0.02 0.11 ostracods 0.01 + 0.01 0.04 gastropod veligers 0.04 + 0.02 0.15 Peridinium 3.59 ± 1.45 13.03 lated with zooplankton density (r = 0.27, P = 0.027). All interpretation of the microzooplankton data was done under the assumption that estimates of volume filtered are accurate. Environmental Variables The mean surface water temperature during this study was 15.2°C. The coldest reading was 8.0°C at station 2 in January; the warmest was 22.5 °C at station 1 in August (Fig. 2). Water temperature near the bottom varied from 8° to 21.5°C (mean = 15.0°C). Mean temperature stratification, the dif- ference between the surface and bottom tempera- tures, was 0.2°C. Stratification was generally pres- ent June through October, especially at station 5. Mean stratification during these months was 0.5°C (Fig. 2). During February and March 1979 the sur- face temperature was lower, on average, than the temperature near the bottom thus showing the influ- ence of air temperature on the surface water tem- perature. Surface salinity varied from 3 to 31°/oo (mean = 23.6%o). Bottom salinity was 14-31%o (mean = 24.8%o). The low readings for both sur- face and bottom salinity occurred at station 5 dur- ing March 1979. Surface salinity at station 1 was usually low, showing the influence of freshwater in- flow at the south end of the Bay (Fig. 2). Salinity at station 6 was relatively high, showing the oceanic influence at the Golden Gate. Surface salinity at other stations reflected their relative positions be- tween these two influences. The lowest surface salinity was always at station 5 due to the Sacra- mento River discharge. During March 1979, salin- ity at stations 4 and 6 also showed the effects of high freshwater discharge which lowered the salinity at station 5 to 3%o. Salinity was slightly lowered this month at station 3 in South Bay also. Surface salin- ity followed a seasonal pattern; it was high from July through January and low in the winter and spring months. Relatively high salinity corresponded to high temperature July through October. Salinity stratification was generally <2%o except at station 5 where the average stratification was 4.7°/oo (Fig. 2). Surface salinity was negatively correlated with salinity stratification, (r = -0.62, P < 0.001), and positively correlated with Secchi depth (r = 0.39, P = 0.001). Salinity stratification was negatively correlated with Secchi depth (r = - 0.29, P = 0.012). Turbidity Light penetration was lowest at stations 1 and 5, and highest at stations 6, 4, and 3 (Fig. 2). The mean depth of light penetration during this study was 1.1 m with a range of 0.1-2.5 m. The data suggest a weak seasonal trend with light transmission higher in summer and lower in winter. The variable with the strongest linear association with Secchi depth was zooplankton density. Light penetration was in- versely related to zooplankton density (r = -0.58). Relationships Among Varibles Northern anchovy egg abundances were positively associated with surface temperature, temperature stratification, and Secchi disk depth and negative- ly correlated with zooplankton density (Table 5). Eggs were positively associated with larvae but this correlation was not significant at the 5% level (P = 0.053). Larvae were positively correlated with surface temperature and zooplankton density (Table 5). They were negatively correlated with Secchi depth. Thus, eggs and larvae both were significantly correlated with zooplankton and Secchi depth but in opposite directions: eggs were associated with clearer water and lower zooplankton density, lar- vae with more turbid water and higher zooplankton density. Stepwise Multiple Regression Surface temperature alone explained 65% of the variability in egg density (r2 = 0.651). The combi- nation of microzooplankton density with surface temperature explains an additional 1.5% of the vari- ability of egg density. The addition of all other variables only increased the amount of variability explained to 68% (r2 = 0.678). The predictive regression model using the independent variables whose addition to the model improved its prediction by more than 1% is E = -2.20 + 0.317T - 0.502M 886 McGOWAN: SPAWNING OF NORTHERN ANCHOVY Table 5.— Bivariate correlations between northern anchovy eggs, larvae, and other variables. EGGS: log (eggs-m-3); LARV: log (larvae -m"3); ZOOP: log (zooplank- tersm-3); MICR: microzooplankton; TEMP: surface water temperature; SALI: sur- face water salinity; TSTR: temperature stratification; SSTR: salinity stratification; SECC: Secchi disk depth. EGGS LARV ZOOP MICR TEMP SALI TSTR SSTR LARV 0.23 + ZOOP -0.33** 0.29* MICR -0.11 -0.02 0.27* TEMP 0.81** 0.31** -0.46** 0.02 SALI 0.18 0.08 -0.20 -0.16 0.19 TSTR 0.40** 0.17 -0.25* -0.02 0.46** 0.06 SSTR -0.10 0.07 0.05 0.09 -0.03 -0.62** 0.12 SECC 0.35** -0.34** -0.58** -0.34** 0.32** 0.39** 0.20 -0.29* * Significant at P = 0.05. "Significant at P = 0.01. * P = 0.053. where E = log (eggs/1,000 m3 + 1) T = surface temperature (°C) M = log (microzooplankton/^ + 1), (Table 6). No single variable explained the majority of the variability in larval density (Table 7). Secchi depth was the single best predictor, accounting for 11% of the variance of larval density (r2 = 0.113). The combination of surface temperature with Secchi depth increased the coefficient of determination to 0.306. All of the variables combined explained just 50% of the variability of larval density (r2 = 0.498). Five variables improved the prediction of the set of independent variables by more than 1% when added to the model. The predictive equation for lar- val density based on using these five is L = -0.842 - 0.591X + 0.1267/ + 0.515Z Table 6. — Stepwise multiple regression: northern anchovy egg density vs. biological and environmental variables. - 0.57LW + 0.029S where L X T Z M S + 1) log (larvae/1,000 m3 Secchi depth (m) surface temperature (°C) log (zooplankton/1,000 m3 log (microzooplankton/^ + surface salinity (%>o). + 1) 1) The results of the multiple regressions show that northern anchovy egg density could be predicted largely by surface water temperature. Larval den- sity could not be predicted well by a single variable or by the five variables which, when combined, ac- counted for only 49% of the variability. Spawning Stock Estimates Based on estimates of egg production, the spawn- Independent Multiple Change in variable r2 Surface temperature 0.651 0.651 Microzooplankton 0.666 0.015 Salinity stratification 0.670 0.004 Surface salinity 0.672 0.002 Secchi depth 0.675 0.003 Zooplankton 0.677 0.002 Temperature stratification 0.678 0.001 Table 7.— Stepwise multiple regression: northern anchovy larval density vs. biological and environmental variables. Independent Multiple Change in variable r2 Secchi depth 0.113 0.113 Surface temperature 0.306 0.194 Zooplankton 0.392 0.085 Microzooplankton 0.459 0.067 Surface salinity 0.486 0.028 Salinity stratification 0.495 0.009 Temperature stratification 0.498 0.003 ing stock biomass of northern anchovies in the part of San Francisco Bay sampled in this study ranged from undetectable in December 1978 and January 1979 (no eggs collected) to 696 t (metric tons) (767 short tons) in July 1978. If the area of the Bay which is <2 m deep were included, the estimate of July biomass would have been 2,030 1 (2,240 short tons). Length Frequencies of Larvae Monthly samples could contain larvae from the current month and 2 previous ones because meta- morphosis is not complete until 35 mm, age 74 days at 16°C (Hunter 1976). However, larvae longer than 15 mm were not taken at the standard stations from August through October, although eggs and smaller 887 FISHERY BULLETIN: VOL. 84, NO. 4 larvae had been abundant since June (Fig. 4). Lar- vae >15 mm long were found over the shoals near station 3 in October and April (Fig. 5). Larvae longer than 15 mm were taken in the channel from Novem- ber through February, months with little or no spawning. Large larvae and juveniles, which had ap- parently overwintered, were present when spawn- ing resumed in March and April. DISCUSSION Previous suggestions that northern anchovy spawn in San Francisco Bay were based on the pres- ence of small larvae (Eldridge 1977; Sitts and Knight 1979), juveniles (Smith and Kato 1979), or the spawning season in the California Current (Hubbs 1925). Anchovy eggs collected in this study provide conclusive evidence that the northern anchovy spawns in San Francisco Bay because eggs could not drift upstream to station 5 or into South Bay as far as station 1 or 2. Peak spawning based on the abun- dance of eggs was May through September when adult anchovies are known to be plentiful in the Bay (Aplin 1967). Spawning in San Francisco Bay differed from an- chovy spawning in the sea. Most spawning of the central subpopulation of northern anchovy in the California Current takes place January- April when the 10 m temperature isl4°-16°C; not June through October when water temperature is 16°-19°C MAY NOVEMBER N=237 N=260 - i 1 I 1 1 i i I I I I I JUNE DECEMBER N=56 — 1 r— N= =32 I I 1 r II ill I 111 JULY JANUARY N=162 N=156 — ~~1— . l l I l I 1 i i — | 1 1 _ AUGUST FEBRUARY - N=483 N=222 . 1 1 i i i 1 l SEPTEMBER i 1 1 1 1 1 1 MARCH N=2313 N= 1 1 =62 100-n ii i i in 1 1 — | 1 O 75_ OCTOBER APRIL < 50- O N=156 N=161 gS 25- — U~1 c ) 5 10 15 20 25 30 35 ST ANC ARC • LE NGT H C LASS (mn l) Figure 4.— Length-class frequencies of larvae and juvenile northern anchovies for each month of the study. 888 McGOWAN: SPAWNING OF NORTHERN ANCHOVY OCTOBER 1978 APRIL 1979 CHANNEL N=97 10 15 20 25 30 STANDARD LENGTH CLASS (mm) 35 Figure 5.— Length-class frequencies of larvae and juvenile northern anchovies for October 1978 and April 1979 showing the different sizes caught in the channel versus those in shallow water. (Smith and Lasker 1978). The northern subpopula- tion spawns off Oregon and Washington from mid- June to mid-August when 1 m temperatures are 14°-17°C (Richardson 1980). These two subpopula- tions overlap at San Francisco (Vrooman et al. 1981) and the spawning season in the Bay overlapped the spawning seasons of both subpopulations. But spawning in the Bay took place at higher tempera- tures than usual for either population in the ocean (13°-18°C, Brewer 1976). Few eggs were taken in the Bay from December 1978 to March 1979 when water temperature was below 13°C. However, at station 3 in March 1979, 477 eggs were taken at a water temperature of 11.5°C. Peak spawning in the Bay was in July, August, and September when the mean water temperature was 19.0°, 19.8°, and 19.2°C, respectively. The highest catch of eggs oc- curred at station 2 in July at 21.0°C. Eggs were also plentiful at station 1 in August at 22.5° C. During June, July, and August, eggs were least abundant at stations 4 and 6, where water temperature was relatively low. During September and October, egg densities at stations 4 and 6 peaked, as did water temperature at these stations. Sitts and Knight (1979) found larvae shorter than 4 mm at 18°-22°C in the Sacramento-San Joaquin estuary in July and August. Although much of the northern anchovy spawning took place in the Bay within the previously reported temperature range and some took place at low temperatures, most occurred in water warmer than in the coastal spawning regions. The strong correlation of egg abundance with temperature in- cludes potential confounding effects of presumed seasonal influx of adults, apparent "preference" for spawning within the Bay, and differences in dilu- tion due to tidal exchange which affected stations 4 and 6 more than the other stations. Therefore the correlations are descriptive, perhaps predictive, but not causal. In the California Current, temperature, upwell- ing, and stable stratification of the water column are thought to interact to produce favorable condi- tions for anchovy larvae (Lasker 1975). In San Fran- cisco Bay there is no upwelling, but salinity or fresh- water outflow variability might influence ecological conditions. Freshwater flow may have an indirect effect by promoting blooms of certain phytoplankton or by retaining particles through estuarine circula- tion (Cloern 1979). Relatively high salinity coincided with warm temperatures at the beginning of the spawning season, but spawning ceased in Novem- ber when water temperature decreased to 13 °C, although salinity remained high until February. Sitts and Knight (1979) found larvae shorter than 10 mm at low salinity (<10%o) and relatively high temper- ature (>18°C). They found only large larvae (>10 mm) in November when water temperature fell below 13°C. In this study, only temperature had a strong direct relationship with abundance of eggs and larvae; 889 FISHERY BULLETIN: VOL. 84, NO. 4 peak abundance tracked the seasonal temperature cycle closely. Temperature stratification was most pronounced in June-October when spawning was greatest, especially at station 5 where salinity stratification was also most noticeable. Offshore transport of eggs and larvae is believed to be one of the environmental hazards to anchovy reproductive success (Bakun and Parrish 1982). Peak spawning in the Bay took place in June- August, the months of greatest offshore directed Ekman transport at the latitude of San Francisco (Parrish et al. 1981). Larvae, retained in San Fran- cisco Bay by estuarine circulation or behavior, would not be subject to offshore drift into areas of low plankton density. Therefore, they may have a higher probability of survival than larvae in the California Current and they might survive during bad years for oceanic larvae. Within San Francisco Bay there were apparent differences between spawning habitat and larval habitat. Eggs and small larvae were more abundant in warm, clear, thermally stratified water with relatively less plankton; large larvae were found in shallow, warm, less stratified, plankton-rich water with reduced light penetration. Negative correla- tions between zooplankton and the eggs of zooplank- tivorous fishes were attributed to predation on the zooplankton by de Ciechomski and Sanchez (1983). Cannibalism on larvae by adult northern anchovies and competition between adults and juveniles are two reasons why separate habitats would be adap- tive. Because spawning and nursery habitats differ in location and environmental properties, it is not surprising that multiple regression variables mea- sured in the spawning habitat did not predict lar- val abundance. It may be that spawning areas are selected by adults, perhaps for feeding (Brewer 1978) or for water clarity, while larger larvae seek different conditions where their survival is deter- mined by other factors than those which affect first- feeding larvae. If variable mortality on the larger larvae determines eventual recruitment, then recruitment may be largely decoupled from spawn- ing and first-feeding conditions. This could explain why predictions of recruitment from larval surveys (which do not adequately sample large larvae and juveniles) have not been reliable. The conditions where larvae were more abundant are more characteristic of shallow nearshore water than of the California Current. Juveniles and young of the year are also relatively more abundant near- shore in California (Parrish et al. 1986). In 1978, when spawning was restricted to nearshore areas, apparent recruitment was high relative to 1979 when spawning was offshore (Hewitt and Methot 1982). The 1978 spawning season for California Cur- rent anchovy was not typical; storms prevented favorable conditions for larvae until March in south- ern California (Lasker 1981). Nearshore areas might be refugia during anomalous years and they could contribute a disproportionate number of recruits every year (Brewer and Smith 1982). It might be argued that the 20-30 mm larvae found nearshore in the Southern California Bight (Brewer and Smith 1982) merely avoided the nets in stan- dard CalCOFI tows, but I found a similar pattern with respect to length frequencies when comparing samples taken in the channels and in shallow water in San Francisco Bay. That is, larger larvae were found in shallower zooplankton-rich areas. Estuaries and nearshore areas may provide conditions favor- able enough for survival of larvae and juveniles to compensate for low mean food density and for occa- sional years of unfavorable oceanographic conditions in the California Current. San Francisco Bay northern anchovy larvae, especially those which overwinter, are subject to dif- ferent ecological conditions than those in the Califor- nia Current, thus they may have slightly different morphology and meristics (Hempel and Blaxter 1961; Blaber et al. 1981). The San Francisco Bay subspecies Engraulis mordax nanus Hubbs (1925) may be an ecotype of E. mordax. A female northern anchovy has enough energy stored as fat for 17 of its 20 annual batches of eggs, but protein for egg production must come from feeding during the spawning season (Hunter and Dorr 1982). The primary food of northern anchovy, zooplankton, was abundant in the Bay. I found a mean density of 1 zooplankter/L with a 0.308 mm mesh net, but this is an underestimate of cope- podites and small copepods because of the relative- ly large mesh size. By comparison, Hutchinson (1981) found at least order of magnitude greater densities at nearby stations over the same time period using 0.080 and 0.064 mm mesh nets. An- chovy feed by biting individual organisms or by filter-feeding if particle density is high enough. The laboratory-determined threshold for filter-feeding is 5-18 particles (0.236 mm wide) per liter (Hunter and Dorr 1982). My zooplankton density estimate, which was biased conservatively, is of the order of magnitude required to stimulate filter-feeding. Therefore, I conclude that zooplankton prey for adult northern anchovies were abundant in the Bay during this study. For the Bay to be a good larval nursery area it should have abundant microzooplankton prey for lar- 890 McGOWAN: SPAWNING OF NORTHERN ANCHOVY vae. I found a mean density of 28.8 per liter using a 0.080 mm mesh net (probably a conservative esti- mate because of net clogging and meter malfunc- tioning). This is higher than would be expected in the California Current using the same mesh size (<1 per liter, Arthur 1977). It is comparable to the 36 per liter found with a finer mesh net (Arthur 1977). It is an underestimate of available prey for larvae because they consume particles as small as 0.040 mm, and there is a peak of biomass of small plankton in the California Current at 0.070 mm (Arthur 1977), just below the mesh size of my net. Sitts and Knight (1979) found a mean density of 32.3 copepod nau- plii/L in a 1-yr study in the Sacramento-San Joaquin estuary using 0.060 mm mesh. Hutchinson (1981) found approximately 10 nauplii/L over the same period of time as this study. (I calculated this value from her data for density of nauplii at 1 m depth at her stations 19 and 30 which correspond to my stations 6 and 2.) My microzooplankton estimates did not adequately represent the rotifers, tintinnids, and other small larval prey which were collected in high numbers with finer mesh nets (Hutchinson 1981). These organisms are known to be eaten by northern anchovy larvae and I observed tintinnids in the guts of some larvae. Larvae reared in the laboratory generally require more than 1,000 prey items/L for good survival, but some survival occurs at lower densities. Houde (1978) obtained 1% survival to metamorphosis of Anchoa mitchilli with a prey density of 27 per liter. Northern anchovy larvae in the sea which obtain enough food to survive also obtain enough to grow rapidly (Methot and Kramer 1979). The existence of dense patches of food has been suggested to ac- count for the discrepancy between average food den- sities observed in the sea and those needed in the laboratory. Dense patches of larval prey might not be needed in the Bay where I found mean prey den- sity higher than that typical of the California Cur- rent. However, dense patches of microzooplankton would be expected in the Bay because blooms of their prey, phytoplankton, occur (Cloern 1982). Dense patches of microzooplankton, undetected by my sampling design, would make San Francisco Bay a very good feeding area for larval northern an- chovies. Because the water was warmer in the Bay than in the California Current, larvae could search a larger volume of water per unit time, they would encounter high densities of prey and would be ex- pected to survive in greater numbers and to grow rapidly. Therefore, San Francisco Bay may be a good feeding area for larvae as well as for spawn- ing adults. To my knowledge, my estimates of spawning bio- mass of northern anchovies in the Bay are the first such estimates. Are they reasonable, and what are the implications of this biomass of anchovies in the Bay? The estimate based on egg abundance assumes that parameters estimated for California Current anchovies apply to San Francisco Bay anchovies. I argue they do because parameters for the estimate were obtained from anchovies at the peak of spawn- ing in the California Current in 1978, the year my study began. I believe these parameter values may be applied to the anchovy population in San Fran- cisco Bay because the seasonal pattern of spawn- ing and abundance of anchovies in the Bay indicates that most of these anchovies are seasonal migrants from the California Current stocks. No actual mea- surements of batch fecundity of anchovy in the Bay have been taken so the values used are the best available. Errors in estimating egg and larval abun- dances are probably more important than small changes in the estimates of batch fecundity. The egg-based estimate could be high if adults leave the Bay immediately after spawning or if they spawn more frequently due to greater food availability. The estimate could be low if they spawn infrequently because the season is later than the regular spawn- ing season in the California Current or if higher temperatures greatly increase metabolic needs. The estimate is conservatively biased because I merely divided the number of eggs caught by the number of days to hatch at the measured tempera- ture without considering mortality. During the months with peak egg abundance the estimated time to hatch was 2 d. If egg mortality was 0.184 da-1 (Picquelle and Hewitt 1984), then the estimate was approximately 25% low. The estimate would be high if eggs were present only in the channel and not over the area used to calculate total abundance. However, station 3, in shallow water near San Bruno Shoal in South San Francisco Bay, had high egg densities; therefore, eggs were distributed in some shallow- water areas. Stations 1 and 2, which had high egg densities, represented small areas, while stations 4 and 6 with low densities represented large areas. San Pablo Bay and the rest of the North Bay were not included in the biomass estimate. Potential biases in the egg-based stock estimate either cancel one another or give a conservative estimate. My estimate is consistent with information from other studies. I found mean values of 3,360 eggs/ 1,000 m3 and 259 larvae/1,000 m3. Hutchinson (1981) found 4,730 eggs/1,000 m3 (my calculations from her stations 19 and 30). Sitts and Knight (1979) calculated a mean larval abundance of 490 per 1,000 891 FISHERY BULLETIN: VOL. 84, NO. 4 m3. The estimates of larval densities are similar to estimates for the Southern California Bight near- shore CalCOFI area in 1978-79 (461 per 1,000 m3, calculated from table 4 of Brewer and Smith 1982, assuming average tow depth = 210 m; two-thirds of the stations were >210 m according to their table 2). The mean density of eggs in the Bay was much higher than in the Southern California Bight near- shore CalCOFI area (310 per 1,000 m3, Brewer and Smith 1982). The seasonal northern anchovies fish- ery in the Bay took approximately 481 tons for frozen and live bait (Smith and Kato 1979). My esti- mate is adequate to permit such a yield. Northern anchovy females need a daily ration of 4-5% of their body weight of copepods per day to support growth and reproduction (Hunter and Leong 1981). Approximately 5% of caloric intake goes into growth. Using these values, 38.35 tons of copepods per day would be consumed by the July biomass of 767 tons of anchovies. Growth would be about 1.92 tons per day. Doing similar calculations for each month and summing for the 12 mo of this study result in an estimate of 3,260 tons of cope- pods consumed and a net annual production of 158 tons of anchovy growth. If the egg estimates based on the area of the Bay, including the shallow areas were used, the consumption of copepods and growth estimates would be approximately doubled. These calculations are a first order estimate of the impact of a carnivorous planktivore on zooplankton in the Bay. The energy converted to anchovy growth would be removed from the Bay, so the estimate of net growth is also a minimum estimate of a sink for Bay production as growth of a transient consumer. In San Francisco Bay where plankton production from a limited area is being consumed by a large, transient anchovy population, grazing by anchovy could conceivably limit zooplankton abundance seasonally. Although it is impossible to distinguish between grazing and interannual differences with- out estimates of zooplankton production, zooplank- ton was more abundant in winter 1978-79 when adult anchovies were absent. A large biomass of planktivores could have other effects on the ecology of the Bay. Selective feeding by clupeoids on larger organisms in lakes can affect the zooplankton community structure (Brooks and Dodson 1965). Northern anchovy schools can also have an impact on nutrient cycling. Smith and Epley (1982) calculated that ambient ammonium concen- tration would be nearly doubled behind an anchovy school in the Southern California Bight. McCarthy and Whitledge (1972) estimated that nitrogen excre- tion by the Peruvian anchoveta is an order of mag- nitude greater than zooplankton excretion, so fish excretion may be the major source of regenerated nitrogen nutrients for phytoplankton production. These high nitrogen inputs would be patchy (Blax- ter and Hunter 1982) and their importance would depend on whether or not background levels of nutrients were limiting. Nutrients may not be limit- ing in San Francisco Bay where light penetration and residence time control phytoplankton dynamics (Cloern 1979). Laboratory studies of copepod pro- ductivity, anchovy predation, and nutrient regenera- tion are needed to define quantitatively the impact of the northern anchovy on plankton dynamics in the Bay. A complete description of the trophic role of anchovy in the Bay should include estimates of zooplankton consumption by larvae, cannibalism by adults, and predation on adult and larval anchovies. CONCLUSION San Francisco Bay is a favorable area for north- ern anchovy spawning because it has abundant food for adults, protection from advective loss for eggs, and abundant food for larvae. There is apparent habitat partitioning between spawning adults and larger larvae which could adaptively reduce preda- tion and competition. Recruitment to the Califor- nia Current stocks may be determined more by events in the nursery habitat of larvae and juveniles than by conditions favorable for spawning adults and first-feeding larvae; therefore, further work in estu- aries and nearshore areas is warranted. ACKNOWLEDGMENTS This study was done under the direction of Mar- garet G. Bradbury as partial fulfillment of the re- quirement for the M.A. in Biology at San Francisco State University. I thank her and the other members of my committee, Robert Berrend and Thomas Niesen, for advice and assistance in completing the work. Michael Hearne supplied the plankton net and assisted in all the field sampling. The figures were drafted by J. Javech. The preparation of the manu- script was supported in part by the National Oceanic and Atmospheric Administration under Cooperative Agreement #NA 84-WC-H-06098. LITERATURE CITED Aplin, J. A. 1967. Biological survey of San Francisco Bay, 1963-1966. Calif. Fish Game, Mar. Resour. Oper. Lab. Ref. No. 67-4, 131 p. 892 McGOWAN: SPAWNING OF NORTHERN ANCHOVY Arthur, D. K. 1976. Food and feeding of larvae of three fishes occurring in the California Current, Sardinops sagax, Engraulis mor- dax, and Trachurus symmetricus. Fish. Bull., U.S. 74: 517-530. 1977. Distribution, size, and abundance of microcopepods in the California Current System and their possible influence on survival of marine teleost larvae. Fish. Bull., U.S. 75: 601-611. Bakun, A., and R. H. Parrish. 1982. Turbulence, transport, and pelagic fish in the Califor- nia and Peru current systems. CalCOFI Rep. 23:99-112. Blaber, S. J., D. P. Cyrus, and A. K. Whitfield. 1981 . The influence of zooplankton food resources on the mor- phology of the estuarine clupeid Gilchristella aestuarius (Gilchrist, 1914). Environ. Biol. Fish. 6:351-355. Blaxter, J. H. S., and J. R. Hunter. 1982. The biology of the clupeoid fishes. Adv. Mar. Biol. 20: 1-223. Brewer, G. D. 1976. Thermal tolerance and resistance of the northern an- chovy, Engraulis mordax. Fish. Bull., U.S. 74:433-445. 1978. Reproduction and spawning of the northern anchovy, Engraulis mordax, in San Pedro Bay, California. Calif. Fish Game 64:175-184. Brewer, G. D., and P. E. Smith. 1982. Northern anchovy and Pacific sardine spawning off Southern California during 1978-1980: preliminary observa- tions on the importance of the nearshore coastal region. CalCOFI Rep. 23:160-171. Brooks, J. L., and S. I. Dodson. 1965. Predation, body size, and composition of plankton. Science 150:28-35. Clarke, T. A. 1983. Comparison of abundance estimates of small fishes by three towed nets and preliminary results of the use of small purse seines as sampling devices. Biol. Oceanogr. 2:311- 340. Cloern, J. E. 1979. Phytoplankton ecology of the San Francisco Bay sys- tem: the status of our current understanding. In T. J. Con- omos (editor), San Francisco Bay: the urbanized estuary, p. 247-264. AAAS Pac. Div., San Franc, CA. 1982. Does the benthos control phytoplankton biomass in south San Francisco Bay? Mar. Ecol. Prog. Ser. 9:191-202. Conomos, T. J., and D. H. Peterson. 1977. Suspended particle transport and circulation in San Francisco Bay: an overview. In M. Wiley (editor), Estuarine processes, Vol. 2, p. 82-97. Acad. Press, N.Y. DE ClECHOMSKI, J. D., AND R. P. SANCHEZ. 1983. Relationship between ichthyoplankton abundance and associated zooplankton biomass in the shelf waters off Argentina. Biol. Oceanogr. 3:77-101. Eldridge, M. B. 1977. Factors influencing distribution of fish eggs and lar- vae over eight 24-hr samplings in Richardson Bay, Califor- nia. Calif. Fish Game 63:100-116. Hempel, G., and J. H. S. Blaxter. 1961. The experimental modification of meristic characters in herring (Clupea harengus L.). J. Cons. Perm. Explor. Mer 26:336-346. Hewitt, R. P., and R. D. Methot Jr. 1982. Distribution and mortality of northern anchovy larvae in 1978 and 1979. CalCOFI Rep. 23:226-245. Houde, E. D. 1978. Critical food concentrations for larvae of three species of subtropical marine fishes. Bull. Mar. Sci. 28:395-411. HUBBS, C. L. 1925. Racial and seasonal variation in the Pacific herring, California sardine, and California anchovy. Calif. Fish Game, Fish Bull. 8, 23 p. Hull, C. H., and N. H. Nie (editors). 1981. SPSS update 7-9. McGraw-Hill Book Co., N.Y. Hunter, J. R. 1976. Culture and growth of northern anchovy, Engraulis mordax, larvae. Fish. Bull., U.S. 74:81-88. Hunter, J. R., and H. Dorr. 1982. Thresholds for filter feeding in northern anchovy, Engraulis mordax. CalCOFI Rep. 23:198-204. Hunter, J. R., and S. R. Goldberg. 1980. Spawning incidence and batch fecundity in northern anchovy, Engraulis mordax. Fish. Bull., U.S. 77:641-652. Hunter, J. R., and R. Leong. 1981 . The spawning energetics of female northern anchovy, Engraulis mordax. Fish. Bull., U.S. 79:215-230. Hutchinson, A. 1981. Zooplankton species composition and abundance in the South Bay, 1978-1979. U.S. Geol. Surv. Open File Rep. 81- 132. Lasker, R. 1975. Field criteria for survival of anchovy larvae: the rela- tion between inshore chlorophyll maximum layers and suc- cessful first feeding. Fish. Bull., U.S. 73:453-462. 1978. The relation between oceanographic conditions and lar- val anchovy food in the California Current: identification of factors contributing to recruitment failure. Rapp. P. -v. Reun. Cons. int. Explor. Mer 177:322-354. 1981. Marine fish larvae. Univ. Wash. Press, Seattle, 131 P- Lund, J. W. G., C. Kipling, and E. D. LeCren. 1958. The inverted microscope method of estimating algal numbers and the statistical basis of estimations by count- ing. Hydrobiologia 11:143-170. McCarthy, J. J., and T. E. Whitledge. 1972. Nitrogen excretion by anchovy (Engraulis mordax and E. ringens) and jack mackerel (Trachurus symmetricus). Fish. Bull., U.S. 70:395-401. McGowan, M. F., and F. H. Berry. 1984. Clupeiformes: development and relationships. In H. G. Moser et al. (editors), Ontogeny and systematics of fishes, p. 108-126. Am. Soc. Ichthyol. Herpetol. Spec. Publ. 1, 760 P- Methot, R. D., Jr., and D. Kramer. 1979. Growth of northern anchovy, Engralis mordax, larvae in the sea. Fish. Bull., U.S. 77:413-423. Miller, D. J., and R. N. Lea. 1972. Guide to the coastal marine fishes of California. Calif. Fish Game, Fish Bull. 157, 235 p. Montgomery, D. C. 1976. Design and analysis of experiments. John Wiley & Sons, N.Y., 418 p. Murphy, G. I., and R. I. Clutter. 1972. Sampling anchovy larvae with a plankton purse seine. Fish. Bull., U.S. 70:789-798. Nie, N. H., C. H. Hull, J. G. Jenkins, K. Steinbrenner, and D. H. Bent. 1975. SPSS: statistical package for the social sciences. 2d ed. McGraw-Hill Book Co., N.Y. Parker, K. 1980. A direct method for estimating northern anchovy, Engraulis mordax, spawning biomass. Fish. Bull., U.S. 78:541-544. 893 Parrish, R. H., D. L. Mallicoate, and K. F. Mais. 1986. Regional variations in the growth and age composition of northern anchovy, Engraulis mordax. Fish. Bull., U.S. 83:483-496. Parrish, R. H., C. S. Nelson, and A. Bakun. 1981. Transport mechanisms and reproductive success of fishes in the California Current. Biol. Oceanogr. 1 : 175-203. PlCQUELLE, S. J., AND R. P. HEWITT. 1984. The 1983 spawning biomass of the northern anchovy. CalCOFI Rep. 25:16-27. Recksiek, C. W., and H. W. Frey. 1978. Biological, oceanographic, and acoustical aspects of the market squid, Loligo opalescens Berry. Calif. Fish Game, Fish Bull. 169:1-185. Richardson, S. L. 1980. Spawning biomass and early life of northern anchovy, Engraulis mordax, in the northern subpopulation off Ore- gon and Washington. Fish. Bull., U.S. 78:855-876. Scura, E. D., and C. W. Jerde. 1977. Various species of phytoplankton as food for larval northern anchovy Engraulis mordax, and relative nutritional value of the dinoflagellates Gymnodinium splendens and Gonyaulax polyedra. Fish. Bull., U.S. 75:577-583. Sitts, R. M., and A. W. Knight. 1979. Plankton ecology in the Sacramento-San Joaquin estu- ary. Water Sci. Eng. Pap. No. 4509, 145 p. Univ. Calif. Davis, Davis, CA. Smith, D. L. 1977. A guide to marine coastal plankton and marine inverte- FISHERY BULLETIN: VOL. 84, NO. 4 brate larvae. Kendall/Hunt Publ. Co., Dubuque, IA. Smith, P. E., and R. W. Eppley. 1982. Primary production and the anchovy population in the Southern California Bight. Limnol. Oceanogr. 27:1-17. Smith, P. E., and R. Lasker. 1978. Position of larval fish in an ecosystem. Rapp. P. -v. Reun. Cons. int. Explor. Mer 173:77-84. Smith, P. E., and S. L. Richardson. 1977. Standard techniques for pelagic fish egg and larva surveys. FAO Fish. Tech. Pap. No. 175, Rome. Smith, S. W., and S. Kato. 1979. The fisheries of San Francisco Bay: past, present, and future. In J. T. Conomos (editor), San Francisco Bay: the urbanized estuary, p. 445-468. AAAS Pac. Div., San Franc, CA. Tyler, J. E. 1968. The Secchi disk. Limnol. Oceanogr. 13:1-6. Vrooman, A. M., P. A. Paloma, and J. R. Zweifel. 1981. Electrophoretic, morphometric, and meristic studies of subpopulations of northern anchovy, Engraulis mordax. Calif. Fish Game 67:39-51. Wang, J. C. 1981. Taxonomy of the early life stages of fishes: fishes of the Sacramento-San Joaquin estuary and Moss Landing Harbor-Elkhorn Slough, California. Ecological Analysts, Inc., Concord, CA., 168 p. Zweifel, J. R., and R. Lasker. 1976. Prehatch and posthatch growth of fishes - a general model. Fish. Bull., U.S. 74:609-621. 894 THE SPAWNING FREQUENCY OF SKIPJACK TUNA, KATSUWONUS PEL AMIS, FROM THE SOUTH PACIFIC J. Roe Hunter,1 Beverly J. Macewicz,1 and John R. Sibert2 ABSTRACT Histological criteria to age postovulatory follicles were developed from examination of laboratory-spawned skipjack tuna; the criteria were used to estimate the frequency of spawning of skipjack tuna from the South Pacific. Examination of 87 skipjack tuna from field collections taken in October-November indicated that spawning occurred nearly every day. The fraction of mature females with postovulatory follicles, <24 hours old, was 0.85 (standard deviation = 0.071) indicating that the mean interval between spawn- ings was only 1.18 days. Estimates of the frequency of spawning of multi- ple spawning fishes are essential for understanding their reproductive biology. To estimate annual reproductive effort or fecundity, and how these variables are related to size or age structure of a population requires knowledge of the frequency of spawning and the number of eggs produced per spawning. Batch fecundity, the number of eggs pro- duced per spawning, has been estimated for skipjack tuna a number of times (see review by Matsumoto et al. 1984) but the spawning rate of the skipjack is unknown. Thus spawning frequency is one of the missing links in an assessment of the reproduction of skipjack populations. It has long been recognized that skipjack tuna spawn more than once in a season because more than one mode of advanced oocytes are found in active ovaries (Brock 1954; Bunag 1956; Joseph 1963; Raju 1964; Simmons 1969; Batts 1972; Cayre 1981; Goldberg and Au 1986). The frequency of occurrence of female black skipjack tuna, Euthyn- nus lineatus, throughout the spawning season with ovaries containing hydrated oocytes led Schaefer (1986) to conclude that the average interval be- tween spawnings of black skipjack in the eastern tropical Pacific was 2.1-5.7 d depending on the region. Over the last 6 years, two methods have been developed for measuring the spawning rate of multi- ple spawning marine fishes: One method is based on the frequency of ovaries containing hydrated Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. 2South Pacific Commission, P.O. Box D5, Noumea, CEDEX, New Caledonia. oocytes and the other is based on the frequency with which they contain postovulatory follicles of known age (Hunter and Macewicz 1985a). These methods have been used to measure the rate of spawning in a number of marine fishes: Engraulis mordax (Hunter and Goldberg 1980; Hunter and Macewicz 1980); Engraulis ringens (Alheit et al. 1984); Hypso- blennius jenkinsi (Present 1985); Sardinella brasi- liensis (Isaac-Nahum et al. 1985); Seriphus politus (DeMartini and Fountain 1981); and Euthynnus lineatus (Schaefer 1986). Postovulatory follicles were used in most studies, but DeMartini and Foun- tain (1981) and Schaefer (1986) used the incidence of females with hydrated oocytes to estimate spawn- ing frequency. The hydrated oocyte method may produce a biased estimate in some species because of increased vulnerability of hydrated females to net- ting gear (Alheit et al. 1984). The objective of this paper was to estimate the spawning rate of South Pacific skipjack tuna by applying some of these techniques. It was not possi- ble to use the hydrated ovary method in our study because fish were not caught during the period of the day when the ovary was hydrated. Instead, we used the incidence of females having ovaries con- taining postovulatory follicles to estimate the fre- quency of spawning of skipjack tuna. This method requires ovaries to be preserved immediately in for- maldehyde solution when the fish is caught, a histological examination of the ovary, and the devel- opment of a staging system for estimating the age of the postovulatory follicle. Our histological classi- fication included not only an assessment of spawn- ing frequency but also an assessment of the extent of ovarian atresia. The atretic condition of the ovary is a sensitive index of the reproductive state of Manuscript accepted June 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 895 FISHERY BULLETIN: VOL. 84, NO. 4 females during the spawning season and can be used to identify females approaching the end of their spawning season as well as those in postspawning condition (Hunter and Macewicz 1985b). METHODS Skipjack tuna were captured either by pole and line or were catches associated with moored fish at- traction devices or free floating natural flotsam. Two sets of collections of skipjack tuna were ana- lyzed: a group of 12 females taken near Noumea, New Caledonia on 23 February 1984; and a group of 87 females taken in 8 different collections at various locations in the South Pacific from 20 Octo- ber to 30 November 1984 (Table 1). Our samples were opportunistically taken and spanned a great latitudinal range (0°-23°S). At present the peak spawning months of skipjack tuna are poorly defined over this range of latitudes. Spawning is known to occur throughout the year in some areas (Nishikawa et al. 1985), but regional differences may exist in the peak months of spawning, and the spawning season also varies with skipjack size (Naganuma 1979). Naganuma concluded from analysis of gono- somatic indices (GSI) that peak spawning period for small skipjack tuna (40-60 cm) in the South Pacific is October to December. Argue et al. (1983) ex- amined 11,000 adult skipjack tuna for cannibalism of juveniles (15-70 mm) and for GSI over the same latitude range as this study, but covering 80° of longitude (140°W-140°E). They found that canni- balism and female GSI was highest between Octo- ber and March in this broad area. More data are needed to identify the regional variation about this general pattern. The 8 collections of gonads (collections 2-9, Table Table 1 .—Characteristics of 9 collections of female skipjack tuna taken in the South Pacific in 1984. Collec- Time of Fork length tion day Mean Range Lat.* Long.2 number Date (h) N (cm) (cm) Gear1 S E 1 2-23-84 0800 12 47 44-51 PL 23.00 167.00 2 10-20-84 0745 7 49 46-50 PS 16. 178-179 3 10-23-84 0700 6 48 46-52 PS 16. 178-179 4 10-24-84 0700 8 49 46-52 PS 16. 178-179 5 10-25-84 0700 7 50 47-52 PS 16. 178-179 6 10-26-84 0700 14 49 45-51 PS 16. 178-179 7 10-27-84 0645 8 48 46-50 PS 16. 178-179 8 11-19-84 0755 25 50 44-62 PS 03.41 144.08 9 11-30-84 1955 12 56 49-60 PS 0.03 147.46 1) were treated statistically as 8 "clusters" of ran- dom samples of unequal size. The mean proportion of postovulatory follicles <24 h old was calculated as the total number of females with such follicles divided by the total number of mature females. Cochran (1977) pointed out that estimation of vari- ance by the simple binomial probability formula can produce serious errors. The variance was calculated by the appropriate formula recommended by Coch- ran (1977). Three female skipjack tuna were spawned in cap- tivity (23°-24°C; June 1985) at the Kewalo Research Facility of the National Marine Fisheries Service using the stress spawning technique of Kaya et al. (1982). One fish (48 cm fork length [FL]) was sacrificed at the time of spawning, another (43.8 cm FL) 12 h later and the third (44 cm FL) 24 h after spawning. The ovaries of these females were used to establish histological criteria for the aging of the postovulatory follicles of the sea-caught females. Ovaries were preserved in 10% Formalin3 and embedded in Paraplast. Histological sections were cut at 5-6 /^m and stained with Harris hematox- ylin followed by eosin-phloxine-B counter stain (H&E). Histological Classification To estimate reproductive condition of skipjack tuna, we used two histological classification systems: one for estimating spawning frequency and the other for assessing the likelihood that a female will continue to spawn (atretic state of the ovary). Each ovary was classified histologically according to both systems. These classification systems were devel- oped for northern anchovy, Engraulis mordax, by Hunter and Goldberg (1980) and Hunter and Macewicz (1980, 1985a, b) and are used here with a few modifications appropriate to skipjack tuna ovarian structure and their rates of postovulatory follicle resorption. The descriptions of postovulatory follicles of different ages are from the three captive Hawaiian skipjack tuna. As these fish resorbed their postovulatory follicles much more rapidly than did the northern anchovy, we used stages of shorter duration. The atretic classification system remains unchanged, except for a few minor details of histo- logical structure based on our observations of sea- caught fish. We believe that the reproductive inter- pretations we associate with the atretic classes are PL = pole and line; PS = purse seine catch of skipjack tuna attracted to either a fish attraction device moored in waters of 350-450 m deep or natural flotsam. Latitude and longitude given in degrees and minutes when available. 3Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 896 HUNTER ET AL.: SPAWNING FREQUENCY OF SKIPJACK TUNA meaningful because the oocyte resorption seems to follow a similar sequence of stages in most teleosts (Bretschneider and Duyvene de Wit 1947; Lambert 1970). The rate a skipjack tuna ovary passes from one atretic state to another is not specified and would require an additional study of captive fish. The characteristics of the two classification systems are outlined below. Spawning Frequency Hydrated and Migratory Nucleus Stages Ovaries with many translucent hydrated oocytes (oocytes enlarged by fluid uptake just prior to ovula- tion) are classified in the hydrated stage. Spawn- ing is considered to be imminent. In northern an- chovy, spawning takes place in <12 h after the onset of hydration. No skipjack tuna with hydrated oocytes were taken in our field collections. Female skipjack tuna were taken with ovaries in the migra- tory nucleus stage. This stage occurs just before the onset of hydration and is characterized by the migra- tion of the nucleus to the animal pole of the oocyte and the beginning of the fusion of its yolk globules (Fig. 1). Age 0-H Postovulatory Follicles Ovaries with new postovulatory follicles with no signs of follicle degeneration are classed as age 0-h postovulatory follicles. Hydrated oocytes may occa- sionally be present. Estimated elapsed time from spawning is 0-2 h. No skipjack tuna taken at sea were in this stage, but from the laboratory speci- men (Fig. 2a, b) we can discern the following histo- logical characteristics: The new postovulatory follicle has an irregular, convoluted shape. The granulosa epithelial cell layer of the follicle appears as an irregularly looped cord of slightly hypertro- phied cuboidal cells with prominent healthy nuclei linearly arranged. The granulosa appears only loose- ly attached to the thecal connective tissue layer. Although the theca is less convoluted than the gran- ulosa layer, it is distinct, contains blood capillaries and appears thicker than the thecal layer seen in northern anchovy. Figure 1.— Skipjack tuna oocyte with migratory nucleus (n) and large oil droplet (o); bar = 0.1 mm. Age 12-H Postovulatory Follicles Twelve-hour-old postovulatory follicles (Fig. 2c, d) show signs of degeneration similar to that ob- served in northern anchovy after about 24 h. Histo- logical characteristics include the follicle which is smaller with fewer convolutions; a lumen which is evident; the degenerating granulosa which is no longer a recognizable unbroken cord of cells, but rather the cells are scattered in clumps in the lumen or may be irregularly attached to the theca; and some pycnotic or irregular nuclei which are evident. The theca has begun to disintegrate although it still remains thick and distinct. Deterioration of the theca is indicated by its overall smaller size, a more filamentous rather than cohesize cellular arrange- ment, and some irregular nuclei. Age 24-H Postovulatory Follicles Ovaries containing 24-h-old postovulatory follicles showed pronounced signs of degeneration similar to that observed in northern anchovy 48 h after spawning. At this stage the follicle is much smaller than that at 12 h but a lumen is still evident (Fig. 2e, f). Only few granulosa cells remain; they usual- ly have pycnotic nuclei and generally are loosely at- tached to the thecal layer. The thecal layer is still fairly thick although it contains some pycnotic 897 FISHERY BULLETIN: VOL. 84, NO. 4 Figure 2.— Degeneration of postovulatory follicles of skipjack tuna spawned in the laboratory. Arrow in left panel indicates the postovulatory follicle that is seen under a higher magnification in right panel, a and b, 0 h after spawning (no deterioration); c and d, 12 h after spawning (pronounced degeneration); and e and f, 24 h after spawning (little remains of the degenerating postovulatory follicle). Bar = 0.1 mm; g = granulosa epithelial cell layer; t = thecal cell layer; b = red blood cell(s); and a = early alpha stage atretic oocytes. 898 HUNTER ET AL.: SPAWNING FREQUENCY OF SKIPJACK TUNA nuclei, and lymphocytes, and has a more filamen- tous composition. Nonspawning (mature) Ovaries with many yolked oocytes and containing no hydrated oocytes or postovulatory follicles were classified as nonspawning. They may contain post- ovulatory follicles in advanced stages of degenera- tion which cannot be readily distinguished from late stage corpora atretica. Elapsed time from spawn- ing was more than 24 h. Also classified as nonspawn- ing (mature) were females in postspawning condi- tion. The ovaries of such females contained no yolked oocytes, but atretic follicles (beta stage) were present indicating that the ovary was active recently (see next section). Immature Ovaries containing no yolked oocytes and no a or ft stage atretic structures were classed as im- mature. Atretic States It is well known in seasonal spawning fishes that a low incidence of atresia (resorption of the oocyte and its follicle) occurs throughout the spawning season, but it becomes marked as the spawning season closes and the remaining advanced oocytes in the ovary are resorbed. During the initial atretic phase (a), the oocyte is resorbed and any yolk globules are broken down and resorbed by the hypertrophying granulosa cells of the follicle (Bret- schneider and Duyvene de Wit 1947; Lambert 1970). In the next stage (/?), all the yolk is gone, and there remains a small, rather compact structure with one or more cavities. The structure is composed of granulosa and theca cells with penetrating blood vessels. Further stages of follicle resorption have been described by the same authors, but the inci- dence and extent of a and (5 stages have proven to be the most useful in the classification of atretic states of ovaries (Hunter and Macewicz 1985b). The characteristics of a and p atretic structures are described and illustrated for northern anchovy by Hunter and Macewicz (1985b) and a atretic oocytes of skipjack tuna are essentially similar. However, P atresia differs from northern anchovy in contain- ing numerous spherical vacuoles scattered through- out the follicle. The vacuoles are the remnants of the oil droplet which takes longer than yolk to resorb and in H&E sections appear empty. Occasionally, a large beta stage follicle may be seen in which the granulosa and thecal cells have proliferated. Listed below are the characteristics of the four atretic states we used to classify skipjack tuna ovaries along with what is known regarding the spawning potential of northern anchovy classed in these states. Atretic State 0 Yolked oocytes present, with no a atresia of yolked oocytes; p stage atresia may be present, but it cannot be distinguished with certainty from late stage postovulatory follicles (>24 h old). Female northern anchovy in this state have a high poten- tial of spawning. Atretic State 1 Less than 50% of the yolked oocytes are in the a stage of atresia. The frequency of spawning for northern anchovy classed in this state is less than half of that for females classed in atretic state 0. Thus, atretic state 1 indicates a decline in spawn- ing rate. Atretic State 2 Fifty percent or more of the yolked oocytes are in the a stage of atresia. The frequency of spawn- ing for female northern anchovy classed in this state is very low and indicates that cessation of spawn- ing is imminent. Atretic State 3 Ovaries contain p stage atresia and no yolked oocytes. Such fish have completed their spawning season since they have no yolked oocytes. The pres- ence of p stage atresia indicates that oocyte resorp- tion has taken place and thereby distinguishes such recently mature but postspawning fish from imma- ture females. In northern anchovy, atretic state 3 may persist for 30 d. RESULTS AND DISCUSSION All postovulatory follicles in sea-caught skipjack were less degenerated than those observed in a laboratory specimen examined 24 h after spawning, indicating that all of those in the sea collections were <24 h old. The fraction of mature females with post- ovulatory follicles <24 h old ([55 + 18J/86, Table 2) was 0.85 with the standard deviation estimated to 899 FISHERY BULLETIN: VOL. 84, NO. 4 Table 2.— Numbers of female skipjack tuna in various spawning and atretic states. The 8 collections taken in the South Pacific between 20 October and 30 November 1984. Age (A) Postovulatory Collec- follicles Total tion number Atretic state1 (n) Non- mature A< 12 12 < A< 24 spawning females 2 0 0 0 2 2 1 1 0 2 3 2 0 0 1 1 3 0 0 1 1 Total 1 0 6 7 3 0 1 0 1 2 1 4 0 0 4 2 0 0 0 0 3 0 0 0 0 Total 5 0 1 6 24 0 3 0 0 3 1 3 0 0 3 2 1 0 0 1 3 0 0 0 0 Total 7 0 0 7 5 0 3 0 0 3 1 2 0 0 2 2 1 1 0 2 3 0 0 0 0 Total 6 1 0 7 6 0 9 0 0 9 1 5 0 0 5 2 0 0 0 0 3 0 0 0 0 Total 14 0 0 14 7 0 3 0 0 3 1 5 0 0 5 2 0 0 0 0 3 0 0 0 0 Total 8 0 0 8 8 0 8 5 0 13 1 5 2 1 8 2 31 0 2 3 3 0 0 1 1 Total 14 7 4 25 9 0 0 44 52 6 1 0 66 0 6 2 0 0 0 0 3 0 0 0 0 Total 0 10 2 12 2-9 0 27 9 5 41 1 25 8 3 36 2 3 1 3 7 3 0 0 2 2 Total 55 18 13 86 'Atretic State 0 State 1 State 2 State 3 no alpha stage atresia of yolked oocytes. alpha stage atresia of yolked oocytes present, but <50% affected. alpha stage atresia present, 50% or more yolked oocytes affected. no yolked oocytes present and beta stage atresia present. One female skipjack tuna in collection 4 was immature. 3A female with hydrated oocytes and age 0 h postovulatory follicles. "Three of these females had oocytes in migratory nucleus stage. 5Two of these females had oocytes in migratory nucleus stage. Five of these females had oocytes in migratory nucleus stage. be 0.071 (Cochran 1977; see methods). This means that the average interval between spawnings (1/0.85) was only 1.18 d. Only one female was imma- ture, reducing the denominator for the above frac- tion spawning from 87 to 86. If we consider only those females with yolked oocytes and no or minor atresia (atretic states 0 and 1) the fraction spawn- ing is 0.90, implying a mean interval of 1.11 d between spawnings. This indicates that the spawn- ing rate of female skipjack tuna in prime reproduc- tive condition is very close to daily. High levels of ovarian atresia were much more common among the 12 females taken in February than those taken in October-November, indicating that the February fish were nearing the end of their spawning season. Females with highly atretic ovaries (state 2) and postspawning ovaries (state 3) constituted 66% of the fish in the February collec- tions (Table 3), but they made up only 10% of the fish taken in October-November. The February col- lection was the only one taken by pole and line. It is possible that pole-and-line fishing may be selec- tive against spawning fish (Iverson et al. 1970; Matsumoto et al. 1984) although some spawning fish were taken in this collection. The most unusual feature of the February collec- tion was that the spawning fraction was high, 0.25 for a group where 50% of the fish were in post- spawning condition, had no yolked oocytes, and were incapable of spawning (atretic state 3). The spawning fraction was 1.0 for the three females with no or minor atresia because all three had postovula- tory follicles. Thus skipjack tuna with active ovaries appear to spawn nearly every day. It appears that those unable to maintain this rate may discontinue spawning and resorb the ovary because females with active ovaries, showing no evidence of spawning, were rare in all collections. Postspawning females Table 3. — Numbers of female skipjack tuna in various spawning and atretic states. This single collection was taken 23 February 1984. Collec- tion number Atretic state1 Postovulatory follicles 12 h 24 h Non- spawning Total mature females 1 0 1 2 3 Total 0 0 0 0 0 0 1 2 6 9 2 2 2 6 12 'Atretic State 0 = no alpha stage atresia of yolked oocytes. State 1 = alpha stage atresia of yolked oocytes present, but <50% affected. State 2 = alpha stage atresia present, 50% or more yolked oocytes affected. State 3 = no yolked oocytes present and beta stage atresia present. 900 HUNTER ET AL.: SPAWNING FREQUENCY OF SKIPJACK TUNA might reactivate their ovary sometime later in the year if their physiological condition favored repro- duction. Evidence for northern anchovy indicates that the transitions from spawning to postspawn- ing states and vice versa can occur rapidly. In the laboratory at 16 °C, northern anchovy can resorb all advanced oocytes within a few weeks (Hunter and Macewicz 1985b) and can produce an active ovary in 30 d (Hunter and Leong 1981). Owing to the higher water temperatures and high metabolism of skipjack tuna they are probably capable of even faster reproductive cycling. Histological examination of females taken late in the day (1955 h, collection 9) provided additional evidence for daily spawning. Eight of 10 females with postovulatory follicles in this collection also had oocytes in the migratory nucleus stage. This stage is the precursor to hydration. Thus, fish which had spawned <24 h before were beginning to hydrate a new batch of eggs which presumably would be spawned in <12 h. The migratory nucleus stage was observed only in this collection probably because it was the only one taken in the evening, whereas all others were taken in the morning (0645-0755). The rarity of females with hydrated oocytes in our col- lections and the age of the postovulatory follicles imply that spawning usually took place at night. Spawning in daylight hours has been observed by fishermen and scientists, however (Iverson et al. 1970; Matsumoto et al. 1984). A single female taken during the morning (collec- tion 8) had small (0.70 mm) early stage hydrated oocytes (hydrated oocytes in which the yolk globules had not fully fused). This female, the only one with hydrated oocytes in our collections, also had new postovulatory follicles despite the fact that the hydrated oocytes were not fully advanced. This female may have been induced to hydrate and spawn by the stress of capture or may be simply an excep- tion to the rule. To capture significant numbers of females with hydrated oocytes would probably re- quire sampling after 2100 h. It is important to cap- ture eventually some females in the hydrated stage because it is the best way to confirm that all oocytes in the most advanced modal group, the group of oocytes considered to be the next spawning batch (Hunter and Goldberg 1980), are in fact spawned. Counts of hydrated eggs are also the easiest and most accurate method of estimating batch fecundity (Hunter et al. 1985). The "stress" spawning technique of Kaya et al. (1982) was used to produce the spawned skipjack tuna for the aging of postovulatory follicles. In this technique females captured at sea and placed in a tank spawn spontaneously, usually about 8 h after capture presumably because of the stress of capture and handling. Spawning typically takes place at about 2400 h, which, by our estimate, appears to be close to the usual time of spawning. It now seems likely that many of these fish are naturally express- ing their daily spawning activity. On the other hand, eggs less than the normal size range, 0.8-1.17 mm (Matsumoto et al. 1984), are occasionally spawned, indicating that stress may induce premature hydra- tion in some individuals. That the skipjack tuna do not continue to spawn in the tanks is due probably to the stress of captivity. Our examination of a cap- tive skipjack 24 h after spawning indicated that nearly all remaining oocytes containing yolk were in the early stages of alpha atresia (Fig. 2e). Similar- ly, female northern anchovy nearly always resorb their advanced oocytes a few days after capture although they will subsequently mature and spawn (Leong 1971; Hunter and Macewicz 1985b). If female skipjack tuna spawn at the frequency we observed (85% of the females per day), the cost of reproduction and annual fecundity will be high because skipjack tuna appear to have a long spawn- ing season. The relative batch fecundity of skipjack (number of eggs per spawning per body weight) is about 100 eggs per gram (Matsumoto et al. 1984; Goldberg and Au 1985). Skipjack tuna eggs are about the same size as those of Scomber japonicus which average in weight 0.04 mg (unpubl. data, Na- tional Marine Fisheries Service, Southwest Fish- eries Center). We estimate the cost of a single spawning (excluding the metabolic cost of egg maturation and reproductive behavior) to be about 2% of the body weight per spawning (Scomber egg dry weight x relative batch fecundity x conversion to wet weight; 4 x 10 ~5 x 100 x 5 = 0.02). If a female spawned every 1.18 d over 3 mo (90 d), it would produce about 7,600 eggs per gram body weight at an average daily cost of 1.7% of the body weight per day; a 4 kg skipjack tuna would spawn about 30 million eggs over this period. If the collections used in this study were an un- biased sample of the South Pacific skipjack tuna population, then little doubt exists that spawning occurs almost daily when they have active ovaries. This preliminary study provides the tools necessary for a population-wide assessment of reproduction. We established the time-specific, histological criteria for assessment of spawning rate, and the method was applied to a small sample. A great deal more remains to be done for a proper assessment of reproduction in skipjack tuna. Specifically, many more samples at different times of day, using a 901 FISHERY BULLETIN: VOL. 84, NO. 4 variety of fishing gears, are needed to insure that sampling biases do not exist; a wide range of skip- jack tuna sizes or ages need to be sampled so that the age-specific reproductive effort can be esti- mated; and females with hydrated oocytes need to be collected to verify that nearly all oocytes in the most advanced mode are hydrated and spawned. The last point seems particularly important because our estimated body weight cost of reproduction is high and is very sensitive to the estimate of batch fecundity. It may never be practical to analyze histo- logically sufficient numbers of specimens to estimate spawning frequency for all months and ages since some spawning occurs the year around (Nishikawa et al. 1985). On the other hand, it may be practical to calibrate the gonosomatic index (GSI) in peak spawning months using histological criteria and to use the GSI as a calibrated index of spawning fre- quency during months of low spawning frequency. We do not intend to continue this work but we en- courage those working on the biology of tunas to include such studies in their research plans. ACKNOWLEDGMENTS We thank the SWFC Honolulu Laboratory of the National Marine Fisheries Service, and particular- ly Christofer Boggs, for providing the samples of recently spawned skipjack tuna from the Kewalo Research Facility. We also thank Robert Gillett and Richard Farman (South Pacific Commission) for col- lecting the wild fish and Robert Kearney and Kurt Schaefer (Inter- American Tropical Tuna Commis- sion) for reading the manuscript. LITERATURE CITED Alheit, J., V. H. Alarcon, and B. J. Macewicz. 1984. Spawning frequency and sex ratio in the Peruvian an- chovy, Engraulis ringens. Calif. Coop. Oceanic Fish. In- vest. Rep. 25:43-52. Argue, A. W., F. Conand, and D. Whyman. 1983. Spatial and temporal distributions of juvenile tunas from the stomachs of tunas caught by pole-and-line gear in the central and western Pacific Ocean. Tuna and Billfish Assessment Programme, Tech. Rep. No. 9, 47 p. South Pacific Commission, Noumea, New Caledonia. Batts, B. S. 1972. Sexual maturity, fecundity, and sex ratios of the skip- jack tuna, Katsuwonus pelamis (Linnaeus), in North Carolina waters. Trans. Am. Fish. Soc. 101:627-637. Bretschneider, L. H., and J. J. Duyvene de Wit. 1947. Sexual endocrinology of non-mammalian vertebrates. Monographs on the Progress of Research in Holland Dur- ing the War, Vol. 11, Elsevier, N.Y., 147 p. Brock, V. E. 1954. Some aspects of the biology of the aku, Katsuwonus -pelamis, in the Hawaiian Islands. Pac. Sci. 8:94-104. Bunag, D. M. 1956. Spawning habits of some Philippine tuna based on diameter measurements of the ovarian ova. Philipp. J. Fish. 4(2):145-177. Cayre, P. 1981. Maturite sexuelle, fecondite et sex ratio du Listao {Kat- suwonus pelamis) des cotes d'Afrique de l'ouest (20°N-0°N) etudies a partir des debarquements thoniers (1977 a 1979) au port de Dakar, (Senegal). Int. Comm. Conserv. Atl. Tunas, Collect. Vol. Sci. Pap. 15:135-149. Cochran, W. G. 1977. Sampling techniques. 3d ed. John Wiley & Sons, N.Y., 428 p. DeMartini, E. E., and R. K. Fountain. 1981. Ovarian cycling frequency and batch fecundity in the queenfish, Seriphus politus: attributes representative of serial spawning fishes. Fish. Bull., U.S. 79:547-560. Goldberg, S. R., and D. W. K. Au. In press. The spawning of skipjack tuna from southeastern Brazil as determined from histological examination of ovaries. In P.E. K. Symons, P. M. Miyake, and G. S. Saka- gawa (editors), Proceedings of the ICCAT Conference on the International Skipjack Year Program. Int. Comm. Conserv. Atl. Tunas, Madrid, Spain. Hunter, J. R., and S. R. Goldberg. 1980. Spawning incidence and batch fecundity in northern anchovy, Engraulis mordax. Fish. Bull., U.S. 77:641-652. Hunter, J. R., and R. Leong. 1981. The spawning energetics of female northern anchovy, Engraulis mordax. Fish. Bull., U.S. 79:215-230. Hunter, J. R., and N. C. H. Lo, and R. J. H. Leong. 1985. Batch fecundity in multiple spawning fishes. In R. Lasker (editor), An egg production method for estimating spawning biomass of pelagic fish: application to the north- ern anchovy (Engraulis mordax), p. 67-78. U.S. Dep. Com- mer., NOAA Tech. Rep. NMFS 36. Hunter, J. R., and B. J. Macewicz. 1980. Sexual maturity, batch fecundity, spawning frequen- cy, and temporal pattern of spawning for the northern an- chovy, Engraulis mordax, during the 1979 spawning season. Calif. Coop. Oceanic Fish. Invest. Rep. 21:139-149. 1985a. Measurement of spawning frequency in multiple spawning fishes. In R. Lasker (editor), An egg production method for estimating spawning biomass of pelagic fish: ap- plication to the northern anchovy (Engraulis mordax), p. 79-94. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 36. 1985b. Rates of atresia in the ovary of captive and wild north- ern anchovy, Engraulis mordax. Fish. Bull., U.S. 83:119- 136. Isaac-Nahum, V. J., R. de D. Cardoso, G. Servo, and C. L. del B. Rossi-Wongschowski. 1985. Some aspects of the spawning biology of the Brazilian sardine, Sardinella brasiliensis, (Clupeidae). International Council for the Exploration of the Sea. CM. 1985/H:63. Pelagic Fish Comm. 12 p. (Mimeo.) Iverson, R. T. B., E. L. Nakamura, and R. M. Gooding. 1970. Courting behavior in skipjack tuna, Katsuwonus pelamis. Trans. Am. Fish. Soc. 99:93. Joseph, J. 1963. Fecundity of yellowfin tuna (Thunnus albacares) and skipjack (Katsuwonus pelamis) from the eastern Pacific Ocean. Inter-Am. Trop. Tuna Comm. 7:255-292. Kaya, C. M., A. E. Dizon, S. D. Hendrix, T. K. Kazama, and M. K. K. Queenth. 1982. Rapid and spontaneous maturation, ovulation, and 902 HUNTER ET AL.: SPAWNING FREQUENCY OF SKIPJACK TUNA spawning of ova by newly captured skipjack tuna, Katsu- wonus pelamis. Fish. Bull, U.S. 80:393-396. Lambert, J. G. D. 1970. The ovary of the guppy, Poecilia reticulata. The atretic follicle, a corpus atreticum or a Corpus luteum praeovula- tionis. Z. Zellforsch 107:54-67. Leong, R. 1971. Induced spawning of the northern anchovy, Engraulis mordax Girard. Fish. Bull., U.S. 69:357-360. Matsumoto, W. M., R. A. Skillman, and A. E. Dizon. 1984. Synopsis of biological data on skipjack tuna, Katsu- wonus pelamis. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circ. 451, 92 p. [Also FAO Fisheries Synopsis No. 136.] Naganuma, A. 1979. On spawning activities of skipjack tuna in the western Pacific Ocean. Bull. Tohoku Reg. Fish. Res. Lab. 40:1-13. NlSHIKAWA, Y., M. HONMA, S. UEYANAGI, AND S. KlKAWA. 1985. Average distribution of larvae of oceanic species of scombroid fishes, 1956-1981. Far Seas Fish. Res. Lab. S series 12, 99 p. Present, T. M. C. 1985. Patterns and processes of energy allocation between growth and reproduction in the marine shore fish, Hypso- blennius jenkinsi. Ph.D. Thesis, Univ. California, San Diego, La Jolla, 250 p. Raju, G. 1964. Studies on the spawning of the oceanic skipjack Kat- suwonus pelamis (Linnaeus) in Minicoy waters. In Pro- ceedings of the Symposium on Scombroid Fishes, p. 744- 768. Mar. Biol. Assoc. India, Symp. Ser. 1. Schaefer, K. M. 1986. Reproductive biology of the black skipjack, Euthynnus lineatus. M.S. Thesis, San Diego State Univ., San Diego, CA, 115 p. Simmons, D. C. 1969. Maturity and spawning of skipjack tuna (Katsuwonus pelamis) in the Atlantic Ocean, with comments on nematode infestation of the ovaries. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 580, 17 p. 903 SURVIVAL AND GROWTH OF STRIPED BASS, MORONE SAXATILIS, AND MORONE HYBRID LARVAE: LABORATORY AND POND ENCLOSURE EXPERIMENTS1 Edward D. Houde2 and Lawrence Lubbers IIP ABSTRACT Survival and growth of striped bass, Morone saxatilis, and its hybrids were compared in the first 30 days after hatching to determine if the reported heterosis of hybrid striped bass is evident in the larval stage. Larvae of striped bass (SB); striped bass x white bass (WBX), M. saxatilis 9 x M . chrysops or; and striped bass x white perch (WPX), M. saxatilis 9x1, americana c, were reared under controlled conditions in the laboratory (19°C, 3°/oo) and under ambient conditions in freshwater pond enclosures. In the laboratory SB had a significantly higher mean survival rate at 30 days of age than either hybrid. In the pond enclosures neither mean survival nor size at 30 days differed significantly among the types of larvae. Mean rates of growth in length, which ranged from 0.28 to 0.36 mm d"1 in the laboratory and from 0.30 to 0.32 mm d"1 in the enclosures did not differ significantly among the types of larvae. Mean rates of growth in weight of 15.0 to 19.0% d"1 were not significantly different in the laboratory, but the rates did differ significantly in the pond enclosures, where the WBX (17.9% d~ x) and WPX (17.3% d"1) rates were significantly higher than the SB (15.5% d"1). If 30-day-old fry were to be reared in hatcheries, there is no clear production advantage for hybrids. A possible initial expression of hybrid vigor, recognized by faster rates of growth in weight, was evident in WBX and WPX at 1 month of age in the pond enclosures but not in the laboratory tanks. A series of recruitment failures (Cooper and Polgar 1981; Boreman and Austin 1985) has stimulated the development of hatcheries to culture juvenile striped bass, Morone saxatilis, or its hybrids for stocking in the Chesapeake Bay region. The striped bass and the striped bass x white bass, M. chrysops, hybrid have been cultured for stocking in freshwater and estuarine systems for several years and also have potential for commercial aquaculture (Bonn et al. 1976; Kerby et al. 1983). A second hybrid, striped bass x white perch, M. americana, has been pro- duced (Bayless 1972; Kerby and Joseph 1979) although its potential is less known. The striped bass x white bass hybrid demonstrates an apparent heterosis and usually grows and survives better dur- ing the first two years of life than does striped bass under similar culture conditions (Logan 1968; Ware 1975; Williams et al. 1981; Kerby et al. 1983). 'Contribution No. 1721, Center for Environmental and Estuarine Studies of the University of Maryland. 2University of Maryland, Center for Environmental and Estu- arine Studies, Chesapeake Biological Laboratory, Solomons, MD 20688-0038. 3University of Maryland, Center for Environmental and Estu- arine Studies, Chesapeake Biological Laboratory, Solomons, MD; present address: Maryland Department of Natural Resources, Tidewater Administration, Tawes State Office Building, Annapolis, MD 21401. The objective of our experiments was to determine if the apparent heterosis of the striped bass x white bass hybrid is established in the larval stage, between hatching and 30 d posthatch. We compared growth and survival of striped bass, striped bass x white bass, and striped bass x white perch (referred to hereafter as "striped bass", "white bass hybrid", and "white perch hybrid") in laboratory experiments and in fine-mesh enclosures within hatchery ponds. METHODS Laboratory Experiments Larvae originated from eggs of a single female striped bass, 15.4 kg, gillnetted in the Patuxent River, transported to the Manning Hatchery, Cedar- ville, MD, on 24 April 1982 and spawned by injec- tion of human chorionic gonadotropin on 27 April. Sperm from 2 male striped bass (Patuxent River), 12 male white bass (Tennessee Fish Commission), and 2 male white perch (Patuxent River) were used to fertilize portions of the spawned eggs. Embryos were incubated in 114 L polyethylene incubation chambers and larvae were held there in 15°-16°C freshwater until 6 d after hatching when some were brought to the Chesapeake Biological Laboratory. Manuscript accepted July 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 905 FISHERY BULLETIN: VOL. 84, NO. 4 Rearing Systems The striped bass and hybrid larvae were reared from 6 to 30 d after hatching in 36 L, rectangular glass aquaria. Each aquarium was lighted by two 61 cm, 40-W fluorescent lights 25 cm overhead on a 12-h light-12-h dark cycle. Immersion heaters con- trolled the temperature. For additional control, the aquaria sat in a shallow, refrigerated waterbath. An airstone in each aquarium provided oxygen and kept food dispersed. Temperature was maintained at 19° ± 1°C. Salin- ity was held at 3°/oo by diluting 5 ^m filtered Patux- ent River water with well water. All larvae were fed Artemia nauplii, eggs of which originated from Shark Bay, Australia. Water quality was maintained by replacing half of the water in each aquarium on alternate days. Feces, dead Artemia, and dead lar- vae were siphoned off each day. Ammonia levels were checked on 13 May (16 d after hatching) and were <0.25 ppm in all tanks. The pH in the nine rear- ing tanks ranged from 8.0 to 8.4 on 11 May (14 d after hatching) and from 8.3 to 8.4 on 26 May (29 d after hatching). Food Levels, Larval Densities, and Sampling Two Artemia nauplii levels, 100 L_1 and 500 Lr1, were tested. The lower level is similar to zoo- plankton densities in Chesapeake Bay subestuaries where striped bass larvae occur (Miller 1978). For each of the larval types duplicate experiments were run at the 500 L_1 level but only a single experi- ment was run at 100 Lr1. Food was first offered at 6 d after hatching when the experiments started. Artemia nauplii concentrations in each aquarium were checked twice daily by counting the num- ber in pipetted 100 cc aliquots. Food levels were maintained and adjusted by adding suspensions of Artemia of known concentration to the aquar- ia. In each aquarium, 144 larvae were stocked at an initial, relatively low density of 4.0 Lr1. Some larvae were preserved in 5% Formalin4 at the start of experiments (6 d after hatching). Three or four larvae from each aquarium were sampled and preserved on days 8, 10, 13, 16, 19, and 25 for growth rate determination. Samples (15-27 larvae) of survivors were preserved at 30 d when ex- periments were terminated. Preserved larvae were 4Reference to trade names does not imply endorsement by the National Marine Fisheries Service. measured and wet- weighed (nearest 0.1 mg after blotting). Analysis The expected number of survivors in each experi- ment is the number that would have survived had no larvae been sampled and preserved during the experiments. If Z = F + M, where Z is instanta- neous total mortality and F is preservation mortal- ity, then M is mortality from all other causes. The expression Nt = N0 e~(F+M)t applies, where Nt is number of survivors at age t (30 d) and N0 is initial number of stocked larvae (144 at 6 d). Knowing N0, Nt , Z, and F, we solved for M and then estimated expected survivors, if no larvae had been preserved, as N't = N0 e~m. Analysis of variance was used to test for survival differences among types of larvae and between food levels. Lengths and weights of the three types of larvae were compared at 6 d after hatching and when ex- periments terminated. In addition, lengths and weights at the 100 L_1 and 500 L_1 food levels were compared to determine if food concentration affected mean sizes. Comparisons were carried out using analysis of variance followed by the SNK multiple comparison test. Growth in length was described by linear regres- sions of standard length on days after hatching, lt = a + bt, where lt is estimated length (mm) at age t and b is daily growth rate (mm day-1). Growth in weight was determined from the exponential regres- sion of wet weight (mg) on days after hatching, Wt = W0 eGt, where Wt is estimated weight at age t and G is the instantaneous daily growth coefficient (day-1). Percent daily weight gains were calculated as 100 (eG - 1). Weight-length relationships were obtained from the power function, W = aLb, where Wis wet weight (mg), I is standard length (mm), and a and b are coefficients from the fitted regression. Enclosure Experiments Cubic enclosures, 1.32 m on each side, open at the top, and constructed of wood frames and 500 yon Nitex mesh, were submerged to a depth of 1.12 m in a 1-acre, freshwater pond of 1.5 m mean depth at the Manning Hatchery. The nine enclosures, each holding 2 m3, were placed in the pond from 3 to 5 d before larvae were stocked. Enclosures were assigned to the striped bass and two hybrids using a linearized Latin-square design (Steel and Torrie 1960) with three replicates for each type of larva. The larvae were progeny of a single 10.4 kg female 906 HOUDE and LUBBERS: SURVIVAL AND GROWTH OF STRIPED BASS striped bass from the Patuxent River. Sperm from Patuxent River male striped bass were used to fer- tilize eggs. The hybrids resulted from fertilization by Tennessee white bass males and Patuxent River white perch males. Larvae were held in hatchery troughs and fed Artemia nauplii from 6 to 8 d after hatching. A total of 2,500 9-d-old larvae were stocked in each en- closure on 12 May 1983. Larvae were sampled by dipnet and preserved in 5% Formalin at 13, 17, 20, 23, and 27 d after hatching. At 30 d all survivors from each enclosure were counted and samples pre- served. Temperatures in the pond ranged from 18.5° to 22.0°C during the course of the experiment. Pond Zooplankton The kinds and abundances of potentially edible zooplankton were sampled on each day that larvae were collected, using a 15 cm diameter, 72 pm mesh plankton net that was lifted vertically in each enclosure. For comparison, zooplankton also was collected in three vertical lifts of the net outside the enclosures. Analysis Survival, lengths and weights at age, growth rates, and weight-length relationships were cal- culated as for the laboratory experiments. Variance, covariance, and regression analyses were used to test for differences in means among the striped bass and two types of hybrid larvae. RESULTS Laboratory Experiments Survival Survival at 30 d after hatching ranged from 45.8 to 85.4% (Table 1). Mean percentage survivals were striped bass, 84.7%; white bass hybrid, 60.4%; and white perch hybrid, 73.1%. The mean expected number of survivors differed significantly among types of larvae (ANOVA, P < 0.05). Mean survival of striped bass was significantly higher than that of the white bass hybrids (SNK multiple comparison procedure, P < 0.05). There were no detectable dif- ferences in mean survival between the two Artemia nauplii feeding levels (ANOVA, P > 0.05). Size-at-Age The white perch hybrid larvae were significantly shorter and weighed less than either striped bass or white bass hybrid larvae when the experiments began at 6 d after hatching, before larvae had been fed (Table 2; ANOVA, P < 0.05). At 30 d after hatching there were some statistical- ly significant differences in mean lengths and weights among the three types of larvae, and be- tween the two food levels, but no clear result was obtained (Table 2). No significant differences among mean lengths or weights of the white bass hybrid larvae were detected between the 100 L_1 and 500 L_1 food levels. But, the striped bass and white Table 1.— Survival at 30 d after hatching of striped bass (SB), striped bass x white bass (WBX), and striped bass x white perch (WPX) larvae in laboratory experiments at two food levels. Larvae and experiment numbers Artemia concentration (number L"1) Number preserved Number of survivors Expected number1 of survivors Expected instantaneous daily mortality rates (Z) Expected percentage survival SB-1 SB-2 SB-3 SB mean WBX-1 WBX-2 WBX-3 WBX mean WPX-1 WPX-2 WPX-3 WPX mean 500 500 100 500 500 100 500 500 100 20 106 123 0.0066 85.4 18 108 123 0.0066 85.4 19 104 122 0.0069 83.3 106.0 2122.0 0.0069 84.7 18 58 66 0.0325 45.8 20 93 108 0.0120 75.0 18 76 87 0.0210 60.4 75.7 87.0 0.0210 60.4 22 85 100 0.0152 69.4 18 100 114 0.0097 79.2 18 89 102 0.0144 70.8 91.3 105.3 0.0130 73.1 'Expected number of survivors is the adjusted number, accounting for samples of larvae that were preserved dur- ing the experiment (see Methods). 2The SB mean differed significantly from the WBX and WPX means (Analysis of variance followed by SNK multiple comparison procedure, P < 0.05). 907 FISHERY BULLETIN: VOL. 84, NO. 4 Table 2.— Mean standard lengths and wet weights of larvae of striped bass (SB), striped bass x white bass (WBX), and striped bass x white perch (WPX) from specimens preserved at 6 d after hatching, immediately before the experiments began and at 30 d after hatching when the experiments were terminated. SIX DAYS Mean length (mm) Mean wet weight (mg) and standard error and standard error Number Larvae preserved X Sj X s* SB 15 5.49 0.06 0.95 0.04 WBX 19 5.29 0.03 0.96 0.03 WPX 17 15.20 0.06 10.85 0.02 THIRTY DAYS Mean wet i _ weiaht Larvae and experiment Artemia concen- tration Number Mean length (mm) and standard error (mg) and standard error number (number L"1) preserved X % X «* SB-3 100 18 212.39 0.17 225.7 1.4 SB-1 500 20 314.26 0.17 "49.1 2.2 SB-2 500 19 314.57 0.17 450.4 1.8 SB mean 13.74 41.7 WBX-3 100 15 13.02 0.30 30.1 2.1 WBX-1 500 18 12.68 0.26 28.8 2.2 WBX-2 500 19 12.73 0.38 29.3 3.6 WBX mean 12.81 29.4 WPX-3 100 21 211.86 0.25 221.1 1.6 WPX-1 500 18 13.22 0.31 33.6 2.2 WPX-2 500 27 13.15 0.22 35.0 2.2 WPX mean 12.74 29.9 'Differ significantly, P < 0.05, from both SB and WBX. ANOVA followed by SNK multiple comparison procedure. 2Differ significantly, P < 0.05, from the 500 L~1 means. ANOVA. 3Differ significantly, P < 0.05, from all WBX and WPX mean lengths. ANOVA followed by SNK multiple comparison procedure. "Differ significantly, P < 0.05, from all WBX and WPX mean weights. ANOVA followed by SNK multiple comparison procedure. perch hybrid larvae were longer and heavier at the 500 L"1 level (ANOVA, P < 0.05). At the 500 L"1 food level the striped bass were significantly heavier than either hybrid (ANOVA and SNK multiple com- parison procedure, P < 0.05). The mean lengths of 30-d-old striped bass at 500 L_1 food level were significantly longer than the mean lengths of the hybrids (Table 2) (ANOVA and SNK multiple com- parison procedure, P < 0.05). Growth Rates From 6 to 30 d after hatching larvae grew in length at mean rates ranging from 0.28 to 0.36 mm d_1 (Table 3, Fig. 1). There were no significant dif- ferences in the growth-in-length rates among the three types of larvae at the 500 L_1 Artemia food level. The exponential regressions of mean weights on age (Table 3, Fig. 2) gave instantaneous growth coefficients ranging from 0.1396 to 0.1739 d_1, equivalent to 15-19% d~: weight gains. None of the coefficients differed significantly from each other (ANCOVA, P > 0.50). There were no significant differences in weight- length relationships among types of larvae or be- tween food levels (ANCOVA, P > 0.50). An average relationship, based on the total regression compo- nent of the ANCOVA, is W = 7.17 x 10 "4 J4-2399. Enclosure Experiments Survival Survival of striped bass and hybrid larvae at 30 d after hatching ranged from 13.1 to 33.8% in the nine enclosures. At 30 d there was no indication that striped bass or either hybrid was superior in sur- vival capability. The mean percentage survivals for the three types of larvae ranged from 22.0 to 28.5% (Table 4B) and did not differ significantly (ANOVA on arcsin mean percent survivals). The mean over- all survival rate for the three kinds of larvae was 25.0%. 908 HOUDE and LUBBERS: SURVIVAL AND GROWTH OF STRIPED BASS Table 3.— Linear regressions describing growth in length and exponential regressions describing growth in weight of striped bass (SB), striped bass x white bass (WBX), and striped bass x white perch (WPX) during the period 6-30 d after hatching. In the linear regression, / is standard length in mm, t equals days after hatching, b equals growth rate in mm, and a is the y-axis intercept. In the exponential regressions, W is wet weight in mg, f equals days after hatching, G is the instantaneous growth coefficient, and W0 is the theoretical weight in mg at time zero. Larvae and experiment number Artemia concentration (number L~1) LENGTH Equation / = a + bt Standard error of b SB-3 100 V = 3.64 + 0.29f 0.01 0.99 SB-1 500 / = 3.13 + 0.36f 0.02 0.99 SB-2 500 / = 3.20 + 0.36r 0.01 0.99 WBX-3 100 / = 3.10 + 0.34f 0.02 0.98 WBX-1 500 / = 3.36 + 0.32f 0.01 0.99 WBX-2 500 / = 3.31 + 0.32f 0.02 0.98 WPX-3 100 V = 3.70 + 0.28f 0.02 0.98 WPX-1 500 / = 3.18 + 0.32r 0.01 0.99 WPX-2 500 / = 3.22 + WEIGHT 0.32f 0.01 0.99 Larvae and Artemia Equation Standard Percent experiment concentration error gain number (number L"1) W = W0 eG' of G r2 (°/o d"1) SB-3 100 W = 0.51 e01472f 0.0158 0.94 15.9 SB-1 500 W = 0.41 e01713' 0.0141 0.96 18.7 SB-2 500 W = 0.42 e° 1739f 0.0146 0.96 19.0 WBX-3 100 W = 0.41 e0158u 0.0147 0.95 17.1 WBX-1 500 w = 0.41 e01576f 0.0145 0.95 17.1 WBX-2 500 w = 0.31 e01645f 0.0154 0.95 17.9 WPX-3 100 w = 0.48 e° 1396( 0.0123 0.96 15.0 WPX-1 500 w = 0.33 e° 1625' 0.0073 0.96 17.6 WPX-2 500 w = 0.35 e° 1520' 0.0091 0.98 16.4 'Differ significantly from SB-1 and SB-2, P< 0.01. comparison procedure. ANCOVA followed by SNK multiple The mean instantaneous mortality rates during the 9-30 d after hatching ranged from 0.0601 to 0.0713 d-\ equivalent to 5.8 to 6.9% d"1 (Table 4B). Cannibalism probably occurred during the last 10 d of the experiment. Some large survivors had small larvae in their stomachs when the experiments ended. Size-at-Age When larvae were stocked in the enclosures at 9 d after hatching, white bass hybrid larvae were significantly heavier (ANOVA, P < 0.01) and slight- ly, but not significantly, longer than white perch hybrid and striped bass larvae (Table 4A). Because all larvae had been fed Artemia nauplii in the hatch- ery for 3 d prior to stocking it is not known if the sizes at stocking reflect the relative weights and lengths of the three kinds of larvae before they began to feed. At 30 d after hatching mean lengths of striped bass and hybrid larvae from the enclosures ranged from 12.58 to 12.96 mm SL (Table 4C). Mean wet weights ranged from 38.38 to 43.28 mg (Table 4C). There were no significant differences in mean lengths or weights among the three types of larvae or among the nine enclosures (ANOVA, P > 0.25). Growth Rates Mean rates of growth in length for the striped bass and hybrid larvae ranged from 0.30 to 0.32 mm d_1 (Table 4D; Fig. 3). There were no significant differ- ences in the rates among types of larvae or among replicate enclosures (ANCOVA, P > 0.10). The common, instantaneous rates of growth in weight were 0.1444 for striped bass (= 15.5% d_1), 0.1650 for white bass hybrids (= 17.9% d"1) and 0.1593 for white perch hybrids (= 17.3% d"1). The rates of growth (Table 4E; Fig. 4) differed signifi- cantly among the three types of larvae (ANCOVA, P < 0.05) but not among enclosures (P > 0.10). The 909 FISHERY BULLETIN: VOL. 84, NO. 4 13 11- 100 Artemia, liter" 1 E E o) 5 c ~^- o -J 15r CO ■D « 13r CO 11 9 500 Artemia, liter-1 _i i i_ 6 8 10 13 16 19 25 Days after Hatching 30 Figure 1.— Mean standard lengths +2 standard errors of striped bass (SB), striped bass x white bass (WBX), and striped bass x white perch (WPX) larvae from 6 to 30 d after hatching, reared at two food levels in the laboratory. white bass hybrid and white perch hybrid rates were significantly higher than that for striped bass lar- vae (SNK multiple comparison procedure, P < 0.05). Weight-Length Relationships The wet weight-standard length relationships dif- fered significantly among the three types of larvae (ANCOVA, P < 0.001). The power coefficient of the white bass hybrid larvae was higher than those of the striped bass and white perch hybrid larvae (SNK multiple comparison test, P < 0.01) (Table 4F). Pond Zooplankton Copepod nauplii and adults (Diaptomus spp. and other calanoid species) and cladocerans (Bosmina, Scapholebris, Ceriodaphnia, and Daphnia) were abundant in the pond and in the enclosures (Fig. 5). The summed cladoceran and copepod densities declined rapidly in the pond from >1,000 L_1, when the larvae were stocked, to approximately 400 L_1 during the last 10 d of the experiment. Den- sities within the enclosures declined from approx- imately 1,000 L_1 at the time larvae were stocked to 100 L_1 when the experiments ended. Samples of 12-20 larvae of each type were ex- amined for stomach contents on day 30. The small- est larvae of each type had eaten cladocerans and copepods. The largest individuals had eaten chirono- mid larvae and zooplankton. Two of 20 individuals of striped bass and white bass hybrids had eaten fish larvae, proof that cannibalism was occurring. DISCUSSION Neither striped bass nor hybrid larvae, in the lab- 910 HOUDE and LUBBERS: SURVIVAL AND GROWTH OF STRIPED BASS Table 4.— Summary of data and analyses from 2 m3 enclosure experiments in the Manning Hatchery pond, 1983. Three replicate enclosures were run for each type of larva: Striped bass (SB), striped bass x white bass (WBX), and striped bass x white perch (WPX). A) Mean standard lengths and wet weights at 9 d after hatching, prior to stocking in enclosures. B) Percent survivals at 30 d after hatch- ing. C) Mean lengths and weights at 30 d after hatching. D) Growth-in-length equations (/, = standard length in mm at age f; f = days after hatching; Sb = standard error of the regression coefficient; r2 = coefficient of determination). E) Exponential, growth-in- weight equations (Wt = wet weight in mg at age f ; f = days after hatching; SG = standard error of the exponential coefficient; r2 = coefficient of determination). F) Power function equations of the wet weight-standard length relationships (W = wet weight in mg; / = standard length in mm; Sb = standard error of the power coefficient; r2 = coefficient of determination). A. Type of larva n Standard length (mm) Wet weight (mg) x s5 D. Type of larva Equation I, = a + bt n sb X s~x r2 SB 13 6.12 0.06 1.15 0.08 SB 1, = 3.09 + 0.30f 253 0.0136 0.66 WBX 13 6.19 0.14 11.66 0.11 WBX It = 3.22 + 0.31 f 245 0.0130 0.70 WPX 15 5.87 0.13 1.33 0.11 WPX k = 2.96 + 0.32f 263 0.0100 0.79 B. Type of larva SB Percent survival x s* 22.1 3.4 Instantaneous mortality rate (d~1) 0.0713 E. Type of larva Equation Wt = WQ eGt n sG r2 SB Wt = 0.37 e° 1444' 253 0.0044 0.81 WBX 22.0 6.2 0.0695 WBX 2W, = 0.26 e° 1650' 245 0.0041 0.87 WPX 28.5 2.1 0.0601 WPX 2Wt = 0.30 e° 1593r 263 0.0037 0.88 C. Type of larva n Standard length (mm) Wet weight (mg) x s-x F. Type of larva Equation W = alb n sb X s* r2 SB 78 12.58 0.24 38.38 3.57 SB w = 6.23 x in-4/.42879 253 0.0469 0.97 WBX 78 12.90 0.23 43.28 3.86 WBX w = 2.27 x 10"4L47114 245 0.0536 0.97 WPX 78 12.96 0.12 39.73 1.23 WPX w = 5.44 x 1(T4L43496 263 0.0512 0.97 'Significant at P < 0.05. ANOVA. zThe WBX and WPX exponential coefficients differed significantly, P < 0.05, from the SB coefficient. ANCOVA followed by SNK multiple comparison procedure. 3The WBX power coefficient differed significantly, P < 0.05, from the SB and WPX coefficients. ANCOVA followed by SNK multiple comparison procedure. oratory and in freshwater pond enclosures, demon- strated clearly superior growth or survival. The apparent heterosis in young-of-the-year and sub- adult white bass hybrids (Logan 1968; Ware 1975; Bonn et al. 1976; Williams et al. 1981; Kerby et al. 1983) was not evident during the first month after hatching. Survival and growth rates of the three types of larvae were relatively high in all of our ex- periments, indicating that striped bass and its hybrids may have near-equal production potential up to 30 d of age. Larvae grew and survived surprisingly well at the relatively low food concentrations that we offered in the laboratory. There was evidence that striped bass and white perch hybrid larvae grew faster at the 500 L_1 than at the 100 L_1 Artemia concen- tration but there was no significant difference in size of white bass hybrid larvae reared at those two food levels. Survival of all three types of larvae did not differ between the two food levels, demonstrating that high survival and favorable growth can be ob- Figure 2.— Mean wet weights ± 2 standard errors of striped bass (SB), striped bass x white bass (WBX), and striped bass x white perch (WPX) larvae from 6 to 30 d after hatching, reared at two food levels in the laboratory. E J? a> "3 5 Days after Hatching 911 FISHERY BULLETIN: VOL. 84, NO. 4 14 12 E £ c o 10 CO ■a | 8 W 50 © SB a WBX A WPX I ±2 SE 9 13 17 20 23 Days after Hatching 27 30 Figure 3.— Mean standard lengths (±2 standard errors) of striped bass (SB), striped bass x white bass (WBX), and striped bass x white perch (WPX) larvae on seven dates in 2 m3 enclosure ex- periments, Manning Hatchery pond. tained for Morone larvae at low Artemia concentra- tions, if those concentrations are maintained at the nominal levels. Our laboratory survival rates at low food levels were higher than those reported for striped bass larvae in the literature (e.g., Doroshev 1970; Miller 1978; Rogers and Westin 1981; Eld- ridge et al. 1981, 1982), which generally had in- dicated that nominal Artemia concentrations nearly an order of magnitude higher than 500 L_1 were required to obtain high survival rates. The laboratory and pond enclosure methods to assess striped bass and hybrid larvae performance differed in many respects and could have influenced results. Besides great differences in enclosed vol- umes (36 L vs. 2 m3), environmental factors and foods differed. Laboratory experiments were run at 19.0 °C and 3%o salinity, because low salinities are known to improve striped bass larvae survival (Bonn et al. 1976; Kerby et al. 1983). Temperature in- creased from 18.5° to 22.5°C in the Manning Hatch- ery freshwater pond during the 3-wk experiment. The laboratory-reared fish were fed only Artemia nauplii at controlled concentrations while enclosure fish had a variable zooplankton diet. 40 ^ 30 E x: O) "3 20 10 © SB a WBX a WPX I ±2 SE 9 13 17 20 23 Days after Hatching 27 30 Figure 4.— Mean wet weights (±2 standard errors) of striped bass (SB), striped bass x white bass (WBX), and striped bass x white perch (WPX) larvae on seven dates in 2 m3 enclosure experiments, Manning Hatchery pond. Survival of all larvae was lower in the pond en- closures than in the laboratory tanks (Tables 1, 4). White bass hybrids had the lowest mean survival rate in the laboratory but they survived as well as striped bass and white perch hybrids in the pond enclosures. At the relatively high 500 L_1 Artemia level laboratory-reared striped bass larvae were longer and heavier than either of the hybrids at 30 d after hatching. In the pond enclosures no signifi- cant differences in mean lengths or weights among the three types of larvae were detected at 30 d. The weight-length relationship of pond enclosure, white bass hybrid larvae had a relatively high exponen- tial coefficient, indicating that they were heavier at a given length than the other types of larvae. Mean weights of both hybrids at 30 d were considerably heavier in the pond enclosures than in the labora- tory tanks. Relatively great size variability in the 912 HOUDE and LUBBERS: SURVIVAL AND GROWTH OF STRIPED BASS Zooplankton Densities at Cedarville Inside Enclosures 1000 17 20 23 Days after Hatching Figure 5.— Mean densities of copepods and cladocerans inside and outside of the 2 m3 enclosures used for striped bass and hybrid larvae experiments in the Manning Hatchery pond. enclosures may have resulted in part from canni- balism and consumption of chironomids by some lar- vae, promoting their relatively rapid growth. Although mean weights and lengths of 30-d-old striped bass and the two hybrids from the pond enclosures did not differ, the instantaneous rate of growth in weight of striped bass larvae was signifi- cantly lower than that of either hybrid (Table 4E). Had the enclosure experiments proceeded for a few more days the hybrids would have attained larger size than the striped bass. For example, at 35 d after hatching the striped bass would have weighed 20 mg less than either hybrid. The heterotic effect may begin to express itself at approximately 1 mo of age. Alternatively, the freshwater environment, increas- ing temperatures, and the prey available in the pond may have selectively favored growth of hybrids dur- ing the last few days of the experiment. If 30-d-old fry are to be produced for stocking, there is no apparent immediate advantage to rear hybrids rather than striped bass. Our laboratory and pond enclosure studies did not demonstrate advan- tages in survival or production of hybrids. The pond enclosure results did suggest that hybrids may begin to achieve an advantage in growth rates just prior to 1 mo of age. Important questions about compar- ative energetics, nutrition, and genetics still remain to be answered to understand the biology of larval M. saxatilis or its hybrids and the consequences of their possible release into natural systems such as Chesapeake Bay. ACKNOWLEDGMENTS The research was supported by contracts F26-82- 003 and F31-83-008 from the Maryland Department of Natural Resources, Tidewater Administration, Tidal Fisheries Division. Assistance in the labora- tory was provided by H. Hornick, W. Roosenburg, and V. Saksena. Larvae were supplied by the Mary- land DNR Manning Hatchery at Cedarville. The assistance of DNR personnel, particularly M. Beaven, H. King, and J. Stringer, is gratefully acknowledged. We appreciate the critiques of early 913 drafts of the manuscript provided by E. J. Chesney, H. King, J. Kraeuter, and L. C. Woods III. LITERATURE CITED Bayless, J. D. 1972. Artificial propagation and hybridization of striped bass, Morone saxatilis (Walbaum). South Carolina Wildl. Mar. Resour. Dep., 135 p. Bonn, E. W., W. M. Bailey, J. C. Bayless, K. E. Erickson, and R. E. Stevens. 1976. Guidelines for striped bass culture. Southern Div., Am. Fish. Soc, 103 p. plus app. BOREMAN, J., AND H. M. AUSTIN. 1985. Production and harvest of anadromous striped bass stocks along the Atlantic coast. Trans. Am. Fish. Soc. 114:3-7. Cooper, J. C, and T. T. Polgar. 1981. Recognition of year-class dominance in striped bass management. Trans. Am. Fish. Soc. 110:180-187. Doroshev, S. I. 1970. Biological features of the eggs, larvae and young of the striped bass [Roccus saxatilis (Walbaum)] in connection with the problem of its acclimatization in the USSR. J. Ichthyol. 10:235-248. Eldridge, M. B., J. A. Whipple, and M. J. Bowers. 1982. Bioenergetics and growth of striped bass, Morons sax- atilis, embryos and larvae. Fish. Bull., U.S. 80:461-474. Eldridge, M. B., J. A. Whipple, D. Eng, M. J. Bowers, and B. M. Jarvis. 1981. Effects of food and feeding factors on laboratory- reared striped bass larvae. Trans. Am. Fish. Soc. 110: 111-120. FISHERY BULLETIN: VOL. 84, NO. 4 Kerby, J. H., and E. B. Joseph. 1979. Growth and survival of striped bass and striped bass x white perch hybrids. Proc. Annu. Conf. Southeastern Assoc. Fish Wildl. Agencies 32:715-726. Kerby, J. H., L. C. Woods III, and M. T. Huish. 1983. Culture of the striped bass and its hybrids: A review of methods, advances and problems. In R.F. Stickney and S. P. Meyers (editors), Proceedings of the Warmwater Fish Culture Workshop, p. 23-54. World Maricult. Soc, Spec. Publ. 3. Logan, H. J. 1968. Comparison of growth and survival rates of striped bass and striped bass x white bass hybrids under controlled en- vironments. Proc. Annu. Conf. Southeast Assoc. Fish Game Comm. 21:260-263. Miller, P. E. 1978. Food habit study of striped bass post yolk-sac larvae. Johns Hopkins Univ., Chesapeake Bay Inst., Spec. Rep. 68, 49 p. Rogers, B. A., and D. T. Westin. 1981. Laboratory studies on effects of temperature and delayed initial feeding on development of striped bass larvae. Trans. Am. Fish. Soc. 110:100-110. Steel, R. G. D., and J. H. Torrie. 1960. Principles and procedures of statistics. McGraw-Hill Book Co., Inc., N.Y., 481 p. Ware, F. J. 1975. Progress with Morone hybrids in fresh water. Proc. Annu. Conf. Southeast Assoc. Fish Game Comm. 28:48- 54. Williams, J. E., P. A. Sandifer, and J. M. Lindbergh. 1981. Net-pen culture of striped bass x white bass hybrids in estuarine waters of South Carolina: a pilot study. J. World Maricult. Soc. 12:98-110. 914 ASPECTS OF THE REPRODUCTIVE BIOLOGY, SPATIAL DISTRIBUTION, GROWTH, AND MORTALITY OF THE DEEPWATER CARIDEAN SHRIMP, HETEROCARPUS LAEVIGATUS, IN HAWAII Murray D. Dailey1 and Stephen Ralston2 ABSTRACT The recent rapid development of fisheries for the Heterocarpus laevigatas in Hawaii and elsewhere in the tropical Pacific has created the need for biological information to manage the resource. This study reports on a 16-month sampling program of commercial shrimp catches in Hawaii, during which the depth of capture, carapace length (CL), sex, and reproductive condition of 7,368 H. laevigatas were determined. The overall sex ratio of H. laevigatus was 1:1.16 in favor of females and depended on the depth sampled; there were relatively fewer females as depth increased. Seasonal variation in sex ratio was evident which may have been due to changing catchability and availability or a sex related dispersion pattern. Sex ratio also depended on size category, displaying a standard pattern with no evidence of protandry. Females mature at 40 mm CL (64% of asymptotic length) and ovigerous individuals are found year round. However, the main reproductive season is from August-February, with over 50% of females carrying eggs from October-January. Mature shrimp may undergo a depth related seasonal migration in synchrony with breeding. Mature males and females were found deeper (700 m) during the reproduc- tive season than not (550 m). Females apparently settle in deep water and migrate gradually to shallower water as they grow. Seasonal length-frequency data suggest H. laevigatus is not semelparous. Separate analyses of CL- frequency distributions of male and female shrimp indicate their von Bertalanffy asymptotic sizes are 57.9 and 62.5 mm CL, respectively. Growth coefficients (K) estimated by modal progression were 0.35 and 0.25 per year for males and females, and total instantaneous mortality rates were 1.51 and 0.73 per year, respectively. The deepwater caridean shrimp, also known as "ono" or smooth nylon shrimp, Heterocarpus laevi- gatus, (Family Pandalidae) occurs throughout the tropical Pacific Ocean, where it is found in benthic deepwater habitats (450-900 m) (Wilder 1977; King 1983). While early trapping surveys in the Hawaiian Islands revealed its local abundance (Clarke 1972; Struhsaker and Aasted 1974), little information was available concerning its biology. These early studies did show, however, that H. laevigatus was poten- tially of commercial importance, with a preliminary maximum sustained yield estimate of 454-907 metric tons (t) derived for the Hawaiian Archipelago (Department of Land and Natural Resources 1979). More recently the Western Pacific Regional Fishery Management Council3 (WPRFMC) has revised this estimate to 400-4,000 t. ■Hawaiian Fishing Research Company, 737 Bishop Street, Suite 2910, Honolulu, HI 96813; present address: Department of Biology, California State University at Long Beach, Long Beach, CA 90840. 2Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI 96822-2396. Manuscript accepted July 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. A commercial trap fishery for this species sub- sequently developed in the Hawaiian Islands, and in 1984 the WPRFMC began the process of develop- ing a fishery management plan for the Heterocar- pus shrimp resources of the region. Landings from the Hawaiian fishery exceeded 135 t in 1983 but have declined sharply since, although commercial interest in the resource remains great (WPRFMC fn. 3). Recent research surveys in Hawaii have now more clearly defined this species' depth, temporal, and geographic distributions (Oishi 1983; Hawaiian Divers 19834; Gooding 1984), although the life history of H. laevigatus remains largely unknown. The only substantive biological studies to date were 3Western Pacific Regional Fishery Management Council. 1984. Status of fisheries assessment of development and management needs for selected crustacean species in the western Pacific region. Unpubl. manuscr., 60 p. Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI 96822-2396. 4Hawaiian Divers. 1983. Deepwater shrimp utilization study for Hawaii. Report prepared under NOAA Cooperative Agree- ment No. 80-ABH-00065 for the Southwest Region, Western Pacific Program Office, National Marine Fisheries Service, NOAA, Honolulu, HI, 47 p. 915 FISHERY BULLETIN: VOL. 84, NO. 4 completed in the Marianas (including Guam) and Fiji (Wilder 1977; King 1983; King and Butler 1985; Moffitt and Polovina5). Evidence also exists to show that this species is highly susceptible to trapping (Ralston 1986) and, according to commercial fishermen, depletion of the resource has occurred over certain fishing grounds in Hawaii (S. Barrows6). Because estimates of the shrimp's productive capacity which are currently available are preliminary at best and a fishery has developed rapidly, this study set out to examine aspects of the life history of the Hawaiian stock of H. laevigatus to obtain information useful in devel- oping a basis for management of the fishery. METHODS All sampling was conducted by commercial fishing vessels owned by the Hawaiian Shrimp Company (Easy Rider, Mokihana, and the Easy Rider Too) over the 16-mo period from August 1983 to Novem- ber 1984. During this time, six 35-60 d cruises were completed and samples were obtained during 9 of the 12 calendar months (Table 1). Fishing was con- ducted throughout much of the Hawaiian Archi- pelago, from Gardner Pinnacles south to the Island of Hawaii (Fig. 1). Samples were collected at all of the seven main islands (Hawaii, Kauai, Lanai, Maui, Molokai, Niihau, and Oahu) and from Necker, French Frigate Shoals, and Gardner Pinnacles in the Northwestern Hawaiian Islands. All shrimp were caught during overnight sets of baited pyramidal traps, which measured 1.5 x 1.8 m with a funnel opening at the top center. Fishing was targeted between depths of 500 and 700 m, 5Moffitt, R. B., and J. J. Polovina. The distribution and yield assessment of the deepwater shrimp resource in the Marianas. Manuscr. in prep. Southwest Fisheries Center Honolulu Labora- tory, National Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI 96822-2396. 6S. Barrows, Hawaiian Shrimp Company, 737 Bishop Street, Suite 2910, Honolulu, HI 96813, pers. commun. 1985. Table 1 .—Temporal and geographic distribution of Heterocarpus laevigatus samples (FFS = French Frigate Shoals). Year Month Location Sample size 1983 Aug. Oahu 79 1983 Sept. Oahu 26 1983 Oct. FFS 188 1983 Nov. FFS 1,942 1984 Jan. Oahu 285 1984 Mar. Niihau, Kauai 530 1984 April Hawaii 631 1984 May Lanai, Maui, Molokai 1,389 1984 June Necker 842 1984 Sept. Gardner Pinnacles, FFS, Necker 1,438 1984 Nov. Necker 18 although some catches were made in both shallower and deeper water because of the trap drift. The best catch rates were found in areas of hard rough bot- tom; otherwise, all sampling sites were to all ap- pearances similar. Systematic subsamples of the catch were taken from every other trap on every second fishing day by randomly scooping approximately 0.9 kg (2 lb) of shrimp from traps prior to emptying. Samples were placed in double bags with tags recording date, location, depth, and condition, and were then frozen and packed for transfer to the laboratory. There all shrimp were identified to species; sexed; examined for embryos on the pleopods; measured to the near- est 0.1 mm for carapace length (CL), carapace width (CW), and total length (TL); and weighed to the nearest 0.1 g on a top loading scale. The data were then keypunched and stored for analysis. Size-frequency distributions of H. laevigatus were analyzed by the regression method of Wetherall et al. (in press) to estimate maximum size (Lm of the von Bertalanffy growth equation) and the ratio of total instantaneous mortality rate (Z) to von Berta- lanffy growth coefficient (K). Additionally, the growth coefficient of H. laevigatus was estimated by following the progression of size modes evident in three large samples taken: 1) 24 October to 6 November 1983, 2) 24 April to 11 May 1984, and 3) 3 September to 6 November 1984. Sample sizes of N = 2,021, 1,991, and 1,438 were obtained in these respective samples, accounting for 74% of all shrimp measured in the study. Modal progression of size distributions was determined by the ELEFAN I computer program of Pauly (1982). RESULTS A total of 7,368 H. laevigatus were measured and examined for CL, sex, and the presence of eggs (Table 1). Of these 3,956 were females (32.6% of which were ovigerous) and 3,412 were males. This corresponds to an overall male to female sex ratio of 1:1.16, departing significantly from equality (P < 0.0001). Measurements of TL, CW, and weight were obtained from 5,920 of the shrimp sampled. Due to an imbalance in sampling, the effects of location and time on the distribution of H. laevigatus could not be completely separated. We therefore assume that all samples were drawn from statis- tically homogeneous locations in order to isolate and examine temporal and depth effects. The strength of this assumption is based largely on our personal observations and those of fishermen that seasonal change seems to account for most major population 916 DAILEY and RALSTON: BIOLOGICAL DEVELOPMENT OF HETEROCARPUS LAEVIGATUS 1 r i 1 3 & / 1 A. / | .i A * 1 / fir J 1 If-' a W 1 U ( 0 s / Sri <> i J, I 1 tf) c '.•/ 1 C? 3 c £ £ _„■£ e P 1 - W I E i ■:>.:■ 1 E $ c o -*J n! .c +-> T3 a c o c o 2 J3 -a C3 a) s-, cS T3 a> '6 s I- 43 03 O* CD ■^ 3 30% of females). In particular over 50% of all sampled females carried eggs from October to January. Relatively few shrimp were caught with eggs dur- ing the period from April to July (<10%). Moreover, Table 2. — Parameter estimates of functional regressions on linear size measurements. All measurements in millimeters and all sample sizes n = 5,920. Dependent Independent Inter- Correlation variable variable Slope cept coefficient Total length Carapace length 2.864 10.182 0.963 Carapace width Carapace length 0.613 -5.562 0.964 Carapace length Total length 0.349 -3.536 0.963 Carapace length Carapace width 1.630 9.098 0.964 Carapace width Total length 0.214 - 7.737 0.902 Total length Carapace width 4.673 36.153 0.902 Table 3. — Functional and predictive length-weight regressions for Heterocarpus laevigatus. The natural logarithm of weight in grams is fitted to the natural logarithm of carapace length in mm. The standard errors of the slope (b) and intercept (a) are given by Sb and Sa respectively. Slope Intercept Sb Sa n r2 Males Females without eggs Females with eggs Predictive Functional Predictive Functional Predictive Functional 2.755 2.910 2.605 2.745 1.815 2.470 - 6.809 -7.358 -6.252 -6.757 -2.986 -5.498 0.0176 0.0629 2,788 0.8976 0.0185 0.0671 2,202 0.8999 0.0550 0.2114 928 0.5401 918 DAILEY and RALSTON: BIOLOGICAL DEVELOPMENT OF HETEROCARPUS LAEVIGATUS 01 o> D/J F/M A/M J/J Time of Year A/S 0/N Figure 2.— Seasonal incidence of ovigerous Heterocarpus laevigatus females in the Hawaiian Islands. Vertical bars represent 95% confidence intervals and sample sizes are presented above. Site locations vary. when the analysis was restricted to mature females only (see next section) the seasonal pattern of egg- bearing was unchanged. From these results we conclude that in Hawaiian waters H. laevigatus reproduces during the fall and winter seasons (August-February). The size at maturity of female shrimp was deter- mined by aggregating the female data into 5 mm CL classes and plotting the incidence of ovigerous females against CL class (Fig. 3). Only samples ob- tained during the reproductive season were included in the analysis. As before, the overall percentage 100 Carapace Length - mm Figure 3.— Size at maturity for female Heterocarpus laevigatus sampled during the reproductive season. Vertical bars represent 95% confidence intervals and sample sizes are presented above. 919 FISHERY BULLETIN: VOL. 84, NO. 4 with 95% confidence limits and sample sizes are provided. The data show that the 55-60 mm CL class en- compassed the largest shrimp observed. Virtually all (95%) females >50 mm CL that were sampled during August-February bore eggs. Conversely, up to 35 mm CL no more than 1% of the shrimp ex- amined were ovigerous. The figure shows further that at a CL of 40 mm the percentage of ovigerous females is one-half its maximum value, with 48% of all sampled females bearing eggs. We conclude that females become sexually mature at this size (Gunderson et al. 1980). We have no data on matura- tion in males. The data presented in Figure 4 show the sex ratio of shrimp as it depends on size (CL mm). Plotted are the percentage females, with 95% confidence intervals and sample sizes, against 5 mm CL size classes. The data clearly show that H. laevigatus maintains a relatively uniform sex ratio from 10 to 45 mm CL, but that females predominate in the largest length categories (45-65 mm CL). Because some studies (Clarke 1972; Wilder 1977) have indicated that Heterocarpus females may experience mass mortality after egg bearing, we ex- amined the relationship of sex ratio to season (Table 4). Presented for each 2-mo sampling period are the number of females and total number of shrimp sampled, the proportion which are female, and the standard error of the proportion. The results show that an unusually high fraction (0.72) of the shrimp sampled during the peak of the reproductive season (December-January) are female. Note that the in- cidence of females in trap samples declines signifi- cantly to a value of 0.45 in April-May as the breeding season wanes. At first inspection these data support the contention that females experience increased mortality after bearing eggs, i.e., that//, laevigatus may be semelparous. Table 4.— Sex ratio of Heterocarpus laevigatus by month sampled. The standard error of the proportion is given by Sp . Number Proportion Month of females of females n sP December-January 207 0.72 285 0.026 February-March 293 0.55 530 0.022 April-May 923 0.45 2,020 0.011 June-July 459 0.54 842 0.017 August-September 857 0.55 1,542 0.013 October-November 1,217 0.56 2,148 0.011 Spatial Distribution The relationship between the sex ratio of H. laevi- gatus and sampling depth is provided in Table 5. These results demonstrate that the relative abun- dance of the two sexes is not independent of depth (X2 = 165.6, df = 16, P < 0.001). As depth in- creases (440-760 m) there is a significant decline in the percentage of females in our samples (P = 0.05). 100 Carapace Length - mm Figure 4.— Sex ratio as a function of carapace length. Vertical bars represent 95% confidence intervals and sample sizes are presented above. 920 DAILEY and RALSTON: BIOLOGICAL DEVELOPMENT OF HETEROCARPUS LAEVIGATUS Table 5.— Sex ratio and size of Heterocarpus laevigatas by depth (M is for males, F0 for females without eggs, and FE for females with eggs). Sp is the standard error of the proportion. Depth (m) Number of females N Proportion of females sP Carapace length All M F0 FE 440 41 71 0.58 0.059 40 34 43 48 60 9 11 0.81 0.116 39 36 40 — 80 65 141 0.46 0.042 38 33 42 49 500 155 230 0.67 0.031 42 34 45 49 20 287 430 0.66 0.023 43 40 45 46 40 419 943 0.44 0.016 37 35 40 49 60 193 487 0.39 0.022 37 36 38 45 80 280 500 0.56 0.022 39 35 42 47 600 282 498 0.56 0.022 38 35 38 . 46 20 284 463 0.61 0.023 39 35 34 48 40 325 581 0.55 0.021 38 37 32 47 60 109 210 0.51 0.034 37 34 35 44 80 95 167 0.56 0.038 39 37 35 48 700 90 222 0.40 0.033 36 36 31 43 20 399 738 0.54 0.018 39 38 32 47 40 68 126 0.53 0.044 40 37 33 48 60 14 46 0.30 0.068 34 34 28 42 The results presented in Table 5 also show the distribution of mean size (CL mm) by depth (m) for all H. laevigatus caught, and for the M, F0, and FE subgroups. For all shrimp combined, average size decreases slightly with increasing depth fished. The trend for decrease in size with increasing depth is not evident in the M subgroup. However, the F0 class demonstrates a strong relationship of decreas- ing mean CL with depth. For the FE category the decline is much less apparent, if at all. Thus the overall decline in mean CL of all shrimp combined, is clearly due to an overriding influence of females without eggs. We interpret these trends, or lack thereof, to indicate that young (i.e., small) females may move from deep to shallow water as they mature. There is some evidence that the depth distribu- tion of H. laevigatus changes with reproductive activity (i.e, season). Figure 5 presents the depth distributions for reproductively competent (>40 mm CL) male and female shrimp, classified into samples taken outside (March- July) and during the reproduc- tive season (August-February). Note that depth distributions of both male and female shrimp are 60 40- 20- > - 40 « cc 20- M ALE S > 40 mm CL 400 FEMALES > 40 mm CL 800 Depth Figure 5.— Seasonal distributions of large (>40 mm) male and female Heterocarpus laevigatus by depth. The dashed line represents the spawning season distribution (May- February) and the solid line represents the distribution during the nonspawning season (March-July). 921 FISHERY BULLETIN: VOL. 84, NO. 4 shifted 150 m deeper when the females are ovig- erous. Although the data are not corrected for what may have been differences in fishing effort by depth, it is true that fishing was targeted to depths of max- imum shrimp abundance. Based on these findings, and the results presented in Table 5, our data are consistent with a hypothesis of gradual movement of small females from deep to shallow water, with mature shrimp moving between depths of 550 and 700 m in synchrony with the ovigerous cycle of females. Growth and Mortality Clarke (1972) and King (1983) have suggested that Heterocarpus spp. may breed once and die. Indeed the results already presented in Table 4 may be con- sidered consistent with the hypothesis that at least female H. laevigatus are semelparous. To further address this question we examined the size struc- ture of male and female shrimp classified as follows: 1) during the latter half of the reproductive season (January-February) and 2) immediately following the reproductive season (March-July). If postrepro- ductive mortality of shrimp was severe, a decrease in the relative abundance of large, breeding adults would be expected as the reproductive season waned. The results presented in Figure 6 conflict with this expectation, where it is apparent that the propor- tional representation of large reproductive in- dividuals (>40 mm CL) is actually greater imme- 20 15 10 - & 15- 10- 5- FEMALES MALES I 10 20 30 40 Carapace Length -mm 60 Figure 6.— Relative size-frequency distributions of male and female Heterocarpus laevigatus during the peak and postreproductive seasons. The solid line represents the peak season (January-February), males N = 78, females N = 207; the dashed line is based on data collected immediately after the peak season (March-July) males AT = 1,717, females N = 1,675. 922 DAILEY and RALSTON: BIOLOGICAL DEVELOPMENT OF HETEROCARPUS LAEVIGATUS diately following than during the latter half of the reproductive season. The total sample CL-frequency distribution of males and females combined was analyzed by the regression method of Wetherall et al. (in press) to estimate Lm and Z/K. When all shrimp are pooled (N = 7,368), an estimate of Lx = 61.7 mm CL results. Further, the ratio of total mortality rate to von Bertalanffy growth coefficient (Z/K) is esti- mated to be 2.6. Calculations were repeated for separate male and female subgroups, where it was found that L = 57.9 and 62.5 mm CL and ZIK = 00 4.3 and 2.9 for males and females, respectively. These results indicate that males generally grow to a smaller size than females. The results of analyzing the progression of CL size modes in frequency distributions of male and female H. laevigatus provided preliminary estimates of K = 0.35 yr-1 for males and 0.25 yr-1 for females. The former result must be viewed with caution, however, because two "solutions" were detected by the computer search algorithm (Pauly 1982) which differed little in fit. One of these, K = 0.70 yr-1, we believe to be unjustifiably high in light of the minor difference (8%) between the L^ of males and the L^ of females obtained from the regression analysis. Note that estimates of K and Lk typical- ly show a strong inverse correlation (Gallucci and Quinn 1979). These results, in conjunction with the estimates of ZIK for male and female shrimp pre- sented earlier, provide the basis for preliminary estimation of total mortality rate. We estimate Z = 1.51 yr-1 for males and 0.73 yr-1 for females, corresponding to annual survivorship fractions of 22% and 48% per year, respectively. These data in- dicate that males grow faster while experiencing a substantially greater total mortality rate than females. DISCUSSION Earlier it was assumed that, aside from depth, all shrimp samples were drawn from locations which are dynamically homogeneous; i.e., the behavior of shrimp populations through time does not vary from site to site. This is clearly a restrictive and simpli- fying assumption and is without doubt the major limitation on the results presented here. Nonethe- less, it was a necessary simplification for us to analyze the commercial fishing data upon which this study was based. Consequently, we view those results which rely upon this assumption as tentative and in need for further validation. Examination of the seasonal trend in the relative abundance of ovigerous females showed that in Hawaii over 50% of H. laevigatus females bear eggs from October to January, with a peak between August and February. Wilder (1977) found a similar- ly timed but more narrowly defined breeding season for H. laevigatus in Guam, where the percentage of ovigerous females in trap catches reached a max- imum during December, but was not particularly high in any other month. Clarke (1972) reported that H. ensifer in Hawaii also reproduces in the winter. The breeding season of these shrimps is unusual among Hawaiian crustaceans and fishes, which typically reproduce during the spring and summer and uncommonly during the winter (Watson and Leis 1974; Lobel 1978; Uchida et al. 1980; Uchida and Tagami 1984; Walsh 1984). Our data also indicate that in Hawaii sexual maturity of female ono shrimp occurs at approx- imately 40 mm CL, a size similar to that reported by King (1983) for shrimp from Fiji, Vanuatu, West Samoa, and Tonga and by Moffitt and Polovina (fn. 5) for samples from the Marianas. Based upon the estimated parameters of the von Bertalanffy growth equation derived here, this corresponds to an age of first maturity of 4 yr. Although we have no data on the maturation of males, we believe they prob- ably mature earlier and at smaller size, perhaps at age 3 when they are 37-38 mm CL. Such a result is consistent with the findings of Moffitt and Polo- vina (fn. 5) who found that male H. laevigatus in the Marianas mature at a smaller size than do females. Wilder (1977) speculated that both H. ensifer and H. laevigatus in Guam are protandrous hermaphro- dites, as did Clarke (1972) for Hawaiian populations of H. ensifer. However, the results presented in King and Moffitt (1984) tend to contradict this con- clusion. These authors studied the morphometry and sexuality of five deep water pandalids, including//. laevigatus, in Fiji and the Marianas. Using the relative length of the appendix masculina expressed as a proportion of CL, they found no tendency toward protandrous hermaphroditism. Moreover, the sex ratio reported in their study was approx- imately 1:1. Our results also indicate that for Hawaiian popula- tions of H. laevigatus, and we speculate for most tropical pandalids, a sex transition does not occur. Wenner (1972) has termed the pattern exhibited in Figure 4 the standard sex ratio pattern, as distin- guished from one of reversal. Due to the large numbers of females in small size classes, these data are generally inconsistent with a protandric herma- phroditic life history, as has been hypothesized by previous workers on Heterocarpus spp. (Clarke 923 FISHERY BULLETIN: VOL. 84, NO. 4 1972; Wilder 1977). King and Moffitt (1984) also argue for dioecy in this species based upon relative changes in the morphology of the appendix masculina. Evidence now exists to suggest that the sex ratio of if. laevigatus undergoes a seasonal change (Table 4), although the reasons for this are at present unknown. A biological alteration in population struc- ture of this order seems unlikely. Rather, the rela- tively high catch of females during the December- January period may be due to seasonal changes in catchability or vulnerability of one or both sexes to the traps. Alternatively, the spatial dispersion of H. laevigatus may depend on sex. If males and females are spatially segregated, the high proportion of females in the December-January sample may have been due to small sample size (N = 207). We have also shown that sex ratio depends strong- ly on the depth sampled (Table 5), with diminishing representation of females as depth increases. This spatial heterogeneity between the sexes may be due to directed movements. Based on size trends of females we conclude that they recruit to deeper water and subsequently migrate to shallower water. We have no evidence for similar movement of males. Studies by King (1983) on Pacific Heterocarpus spp. showed cyclic migrations in these shrimps, sug- gesting that depth distribution may change season- ally, with an annual migration up and down the slope of the sea floor. The data presented in Figure 5 in- dicate that mature H. laevigatus in Hawaii do migrate seasonally, demonstrating distinct shifts in the depth distributions of both sexes during the reproductive season. Because this result is con- founded by what may be a location effect, however, we view them as preliminary and in need of further confirmation. King (1983) also reported that Hetero- carpus spp. were found in stomachs of tuna in Fiji, indicating perhaps some type of vertical migration in the water column. King (1985), based on work completed in Fiji, ex- amined the question of iteroparity and semelparity in several genera of pandalid shrimp (Plesionika, Saron, Parapandalus, and Heterocarpus). Based on the difference between length at sexual maturity and maximum length, he concluded that shallow- water species (e.g., H. ensifer) are semelparous. He states that deepwater Heterocarpus spp. "have an extended reproductive lifespan, the length of which may be taken to indicate the number of spawnings." We conclude, based on the relative size-frequency distributions of males and females during peak and postreproductive seasons, that both sexes survive well after reproducing— evidence in favor of itero- parity. Although a high mortality of shrimp follow- ing the breeding season would be evidence consis- tent with a semelparous life history, it is not a sufficient result to prove it. This is because each female, before dying, could have sequential multi- ple clutches during the October-February ovigerous period. Nonetheless, good survival oiH. laevigatus females after carrying eggs (Fig. 6) is indicative of iteroparous reproduction. The regression technique of Wetherall et al. (in press) produced estimates of the ratio of mortality to growth coefficient of 2.9 and 4.3 for females and males respectively. Moffitt and Polovina (fn. 5), using similar methods, estimated Lm = 55.2 mm CL and ZIK = 2.5 for combined male and female samples of H. laevigatus from essentially unfished stocks in Guam and the Marianas. Ralston (1986) also reported that the ZIK ratio of an unexploited population of H. laevigatus at Alamagan in the Marianas was about 2.0. The differences between estimates may therefore relate to differences in levels of exploitation. Moreover, the higher mortal- ity rate of male shrimp when compared with females (1.51 versus 0.73 yr_1) may explain the somewhat biased sex ratio in favor of females. ACKNOWLEDGMENTS We acknowledge the help of the fishing vessels Easy Rider, Easy Rider Too, and Mokihana. Special thanks go to Jack Klein of the Mokihana crew for his collection of material. Also, we thank Robert Richlynski and Patricia M. Van Nuis for their technical help during the collection of data. LITERATURE CITED Clarke, T. A. 1972. Exploration for deep benthic fish and crustacean resources in Hawaii. Univ. Hawaii, Hawaii Inst. Mar. Biol. Tech. Rep. 29:1-18. BMDP. 1977. BMDP Biomedical Computer Programs, P-Series. Univ. Calif. Press, Los Ang., 880 p. Department of Land and Natural Resources. 1979. Hawaii fisheries development plan. Department of Land and Natural Resources, State of Hawaii, 297 p. Gallucci, V. F., and T. J. Quinn II. 1979. Reparameterizing, fitting, and testing a simple growth model. Trans. Am. Fish. Soc. 108:14-25. Gooding, R. M. 1984. Trapping surveys for the deepwater caridean shrimps, Heterocarpus laevigatus and H. ensifer, in the Northwestern Hawaiian Islands. Mar. Fish. Rev. 46(2): 18-26. Gunderson, D. R., P. Callahan, and B. Goiney. 1980. Maturation and fecundity of four species of Sebastes. Mar. Fish. Rev. 42(3-4):74-79. 924 DAILEY and RALSTON: BIOLOGICAL DEVELOPMENT OF HETEROCARPUS LAEVIGATUS King, M. G. 1983. The ecology of deepwater caridean shrimps (Crustacea: Decapoda: Caridea) near tropical Pacific Islands with par- ticular emphasis on the relationship of life history patterns to depth. Ph.D. Thesis, Univ. South Pacific, Suva, Fiji, 258 P- King, M. G., and A. J. Butler. 1985. Relationship of life-history patterns to depth in deep- water caridean shrimps (Crustacea: Natantia). Mar. Biol. (Berl.) 86:129-138. King, M. G., and R. B. Moffitt. 1984. The sexuality of tropical deepwater shrimps (Decapoda: Pandalidae). J. Crustacean Biol. 4:567-571. Lobel, P. S. 1978. Diel, lunar, and seasonal periodicity in the reproduc- tive behavior of the pomacanthid fish, Centropyge potteri, and some other reef fishes in Hawaii. Pac. Sci. 32:193-207. OlSHl, F. 1983. Pacific Tuna Development Foundation - State of Hawaii shrimp industry development project. State Hawaii, Dep. Land Nat. Resour., Div. Aquat. Resour., 22 p. Pauly, D. 1982. Studying single-species dynamics in a tropical multi- species context. In D. Pauly and G. I. Murphy (editors), Theory and management of tropical fisheries. ICLARM and CSIRO, Manila, 360 p. Richer, W. E. 1973. Linear regressions in fishery research. J. Fish. Res. Board Can. 30:409-434. Struhsaker, P., and D. C. Aasted. 1974. Deepwater shrimp trapping in the Hawaiian Islands. Mar. Fish. Rev. 36(10):24-30. Uchida, R. N., and D. T. Tagami. 1984. Biology, distribution, population structure, and pre- exploitation abundance of spiny lobster, Panulirus mar- ginatus (Quoy and Gaimard), in the Northwestern Hawaiian Islands. In R. W. Grigg and K. T. Tanoue (editors), Pro- ceedings of the Symposium on Resource Investigations in the Northwestern Hawaiian Islands, Vol. 1, April 24-25, 1980, University of Hawaii, Honolulu, Hawaii, p. 157-198. UNIHI-SEAGRANT-MR-84-01. Uchida, R. N., J. H. Uchiyama, D. T. Tagami, and P. M. Shiota. 1980. Biology, distribution, and estimates of apparent abun- dance of the spiny lobster, Panulirus marginatus (Quoy and Gaimard), in waters of the Northwestern Hawaiian Islands: Part II. Size distribution, legal to sublegal ratio, sex ratio, reproductive cycle, and morphometric characteristics. In R. W. Grigg and R. T. Pfund (editors), Proceedings of the Symposium on Status of Resource Investigations in the Northwestern Hawaiian Islands, April 24-25, 1980, Univer- sity of Hawaii, Honolulu, Hawaii, p. 131-142. UNIHI- SEAGRANT-MR-80-04. Walsh, W. J., III. 1984. Aspects of nocturnal shelter, habitat space, and juvenile recruitment in Hawaiian coral reef fishes. Ph.D. Thesis, Univ. Hawaii, Honolulu, 475 p. Watson, W., and J. M. Leis. 1974. Ichthyoplankton of Kaneohe Bay, Hawaii. Univ. Hawaii Sea Grant Publ. UNIHI-SEAGRANT-TR-01, 178 p. Wenner, A. M. 1972. Sex ratio as a function of size in marine Crustacea. Am. Nat. 106-321-350. WETHERALL, J. A., J. J. POLOVINA, AND S. RALSTON. In press. Estimating growth and mortality in steady state fish stocks from length-frequency data. ICLARM/KISR Conf. Theory and Application of Length-Based Methods for Stock Assessment. Wilder, M. N. 1977. Biological aspects and the fisheries potential of two deepwater shrimps, Heterocarpus ensifer and H. laevigatus in waters surrounding Guam. M.S. Thesis, Univ. Guam, Agana, Guam, 79 p. 925 AN INTENSIVE FISHING EXPERIMENT FOR THE CARIDEAN SHRIMP, HETEROCARPUS LAEVIGATUS, AT ALAMAGAN ISLAND IN THE MARIANA ARCHIPELAGO Stephen Ralston1 ABSTRACT During January 1984 an intensive fishing experiment for the deepwater caridean shrimp, Heterocarpus laevigatus, was conducted near Alamagan Island in the Mariana Archipelago. Twenty standard shrimp traps were set daily, producing a significant decline in the average catch rate from 3.33 to 1.82 kg/trap- night over a 16-day period. This drop was associated with a removal of 776 kg of shrimp from the study site. Resampling the area 4 months later showed that the catch rate remained depressed. Length-frequency data demonstrate that the decrease in catch per unit effort was due to a decline in the number of shrimp caught. An initial population size of 1,714 kg from 312 ha habitat is estimated, corresponding to one exploitable shrimp per 51 m2. The estimate of catchability (0.001945 trap-night"1) indicates that H. laevigatus may be easily overfished by trapping. Intensive fishing experiments can provide the ideal complement to resource surveys using catch per unit effort (CPUE) to estimate the relative abundance of exploitable stock. Whereas values of CPUE are usually adequate for studying spatial and temporal variation in resource abundance, often an absolute estimate of exploitable biomass is required. This is particularly true of yield assessments. Due to the relative nature of CPUE statistics, a conversion fac- tor is necessary to translate catch rates into absolute units of biomass. This proportionality is termed catchability, typically a constant parameter (but see Schnute 1983; Polovina 1986) which can be esti- mated from the results of intensive fishing ex- periments (Ricker 1975). The advantages of intensive fishing over alterna- tive methods of estimating the catchability coeffi- cient (q) are several. Foremost is that no history of either catch or effort data is needed. This character- istic makes methods of fishing success (Ricker 1975) or survey-removal (Schnute 1983) particularly at- tractive for use in assessments involving exploratory survey data, as well as for studying emerging new fisheries. A second advantage is that results can be obtained rapidly. Because fishing is, by definition, conducted intensively over a short time period and the necessary computations are quite simple, an estimate of q is quickly realized. Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI 96822-2396. Although these advantages recommend the ap- proach, two restrictive assumptions must be made in analyzing the data. One must assume, in the absence of information to the contrary, that the population fished is closed, or equivalently, that additions exactly balance removals other than those due to fishing. The basis of this assumption can be strengthened if the intensive fishing site is located in a naturally isolated area. For example, Polovina (1986) performed an intensive fishing experiment on a small pinnacle 5.5 km in circumference which was isolated by 75 km of deep water from the near- est similar habitat. A second assumption is that fish- ing removals account for all changes in stock biomass, i.e., natural mortality, growth, and recruit- ment are negligible during the period of fishing. For this reason, removals are carried out intensively over as short a time interval as possible. If both assumptions hold then q can be estimated directly by the slope of the linear regression of either CPUE on cumulative catch (Leslie and Davis 1939) or log{CPUE} on log{cumulative effort} (DeLury 1947). Refinements to these two basic methods have been proposed by Braaten (1969), Crittenden (1983), Schnute (1983), and Polovina (1986) among others. Generally, estimators have been found to be most sensitive to a departure from the assumption of con- stant catchability. A variety of adjustments have been used to correct this and other statistical prob- lems which often occur with real data. The work reported here is an application of the intensive fishing method to estimate the catchabil- Manuscript accepted July 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 927 FISHERY BULLETIN: VOL. 84, NO. 4 ity and population density of a deepwater caridean shrimp, Heterocarpus laevigatas. This circumglobal species is found in depths of 400-950 m in subtropical and tropical latitudes (King 1984). Experimental trapping surveys have shown it to be abundant at widespread localities in the central and western Pacific (King 1983), and a developing commercial fishery for this species has emerged in the Hawaiian Islands (Gooding 1984). Interest by Pacific island na- tions in promoting the harvest of this shrimp is great (King 1981), providing the impetus for an assess- ment of the Heterocarpus resource in the Mariana Archipelago. Additional results of this research program are reported elsewhere (Moffitt and Polovina2). MATERIALS AND METHODS Intensive fishing for Heterocarpus laevigatus was conducted in an area 3.5 km off the north end of Alamagan Island in the western Pacific (lat. 17°39'N, long. 145°50'E). Alamagan is part of the Commonwealth of the Northern Mariana Islands, lying 450 km north of Guam (Fig. 1). It is small, uninhabited, and of recent volcanic origin. While the ocean bottom slopes steeply away from the island at an angle of 25° to the east, south, and west, a broad shelf, approximately 6.5 km2 in area and lying 600-800 m deep, extends well off the north end of the island. This shelf was selected as a study site because 1) good catches of H. laevigatus were pre- viously obtained in the area, 2) 700 m is an ideal target depth for trapping this species (Moffitt and Polovina fn. 2), 3) the relatively uniform bottom topography would facilitate setting and retrieving fishing gear, and 4) the area had no known history of prior exploitation. Fishing was conducted over a 16-d period, 9-24 January 1984, from the NOAA ship Townsend Cromwell. Shrimp traps of standard Honolulu Lab- oratory design were set daily in four strings of five traps each. All traps were half round in shape (91 x 66 x 46 cm), with a frame constructed of 1.27 cm reinforcing steel, covered with 1.27 x 2.54 cm mesh hardware cloth (illustrated in figure 3 of Gooding 1984). Individual traps within a set were spaced 40 m apart and were baited with three chopped Pacific mackerel, Scomber japonicus. All traps were set between 1100 and 1300 h in 600-800 2Moffitt, R. B., and J. J. Polovina. In prep. Distribution and yield of the deepwater shrimp resource in the Marianas. South- west Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI 96822-2396. m and were retrieved the following day between 0800 and 1100 h. In addition, a large (150 x 150 x 150 cm) pyramidal commercial shrimp trap was sometimes deployed alone. When fishing gear was recovered, the traps were individually emptied and the contents sorted, counted, and weighed to the nearest 0.01 kg by species lot. On three occasions a random length- frequency sample of trap-caught H. laevigatus was saved for later study. All shrimp in these samples were measured to the nearest 0.1 mm CL (carapace length) with dial calipers. To accurately delimit the bottom topography of the study area, an unregistered reconnaissance hydrographic survey was conducted over the site on 9 February 1984, with the Townsend Cromwell. Depth soundings from a Raytheon3 fathometer were recorded every 3 min over an 8.5-h period as the vessel ran a predefined cruise track which covered the entire study area. The position of the vessel was recorded to the nearest 0.01 min at each sounding. The Townsend Cromwell returned to the study site again from 12 to 16 May 1984, to assess the recovery of the H. laevigatus population in the study area and to determine the effect of different baiting practices on CPUE. Four sets of six traps each were set over- night on each of four occasions. Half these traps con- tained three chopped Pacific mackerel whereas the other half (i.e., every other one) contained two whole Pacific mackerel. The catch was sorted and treated as discussed previously. RESULTS Hydrographic Survey A total of 164 depth soundings were obtained over the study site. The data were contoured using the GCONTOUR procedure of SAS/GRAPH (SAS 1981) and the resulting chart is presented in Figure 2. Solid lines represent isobath contours spaced at 200 m depth intervals. Note that the shrimp study site is a saddle point; concave upwards along the north- south axis and concave downwards from east-west. The hydrographic survey revealed a small but steep pinnacle and a deep canyon immediately adjacent to the study area. In the figure the locations of each string of five standard traps are shown as open circles (n = 60) whereas single sets of the large pyramid trap are given as closed circles (n = 8). Fishing effort was 3Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 928 RALSTON: FISHING EXPERIMENT FOR CARIDEAN SHRIMP Bank C ■°Maug I. o Asuncion I. DAgrihan I P Pagan I. -20°- ■19*- -18°- I LU 3 a U 2.0- 1.0- 200 400 600 Cumulative Catch - kg 800 Figure 3.— Leslie model applied to Heterocarpus laevigatus at Alamagan. Each point represents 1 day of fishing. Data from Table 1. randomly sequenced (P > 0.40). This result supports the assumption of constant catchability. At the time the experiment was terminated, 776 kg of shrimp had been removed by trapping. An estimate of the concomitant catch rate can be cal- culated from the regression equation of Figure 3. This estimate of CPUE is 1.82 kg/trap-night. When the Townsend Cromwell returned to the study site, 4 mo later, the mean catch rate was 1.91 kg/trap- night (42 effective standard trap-nights of effort, s = 1.33), this based on a total catch of 80.08 kg H. laevigatus. The preceding calculations include only those traps which were baited comparably to the experimental traps (three chopped Pacific mackerel). Traps with two whole baits (n = 42) yielded an average catch rate of 1.39 kg/trap-night (s = 1.09). Length-Frequency Distributions Examination of size-composition data can help interpret changes in weight CPUE. Declining trap catch rates could, for example, represent fewer in- dividuals of the same size. Conversely, a decline in the average size of individuals caught with no change in numbers would also result in declining CPUE. The three length-frequency distributions of H. laevigatus sampled during the period of experimen- tal fishing are presented in Figure 4. For each dis- tribution the date of capture, depth of capture, sam- ple size, and mean carapace length are provided. Although appearing superficially similar, the results of ANOVA show that significant differences exist in size composition among the three samples (F = 10.03, df = 2, 343, P < 0.001). These differences, however, do not explain the decline in CPUE. The data in the figure show that the mean size of H. laevigatus actually increased over time, and that the overall decline in CPUE observed in Figure 2 must therefore have been due to a decrease in the number of shrimps caught. DISCUSSION Powell (1979) has shown that the shape of the descending limb of length-frequency distributions can provide useful information concerning the rela- tionship between mortality and growth. Specifical- ly, the ratio of Z (instantaneous total mortality rate) to K (von Bertalanffy growth coefficient) is defined in a simple way by the interrelationship of the least size when fully vulnerable to the gear, the mean size in the catch of fully recruited individuals, and the 931 FISHERY BULLETIN: VOL. 84, NO. 4 15 10- 5- 15 10- ■ 5- 15 1-9-84 695 m n = 132 X = 37.4 fl/ 1-16-84 530 m n = 103 X = 40.2 N ru _DL ^ Carapace Length -mm Figure 4.— Length-frequency distributions of Heterocarpus laevigatas taken in shrimp traps. von Bertalanffy asymptotic size (L^). This is true if the following conditions hold: 1) the growth of in- dividuals follows a deterministic von Bertalanffy growth curve, 2) mortality is constant and uniform for all ages, and 3) recruitment is constant and con- tinuous over time (Beverton and Holt 1956). Results presented in Dailey and Ralston (1986) provide the basis for estimating the minimum CL when H. laevigatus becomes fully recruited to the trap fishery. They provide a regression equation relating carapace width (CW) to CL. In this study the smallest mesh dimension of standard shrimp traps was 1.27 cm. This provides a logical cutoff point for measurement of least CW for shrimp that are fully vulnerable to the gear. Based on their func- tional regression this corresponds to 30 mm CL. It is evident from the three panels in Figure 4 that the size distribution of H. laevigatus above 30 mm CL is characterized by both rising and descending portions. As shown by Powell (1979) this indicates a ZIK ratio of less than unity (i.e., instantaneous mortality rate is less than the growth coefficient). Alternatively, it is possible that the rising portions of the length distributions are not representative of the population sampled, but are instead a reflection of behavioral interactions among shrimp of differ- ent sizes. Chittleborough (1974), for example, has shown that the presence of large individuals of the decapod crustacean Panulirus cygnus inhibited smaller conspecifics from entering baited traps, even though smaller lobsters were vulnerable to the traps in the absence of large ones. If this kind of behavioral interaction was also in evidence here, the effective least size of H. laevigatus when fully vulnerable to the traps may be as large as 41 mm CL, the mode of the pooled length-frequency distri- bution. This would indicate a ZIK ratio of 2.0 because of the linearity of the descending portions of the size-frequency distributions. Only further experimentation will resolve this issue. With respect to the intensive fishing experiment it is useful to consider whether or not the basic assumptions of the Leslie model were violated dur- ing the course of the study. The first of these was closure of the population. Two factors support the contention that the study population was effective- ly isolated and that the effects of immigration and emigration were negligible. First, the hydrographic survey showed that the study site comprised a semi- isolated extension of the main island. Continuity of prime habitat (600-800 m depth) with the island proper extended along two narrow corridors to the southeast and southwest. The shrimp has been taken as shallow as 400 m and as deep as 950 m in the Mariana Archipelago, but the 600-800 m depth range encompasses the preponderance of the region's shrimp stock (Moffitt and Polovina fn. 2), although elsewhere (e.g., Fiji, Vanuatu, and Samoa) the depth distribution apparently extends into some- what shallower water (King 1984). The second fac- tor arguing for closure is that the catch rate of H. laevigatus remained low after a 4-mo hiatus in fish- ing. If movements or migrations of shrimp were biologically significant over this time interval, a larger change in CPUE would be expected. It is tempting to attribute the small increase in catch rate (4.9%) to some type of biological recovery, but the estimate of mean squared error in CPUE from Figure 3 (0.3754 kg2/trap-night2) indicates that background variation is too large for the observed difference to be significant. Regardless, the data support the assumption that the population is closed. The second assumption was that growth, natural mortality, and recruitment are negligible factors in accounting for changes in CPUE. That the experi- ment was completed in only 16 d and the popula- 932 RALSTON: FISHING EXPERIMENT FOR CARIDEAN SHRIMP tion was reduced an estimated 45.3% are persua- sive elements here. Additionally, the size-frequency data show no indication of a major alteration in population structure. As long as the selective prop- erties of the fishing gear remain unchanged, alter- ations in the length composition of the catch are not expected over short time intervals, at least due to the direct effects of fishing. Further, no recruitment of small shrimp is evident. That the mean size of H. laevigatus seemed to increase as the experiment pro- gressed might support the hypothesis that growth of the stock was significant. An alternate explana- tion, however, is that size structure varies with depth of capture. Results from the Hawaiian Islands (Gooding 1984; Dailey and Ralston 1986) have now demonstrated this. The three samples presented in Figure 4 are confounded by this variable; other unknown factors may also have affected the shrimp size-frequency data (e.g., sexual dimorphism, con- tagious dispersion, sampling error, etc.). In addition, the estimated growth rate from the data (3.9 mm CL over 8 d = 0.49 mm/d) is biologically unten- able. Other investigators, notably Schnute (1983) and Crittenden (1983), have cautioned against the effects of changing catchability and unequal variance on Leslie model estimates. From the data gathered, there is little statistical evidence to suggest that these factors affected parameter estimates and I therefore assume that 0.001945 trap-night-1 and 1,714 kg are reasonable estimates of standard trap catchability and virgin population size, respec- tively. Given that the virgin biomass of H. laevigatus in the study area was 1,714 kg, the next question is: How large an area was intensively fished? From Figure 2 it is clear that there is no simple answer to this question. A number of sets were located in areas peripheral to the main trapping area. Desig- nating the stipple bordered area as the effective area fished is arbitrary, but provides a useful starting point to allow calculation of shrimp densities. This area was calculated to be 312 ha, corresponding to a projected density of 5.5 kg of exploitable H. laevi- gatus per hectare. Since individuals weighed 28 g each, on average, this is equivalent to 1 exploitable shrimp/51 m2 of bottom, a remarkably low density. Furthermore, a catchability coefficient of 0.001945 trap-night-1 indicates that one unit of standard trap effort can reduce a 312-ha population of shrimp by about 0.2%. This is certainly a significant impact. The vulnerability to trapping that this species demonstrates is cause for attention and careful resource management. ACKNOWLEDGMENTS This paper is the result of the Resource Assess- ment Investigation of the Mariana Archipelago at the Southwest Fisheries Center Honolulu Labora- tory, National Marine Fisheries Service. I would like to thank the crew of the Townsend Cromwell for their help in completing this study and Samuel G. Pooley, Victor A. Honda, Leigh Neil, and Ahser Edward for their tireless efforts and good spirits while at sea. This paper benefited greatly from a review provided by C. D. Knechtel. LITERATURE CITED Beverton, R. J. H., and S. J. Holt. 1956. A review of methods for estimating mortality rates in fish populations, with special reference to source of bias in catch sampling. Rapp. P.-v. Reun. Cons. Perm. int. Explor. Mer 140:67-83. Braaten, D. O. 1969. Robustness of the DeLury population estimator. J. Fish. Res. Board Can. 26:339-355. Chittleborough, R. G. 1974. Home range, homing, and dominance in juvenile west- ern rock lobsters. Aust. J. Mar. Freshwater. Res. 25: 227-234. Crittenden, R. N. 1983. An evaluation of the Leslie-DeLury method and a weighted method for estimating the size of a closed popula- tion. Fish. Res. (Amst.) 2:149-158. Dailey, M. D., and S. Ralston. 1986. Aspects of the reproductive biology, spatial distribu- tion, growth, and mortality of the deepwater caridean shrimp, Heterocarpus laevigatus, in Hawaii. Fish. Bull., U.S. 84:915-925. DeLury, D. B. 1947. On the estimation of biological populations. Bio- metrics 3:145-167. Gooding, R. M. 1984. Trapping surveys for the deepwater caridean shrimps, Heterocarpus laevigatus and H. ensifer, in the Northwestern Hawaiian Islands. Mar. Fish. Rev. 46(2): 18-26. King, M. G. 1981. Increasing interest in the tropical Pacific's deepwater shrimps. Aust. Fish. 40(6):33-41. 1983. The ecology of deepwater caridean shrimps (Crustacea: Decapoda: Caridea) near tropical Pacific islands with par- ticular emphasis on the relationship of life history patterns to depth. Ph.D. Thesis, Univ. South Pacific, Suva, Fiji, 258 P- 1984. The species and depth distribution of deepwater cari- dean shrimps (Decapoda, Caridea) near some Southwest Pacific islands. Crustaceana 47:174-191. Leslie, P. H., and D. H. S. Davis. 1939. An attempt to determine the absolute number of rats on a given area. J. Anim. Ecol. 8:94-113. Polovina, J. J. 1986. A variable catchability version of the Leslie model with application to an intensive fishing experiment on a multi- species stock. Fish. Bull, U.S. 84:423-428. 933 FISHERY BULLETIN: VOL. 84, NO. 4 Powell, D. G. Schnute, J. 1979. Estimation of mortality and growth parameters from 1983. A new approach to estimating populations by the the length frequency of a catch. Rapp. P. -v. Reun. Cons. removal method. Can. J. Fish. Aquat. Sci. 40:2153-2169. int. Explor. Mer 175:167-169. Tate, M. W., and R. C. Clelland. Ricker, W. E. 1957. Nonparametric and shortcut statistics in the social, 1975. Computation and interpretation of biological statistics biological, and medical sciences. Interstate Printers Publ., of fish populations. Fish. Res. Board Can., Bull. 191, Inc., Danville, IL., 171 p. 382 p. Von Geldern, C. E., Jr. SAS Institute, Inc. 1961. Application of the DeLury method in determining the 1981. SAS/GRAPH user's guide. 1981 ed. SAS Inst., Inc., angler harvest of stocked catchable-sized trout. Trans. Am. 126 p. Fish. Soc. 90:259-263. 934 ICHTHYOPLANKTON IN NERITIC WATERS OF THE NORTHERN GULF OF MEXICO OFF LOUISIANA: COMPOSITION, RELATIVE ABUNDANCE, AND SEASONALITY James G. Ditty1 ABSTRACT Ichthyoplankton samples were collected monthly between November 1981 and October 1982 in neritic continental shelf waters off Louisiana. The survey provided the first quantitative data on the abundance and seasonal occurrence of larval fishes from open coastal waters of this area. At least 48 families of fishes were represented in samples that included 107 taxa, 54 of which were identified to species. Larval densities were lowest during the winter and highest during the summer with a mean monthly density of 208/100 m3. Five families accounted for about 90% of total larvae: Engraulidae, Sciaenidae, Clupeidae, Carangidae, and Bothidae. The five most abundant taxa overall, in order of decreasing abundance, were anchovies (Engraulidae); Atlantic croaker, Micropogonias undulatus; Atlantic thread herring, Opistho- nema oglinum; gulf menhaden, Brevoortia patronus; and Atlantic bumper, Chloroscombrus chrysurus. These taxa accounted for 82% of all larvae collected. Comparison of ichthyoplankton surveys throughout the Gulf of Mexico showed that the 10 most abundant families contributed over 90% of total larval abun- dance in coastal surveys but less than 70% in offshore surveys. Likewise, the five most abundant taxa contributed over 80% of total larval abundance in all but one of the coastal surveys but less than 40% in the offshore surveys. These data suggest that compared with offshore waters, there are relatively fewer dominant taxa among the ichthyoplankton in neritic waters of the Gulf of Mexico. The northern Gulf has traditionally been one of the most productive fishery areas in North America (Gunter 1967), yet seasonality and abundance of larval fishes from open waters are poorly known. Previous studies of early life history stages in this area have mainly been focused either on select taxa (Turner 1969; Fore 1970, 1971; Christmas and Waller 1975; Fruge 1977; Ditty 1984; Cowan 1985; Shaw et al. 1985) or to surveys limited in temporal and areal coverage (Walker 1978; Ditty and Trues- dale 1984). Stuck and Perry (1982) surveyed the ichthyoplankton community adjacent to Mississippi Sound, while Marley (1983) conducted an egg survey and Williams (1983) a larval fish survey of lower Mobile Bay, AL. The most comprehensive studies of the offshore larval ichthyofauna in the Gulf of Mexico and adjacent areas were those of Finucane et al. (1977) from the south Texas outer continen- tal shelf; Houde et al. (1979) from the eastern Gulf of Mexico off Florida; Richards (1984) from the Caribbean Sea; and Powles and Stender (1976) from the South Atlantic Bight area off the east coast of the United States. The objective of this paper is to provide quantitative data on the abundance and seasonal occurrence of larval fishes from open Louisiana Department of Wildlife and Fisheries, Seafood Divi- sion, P.O. Box 15570, Baton Rouge, LA 70895. coastal waters of the northern Gulf of Mexico off Louisiana. MATERIALS AND METHODS Plankton samples were collected monthly between November 1981 and October 1982 (except March 1982) in neritic continental shelf waters off Louisi- ana. Samples were collected at six stations in a 3.2 km2 area located about 12.9 km south-southwest of Caminada Pass, in depths of 10-12 m (Fig. 1). Col- lections were made with a 60 cm paired-net, open- ing and closing bongo-type BNF-1 sampler2, each net was of 0.363 mm Nitex3 mesh. Nets were lowered to depth, opened, and towed simultaneous- ly, in series, at discrete depths (surface, middepth, and near-bottom) for about 3 min, at a ship speed of approximately 1.5 kn; all samples were collected during the day. A General Oceanics (Model 2030) flowmeter was placed in the mouth of each net to estimate volume filtered. Samples were preserved in seawater with buffered Formalin and returned to the laboratory for sorting. Fish larvae were removed from each net and identified to the lowest Manuscript accepted July 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 2Tareq and Co., 8460 S.W. 68th Street, Miami, FL 33143. 3Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 935 FISHERY BULLETIN: VOL. 84, NO. 4 94°00W 93 00 W 92°00W I 9100 W i 90°00W I 89 OOW — J— 31°00N 30 OON- 29 OON- GULF OF MEXICO Figure 1.— Location of study area. possible taxon, and standard length was measured with an ocular micrometer; all specimens were sub- sequently archived in 70% ethanol. Hydrographic profiles of the water column were taken at approx- imately 1-1.5 m intervals with a Martek Mark VI water quality monitor, except during January, Feb- ruary, and September 1982 when the Martek unit was inoperable. During these 3 mo, water temper- atures and salinities were measured with a Beckman RS-5 inductive salinometer near surface, middepth, and bottom. Estimates of monthly mean larval den- sities were calculated by dividing total larvae by total volume filtered at each depth and integrated over depth. Densities are expressed as number/ 100 m3. Seasonal designations were based primarily on mean surface water temperatures during the year: <20°C (Winter: December-February); 20°-25°C (Spring: April-May); >25°C (Summer: June- August); and rapidly declining surface water temperatures (Fall: September-November). Additional data on larval occurrence and season- ality only were compiled from surface-towed meter net (0.363 mm mesh) collections at stations sampled between January 1981 and December 1982. These data consisted of four nearshore stations located adjacent to the bongo stations and were sampled monthly. Two additional groups of stations, one of four and the other of five stations, were located about 24 km south of the nearshore stations in depths of about 30 m. Each group of offshore sta- tions was sampled quarterly but on consecutive months during 1981; thereafter, monthly samples were collected only at the four station group. These seasonality data are not discussed but are included in the Appendix Table. Ancillary occurrence and seasonality data on lar- val bothids, scombrids, and sciaenids collected off Louisiana during the spring and early summer of 1982 were compiled from surface-towed 0.5 m ring net (0.505 mm mesh), 60 cm bongo net (0.333 mm mesh), and surface-towed 1 x 2 m neuston net (0.946 mm mesh) samples (SEAMAP 1983). Bongo tows were oblique and from the surface to 200 m or within 5 m of the bottom at shallower depths. Seasonality data for these taxa were compiled only from stations located between long. 88°30'W and 93°30'W and shoreward of lat. 27°00'N and, al- though not discussed, are also included in the Appendix Table. Additional station and cruise data are provided in Richards et al. (1984). 936 DITTY: ICHTHYOPLANKTON IN NERITIC WATERS RESULTS Taxonomic Problems Larvae of many fishes in the northern Gulf of Mexico are poorly known and taxonomic problems are common, even in some of the most abundant taxa. No attempt was made to identify blennies, gobies, myctophids, synodontids, or cynoglossids to species because of the paucity of literature on lar- val development for these taxa. Little is also known about the taxonomy and morphological development of engraulid larvae. At least five species of engrau- lids are known to occur as adults in the north-central Gulf: Anchoa mitchilli, A. hepsetus, A. lyolepis, A. cubana, Anchoviella perfasciata (Modde and Ross 1981), and possibly Engraulis eurystole (Hastings 1977). Anchoa mitchilli, A. hepsetus, and A. lyolepis probably account for most of the engraulid larvae collected. Larvae of A. hepsetus and A. mitchilli in the Chesapeake Bay Region can be distinguished from each other primarily on placement of dorsal and anal fins (Manseuti and Hardy 1967), but this character is insufficient to separate reliably the addi- tional species of anchovy that may occur in this area. Separation of menhaden larvae is also difficult. Three species of menhaden are known to occur as adults in this area: Brevoortia smithi (Chandeleur Sound, LA, eastward), B. patronus (Tampa Bay, FL, westward to Veracruz, Mexico) and B. gunteri (Mississippi Sound, MS, westward) (Christmas and Gunter 1960; Springer and Woodburn 1960; Dahl- berg 1970; Turner 1971). Published descriptions are available for laboratory-reared larvae of B. smithi (Houde and Swanson 1975) and B. patronus (Hettler 1984) only. Brevoortia gunteri have never been described nor positively identified from the north- ern Gulf. Although the congeners have spawning seasons that reportedly overlap, the center of spawning of B. patronus is apparently off Louisiana between the Mississippi and Atchafalaya River Deltas (Turner 1969; Fore 1970; Christmas and Waller 1975). Since Brevoortia larvae collected dur- ing this study appear similar to that described as B. patronus (Hettler 1984) and because I have recog- nized in subsequent samples (at sizes >7 mm SL) a second morph that could be B. gunteri, all larvae were considered B. patronus. Published descriptions of sciaenid larvae are in- adequate to reliably distinguish between small larvae of the species of Menticirrhus (M. ameri- canus, M. littoralis, and M. saxatilis) or between small Cynoscion arenarius and C. nothus. Two types of C. arenarius larvae were recognized primarily on the absence (Type A) or presence (Type B) of pig- ment in the dorsal midline immediately above the enlarged melanophore located in the ventral mid- line about midway along the anal fin base. Additional data on the separation of these types are provided in Cowan (1985). Small carangid larvae (<5 mm SL) of certain taxa are also difficult to identify and were referred to a morphological type when a generic or specific epithet could not be assigned. The taxonomy and/or larval development of some of the other monthly dominants (e.g., Lepophidium spp., Ophi- dion spp., Auxis spp., and Ariomma sp.) are poorly understood. Hydrography Water temperatures between November 1981 and October 1982 ranged from 16°C in January and Feb- ruary to 31° C in June and were below 20 °C from December through February and above 25 °C from May through October. There was little thermal stratification except during the summer, with stratification most pronounced in June (Fig. 2A). Salinity stratification was most pronounced from February through August, with little stratification from September through January. Salinities were lowest near the surface, increased with depth, and ranged from <20%o near the surface in February to 32%o in December; salinities near the bottom ranged from 31%o in September to 36%o in the spring and early summer. Salinities near the sur- face steadily decreased from April through July and increased thereafter, whereas those near the bot- tom were comparatively more stable throughout the study period (Fig. 2B). In February, there was a distinct salinity gradient within the upper 6 m of the water column that ranged from 18°/oo at the sur- face to 30°/oo near middepth. In June, two distinct water masses were present with a halocline near middepth. Salinities of these two water masses dif- fered by about 10%o with the less saline waters above middepth (Fig. 2B). Further information on water temperature and salinity variability and the physical processes that affect the hydrography of the study area are provided in Wiseman et al. (1982). Seasonal Composition and Abundance At least 48 families of fishes were represented in bongo net samples that included 107 taxa, 54 of which were identified to species. About 36,500 lar- vae were collected, with <5% (primarily damaged or yolk-sac larvae) unidentifiable to family. The majority of larvae collected were <5 mm SL except 937 FISHERY BULLETIN: VOL. 84, NO. 4 N M N M J J A Figure 2.— Profiles of water temperature and salinity (November 1981-October 1982) at a representative station from the study area located in neritic waters of the northern Gulf of Mexico off Louisiana. A. Water temperature, B. Salinity. X indicates sampling depths. Collection dates were scaled by Julian calendar. larvae of clupeiform fishes; these were usually <10 mm SL. Generally, seasonal larval densities followed water temperatures (T) and were lowest during winter (X = 51/100 m3 at T <20°C), increased during the spring (X = 207/100_m3 at T <25°C), peaked during the summer (X = 394/100 m3 at T near 30°C), and declined during the fall (X = 179/100 m3 at rapidly declining T). Approximately 6% of all fish larvae were collected during the winter and 47.5% during the summer. Larval densities were lowest in December and highest in June, with a mean monthly density of 208/100 m3 (Fig. 3). Over- all, December had the fewest taxa (13) and Septem- ber the most (37). Five families accounted for about 90% of total larvae: Engraulidae, Sciaenidae, Clupeidae, Carangidae, and Bothidae. The five most abundant taxa overall, in order of decreasing abun- dance, were anchovies (Engraulidae); Atlantic croaker, Micropogonias undulatus; Atlantic thread herring, Opisthonema oglinum; gulf menhaden, Brevoortia patronus; and Atlantic bumper, Chloro- scombrus chrysurus. These taxa accounted for about 82% of all larvae taken. Thirty-eight taxa occurred in sufficient numbers that they were within the 10 most abundant taxa collected in at least one month. Densities of these taxa are presented in Table 1. Anchovies accounted for about 49% of all larvae 938 DITTY: ICHTHYOPLANKTON IN NERITIC WATERS and were collected throughout the year, but were most abundant in June and least abundant in November (Table 1). Anchovies accounted for 65% of all larvae taken during the spring and 69% dur- ing the summer, but declined to about 6% of all lar- vae collected during the fall and winter, respectively; anchovies were the second most abundant taxon col- lected during the winter and were fourth during the fall. Most anchovy larvae were collected near the surface and middepths; only 11% were collected near the bottom. A few flat anchovy, Anchoviella perfasciata, postlarvae were collected in February only. Atlantic croaker accounted for 66% of all sciaenid larvae and were most abundant in November (Table 1). This species accounted for 58% of all larvae taken during the fall and for 14% of larvae overall. Most Atlantic croaker (65%) were collected near middepth with only 1% collected near the surface. Two types of sand seatrout, Cynoscion arenarins, were recog- nized with Type A collected from April to September and Type B from April to October. Of all sand seatrout larvae taken, 60% were Type A and 40% Type B, with Type A the second and Type B the third most abundant of all sciaenid larvae. Density of Type A exceeded that of Type B until September o o 2 uj Q < > M A MONTH Figure 3.— Density of ichthyoplankton (no./lOO m3) by month, from neritic Gulf of Mexico waters off Louisiana, November 1981- October 1982. and October when Type B were more abundant (Table 1). Most Type A (66%) and Type B (56%) lar- vae were collected near the bottom with <5% of Type A and of Type B larvae, respectively, collected near the surface. Larvae of red drum, Sciaenops ocellatus, were taken during the fall only and were most abundant in September, whereas Menticirrhus spp. were collected in all months except December and January and were most abundant in October (Table 1). Larval densities of other less abundant sciaenids that included black drum, Pogonias cromis; banded drum, Larimus fasciatus; spot, Leiostomus xanthurus; silver perch, Bairdiella chrysoura; and silver seatrout, Cynoscion nothus, never exceeded 1/100 m3. Densities of star drum, Stellifer lance- olatus, and spotted seatrout, C. nebulosus, were <2/100 m3 for any month. Larvae of both the scaled sardine, Harengula jaguana, and Atlantic thread herring were collected from April to October, whereas gulf menhaden were collected from October to February and round her- ring, Etrumeus teres, only in January and February. No larvae of Spanish sardine, Sardinella sp., were identified. Densities of Atlantic thread herring were greatest in June, scaled sardine in July, and gulf menhaden in January. Densities of Atlantic thread herring accounted for about 58% of all clupeid larvae and for 9% of larvae overall; gulf menhaden accounted for 34% of all clupeids and for 5% of lar- vae overall. Scaled sardine accounted for 8% of all clupeid larvae. Atlantic thread herring was the sec- ond most abundant taxon collected in each season except winter, and accounted for 88% of all clupeid larvae collected between April and October; gulf menhaden accounted for 73% of all winter larvae. Over 99% of Atlantic thread herring and 80% of scaled sardine were collected when surface water temperatures were above 25°C; 90% of gulf men- haden were taken at water temperatures below 20°C. Most scaled sardine (79%) larvae were taken near the surface and only 2% near the bottom. Men- haden larvae were abundant at all depths with 37% collected near the surface and 24% near the bottom. Atlantic thread herring were most abundant near middepth (62%) and least abundant near the surface (6%). Larvae of Atlantic bumper were collected from June to October but were most abundant in July. This species accounted for about 5% of all larvae and was the third most abundant taxon collected during both the summer and fall months. Atlantic bumper accounted for about 94% of all carangid lar- vae with most bumper (94%) collected when surface water temperatures averaged 30°C. Atlantic 939 FISHERY BULLETIN: VOL. 84, NO. 4 Table 1 .—Densities (no./1 00 m3) of abundant taxa from neritic waters of the northern Gulf of Mexico off Louisiana, November 1981 -October 19821. Taxa Nov. Dec. Jan. Feb. Apr. May June July Aug. Sept. Oct. Engraulidae 0.8 4.0 1.6 2.9 193.8 74.3 598.1 213.4 3.0 27.6 3.3 Brevoortia patronus 7.3 2.7 67.1 41.0 0.8 1.4 Etrumeus teres — — 0.4 0.2 — — — — — — — Opisthonema oglinum — — — — 0.3 52.1 71.9 2.3 16.6 62.5 (2) Harengula jaguana — — — — 5.7 0.2 — 11.9 9.5 0.2 (2) Synodontidae (2) — (2) 0.6 — 0.3 — — — — — Myctophidae (2) — 1.0 7.2 0.4 Bregmaceros canton 0.4 (2) — 0.1 0.3 0.5 — — — — 0.1 Lepophidium spp. 0.9 Ophidion spp. 0.6 Membras martinica — — — (2) 1.0 (2) Carangidae Type A — — — — — 1.5 — — — — — Chloroscombrus chrysurus 7.8 48.3 13.5 39.2 6.7 Oligoplites saurus 1.3 2.8 — — — Trachurus lathami — ; — — 0.4 — — — — — — — Orthopristis chrysoptera — — — (2) 1.0 — — — — — — Archosargus probatocephalus — — — — 2.9 — — — — — — Lagodon rhomboides (2) — 0.3 0.4 Cynoscion arenarius (Type A) — — — — 22.5 6.3 7.2 17.2 10.3 1.9 — Cynoscion arenarius (Type B) — — — — 10.6 0.8 6.7 11.1 10.2 5.7 1.5 Leiostomus xanthurus 0.8 0.5 0.1 0.2 Menticirrhus spp. (2) — — (2) 1.2 0.4 1.4 2.5 0.1 2.0 5.2 Micropogonias undulatus 182.8 2.8 — 2.2 126.4 Sciaenops ocellatus 12.8 6.3 Stellifer lanceolatus — — — — 1.4 0.1 1.6 0.3 0.3 0.3 0.2 Chaetodipterus faber (2) 0.6 5.5 0.1 1.3 — Mugil cephalus — — 0.4 (2) — — — — — — — Blennidae 0.3 1.6 0.6 (2) 2.8 9.0 1.2 2.0 0.2 0.9 0.1 Gobiidae 0.4 1.0 — 0.4 0.6 0.8 0.2 0.2 — 0.2 0.4 Auxis spp. — — — — 0.2 1.0 — — — 0.5 0.1 Scomberomorus maculatus — — — — 0.5 0.1 1.7 5.5 4.8 1.5 — Ariomma sp. — — 0.7 — — — — — — — — Peprilus burti 1.4 0.1 0.5 0.9 0.1 0.2 0.1 — — — 0.5 Peprilus paru 0.4 0.6 0.5 0.6 4.4 (2) Etropus crossotus 0.4 — — — — 9.0 9.4 1.9 (2) 0.7 0.5 Citharichthys spilopterus 0.4 (2) 0.4 0.6 0.1 — — (2) — — — Symphurus spp. 0.5 0.1 — — 0.1 1.5 2.8 1.4 0.1 0.5 0.2 Myrophis punctatus (2) 0.1 0.1 0.2 Wo data for March 1982. 2Density <0. 1/100 m3. bumper were most abundant near middepth (60%) and least abundant near the bottom (9%). Other abundant carangids included leatherjacket, Oligo- plites saurus; rough scad, Trachurus lathami; and carangid Type A larvae. All carangid Type A lar- vae were <4 mm SL and appear similar to that described as the round scad, Decapterus punctatus, by Aprieto (1974). Larvae of gulf butterfish, Peprilus burti, occurred from October to June and harvestfish, P. paru, from May to October (Table 1). Most gulf butterfish (85%) larvae were collected when surface water tempera- tures were <25°C whereas all harvestfish were col- lected when surface water temperatures were above 25°C. Spanish mackerel, Scomberomorus macula- tus, larvae occurred from April to September but were most abundant in July; most (96%) were col- lected when surface water temperatures exceeded 25°C. Most Spanish mackerel (74%) larvae were col- lected near middepth; only 5% were collected near the bottom. King mackerel, 5. cavalla, larvae were collected only in September and at a density <0.5/100 m3. Many taxa occurred in relatively low abundance, and although not included in Table 1, provided addi- tional data on seasonality. These data are presented in the Appendix Table. Only taxa with larvae <10 mm SL for a given month (except anguilliform lepto- cephali or sygnathids) were included in the Appen- dix Table, except where noted. DISCUSSION Data on peak seasonal occurrence of many of the abundant taxa from the present study agree with those of other coastal surveys from the north-central Gulf of Mexico off Mississippi (Stuck and Perry 1982) and off Alabama (Williams 1983). During 1982, greatest densities of larval menhaden off central Louisiana (the present study) occurred in 940 DITTY: ICHTHYOPLANKTON IN NERITIC WATERS January-February and off western Louisiana (Shaw et al. 1985) in February-March. Stuck and Perry (1982) found larval menhaden most abundant be- tween January and March adjacent to Mississippi Sound. These data agree with past studies (Fore 1970; Christmas and Waller 1981) from this area that reported high densities of menhaden eggs between December and February. All three of the north-central Gulf studies (Stuck and Perry 1982; Williams 1983; and the present study) reported greatest densities of Atlantic croaker during Octo- ber and November; densities of sand seatrout were greatest in April, with a second smaller peak in den- sity during either July or August. Both Atlantic bumper and Spanish mackerel were most abundant from July to September in each of these three studies. Stuck and Perry and the present study also found the greatest density of red drum in Septem- ber; Williams did not sample in September. In the present study, Atlantic thread herring were most abundant in June, with a second peak in September; scaled sardine were most abundant during July and August. Few scaled sardine and Atlantic thread her- ring larvae were collected by Williams; no Atlantic thread herring and few scaled sardine were collected by Stuck and Perry. All three of these north-central Gulf studies also reported a bimodal peak in abun- dance of engraulids but differed slightly in month of peak density. Stuck and Perry, and Williams found greatest densities in April, with a second smaller peak in August. The smaller of the two peaks in abundance of engraulids occurred in April, with the greatest density in June in the present study (Table 1). Comparison of dominant families and taxa col- lected overall in the present study with those of other ichthyoplankton surveys throughout the Gulf of Mexico are presented in Tables 2 and 3. Lower bay/coastal surveys were those conducted primarily inside the 10 m depth contour, except for Hoese (1965), who had a single transect of six stations out to 50 m. Offshore surveys were those conducted mainly in waters deeper than 10 m but shoreward of the edge of the continental shelf. Although not all the data listed in Tables 2 and 3 are directly com- parable because of differences in gear type, mesh size, or tow, these studies provide general informa- tion on larval composition and abundance. Most of the surveys from coastal waters (Hoese 1965; Blanchet 1979; Williams 1983; Collins and Finucane 1984; and the present study) found that engraulids dominated the summer ichthyoplankton, whereas Stuck and Perry (1982) reported engraulids second to Atlantic bumper in abundance. However, Stuck and Perry may have undersampled small engraulid and clupeid larvae because of the large mesh (1.050 mm) of their nets. Menhaden dominated the winter ichthyoplankton in all of the aforemen- tioned coastal surveys, except Collins and Finucane (1984). These authors found that pigfish, Orthopris- tis chrysoptera, larvae were the most abundant taxa during the winter in waters off the Everglades of south Florida. All of these surveys also consistent- ly placed engraulids and sciaenids at or near the top in total larval abundance. Overall, clupeids were relatively more abundant off south Florida (Collins and Finucane 1984) than in the other coastal surveys, except Hoese (1965), who sampled only the Table 2.— Comparison of the five most abundant families collected overall from neritic waters off Louisiana with other ichthyoplankton surveys throughout the Gulf of Mexico. Engraulidae Sciaenidae Clupe idae Carangidae Bothidae Gear type, mesh size, depth of tow, Study Rank % Rank % Rank °/o Rank °/o Rank % Location and region1 Present study 1 49.0 2 19.0 3 16.0 4 5.5 5 1.5 Coastal 1,6,9,10,11,15 Hoese 1965 2 42.0 3 7.0 1 45.0 — 0.5 — 0.5 Coastal 2,4,9,14 Stuck and Perry 2 19.7 3 18.2 6 3.4 1 38.8 4 5.6 Coastal 3,8,9,11,15 1982 Williams 1983 1 69.3 2 14.5 3 4.5 4 2.8 — 0.5 Lower Mobile Bay/Coastal 3,7,9,11,15 Blanchet 1979 1 75.8 2 4.9 8 1.9 7 2.2 <0.1 Lower Apalachicola Bay/Coastal 2,6,7,9,16 Collins and 2 22.5 4 6.9 1 24.1 5 6.1 — <0.1 Coastal 2,7,9,12,18 Finucane 1984* Finucane et al. 19793 5 6.2 — <0.1 3 8.1 8 3.7 6 6.1 Offshore 1,7,13,14 Houde et al. 1979 12 2.0 30 0.3 1 20.5 6 3.9 3 6.4 Offshore 1,2,7,12,17 '1 60 cm bongo 7 0.505 mm 13 double-obliq ue 2inshore data only 2 1 m 8 1 .050 mm 14 west-central 31977 bongo net data only 3 1 x 0.5 m rectangu lar 9 surface 15 north-centra I 4 0.086 mm 10 middepth 16 north-east 5 0.333 mm 11 bottom 17 east-central 6 0.363 mm 12 oblique 18 south-east 941 Table 3.- FISHERY BULLETIN: VOL. 84, NO. 4 -Comparison of five most abundant taxa from neritic waters off Louisiana with ichthyoplankton surveys throughout the Gulf of Mexico. Stuck Collins and and Finucane Houde Present Hoese Perry Williams Blanchet Finucane et al. et al. study 1965 1982 1983 1979 19841 19792 1979 Taxa % % % % % % % % Engraulidae 49.0 19.7 69.0 75.8 22.5 7.1 Micropogonias undulatus 14.0 5.8 Opisthonema oglinum 9.0 4.5 7.9 Brevoortia patronus 5.0 15.6 4.3 Chloroscombrus chrysurus 5.0 38.4 2.8 1.8 4.8 Harengula jaguana 29.1 Anchoa hepsetus 24.0 Anchoa mitchilli 17.7 Menticirrhus spp. 2.6 Cynoscion arenarius 12.0 8.2 Citharichthys-Etropus complex 5.6 Symphurus spp. 3.8 Atherinidae 3.9 Gobiesox strumosus 3.2 Gobiosoma spp. 2.7 Microgobius spp. 9.1 Orthopristis chrysoptera 4.8 Gobiidae 15.8 15.1 Bregmaceros atlanticus 7.1 Saurida spp. 6.1 Syacium spp. 4.1 Sardinella anchovia 8.6 Decapterus punctatus 3.1 Diplectrum formosum 2.8 1 1nshore data only. 21977 bongo net data only. surface waters of his offshore transect (Table 2). Offshore, Houde et al. (1979) found that clupeids (Spanish sardine and Atlantic thread herring), gobiids, and bothids (mostly dusky flounder, Sya- cium papillosum) dominated summer ichthyoplank- ton in the eastern Gulf of Mexico off Florida, where- as clupeids (round herring and Spanish sardine), bothids (mostly gray flounder, Etropus rimosus), and bregmacerotids dominated the winter. In the western Gulf of Mexico off the south Texas coast, Finucane et al. (1979) reported that, during 1977, clupeids (mostly scaled sardine) and bothids (most- ly Syacium spp.) dominated the summer and breg- macerotids and clupeids (menhaden) the winter ichthyoplankton. In the northern Gulf of Mexico off Louisiana, Ditty and Truesdale (1984) found that engraulids and carangids (mostly Atlantic bumper) dominated the summer (July 1976), whereas larvae of clupeids (mostly gulf menhaden) and gobiids dominated the winter (January-February 1976). The most abundant families collected overall off Florida were clupeids and gobiids (35.6% of all larvae), and off south Texas were gobiids and synodontids (26.7% of all larvae). The kinds of larvae (gobiids, bothids, clupeids, and bregmacerotids) that domi- nated these two offshore surveys were similar, but with clupeids and bothids relatively more abundant off Florida than Texas; engraulids were relatively more abundant off south Texas than off Florida (Table 2). Ditty and Truesdale (1984) found clupeids and engraulids most abundant overall (67.7% of all larvae), but their surveys were too limited temporal- ly and in areal coverage for adequate comparison to the other two offshore surveys. The 10 most abundant families accounted for 66.6% of all larvae collected off Florida (Houde et al. 1979) and for 68.6% off south Texas (Finucane et al. 1979). In contrast, the top 10 families in each of the coastal surveys contributed over 90% of all larvae collected. Likewise, the five most abundant taxa contributed over 80% of all larvae collected in all but one (Collins and Finucane 1984) of the coastal surveys but <40% in the two offshore surveys (Table 3). In conclusion, there was general agreement among all three coastal surveys from the north- central Gulf of Mexico on peak seasonal occurrence of many of the abundant taxa and on the dominant families in overall larval abundance. Comparison of other coastal and offshore ichthyoplankton surveys throughout the Gulf of Mexico with the present study suggests that, when compared with offshore waters, there are relatively fewer dominant taxa 942 DITTY: ICHTHYOPLANKTON IN NERITIC WATERS among the ichthyoplankton in neritic waters of the Gulf of Mexico. ACKNOWLEDGMENTS This study represents a portion of an ongoing multiyear synoptic environmental assessment of the Louisiana Offshore Oil Port (LOOP, Inc.) and related facilities conducted by the Louisiana Department of Wildlife and Fisheries. I would like to thank Robert Ganczak and R. Harry Blanchet for their support and advice; Carlos Garces for statistical guidance; Jill Onega for typing the various drafts of the manu- script; Ron Gouguet for computer programming expertise, collection of hydrographic data, and generating the hydrographic profile plots; and to acknowledge Frank M. Truesdale, R. Harry Blan- chet, and the other reviewers for their valuable com- ments and suggestions for manuscript improve- ment. Thanks to the captain and crew of the LOOP Vigilance, to LOOP Inc., and to the Louisiana Department of Wildlife and Fisheries for additional support. Thanks also to the Southeast Area Monitor- ing and Assessment Program (SEAMAP) for pro- viding data on the bothids, sciaenids, and scombrids collected off Louisiana during 1982 and to William J. Richards and Tom Potthoff for identifying the SEAMAP scombrids. LITERATURE CITED Aprieto, V. L. 1974. Early development of five carangid fishes of the Gulf of Mexico and the South Atlantic coast -of the United States. Fish. Bull., U.S. 72:415-443. Blanchet, R. H. 1979. The distribution and abundance of ichthyoplankton in the Apalachicola Bay, Florida area. Master's Thesis, Florida State Univ., Tallahassee, 143 p. Christmas, J. Y., and G. Gunter. 1960. Distribution of menhaden, genus Brevoortia, in the Gulf of Mexico. Gulf Coast Research Laboratory, Ocean Springs, MS, 20 p. Collins, L. A., and J. H. Finucane. 1984. Ichthyoplankton survey of the estuarine and inshore waters of the Florida Everglades, May 1971 to February 1972. U.S. Dep. Commer., NOAA Tech. Rep., NMFS 6, 75 P- Cowan, J. H., Jr. 1985. The distribution, transport and age structure of drums (Family Sciaenidae) spawned in the winter and early spring in the continental shelf waters off western Louisiana. Ph.D. Thesis, Louisiana State Univ., Baton Rouge, 182 p. Dahlberg, M. D. 1970. Atlantic and Gulf menhadens, genus Brevoortia (Pisces: Clupeidae). Bull. Fla. State Mus. Biol. Sci. 15:91-162. Ditty, J. G. 1984. Seasonality of sciaenids in the northern Gulf of Mexico. Assoc. Southeastern Biol. Bull. 31(2):55. Ditty, J. G., and F. M. Truesdale. 1984. Ichthyoplankton surveys of nearshore Gulf waters off Louisiana: January - February and July, 1976. Assoc. Southeastern Biol. Bull. 31(2):55-56. Finucane, J. H., L. A. Collins, L. E. Barger, and J. B. McEachran. 1979. Ichthyoplankton/mackeral eggs and larvae. NOAA Final Report to BLM. Environmental Studies of the South Texas Outer Continental Shelf 1977. BLM Contract AA550-1A7-21, 504 p. Fore, P. L. 1970. Oceanic distribution of eggs and larvae of the Gulf menhaden. In Report of the Bureau of Commercial Fish- eries Biological Laboratory, Beaufort, N.C., for the fiscal year ending June 30, 1968, p. 11-13. U.S. Fish Wildl. Serv. Cir. 341. 1971. The distribution of the eggs and larvae of the round herring, Etrumeus teres, in the northern Gulf of Mexico. Assoc. Southeastern Biol. Bull. 18(1):34. Fruge, D. J. 1977. Larval development and distribution of Micropogonias undulatus and Leiostomus xanthurus and larval distribution of Mugil cephalus and Bregmaceros atlanticus of the south- eastern Louisiana coast. Master's Thesis, Louisiana State Univ., Baton Rouge, 75 p. Gunter, G. 1967. Some relationships of estuaries to the fisheries of the Gulf of Mexico. In G. Lauff (editor), Estuaries, p. 621-638. Am. Assoc. Adv. Sci. Spec. Publ. No. 83. Hastings, R. W. 1977. Notes on the occurrence of the silver anchovy, Engrau- lis eurystole, in the northern Gulf of Mexico. Northeast Gulf Sci. 1(2):116-118. Hettler, W. F. 1984. Descriptions of eggs, larvae, and early juveniles of Gulf menhaden, Brevoortia patronus, and comparisons with Atlantic menhaden, B. tyrannus, and yellowfin, B. smithi. Fish. Bull., U.S. 82:85-95. Hoese, H. D. 1965. Spawning of marine fishes in the Port Aransas, Texas area as determined by the distribution of young and larvae. Ph.D. Thesis, Univ. Texas, Austin, 144 p. Houde, E. D., J. C. Leak, C. E. Dowd, S. A. Berkeley, and W. J. Richards. 1979. Ichthyoplankton abundance and diversity in the east- ern Gulf of Mexico. Report to BLM, Contract No. AA550- CT7-28, 546 p. Houde, E. D., and L. J. Swanson, Jr. 1975. Description of eggs and larvae of yellowfin menhaden, Brevoortia smithi. Fish. Bull., U.S. 73:660-673. Manseuti, A. J., and J. D. Hardy, Jr. 1967. Development of fishes of the Chesapeake Bay region: an atlas of egg, larval, and juvenile stages. Nat. Resourc. Inst., Univ. Maryland, Baltimore, 202 p. Marley, R. D. 1983. Spatial distribution patterns of planktonic fish eggs in lower Mobile Bay, Alabama. Trans. Am. Fish. Soc. 112: 257-266. Modde, T., and S. T. Ross. 1981. Seasonality of fishes occupying a surf zone habitat in the northern Gulf of Mexico. Fish. Bull., U.S. 78:911-922. Powles, H., and B. W. Stender. 1976. Observations on composition, seasonality and distribu- tion of ichthyoplankton from MARMAP cruises in the South Atlantic Bight in 1973. South Carolina Mar. Resour. Cent., Tech. Rep. Ser. No. 11, 47 p. 943 Richards, W. J. 1984. Kinds and abundances of fish larvae in the Caribbean Sea and adjacent areas. U.S. Dep. Commer., NOAA Tech. Rep., NMFS SSRF-776, 54 p. Richards, W. J., T. Potthoff, S. Kelley, M. F. McGowan, L. Ejsymont, J. H. Power, and R. M. Olvera L. 1984. SEAMAP 1982 - Ichthyoplankton: Larval distribution and abundance of Engraulidae, Carangidae, Clupeidae, Lut- janidae, Serranidae, Coryphaenidae, Istiophoridae, Xiphii- dae, and Scombridae in the Gulf of Mexico. U.S. Dep. Com- mer., NOAA Tech. Mem., NMFS-SEFC-144. SEAMAP. 1983. (plankton). ASCII characters. Data for 1982. [Fish- eries-independent survey data]/National Marine Fisheries Service; Southeast Fisheries Center: Gulf States Marine Fisheries Commission [producer]. Shaw, R. F., J. H. Cowan, Jr., and T. L. Tillman. 1985. Distribution and density of Brevoortia patronus (Gulf menhaden) eggs and larvae in the continental shelf waters of western Louisiana. Bull. Mar. Sci. 36:96-103. Springer, V. G., and K. D. Woodburn. 1960. An ecological study of the fishes of the Tampa Bay area. Fla. State Board Oonserv., Mar. Lab., Prof. Pap. Ser. 1, 104 p. FISHERY BULLETIN: VOL. 84, NO. 4 Stuck, K. C, and H. M. Perry. 1982. Ichthyoplankton community structure in Mississippi coastal waters. In Fishery Monitoring and Assessment Completion Report, 1 January 1977 to 31 December 1981, p. VI-I-1 thru VI-I-53. Proj. No. 2-296-R, Gulf Coast Research Laboratory, Ocean Springs, MS. Turner, W. R. 1969. Life history of menhadens in the eastern Gulf of Mex- ico. Trans. Am. Fish. Soc. 98:216-224. 1971. Occurrence of Brevoortia gunteri in Mississippi Sound . Q. J. Fla. Acad. Sci. 33:273-274. Walker, H. J., Jr. 1978. Ichthyoplankton survey of nearshore Gulf waters be- tween Barataria Bay and Timbalier Bay, Louisiana, during July, August, and December, 1973. Master's Thesis, Loui- siana State Univ., Baton Rouge, 59 p. Williams, L. W. 1983. Larval fish assemblages of lower Mobile Bay. Master's Thesis, Univ. Southern Alabama, Mobile, 55 P- Wiseman, W. J., Jr., S. P. Murray, J. M. Bane, and M. W. Tubman. 1982. Temperature and salinity variability within the Loui- siana Bight. Contrib. Mar. Sci. 25:109-120. 944 DITTY: ICHTHYOPLANKTON IN NERITIC WATERS Appendix Table.— Seasonality of larval fishes in the northern Gulf of Mexico off Louisiana, January 1981 -December 1982. Taxa Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Neoconger mucronatus .... Muraenidae Gymnothorax sp. Hoplunnis sp. Congridae .... Ophichthidae .... .... .... Bascanichthys bascanium Myrophis punctatus Ophichthus gomesii .... .... Pseudomyrophis 'D' .... Brevoortia patronus Etrumeus teres Harengula jaguana Opisthonema oglinum Sardinella sp. .... Engraulidae Anchoviella perfasciata^ .... Gonostomatidae .... Cyclothone sp. .... Vinciquerria nimbaria .... Synodontidae .... .... Paralepidae .... Lestidiops affinis Myctophidae Centrobranchus nigriocellatus .... Diaphus sp. Diogenichthys atlanticus .... Hygophum sp. .... Lampanyctus sp. .... Gobiesox strumosus Ceratiodei .... .... Antennariidae .... .... Gigantactinidae Bregmaceros cantori Bregmaceros atlanticus Urophycis spp. Ophidiidae .... .... Brotula barbata Lepophidium spp. Ophidion spp. Ophidion welshilgrayi .... .... Ophidion selenops .... .... Exocoetidae .... .... .... Hyporhamphus unifasciatus Atnerinidae .... .... .... Membras martinica Holocentrus sp. .... Macrorhamphosus scolopax Syngnathus spp. .... .... Serranidae Anthinae .... .... Hemanthias leptus .... Grammistinae .... Rypticus maculatus .... Serraninae .... Serraniculus pumilio .... Pomatomus saltatrix .... Carangidae Type A2 .... Carangidae Type B2 .... .... .... Carangidae Type C2 Carangidae Type D2 .... Chloroscombrus chrysurus Oligoplites saurus Selene sp. Trachurus lathami Coryphaena equiselis .... Lutjanus sp. Gerreidae .... Orthopristis chrysoptera . . . . .... Archosargus probatocephalus 945 FISHERY BULLETIN: VOL. 84, NO. 4 Appendix Table.— Continued. Taxa Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Lagodon rhomboides Bairdiella chrysoura Cynoscion arenarius Type A C. arenarius Type B C. nebulosus C. nothus Larimus fasciatus Leiostomus xanthurus Menticirrhus spp. Micropogonias undulatus Pogonias cromis Sciaenops ocellatus Stellifer lanceolatus Mullidae Chaetodipterus faber Labridae Scaridae Mugil cephalus • Mugil curema Sphyraena spp. Blennidae Callionymus pauciradiatus Gobiidae Gobionellus hastatus Microdesmus spp. Diplospinus multistriatus Trichiurus lepturus Auxis sp. Euthynnus alletteratus E. pelamis Scomber japonicus Scomberomorus cavalla S. maculatus Thunnus albacares T. atlanticus T. thynnus Ariomma sp. Cubiceps pauciradiatus Nomeus gronovii Peprilus burti P. paru Scorpaena spp. Prionotus spp. Dactylopterus volitans Bothus sp. Citharichthys sp. Citharichthys sp. Type A Citharichthys sp. Type B Citharichthys sp. Type C C. cornutus C. gymnorhinus C. spilopterus Cyclopsetta sp. Engyophrys senta Etropus crossotus Monolene sessilicauda Paralichthys sp. Syacium sp. S. gunteri S. papillosum Trichopsetta ventralis Achirus lineatus Trinectes maculatus Symphurus spp. Monacanthus setifer Sphoeroides spp. 'Juveniles. 2Morph Type A may represent Decapterus lEIagatis; Type B - Selar crumenopthalamus; Type C - Seriola spp.; Type D - Caranx spp. 946 STOMACH CONTENTS AND FOOD CONSUMPTION ESTIMATES OF PACIFIC HAKE, MERLUCCIUS PRODUCTUS1 Eric A. Rexstad2 and Ellen K. Pikitch3 ABSTRACT Analysis of 466 stomachs of Pacific hake, Merluccius productus, collected during August 1983 off the coasts of Washington and Oregon indicates euphausiids comprise the most important food resource in terms of percent by weight, numbers, and frequency of occurrence for the species at that time of year. The importance of fish in the Pacific hake diet increases with the size of the hake, constituting 87% of the diet by weight in the largest individuals. Weak evidence of a nocturnal feeding pattern was observed. This indistinct nocturnal feeding pattern could have been caused by poor food availability due to El Nino. Estimates of food consumption by Pacific hake indicate that this species may have a substantial impact on some commercially valuable species such as pink shrimp, Pandalus jordani, even though pink shrimp is a fairly minor component of the diet. A statistically significant negative relationship between Pacific hake catch-per-unit-effort (CPUE) and pink shrimp CPUE off the west coast of the United States, using a lag of 2 years, was found. Pacific hake, Merluccius productus, constitute an important component of the California Current ecosystem off the west coast of North America. It is estimated that a standing stock of approx- imately 1.5 million metric tons (t) exists off the Pacific coast between central California and Van- couver Island (Bailey et al. 1982). This biomass represents a substantial prey base for a variety of fish in the ecosystem: great white sharks, Car- charodon carcharias; soupfin sharks, Galeorhinus zyopterus; Pacific electric ray, Torpedo californica; bonito, Sarda chiliensis; albacore, Thunnus alalunga; bluefin tuna, Thunnus thynnus; rock- fishes, Sebastes spp.; sablefish, Anoplopoma fimbria; lingcod, Ophiodon elongatus; dogfish, Squalus acan- thias; and arrowtooth flounder, Atheresthes stomias (Bailey et al. 1982). Pacific hake also constitute a major prey item for a number of marine mammals, including the California sea lion, Zalophus califor- nianus; northern sea lion, Eumetopias jubatus; northern fur seal, Callorhinus ur sinus; saddleback dolphin, Delphinus delphis; Pacific whiteside dolphin, Lagenorhynchus obliquidens; and northern right whale dolphin, Lissodelphis borealis (Fiscus 1979). 'Technical paper No. 7718, Oregon Agricultural Experimental Station. department of Fisheries and Wildlife, Hatfield Marine Science Center, Oregon State University, Newport, OR 97365; present address: Colorado Cooperative Fish and Wildlife Research Unit, Department of Fishery and Wildlife Biology, Colorado State University, Ft. Collins, CO 80523. department of Fisheries and Wildlife, Hatfield Marine Science Center, Oregon State University, Newport, OR 97365. Pacific hake also have an important impact on species below them in the food chain. Best (1963) described Pacific hake as opportunistic feeders. Their diet includes numerous species of Crustacea, particularly euphausiids, several genera of shrimp, crab megalopae, and a variety of fish including Pacific herring, Clupea harengus pallasi; rockfish; sablefish; and flatfish (Livingston 1983). Pacific hake may compete for food resources with a host of other species that feed on the abundant euphausiid resource (Tyler and Pearcy 1975; Karpov and Cailliet 1978; Brodeur and Pearcy 1984), including commercially prized salmonids (Peterson et al. 1982). At the top of the trophic structure is the commer- cial fishing fleet, comprised mainly of foreign joint- venture fishing boats that have harvested, on average, 127,000 t of Pacific hake per year since 1966 (R. C. Francis4). Pacific hake migrate seasonally along the west coast of North America (Swartzman et al. 1983) and spawn in winter in the warm waters off southern California and the Baja peninsula. During the spring and summer, the adults migrate as far north as Van- couver Island to feed. The Pacific hake tend to stratify along the coast by size, with the largest in- dividuals traveling farthest from the spawning areas and smaller juveniles remaining off the coast of California. In autumn, the adults return to the south- ern spawning areas (Bailey et al. 1982). 4R. C. Francis, Fisheries Research Institute, University of Wash- ington, Seattle, WA 98195, pers. commun. May 1985. Manuscript accepted July 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 947 FISHERY BULLETIN: VOL. 84, NO. 4 The pink shrimp, Pandalus jordani, fishery off Oregon was one of the most economically viable fish- eries during the late 1970s with landings in excess of 26,000 t in 1978. Subsequent to that time, pink shrimp landings have declined, with slightly over 2,000 t being landed in 1984 (Saelens and Zirges 1985). The purpose of this study was to describe the dietary habits of the Pacific hake and, in particular, to determine whether predation by Pacific hake on pink shrimp could explain some of the fluctuations seen in pink shrimp landings. MATERIALS AND METHODS In August and September 1983, during the Na- tional Marine Fisheries Service (NMFS) West Coast Groundfish Survey, Pacific hake stomachs were sampled from 41 hauls taken during daylight hours between Coos Bay, OR, and Grays Harbor, WA (Fig. 1). Tows were of 0.5-h duration using a Nor'eastern5 high-opening bottom trawl equipped with roller gear which has an approximate horizontal opening of 13.4 m and vertical opening of 8.8 m. Further details of the sampling regime can be found in Gunderson and Sample (1980) and Weinberg et al. (1984). Between 5 and 15 individuals of each sex from a 5 cm size class (30-34 cm, 35-39 cm, 40-44 cm, 45-49 cm, 50-54 cm, 55 + cm) were sampled from each haul where practical. A total of 466 stomachs were extracted at sea and placed in cheesecloth bags. Stomachs with evidence of regurgitated contents were not included in the sample. Stomachs were preserved in a 10:1 solution of seawater to Formalin. Stomach Content Analysis In the laboratory, stomachs were transferred to ethyl alcohol and examined under a dissecting micro- scope. Stomach fullness and degree of digestion were visually estimated and given a qualitative rating (0-4 from empty to distended, and from un- recognizable to recently consumed). Contents were identified to the lowest taxon and enumerated. Wet weight of each taxon was also determined. Diet composition was characterized by percent of total number of food items (%N), percent of total diet by weight (%VF), and frequency of occurrence in nonempty stomachs (FO). An index of relative im- portance (IRI) was then derived from these values IRI = FO (%N + %W) (Pinkas et al. 1971). The data were further stratified by sex, time of 125 124' 5Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Figure 1.— Stations where Pacific hake stomachs were taken dur- ing 1983 NMFS West Coast Groundfish Survey; 100 and 200 m isobaths are also shown. 948 REXSTAD and PIKITCH: FOOD CONSUMPTION OF PACIFIC HAKE collection (morning, afternoon, and evening), depth of capture (0-100 m, 100-200 m, >200 m), and size. Chi-square tests of homogeneity (Ostle and Mensing 1975) were performed on the frequency of occur- rence data for each prey species in these categories. Consumption Estimates Using the size-specific prey consumption informa- tion derived from this study, Pacific hake popula- tion abundance estimates from the 1983 NMFS survey (Weinberg et al. 1984; Francis fn. 4) and bioenergetics estimates from Francis (1983), trophic calculations were performed to estimate ecosystem- level impacts of prey consumption by Pacific hake in the Columbia INPFC (International North Pacific Fisheries Commission) statistical area in 1983. Biomass estimates were derived from two distinct surveys. The bottom trawl survey estimated the ben- thic component of the population. Details of these estimates can be found in Weinberg et al. (1984). The pelagic component of the population was esti- mated by hydroacoustic methods. Size composition of the pelagic segment of the population was esti- mated from companion midwater trawls conducted from the hydroacoustic vessel. Biomass estimates for each of the five size classes sampled were deter- mined from estimated numbers in each centimeter size interval and length-weight regressions (Fran- cis fn. 4). Using a mean body weight for each size class, the percent of total body weight consumed daily was calculated based on the equations of Francis (1983). This total biomass consumption was then broken down into the constituent prey categories found in the stomachs of fish sampled using the percent of the diet by weight. These calculations were repeated for each of the five classes, and both the pelagic and benthic components of the population, to derive daily consumption estimates. Residence times provided by Francis (1983) for each age class within each statistical area were con- verted to residence time by size class to account for the migratory behavior of Pacific hake. This pro- vided consumption rate estimates summed over the length of time Pacific hake are found in the Colum- bia statistical area. An example of the calculations used to estimate total consumption of each prey item category is shown in Table 1. Pacific Hake-Pink Shrimp Interaction The relationship between the abundance of Pacific hake and pink shrimp was examined via regression Table 1.— Calculations used to compute total consumption of Thysanoessa spinifera. Column E1 = A x B/100 x C/1 00. Column E2 = E1 x D. Biomass is combined benthic and pelagic com- ponents of the population, BWD is percent body weight consumed per day, W is percent of the diet by weight composed of T. spini- fera, and Days is number of days each size class resides in the Columbia INPFC Area. Note total biomass differs from value given in text due to biomass of population <35 cm in length. (A) (B) (C) (D) (E1) (E2) Size Consumption class Biomass Daily Seasonal (cm) (1,000 t) %BWD %W Days (1,000 t) 35-39 331.252 1.10 14.7 80 0.540 42.85 40-44 69.126 0.98 9.4 69 0.060 4.20 45-49 21.272 0.84 17.2 45 0.030 1.38 50-54 17.640 0.65 14.9 42 0.020 0.72 55 + 8.936 0.40 6.4 41 0.002 0.09 Total 448.226 0.652 49.24 analysis. Data from Francis et al. (unpubl. data) on Pacific hake catches in U.S. waters from 1967 to 1982 were converted to catch per unit effort (CPUE) based on the number of days of effort of foreign stern-trawling factory ships (BMRTs). Pounds per hour of pink shrimp taken in the equivalent of single- rigged shrimp trawls (SRE) in California, Washing- ton, and Oregon from 1968 to 1984 (Saelens and Zirges 1985) were used as the dependent variable in regression analyses. Two regressions were performed. The first used hake CPUE in year i to predict shrimp CPUE in year i, while the second involved a 2-yr lag (i.e., Pacific hake CPUE in year i versus shrimp CPUE in year i + 2). RESULTS Stomach Content Analysis A breakdown of the stomach contents by size class of Pacific hake is presented in Table 2. Euphausiids dominate the diet of small hake while decapods and fish become increasingly important as Pacific hake increase in size. Considering percent of the diet by weight, the importance of euphausiids monotonically decreases from 100 to 7.9% with increasing predator size. Likewise, the importance of fish rises from 0 to 87.1% with increasing predator size. Pink shrimp comprise only a minor portion of the diet, the largest percentage being 4.9% for the largest size class. Commercially important herring comprise nearly one-third of the diet of the larger size classes. A previously unreported prey item, the ghost shrimp, Callianassa sp., appeared in the diet of the Pacific hake sampled in this study. These burrow- 949 FISHERY BULLETIN: VOL. 84, NO. 4 Table 2.— Summary of stomach contents of Merluccius productus collected during the 1983 NMFS Pacific Coast Groundfish Survey. T = <0.1%. 30-34 cm 35-39 cm 40-44 cm 45-49 cm 50-54 cm >55 cm Prey category FO1 %N2 %W3 FO %N %W FO %N %W FO %N %W FO %N %W FO %N %W Euphausiacea Thysanoessa spinifera — — — 14.9 15.0 14.7 27.5 10.6 9.4 32.5 79.6 17.2 52.4 65.1 14.9 61.9 62.3 6.4 Euphausia pacifica 25.0 16.7 17.1 56.4 73.7 72.7 68.1 78.5 70.8 10.4 9.0 2.2 19.5 18.7 3.9 33.3 23.4 1.3 Unidentified 100.0 83.3 82.9 33.0 10.4 6.2 33.3 10.3 10.7 31.2 9.2 2.5 35.4 12.4 1.9 19.0 9.0 0.2 Decapoda Pandalus jordani — — — 1.1 T 0.2 — — — 1.3 0.1 0.1 — — — 19.0 2.3 4.9 Sergestes similis — — — 2.1 0.7 2.9 5.8 0.3 2.5 — — — — — — — — — Pasiphaea pacifica — — — — — — 1.5 T 2.0 — — — — — — 14.3 1.5 0.4 Crangon sp. — — — — — — — — — 2.6 0.1 0.2 4.9 1.1 12.0 — — — Callianassa sp. ___ ______ 11.7 o.7 13.4 8.5 1.1 14.5 — — — Osteichthyes Engraulis mordax — — — — — — — — — 1.3 0.1 1.5 2.4 0.3 2.7 — — — Clupea harengus — — — — — — 2.9 0.1 3.8 3.9 0.2 34.7 3.7 0.2 28.4 — — — Thaleichthys pacificus — — — — — — 1.4 T 0.6 9.1 0.4 23.0 1.2 0.1 T — — — Osmeridae — — — — — — — — — 1.3 0.1 T 1.2 0.1 2.8 — — — Gadidae — — — ———— — — 1.3 0.1 1.2 2.4 0.2 2.4 4.8 0.2 83.6 Pleuronectidae — — — — — — — — — 1.3 0.2 0.1 2.4 0.2 14.5 9.5 0.6 3.4 Agonidae ___ ___ ___ ___ 1.2 0.1 0.4 — — — Myctophidae ___ 0.8 T 1.5 — — —— — —— — —— — _ Unidentified — — — 5.3 0.2 1.8 5.8 0.2 0.3 9.1 0.4 3.6 7.3 0.5 1.8 14.3 0.6 0.1 Number of stomachs (empty) 11(7) 120(26) 97(28) 118(41) 93(11) 27(6) Number of prey items 6 2,006 2,029 1,921 1,206 478 Weight of stomach contents (g) 0.4 77.0 78.0 376.5 319.3 293.0 1 Frequency of occurrence in non-empty stomachs. 2Percent of diet by number of items. 3Percent of diet by weight of stomach contents. ing animals were found in stomachs of Pacific hake taken at towing stations between 8.3 and 15.6 km (4.5 and 8.5 mi) offshore but not in immediate prox- imity to estuaries where ghost shrimp are most often found. Chi-square tests (Table 3) illustrate the patterns in prey consumption by various stratifications of the data. There was little statistical difference in stomach contents of males compared with females. The analysis of prey categories by depth is essen- tially an inshore-offshore comparison as isobaths run roughly parallel to the coastline in the study area. Statistically significant differences were found in depth of capture for both species of euphausiids found in this study. Thysanoessa spinifera was more important in the diet of fish taken close to shore whereas Euphausia pacifica was important for fish taken futher offshore. Eulachon, Thaleichthys pacif- icus, was found in stomachs more often in shallow waters than at depth. These animals, being anad- romous, are often found in bays and estuaries, i.e., close to shore. A significant difference exists in the presence of the two species of euphausiids in stomachs collected at different times of the day. The data collected in this study show that T. spinifera were seldom found in stomachs collected after 1600 h while E. pacifica Table 3.— Chi-square analysis of difference in stomach content by prey category and various factors. Factor Sex Depth Time Size Prey category df = 1 df = 2 df = 2 df = 4 Thysanoessa spinifera 2.48 15.65*** 23.67*** 36.69*** Euphausia pacifica 1.42 17.45*** 6.23* 76.90*** Pandalus jordani 0.48 1.02 3.53 39.60*** Sergestes similis 0.02 28.07*** 0.12 9.86* Pasiphaea pacifica 0.80 17.47*** 3.92 34.39*** Crangon sp. 0.02 6.40* 3.07 8.27 Callianassa sp. 0.05 3.69 9.14* 20.30*** Engraulis mordax 0.46 0.68 3.87 3.99 Clupea harengus 0.76 3.95 10.14** 4.30 Thaleichthys pacificus 0.72 9.35** 4.14 16.71** Osmeridae 2.24 1.44 2.49 2.33 Gadidae 0.01 0.16 2.20 5.44 Pleuronectidae 4.55* 1.45 2.21 12.49* ' = P<, 0.05, = P*S 0.01, *** = P< 0.001. were often found in stomachs collected during that time (Fig. 2a). To further examine the diel feeding pattern of Pacific hake, the percent of all stomachs in each of two fullness categories (<25% full; > 75% full) was calculated by time of day. A three-point moving average was computed for each fullness category, and the resulting averages plotted (Fig. 3). There 950 REXSTAD and PIKITCH: FOOD CONSUMPTION OF PACIFIC HAKE CD O c 05 O Q. E "cd DC X CD ~o c S> Prey Consumption by Time of Day - 70 1 60 CD 50 ■ 05 ■4— « L. O Q_ 40 ■ E "CD 30 ■ DC X CD "D 20 - _C 10 ■ >> 0 A Thysanoessaspinifera B Euphausia pacifica C Pandalus Jordan i D Sergestessp. E Pasiphaea pacifica F C rang on sp. G Callianassasp. H Engraulis mordax I Clupeaharengus J Thaleichthys pacificus K Osmeridae L Gadidae M Pleuronectidae 800-1200 1200-1600 1600-2000 jj_* t^TU i-l- Ixl - fca A B C D E F G — . n S. 1 Prey Category Prey Consumption by Size Class of Predator 100 - 2 80 - 60 - 40 - 20 ~ff . - j n .L I A Thysanoessaspinifera B Euphausia pacifica C Pandalus jordani D Sergestessp. E Pasiphaea pacifica F Crangonsp. G Callianassasp. H Engraulis mordax I Clupeaharengus J Thaleichthys pacificus K Osmeridae L Gacf/c/ae M Pleuronectidae 35-39cm 40 -44cm 45 -49cm 50 -54cm 55 + cm jfl *_[]_ ■ I - ji K L M JI A B D E F G H I Prey Category K L M Figure 2.— Index of relative importance for major prey categories by a) time of collection and b) size of Pacific hake. Square root transformation used for scaling purposes. 951 FISHERY BULLETIN: VOL. 84, NO. 4 100* Percent Stomachs Mostly Full/Empty 75* c V o 1_ V 0- 50* - 25* 0* Stomochs less than 25* full 54 40 80 31 30 25 59 50 15 34 — i 1 1 1 1 1 — 10 11 12 13 14 15 Time (hours) 16 17 18 19 Figure 3.— Diel pattern of stomach contents of Pacific hake as demonstrated by percent of stomachs <25% full (upper curve) and >75% full (lower curve). Three-point moving average used to smooth the curves. Sample sizes shown above x-axis. is a weak indication that these fish exhibit a pattern of feeding more heavily at night than during the day. For a predator feeding nocturnally, the expected pattern of this curve would be low percentages of empty stomachs early and late in the day, and high percentages of empty stomachs at midday. No tows were made between the hours of 2000 and 0700 thus direct evidence of nocturnal feeding was not avail- able. Comparison of stomach contents by size class showed the greatest amount of variation (Table 3, Fig. 2b) because of the shift in diet composition from euphausiids in early life stages to fishes in later stages. The estimated consumption by Pacific hake in the Columbia statistical area over all prey categories is 4,651 t/d (Table 4). The amount of euphausiids con- sumed (over 4 kt/d), exceeds that of all other prey categories combined, but several commercially valuable species are also consumed in significant quantities. Consumption of pink shrimp is estimated at over 9.2 t/d, and almost 120 t/d of herring are consumed. Residence time for each size class of Pacific hake was derived from data presented by Francis (1983) (size class 1: 80 d; size class 2: 69 d; size class 3: 45 d; size class 4: 42 d; and size class 5: 41 d) to extrapolate estimates of annual prey con- sumption from the daily consumption rate in the Columbia area. The annual consumption of pink shrimp, based on these data, is estimated at 659.3 1. Pacific Hake-Pink Shrimp Interaction The regression of Pacific hake CPUE versus pink shrimp CPUE resulted in a nonsignificant correla- tion (r2 = 0.114, df = 15, P = 0.185). However, the regression performed with a 2-yr lag (hake CPUE in year i versus shrimp CPUE in year i + 2) showed a significant negative correlation between the variables (r2 = 0.418, df = 15, P = 0.005). Note that the significance of the latter analysis stems largely from data obtained in recent years (Fig. 4). DISCUSSION One of the most striking patterns found in the data is the distinct change in diet composition that Pacific 952 REXSTAD and PIKITCH: FOOD CONSUMPTION OF PACIFIC HAKE Table 4.— Diet composition by size class on a daily basis (t) and on seasonal basis (kt). Values based on biomass tor the Columbia INPFC area estimated from bottom trawl survey (Weinberg et al. 1984) and hydroacoustic survey (Francis, see text fn. 4). T - <0.1 t/d or 0.05 kt seasonally. Size 1 Size 2 Si ze 3 Size 4 Size 5 Totals Prey category Daily Season Daily Season Daily Season Daily Season Daily Season Daily Season Euphausiacea Thysanoessa spinifera 535.1 42.5 64.0 4.4 30.6 1.4 17.2 0.7 2.3 .0.1 649.2 49.1 Euphausia pacifica 2,646.4 210.4 481.8 33.2 3.9 0.2 4.5 0.2 0.5 T 3,137.1 244.0 Unid. euphausiid 225.7 17.9 72.8 5.0 4.5 0.2 2.2 0.1 0.1 T 305.2 23.3 Total euphausiid 3,407.3 270.9 618.6 42.7 39.0 1.7 23.9 1.0 2.8 0.1 4,091.6 316.4 Decapoda Pandalus jordani 7.3 0.6 0.0 0.0 0.2 T 0.0 0.0 1.8 0.1 9.2 0.7 Sergestes sp. 105.6 8.4 17.0 1.2 0.0 0.0 0.0 0.0 0.0 0.0 122.6 9.6 Pasiphaea pacifica 0.0 0.0 13.6 0.9 0.0 0.0 0.0 0.0 0.1 T 13.8 1.0 Crangon sp. 0.0 0.0 0.0 0.0 0.4 T 13.8 0.6 0.0 0.0 14.2 0.6 Callianassa sp. 0.0 0.0 0.0 0.0 23.9 1.1 16.7 0.7 0.0 0.0 40.6 1.8 Osteichthyes Engraulis mordax 0.0 0.0 0.0 0.0 2.7 0.1 3.1 0.1 0.0 0.0 5.8 0.2 Clupea harengus 0.0 0.0 25.9 1.8 61.8 2.8 32.8 1.4 0.0 0.0 120.5 5.9 Thaleichthys pacificus 0.0 0.0 4.1 0.3 41.0 1.8 0.1 T 0.0 0.0 45.1 2.1 Osmeridae 0.0 0.0 0.0 0.0 0.4 T 3.2 0.1 0.0 0.0 3.6 0.2 Gadidae 0.0 0.0 0.0 0.0 2.1 0.1 2.8 0.1 30.2 1.2 35.1 1.4 Pleuronectidae 0.0 0.0 0.0 0.0 0.2 T 16.7 0.7 1.2 0.1 18.1 0.8 Agonidae 0.0 0.0 0.0 0.0 0.0 0.0 0.5 T 0.0 0.0 0.5 T Myctophidae 54.6 4.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 54.6 4.3 Unid. fish 65.5 5.2 2.0 0.1 6.4 0.3 2.1 0.1 T T 76.1 5.7 Grand total 3,640.2 289.4 681.2 47.0 178.0 8.0 115.7 4.8 36.2 1.5 4,651.4 350.7 Shrimp CPUE (year i+2), Hake CPUE (year i) 800 700 600 in 500 UJ z> 0. o 400 a. E If) 300 - 200 - 100 D 75-77 68-70 D D 66-68 70-72 a D 73-75 - D 71-73 76-78 a 67-69 D 69-71 □ □ 72-74 74-76 a - 77-79 □ i i , . ... ,_, 80-82 79-81 u 81-83 -, P- 78-80 D 82-84 □ r 1 12 16 20 24 Hake CPUE (t/BMRT day) 28 32 36 Figure 4.— Pink shrimp CPUE in year i + 2 (y-ax\s) plotted against Pacific hake CPUE in year i (z-axis). Regression expression is y = 1029 - 23.23x (r2 = 0.418). Numbers on the plot represent the years of the CPUE data. 953 FISHERY BULLETIN: VOL. 84, NO. 4 hake undergo as they increase in size. Thysanoessa spinifera appears to be more important to larger hake whereas Euphausia pacifica is more important to smaller individuals. Pink shrimp and glass shrimp, Pasiphaea pacifica, were consumed almost exclu- sively by fish >55 cm. Eulachon and pleuronectids were also predominantly consumed by larger hake. Cannibalism was also observed among larger individuals. Diel Feeding Pattern A number of previous researchers have postulated that species in the genus Merluccius exhibit a diel feeding pattern (Outram and Haegele 1972; Bow- man and Bowman 1980). Alton and Nelson (1970) as well as Livingston (1983) described Pacific hake as nocturnal predators that migrate vertically to feed near the surface during hours of darkness and dive to deeper water during daylight hours. Brin- ton (1967) and Alton and Blackburn (1972) showed that this same vertical migration pattern exists for the two species of euphausiids found in this study. If the Pacific hake follow the euphausiids on their vertical diel migration, the expectation is that the relative proportion of both species of euphausiids in the diet should not vary significantly by time of day. As reported above, our findings conflict with this expectation. To further examine this apparent deviation, we considered potentially confounding factors; such as differences in the distributions of the euphausiid species and various size classes of Pacific hake. Brin- ton (1962) reported that T. spinifera is a neritic species and E. pacifica is a more oceanic species. Analysis of length-frequency data from the cruise during which this study was conducted shows that Pacific hake of different size classes segregate by depth. Pacific hake <40 cm in length made up 37% of the catch in <100 m of water, but these same size classes comprised 62% of the catch taken in MOO m of water. Hence, the smaller individuals were found in greater abundance in the habitat associated with E. pacifica. This phenomenon of smaller fish occurring in deeper water and consequently consuming greater quantities of E. pacifica explains the apparent dif- ference in importance of the two species of euphau- siids by time of day (Fig. 2a). Only 7% of the non- empty stomachs taken before 1200 h were from fish <40 cm in length whereas of the fish sampled after 1600 h, 34% were <40 cm in length. Thus, we regard the observed differences in consumption of T. spini- fera and increasing importance of E. pacifica by time of day as spurious, confounded by the differ- ences in the diets and distributions of various size classes of Pacific hake. This study coincided with the strong presence of El Nino in 1983 which may have altered the nor- mal migration pattern of Pacific hake and conse- quently the residence time estimates, and may also have affected the abundance of the prey base. Hence, there may be some error in the consump- tion estimates presented herein. Miller et al. (1984) noted a decline in the relative abundance of T. spinifera off the Oregon coast during 1983 in com- parison with other years. Thus, feeding to satiation during evening hours may have been impossible; consequently, feeding occurred whenever euphau- siids were encountered. Additional circumstantial evidence of aberrant feeding behavior of Pacific hake in 1983 is their severely depressed growth (Francis and Hollowed 1985). Food resources may only have been sufficient for maintenance metabo- lism with little energy remaining for growth. These observations may explain why the diel feeding pat- tern observed was weak. Trophic Interaction The seasonal migration pattern and consequent latitudinal stratification of Pacific hake stocks by size class makes it difficult to compare food habit studies conducted at different times of the year and at different locations on the Pacific coast. Nonethe- less, examining only the role of pink shrimp in the diet, we find first mention of Pacific hake preying on this species by Gotshall (1969a, b). Analyzing Pacific hake stomachs collected off California be- tween 1966 and 1969, Gotshall found high incidences (54% frequency of occurrence) of pink shrimp dur- ing late summer and early fall, particularly in Pacific hake collected over shrimp beds. The study was an attempt to use Pacific hake as biological samplers to estimate pink shrimp abundance, focusing sam- pling effort on known pink shrimp beds, and, as such, the sampling design was quite different from other studies. Outram and Haegele (1972) reported that 3% of the Pacific hake stomachs collected off the coast of British Columbia contained pink shrimp. Pink shrimp were found in 5.7% of the Pacific hake stomachs collected during the summers of 1965 and 1966 of f Washington and Oregon (Alton and Nelson 1970). Livingston and Alton (1982) found that pan- dalid shrimp constituted 0.3% by weight of the con- tents of the 1,430 stomachs of Pacific hake taken off the coasts of Washington and Oregon during the 954 REXSTAD and PIKITCH: FOOD CONSUMPTION OF PACIFIC HAKE summer of 1967. From 204 stomachs collected dur- ing the 1980 NMFS West Coast Groundfish Survey off the coasts from Oregon to Vancouver Island, Livingston (1983) found pink shrimp constituted 0.7% by weight of the Pacific hake diet. Pink shrimp occurred in 1.7% of the Pacific hake stomachs col- lected in the study described in this paper. Thus, with the exception of Gotshall's work, studies of the food habits of Pacific hake have shown pink shrimp generally comprise well under 10% of the Pacific hake diet, and thus do not appear to be an impor- tant food source for hake. However, due to the large biomass of Pacific hake in the North Pacific, it is possible that Pacific hake may represent a signifi- cant source of mortality even for those species, in- cluding pink shrimp, that are not significant com- ponents of the Pacific hake diet (Francis 1983). The estimated consumption of 659.3 t/season of pink shrimp compares with a commercial catch of 2,197 1 of pink shrimp landed in Oregon during 1984 by 59 vessels (Saelens and Zirges 1985). It is con- ceivable that the magnitude of Pacific hake preda- tion on pink shrimp may increase in the near future. Small Pacific hake, preying mainly on euphausiids, constituted the bulk of the consumers in this study. The strong 1980 year class of Pacific hake, seen as the 35-39 cm size class in these 1983 data, will have substantially greater impact on commercially valu- able species upon reaching larger sizes when these valuable species comprise a larger fraction of the diet. Francis (1983) inferred, from catch statistics of Pacific hake and pink shrimp, that increased catches of Pacific hake since the inception of the foreign and subsequent joint- venture fisheries may have con- tributed to the dramatic increase in the landings of pink shrimp during the late 1970s. The causal mechanism inferred is the release of predation pressure on the pink shrimp population as a result of decreased Pacific hake abundance due to fishing. This "surplus" in the pink shrimp population was harvested by the increasingly vigorous shrimp fishery. This contention is disputed by Livingston and Bailey (1985). Their analysis focuses on pink shrimp CPUE during two time periods: 1952-65 during which Pacific hake were unexploited and 1966-77 during which a substantial joint-venture fishery occurred. They found no appreciable change in aver- age pink shrimp CPUE between the two periods. Extending their analysis to include the most recent catch statistics, we also fail to find the existence of a significant difference between the periods 1957-65 and 1966-84 (t = 1.05, 26 df, P = 0.303). However, if pink shrimp have constituted a fair- ly constant proportion of the Pacific hake diet over time, as suggested by this and previous Pacific hake food habit studies, then there may indeed be a rela- tionship between the release of predator pressure by the Pacific hake and increased catches of pink shrimp. The regression-correlation analysis pre- sented above has an advantage over the average pink shrimp CPUE analysis because it incorporates information about both hake and shrimp abun- dances. The regression-correlation results provide weak statistical support to Francis' contention that there is a relationship between Pacific hake and pink shrimp population dynamics. However, further ob- servations are needed to obtain greater confidence in this relationship. In particular, it will be interest- ing to note that the impact of the strong 1980 year class Pacific hake on pink shrimp catches in the near future. CONCLUSION Pacific hake occupy a unique trophic position, serving not only as predators but also as prey for a variety of species carrying valuations other than those of an economic nature (endangered species and species managed under the Marine Mammal Protection Act). Euphausiids constitute the primary source of food for Pacific hake in the North Pacific. However, as Pacific hake mature, euphausiids decrease in importance and fish take on greater im- portance. Owing to the vast quantity of hake bio- mass living in the North Pacific, it has been shown that Pacific hake may consume large quantities of several commercially valuable species, even though these species comprise a fairly small percentage of the diet. It has also been demonstrated that a statistically significant relationship exists between CPUE of Pacific hake and pink shrimp. Additional years of data are required to have a clearer under- standing of this relationship. ACKNOWLEDGMENTS Bill Barss and Mark Saelens of the Oregon Department of Fish and Wildlife (ODFW) assisted in the collection of the Pacific hake stomachs at sea and with the identification of decapods. Leslie Lutz of ODFW helped with the laboratory analysis and Rick Brodeur, Chris Wilson, and Bruce Mundy of Oregon State University aided in the identification of fish remains. Rick Brodeur also gave suggestions on statistical analysis. Helpful comments were pro- vided by Mac Zirges, Mark Saelens, Robert Fran- 955 FISHERY BULLETIN: VOL. 84, NO. 4 cis, Rick Brodeur, Chuck Harding, Barb Knopf, Chris Wilson, David Erickson, and two anonymous reviewers. Robert Francis, Northwest and Alaska Fisheries Center, NMFS, provided data on histori- cal Pacific hake catch data and hydroacoustic survey estimates. This publication is the result, in part, of research sponsored by NOAA, Office of Sea Grant, Department of Commerce, under contract No. NA81AA-D-00086 (Project No. R/OFP-20), and by the Oregon Department of Fish and Wildlife. LITERATURE CITED Alton, M. S., and C. J. Blackburn. 1972 . Diel changes in the vertical distribution of the euphau- siids, Thysanoessa spinifera Holmes and Euphausia pacifica Hansen in coastal waters of Washington. Calif. Fish Game 58:179-190. Alton, M. S., and M. 0. Nelson. 1970. Food of Pacific hake, Merluccius productus, in Wash- ington and northern Oregon coastal waters. In Pacific hake, p. 35-42. U.S. Fish Wildl. Serv., Circ. 332. Bailey, K. M., R. C. Francis, and P. R. Stevens. 1982. The life history and fishery of Pacific whiting, Merluc- cius productus. Calif. Coop. Oceanic Fish. Invest. Rep. 23: 81-98. Best, E. A. 1963. Contribution to the biology of the Pacific hake, Merluc- cius productus (Ayres). Calif. Coop. Oceanic Fish. Invest. Rep. 9:51-56. Bowman, R. E., and E. W. Bowman. 1980. Diurnal variation in the feeding intensity and catch- ability of silver hake {Merluccius bilinearis). Can. J. Fish. Aquat. Sci. 37:1565-1572. Brinton, E. 1962. The distribution of Pacific euphausiids. Bull. Scripps Inst. Oceanogr., Univ. Calif. 8(2):51-269. 1967. Vertical migration and avoidance capability of euphau- siids in the California Current. Limnol. Oceanogr. 12: 451-483. Brodeur, R. D., and W. G. Pearcy. 1984. Food habits and dietary overlap of some shelf rock- fishes (genus Sebastes) from the northeastern Pacific Ocean. Fish. Bull., U.S. 82:269-293. Fiscus, C. H. 1979. Interactions of marine mammals and Pacific hake. Mar. Fish. Rev. 41(10):l-9. Francis, R. C. 1983. Population and trophic dynamics of Pacific hake (Merluccius productus). Can. J. Fish. Aquat. Sci. 40:1925- 1943. Francis, R. C, and A. Hollowed. 1985. Status of the Pacific hake resource and recommenda- tions for management in 1985. Status of stocks document to Pacific Fisheries Management Council, Portland, OR. Gotshall, D. W. 1969a. Stomach contents of Pacific hake and arrowtooth flounder from northern California. Calif. Fish Game 55: 75-82. 1969b. The use of predator food habits in estimating relative abundance of the ocean shrimp, Pandalus jordani Rathbun. FAO Fish. Rep. 57, p. 667-685. GUNDERSON, D. R., AND T. M. SAMPLE. 1980. Distribution and abundance of rockfish off Washington, Oregon, and California during 1977. Mar. Fish. Rev. 42 (3-4):2-16. Karpov, K. A., and G. M. Cailliet. 1978. Feeding dynamics of Loligo opalescens. In C. W. Recksiek and H. W. Frey (editors), Biological, oceanograph- ic, and acoustic aspects of the market squid, Loligo opales- cens Berry, p. 45-65. Calif. Dep. Fish Game, Fish Bull. 169. Livingston, P. A. 1983. Food habits of Pacific whiting, Merluccius productus, off the west coast of North America, 1967 and 1980. Fish. Bull. U.S. 81:629-636. Livingston, P. A., and M. S. Alton. 1982. Stomach contents of Pacific whiting off Washington and Oregon, April-July 1967. U.S. Dep. Commer., NOAA Tech. Memo, NMFS F/NWC-32. Livingston, P. A., and K. M. Bailey. 1985. Trophic role of the pacific Whiting, Merluccius pro- ductus. Mar. Fish. Rev. 47(2):16-22. Miller, C. B., H. P. Batchelder, R. D. Brodeur, and W. G. Pearcy. 1984. Response to the zooplankton and ichthyoplankton off Oregon to the El Nino event of 1983. In W. S. Wooster and D. L. Fluharty (editors), El Nino North: Nino effects in the eastern subarctic Pacific Ocean, p. 185-187. Wash. Sea Grant Prog., Univ. Wash., Seattle. Ostle, B., and R. W. Mensing. 1975. Statistics in research. 3d ed. Iowa State Univ. Press, Ames, 596 p. OUTRAM, D. N., AND C. HAEGELE. 1972. Food of Pacific hake (Merluccius productus) on an off- shore bank southwest of Vancouver Island, British Colum- bia. J. Fish. Res. Board Can. 29:1792-1795. Peterson, W. T., R. D. Brodeur, and W. G. Pearcy. 1982. Food habits of juvenile salmon in the Oregon coastal zone, June 1979. Fish. Bull., U.S. 80:841-851. Pinkas, L., M. S. Oliphant, and I. L. K. Iverson. 1971. Food habits of albacore, bluefin tuna, and bonito in California waters. Calif. Dep. Fish Game, Fish Bull. 152: 1-105. Saelens, M. R., and M. H. Zirges. 1985. The 1984 Oregon shrimp fishery. Oreg. Dep. Fish Wildl. Inf. Rep. 85-6. Swartzman, G. L., W. M. Getz, R. C. Francis, R. T. Haar, and K. Rose. 1983. A management analysis of Pacific whiting (Merluccius productus) fishery using an age-structured stochastic recruit- ment model. Can. J. Fish. Aquat. Sci. 40(4):524-539. Tyler, H. R., Jr., and W. G. Pearcy. 1975. The feeding habits of three species of lanternfishes (family Myctophidae) off Oregon, USA. Mar. Biol. (Berl.) 32:7-11. Weinberg, K. L., M. E. Wilkins, and T. A. Dark. 1984. The 1983 Pacific west coast bottom trawl survey of groundfish resources: estimates of distribution, abundance, age and length composition. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-70. 956 DIET OF NORTHERN FUR SEALS, CALLORHINUS URSINUS, OFF WESTERN NORTH AMERICA Michael A. Perez1 and Michael A. Bigg2 ABSTRACT Data recorded from the stomach contents of 18,404 northern fur seals, Callorhinus ursinus, mostly females aged >3 years collected off western North America during 1958-74, were analyzed to determine the relative importance of each prey species by region, subregion, and month. When weighted for energy content, the primary food species were small schooling fishes. Between western Alaska and California from December to August the most significant prey species were northern anchovy, Engraulis mordax (20%); Pacific herring, Clupea harengus pallasi (19%); capelin, Mallotus villosus (8%); Pacific sand lance, Ammodytes hexapterus (8%); Pacific whiting, Merluccius productus (7%); salmon, Oncorhynchus spp. (6%); Pacific saury, Cololabis saira (4%); and rockfishes, Sebastes spp. (4%). Other food species eaten in this area consisted of a wide variety of squids (17%) and other fishes (7%). In the eastern Bering Sea the main prey species from June to October were juvenile walleye pollock, Theragra chalcogramma (35%); capelin (16%); Pacific herring (11%); and squids, Berryteuthis magister and Gonatopsis borealis, which comprise most (30%) of the remaining diet of northern fur seals in this region. In all areas off western North America, fishes were the main food species of these pinnipeds in neritic waters, while squids were the most important prey in oceanic waters. Typically three prey species comprised 80% of their diet in any one area, although the composition of the diet varied in type and importance by region and month. The northern fur seal, Callorhinus ursinus, is found in the Bering Sea, Sea of Okhotsk, and throughout the North Pacific Ocean, north of approximately lat. 32 °N off western North America and lat. 36 °N off Asia (Baker et al. 1970; Fiscus 1978). Although its pelagic distribution is extensive, the main concen- trations lie over the continental shelf. There are three main stocks of this species. The largest stock breeds on the Pribilof Islands in the eastern Bering Sea and migrates primarily to coastal waters be- tween the Gulf of Alaska and California. The other two stocks breed on the Commander Islands in the western Bering Sea and on Robben Island off north- ern Japan. Both stocks migrate primarily along the Asian coast. To determine the diet of the Pribilof Islands population, the United States and Canada, under the auspices of the North Pacific Fur Seal Commission, conducted annual pelagic studies dur- ing 1958-74 to collect stomach contents and other biological information. The results of research on the diet of northern fur seals by the United States and Canada during 1958-74 have been presented in many annual and 2-6 yr summaries submitted by each country to the North Pacific Fur Seal Commission. Kajimura (1984) cited most of these reports. Spalding (1964), Stroud et al. (1981), and Kajimura (1985) also published reports on diet collected since 1958. Studies on the food habits of the northern fur seal prior to 1958 include Lucas (1899), Clemens and Wilby (1933), Clemens et al. (1936), Schultz and Rafn (1936), May (1937), Wilke and Kenyon (1952, 1954, 1957), Taylor et al. (1955), and Kenyon (1956). Investigations to date have reported that north- ern fur seals eat a wide variety of fishes and squids. However, the relative importance of each prey species has remained uncertain because substantial differences often existed between values of relative importance derived by volumetric measure and those derived by frequency of occurrence. For ex- ample, squids were important (averaging 39%) in the diet using frequency of occurrence but not significant (15%) using volume (Bigg and Fawcett 1985; Perez and Bigg3). The long-suspected reason for this difference was that squid beaks accumulated Northwest and Alaska Fisheries Center, National Marine Mam- mal Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA 98115. department of Fisheries and Oceans, Pacific Biological Station, Nanaimo, British Columbia V9R 5K6, Canada. 3Perez, M. A., and M. A. Bigg. 1980. Interim report on the feeding habits of the northern fur seal in the eastern North Pacific Ocean and eastern Bering Sea. In H. Kajimura, R. H. Lander, M. A. Perez, A. E. York, and M. A. Bigg, Further analysis of pelagic fur seal data collected by the United States and Canada during 1958-74, Part 2, p. 4-172. Unpubl. rep. Northwest and Alaska Fisheries Center, National Marine Mammal Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA 98115. Manuscript accepted July 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 957 FISHERY BULLETIN: VOL. 84, NO. 4 in stomachs of seals thereby inflating the importance of squid (Scheffer 1950; Spalding 1964; Bigg and Fawcett 1985). Recent experimental studies con- firmed that squid beaks accumulate in fur seal stomachs (Bigg and Fawcett 1985). To date, no reports have been published on the diet of northern fur seals that take this bias into ac- count. However, Bigg and Perez (1985) suggested a method, called modified volume, which reduces the bias and also accounts for differences in digestion rates between fish and squid. In this method, evidence of diet based on trace remains, such as squid beaks and fish bones, is omitted in the analyses, and a combination of the frequency of oc- currence and volumetric methods is used to estab- lish the relative importance of individual prey species. We use the modified volume method in this report to analyze data from the stomach contents collected by the United States and Canada during 1958-74. We will describe the annual diet of northern fur seals in the eastern North Pacific and eastern Bering Sea by region and subregion. We also incorporate the energy content of important prey species to deter- mine whether this might affect relative importance, a procedure not tried previously with this seal. METHODS Lander (1980) and Kajimura (1984, 1985) de- scribed the methods used to take northern fur seals at sea during 1958-74 and to identify- and measure the prey items found in their stomachs by volume and frequency of occurrence. A total of 18,404 stomachs were collected of which 7,373 contained food and an additional 3,326 had only trace remains. Perez and Bigg (fn. 3) summarized the data on volume and frequency of occurrence for all species of northern fur seal prey by month and region. Perez and Bigg4 and Bigg and Perez (1985) gave a detailed discussion of the procedure used to cal- culate modified volume values. First, prey species represented in any stomach only by trace amounts (<10 cc) were omitted. Second, the proportions of total fish and total squid in the diet by subregion, region, and month were then determined by non- trace frequency of occurrence. Third, the ratio of each species within only the fish category and within only the squid category was determined by volume. The taxonomic groupings recorded in the original data which overlapped each other were either pooled with higher taxa or were proportionally divided among component species depending upon which level of taxa had the most data. This prevented food groupings from being partially compared against themselves. Next, the volumetric ratios for in- dividual fish and squid species were adjusted to sum, respectively, to the total proportion of fish and squid in the diet. Finally, all values were readjusted to total 100%. The relative importance of prey species has been presented in this report in two ways: 1) modified volume values for each region by month, and for each subregion with data from all months pooled; and 2) modified volume values for each region based on combined months data which were weighted for Table 1 .—Estimated energy values (wet mass) for important north- ern fur seal prey. C = bomb calorimetry combustion value; P = proximate analysis value1; muscle = edible portion only of raw material; whole = raw material from entire specimen. Energy Analysis value and Prey (kcal/g) tissue Reference American shad 2.08 P, muscle Sidwell (1981) Pacific herring 2.17 P, whole Sidwell (1981); Bigg et al. (1978) Northern anchovy 1.79 P, whole Sidwell (1981) Salmonids 2.01 P, muscle Sidwell (1981) Capelin 1.31 C, whole Miller2 Eulachon 1.41 P, muscle Stansby (1976) Deep-sea smelts 0.76 P, whole Childress and Nygaard (1973) Myctophiform 1.58 P, whole Childress and Nygaard fishes (1973)3 Pacific saury 2.20 P, muscle Sidwell (1981) Jacksmelt 1.24 P, muscle Watt and Merrill (1963) Pacific cod 1.00 P, muscle Sidwell (1981) Pacific whiting 1.17 P, whole Sidwell (1981) Walleye pollock 1.41 C, whole Miller2 Threespine stickleback 1.15 C, whole Wootton (1976) Jack mackerel 1.24 P, whole Sidwell (1981) Rockfishes 1.17 P, muscle Sidwell et al. (1974) Sablefish 2.17 P, muscle Sidwell (1981) Atka mackerel 1.58 P, muscle Kizevetter (1971) Pacific sand lance 1.22 P, muscle Sidwell (1981) Flounders 1.20 P, muscle Sidwell (1981)4 Market squid 1.15 P, muscle Sidwell (1981) Onychoteuthid squids 1.29 Perez5 Gonatid squids 1.27 Perez5 4Perez, M. A., and M. A. Bigg. 1981. Modified volume: a two- step frequency-volume method for ranking food types found in stomachs of northern fur seals. Unpubl. rep., 25 p. Northwest and Alaska Fisheries Center, National Marine Mammal Labora- tory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA 98115. 'Values were calculated with the following energy factors derived from Watt and Merrill (1963): 9.50, 5.65 and 4.00 kcal/g respectively for fat, protein and carbohydrate. 2Miller, L. K. 1978. Energetics of the northern fur seal in relation to climate and food resources of the Bering Sea. U.S. Mar. Mammal Comm. Rep. MMC-75/08, 27 p. 3Myctophidae and Paralepididae. 4Pleuronectidae. 5Perez, M. A., Natl. Mar. Mammal Lab., Northwest and Alaska Fish. Cent., Natl. Mar. Fish. Serv., NOAA, 7600 Sand Point Way N.E., Seattle, WA 981 15, unpubl. data, 1984. 958 PEREZ and BIGG: DIET OF NORTHERN FUR SEALS the energy value of prey. Data from all years of collection were pooled. We assumed that the impor- tance of a prey species to northern fur seals de- pended, at least in part, on its energy content. Table 1 lists the estimated caloric values for prey species consumed most often. These estimates are provi- sional because little is known about changes in energy content within each species by season. Energy values for squids tend to be lower than those for fishes, although large variability exists among fish species. No attempt was made to describe diet by age, sex, and reproductive condition. In our sample, 88% of the northern fur seals were females aged >3 yr, of which 53% were pregnant and 29% were nonpreg- nant. Thus, the diet described is primarily that for pregnant and nonpregnant females aged >3 yr. The eastern North Pacific Ocean and eastern Ber- ing Sea were divided into 7 regions and 21 sub- regions (Fig. 1). The boundaries for the seven regions were those which have been traditionally EASTERN BERING SEA SEALASKA BEROFF ^V^ 62° N .- y ( UNIMAK WESTERN ALASKA GULF OF ALASKA HECATE WASHOFF WASHINGTON •BCIMLETS WESTVAN . PEROUSE OREGON WASHNO CALIFORNIA NORTH PACIFIC OCEAN NCALOFF CCALOFF SCALOFF CCALIN SCALIN \ 52° 42° 32° 175° W 155° 135° 115° Figure 1.— Seven regions (denoted by darker lines) and 21 subregions used in the northern fur seal analyses: 1) California comprised of subregions SCALIN (southern California, inshore), SCALOFF (southern California, offshore), CCALIN (central California, inshore), CCALOFF (central California, offshore), NCALIN (northern California, inshore), and NCALOFF (northern California, offshore); 2) Oregon which includes half of subregion WASHNO (southern Washington and northern Oregon, inshore); 3) Washington which includes subregions PEROUSE (area west of Juan de Fuca Strait from Barkley Sound to Cape Flattery, including La Perouse Bank and Swift- sure Bank; inshore), WASHOFF (Washington, offshore), and half of WASHNO; 4) British Columbia which includes subregions BCINLETS (inside passages and inlets of B.C., inshore), WESTVAN (area west of Vancouver Island and Queen Charlotte Strait, inshore), HECATE (Hecate Strait area, inshore), and BCOFF (British Columbia, offshore); 5) the Gulf of Alaska which includes subregions SEALASKA (southeast Alaska, inshore), NOGULF (northern Gulf of Alaska, including Fairweather Bank; inshore), KODIAK (area around Kodiak Island, including Portlock Bank and Albatross Bank; inshore), and CENGULF (oceanic region of the Gulf of Alaska, offshore); 6) western Alaska which includes part of subregion UNIMAK (Unimak Pass area); and 7) the eastern Bering Sea comprised of subregions BERIN (Bering Sea shelf, inshore) and BEROFF (Bering Sea basin, offshore), and also includes subregions PRIBILOF (area around the Pribilof Islands) and most of UNIMAK. Subregions in which >50% of the area is <100 fathoms are noted as inshore; the remainder are noted as offshore. 959 FISHERY BULLETIN: VOL. 84, NO. 4 used in analyses of pelagic data for northern fur seals. The subregions were selected to compare diet between inshore (neritic) and offshore (oceanic) areas and to indicate diet in certain localities where collection effort was relatively high. Inshore areas were defined as those generally occurring on the continental shelf (depths up to 100 fathoms) and off- shore areas as those beyond the continental shelf. RESULTS The cruise tracks (Fig. 2) taken by research vessels of the United States and Canada for the col- lection of northern fur seals during 1958-74 indicate the relative distribution of research effort. Most col- lections were made in the coastal areas between California and British Columbia, off Kodiak Island, and in the eastern Bering Sea between Unimak Pass and the Pribilof Islands. Few specimens were taken more than 160 km from shore. Diet by Region and Month An examination of the number of prey species that made up the diet indicates that at least nine species may be consumed within any one subregion. How- ever, typically only three prey species made up about 80% of the diet (Fig. 3). Thus, relatively few species of food are of primary importance in any one local- ity. As will be made clear in the following regional and subregional accounts, the primary food species can change among localities. Our interpretation of Figures 4-11 which follow requires clarification. These figures show modified volume values only for those individual species that we felt were important and that had sufficient sample sizes to be reliable. Thus, we arbitrarily pre- sented only those species that were of >5% in im- portance for samples with at least 20 stomachs con- taining food. Species of less importance were pooled either as miscellaneous fishes or squids. Also, be- 62° N 175° W 155° 135° 115° Figure 2.— Cruise tracks of northern fur seal research vessels from the United States and Canada during 1958-74. 960 PEREZ and BIGG: DIET OF NORTHERN FUR SEALS 4 6 Number of prey species 10 Figure 3.^The cumulative percentage distribution of the number of prey species eaten in the total diet of northern fur seals taken during 1958-74. Data for each of the 21 subregions are plotted, although only the average relationship is graphed. cause the prey consumed by month within sub- regions were not presented here, we take these data from Perez and Bigg5 in our interpretation of sub- regional data. California Northern anchovy, Engraulis mordax, was the most important food eaten by the northern fur seals off California (Fig. 4A) whether its energy content was considered or not. However, it was more im- portant when its caloric value was taken into ac- count. Northern anchovy was eaten mainly during January to March in inshore and offshore waters of central and southern California (Fig. 4B, C). Pacific whiting, Merluccius productus, was second in im- portance (Fig. 4A) and was preyed upon in all areas of California, although primarily during April and May (Fig. 4B, C). Market squid, Loligo opalescens, was eaten from January to June, but only in neritic locations (Fig. 4B, C). Onychoteuthid squids (Ony- choteuthidae) were eaten offshore and were the more important squid species consumed in the south- ern areas off California (Fig. 4C). Other prey types were of relatively minor importance, although some were locally significant, such as Pacific saury, Colo- labis saira, mainly in oceanic areas off northern and central California (Fig. 4 A, B, C). 6Perez, M. A., and M. A. Bigg. 1981. An assessment of the feeding habits of the northern fur seal in the eastern North Pacific Ocean and eastern Bering Sea. Unpubl. draft rep., 146 p. North- west and Alaska Fisheries Center, National Marine Mammal Laboratory, National Marine Fisheries Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA 98115. 100 NCALIN NCAL0FF CCALIN CCAL0FF SCALIN SCAL0FF 116 53 1105 353 78 106 Figure 4.— Composition (percent) of diet of northern fur seals by prey species off California during 1958-74 (A) for pooled January- June samples (N = 1,811), using modified volume (dark bars) and energy-adjusted modified volume; (B) by month using modified volume; and (C) by subregion with pooled January-June samples using modified volume. A dark line separates squid and fish categories in the latter two figures. Key: ANC = northern an- chovy; GON = gonatid squids; JAC = jack mackerel; JCK = jacksmelt; MAR = market squid; MF = miscellaneous fish species; MS = miscellaneous squid species; MYC = myctophiform fishes; ONY = onychoteuthid squids; OP = other prey; SBL = sablefish; SRY = Pacific saury; WHI = Pacific whiting. 961 FISHERY BULLETIN: VOL. 84, NO. 4 Oregon Only 69 northern fur seals with food in their stomachs were collected in Oregon between Janu- ary and May during 1958-74, with 58 of these taken during April. Thus, diet could not be determined by month or by inshore and offshore areas. As in California, the main food was northern anchovy (Fig. 5). Other important prey were market squid, onychoteuthid squids, Pacific whiting, and rock- fishes (Sebastes spp.). Washington Pacific herring, Clupea harengus pallasi, was the most important food for northern fur seals off Washington, particularly when energy content was considered (Fig. 6A). It was only slightly more sig- nificant than rockfishes, salmonids (Salmonidae, primarily Oncorhynchus spp.), and northern anchovy when caloric values were not incorporated. Pacific herring was eaten from December to June but only SRY WHI Prey species Figure 5.— Composition (percent) of diet of northern fur seals by prey species off Oregon during 1958-74 for pooled January- June samples (N = 69), using modified volume (dark bars) and energy- adjusted modified volume. Key: ANC = northern anchovy; MAR = market squid; ONY = onychoteuthid squids; OP = other prey; ROC = rockfishes; SRY = Pacific saury; WHI = Pacific whiting. MAR ONY GON SHA HER ANC SAL CAP EUL ROC SBL OP Prey species B 100 8 50 D J F M A M J 42 179 152 343 757 425 20 N WASHNO PEROUSE WASHOFF 746 961 234 Figure 6.— Composition (percent) of diet of north- ern fur seals by prey species off Washington dur- ing 1958-74 (A) for pooled December-June samples (N = 1,918), using modified volume (dark bars) and energy-adjusted modified volume; (B) by month using modified volume; and (C) by subregion with pooled December-June samples using modified volume. A dark line separates squid and fish categories in the latter two figures. Key: ANC = northern anchovy; CAP = capelin; EUL = eula- chon; GON = gonatid squids; HER = Pacific her- ring; MAR = market squid; MF = miscellaneous fish species; MS = miscellaneous squid species; ONY = onychoteuthid squids; OP = other prey; ROC = rockfishes; SAL = salmonids; SBL = sablefish; SHA = American shad; WHI = Pacific whiting. 962 PEREZ and BIGG: DIET OF NORTHERN FUR SEALS in neritic areas (Fig. 6B, C). Rockfishes, salmonids, and northern anchovy were also consumed by seals during this time, both inshore and offshore. North- ern anchovy was primarily important in the south- ern area of the region (Fig. 6C). The main food in oceanic waters consisted of two families of squids, Onychoteuthidae and Gonatidae (Fig. 6C). Market squid was the primary squid species preyed upon in neritic areas. British Columbia As in Washington, Pacific herring was the prim- ary food of the northern fur seals from February to June in most inshore areas, particularly when energy content was taken into account (Fig. 7 A, B, C). It was mainly consumed by northern fur seals off the west coast of Vancouver Island and in Hecate Strait. In coastal inlets, market squid was impor- tant, but not significantly for the region as a whole. The diet of northern fur seals in oceanic waters dur- ing May and June was almost exclusively onycho- teuthid squids and salmonids (Fig. 7B, C). Other prey species were relatively insignificant (Fig. 7A). However, because the coastline of British Colum- bia is complex, and sample sizes were small, addi- tional local differences in diet may exist in inshore areas (Fig. 7C). MAR HER SAL Prey species B 100 £ 50 - N WESTVAN BCINLETS HECATE BCOFF 140 41 79 78 Figure 7.— Composition (percent) of diet of northern fur seals by prey species off British Columbia during 1958-74 (A) for pooled January- June samples (N = 354), using modified volume (dark bars) and energy-adjusted modified volume; (B) by month using modified volume; and (C) by subregion with pooled January-June samples using modified volume. A dark line separates squid and fish categories in the latter two figures. Key: COD = Pacific cod; EUL = eulachon; GAD = gadid fishes; GON = gonatid squids; HER = Pacific herring; MAR = market squid; MF = miscellaneous fish species; ONY = onychoteuthid squids; OP = other prey; POL = walleye pollock; ROC = rockfishes; SAL = salmonids; SBL = sablefish; US = unidentified squid; WHI = Pacific whiting. 963 FISHERY BULLETIN: VOL. 84, NO. 4 Gulf of Alaska Based on all samples collected in the Gulf of Alaska, the main diet of northern fur seals was Pacific herring when energy content was con- sidered, but Pacific sand lance, Ammodytes hexap- terus, was most important when caloric values were not considered (Fig. 8A). However, there were subregional differences in diet. Off southeastern Alaska, collections were made in Sitka Sound dur- ing February and March where the diet was almost exclusively Pacific herring (Fig. 8B, C). In the north- ernmost area of the region the diet consisted chief- ly of capelin, Mallotus villosus, but also to a lesser degree of both walleye pollock, Theragra chalco- gramma, and Pacific sand lance (Fig. 8C). Off Kodiak Island during April to July, the diet was mainly Pacific sand lance and capelin (Fig. 8B, C). Gonatid squids (Gonatidae) were the primary foods of northern fur seals in oceanic waters of this region from April to June. Rockfishes and salmonids were also eaten by northern fur seals in offshore and northern inshore areas of the region (Fig. 8C). Western Alaska Of the 309 stomachs with food collected in this region from May to October 1958-74, 239 were taken during June, with most of these collected south of Unimak Pass. The main foods of the north- ern fur seals were Pacific sand lance and capelin, as off Kodiak Island, with the energy content of each having little effect on their relative importance (Fig. 9). Other important prey were Atka mackerel, Pleurogrammus monopterygius, salmonids, walleye pollock, and the squid Berryteuthis magister. Sable- fish, Anoplopoma fimbria, and Pacific herring were also eaten by northern fur seals south of Unimak Pass during summer months. GON HER CAP .POL Prey species SND OP B 100 75 - S 50 Figure 8.— Composition (percent) of diet of northern fur seals by prey species in the Gulf of Alaska during 1958-74 (A) for pooled February-July samples (N = 1,163), using modified volume (dark bars) and energy- adjusted modified volume; (B) by month using modified volume; and (C) by subregion with pooled February- July samples using modified volume. Key: CAP = capelin; GON = gonatid squids; HER = Pacific her- ring; MF = miscellaneous fish species; MS = miscellaneous squid species; OP = other prey; POL = walleye pollock; ROC = rockfishes; SAL = salmonids; SND = Pacific sand lance; US = unidentified squid. 100' 75- 50 25- HER • • • • '.SND*. ■ ■ ■_■_•_■. *p6l>: HER CAP Mv/;i*y^ CAP SND F M A M J J SEALASKA NOGULF KODIAK CENGULF N 32 196 203 452 260 20 N 242 115 733 73 964 PEREZ and BIGG: DIET OF NORTHERN FUR SEALS 30,- BER SAL CAP POL ATK Prey species SND OP Figure 9.— Composition (percent) of diet of northern fur seals by prey species in western Alaska during 1958-74 for pooled May-October samples (N = 309), using modified volume (dark bars) and energy-adjusted modified volume. Key: ATK = Atka mackerel; BER = Berryteuthis magister; CAP = capelin; OP = other prey; POL = walleye pollock; SAL = salmonids; SND = Pacific sand lance. Eastern North Pacific Northern anchovy (20%) and Pacific herring (19%) were the main species eaten by the northern fur seals in the eastern North Pacific when data from all regions and months were pooled (Fig. 10). These prey were the most important whether energy con- tent was considered or not, although importance in- creased when the caloric values were included. Salmonids (6%), capelin (8%), Pacific whiting (7%), walleye pollock (2%), Pacific sand lance (8%), and rockfishes (4%) were also commonly eaten. The re- maining diet was made up of a wide variety of squids (mainly market squid, 6%; onychoteuthid squids, 6%; and gonatid squids, 5%) and other fishes (mainly Pacific saury, 4%; sablefish, 2%; and Atka mackerel, 2%). Squids were the primary food species in oceanic waters between California and the Gulf of Alaska, and fishes were the main prey in the neritic areas. Although not eaten in large amounts, salmonids and rockfishes were the main fishes consumed in oceanic areas between Washington and the Gulf of Alaska (Figs. 6C, 7C, 8C). O ■D > n. £ > ° *o-= Ox Z O uiZ I cc o» 27) Pelagic (l,S) Northern anchovy <18 9-18 (7,27) Pelagic (l-O.S) Salmonids3 <80 15-41 (22,>26) Pelagic (l-0,A) Capelin 22 7-14 (7,64) Pelagic (l,S) Eulachon 23-30 12-21 (3,11) Pelagic (l,S,A) Deep-sea smelts 2-18 8-12(6,986) Pelagic (0,S) Myctophiform fishes 13-20 — Pelagic (0,S) Pacific saury 10-32 25 (1 ,4) Pelagic (0,S) Pacific whiting 66-76 15(1,2) Pelagic and semidemersal (l-O.S) Walleye pollock <90 4-40 (71,1721) Pelagic and semidemersal (l-O.S) Rockfishes 30-53 11-31 (6,>19) Demersal (l-0,S) Sablefish 57-60 20-31 (3,>3) Pelagic and semidemersal (l-O.S) Atka mackerel <120 15-23 (5,>5) Pelagic and semidemersal (l,S) Pacific sand lance 20 — Demersal (l,S) Market squid 14-17 7-15 (6,43) Pelagic (I) Onychoteuthid squids4 10-37 14-22 (3,>3) Pelagic (l-O) Gonatid squid 12-32 5-24 (10,>59) Pelagic (l-O) 7McAlister, W. B., and M. A. Perez. 1977. Ecosystem dynamics— birds and marine mammals. Part 1: preliminary esti- mates of pinniped-finfish relationships in the Bering Sea (final report). In Environmental assessment of the Alaskan continen- tal shelf, Annual Report 12, p. 342-371. U.S. Department of Com- merce, Environmental Research Laboratory, Boulder, CO. 'Data on average lengths of prey and ecology were compiled from Aki- mushkin (1 963), Bakkala et al. (1 981 ), Baxter (1 967), Baxter and Duffy (1 974), Carl (1964), Childress and Nygaard (1973), Childress et al. (1980), Fields (1965), Fitch (1974), Fitch and Lavenberg (1968, 1971, 1975), Hart (1973), lnada(1981), Miller and Lea (1976), Naitoetal. (1977), Niggol (1982), Pearcy (1965), Pearcy et al. (1979), Smith (1981), Taka et al. (1980), and Wespestad and Barton (1981). 2Total length for fish and dorsal mantle length for squid. The first number in parentheses is the number of fur seal stomachs examined, and the second number in parentheses is the number of prey specimens measured. These data were derived from an analysis of the original unpublished 1 958-74 data. 3Maximum size of salmonids found at sea. Adults in freshwater are larger (to 147 cm) depending upon species. 4Does not include size range of Moroteuthis (<140 cm) which has been taken by northern fur seals, but rarely off North America. japonicus, in addition to walleye pollock and squid (Taylor et al. 1955; Lander and Kajimura 1980). Of interest is the fact that in recent years the Japanese sardine, Sardinops melanosticta, has become more important in the diet of northern fur seals off Asia (Yoshida et al.89; Yoshida and Baba1011). This sar- dine was depleted during the 1930's and 1940's and 8Yoshida, K., N. Okumoto, and N. Baba. 1979. Japanese pelagic investigation on fur seals, 1978. Far Seas Fish. Res. Lab., Shimizu, Jpn., Fur Seal Resour. Sect., Contrib. No. 41-9, 66 p. 968 PEREZ and BIGG: DIET OF NORTHERN FUR SEALS recovered only recently (Kondo 1980). The northern fur seal appears to have reacted to this recovery by eating more sardines. A similar change in diet may have taken place off California during the past 50 years. The Pacific sardine, Sardinops sagax, was once the most abundant small, schooling fish off California, whereas now northern anchovy is (Mur- phy 1966; Smith 1972; Mais 1974). The Pacific sar- dine population was drastically reduced during the 1940's mainly because of fishing pressure and has remained at a relatively low level since, while the northern anchovy increased in abundance during the 1950's and the 1960's (Vrooman and Smith 1971; Hart 1973; Wolf and Smith 1985). The Pacific sar- dine may undergo long-term periodic fluctuations in population size (Thompson 1921), and it may now once again be increasing in biomass (Wolf and Smith 1985). Northern fur seals have not eaten Pacific sar- dine in recent years, but perhaps they fed on this species prior to the 1940's. The seal may have changed its diet from largely Pacific sardine to northern anchovy. Unfortunately, the stomach con- tents of only two northern fur seals were collected from California prior to the 1950's (Scheffer 1950). Clemens and Wilby (1933) gave the only evidence that sardines were once consumed by these seals in the eastern North Pacific Ocean. They found that sardines were commonly eaten during 1931 off southwestern Vancouver Island. An interesting speculation regarding the signifi- cance of small schooling fish to northern fur seals is the relationship between diet and the migration route of the seal. Small schooling fish could be im- portant just because they are abundant and lie along the coastal migration path of northern fur seals. Ka- jimura (1985) argued for this possibility. He sug- gested that the migration pattern of northern fur seals is genetically established and that the seal feeds opportunistically upon whatever prey species are most abundant in its path. He believes that, although food is not a major factor in determing the migration route of northern fur seals, the move- ments of prey species can still alter the local distribu- tion of fur seals. An alternative possibility is that the seals learn the location of the main foods and then selects its migration route to include them. 9Yoshida, K., N. Okumoto, and N. Baba. 1981. Japanese pelagic investigation on fur seals, 1979-1980. Far Seas Fish. Res. Lab., Shimizu, Jpn., Fur Seal Resour. Sect., Contrib. No. 41-10, 150 p. 10Yoshida, K., and N. Baba. 1983. Japanese pelagic investiga- tion on fur seals, 1981-1982. Far Seas Fish. Res. Lab., Shimizu, Jpn., Fur Seal Resour. Sect, Contrib. No. 41-11, 118 p. "Yoshida, K., and N. Baba. 1984. Japanese pelagic investiga- tion on fur seals, 1983. Far Seas Fish. Res. Lab., Shimizu, Jpn., Fur Seal Resour. Sect., Contrib. No. 41-12, 67 p. Baker (1978) argued for this alternative. He pro- posed that, while some inherited factors may be in- volved in migration, northern fur seals could main- ly search the North Pacific Ocean for the most preferred or abundant food, and thereafter estab- lish the migration route. Such being the case, perhaps inexperience explains why 1-2 yr-old seals are rarely seen inshore feeding with older seals. Also, perhaps squid is not a preferred or sufficiently available food for northern fur seals offshore, because most seals older than 1-2 yr feed inshore on fish. However, at this stage, not enough is known about the factors that control migration of the north- ern fur seal to establish which alternative is true. As Kajimura (1985) has pointed out, factors other than diet are no doubt involved as indicated by the fact that males do not migrate as far south as females. ACKNOWLEDGMENTS The following people assisted us with the prepara- tion of data for analysis, and with the construction of tables and figures presented in unpublished pre- liminary reports of this study: Julia Bosma, Patricia Bouthillette, Laurie Briggs, Carl Brooks, David Crystal, Ian Fawcett, Gary Fidler, Job Groot, Carol Hastings, Marta Hladyschevsky, Kerry Hobbs, Gerald Hornof, Marilyn Marshall, Elizabeth Mooney, R. Perez, Kenneth Pierce, and Marsha Schad. LITERATURE CITED Akimushkin, I. I. 1963. Cephalopods of the seas of the U.S.S.R. [In Russ.] Izd. Akad. Nauk, Mosc. (Transl. by Isr. Prog. Sci. Transl., 1965, avail. U.S. Dep. Commer., Natl. Tech. Inf. Serv., Springfield, Va., as TT65-50013, 223 p.) Baker, R. C, F. Wilke, and C. H. Baltzo. 1970. The northern fur seal. U.S. Fish Wildl. Serv., Circ. 336, 20 p. Baker, R. R. 1978. The evolutionary ecology of animal migration. Holmes and Meier Publ., N.Y., 1012 p. Bakkala, R., K. King, and W. Hirschberger. 1981. Commercial use and management of demersal fish. In D. W. Hood and J. A. Calder (editors), The eastern Ber- ing Sea shelf: Oceanography and resources, vol. 2, p. 1015- 1036. Univ. Wash. Press, Seattle, WA. Baxter, J. L. 1967. Summary of biological information on the northern an- chovy Engraulis mordax Girard. Calif. Coop. Oceanic Fish. Invest. Rep. 11:110-116. Baxter, J. L., and J. M. Duffy. 1974. Inshore fishes of California, 4th revision. Calif. Dep. Fish Game, Sacramento, CA, 78 p. Bigg, M. A., and I. Fawcett. 1985. Two biases in diet determination of northern fur seals 969 FISHERY BULLETIN: VOL. 84, NO. 4 (Callorhinus ursinus). In J. R. Beddington, R. J. H. Bever- ton, and D. M. Lavigne (editors), Marine mammals and fisheries, p. 284-291. George Allen & Unwin (Publ.), Lond. Bigg, M. A., I. B. MacAskie, and G. Ellis. 1978. Studies on captive fur seals. Progress Rep. No. 2. Can. Fish. Mar. Serv. Manuscr. Rep. 1471, 21 p. Nanaimo, B.C. Bigg, M. A., and M. A. Perez. 1985. Modified volume: a frequency-volume method to assess marine mammal food habits. In J. R. Beddington, R. J. H. Beverton, and D. M. Lavigne (editors), Marine mammals and fisheries, p. 277-283. George Allen & Unwin (Publ.), Lond. Carl, G. C. 1964. Some common marine fishes of British Columbia. B.C. Prov. Mus., Victoria, B.C., Handb. No. 23, 86 p. Childress, J. J., and M. H. Nygaard. 1973. The chemical composition of midwater fishes as a func- tion of depth of occurrence off southern California. Deep- Sea Res. 20:1093-1109. Childress, J. J., S. M. Taylor, G. M. Cailliet, and M. H. Price. 1980. Patterns of growth, energy utilization and reproduc- tion in some meso- and bathypelagic fishes off southern California. Mar. Biol. 61:27-40. Clemens, W. A., J. L. Hart, and G. V. Wilby. 1936. Analysis of stomach contents of fur seals taken off the west coast of Vancouver Island in April and May 1935. Can. Dep. Fish., Ottawa, 20 p. Clemens, W. A., and G. V. Wilby. 1933. Food of the fur seal off the coast of British Columbia. J. Mammal. 14:43-46. Fields, W. G. 1965. The structure, development, food relations, reproduc- tion, and life history of the squid Loligo opalescens Berry. Calif. Dep. Fish Game, Fish Bull. 131, 108 p. Fiscus, C. H. 1978. Northern fur seal. In D. Haley (editor), Marine mam- mals of eastern North Pacific and arctic waters, p. 152- 159. Pac. Search Press, Seattle, WA. Fitch, J. E. 1974. Offshore fishes of California. 5th rev. Calif. Dep. Fish Game, Sacramento, CA, 80 p. Fitch, J. E., and R. J. Lavenberg. 1968. Deep-water teleostean fishes of California. Univ. Calif. Press, Berkeley, CA, 155 p. 1971. Marine food and game fishes of California. Univ. Calif. Press, Berkeley, CA, 179 p. 1975. Tidepool and nearshore fishes of California. Univ. Calif. Press, Berkeley, CA, 156 p. Hart, J. L. 1973. Pacific fishes of Canada. Bull. Fish. Res. Board Can. 180, 740 p. Inada, T. 1981. Studies on the Merlucciid fishes. Bull. Far Seas Fish. Res. Lab. (Shimizu) 18:1-172. Kajimura, H. 1984. Opportunistic feeding of the northern fur seal Callo- rhinus ursinus, in the eastern North Pacific Ocean and eastern Bering Sea. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-779, 49 p. 1985. Opportunistic feeding by the northern fur seal, (Callo- rhinus ursinus). In J. R. Beddington, R. J. H. Beverton, and D. M. Lavigne (editors), Marine mammals and fisheries, p. 300-318. George Allen & Unwin (Publ.), Lond. Kenyon, K. W. 1956. Food of fur seals taken on St. Paul Island, Alaska, 1954. J. Wildl. Manage. 20:214-215. KlZEVETTER, I. V. 1971. Chemistry and technology of Pacific fish. Dal'izdat, Vladivostok. (Transl. by Isr. Prog. Sci. Transl., 1973, avail. U.S. Dep. Commer., Natl. Tech. Inf. Serv., Springfield, VA, as TT72-50019, 304 p.) Kondo, K. 1980. The recovery of the Japanese sardine - the biological basis of stock-size fluctuations. Rapp. P. -v. Reun. Cons. int. Explor. Mer 177:332-354. Lander, R. H. (editor). 1980. Summary of northern fur seal data and collection pro- cedures. Vol. 2: Eastern Pacific pelagic data of the United States and Canada (excluding fur seals sighted). U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-4, 541 p. Lander, R. H., and H. Kajimura (editors). 1980. Summary of northern fur seal data and collection pro- cedures. Vol. 3: Western Pacific pelagic data of the Soviet Union and Japan, 1958-1978 (excluding fur seals sighted). U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-5, 304 p. Lucas, F. A. 1899. The food of the northern fur seals. In D. S. Jordan et al. (editors), The fur seals and fur-seal islands of the North Pacific Ocean, Part 3, p. 59-68. Gov. Print. Off., Wash., DC. Mais, K. F. 1974. Pelagic fish surveys in the California Current. Calif. Dep. Fish Game, Fish Bull. 162, 79 p. May, F. H. 1937. The food of the fur seal. J. Mammal. 18:99-100. Miller, D. J., and R. N. Lea. 1976. Guide to the coastal marine fishes of California. Calif. Dep. Fish Game, Fish Bull. 157, 249 p. [1972, addendum added 1976.] Murphy, G. I. 1966. Population biology of the Pacific sardine (Sardinops caerulea). Proc. Calif. Acad. Sci. (4th Ser.) 34:1-84. Naito, M., K. Murakami, and T. Kobayashi. 1977. Growth and food habit of oceanic squids (Ommastrephes bartrami, Onychoteuthis borealijaponicus, Berryteuthis magister and Gonatopsis borealis) in the western subarctic Pacific region. [In Jpn., Engl, summ.] In Fisheries bio- logical production in the subarctic Pacific region, p. 339-351. Res. Inst. North Pac. Fish. Fac. Fish., Hokkaido Univ., Spec. Vol. 1977. NlGGOL, K. 1982. Data on fish species from Bering Sea and Gulf of Alaska. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-29, 125 p. Pearcy, W. G. 1965. Species composition and distribution of pelagic cepha- lopods from the Pacific Ocean off Oregon. Pac. Sci. 19:261-266. Pearcy, W. G., T. Nemoto, and M. Okiyama. 1979. Mesopelagic fishes of the Bering Sea and adjacent northern North Pacific Ocean. J. Oceanogr. Soc. Jpn. 35: 127-134. Robbins, C. T. 1983. Wildlife feeding and nutrition. Acad. Press, N.Y., 343 P- SCHEFFER, V. B. 1950. The food of the Alaska fur seal. U.S. Fish Wildl. Serv., Bur. Commer. Fish., Wildl. Leafl. 329, 16 p. 970 PEREZ and BIGG: DIET OF NORTHERN FUR SEALS SCHULTZ, L. P., AND A. M. RAFN. 1936. Stomach contents of fur seals taken off the coast of Washington. J. Mammal. 17:13-15. Sidwell, V. D. 1981. Chemical and nutritional composition of finfishes, whales, crustaceans, mollusks, and their products. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/SEC-11, 432 p. Sidwell, V. D., P. R. Foncannon, N. S. Moore, and J. C. Bonnet. 1974. Composition of the edible portion of raw (fresh or frozen) crustaceans, finfish and mollusks. I. Protein, fat, moisture, ash, carbohydrate, energy value, and cholesterol. Mar. Fish. Rev. 36(3):21-35. Smith, G. B. [1981.]. The biology of walleye pollock. In D. W. Hood and J. A. Calder (editors), The eastern Bering Sea shelf: Ocean- ography and resources, Vol. 1, p. 527-551. U.S. Gov. Print. Off., Wash., DC. Smith, P. E. 1972. The increase in spawning biomass of northern anchovy, Engraulis rrwrdax. U.S. Natl. Mar. Fish. Serv., Fish. Bull. 70:849-874. Spalding, D. J. 1964. Comparative feeding habits of the fur seal, sea lion, and habour seal on the British Columbia coast. Bull. Fish. Res. Board Can. 146, 52 p. Stansby, M. E. 1976. Chemical characteristics of fish caught in the northeast Pacific Ocean. Mar. Fish. Rev. 38(9):1-11. Stroud, R. K., C. H. Fiscus, and H. Kajimura. 1981. Food of the Pacific white-sided dolphin, Lagenorhyn- chus obliquidens, Dall's porpoise, Phocoenoides dalli, and northern fur seal, Callorhinus ursinus, off California and Washington. U.S. Natl. Mar. Fish. Serv., Fish. Bull. 78: 951-959. Taka, S., M. Kitakata, and T. Wada. 1980. Food organisms of saury, Cololabis saira (Brevoort) and vertical distribution of zooplankton in the southeast waters off Kuril Islands in July, 1976-1978. [In Jpn., Engl, summ.] Bull. Hokkaido Reg. Fish. Res. Lab. 45:15-41. Taylor, F. H. C, M. Fujinaga, and F. Wilke. 1955. Distribution and food habits of the fur seals of the North Pacific Ocean. U.S. Dep. Inter., Fish Wild!. Serv., Wash., DC, 86 p. Thompson, W. F. 1921. The future of the sardine. Calif. Fish Game 7:38- 41. Vrooman, A. M., and P. E. Smith. 1971. Biomass of the subpopulations of northern anchovy Engraulis mordax Girard. Calif. Coop. Oceanic Fish. Invest. Rep. 15:49-51. Watt, B. K., and A. L. Merrill. 1963. Composition of foods. . .raw, processed, prepared. U.S. Dep. Agric, Agric. Handb. 8, 190 p. Wespestad, V. G., and L. H. Barton. [1981]. Distribution, migration, and status of Pacific herring. In D. W. Hood and J. A. Calder (editors), The eastern Ber- ing Sea shelf: Oceanography and resources, Vol. 1, p. 509- 525. U.S. Gov. Print. Off., Wash., DC. Wilke, F., and K. W. Kenyon. 1952. Notes on the food of fur seal, sea lion and harbor por- poise. J. Wildl. Manage. 16:396-397. 1954. Migration and food of the northern fur seal. Trans. North Am. Wildl. Nat. Resour. Conf. 19:430-440. 1957. The food of fur seals in the eastern Bering Sea. J. Wildl. Manage. 21:237-238. Wolf, P., and P. E. Smith. 1985. An inverse egg production method for determining the relative magnitude of Pacific sardine spawning biomass off California. Calif. Coop. Oceanic Fish. Invest. Rep. 26:130- 138. Wootton, R. J. 1976. The biology of the sticklebacks. Acad. Press, Lond., 387 p. 971 INSTAR IDENTIFICATION AND LIFE HISTORY ASPECTS OF JUVENILE DEEPWATER SPIDER CRABS, CHIONOECETES TANNERI RATHBUN Patricia A. Tester1 and Andrew G. Carey, Jr.2 ABSTRACT For the deepwater spider crab, Chionoecetes tanneri, seven instars from first crab stage (3.8 mm carapace width (CW)) to instar VII (26.8 mm CW) are identified from size-frequency histograms. The average growth per molt for the first seven instars is 39% and the time from egg to instar VII is estimated to be 20 months. Measurements of chela length, abdomen width and carapace width were used to define two growth phases for C. tanneri and to determine size at maturity for males (142.7 mm CW) and females (102.3 mm CW). The unequal sex ratio of adults (29% males) and presence of chitinoclastic lesions on 76% of the adult females as compared with only 29% of the adult males suggest that adult females are anecdysic. In this study of material collected off the southern Oregon coast, the mean adult carapace widths for males and females is very close to the sizes reported for adult males and females (148.9 and 102.5 mm CW respectively) from the northern Oregon coast. The similarity in size extends to the material collected from near the type location (Gulf of the Farallons) where instars VI and VII are 19.4 and 27.3 mm CW compared with 19.8 and 26.7 mm CW for the same instars from the southern Oregon coast. The biotic stability at depths of maximum abundance (500-775 m) contributes to this uniformity. The spider (or tanner) crab, Chionoecetes tanneri Rathbun, is similar in size and morphology to the better known and commercially harvested species C. bairdi and C. opilio. Unlike C. bairdi and C. opilio which are typically encountered in shallow waters and are not reported deeper than 400 m in the eastern Pacific, C. tanneri is a deep-water species which ranges to 1,925 m and has its maximum abun- dance at 500-775 m (Pereyra 1972). Although C. tanneri is not likely to be fished commercially because of its deep-water habitat and certain aspects of its biology, Somerton (1981) sug- gested that fluctuating supplies of Alaskan crab species might promote more economical methods for fishing in deep water. Red crab, Geryon quin- quedens, taken from depths of 257-1,000 m between Georges Bank and Cape Hatteras are landed com- mercially in limited numbers on the eastern sea- board (Lux et al. 1982; U.S. National Marine Fisheries Service Fisheries Statistics 1985). In part, because of its deep-water habitat, certain life history aspects of C. tanneri are not well known. Pereyra (1966, 1968) determined size at maturity 1 College of Oceanography, Oregon State University, Corvallis OR 97331; present address: Southeast Fisheries Center Beaufort Laboratory, National Marine Fisheries Service, NOAA, Beaufort, NC 28516-9722. 2College of Oceanography, Oregon State University, Corvallis OR 97331. and described the seasonal distribution of adult and late juvenile crabs. Egg development follows a year- ly cycle with release of matured eggs and ovulation of new eggs during the winter (Pereyra 1966). After hatching, the total larval (pelagic) phase (prezoea, zoea I, II and megalopa) is estimated to be 80 d (Lough 1974). Samples collected mainly from a series of cruises off the Oregon coast from 1972 to 1975 have provided us with C. tanneri specimens from the first crab stage to adult. These specimens have made it possible to identify a series of early instars and to determine juvenile growth rates; they also provided life history information on size at maturity and adult and juvenile sex ratios for comparison with earlier work. In addition, observa- tions of the carapace condition of adults helped to substantiate the anecdysic condition of adult females. METHODS Sampling Samples of C. tanneri were collected off the con- tinental shelf and slope areas adjacent to Coos Bay, OR (lat. 42°25'N, long. 124°50'W) in depths rang- ing from 300 to 1,200 m during 10 cruises between April 1973 and March 1975. A total of 1,625 crabs Manuscript accepted February 1986. FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 973 FISHERY BULLETIN: VOL. 84, NO. 4 of both sexes ranging in size from 10 to 165 mm carapace width (CW) were captured using two types of trawl gear: a 9 m semi-baloon Gulf of Mexico shrimp trawl and a 3 m beam trawl (Carey and Heyamoto 1972). The stretched dimension of the mesh for both trawl nets was 38 mm (1.5-in), and the cod ends were lined with 12.7 mm (0.5-in) mesh. In addition 47 of the smallest crabs (3-10 mm CW) were found in the gut contents of sable fish, Ano- plopoma fimbria, and Dover sole, Microstomias pacificus, caught in these trawls. The Smithsonian Institution provided another 306 juvenile tanner crabs taken near the type location for C. tanneri west of the Farallon Islands (lat. 37°30'N, long. 122°59'W) (Rathbun 1925) at 500-783 m. Size at Maturity The size at maturity for both male and female C. tanneri was based on allometric measurements. Allometry compares the difference in the propor- tions of specific body parts with changes in absolute size of a major body axis (Gould 1966). In Brachyura the allometric growth of secondary sex characters is well documented (Tessier 1960; Hartnoll 1969). In the genus Chionoecetes it takes the form of dif- ferential enlargement of the abdomen and modifica- tion of pleopods in females whereas the size and shape of the chelae are modified in the males (Watson 1970; Brown and Powell 1972). Carapace width for both male and female crabs was measured at its widest part (mesobranchial region) exclusive of spines (Fig. 1A). The male carapace width was compared with the length of the chelar propodus (CPL) which is measured from the joint between the carpus to the tip of the fixed finger of the propodus (Fig. IB), whereas the female carapace width was compared with the width of the abdomen (AW) which is measured at its widest part (fifth segment) (Fig. 1C). Males with worn or broken chelipeds were not used. All measurements were made to the nearest 0.01 mm using precision dial calipers, and numbers were rounded to the first decimal for plotting. Plots of the measurements of CW vs. CPL and CW vs. AW were used to identify size at maturity for males and females. Size-Frequency Histograms, Growth, and Sex Ratio Measurements of the carapace width were taken from the 1,978 crabs available. Size-frequency histo- grams were constructed and seven juvenile instars were identified from dominant modes. Adult C. tan- FlGURE 1.— Body dimensions of Chionoecetes tanneri measured for size-frequency and allometric analyses. (A) Carapace width (o* and 9) measured at its widest part across the mesobranchial region and exclusive of spines. (B) Chelar propodus length (cr) measured from the joint between the carpus and the tip of the fixed finger of the propodus. (C) Abdomen width (9) measured at its widest part, across the fifth segment. neri are sexually dimorphic with respect to body size (Pereyra 1972). Since we did not know at which molt this size dimorphism was first evident, the data for males and females was shaded differently in the size- frequency histogram. The juvenile sex ratio was 974 TESTER and CAREY: INSTAR IDENTIFICATION OF SPIDER CRABS calculated for each instar, and when it was clear from the equal sex ratio that juvenile males and females were not dimorphic, the percent increase in carapace width per molt was computed as Percent increase in CW (mm) = Postmolt CW (mm) - Premolt CW (mm) Premolt CW (mm) x 100 In one series of size-frequency histograms from samples collected in June, July, and August 1974 and January and March 1975, the progression of modes (representative of instars of the small juveniles from fish gut contents) was used to estimate growth rate. The next larger instar (CW = 10 mm) was the first to be consistently sampled by the trawl gear. Starting with the 10 mm CW in- star from April 1973, growth of juvenile tanner crabs was followed through August, October, and November 1973 and March 1974. Carapace Condition Detailed observations were made of the carapace on each specimen and included hardness, amount of attached fauna, and general condition. Darkened and softened or weakened areas on the carapaces were similar to those caused by chitinoclastic bac- teria (Sindermann 1970) and were thought to be associated with age. Adult female C. tanneri were especially subject to carapace deterioration. RESULTS AND DISCUSSION Since a high degree of correlation between gonad maturity and external morphology has been shown for the genus Chionoecetes (Brown and Powell 1972; Donaldson et al. 1981), a plot of carapace width and chela length (Fig. 2) was used to define adult males. Specimens with chelae longer than 85 mm (corre- sponding to carapace width >118 mm) were as- sumed to be sexually mature males. Those females E E D) C CD CO D T5 O a o 05 CD .C O 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 N=760 .**■"' 0 10 20 30 40 50 60 70 80 90 100 1 10 120 130 140 150 160 170 180 Carapace Width (mm) Figure 2.— Relationship between carapace width and chelar propodus length for juvenile and adult male Chionoecetes tanneri. 975 FISHERY BULLETIN: VOL. 84, NO. 4 with abdomen widths > 50 mm (corresponding to carapace widths >85 mm) form a well-defined group (Fig. 3) of adults. The mean carapace width for adult male and female crabs in this study was 142.7 and 102.3 mm respectively and was compared with the mean carapace widths of 148.9 and 102.5 mm for females given by Pereyra (1972) for adult C. tan- neri collected south of the Columbia River mouth. Brown and Powell (1972) reported a similar corre- spondence in adult carapace widths for C. bairdi collected from locations in Alaska. The large varia- tion in size of mature male C. bairdi in the eastern Bering Sea was clearly related at the clinal varia- tion temperature (Somerton 1981). Seven modes representing juvenile instars are evi- dent from the size-frequency histograms (Fig. 4). The mean carapace widths for each juvenile instar were calculated and subsequently the increase in CW per molt was computed (Table 1). The average increase at each molt for instars I-VII is 39% and there is no difference in growth increment of juve- nile males and females In a laboratory study using C. opilio, Miller and Watson (1976) reported that growth per molt for immature females was signifi- cantly greater than for immature males. But the findings of Hilsinger et al. (1975) agree with ours. They found no difference in growth rate for juvenile male and female C. bairdi and reported a constant growth rate of 27% for juvenile females. The change in the slope of the regression lines of the log-log plots of the allometric measurements of C. tanneri (Fig. 5) indicated a change in the rate of growth only at sexual maturity. Chionoecetes tanneri, like C. opilio (Watson 1970), showed two growth phases, one for juveniles and one for adults. If C. tanneri eggs hatch predominantly in winter (January-March) and the total larval life is 80 d, the recruitment of the smallest crab stage (CW = 4) to the population in June- July is in agreement with our findings. Instars can be followed from 4 mm CW (instar I) in June and July 1974, to 5.5 cm CW (in- star II) in August 1974, to 7.5 mm CW (instar III) in January 1975 (Fig. 6). The smallest specimens sampled by the trawls were about 10 mm CW, and there were relatively large numbers of these instar IV specimens in April 1973 which molted to instar V by August and to instar VI in October 1973 (Fig. 7). No growth of these instar VI crabs is evident from the November 1973 or March 1974 data. We estimate approximately 20 mo from egg hatching to instar VII (CW = 26.8 mm) (Fig. 8). Observations on general carapace condition and abundance of epifauna indicate that adult male C. tanneri do molt frequently enough to maintain their E E T3 CD E o < 70i 60 50 2 40 30 20 10 N=851 ♦ ♦ ♦ .♦♦% ♦♦ ♦**♦ «♦♦ 10 20 30 40 50 60 70 80 90 100 110 120 130 Carapace Width (mm) Figure 3.— Relationship between carapace width and abdomen width for juvenile and adult female Chionoecetes tanneri. 976 TESTER and CAREY: INSTAR IDENTIFICATION OF SPIDER CRABS 155 c c en W 0) CD u 0) o c g O d) .O e 3 Carapace Width (mm) Figure 4.— Size-frequency histograms representing all specimens of Chionoecetes tanneri collected off the Oregon coast 1973-75 and from the Gulf of the Farallons. Instars I-VII are indicated. Males are shown in solid color. carapaces relatively free of epifauna and lesions caused by bacterial infections (Baross et al. 1978). Of the 290 adult female specimens examined, 87% showed exoskeleton lesions and these adult females also had the highest diversity and abundance of epi- fauna on their exoskeletons. Only 29% of the 124 adult males observed showed the effect of chitino- clastic bacterial infection. No lesions or epifauna were found on any of the 1,447 juveniles examined. In contrast to the findings of Hartnoll (1969) who worked with shallow-water spider crabs, observa- tions of the carapace condition of adult male and female C. tanneri suggests adult males continue to molt after maturity while adult females are anec- dysic, a finding consistent with Watson's (1970) data for C. vpilio. The unequal adult sex ratio (29% males, Table 1) is also an indication that males may be sub- jected to the differential mortality of continued molting. The agreement of mean CW for adults collected off the Oregon coast in the study and that of Pereyra's (1966) work has an interesting corollary in the material collected from near the Farallon Islands. The mean carapace width of instars VI and VII for C. tanneri collected west of the Farallon Islands is 19.4 and 27.3 mm respectively. The cara- pace widths for the same instars collected from Oregon is 19.8 and 26.7 mm. Childress and Price (1978) credited the constant increase in size between each pair of instars in the deep-living, midwater 977 FISHERY BULLETIN: VOL. 84, NO. 4 Figure 5.— Growth phases for juvenile and adult Chionoecetes tanneri. CPL = Chelar propodus length (mm); CW = Carapace width (mm); AW = Abdomen width (mm). 200 E E •g c 3 •o o Q. o sz O 40 30 20- 10 Adult CPL = 1.21 CW - 57.16 r2.91 Adult AW * 0.54 CW - 3.07 jfj 2 Juvenile CPL = 0.56 CW - 3.37 X / / 2 X * r .97 X / / / / / * Juvenile 0 / T AW = 0.46 CW - 3.33 / ,2.99 20 — i 1 1 — 30 40 50 100 150 Carapace width mm c c (D (1) O Q) O C g o a> E 3 Z 4- 2- 0 4 2 0 4 2 -i r November 1973 N = 1 i — I — r-^| 1 r June 1974 N=10 -i 1 ra*T i — r July 1974 N=4 -i 1 1 1 N August 1974 y '■■* -i r January 1975 N=14 \ in T 1 1 1 "■ "f""\> March 1975 N=3 -i r 2 "1 1 i 1 — "7— 6 8 10 Table 1 . — Percent increase in mean carapace width and sex ratio (males %) for successive instars of Chionoecetes tanneri. Increase in carapace width (%) 48.4 35.8 31.7 43.0 36.5 36.6 N Males (0/0) Carapace width (mm) Instar Mean s2 I 16 53 3.80 0.25 II 19 — 5.64 0.62 III 18 53 7.66 0.53 IV 175 49 10.09 0.57 V 281 49 14.43 0.68 VI 499 50 19.69 1.18 VII 268 53 26.83 3.40 Adults 411 29 Carapace Width (mm) Figure 6.— Size- frequency histograms representing early juveniles with carapace widths < 9 mm. These samples were collected from stomachs of benthic fish. The dashed line represents the progres- sion of instars through time with first crab stage in June to instar II in August and instar III in January. 978 TESTER and CAREY: INSTAR IDENTIFICATION OF SPIDER CRABS Cruise 1 April 1973 N=93 Figure 7.— Size-frequency histograms representing juveniles with carapace widths 10-50 mm wide. The dashed line represents progression of instars through time with instar IV in April, instar V in August and instar VI in October through March. c c (0 •4-+ CO CO o <0 o c o !E O .Q E 3 Cruise 3 October 1973 N=143 Cruise 4 November 1973 N= 102 Carapace Width (mm) E E »-* X h- Q 30 -r 25 20- 15 LU o < tt 10- < o megalopoa «S9» n — i — i — i — i — i — i — i — i — i — i — | — i — i — i — i — r January January TIME (month) T 1 1 — I 1 — J January Figure 8.— Growth rate of Chionoecetes tanneri from egg to seventh instar is estimated to be at least 20 mo. Dotted lines indicate standard deviation. 979 FISHERY BULLETIN: VOL. 84, NO. 4 mysid, Gnathophausis ingens, to the physical and biotic stability of this species' environment. Various environmental factors can alter both the dimensions and the number of molts in many species of crus- taceans. At depths of maximum abundance (500-775 m) of C. tanneri, the annual water ranges from 2.3° to 5.6°C and certainly this uniform environment contributes to the consistency of instar size and size at maturity. ACKNOWLEDGMENTS We appreciate the assistance of Brian Oliver. Howard Horton and David Colby reviewed this manuscript and their time and efforts have greatly contributed to it. Robert S. Carney was responsible for the loan of Chionoecetes tanneri from the Smith- sonian Institution. This research was funded by NOAA (maintained by the U.S. Department of Com- merce) Sea Grant Institution Grant Nos. NOAA 04-3-158-1 and NOAA 04-5-158-2. Data analysis was facilitated by a grant from the Oregon State Uni- versity Computer Center. LITERATURE CITED Baross, J. A., P. A. Tester, and R. Y. Morita. 1978. Incidence, microscopy, and etiology of exoskeleton lesions in the tanner crab, Chionoecetes tanneri. J. Fish. Res. Board Can. 35:1141-1149. Brown, R. B., and G. C. Powell. 1972. Size at maturity in the male Alaskan tanner crab, Chio- noecetes bairdi, as determined by chela allometry, reproduc- tive tract weights, and size of precopulatory males. J. Fish. Res. Board Can. 29:423-427. Carey, A. G., and H. Heyamoto. 1972. Techniques and equipment for sampling benthic organ- isms. In A. T. Pruter and D. L. Alverson (editors), Colum- bia River estuary and adjacent ocean waters: Bioenviron- mental studies, p. 378-412. Univ. Wash. Press, Seattle. Childress, J. J., and M. H. Price. 1978. Growth rate of the bathypelagic crustacean Gnatho- phausia ingens (Mysidacea: Lophogastridae). I. Dimen- sional growth and population structure. Mar. Biol. (Berl.) 50:47-62. Donaldson, W. E., R. T. Cooney, and J. R. Hilsinger. 1981. Growth, age and size at maturity of tanner crab, Chio- noecetes bairdi M. J. Rathbun, in the northern Gulf of Alaska (Decapoda, Brachyura). Crustaceana 40:286-302. Gould, S. J. 1966. Allometry and size in ontogeny and phylogeny. Biol. Rev. 41:587-640. Hartnoll, R. G. 1969. Mating in the Brachyura. Crustaceana 16:161-181. Hilsinger, J. R., W. E. Donaldson, and R. T. Cooney. 1975. The Alaskan snow crab, Chionoecetes bairdi, size and growth. Univ. Alaska Sea Grant Rep. No. 75-12, Inst. Mar. Sci., 75-6 p. Lough, R. G. 1974. Dynamics of crab larvae (Anomura, Brachyura) off the central Oregon coast, 1969-1971. Ph.D. Thesis, Oregon State Univ., Corvallis, 299 p. Lux, F. E., A. R. Ganz, and W. F. Rathjen. 1982. Marking studies on the red crab (Geryon quinquedens) Smith off southern New England. J. Shellfish. Res. 2:71- 80. Miller, R. J., and J. Watson. 1976. Growth per molt and limb regeneration in the spider crab, Chionoecetes opilio. J. Fish. Res. Board Can. 33: 1644-1649. Pereyra, W. T. 1966. The bathymetric and seasonal distribution, and repro- duction of adult tanner crabs, Chionoecetes tanneri Rathbun (Brachyura: Majidae), off the northern Oregon coast. Deep- Sea Res. 13:1185-1205. 1968. Distribution of juvenile tanner crabs (Chionoecetes tan- neri) Rathbun, life history model, and fisheries management. Proc. Natl. Shellfish. Assoc. 58:66-70. 1972. Bathymetric and seasonal abundance and general ecology of the tanner crab, Chionoecetes tanneri Rathbun (Brachyura: Majidae) off the northern Oregon coast. In A. T. Pruter and D. L. Alverson (editors), Columbia River estuary and adjacent ocean waters: Bioenvironmental studies, p. 538-582. Univ. Wash. Press, Seattle. Rathbun, M. J. 1925. The spider crabs of America. Bull. U.S. Natl. Mus. 129, 613 p. SlNDERMANN, C. J. 1970. Principal diseases of marine fish and shellfish. Acad. Press, N.Y., 369 p. SOMERTON, D. A. 1981 . Regional variation in the size of maturity of two species of tanner crab (Chionoecetes bairdi and C. opilio) in the eastern Bering Sea, and its use in defining management subareas. Can. J. Fish. Aquat. Sci. 38:163-174. Tessier, G. 1960. Relative growth. In T. Waterman (editor), The phys- iology of Crustacea, Vol. 1, p. 537-560. Acad. Press, N.Y. U.S. National Marine Fisheries Service. 1985. Fisheries of the United States, 1984. U.S. Natl. Mar. Fish. Serv., Curr. Fish. Stat. 8360, 121 p. Watson, J. 1970. Maturity, mating, and egg laying in the spider crab, Chionoecetes opilio. J. Fish. Res. Board Can. 27:1607-1616. 980 NOTES COMPARISON OF CATCHES IN 4.3 M AND 12.2 M SHRIMP TRAWLS IN THE GULF OF MEXICO Shrimp trawls used to assess shrimp and fish popula- tions in the southern United States have varied in length, width, and basic design, making comparisons of results among studies difficult. Fishery manage- ment plans by State and Federal agencies emphasize the need for data that can be reliably compared. Techniques and equipment necessary to measure trawl performance so that data collected with dif- ferent trawls can be compared is costly and time consuming (Watson 1976; Loesch et al. 1976; Wathne 1977; Kjelson and Johnson 1978). Recent emphasis has been placed on standardizing gear and sampling methods (Watson and Bane 1985) and determining the effects on catch and mean length of organisms for different tow durations, mesh sizes, trawl widths, and towing vessels (Clark 1963; Chit- tenden and Van Engle 1972; Green and Benefield 1982; Matthews 1982; Cody and Fuls 1985). How- ever, sample sizes generally have been small and only selected species have been analyzed. The present study evaluates small trawls as popu- lation sampling devices for penaeid shrimp and other organisms in the Gulf of Mexico. The objective of this study was to compare the catch rates and mean lengths of organisms caught with 4.3 m and 12.2 m trawls pulled during day and night. Materials and Methods The study area was the Gulf of Mexico off Texas between the Colorado River and Port Mansfield in depths from 7 m to 24 m (Fig. 1). Sample sites were established in 1° latitude by 1° longitude grids within the study area. Twenty randomly selected sites were sampled monthly from November 1982- February 1983. Samples were equally and random- ly distributed between day and night. At each site two trawls were towed simultaneous- ly for 15 min at approximately 3 kn from the Texas Parks and Wildlife Department (TPWD) RV Western Gulf, a double-rigged 21.9 m steel-hull shrimp trawler. The 4.3 m trawl (small net) was spread by wooden trawl doors 0.4 m high and 0.8 m long and the 12.2 m wide trawl (large net) was spread by wooden trawl doors 0.9 m high and 2.1 m long. Both nets had 5.1 cm stretched mesh web- bing in the body, 4.4 cm mesh in the bag, and were equipped with tickler chains. Trawl catches weighing <10 kg were processed by identifying and counting all organisms in the catch. For larger catches a 10 kg subsample was ran- domly selected from the total catch, and the total number for each species was estimated by dividing subsample counts by the proportion of subsample weight to total weight. Total lengths were measured on no more than 50 individuals of each Penaeus shrimp species and no more than 20 individuals of all other species. The arithmetic mean for length data was calculated for each species in each sample. The relationship between number caught (or mean length) in the two trawls was tested for linear, mul- tiplicative and exponential models, and log and square root transformations (Sokal and Rohlf 1981). No significant improvement was found over a linear regression with no transformation. Mean length regressions were developed for species with 10 or more pairs of mean length data (>2 measurements) in each size of trawl (Fig. 2). Catch regressions were developed for those species that were present in at least 20 samples in the large net and were repre- sented by at least 5 samples with >20 individuals in the small net. This insured a sufficiently wide distribution to yield meaningful results (Fig. 3). Differences (P < 0.01) between day and night regressions for each species were evaluated using analysis of co variance (Snedecor and Cochran 1980). Results Small trawls can be used to obtain trend data on mean lengths of species caught in offshore waters. Relationships exist between the catch in the 4.3 m trawl vs. the catch in the 12.2 m trawl. No signifi- cant differences were found in the day-night regres- sions of mean length for any species tested. There was no difference in the day-night catch vs. catch relationship for total organisms or Penaeus setiferus but one did exist for Trachypenaeus sp. and Squilla empusa. Mean lengths in the two trawls were directly cor- related for all species that met criteria for regres- sion analysis (Table 1). The regressions of the mean length of fish caught in one net vs. the other for day and night were not significantly different for any FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 981 Figure 1.— Gulf of Mexico sampling area off the Texas coast for 4.3 m and 12.2 m trawls towed simultaneously during November 1982-February 1983. 982 of the species tested (Table 2). The combined regres- sions had significant positive correlations (0.51-0.89) explaining 26-79% of the variation. Catch per tow in the two trawls was positively cor- related. Correlation coefficients (0.48-0.93) were significant for all species meeting the criteria for analysis (Table 3). The percent of variation explained (r2) varied from 23 to 86%. There were no significant differences in the day- night catch vs. catch relationships for total organ- Table 1.— Linear regression results of 4.3 m trawl mean length (X,) versus the 12.2 m trawl length (V,) for selected species. Species Time Range of Number /-intercept Slope (b) Correlation coefficient S2 Y ■ X 95% confidence interval of b Penaeus setiferus Day Night Combined 93-135 94-164 93-164 29 32 61 12.31 16.29 14.48 0.91 0.87 0.89 0.85** 0.93** 0.88** 61.33 22.98 39.86 0.68-1.13 0.74-1.00 0.76-1.01 Stellifer lanceolatus Day Night Combined 44-125 44-115 44-125 11 24 35 26.35 28.32 27.42 0.70 0.67 0.68 0.91** 0.88** 0.89** 88.77 65.60 68.19 0.45-0.95 0.51-0.83 0.56-0.80 Trachypenaeus sp. Day Night Combined 50-78 50-84 50-84 22 36 58 38.03 43.03 41.77 0.47 0.40 0.42 0.61** 0.61** 0.62** 17.55 19.28 18.03 0.18-0.76 0.22-0.58 0.28-0.56 Portunus gibbesii Day Night Combined 30-48 30-55 30-55 14 30 44 26.44 23.81 25.43 0.39 0.41 0.38 0.53* 0.62** 0.56** 11.78 9.70 10.53 -0.01-0.78 0.21-0.61 0.21-0.56 Squilla empusa Day Night Combined 77-104 48-132 48-132 10 31 41 49.37 69.43 65.78 0.46 0.32 0.34 0.62ns 0.51** 0.51** 48.08 86.95 83.87 0.04-0.89 0.11-0.52 0.15-0.53 Cynoscion nothus Day Night Combined 62-110 70-122 62-122 25 21 46 52.02 45.59 46.18 0.42 0.44 0.46 0.63** 0.71** 0.67** 43.87 30.64 42.16 0.20-0.64 0.23-0.65 0.31-0.62 *P< 0.05. "P<0.01. Table 2.— Summary of ANCOVA for mean length of selected species. Calculated Calculated Calculated Fsfor Fsfor Fsfor Species df W0:o, = o2 df H0:p, = p2 df HQ:ct, = a2 Penaeus setiferus (27,30) 2.67 ns (1,57) 0.04 ns (1,58) 0.07 ns Stellifer lanceolatus (9,22) 1.35 ns (1,31) 0.03 ns (1,32) 0.04 ns Trachypenaeus sp. (34,20) 1.10 ns (1,54) 0.09 ns (1,55) 0.00 ns Portunus gibbesii (12,28) 1 .22 ns (1,40) 0.00 ns (1,41) 2.89 ns Squilla empusa (29,8) 1.81 ns (1,37) 0.15 ns (1,38) 4.46 ns Cynoscion nothus (23,19) 1 .43 ns (1,42) 0.01 ns (1,43) 7.12 ns Table 3.— Linear regression results of 4.3 m trawl catch/tow (X,) versus the 12.2 m trawl catch/tow (V,) for total organisms and selected species. Species Time Range of Number /-intercept Slope (b) Correlation coefficient S2YX 95% confidence interval of b Total organisms Day Night Combined 16-212 43-210 16-212 40 40 80 352.71 593.65 420.42 5.83 6.04 6.53 0.58** 0.48** 0.55** 143,234.17 310,412.80 237,569.91 3.18-8.47 2.45-9.64 4.31-8.75 Penaeus setiferus Day Night Combined 0-55 0-51 0-55 40 39 79 12.21 -0.87 7.40 5.37 6.96 6.14 0.90** 0.87** 0.88** 1,129.54 2,291.77 1,757.99 4.51-6.23 5.65-8.26 5.38-6.90 Squilla empusa Day Night 0-28 0-37 40 39 6.03 -5.58 4.50 6.81 0.93** 0.85** 139.44 1,438.75 3.92-5.07 5.38-8.25 Trachypenaeus sp. Day Night 0-45 0-43 40 40 20.63 60.23 13.38 19.51 0.80** 0.73** 13,040.70 40,354.41 10.15-16.60 13.46-25.67 Portunus gibbesii Night 0-114 40 24.65 5.92 0.78** 9,961.84 4.40-7.45 Lolliguncula brevis Day 0-42 40 9.92 1.66 0.72** 292.57 1.14-2.19 **P< 0.01. 983 170 160 150 140 130 -J 120 £ 175 250 200 175 150 - 125 - 100 75 50 - 10 15 NUMBER PER TOW IN 4.3 M TRAWL Figure 3.— Regression of catch per tow in 12.2 m trawl (Yt) on catch per tow in 4.3 m trawl (AT,) for comparative tows during November 986 1 200- TRACHYPENA£US5/tow in either net. tionships between day and night catches in a fishery independent assessment program can increase sam- pling frequency and decrease the cost of sampling by reducing processing time, manpower require- ments, and variability caused by subsampling large catches. Samples from the small trawl could be pro- cessed in approximately 25% of the time required for sample processing from the large trawl. The small trawl required no subsampling. Management agencies should consider these findings when plan- ning long-term programs. Acknowledgments We would like to express our appreciation to each member of the Gulf Research Program who so conscientiously collected scheduled samples. Thanks are extended to the Texas Parks and Wildlife Department review committee and an unknown reviewer for their valuable comments. Nancy Ziegler prepared the manuscript. This study was conducted with partial funding from the U.S. Department of Commerce; National Oceanic and At- mospheric Administration, National Marine Fish- eries Service, under P.L. 88-309 (Project 2-385- R). Literature Cited Chittenden, M. E., Jr., and W. A. Van Engel. 1972. Effect of a tickler chain and tow duration on trawl catches of the blue crab, Callinectes sapidus. Trans. Am. Fish. Soc. 101:732-734. Christmas, J. Y., and D. J. Etzold. 1977. The shrimp fishery of the Gulf of Mexico United States: a regional management plan. Gulf Coast Res. Lab., Ocean Springs, MS, Tech. Rep. Ser. No. 2, 128 p. Clark, J. R. 1963. Size selection of fish by otter trawls. Results of recent experiments in the Northwest Atlantic. D. Effect of dura- tion of tow on codend escapement. Int. Comm. Northwest Atl. Fish. Spec. Pub. 5, p. 55-58. Cody, T. J., and B. E. Fuls. 1985. Comparison of the catch rates of three trawls in off- shore Texas waters. In J. W. Watson and N. Bane (editors), Proceedings of the SEAMAP shrimp and bottomfish sam- pling gear workshop, p. 19-29. Gulf States Mar. Fish. Comm. No. 12. Green, A. W., and R. L. Benefield. 1982. Mesh size selectivity study of penaeid shrimp trawled from Galveston Bay, Texas May 1981. Tex. Parks Wildlf. Dep., Coast. Fish. Branch, Manage. Data Ser. No. 40, 9 p. Kjelson, M. A., and G. N. Johnson. 1978. Catch efficiencies of a 6.1-meter otter trawl for estu- arine fish populations. Trans. Am. Fish. Soc. 107:246-254. Loesch, H., J. Bishop, A. Crowe, R. Kuckyr, and P. Wagner. 1976. Technique for estimating trawl efficiency in catching brown shrimp (Penaeus aztecus), Atlantic croaker (Micro- pogon undulatus) and spot (Leiostomus xanthurus). Gulf 990 Res. Rep. 5(2):29-33. Matthews, G. A. 1982. Relative abundance and size distributions of commer- cially important shrimp during the 1981 Texas closure. Mar. Fish. Rev. 44(9-10):5-15. Snedecor, G. W., and W. G. Cochran. 1980. Statistical methods. 7th ed. Iowa State Univ. Press., Ames, 507 p. SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry: The principles and practices of statistics in biological research. 2d ed. W. H. Freeman and Co., San Francisco, 859 p. Wathne, F. 1977. Performance of trawls used in resource assessment. Mar. Fish. Rev. 39(6):16-23. Watson, J. W., Jr. 1976. Electric shrimp trawl catch efficiency for Penaeus duorarum and Penaeus aztecus. Trans. Am. Fish. Soc. 105: 135-148. Watson, J. W., Jr., and N. Bane (editors). 1985. Proceedings of the SEAMAP shrimp and bottomfish sampling gear workshop. Gulf States Mar. Fish. Comm. No. 12, 80 p. Terry J. Cody Billy E. Fuls Texas Parks and Wildlife Department Coastal Fisheries Branch 4200 Smith School Road Austin, TX 787U Long Island to Chesapeake Bay, spawning occurs in offshore coastal waters from October to Decem- ber and from March to May. From North Carolina to Florida, spawning occurs in offshore coastal waters from October through March and this spawn- ing population consists of fish that have migrated from the north and contains all age groups (Nichol- son 1978). The gulf menhaden, which is distributed zonally, is restricted to the Gulf of Mexico and ranges from Cape Sable, FL, to Vera Cruz, Mexico (Reintjes 1969). Their maximum reported age is ap- proximately 4 yr, and they may spawn for approx- imately 2 yr (Lewis and Roithmayr 1981). They spawn from October through March in nearshore and offshore waters within the 110 m depth contour (Christmas and Waller 1975). Both species use estu- aries as nursery areas for more than half their first year of life. The major objectives of this study were to examine and compare early life history characteristics of these two menhadens and to investigate the effects of temperature on developmental processes. Char- acteristics examined were egg size, size at hatching, yolk utilization rates, yolk volume at first feeding, size and age at first feeding, and growth. EARLY LIFE HISTORY OF ATLANTIC MENHADEN, BREVOORTIA TYRANNUS, AND GULF MENHADEN, B. PATRONUS Atlantic menhaden, Brevoortia tyrannus, and gulf menhaden, B. patronus, are allopatric, morphologi- cally similar clupeids with contrasting distributional patterns and reproductive traits. The Atlantic men- haden has a meridional distribution and encounters variable environmental conditions during its life- time. It occurs along the eastern coast of North America from Nova Scotia to Florida, and its dis- tribution is stratified by age and size, with the older and larger fish ranging farther north (Nicholson 1978). Atlantic menhaden are a relatively long-lived clupeid. Their maximum reported age is approx- imately 10 yr, and they may spawn for approximate- ly 7 yr (Higham and Nicholson 1964; Nicholson 1975). The spatial and temporal spawning habits of Atlantic menhaden are more complex than those of its congener. In Long Island Sound and New Eng- land waters, limited spawning occurs in inshore waters during the summer and early fall. From Methods Atlantic menhaden were collected with a commer- cial purse seine from the Newport River, NC, dur- ing the summer. Fish were held in the laboratory at ambient temperatures for approximately 4 mo before spawning. Gulf menhaden were collected in late September by cast net near Gulf Breeze, FL, and transported to the laboratory by methods devel- oped by Hettler (1983). They were held in the lab- oratory at ambient temperatures for about 1 mo before spawning. For each spawning, about 10 men- haden were induced to spawn by methods described by Hettler (1981, 1983). Eggs were spawned in approximately 20 °C water during the night and col- lected the following morning. All experiments ex- cept those dealing specifically with growth were con- ducted in 10 L rearing tanks; growth experiments were conducted in 60 L rearing tanks. Tanks were set in a temperature controlled water bath with two 40-W fluorescent lamps positioned 40 cm above each tank, and the tanks were illuminated for 12 h daily. Temperatures were controlled to approximately ±0.5°C. Salinities ranged from 28%>o to 32%o. Rotifers, Brachionus plicatilis, were used as food for about the first 10 d, and Artemia nauplii and rotifers were used thereafter. Feeding levels were not controlled, but, based on experience, we pro- FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 991 vided food in densities we felt would not limit growth. Growth in standard length (SL) from the time lar- vae begin feeding to age 21 d at 20° C was modeled by an exponential equation. All measurements were made on eggs and larvae that were preserved in 5% sodium acetate buffered Formalin1. Volumes (V) of the elliptically shaped yolk mass were calculated using the formula for a prolate spheroid V = (n/6) lh2, where I is the length and h is the height of the yolk mass (Blaxter and Hempel 1963). We were unable to treat the two species the same in most experiments. The gulf menhaden was sub- jected to a greater number of treatments than the Atlantic menhaden. Experiments dealing with starvation and yolk utilization rates were conducted only on the gulf menhaden. In addition, the lack of replications for some experiments limited the ap- plication of statistical tests (e.g., ANOVA) and, as a result the differences or similarities between the two menhadens, should be considered tentative. Results and Discussion Based on a sample of eggs from the single spawn of a group of approximately five females from each species, Atlantic menhaden had significantly (P < 0.001) larger eggs (1.6 mm diameter, N = 20) than gulf menhaden (1.3 mm diameter, N = 20). Egg sizes for both these species that have been reported (Houde and Fore 1973; Jones et al. 1978; Hettler 1984) support our observations that Atlantic men- haden eggs are larger than gulf menhaden eggs. Atlantic menhaden larvae measured at hatching also were larger than gulf menhaden (Fig. 1) and sup- ports Blaxter and Hunter's (1982) view that egg size greatly influences the size of larvae at hatching. Temperature did not affect the size at hatching of gulf menhaden (Fig. 1), but the rate of yolk utiliza- tion was affected by temperature and was roughly 2.5 times faster at the highest temperature (24 °C) than at the lowest temperature (14°C) (Table 1). The instantaneous rate of yolk utilization increased linearly with increasing temperature (Fig. 2). The volume of yolk at the onset of exogenous feeding (first feeding) was approximately similar at all tem- peratures (Table 1) and was not affected by temper- ature (ANOVA, P = 0.13). The size of gulf menhaden at first feeding was in- dependent of temperature (Fig. 3) (ANOVA P = 0.15) and, although data are limited, the size of Atlantic menhaden also was independent of tem- perature. The age at first feeding, however, was dependent on temperature (Fig. 3). An ANCOVA (log transformed ages on temperature) revealed that the regression slopes were similar (P = 0.37), in- i- < E E O z ai DW s« ZO - ~ 2 2 HI < CO 3 o UJ z < z o I- < N < H I- z> CO z 3.00 2.50 2.00 1.50 1.00- C 0.50 - 0.00- Gulf Menhaden Y= 0.736+0. 12426(X) ra = 0.95 _L 14 16 18 20 22 24 TEMPERATURE °C Figure 2.— The effects of temperature on the instantaneous rate of yolk utilization for gulf menhaden. dicating a similar response to temperature by both species. But the Y-intercepts differed significantly (P = 0.02) indicating that, over the range of tem- peratures tested, the Atlantic menhaden fed at a significantly earlier age than gulf menhaden. For both species the age at first feeding declined ex- ponentially with increasing temperatures. Atlantic menhaden were larger than the gulf menhaden at first feeding (Fig. 3). At 20° C, Atlantic and gulf menhaden growth rates were similar (ANCOVA, P = 0.36), but Atlantic menhaden maintained a size advantage during the early larval period (Table 2). This difference was attributed to differential size and age at first feeding. The ability of early larvae to withstand the depri- vation of food was influenced by temperature (Table 3). Although at 20 °C mortalities may be attributed to causes other than starvation (compare control and starved), at progressively higher temperatures lar- vae are less able to withstand the deprivation of food. For example, at 24°C, gulf menhaden must find food within three days after the onset of first E E I I- CD O Z z III Q 1 HI Q DC < HI 1- rn UJ CO CM z rr +i < LL 1- H co < z < Lit 2 o z CO Q >» UJ « UJ n Ll_ *w h- UJ fO O rr < LL 1- < 5.5- 5.0- 4.5- 4.0 8 6 4 2 {Atlantic Menhaden - o Gulf Menhaden - t - { 1 { a • i i 6 X 1 { 1 • Atlantic Menhaden o Gulf Menhaden / Y=30.7e-°-11454 r2=0.99 _l L Y=25.1e / r2 = 0.96 -0.0929(X) 14 16 18 20 22 24 TEMPERATURE °C Figure 3.— The size and age when gulf and Atlantic menhadens begin feeding on exogenous food sources at different temperatures. Each point represents a sample of about 10 fish. Replicate experiments were only conducted for gulf menhaden and only at 14°, 20°, 22°, and 24°C. 993 Table 2.— Growth of larval gulf and Atlantic menhadens from time of first feeding to age 21 d at 20°C. W1 Growth parameters2 a b r2 Estimated SL (mm) Species First Age feeding 21 d Gulf menhaden Atlantic menhaden 11 16 3.36 4.38 0.04640 0.04267 0.97 0.95 4.0 8.9 5.0 10.7 'Number of samples; about 10 fish per sample. 2SL (mm) = a x exp b (age in d). tuating environment producing more reproductive uncertainty (Murphy 1968; Stearns 1976). This in- formation suggests to us that the subtle differences we observed may indicate a fine tuning of reproduc- tive strategies that allow these menhadens to per- sist in their particular environments. A more rigor- ous comparative study is required before we can understand how menhaden life history character- istics are adapted to their particular environments. Such a study is presently underway by the senior author. Table 3.— The survival (%) of first-feeding gulf menhaden larvae deprived of food (starved) in relation to temperature. The fed treat- ment represents the control group. Temper- ature (°C) Treatment N Days past time of first feeding 1 2 3 4 5 6 7 18 Starved 25 100 100 100 100 92 32 0 Fed 25 100 100 100 100 96 92 92 20 Starved 25 92 76 72 48 8 4 0 Fed 25 88 84 84 84 72 68 68 22 Starved 20 100 100 80 75 0 Fed 20 100 100 100 100 100 24 Starved 25 56 40 40 4 0 Fed 25 96 96 96 96 96 feeding or high mortalities will occur; whereas at 18 °C they can survive without food for 5 d without incurring high mortalities (Table 3). The gulf men- haden's response to starvation in relation to tem- perature is comparable to numerous temperate zone, pelagic fish larvae (McGurk 1984). In conclusion, although temperature is an impor- tant factor in controlling the development of marine fish larvae (Blaxter 1970), we observe that temper- ature was not a determinant of size at hatching, size at first feeding, and yolk volume remaining at first feeding. These data suggest that age is not a good correlate of these developmental events. On the other hand, temperature had an effect on the rate of yolk utilization, the time between hatching and exogenous feeding, and the ability of larvae to with- stand the deprivation of food. Our observations, although limited by a lack of rigorous statistical testing, suggest that, relative to gulf menhaden, Atlantic menhaden produced larger eggs, were larger at hatching, were larger and younger at time of first feeding, and appeared to maintain a larger size throughout the early larval period. We tried to interpret these differences in the context of their entire life history. Relative to gulf menhaden, Atlantic menhaden exhibit life history traits (later maturity, longer life, and more repro- ductive years) that may be adapted to a more flue- Acknowledgments Sincere appreciation is extended to J. Govoni, D. Peters, and two anonymous reviewers for their critical review of the manuscript. W. Hettler and C. Lewis provided technical support during various phases of the study. This research was supported by a contract from the Ocean Assessments Division, National Ocean Service, National Oceanic and At- mospheric Administration. Literature Cited Blaxter, J. H. S. 1970. Development: eggs and larvae. In W. S. Hoar and D. J. Randall (editors), Fish physiology, Vol. 3, p. 178-252. Acad. Press, Inc. N.Y. Blaxter, J. H. S., and G. Hempel. 1963. The influence of egg size on herring larvae (Clupea harengus L.). J. Cons., Cons. Int. Explor. Mer 28:211-240. Blaxter, J. H. S., and J. R. Hunter. 1982. The biology of clupeoid fishes. Adv. Mar. Biol. 20: 1-223. Christmas, J. Y., and R. S. Waller. 1975. Location and time of menhaden spawning in the Gulf of Mexico. Gulf Coast Res. Lab., Ocean Springs, MS, 20 p. Hettler, W. F. 1981. Spawning and rearing Atlantic menhaden. Prog. Fish- Cult. 43:80-84. 1983. Transporting adult and larval gulf menhaden and tech- niques for spawning in the laboratory. Prog. Fish-Cult. 45:45-48. 1984. Description of eggs, larvae, and early juveniles of gulf menhaden, Brevoortia patronus, and comparisons with Atlantic menhaden, B. tyrannus, and yellowfin menhaden, B. smithi. Fish. Bull., U.S. 82:85-95. HlGHAM, J. R., AND W. R. NICHOLSON. 1964. Sexual maturation and spawning of Atlantic menhaden. U.S. Fish Wildl. Serv., Fish. Bull. 63:255-271. Houde, E. D., and P. L. Fore. 1973. Guide to the identity of eggs and larvae of some Gulf of Mexico clupeid fishes. Fla. Dep. Nat. Resour., Mar. Res. Lab., Leafl. Ser. 4(23):1-14. Jones, P. W., F. D. Martin, and J. D. Hardy, Jr. 1978. Development of fishes of the Mid- Atlantic Bight. Vol. 1, Acipensaridae through Ictaluridae. U.S. Fish. Wildl. Serv., Biol. Serv. Program FWS/OBS-78/12, 314 p. 994 Lewis, R. M., and C. M. Roithmayr. 1981. Spawning and sexual maturity of gulf menhaden, Brevoortia patronus. Fish. Bull., U.S. 78:947-951. McGurk, W. D. 1984. Effects of delayed feeding and temperature on the age of irreversible starvation and on the rates of growth and mortality of Pacific herring larvae. Mar. Biol. 84:13-26. Murphy, G. I. 1968. Pattern in life history and the environment. Am. Nat. 102:391-403. Nicholson, W. R. 1975. Age and size composition of the Atlantic menhaden, Brevoortia tyrannus, purse seine catch, 1963-71, with a brief discussion of the fishery. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-684, 28 p. 1978. Movements and population structure of Atlantic men- haden indicated by tag returns. Estuaries 1:141-150. Reintjes, J. 1969. Synopsis of biological data on the Atlantic menhaden, Brevoortia tyrannus. U.S. Fish Wildl. Serv., Circ. 320, 30 P- Stearns, S. C. 1976. Life-history tactics: a review of the ideas. Q. Rev. Biol. 51:3-47. Allyn B. Powell Southeast Fisheries Center Beaufort Laboratory National Marine Fisheries Service, NOAA Beaufort, NC 28516 USA Germano Phonlor Fundacao Universidade do Rio Grande Departamento de Oceanografia Caixa postal k7k 96200 Rio Grande - RS, BRAZIL SEASONALITY OF BLUE MUSSEL, MYTILUS EDULIS L., LARVAE IN THE DAMARISCOTTA RIVER ESTUARY, MAINE, 1969-771 (Engle and Loosanoff 1944; Stubbings 1954; Baird 1966; Bohle 1971; Rasmussen 1973; Jorgensen 1981; Kautsky 1982). Seed (1975) summarized reproduction in Euro- pean mussel populations and found that spawning in M. edulis varies with latitude, occurring earlier in warm waters and progressively later in cooler, northern waters. However, Newell et al. (1982) reported no latitudinal variation of spawning among mussel populations along the northwestern Atlan- tic coast. Such geographic variation has been attrib- uted to the existence of physiological races (Stauber 1950; Loosanoff and Nomejko 1951). Newell et al. (1982) and Fell and Belsamo (1985) also found that mussel populations at the same latitude in Long Island Sound spawn at different temperatures and times of the year. They surmised that food avail- ability, rather than temperature, dictates when spawning occurs. Factors which are important in the timing and intensity of spawning can be determined by moni- toring spawning activity. This may be achieved directly, by examination of gonad development in seasonally collected samples, or indirectly, by ob- serving the presence or absence of M. edulis larvae in plankton samples (Chipperfield 1953). While the direct method is preferable, the indirect method does allow one to use long-term plankton records. These provide an estimate of the variation in both the timing and intensity of spawning. Since the source of the larvae is not certain, some caution should be used in the interpretation of the results (Seed 1975). An 8-yr plankton record of Mytilus larval abun- dance presents an unusual opportunity to observe long-term variability in spawning and larval occur- rence. Specifically, the data were examined with the following goals: The spawning of the blue mussel, Mytilus edulis L., has been the subject of many studies (see Bayne 1976 for partial review). In an early paper Field (1922) reported that gametogenesis and spawning were influenced by water temperature, though he provided no data. Chipperfield (1953) found that mussels spawn over a specific range of water tem- perature (9.5°-12.5°C). In addition, Chipperfield noted that the rate of temperature change prior to spawning influences intensity. Other investigators have found that mussels spawn over a specific tem- perature range, which may vary among locales Contribution No. 183, Ira C. Darling Center, University of Maine, Orono, ME. 1) Determination of the initiation and the dura- tion of the spawning season and degree of tem- poral variation between years; 2) Determination of the variation in larval abun- dances within and between seasons; 3) Examination of the possible correlation of en- vironmental variables (temperature, phyto- plankton abundance, degree days, calendar date, and lunar cycles) with spawning activity. Materials and Methods The study site was the Damariscotta River estuary (Fig. 1), a narrow embayment, 29 km long, which receives a limited amount of freshwater. The estu- FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 995 arine portion has a MLW (mean low water) volume of 123.4 x 106 m3, a tidal volume of 56.2 x 106 m3, and a mean summer flushing time of 4-5 wk DAMARISCOTTA DAMARISCOTTA RIVER LOCATION MAP 1 ,5 0 1 NAUTICAL MILES 1 .5 0 1 2 KILOMETERS - 44" 00' - 43* 55' 43* 50' (McAlice 1977). The estuary is stratified near its head but approaches a well-mixed condition further seaward. Tides are semi-diurnal with a mean range of 2.7 m and a tidal excursion of about 2.8 km (Lee and McAlice 1979). Monthly plankton samples were collected during daylight at station D7 (Fig. 1) from October 1969 to June 1970 and then biweekly until September 1977. Plankton tows were 10-15 min oblique hauls with #20 mesh (76 /im) nets of 0.5 m mouth diam- eter equipped with centrally mounted flowmeters. Maximum depths of tows were 10-15 m (4-5 m above the bottom). Boat speed was 1-2 m s-1. Samples were immediately fixed in 4% buffered Formalin2. Laboratory subsampling followed the method recommended by Frolander (1968). The concen- trated plankton was diluted to a known volume, thoroughly stirred, and a 1 mL aliquot removed with a Stempel pipette. Initial counts on samples taken from June 1974 to September 1977 did not distin- guish among taxa of larval bivalves. We therefore took an additional subsample, determined the per- centage of Mytilus in 50 bivalve larvae, and multi- plied this by the total veliger abundance to obtain Mytilus densities for each sampling period. Several key publications (Loosanoff et al. 1966; Chanley and Andrews 1971; DeScweinitz and Lutz 1976; Lutz and Hidu 1979) containing photomicro- graphs and descriptions were used to identify Mytilus edulis larvae. The differentiation of Mytilus edulis larvae from other mytilid larvae (Modiolus modiolus and Geukensia demissa) at the straight hinge stage was achieved by comparing the length of the hinge line as well as total shell length and height. The early and late umbo larvae of Geuken- sia were easily distinguishable by their elongated appearance; Mytilus larvae tended to be less elongate, though pointed anteriorly (Chanley and Andrews 1971). The differentiation of Modiolus modiolus larvae and Mytilus edulis larvae was based mainly on the characteristics described by DeSchweinitz and Lutz (1976); hinge line lengths, total shell length in the 95-105 ycm range, shell shape of umbo stage larvae, presence of an eye spot in specimens <270 yxn, and the presence of a functional foot in larvae <295 \xm. Further positive identifica- tion of late stage Mytilus larvae was achieved by examining the hinge teeth of disarticulated valves (Lutz and Hidu 1979). Spawning dates were estimated by subtracting the approximate age of the larvae from the sampling 69' 35' 69'30' Figure 1.— Location map. 2Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 996 date. Larval age was estimated using photomicro- graphs of larvae of known age and size for com- parison, available in Chanley and Andrews (1971). The initial occurrence of larvae in each year was dominated by early straight hinge larvae. The spawning season was defined by larval abundances >10 m~3. This level was chosen arbitrarily to distinguish major spawning from occasional low lar- val abundances (<10 m~3). Environmental variables that were examined for correlations with the initiation of spawning and lar- val abundance included water temperature, phyto- plankton abundance, degree days, calendar date, and lunar cycles. Water temperatures were taken concurrently with the plankton samples. Phyto- plankton abundances from July 1974 to August 1977 were available for the Damariscotta River (McAlice unpubl. data). Data from the neighboring Sheepscot River estuary (McAlice and Denniston3) were sub- stituted for the period October 1969 to June 1974. The decision to use the Sheepscot data was based on the highly significant Spearman's rank correla- tion (Zar 1984) (r = 0.67, P < 0.001) between the Damariscotta and Sheepscot phytoplankton abun- dances from July 1974 to August 1977. Degree days were calculated in the manner described by Thiesen (1973). For each year, degree days were summed from the time of peak larval abundance the previous year to the initiation of spawning. Lunar cycle in- formation was obtained from tide tables published by NOAA (1969-76). once initiated, probably continued throughout the summer as indicated by the persistence of early stage mussel larvae. Spawning appeared to cease as temperatures fell to 9° -14° C in September and October (Fig. 2), when only late stage larvae were present. Maximal larval abundances were observed in mid- to late June, shortly after spawning began. At this time, straight hinge larvae, <6 d old, were dominant. Maximum values for the period 1970-75 ranged from 787 larvae m~3 to 5,400 larvae m~3. In 1976 and 1977, maximum abundances were an order of magnitude larger (3.16 x 104 m~3 and 6.09 x 104 m-3, respectively). Following the peaks in June, larval densities generally declined through 1 to 3 successively smaller peaks (Fig. 2). Mussel larvae appeared well after phytoplankton abundances had begun to increase from low winter values to generally high summer values (Fig. 3). Lar- vae usually disappeared before phytoplankton abun- dances fell to typically low winter levels. In addition to the larvae of Mytilus edulis, those of Anomia simplex, Geukensia demissa, Modiolus modiolus, and what was probably a complex of My a arenaria, Hiatella arctica, and possibly Sphenia sincera (Hanks and Packer 1985) larvae were also identified. Amonia simplex occurred most commonly from September through December, though never in great numbers. The Mya-Hiatella-Sphenia group was often very abundant, and occurred from early May through September. Geukensia and Modiolus were never common. Results Examination of the age and abundance of mussel larvae from December 1969 to September 1977 in- dicated that spawning began in late May or early June when temperatures reached 10° -12.5°C (Fig. 2). The average date when spawning began was 4 June, with a standard deviation of approximately 7 d. The average number of degree days prior to spawning was 2,853, with a standard deviation of 368. No significant relationship was found between degree days and commencement of spawning or degree days and maximum larval abundances. Commencement of spawning may be related to the time of spring tides (Table 1). In 7 of the 8 yr ex- amined, spawning began within 5 d, before and after, a spring tide. On four occasions spawning commenced within 2 d of a spring tide. Spawning, 3McAlice, B. J., and F. D. Denniston. Dominance and diversity of Sheepscot River estuary phytoplankton. Manuscr. in prep. Ira C. Darling Center, University of Maine, Walpole, ME 04573. Table 1 .—Estimated dates and temperatures of the initiation and cessation of spawning for Mytilus edulis in the Damariscotta River estuary, and dates of nearest spring tides, 1970-77. Date of Estimated date and (°C) when spawning „„,„, spring Year began ended tide 1970 2 June (10.0°-13.2°C) 2 Oct. (14.6°-13.9°C) June 4 1971 8 June (10.2°-12.2°C) 18 Oct. (14.3°-12.8°C) June 9 1972 16 June (10.5°-12.1°C) 20 Oct. (13.1°-9.0°C) June 11 1973 12 June (10.0°-14.0°C) 10 Oct. (12.7°-11.0°C) June 15 1974 12 June (10.7°-12.3°C) 8 Oct. (12.4°-9.1°C) June 4 1975 27 May (10.3°-12.5°C) 24 Sept. (16.9°-13.7°C) May 25 1976 24 May (10.1°-12.8°C) 22 Oct. (14.6°-11.0°C) May 29 1977 2 June (9.3°-10.2°C) — June 1 Discussion A temperature threshold for spawning was in- dicated by the appearance of Mytilus larvae when water temperatures exceeded 10°-12.5°C and the subsequent disappearance of larvae when tempera- tures fell below 9° -14°C. A number of studies have reported the initiation of spawning in Mytilus edulis 997 IO 1970 iij o z < a z O < B 4.0- 3.0- 2.0- < > < o IO~ 0.0- J' • '0' ' |j" 'A1 ' 'j1 ■ b1 ' |J ' 'a1 ' 'J1 ' 'o1 ' p" 'a' ' 'j 1974 1975 1976 1977 Figure 2.— Abundance ofMytilus edulis larvae (solid line) and water temperature (broken line) at station D7: A) 1969-73; B) 1974-77. at temperatures of 10°-13°C or higher while few studies have reported spawning at lower tempera- tures (Table 2), which also suggests a thermal threshold for spawning. The significance of this threshold may be linked to gametogenesis. Bayne (1965) found that mussels with fully developed gametes would not spawn when held at 5°C under high food concentrations. However, if temperatures were raised to 12°-14°C, gametes matured and spawning ensued. Similarly, Sastry (1968) found that in the bay scallop, Aequipecten irradians, oogonia and spermatozoa formed at 15 °C and 20 °C, but that temperatures higher than 20 °C were necessary for oocytes to reach a fertilizable stage. Therefore, the apparent correlation between a par- ticular temperature and the initiation of spawning may actually reflect the maturation of gametes followed by induction of spawning by any of a num- ber of stimuli. Given the predictable rise in temper- ature each spring, this may explain the initiation of spawning at approximately the same time each year. Use of degree days to predict the time of spawn- ing does not appear to be useful. This is due to a very regular pattern of rising and falling water temperatures each year. As a result, the sum of degree days between spawning periods conveyed no more information than did elapsed time. Newell et al. (1982) arrived at a similar conclusion for mussel populations in Long Island Sound. They found that one Long Island Sound mussel population spawned 3 mo later than another, despite nearly identical temperature conditions, difference in degree days due solely to a difference in elapsed calendar days. Bayne (1975), however, did find a relationship be- tween rate of gametogenesis and degree days, but not calendar days. 998 UJ 6.0- w UJ o o 5.0- 4.0- 3.0 [J ' 'A1 ' 7 ' 'o' ' P ' 'A' ' 'J1 ' 'o' ' |J' ' 'A1 ' 'j' ' '01 ' |J' ' 'A1 ' 'J1 ' 01 ' 1970 1971 1972 1973 3.0 |0i i ,Ai i .j. , lQl i jji i iai i iji i i0. i |ji i iai i iji i b. . jji" i 'iai i iji i 1974 1975 1976 1977 Figure 3.— Abundance of phytoplankton in the lower Sheepscot River estuary: A) 1969-73; B) January 1974-June 1974 and at station D7, July 1974-August 1977. Table 2.— Reported spawning temperatures and periods of Mytilus edulis. Tempera- tures Major spawning Location (°C) period Reference Europe Norway 8 early May Bohle 1971 Denmark 7-16 May Jorgensen 1981 England 9.5-12.5 May Chipperfield 1953 Sweden 12 mid-May-early June Kautsky 1982 England 13 early May Baird 1966 Denmark 13-14 May-June Rasmussen 1973 United States Damariscotta 10-13 late May-mid-June This study River, ME Milford, CT 15-16 May Engle and Loosanoff 1944 Branford, CT 14-16 late May-early June Fell and Balsimo 1985 Stony Brook, NY 11-15 late April-early June Newell et al. 1982 Shinnecock, NY 16-22 August-October Newell et al. 1982 Spawning in response to lunar cycles is also a possibility. Korringa (1947) noted that the European oyster, Ostrea edulis, spawns around the period of spring tides and attributed this to increased hydro- static pressure. Chipperfield (1953) also observed 0. edulis at several sites in Great Britain shortly after 999 the occurrence of a spring tide. In our study, spawn- ing began around the time of spring tides, but in- duction of spawning by hydrostatic pressure has not been reported in mussels. Alternatively, spawning may be induced by other factors associated with spring tides, such as increased temperature fluctua- tions, air exposure, and water movement. Temper- ature fluctuations have been shown to induce labor- atory spawning in Mytilus edulis (Bayne 1976). While a temperature threshold is suggested, time of year may also be important as indicated by the spawning periods in Table 2. Of the 10 studies ex- amined, all but one reported the initiation of spawn- ing from May to June. Aside from temperature, the initiation of spawning may be influenced by another cyclic phenomena such as photoperiod. Light and photoperiod in particular have been shown to affect the timing of reproduction in a number of marine invertebrates (Segal 1970). While adult mussels are sensitive to changes in light intensity (Bayne et al. 1976), the ability to detect changing photoperiod has not been demonstrated. The results of this study have been attributed to annual temperature cycles, but until light response of mussels is more fully ex- amined photoperiod cannot be ruled out. Variations in larval abundance from year to year do not appear to be linked to temperature, nor to availability of food energy. Kautsky (1982) reported that Baltic Sea mussel populations were limited to one major spawning by reduced food availability dur- ing the remainder of the year. Similarly, Thompson (1979) attributed annual variation in reproductive condition and fecundity of mussels along the coast of Nova Scotia to annual variations in food supply. Bayne (1975) noted that while poor nutrition does not significantly alter the timing of gametogenesis, it can result in resorption of gametes prior to spawn- ing. Newell et al. (1982) suggested that the cycle of food availability could affect both the nutrient storage cycle and the timing of gametogenic events, including spawning. In every year of our study the spring augmentation of phytoplankton was well under way by March or April, with densities >105 cells 1_1. Significant numbers of mussel larvae were first detected between late May and early June. Thus, it appears that food is not limiting to either adult or larval mussel populations in our area. Our phytoplankton data, however, do not include the smaller naked nanoplankton which, together with particulate organic matter, could account for more than half of the available energy in the Damariscotta River (Incze 1979). This fraction would be a better index of food available to mussel larvae and should be included in studies attempting to link abundance or setting success of larvae to their food supply. Onset of spawning in Damariscotta River mussel populations is predictable from year to year. It oc- curs when water temperature exceeds 10°-12.5°C, and near the spring tide portion of the neap-spring cycle. Food does not appear to be limiting to either gametogenesis or the development of larvae. Acknowledgments We thank H. Hidu for stimulating discussions and for criticizing an earlier draft of the manuscript. E. S. Gardella and A. L. Heinig contributed greatly to the sampling efforts. Greg Podniesinski was sup- ported by UMO-UNH Sea Grant R/FD-99 awarded to H. Hidu. Literature Cited Baird, R. H. 1966. Factors affecting the growth and condition of mussels (Mytilus edulis L.). Fish. Invest., Lond., Ser. II, 25:1-33. Bayne, B. L. 1965. Growth and the delay of metamorphosis of the larvae of Mytilus edulis (L.). Ophelia 2:1-47. 1975. Reproduction in bivalve molluscs under environmen- tal stress. In F. J. Vernberg (editor), Physiological ecology of estuarine organisms, p. 259-277. Univ. South Carolina Press. 1976. The biology of mussel larvae. In B. L. Bayne (editor), Marine mussels: their ecology and physiology, p. 81-120. Cambridge Univ. Press, N.Y. Bayne, B. L., J. Widdows, and R. J. Thompson. 1976. Physiology II. In B. L. Bayne (editor), Marine mussels: their ecology and physiology, p. 207-260. Cam- bridge Univ. Press, N.Y. Bohle, B. 1971. Settlement of mussel larvae Mytilus edulis on sus- pended collectors in Norwegian waters. 4th Eur. Mar. Biol. Symp., p. 63-69. Chanley, P., and J. D. Andrews. 1971. Aids for identification of bivalve larvae of Virginia. Malacologia 11:45-119. Chipperfield, P. N. J. 1953. Observations on the breeding and settlement of Mytilus edulis (L.) in British waters. J. Mar. Biol. Assoc. U.K. 32:449-476. DeSchweinitz, E. H., and R. A. Lutz. 1976. Larval development of the northern horse mussel, Modiolus modiolus (L.), including a comparison with the lar- vae of Mytilus edulis (L.) as an aid in planktonic identifica- tion. Biol. Bull. (Woods Hole) 150:348-360. Engle, J. B., and V. L. Loosanoff. 1944. On season of attachment of larvae of Mytilus edulis. Ecology 25:433-440. Fell, P. E., and A. M. Balsamo. 1985. Recruitment of Mytilus edulis L. in the Thames estuary, with evidence of differences in the time of maximal settling along the Connecticut shore. Estuaries 8:68-75. Field, I. A. 1922. Biology and economic value of the sea mussel Mytilus 1000 edulis. Fish. Bull, U.S. 38:127-260. Frolander, H. F. 1968. Statistical variation in zooplankton numbers from sub- sampling with a Stempel pipette. J. Water Pollut. Control Fed. 40:R82-R88. Hanks, R. W., and D. B. Packer. 1985. A new species of Sphenia (Bivalvia:Myidae) from the Gulf of Maine. Veliger 27:320-330. Incze, L. S. 1979. Relationships between environmental temperatures, seston, and the growth and survival of Mytilus edulis, L. in a north temperate estuary. M.S. Thesis, Univ. Maine, Orono. Jorgenson, C. B. 1981. Mortality, growth and grazing impact of a cohort of bivalve larvae, Mytilus edulis. (L.). Ophelia 20:185-192. Kautsky, N. 1982. Quantitative studies on gonad cycle, fecundity, repro- ductive output and recruitment in a Baltic Mytilus edulis population. Mar. Biol. (Berl.) 58:143-160. Korriga, P. 1947. Relations between the moon and periodicity in the breeding of marine animals. Ecol. Monogr. 17:347-381. Lee, W. Y., and B. J. McAlice. 1979. Sampling variability of marine zooplankton in a tidal estuary. Estuarine Coastal Mar. Sci. 8:565-582. Loosanoff, V. L., H. C. Davis, and P. E. Chanley. 1966. Dimensions and shapes of larvae of some marine bivalve mollusks. Malacologia 4:351-435. Loosanoff, V. L., and C. A. Nomejko. 1951 . Existence of physiologically different races of oysters, Crassostrea virginica. Biol. Bull. (Woods Hole) 101:151- 156. Lutz, R. A., and H. HlDU. 1979. Hinge morphogenesis in the shells of larval and early post-larval mussels (Mytilus edulis and Modiolus modiolus L.). J. Mar. Biol. Assoc. U.K. 59:111-121. McAlice, B. J. 1977. A preliminary oceanographic survey of the Damari- scotta River estuary, Lincoln County, Maine. Maine Sea Grant Tech. Rep. 13, 27 p. Newell, R. I. E., T. J. Hilbish, R. K. Koehn, and C. J. Newell. 1982. Temporal variation in the reproductive cycle of Mytilus edulis L. (Bivalvia, Mytilidae) from localities on the east coast of the United States. Biol. Bull. (Woods Hole) 162: 299-310. NOAA. 1969-1976. Tide Tables: East Coast of North and South America, including Greenland. Tide Tables Natl. Ocean Surv. (U.S.) 1969-1976. Rasmussen, E. 1973. Systematics and ecology of the Isefjord marine fauna (Denmark) with a survey of the eelgrass (Zostera) vegeta- tion and its communities. Ophelia 11:1-495. Sastry, A. N. 1968. The relationships among food, temperature and gonad development of the bay scallop, Aequipecten irradians Lamarck. Physiol. Zool. 41:44-53. Seed, R. 1975. Reproduction in Mytilus (Mollusca-.Bivalvia) in Euro- pean waters. Publ. Stn. Zool. Napoli 39(suppl. l):317-334. Segal, E. 1970. Light, animals, invertebrates. In O. Kinne (editor), Marine ecology, Vol. I, Part I, p. 159-211. Wiley-Inter- science, N.Y. Stauber, L. A. 1950. The problem of physiological species with special reference to oysters and oyster drills. Ecology 31:109-118. Stubbings, H. G. 1954. The biology of the common mussel in relation to foul- ing problems. Research 7:222-229. Theisen, B. F. 1973. The growth of Mytilus edulis L. (Bivalvia) from Disko and Th:ile district, Greenland. Ophelia 12:59-77. Thompson, R. J. 1979. Fecundity and reproductive effort in the blue mussel (Mytilus edulis), the sea urchin (Stronglyocentrotus droe- bachiensis) and the snow crab (Chionocetes opilio) from populations in Nova Scotia and Newfoundland. J. Fish. Res. Board Can. 36:955-964. Zar, J. H. 1984. Biostatistical analysis. 2d ed. Prentice-Hall, Engle- wood Cliffs, N.J. Greg S. Podniesinski Department of Zoology University of Maine, Orono, ME Mailing address: Ira C. Darling Center University of Maine Walpole, ME 04573 Bernard J. McAlice Department of Botany and Plant Pathology University of Maine Orono, ME 04573 1001 INDEX Fishery Bulletin Vol. 84, No. 1-4 "The abundance and distribution of the family Macrouridae (Pisces: Gadiformes) in the Norfolk Canyon area," by Robert W. Middleton and John A. Musick 35 "Abundance, size, and sex ratio of adult sea-run sea lam- preys, Petromyzon marinus, in the Connecticut River," by Kathleen Stier and Boyd Kynard 476 "Age and growth of the marine catfish, Netuma barba (Siluriformes, Ariidae), in the estuary of the Patos Lagoon (Brasil)," by Enir Girondi Reis 679 "Age dependent fecundity, number of spawnings per year, sex ratio, and maturation stages in northern anchovy, Engraulis mordax," by Richard H. Parrish, Donna L. Mallicoate, and Richard A. Klingbeil 503 Aging studies larval fish 91 sailfish, Atlantic 493 AHRENHOLZ, DEAN W.-see NELSON and AHRENHOLZ Albacore chromosomal analysis 469 Ammodytes americanus—see Eel, sand Ammodytes hexapterus—see Sand lance Ampelisca agassizi—see Amphipods, gammaridean Amphipods, benthic parasites of 204, 605 "Anatomical trauma to sponge-coral reef fishes captured by trawling and angling," by S. Gordon Rogers, Hiram T. Langston, and Timothy E. Targett 697 Anchovy, northern fecundity and spawning 503 drift in the California Current 585 life-stage-specific instantaneous mortality rates 395 spawning in San Francisco Bay 879 vulnerability to predation 859 "An approach to yield assessment for unexploited resources with application to the deep slope fishes of the Marianas," by Jeffrey J. Polovina and Stephen Ralston . . 759 Anglerfish, lophiid early development 429 Angling trauma to sponge-coral reef fishes 697 Anguilla rostrata—see Eel, American "Annual production of eviscerated body weight, fat, and gonads of Pacific herring, Clupea harengus pallasi, near Auke Bay, southeastern Alaska," by Jay C. Quast .... 705 Anoplopoma fimbria— see Sablefish Anthozoans ecology of Ceriantharia from Cape Hatteras to Nova Scotia 625 Arctica islandica—see Quahog, ocean "Arrival of northern fur seals, Callorhinus ursinus, on St. Paul Island, Alaska," by Michael A. Bigg 383 Ascarophis sp. parasites in American lobster 197 "Aspects of the reproductive biology, spatial distribution, growth, and mortality of the deepwater caridean shrimp, Heterocarpus laevigatus, in Hawaii," by Murray D. Dailey and Stephen Ralston 915 Atheresthes evermanni—see Flounder, Kamchatka Atheresthes stomias—see Flounder, arrowtooth ATKINSON, C. ALLEN, "Discrete-time difference model for simulating interacting fish population dynamics" . . . 535 Backdown dolphin-releasing procedure 27 Balaenoptera edeni—see Whale, Bryde Balaenoptera physalus—see Whale, finback BALTZ, DONALD M.-see MOYLE et al. BARBER, RICHARD T.-see FORWARD et al. BARLOW, J.-see MYRICK et al. -see REILLY and BARLOW Bass, black sea anatomical trauma from angling 697 contributions to life history 723 Bass, striped feeding below a hydroelectric dam 220 survival and growth 905 BAYER, RANGE D., "Seabirds near an Oregon estuarine salmon hatchery in 1982 and during the 1983 El Nino" . . 279 BEACHAM, TERRY D., "Type, quantity, and size of food of Pacific salmon (Oncorhynchus) in the Strait of Juan de Fuca, British Columbia" 77 1003 BIANCHINI, MARCO L.-see SORENSEN et al. BIGG, MICHAEL A., "Arrival of northern fur seals, Callorhinus ursinus, on St. Paul Island, Alaska" 383 BIGG, MICHAEL A.-see PEREZ and BIGG BODKIN, JAMES LEE, "Fish assemblages in Macrocystis and Nereocystis kelp forests off central California" .... 799 BOEHM, PAUL D.-see STEIMLE et al. BONNELL, MICHAEL L.-see DOHL et al. Boston Ex-vessel price in New England fishing industry .... 437 BOTSFORD, LOUIS W.-see SYKES and BOTSFORD Bottomfish affect of hypoxia on abundance and distribution 19 Brevoortia patronus—see Menhaden, gulf Brevoortia tyrannus—see Menhaden, Atlantic Briarosaccus callosus—see Rhizocephalan BRUNO, RALPH A.-see STEIMLE et al. BRUSHER, HAROLD A.-see FINUCANE et al. BURCH, RAYMOND, K.-see UCHIYAMA et al. Calanus pacificus abundance, chemical composition, distribution, and size 157 California Current 157 California Cooperative Oceanic Fisheries Investigation northern anchovy drift studies 587 California Current Calanus pacificus 157 northern anchovy 585 Callorhinus ursinus— see Seal, northern fur CAREY, ANDREW G, JR.-see TESTER and CAREY "Cartilage and bone development in scombroid fishes," by Thomas Potthoff, Sharon Kelley, and Joaquin C. Javech . . 647 Catfish, sea age and growth in Patos Lagoon, Brasil 679 Centrapristis striata— see Bass, black sea Ceriantharia— see Anthozoans "Cetacean high-use habitats of the northeast United States continental shelf," by Robert D. Kenney and Howard E. Winn 345 Cetaceans high-use habitats 345 1004 sighted by CETAP aerial and POP surveys 349 CETAP aerial surveys 345 "Chinook salmon, Oncorhynchus tshawytscha, spawn- ing escapement based on multiple mark-recapture of carcasses," by Stephen D. Sykes and Louis W. Bots- ford 261 Chionoecetes tanneri—see Crab, spider Chlorophyll sea surface concentration in the tropical Pacific 687 "Chromosomal analysis of albacore, Thunnus alalunga, and yellowfin, Thunnus albacares, and skipjack, Kat- suwonus pelamis, tuna," by F. J. Ratty, Y. C. Song, and R. M. Laurs 469 Ciliates parasites of benthic amphipods 204 Clam, soft shell description of poecilostomatoid copepods 227 occurrence of epizootic sarcoma in Chesapeake Bay . . 851 Clupea harengus pallasi—see Herring, Pacific Cod ex-vessel price in New England fishing industry .... 437 CODY, TERRY J., and BILLY E. FULS, "Comparison of catches in 4.3 m and 12.2 m shrimp trawls in the Gulf of Mexico" 981 Coelorinchus c. carminatus abundance and distribution in Norfolk Canyon 37 COLLETTE, BRUCE B., "Resilience of the fish assem- blage in New England tidepools" 200 COLLINS, L. ALAN-see FINUCANE et al. "Community studies in seagrass meadows: A comparison of two methods for sampling macroinvertebrates and fishes," by Kenneth M. Leber and Holly S. Greening . . 443 "Comparison of catches in 4.3 m and 12.2 m shrimp trawls in the Gulf of Mexico," by Terry J. Cody and Billy E. Fuls 981 "Comparison of visceral fat and gonadal fat volumes of yellowtail rockfish, Sebastesflavidus, during a normal year and a year of El Nino conditions," by William H. Lenarz and Tina Wyllie Echeverria 743 Connecticut River spawning migration of sea lampreys 749 "Contributions to the life history of black sea bass, Cen- tropristis striata, off the southeastern United States," by Charles A. Wenner, William A. Roumillat, and C. Wayne Waltz 723 COOPER, RICHARD A.-see SHEPARD et al. "Copepodids and adults of Leptinogaster major (Williams, 1907), a poecilostomatoid copepod living in Mya arenaria L. and other marine bivalve mollusks," by Arthur G. Humes 227 Copepods size and chemical composition 165 Copepods, poecilostomatoid Leptinogaster major living in mollusks 227 taxonomic history 227 Coryphaena equiselis—see Dolphin, pompano Coryphaena hippurus—see Dolphin (fish) Coryphaenoides armatus abundance and distribution in Norfolk Canyon 51 Coryphaenoides carapinus abundance and distribution in Norfolk Canyon 51 Coryphaenoides rupestris abundance and distribution in Norfolk Canyon 48 COX, JAMES L.-see WILLASON et al. Crab, blue king Rhizocephalan infection 177 Crab, golden king distribution and reproductive biology in eastern Bering Sea 571 Crab, king comparison of blue and golden king crabs 327 Crab, spider instar identification 973 life history 973 Culture studies squid, market 771 CUMMINGS, WILLIAM C, PAUL 0. THOMPSON, and SAMUEL J. HA, "Sounds from Bryde, Balaenoptera edeni, and finback, B. physalus, whales in the Gulf of California" 359 DAILEY, MURRAY D, and STEPHEN RALSTON, "Aspects of the reproductive biology, spatial distribution, growth, and mortality of the deepwater caridean shrimp, Heterocarpus laevigatus, in Hawaii" 915 Dams, hydroelectric striped bass feeding area 220 DANDONNEAU, YVES, "Monitoring the sea surface chlorophyll concentration in the tropical Pacific: conse- quences of the 1982-83 El Nino" 687 DANIELS, ROBERT A.-see MOYLE et al. DEAN, JOHN M.-see PRINCE et al. Delphinus delphis—see Dolphin, common de MENDIOLA, BLANCA ROJAS-see FORWARD et al. "Determining age of larval fish with the otolith increment technique," by Cynthia Jones 91 "Development and evaluation of methodologies for assess- ing and monitoring the abundance of widow rockfish, Sebastes entomelas," by Mark E. Wilkins 287 Developmental studies scombroid fishes 647 "Diel foraging activity of American eels, Anguilla rostrata (LeSueur), in a Rhode Island estuary," by Peter W. Sorensen, Marco L. Bianchini, and Howard E. Winn . . 746 "Diet of northern fur seals, Callorhinus ursinus, off western North America," by Michael A. Perez and Michael A. Bigg 957 "Differentiation of Prionotus carolinus and Prionotus evolans eggs in Hereford Inlet estuary, southern New Jersey, using immunodiffusion," by Walter J. Keirans, Sidney S. Herman, and R. G. Malsberger 63 Dinoflagellates swimming speed of Gymnodinium splendens 461 parasites of benthic amphipods 605 "Discrete-time difference model for simulating interacting fish population dynamics," by C. Allen Atkinson 535 Disease epizootic sarcoma in soft shell clams 851 Dissolved oxygen concentration effect on shrimp and bottomfish in Louisiana coastal waters 19 "Distribution and abundance of common dolphin, Del- phinus delphis, in the Southern California Bight: a quan- titative assessment based upon aerial transect data," by Thomas P. Dohl, Michael L. Bonnell, and R. Glenn Ford . . 333 "Distribution and reproductive biology of the golden king crab, Lithodes aequispina, in the eastern Bering Sea," by David A. Somerton and Robert S. Otto 571 "The distribution of the humpback whale, Megaptera novaeangliae, on Georges Bank and in the Gulf of Maine in relation to densities of the sand eel, Ammodytes americanus," by P. Michael Payne, John R. Nicolas, Loretta O'Brien, and Kevin D. Powers 271 DITTY, JAMES G., "Ichthyoplankton in neritic waters of the northern Gulf of Mexico off Louisiana: composition, relative abundance, and seasonality" 935 DOHL, THOMAS P., MICHAEL L. BONNELL, and R. GLENN FORD, "Distribution and abundance of common dolphin, Delphinus delphis, in the Southern California Bight: a quantitative assessment based upon aerial transect data" 333 Dolphin, common distribution and abundance in southern California . . . 333 1005 Dolphin (fish) growth in Hawaiian waters by daily increments 186 stock structure in western central Atlantic 451 Dolphin, spotted reproductive biology in eastern tropical Pacific 247 Dolphins increase in population size 527 mortality due to tuna purse seine fishery 27 in eastern tropical Pacific tuna fishery 559 "Early development of the Lophiid anglerfish, Lophius gastrophysus," by Yasunobu Matsuura and Nelson Takumi Yoneda 429 EARLY, GREG-see SELZER et al. "Early life history of Atlantic menhaden, Brevoortia tyran- nus, and gulf menhaden, B. patronus" by Allyn B. Powell and Germano Phonlor 991 ECHEVERRIA, TINA WYLLIE-see LENARZ and ECHEVERRIA Echo integration assessing widow rockfish abundance 287 "An ecological survey and comparison of bottom fish resource assessments (submersible versus handline fishing) at Johnston Atoll," by Stephen Ralston, Reginald M. Gooding, and Gerald M. Ludwig 141 Ecology anthozoans 625 bottom fish resource assessment at Johnston Atoll . . 141 community studies in seagrass meadows 443 "Ecology of Ceriantharia (Coelenterata, Anthozoa) of the northwest Atlantic from Cape Hatteras to Nova Scotia," by Andrew N. Shepard, Roger B. Theroux, Richard A. Cooper, and Joseph R. Uzmann 625 Economic studies spiny lobster 69, 74 Ecosystems fish population dynamic simulations 535 Eel, American diel foraging activity 746 Eel, sand relationship to humpback whale 271 "Effects of exposure and confinement on spiny lobsters, Panulirus argus, used as attractants in the Florida trap fishery," by John H. Hunt, William G. Lyons, and Frank S. Kennedy, Jr 69 "Effects of temperature on swimming speed of the dinoflagellate, Gymnodinium splendens," by Richard B. Forward, Jr., Blanca Rojas de Mendiola, and Richard T. Barber 460 1006 El Nino chlorophyll concentration in tropical Pacific 687 correlations of seabirds and salmon smolts 279 fat volume of yellowtail rockfish 743 Engraulis mordax—see Anchovy, northern Euphausia pacifica 161 Euphausiids distribution and abundance in California Current .... 161 invertebrate prey of Pacific salmon 77 "Ex-vessel price linkages in the New England fishing in- dustry," by Dale Squires 437 FARLEY, C. A., S. V. OTTO, and C. L. REINISCH, "New occurrence of epizootic sarcoma in Chesapeake Bay soft shell clams, Mya armaria 851 FAVUZZI, JC N-see WILLASON et al. "Fecundity of northern shrimp, Pandalus borealis, (Crustacea, Decapoda) in areas of the Northwest Atlantic," by D. G. Parsons and G. E. Tucker 549 "Fecundity of the Pacific hake, Merluccius productus, spawning in Canadian waters," by J. C. Mason 209 FINUCANE, JOHN H., L. ALAN COLLINS, HAROLD A. BRUSHER, and CARL H. SALOMAN, "Reproductive biology of king mackerel, Scomberomorus cavalla, from the southeastern United States" 841 FIORELLI, PATRICIA M.-see SELZER et al. "First record of the longfin mako, Isurus paucus, in the Gulf of Mexico," by Kristie Killam and Glenn Parsons. . 748 Fish. assemblages in kelp forests 799 resilience in New England tidepools 200 "Fish assemblages in Macrocystis and Nereocystis kelp forests off central California," by James Lee Bodkin . . 799 Fish population studies food consumption estimates 827 simulating population dynamics 535 Fishery crab, golden king 571 hypoxia in Louisiana coastal waters 19 management 697 menhaden, gulf 311 rockfish, commercial 409 spiny lobster, Florida 69 tuna purse seine and dolphin mortality 27 tuna, yellowfin 247, 559 Fishes community studies in seagrass meadows 443 distribution and abundance in Suisan Marsh 105 Fishes, reef trauma from trawling and angling 697 Fishing ex-vessel price in New England fishing industries . . . 437 multispecies intensive fishing experiment 423 shrimp caridean 927 submersible versus handline 141 Fistulicola plicatus—see Tapeworm Flounder ex-vessel price linkages in New England 437 Flounder, arrowtooth food habits in eastern Bering Sea 615 genetic confirmation of specific distinction 222 Flounder, Kamchatka food habits in eastern Bering Sea 615 genetic confirmation 222 Flounder, yellowtail statistical methods for estimating abundance 519 FOLKVORD, ARILD, and JOHN R. HUNTER, "Size- specific vulnerability of northern anchovy, Engraulis mor- dax, larvae to predation by fishes" 859 Food habits bass, striped 220 fish consumption estimates 615 flounder 615, 827 hake, Pacific 947 salmon, Pacific 77 seal, northern fur 957 "Food habits and diet overlap of two congeneric species, Atheresthes stomias and Atheresthes evermanni, in the eastern Bering Sea," by M. S. Yang and P. A. Livingston. . 615 FORD, R. GLENN-see DOHL et al. FORWARD, RICHARD B., JR., BLANCA de MENDIOLA, and RICHARD T. BARBER, "Effects of temperature on swimming speed of the dinoflagellate, Gymnodinium splen- dens" 460 FROST, KATHRYN J., and LLOYD F. LOWRY, "Sizes of walleye pollock, Theragra chalcogramma, consumed by marine mammals in the Bering Sea" 192 FULS, BILLY E.-see CODY and FULS Gammaridean amphipods 204 "Genetic confirmation of specific distinction of arrowtooth flounder, Atheresthes stomias, and Kamchatka flounder, A. evermanni" by Carol L. Ranck, Fred M. Utter, George B. Milner, and Gary B. Smith 222 Genetic studies chromosomal analysis of tuna 469 specific distinction of flounder 222 Georges Bank distribution of humpback whales 271 GIBSON, DARCY L.-see GRAHAM et al. Gloucester ex-vessel price in New England fishing industry .... 437 GOLDBERG, STEPHEN R.-see WEBER and GOLDBERG GOODING, REGINALD M.-see RALSTON et al. GRAHAM, JEFFREY B., RICHARD H. ROSENBLATT, and DARCY L. GIBSON, "Morphology and possible swim- ming mode of a yellowfin tuna, Thunnus albacares, lack- ing one pectoral fin" 463 GREENING, HOLLY S.-see LEBER and GREENING Groupers unexploited resources in the Marianas 759 GROVER, JILL J., and BORI L. OLLA, "Morphological evidence for starvation and prey size selection of sea- caught larval sablefish, Anoplopoma fimbria" 484 "Growth, behavior, and sexual maturation of the market squid, Loligo opalescens, cultured through the life cycle," by W. T. Yang, R. F. Hixon, P. E. Turk, M. E. Krejci, W. H. Hulet, and R. T. Hanlon 771 "Growth of dolphins, Coryphaena hippurus, and C. equiselis, in Hawaiian waters as determined by daily in- crements on otoliths," by James H. Uchiyama, Raymond K. Burch, and Syd A. Kraul, Jr 186 Growth rates anchovy, northern 503 bass, black sea 723 bass, striped 905 catfish, sea 679 dolphin fishes in Hawaiian waters 186 herring, Pacific 705 shrimp, caridean 915 squid, market 771 Gulf of California sounds of Bryde and finback whales 359 Gulf of Maine distribution of humpback whales 271 Gulf of Mexico first record of longfin mako 748 ichthyoplankton 935 Gymnodinium splendens—see Dinoflagellate HA, SAMUEL J. -see CUMMINGS et al. Habitat studies cetaceans of the northeast United States . Haddock ex-vessel price in New England fishing industry 345 437 1007 Hake, Pacific fecundity in Canadian waters 209 stomach contents and food consumption 947 HANLON, R. T.-see YANG et al. HARRIS, R. E., JR.-see VAN ENGEL et al. Hatcheries growth and survival of striped bass 905 HAWKES, CLAYTON R, THEODORE R. MEYERS, and THOMAS C. SHIRLEY, "Length-weight relationships of blue, Paralithodes platypus, and golden, Lithodes aequispina, king crabs parasitized by the rhizocephalan, Briarosaccus callosus Boschma" 327 HERBOLD, BRUCE-see MOYLE et al. HERMAN, SIDNEY S.-see KEIRANS et al. Herring, Pacific annual production 705 Heterocarpus laevigatas— see Shrimp, caridean Heterostichus rostratus—see Kelpfish, giant Histology starvation induced mortality 1 HIXON, R. F.-see YANG et al. HOGANS, W. E., and P. C. F. HURLEY, "Variations in the morphology of Fistulicola plicatus Rudolphi (1802) (Cestoda: Pseudophyllidea) from the swordfish, Xiphias gladius L., in the northwest Atlantic Ocean" 754 HOHN, A. A. -see MYRICK et al. Homarus americanus—see Lobster, American HOUDE, EDWARD D, and LAWRENCE LUBBERS III, "Survival and growth of striped bass, Morone saxatilis, and Morone hybrid larvae: laboratory and pond enclosure experiments 905 HULET, W H.-see YANG et al. HUMES, ARTHUR G., "Copepodids and adults of Lep- tinogaster major (Williams, 1907), a poecilostomatoid copepod living in Mya arenaria L. and other marine bivalve mollusks" 227 HUNT, JOHN H., WILLIAM G. LYONS, and FRANK S. KENNEDY, JR., "Effects of exposure and confinement on spiny lobsters, Panulirus argus, used as attractants in the Florida trap fishery" 69 HUNTE, WAYNE-see OXENFORD and HUNTE HUNTER, J. ROE, BEVERLY J. MACEWICZ, and JOHN R. SIBERT, "The spawning frequency of skipjack tuna, Katsuwonus pelamis, from the South Pacific" 895 HUNTER, JOHN R.-see FOLKVORD and HUNTER 1008 HURLEY, P. C. F.-see HOGANS and HURLEY Hybrids, bass striped survival and growth 905 "Hypoxia in Louisiana coastal waters during 1983: implica- tions for fisheries," by Maurice L. Renaud 19 "Ichthyoplankton in neritic waters of the northern Gulf of Mexico off Louisiana: composition, relative abundance, and seasonality," by James G. Ditty 935 Immunodiffusion differentiation of Prionotus eggs 63 "An improved otter surface sampler," by J. C. Mason and A. C. Phillips 480 "Incidental dolphin mortality in the eastern tropical Pacific tuna fishery, 1973 through 1978," by Bruce E. Wahlen . . 559 "Incidental mortality of dolphins in the eastern tropical Pacific, 1959-72," by N. C. H. Lo and T. D. Smith .... 27 "Increased food and energy consumption of lactating northern fur seals, Callorhinus ursinus," by Michael A. Perez and Elizabeth E. Mooney 371 "Instar identification and life history aspects of juvenile deepwater spider crabs, Chionoecetes tanneri Rathbun," by Patricia A. Tester and Andrew G. Carey, Jr 973 "An intensive fishing experiment for the caridean shrimp, Heterocarpus laevigatas, at Alamagan Island in the Mariana Archipelago," by Stephen Ralston 927 Istiophorus platypterus—see Sailfish, Atlantic Isurus paucus—see Shark, longfin mako JAVECH, JOAQUIN C.-see POTTHOFF et al. JOHNSON, P. T, R. A. MacINTOSH, and D. A. SOMER- TON, "Rhizocephalan infection in blue king crabs, Paralithodes platypus, from Olga Bay, Kodiak Island, Alaska" 177 JOHNSON, PHYLLIS T., "Parasites of benthic amphi- pods: ciliates" 204 JOHNSON, PHYLLIS T., "Parasites of benthic amphi- pods: dinoflagellates (Duboscquodinida: Syndinidae) . . . 605 Johnston Atoll resource assessment 141 Jolly-Seber spawning escapement of chinook salmon 261 JONES, CYNTHIA, "Determining age of larval fish with the otolith increment technique" 91 Katsuwonus pelamis— see Tuna, skipjack KEIRANS, WALTER J., SIDNEY S. HERMAN, and R. G. MALSBERGER, "Differentiation of Prionotus carolinus and Prionotus evolans eggs in Hereford Inlet estuary, southern New Jersey, using immunodiffusion" . . 63 KELLEY, SHARON-see POTTHOFF et al. Kelp, bull fish assemblages in kelp forests 799 Kelp forests fish assemblages 799 Kelp, giant fish assemblages in kelp forests 799 Kelpfish, giant life history and larval development 809 KENNEDY, FRANK S., JR.-see HUNT et al. KENNEY, ROBERT D, and HOWARD E. WINN, "Ceta- cean high-use habitats of the northeast United States con- tinental shelf" ■ 345 KILLAM, KRISTIE, and GLENN PARSONS, "First record of the longfin mako, Isurus paucus, in the Gulf of Mexico" 748 KLINGBEIL, RICHARD A.-see PARRISH et al. KRAUL, SYD A., JR.-see UCHIYAMA et al. KREJCI, M. E.-see YANG et al. KRYGIER, E. E, and W G PEARCY, "The role of estuarine and offshore nursery areas for young English sole, Parophrys vetulus Girard, of Oregon" 119 KYNARD, BOYD-see STIER and KYNARD -see WARNER and KYNARD Lampreys, sea abundance, size, and sex ratio 476 movement in the Connecticut River 749 LANGSTON, HIRAM T.-see ROGERS et al. Larvae anchovy, northern in the California Current 585 instantaneous mortality rates 395 spawning and predation in San Francisco Bay 879 vulnerability of 859 anglerfish, lophiid early development and comparison with other lophiid species 429 bass, striped survival and growth 905 fish age determination 91 kelpfish, giant 809 mackerel, jack 1 mussels, blue spawning and seasonality 995 sablefish starvation and prey size selection 484 scombroid cartilage and bone development 647 LAURS, R. M.-see RATTY et al. LEBER, KENNETH M., and HOLLY S. GREENING, "Community studies in seagrass meadows: A comparison of two methods for sampling macroinvertebrates and fishes" 443 LEE, DENNIS W.-see PRINCE et al. LENARZ, WILLIAM H., and TINA WYLLIE ECHE- VERRIA, "Comparison of visceral fat and gonadal fat volumes of yellowtail rockfish, Sebastesflavidus, during a normal year and a year of El Nino conditions" 743 "Length-weight relationships of blue, Paralithodes platy- pus, and golden, Lithodes aequispina, king crabs para- sitized by the rhizocephalan, Briarosaccus callosus Boschma," by Clayton R. Hawkes, Theodore R. Meyers, and Thomas C. Shirley 327 Leptinogaster major— see Copepod, poecilostomatoid Leslie model dolphin population studies 527 fish population studies 535 variable catchability version 423 "Life history and larval development of the giant kelpfish, Heterostichus rostratus Girard, 1854," by Carol A. Stepien 809 Life history studies bass, black sea 723 crab, spider 973 kelpfish, giant 809 menhaden, Atlantic and gulf 991 Limanda ferruginea—see Flounder, yellowtail Line intercept survey assessing widow rockfish abundance 287 Line transect survey assessing widow rockfish abundance 287 Literature review daily deposition of otolith increments 91 Lithodes aequispina— see Crab, golden king LIVINGSTON, P. A.-see YANG and LIVINGSTON LO, NANCY C. H., "Modeling life-stage-specific instan- taneous mortality rates, an application to northern anchovy, Engraulis mordax, eggs and larvae" 395 LO, N. C. H, and T D SMITH, "Incidental mortality of dolphins in the eastern tropical Pacific, 1959-72" 27 Lobster, American occurrence of parasites 197 1009 Lobster, spiny in Florida trap fishery 69 Loligo opalescens—see Squid, market "Longevity and age validation of a tag-recaptured Atlan- tic sailfish, Istiophoms platypterus, using dorsal spines and otoliths," by Eric D. Prince, Dennis W. Lee, Charles A. Wilson, and John M. Dean 493 Lophius gastrophysus—see Anglerfish, lophiid Louisiana hypoxia in coastal waters 19 LOWRY, LLOYD F.-see FROST and LOWRY LUBBERS, LAWRENCE Ill-see HOUDE and LUBBERS LUDWIG, GERALD M.-see RALSTON et al. LYONS, WILLIAM G.-see HUNT et al. MACEWICZ, BEVERLY J.-see HUNTER et al. MacINTOSH, R. A.-see JOHNSON et al. Mackerel, jack histological analysis 2 morphological analysis 2 offshore starvation 1 reproductive biology 841 starvation-induced mortality 1 Macrocytis pyrifera—see Kelp, giant Macroinvertebrates community studies in seagrass meadows 443 Macrouridae abundance and distribution in Norfolk Canyon 35 Makalii— see Submersible Mako, longfin— see Shark, longfin mako MALLICOATE, DONNA L.-see PARRISH et al. MALSBERGER, R. G.-see KEIRANS et al. Manly-Parr model for spawning escapement of chinook salmon . . 261 Mariana Archipelago fishing experiment for caridean shrimp 927 Marine mammals sizes of walleye pollock consumed 192 MASON, J. C, "Fecundity of the Pacific hake, Merluccius productus, spawning in Canadian waters" 209 MASON, J. C, and A. C. PHILLIPS, "An improved otter surface sampler" 480 1010 MATSUURA, YASUNOBU, and NELSON TAKUMI YONEDA, "Early development of the lophiid anglerfish, Lophius gastrophysus" 429 McALICE, BERNARD, J.-see PODNIESINSKI and McALICE McGOWAN, MICHAEL F., "Northern anchovy, Engraulis mordax, spawning in San Francisco Bay, California, 1978-79, relative to hydrography and zooplankton prey of adults and larvae" 879 Megaptera novaeangliae—see Whale, humpback Menhaden, Atlantic early life history 991 Menhaden, gulf early life history 991 population and fishery characteristics 311 Merluccius productus— see Hake, Pacific MERRINER, JOHN V.-see SMITH, JOSEPH W "Methodological problems in sampling commercial rockfish landings," by A. R. Sen 409 MEYERS, THEODORE R.-see HAWKES et al. Micropogonius undulatus—see Croaker, Atlantic MIDDLETON, ROBERT W., and JOHN A. MUSICK, "The abundance and distribution of the family Macrouridae (Pisces: Gadiformes) in the Norfolk Canyon area" 35 MILNER, GEORGE B.-see RANCK et al. "A model of the drift of northern anchovy, Engraulis mor- dax, larvae in the California Current," by James H. Power. . 585 "Modeling life-stage-specific instantaneous mortality rates, an application to northern anchovy, Engraulis mordax, eggs and larvae," by Nancy C. H. Lo 395 Models estimating food consumption of fish populations 827 Leslie 423 life-stage-specific instantaneous mortality rates 395 Manly-Parr 261 "Monitoring the sea surface chlorophyll concentration in the tropical Pacific: consequences of the 1982-83 El Nino," by Yves Dandonneau 687 MOONEY, ELIZABETH E.-see PEREZ and MOONEY Morone saxatilis—see Bass, striped "Morphological evidence for starvation and prey size selec- tion of sea-caught larval sablefish, Anoplopoma fimbria," by Jill J. Grover and Bori L. Olla 484 "Morphology and possible swimming mode of a yellowfin tuna, Thunnus albacares, lacking one pectoral fin," by Jeffrey B. Graham, Richard H. Rosenblatt, and Darcy L. Gibson 463 "Movement of sea-run sea lampreys, Petromyzon marinus, during the spawning migration in the Connecticut River," by Kathleen Stier and Boyd Kynard 749 MOYLE, PETER B., ROBERT A. DANIELS, BRUCE HERBOLD, and DONALD M. BALTZ, "Patterns in distribution and abundance of a noncoevolved assemblage of estuarine fishes in California" 105 MUSICK, JOHN A.-see MIDDLETON and MUSICK Mussel, blue larvae in Damariscotta River estuary, Maine 995 Mya arenaria—see Clam, soft shell MYRICK, A. C, JR., A. A. HOHN, J. BARLOW, and P. A. SLOAN, "Reproductive biology of female spotted dolphins, Stenella attenuata, from the eastern tropical Pacific" . . 247 Mytilus edulis—see Mussel, blue NELSON, WALTER R, and DEAN W AHRENHOLZ, "Population and fishery characteristics of gulf menhaden, Brevoortia patronus" 311 Nematoscelis difficilis 157 Nereocystis leutkeana—see Kelp, bull Netuma barba—see Catfish, sea Neuston sampler an improved otter surface sampler 480 New Bedford ex-vessel price in New England fishing industry .... 437 New England ex-vessel price in New England fishing industry .... 437 "New occurrence of epizootic sarcoma in Chesapeake Bay soft shell clams, Mya arenaria," by C. A. Farley, S. V. Otto, and C. L. Reinisch 851 Nezumia aequalis abundance and distribution in Norfolk Canyon 35 Nezumia bairdii abundance and distribution in Norfolk Canyon 35 NICHOLAS, JOHN R.-see PAYNE et al. "Northern anchovy, Engraulis mordax, spawning in San Francisco Bay, California, 1978-79, relative to hydrography and zooplankton prey of adults and larvae," by Michael F. McGowan 879 Nursery studies estuary studies 119 "Observations on the reproductive biology of the cownose ray, Rhinoptera bonasus in Chesapeake Bay," by Joseph W. Smith and John V. Merriner 871 O'BRIEN, LORETTA-see PAYNE et al. "Occurrence of some parasites and a commensal in the American lobster, Homarus americanus, from the Mid- Atlantic Bight," by W. A. Van Engel, R. E. Harris, Jr., and D. E. Zwerner 197 OLLA, BORI L.-see GROVER and OLLA Oncorhynchus gorbuscha—see Salmon, pink Oncorhynchus kisutch—see Salmon, coho Oncorhynchus nerka—see Salmon, sockeye Oncorhynchus tshawytscha—see Salmon, chinook Oregon nursery areas for English sole 119 "Organic and trace metal levels in ocean quahog, Arctica islandica Linne, from the northwestern Atlantic," by Frank W. Steimle, Paul D. Boehm, Vincent S. Zdanowicz, and Ralph A. Bruno 133 Otoliths age validation in Atlantic sailfish 493 catfish, sea 679 disposition rates 91 dolphin fishes in Hawaiian waters 186 increment technique for aging larval fishes 493 Otter surface sampler 480 OTTO, ROBERT S.-see SOMERTON and OTTO OTTO, S. V.-see FARLEY et al. OXENFORD, HAZEL A., and WAYNE HUNTE, "A pre- liminary investigation of the stock structure of the dolphin, Coryphaena hippurus, in the western central Atlantic" . . 451 Pacific eastern tropical reproductive biology of the spotted dolphin 247 Pandalus borealis—see Shrimp, northern Panulirus argus—see Lobster, spiny Paralithodes platypus— see Crab, king Parasite studies American lobster 197 benthic amphipods infected with dinoflagellates 605 rhizocephalan infection in king crab 327 swordfish tapeworm 754 "Parasites of benthic amphipods: ciliates," by Phyllis T. Johnson 204 "Parasites of benthic amphipods: dinoflagellates (Dubosc- quodinida: Syndinidae)," by Phyllis T. Johnson 605 Parathemisto (hyperiid amphipod) food items of Pacific salmon 77 1011 Parophrys vetulus—see Sole, English PARRISH, RICHARD H., DONNA L. MALLICOATE, and RICHARD A. KLINGBEIL, "Age dependent fecundity, number of spawnings per year, sex ratio, and maturation stages in northern anchovy, Engraulis mordax" 503 PARSONS, D. G., and G. E. TUCKER, "Fecundity of north- ern shrimp, Pandalus borealis, (Crustacea, Decapoda) in areas of the Northwest Atlantic" 549 PARSONS, GLENN-see KILLAM and PARSONS "Patchiness and nutritional condition of zooplankton in the California Current," by Stewart W. Willason, John Favuzzi, and James L. Cox 157 "Patterns in distribution and abundance of a noncoevolved assemblage of estuarine fishes in California," by Peter B. Moyle, Robert A. Daniels, Bruce Herbold, and Donald M. Baltz 105 PAULY, DANIEL, "A simple method for estimating the food consumption of fish populations from growth data and food conversion experiments" 827 PAYNE, P. MICHAEL, JOHN R. NICOLAS, LORETTA O'BRIEN, and KEVIN D. POWERS, "The distribution of the humpback whale, Megaptera novaeangliae, on Georges Bank and in the Gulf of Maine in relation to densities of the sand eel, Ammodytes americanus" 271 PAYNE, P. MICHAEL-see SELZER et al. PCB's organic and trace metals in ocean quahog 133 PEARCY, W. G.-see KRYGIER and PEARCY Penaeus aztecus—see Shrimp, brown Penaeus setiferus—see Shrimp, white PENNINGTON, MICHAEL, "Some statistical techniques for estimating abundance indices from trawl surveys" . . 519 PEREZ, MICHAEL A., and MICHAEL A. BIGG, "Diet of northern fur seals, Callorhinus ursinus, off western North America" 957 PEREZ, MICHAEL A., and ELIZABETH E. MOONEY, "Increased food and energy consumption of lactating northern fur seals, Callorhinus ursinus" 371 Petromyzon marinus—see Lampreys, sea PHILLIPS, A. C.-see MASON and PHILLIPS Phoca vitulina concolor—see Seal, harbor PHONLOR, GERMANO-see POWELL and PHONLOR Phytoplankton zooplankton biomass and nutritional parameters 157 PIKITCH, ELLEN K.-see REXSTAD and PIKITCH 1012 Plankton identification of Prionotus 63 PODNIESINSKI, GREG S., and BERNARD J. McALICE, "Seasonality of blue mussel, Mytilus edulis L., larvae in the Damariscotta River estuary, Maine, 1969-77" 995 Pollock ex-vessel price in New England fishing industry .... 437 Pollock, walleye sizes consumed by marine mammals in the Bering Sea 192 Pollution benthic animals as indicator species 133 POLOVINA, JEFFREY J., "A variable catchability ver- sion of the Leslie model with application to an intensive fishing experiment on a multispecies stock" 423 POLOVINA, JEFFREY J., and STEPHEN RALSTON, "An approach to yield assessment for unexploited resources with application to the deep slope fishes of the Marianas" 759 POP surveys cetacean high-use habitats 345 "Population and fishery characteristics of gulf menhaden, Brevoortia patronus," by Walter R. Nelson and Dean W. Ahrenholz 311 Population sampling devices shrimp trawls 981 Population studies bass, black sea 723 dolphin 527 fish 535 menhaden, gulf 311 sardine 540 POTTHOFF, THOMAS, SHARON KELLEY, and JOA- QUIN C JAVECH, "Cartilage and bone development in scombroid fishes" 647 POWELL, ALLYN B„ and GERMANO PHONLOR, "Early life history of Atlantic menhaden, Brevoortia tyran- nus, and gulf menhaden, B. patronus" 991 POWER, JAMES H., "A model of the drift of northern anchovy, Engraulis mordax, larvae in the California Current" 585 POWERS, KEVIN D.-see PAYNE et al. Predation of northern anchovy 859 "A preliminary investigation of the stock structure of the dolphin, Coryphaena hippurus, in the western central Atlantic," by Hazel A. Oxenford and Wayne Hunte 451 PRESCOTT, ROBERT-see SELZER et al. PRINCE, ERIC D., DENNIS W. LEE, CHARLES A. WILSON, and JOHN M. DEAN, "Longevity and age validation of a tag-recaptured Atlantic sailfish, Istiophorus platypterus, using dorsal spines and otoliths" 493 Prionotus carolinus—see Searobin, northern Prionotus evolans—see Searobin, striped Quahog, ocean organic and trace metal levels 133 QUAST, JAY C, "Annual production of eviscerated body weight, fat, and gonads by Pacific herring, Clupea haren- gus pallasi, near Auke Bay, southeastern Alaska" .... 705 RALSTON, STEPHEN-see DAILEY and RALSTON -see POLOVINA and RALSTON RALSTON, STEPHEN, "An intensive fishing experiment for the caridean shrimp, Heterocarpus laevigatus, at Alamagan Island in the Marina Archipelago" 927 RALSTON, STEPHEN, REGINALD M. GOODING, and GERALD M. LUDWIG, "An ecological survey and com- parison of bottom fish resource assessments (submersible versus handline fishing) at Johnston Atoll" 141 RANCK, CAROL L., FRED M. UTTER, GEORGE B. MILNER, and GARY B. SMITH, "Genetic confirmation of specific distinction of arrowtooth flounder, Atheresthes stomias, and Kamchatka flounder, A. evermanni" 222 "Rates of increase in dolphin population size," by Stephen B. Reilly and Jay Barlow 527 RATTY, F. J., Y. C. SONG, and R. M. LAURS, "Chromosomal analysis of albacore, Thunnus alalwnga, and yellowfin, Thunnus albacares, and skipjack, Katsu- wonus pelamis, tuna" 469 Ray, cownose reproductive biology 871 Red tide effects on Gymnodinium splendens swimming speed 460 REINISCH, C. L.-see FARLEY et al. REILLY, STEPHEN B., and JAY BARLOW, "Rates of increase in dolphin population size" 527 REIS, ENIR GIRONDI, "Age and growth of the marine catfish, Netuma barba (Siluriformes, Ariidae), in the estuary of the Patos Lagoon (Brasil)" 679 RENAUD, MAURICE L., "Hypoxia in Louisiana coastal waters during 1983: implications for fisheries" 19 "Reproductive biology of female spotted dolphins, Stenella attenuata, from the eastern tropical Pacific," by A. C. Myrick, Jr., A. A. Hohn, J. Barlow, and P. A. Sloan 247 "Reproductive biology of king mackerel, Scomberomorus cavalla, from the southeastern United States," by John H. Finucane, L. Alan Collins, Harold A. Brusher, and Carl H. Saloman 841 Reproductive studies anchovy, northern 503, 879 bass, black sea 723 crab, golden king 571 dolphin, spotted 247 mackerel, king 841 mussel, blue 995 ray, cownose 871 shrimp, caridean 915 squid, market 771 tuna, skipjack 895 "Resilience of the fish assemblage in New England tidepools," by Bruce B. Collette 200 Resource assessment techniques hydroacoustic echo integration 287 line intercept survey 287 line transect survey 287 abundance of widow rockfish 287 REXSTAD, ERIC A, and ELLEN K. PIKITCH, "Stomach contents and food consumption estimates of Pacific hake, Merluccius productus" 947 Rhinoptera loonasus—see Ray, cownose "Rhizocephalan infection in blue king crabs, Paralithodes platypus, from Olga Bay, Kodiak Island, Alaska," by P. T. Johnson, R. A. Macintosh, and D. A. Somerton 177 Rhizocephalans length-weight relationships of king crab 327 Rockfish problems in sampling commercial landings 409 Rockfish, widow behavior studies 287 methodologies for assessing abundance 287 Rockfish, yellowtail fat volume comparisons 743 ROGERS, S. GORDON, HIRAM T LANGSTON, and TIMOTHY E. TARGETT, "Anatomical trauma to sponge- coral reef fishes captured by trawling and angling" . . . 697 "The role of estuarine and offshore nursery areas for young English sole, Parophrys vetulus Girard, of Oregon," by E. E. Krygier and W. G. Pearcy 119 ROSENBLATT, RICHARD H.-see GRAHAM et al. ROUMILLAT, WILLIAM A.-see WENNER et al. Sablefish, larval starvation prey size selection 484 484 1013 Sailfish, Atlantic longevity and age validation 493 Salmon, chinook methods used to estimate spawning escapement 261 Salmon, Pacific food habits in the Strait of Juan de Fuca 77 Salmon, smolts correlation with seabirds 279 SALOMAN, CARL H.-see FINUCANE et al. Samplers improved otter surface sampler 480 Sarcoma occurrence in soft shell clams 851 Sardine population collapse 535 population studies 535 "Scavenger feeding by subadult striped bass, Morone sax- atilis, below a low-head hydroelectric dam," by John Warner and Boyd Kynard 220 Scomberomorus cavallasee Mackerel, king Scombroid fish cartilage and bone development 647 Sea surface studies monitoring chlorophyll concentration in the tropical Pacific 687 Seabirds correlations with salmon smolts 279 "Seabirds near an Oregon estuarine salmon hatchery in 1982 and during the 1983 El Nino," by Range D. Bayer . . 279 Seagrass meadows community studies 443 comparison of two sampling methods 443 Seal, harbor in southern New England 217 Seal, northern fur arrival times and numbers on St. Paul Island, Alaska. . 383 diet 957 food and energy consumption of lactating females . . . 371 Searobin, northern differentiation of Prionotus eggs 63 Searobin, striped differentiation of Prionotus eggs 63 "Seasonality of blue mussel, Mytilus edulis L., larvae in the Damariscotta River estuary, Maine, 1969-77," by Greg S. Podniesinski and Bernard J. McAlice 995 Sebastes entomelas—see Rockfish, widow 1014 Sebastes flavidus— see Rockfish, yellowtail SELZER, LAWRENCE, A, GREG EARLY, PATRICIA M. FIORELLI, P. MICHAEL PAYNE, and ROBERT PRESCOTT, "Stranded animals as indicators of prey utilization by harbor seals, Phoca vitulina concolor, in southern New England" 217 SEN, A. R., "Methodological problems in sampling com- mercial rockfish landings" 409 "The sex ratio and gonad indices of swordfish, Xiphias gladius, caught off the coast of Southern California in 1978," by Earl C. Weber and Stephen R. Goldberg ... 185 Shark, longfin mako first record in the Gulf of Mexico 748 SHEPARD, ANDREW N„ ROGER B. THEROUX, RICHARD A. COOPER, and JOSEPH R. UZMANN, "Ecology of Ceriantharia (Coelenterata, Anthozoa) of the northwest Atlantic from Cape Hatteras to Nova Scotia" . . 625 SHIRLEY, THOMAS C.-see HAWKES et al. Shrimp affect of hypoxia on abundance and distribution 19 Shrimp, caridean an intensive fishing experiment 927 reproduction, distribution, and growth 915 Shrimp, northern fecundity studies in northwest Atlantic 549 Shrimp, penaid population sampling 981 Shrimp, pink— see Shrimp, northern SIBERT, JOHN R.-see HUNTER et al. "A simple method for estimating the food consumption of fish populations from growth data and food conversion ex- periments," by Daniel Pauly 827 "Sizes of walleye pollock, Theragra chalcogramma, con- sumed by marine mammals in the Bering Sea," by Kathryn J. Frost and Lloyd F. Lowry 192 "Size-specific vulnerability of northern anchovy, Engraulis mordax, larvae to predation by fishes," by Arild Folkvord and John R. Hunter 859 SLOAN, P. A. -see MYRICK et al. SMITH, GARY B.-see RANCK et al. SMITH, JOSEPH, W., and JOHN V. MERRINER, "Obser- vations on the reproductive biology of the cownose ray, Rhinoptera bonasus, in Chesapeake Bay" 87 SMITH, T. D.-see LO and SMITH Snappers yield assessment in the Marianas 759 Sole, English estuarine and offshore nursery areas 119 SOMERTON, D. A.-see JOHNSON et al. SOMERTON, DAVID A., and ROBERT S. OTTO, "Distri- bution and reproductive biology of the golden king crab, Lithodes aequispina, in the eastern Bering Sea" 571 "Some statistical techniques for estimating abundance in- dices from trawl surveys," by Michael Pennington .... 519 SONG, Y. C.-see RATTY et al. SORENSEN, PETER W., MARCO L. BIANCHINI, and HOWARD E. WINN, "Diel foraging activity of American eels, Anguilla rostrata (LeSueur), in a Rhode Island estuary" 746 "Sounds from Bryde, Balaenoptera edeni, and finback, B. physalus, whales in the Gulf of California," by William C. Cummings, Paul 0. Thompson, and Samuel J. Ha 359 Southern California Bight distribution and abundance of common dolphin 333 Squid, market growth, behavior, and sexual maturation 771 SQUIRES, DALE, "Ex-vessel price linkages in the New England fishing industry" 437 "The spawning frequency of skipjack tuna, Katsuwonus pelamis, from the South Pacific," by J. Roe Hunter, Beverly J. Macewicz, and John R. Sibert 895 "Starvation-induced mortality of young sea-caught jack mackerel, Trachurus symmetricus, determined with histological and morphological methods," by Gail H. Theilacker 1 Statistical methods estimating abundance 519 STEIMLE, FRANK W, PAUL D BOEHM, VINCENT S. ZDANOWICZ, and RALPH A. BRUNO, "Organic and trace metal levels in ocean quahog, Arctica islandica Linne, from the northwestern Atlantic" 133 Stenella attenuata—see Dolphin, spotted STEPIEN, CAROL A., "Life history and larval develop- ment of the giant kelpfish, Heterostichus rostratus Girard, 1854" 809 STIER, KATHLEEN, and BOYD KYNARD, "Abundance, size, and sex ratio of adult sea-run sea lampreys, Petromyzon marinus, in the Connecticut River" 476 STIER, KATHLEEN, and BOYD KYNARD, "Movement of sea-run sea lampreys, Petromyzon marinus, during the spawning migration in the Connecticut River" 749 "Stomach contents and food consumption estimates of Pacific hake, Merluccius productus," by Eric A. Rexstad and Ellen K. Pikitch 947 "Stranded animals as indicators of prey utilization by harbor seals, Phoca vitulina concolor, in southern New England," by Lawrence A. Selzer, Greg Early, Patricia M. Fiorelli, P. Michael Payne, and Robert Prescott 217 Submersibles resource assessment at Johnston Atoll 141 "Survival and growth of striped bass, Morone saxatilis, and Morone hybrid larvae: laboratory and pond enclosure experiments," by Edward D. Houde and Lawrence Lubbers III 905 Swordfish sex ratio and gonad indices 185 morphology of Fistulicola plicatus 754 SYKES, STEPHEN D., and LOUIS W BOTSFORD, "Chinook salmon, Oncorhynchus tshawytscha, spawning escapement based on multiple mark-recapture of carcasses" 261 TARGETT, TIMOTHY E.-see ROGERS et al. Taxonomy arrowtooth and Kamchatka flounders 222 TESTER, PATRICIA A., and ANDREW G. CAREY, JR., "Instar identification and life history aspects of juve- nile deepwater spider crabs, Chionoecetes tanneri Rathbun" 973 THEILACKER, GAIL H., "Starvation-induced mortality of young sea-caught jack mackerel, Trachurus sym- metricus, determined with histological and morphological methods" 1 Theragra chalcogramma—see Pollock, walleye THEROUX, ROGER B.-see SHEPARD et al. THOMPSON, PAUL O.-see CUMMINGS et al. Thunnus alalunga—see Albacore Thunnus albacares—see Tuna, yellowfin Tidepools fish assemblages in New England 200 Townsend Cromwell resource assessment at Johnston Atoll 141 Trachurus symmetricus— see Mackerel, jack Trawls, shrimp population sampling devices 981 Trawl surveys statistical methods for estimating abundance 519 Trawling anatomical trauma to sponge-coral reef fishes 697 TUCKER, G. E.-see PARSONS and TUCKER 1015 Tuna, skipjack chromosomal analysis 469 spawning frequency 895 Tuna, yellowfin chromosomal analysis 469 incidental dolphin mortality 559 morphology and swimming mode 463 TURK, P. E.-see YANG et al. "Type, quantity, and size of food of Pacific salmon (Oncorhynchus) in the Strait of Juan de Fuca, British Columbia," by Terry D. Beacham 77 UCHIYAMA, JAMES H., RAYMOND K. BURCH, and SYD A. KRAUL, JR., "Growth of dolphins, Coryphaena hippurus and C. equiselis, in Hawaiian waters as deter- mined by daily increments on otoliths" Underwater sounds Bryde and finback whales 186 359 UTTER, FRED M.-see RANCK et al. UZMANN, JOSEPH R.-see SHEPARD et al. WEBER, EARL C., and STEPHEN R. GOLDBERG, "The sex ratio and gonad indices of swordfish, Xiphias gladius, caught off the coast of Southern California in 1978" 185 WENNER, CHARLES A., WILLIAM A. ROUMILLAT, and C. WAYNE WALTZ, "Contributions to the life history of black sea bass, Centropristis striata, off the south- eastern United States" 723 Whales, Bryde underwater sounds in the Gulf of California 359 Whales, finback underwater sounds in the Gulf of California 359 Whales, humpback distribution in relation to sand eels 271 WILKINS, MARK E., "Development and evaluation of methodologies for assessing and monitoring the abundance of widow rockfish, Sebastes entomelas" 287 WILLASON, STEWART W, JOHN FAVUZZI, and JAMES L. COX, "Patchiness and nutritional condition of zooplankton in the California Current" 157 WILSON, CHARLES, A.-see PRINCE et al. VAN ENGEL, W A., R. E. HARRIS, JR., and D. E. ZWERNER, "Occurrence of some parasites and a com- mensal in the American lobster, Homarus americanus, from the Mid-Atlantic Bight" 197 "A variable catchability version of the Leslie model with application to an intensive fishing experiment on a multispecies stock," by Jeffrey J. Polovina 423 "Variations in the morphology of Fistulicola plicatus Rudolphi (1802) (Cestoda: Pseudophyllidea) from the swordfish, Xiphias gladius L., in the northwest Atlantic Ocean," by W. E. Hogans and P. C. F. Hurley 754 WAHLEN, BRUCE E., "Incidental dolphin mortality in the eastern tropical Pacific tuna fishery, 1973 through 1978" 559 WALTZ, C. WAYNE-see WENNER et al. WARNER, JOHN, and BOYD KYNARD, "Scavenger feeding by subadult striped bass, Morone saxatilis, below a low-head hydroelectric dam" 220 WINN, HOWARD E.-see KENNEY and WINN -see SORENSEN et al. Xiphias gladius— see Swordfish YANG, M. S., and P. A. LIVINGSTON, "Food habits and diet overlap of two congeneric species, Atheresthes stomias and Atheresthes evermanni, in the eastern Bering Sea" 615 YONEDA, NELSON TAKUMI-see MATSUURA and YONEDA ZDANOWICZ, VINCENT S.-see STEIMLE et al. Zooplankton California Current 157 ZWERNER, D E.-see VAN ENGEL et al. 1016 NOTICES NOAA Technical Reports NMFS published during first 6 months of 1986. 39. Survey of fish protective facilities at water withdrawl sites on the Snake and Columbia Rivers. By George A. Swan, Tommy G. Withrow, and Donn L. Park. April 1986, iii + 34 p., 26 figs., 6 tables. Some NOAA publications are available by purchase from the Superintendent of Documents, U.S. Govern- ment Printing Office, Washington, DC 20402. Gulf of Mexico Ichthyoplankton Samples The Gulf States Marine Fisheries Commission wishes to announce the availability of Gulf of Mexico ichthyoplankton samples for loan to qualified researchers. Samples have been and are continuing to be collected for SEAMAP (Southeast Area Monitoring and Assessment Program), a multi-year international federal/state/university program of the GSMFC. Neuston and bongo nets were employed for specimen collection in a one degree latitude/longitude grid over the entire Gulf from 26 °N northward and sorted and preliminarily identified by the Plankton Sorting and Identification Center, Szczecin, Poland. At present samples from 1982 (7057 lots, 93 families), 1983 (8351 lots, 106 families) and material from one summer cruise in 1984 (4155 lots, 75 families) are available for loan. Lots of unsorted fish eggs are also available from these years. Most samples are sorted to the family level, although many have identification to generic or species level. Additional 1984 samples are expected to become available by the end of 1986. Specimens are available for loan on a 6-month renewable basis. Researchers interested in obtaining additional information can contact either SEAMAP Ichthyoplankton Curator, Florida Department of Natural Resources, Bureau of Marine Research, St. Petersburg, FL 33701, or SEAMAP Coordinator, Gulf States Marine Fisheries Commission, P.O. Box 726, Ocean Springs, MS 39564. 1017 U.S. POSTAL SERVICE STATEMENT OF OWNERSHIP, MANAGEMENT AND CIRCULATION (Required by }» U.S.C. 36SS) 1. TTTUI Of PUBLICATION Fishery Bui letin A PUBLICATION NO. 2. DATE OF FILINQ 1 October 1986 S. FREQUENCY OF ISSUE Quarterly A. NO. OF ISSUES PUBLISHED ANNUALLY B. ANNUAL SUBSCRIPTION PRICE $21.00 4 COMPLETE MAILING AOORESS OF KNOWN OFFICE OF PUBLICATION (Stittt. City. County, Slate and ZIP Cod*) (Not prlnten) Scientific Publications Office, NMFS 7600 Sand Point Way N.E., BIN C15700 (Bldg. 1), Seattle, WA 98115 3 COMPLETE MAILING ADDRESS OF THE HEADOUARTERS OR GENERAL BUSINESS OFFICES OF THE PUBLISHERS (Not print, n) National Marine Fisheries Service, NOAA, DOC Universal Building, Washington, DC 20235 6. FULL NAMES AND COMPLETE MAILING ADDRESS OF PUBLISHER. EDITOR, AND MANAGING EDITOR (Thtt Item MUST NOT be blenk) PUBLISHER (Name and Complete Milling Addreai) Jack McCormick, Scientific Publications Office, NMFS 7600 Sand Point Way N.E., BIN C15700 (Bldg. 1), Seattle, WA 98115 EDITOR (Name and Compnn Mailing Addreai) Mary Fukuyama, Scientific Publications Office, NMFS 7600 Sand Point Way N.E., BIN C15700 (Bldg. 1), Seattle, WA 98115 MANAGING EDITOR (Nama and Complata Mailing Addreaa) Mary Fukuyama, Scientific Publications Office, NMFS 7600 Sand Point Way N.E., BIN C17500 (Bldg. 1), Seattle, WA 98115 7. OWNER tit owned by a corporation, its nama and address must be stated and also immediately thereunder the names and addresses of stockholders owning or holding I percent or more of total amoun r of stock. If not owned by a corporation, the names and addresses of the individual owners must be given. If owned by a partnership or other unincorporated firm, its name and address, as well as that of each individual must be given. If the publica- tion is published by a nonprofit organization, its name and address must be stated.) {Item must be completed) FULL NAME COMPLETE MAILING ADDRESS lAth St., NW, Washington, DC 20230 U.S. Department of Commerce t KNOWN BONDHOLDERS MORTGAGEES. AND OTHER SECURITY HOLDERS OWNING OR HOLDING 1 PERCENT OR MORE OF TOTAL AMOUNT OF BONDS. MORTGAGES OR OTHER SECURITIES (II there are none, «o Hale) FULL NAME COMPLETE MAILING ADDRESS None 9. FOR COMPLETION BY NONPROFIT ORGANIZATIONS AUTHORIZED TO MAIL AT SPECIAL RATES (Section 411.3, DMM only) The purpose, (unction and nonprolit status of this organization and the exempt status lor Federal income lax purposes [Check one) m ID HAS NOT CHANGED DURING PRECEDING 12 MONTHS □ m HAS CHANGED DURING PRECEDING 12 MONTHS (It changed, publisher mutt submit explanation of change with this statement.) 10. EXTENT AND NATURE OF CIRCULATION AVERAGE NO COPIES EACH ISSUE DURING PRECEDING 12 MONTHS ACTUAL NO COPIES OF SINGLE ISSUE PUBLISHED NEAREST TO FILING DATE A. TOTAL NO. COPIES (Mel Am Run) 2,197 2,066 i. paio circulation (handled by U.S. 1. SALES THROUGH DEALERS AND CARRIERS. STREET ' .*. VENDORS AND COUNTER SALES Wash. . DC 20240) 1P0 J MAIL SUBSCRIPTION c total paid circulation rsum of imi«m« i«2i(pri n ted for sales 750 600 O.FREE DISTRIBUTION BY MAIL, CARRIER OR OTHER MEANS ( TeqUeS ted SAMPLES. COMPLIMENTARY, AND OTHER FREE COPIES C OD i 6 S ) 1.M.7 1,466 E TOTAL DISTRIBUTION (Sum el C and 01 2,197 2,066 F COPIES NOT DISTRIBUTED 1. OFFICE USE. LEFT OVER. UNACCOUNTED. SPOILED AFTER PRINTING 1 RETURN FROM NEWS AGENTS Q. TOTAL (Sum of E. fl end 2 • thoute equal net prms run ihown m A) 2,197 2.066 I certify that the statements made by me above are correct and complete SIGNATURE AND TITLE OF EDITOR. PUBLISHES. BUSINESS MANAGER. OR OWNER . /Tl ' /) July 1981 on on reverie) (Page 1) 032 INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN Manuscripts submitted to the Fishery Bulletin will reach print th mform to the following instructions These are not absolute requirements, of course, but desiderata. sibility of tl and serials shoi Data Base. {Ci< was developed h CONTENT OF MANUSCRIPT The title page should give only the title of the paper, the author's name, his affiliation, and mailing address, in- cluding ZIP code thor. Abbrevial ions of names of periodicals riform U > & rial Sources for the BIOSIS - bstracts also uses this system, which Standai Association.) The abstract should not exceed one double-spaced page In the text, Fishery Bulletin style, for the most part, follows that of the U.S. Government Printing Office Style Manual. Fish names follow the style of the American Fisheries Society Special Publication No. 12, A List of Com- mon and Scientific Names of Fishes from the United States and Canada, Fourth Edition, 1980. Text footnotes should be typed separately from the text. Figures and tables, with their legends and headings, should be self-explanatory, not requiring reference to the text. Their placement should be indicated in the right-hand margin of the manuscript. Preferably figures should be reduced by photography to 53/« inches (for single-column figures, allowing for 50% reduction in printing), or to 12 inches (for double-column figures). The maximum height, for either width, is 14 inches. Photographs should be printed on high quality or glossy paper. Do not send original drawings to the Scientific Editor; if they, rather than the photographic reductions, are needed by the printer, the Scientific Publications Office will request them. Each table should start on a separate page Consistency in headings and format is desirable Vertical rules should be avoided, as they make the tables more expensive to print. Footnotes in tables should be numbered sequentially in arabic numerals. To avoid confusion with powers, they should be placed to the left of numerals. Acknowledgments, if included, are placed at the end of the text. Literature is cited in the text as: Lynn and Reid (1968) or (Lynn and Reid 1968). All papers referred to in the text should be listed alphabetically by the senior author's sur- name under the heading "Literature Cited." Only the author's surname and initials are required in the literature cited. The accuracy of the literature cited is the respon- Common abbreviation and symbols, such < ..k g, L, mL, mg, °C (for be used. Abbreviate with numerals. 1 tions. We prefer that measi other equivalent units as mm, m, %, °/00, and so forth, should " measure only when used • rarely used with abbrevia- given in metric unit in parentht w FORM OF THL The original of the man; spaced, on white bond paper. 1 headings. We would rather re of manuscripts than carbon copn material should be: TITLE PAGE ABSTRACT TEXT LITERATURE CITED TEXT FOOTNOTES APPENDIX TABLES (Each table should I arabic numeral and heading pr< LIST OF FIGURES (Entire fi< FIGURES (Each figure sb arabic numeral; legends ai ADDITIONAL INFORMATION Send the ribbon copy and two duplicated or carbon copies of the manuscript to: Dr. William J. Richards, Scientific Editor Fishery Bulletin Southeast Fisheries Center Miami Laboratory National Marine Fisheries Service, NOAA 75 Virginia Beach Drive Miami, FL 33149-1099 Fifty separates will be supplied to an author free of charge and 50 supplied to his organization. No covers will be supplied. $,. Onteryts— Contim REXSTAD, ERIC A., and ELLEN K. PUCITCH. Stomach contents and food consump- tion estimates of Pacific hak< us productus 947 PEREZ, MICHAEL A and MV i ' iEL' A. BIGG. Diet of northern fur seals, CaUorhinvs ursinus, off western D >'rica 957 TESTER, PATRICIA A NTPREW G. CAREY, JR. Instar identification and life history aspects of ju\ deepvater spider crabs, Chionoecetes tanneri Rathbun . . . 973 jar Jfij Notes CODY TERRY i E. FULS. Comparison of catches in 4.3 m and 12.2 m shrimp trawls in th< M 'xico 981 POWELL, AT B. 4 GERMANO PHONLOR. Early life history of Atlantic men- haden, Bre . ' ' innnus, and gulf menhaden, B. patronus 991 PODNL G s-> antl BERNARD J. McALICE. Seasonality of blue mussel, arvae in the Damariscotta River estuary, Maine, 1969-77 995 1003 1017 LIBRARY MARIN€ 'B'JOLOGICAL LABORATORY *>;\n/L 02543 • GPO 791-008 MBL WHOI LIBRARY UH 1 UC H