MBL ..< °7% ^^ATESO^^ Vol. 87, No. 1 Marine Biological Laboratory LIBRARY AUG 1 5 1989 Woods Hole, Mass. January 1989 / RALSTON, STEPHEN, and HAPPY A. WILLIAMS. Numerical integration of daily growth increments: an efficient means of ageing tropical fishes for stock ^ 3.ss6SsmGnt BECKMAN, DANIEL W., CHARLES A. WILSON, and A. LOUISE STANLEY. Age and growth of red drum, Sciaenops ocellatus, from offshore waters of the northern Gulf of Mexico 1'^ THRESHER, R. E., B. D. BRUCE, D. M. FURLANI, and J. S. GUNN. Distribution, advection, and growth of larvae of the southern temperate gadoid, Macruronus novae- zelandiae (Teleostei: Merlucciidae), in Australian coastal waters 29 COLLINS, MARK R., DAVID J. SCHMIDT, C. WAYNE WALTZ, and JAMES L. PICKNEY. Age and growth of king mackerel, Scomberomorus eavalla, from the Atlantic coast of the United States 49 MARGULIES, DANIEL. Effects of food concentration and temperature on develop- ment, growth, and survival of white perch, Morone americana, eggs and larvae. . . 63 THORROLD, SIMON R. Estimating some early life history parameters in a tropical clupeid, Herklotsichthys castelnaui, from daily growth increments in otoliths .... 73 FELDKAMP, STEVEN D., DANIEL R COSTA, and GREGORY K. DeKREY. Ener- getic and behavioral effects of net entanglements on juvenile northern fur seals, Callorhinns ursinus 85 EMMETT, B., and G. S. JAMIESON. An experimental transplant of northern abalone, Haliotis kamtschatkana, in Barkley Sound, British Columbia 95 WILLIAMS, AUSTIN B., and C. ALLAN CHILD. Comparison of some genera and species of box crabs (Brachyura: Calappidae), southwestern North Atlantic, with description of a new genus and species 105 POLACHECK, TOM. Yellowfin tuna, Thunnus albacares, catch rates in the western Pacific 123 LAUTH, ROBERT R. Seasonal spawning cycle, spawning frequency, and batch fecun- dity of the cabezon, Sco-rpaenichthys marmoratus, in Puget Sound, Washington . . . 145 WENNER, ELIZABETH L., and CHARLES A. WENNER. Seasonal composition and abundance of decapod and stomatopod crustaceans from coastal habitats, southeast- ern United States 155 (Continued on back cover) Seattle, Washington U.S. DEPARTMENT OF COMMERCE Robert Mosbacher, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION William E. Evans, Under Secretary for Oceans and Atmosphere NATIONAL MARINE FISHERIES SERVICE James W. Brennan, Assistant Administrator for Fisheries 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 quarter- ly. 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. Andrew E. Dizon Southwest Fisheries Center La Jolla Laboratory National Marine Fisheries Service, NOAA P.O. Box 271 La Jolla, CA 92038 Editorial Committee Dr. Jay Barlow National Marine Fisheries Service Dr. William H. Bayliff Inter-American Tropical Tuna Commission Dr. George W. Boehlert National Marine Fisheries Service Dr. Robert C. Francis University of Washington Dr. James R. Kitchell University of Wisconsin Dr. William J. Richards National Marine Fisheries Service Dr. Bruce B. Collette National Marine Fisheries Service Dr. Tim D. Smith National Marine Fisheries Service Mary S. Fukuyama, Managing Editor The Fishxry 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 Office, Washington. DC 20402. Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated. The Secretary of Commerce has determined that the publication of this periodical is necessary in the transaction of the public business required by law of this Department. Use of funds for printing of this periodical has been approved by the Director of the Office of Management and Budget. Fishery Bulletin CONTENTS Vol. 87, No. 1 January 1989 RALSTON, STEPHEN, and HAPPY A. WILLIAMS. Numerical integration of daily growth increments: an efficient means of ageing tropical fishes for stock assessment 1 BECKMAN, DANIEL W., CHARLES A. WILSON, and A. LOUISE STANLEY. Age and growth of red drum, Sciaenops ocellatus, from offshore waters of the northern Gulf of Mexico 17 THRESHER, R. E., B. D. BRUCE, D. M. FURLANI, and J. S. GUNN. Distribution, advection, and growth of larvae of the southern temperate gadoid, Macruronus rwvae- zelandiae (Teleostei: Merlucciidae), in Australian coastal waters 29 COLLINS, MARK R., DAVID J. SCHMIDT, C. WAYNE WALTZ, and JAMES L. PICKNEY Age and growth of king mackerel, Scomberomorus cavalla, from the Atlantic coast of the United States 49 MARGULIES, DANIEL. Effects of food concentration and temperature on develop- ment, growth, and survival of white perch, Morone americana, eggs and larvae. . . 63 THORROLD, SIMON R. Estimating some early life history parameters in a tropical clupeid, Herklotsichthys castelnaui, from daily growth increments in otoliths .... 73 FELDKAMP, STEVEN D., DANIEL P COSTA, and GREGORY K. DeKREY Ener- getic and behavioral effects of net entanglements on juvenile northern fur seals, Callorhinv^ ursinus 85 EMMETT, B., and G. S. JAMIE SON. An experimental transplant of northern abalone, Haliotis kamtschatkana, in Barkley Sound, British Colimibia 95 WILLIAMS, AUSTIN B., and C. ALLAN CHILD. Comparison of some genera and species of box crabs (Brachyura: Calappidae), southwestern North Atlantic, with description of a new genus and species 105 POLACHECK, TOM. Yellowfin tuna, Thunnns albacares, catch rates in the western Pacific 123 LAUTH, ROBERT R. Seasonal spawning cycle, spawning frequency, and batch fecun- dity of the cabezon, Scorpaenichthys marmoratus, in Puget Sound, Washington . . . 145 WENNER, ELIZABETH L., and CHARLES A. WENNER. Seasonal composition and abundance of decapod and stomatopod crustaceans from coastal habitats, southeast- ern United States 155 (Continued on next page) Seattle, Washington 1989 Marine Biological Laboratory LIBRARY AUG 1 5 1989 For sale by the Superintendent of Documents. U.S. Government Printing Offic . WasM^© CJS HolS, MSSS. DC 20402— Subscription price per year: $16.00 domestic and $20.00 foreign. Cost per smgle issue: $9.00 domestic and $11.25 foreign. — ^■■^■••mi^^^^^^mm^ (Contents— Continued) HIGHTOWER, JOSEPH E., and GARY D. GROSSMAN. Status of the tilefish, Lopho- latilus chamaeleonticepes, fishery off South Carohna and Georgia and recommenda- tions for management 177 SMITH, INGRID, MARK S. FONSECA, JOSE A. RIVERA, and KEITH A. RITT- MASTER. Habitat value of natural versus recently transplanted eelgrass, Zostera marina, for the bay scallop, Argopecten irradians 189 ROPES, JOHN W. The food habits of five crab species at Pettaquamscutt River, Rhode Island 197 CRAWFORD, MAURICE K, CHURCHILL B. GRIMES, and NORMAN E. BUROKER. Stock identification of weakfish, Cynoscion regalis, in the Middle Atlantic region . . . 205 GARTNER, JOHN V., JR., WALTER J. CONLEY, and THOMAS L. HOPKINS. Escape- ment by fishes from midwater trawlers: a case study using lanternfishes (Pisces: Myctophidae) 213 AGNELLO, RICHARD J. The economic value of fishing success: an application of socioeconomic survey data 223 Notes MORING, JOHN R. Food habits and algal associations of juvenile lumpfish, Cycloptents lumpus L., in intertidal waters 233 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. NUMERICAL INTEGRATION OF DAILY GROWTH INCREMENTS: AN EFFICIENT MEANS OF AGEING TROPICAL FISHES FOR STOCK ASSESSMENT Stephen Ralston' and Happy A. Williams^ ABSTRACT For an objective, cost-effective ageing methodologi,' applicable to tropical species, a new approach to estimating parameters of the von Bertalanffy growth equation through the study of otolith microstruc- ture was developed and applied to Pristipomoides zonatus. a deepwater snapper widely distributed throughout the Indo-Pacific region. The average width of sagittal daily growth increments was used to measure otolith growth rate, which was then related to the size of the otolith. The data were numerically integrated, providing estimates of age (in years) at regular 500 |jm increments to otolith length, which was then used to predict fork length (FL mm) at age with regression analysis. The data were fitted to the von Bertalanffy growth model, resulting in FL = 442 (1 - exp(- 0.234 (Age + 0.892))). The method was critically examined and validated through the study of 1) annual hyaline and opaque markings that appear in the otoliths, 2) Monte Carlo simulation. 3) length-frequency analysis. 4) ex- amination of spawning seasonality relative to back-calculated birth date taken from the time of first annulus formation, and 5) empirical comparisons with the literature concerning snapper growth. Developing stock-assessment models tailored to the characteristics and needs of tropical fisheries is an area of active and productive research. In particular, significant progress has been made over the last several years in the area of length-based methods (Schnute and Fournier 1980; Jones 1981; Pauly 1982, 1987b; Fournier and Breen 1983; Fournier and Doonan 1987; Schnute 1987). With these ad- vances, a powerful array of biologically realistic models is now available for analyzing length-fre- quency data. Although tremendous strides have been made in developing these new length-based methods, the im- portance of acquiring other information besides length-frequency data and total catch statistics is all the more evident. Ancillary information usually helps to stabilize and improve the estimation of model parameters (Schnute and Fournier 1980; Fournier and Doonan 1987). Foremost is develop- ing an independent knowledge of growth dynamics (Gulland 1987; Morgan 1987). It is now generally accepted that the analysis of length-frequency data, 'Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI 96822-2396; present address: Southwest Fisheries Center Tiburon Laboratory, National Marine Fisheries Service, NOAA, 3150 Para- dise Drive, Tiburon, CA 94920. ^Southwest Fisheries Center Honolulu Laboratory, National Marine Fisheries Service. NOAA, 2570 Dole Street, Honolulu, HI 96822-2396. Manuscript accepted September 1988. Fishery Bulletin. U.S. 87:1-16. in conjunction with age estimates derived from the study of hard parts, represents the most promising avenue for future assessment work on exploited tropical species (Pauly 1987a). Nonetheless, estimating growth rates of tropical species by using otoliths has been a difficult and per- sistent problem. Investigators have often failed in their efforts, either because of an absence of con- ventional hyaline and opaque markings, as is true of most tropical species, or because of an aversion to direct enumeration of daily otolith increments. The latter can be an extremely difficult, time con- suming, and tedious process. Since Pannella (1971) first discovered the exis- tence of daily otolith increments, a large body of work has developed on the subject. While many investigators have touted the potential benefits of ageing tropical species by using otolith microstruc- ture, few have attempted to develop growth curves with assessment goals specifically in mind. Instead, most work to date has dealt with ageing larval forms (Jones 1986) and elucidating endogenous and environmental effects on increment formation (Cam- pana and Neilson 1985). Although much useful in- formation has been gained, daily increments have yet to fulfill their promise with respect to applica- tions in the area of juvenile and adult population dynamics. The purpose of this study was to develop a gen- eral method of ageing tropical fishes by using daily 1 FISHERY BULLETIN: VOL. 87, NO. 1 growth increments, specifically with stock-assess- ment applications in mind. In this regard, the von Bertalanffy growth equation (Ricker 1979) is of fun- damental importance. Due to its widespread use in assessment models (e.g., the Beverton and Holt [1957] yield formulation), parameter estimates for this equation provide an ideal complement to many of the length-based methods that are currently in use (e.g., Morgan 1987). The ultimate goal of this study was, therefore, to develop a methodology to estimate the von Bertalanffy growth parameters K and L„ from the study of daily increments. Ideal- ly, the approach developed should be general in its application, easy to implement, simple in its tech- nical requirements, and cost effective. Were such a uniform framework to the study of age and growth of exploited tropical species developed, it would assist routine assessment work greatly. MATERIALS AND METHODS As part of a larger program to assess stocks of deep slope fishes in the Mariana Archipelago (Polo- vina 1985; Polovina and Ralston 1986; Ralston in press b), a study of the age and growth of gindai, Pristipomoides zotmtus, was initiated. This commer- cially important eteline snapper (Lutjanidae) is wide- ly distributed in the Indo-Pacific region (Allen 1985) and is the most commonly caught species in Guam's deepwater hook-and-line fishery (Polovina 1986). Field sampling for gindai specimens was con- ducted from the NOAA ship Townsend Cromwell during the 2 yr period spanning April 1982 to May 1984. During this time, six 40 d cruises were com- pleted, such that samples of gindai were obtained during all months of the year except March, Sep- tember, and October. All gindai were caught during daylight hours by using hydraulic fishing reels equipped with circle fish hooks. When landed, fish were measured to the nearest millimeter fork length (FL) with a measur- ing board and weighed to the nearest 0.01 kg on a beam balance. Specimens were sexed at the time of capture by gross examination of the gonads. In addition, a rep- resentative selection of the gonads was frozen for more detailed examination in the laboratory. There, they were preserved in a solution of 10% buffered formalin and weighed to the nearest 0.1 mg. Ova- ries were staged with the classification of Everson (1984), developed for Etelis carbunculus. a related deepwater eteline lutjanid. His classification recog- nizes seven stages based on egg size, shape, and yolk content, i.e., (I) primordial, (II) early developing, (III) developing, (IV) advanced developing, (V) early ripe, (VI) ripe, and (VII) residual. Gonadosomatic indexes (gonad weight expressed as a percentage of body weight) also were calculated where possible. Otoliths At the time of capture, sagittal otoliths were col- lected, by frontal section through the cranium, from certain individuals sampled uniformly from the full size range of gindai captured. The otoliths were rinsed in fresh water to remove adhering mem- branes and endolymph and were stored dry in glass vials. Later in the laboratory, they were examined with a dissecting microscope for the presence of hyaline (i.e., translucent) and opaque markings while illuminated with reflected light against a dark back- ground. When markings were present, the distance from the focus to the beginning of each opaque zone was measured along the postrostral growth axis by using a calibrated ocular micrometer. Total otolith length (focus to postrostrum) also was recorded. A random subsample of gindai otoliths was taken, and their microstructure examined for the presence of daily increments (Campana and Neilson 1985). To prepare the otoliths, they were first embedded in casting resin, which was allowed to harden com- pletely. Cast otoliths were sectioned on a Buehler^ ISOMET low speed jewelry saw. Thin (0.70 mm) sec- tions were made through the focus along a frontal plane to the most distal portion of the postrostrum. Sections were polished sequentially on a Buehler ECOMET polisher/grinder with 180 and 600 grit abrasive disks. Samples were then briefly etched for 5-30 seconds in a dilute solution of 1% HCl, washed in water, and dried. Prepared sections were mounted on glass slides with Euparol or Flotexx and cover slips and allowed to clear and harden com- pletely prior to viewing (approximately 2 weeks). Mounted otolith sections were examined with a compound binocular microscope by using trans- mitted light at a magnification of 200 or 400 x . Total lengths of the otoliths (i.e., the distance in micro- meters between the focus and the postrostral margin) were measured (N = 94) and individual readings were made at selected points along the postrostral growth axis, v/herever it was possible to distinguish the characteristic bipartite structure of daily increments. At each site sampled the aver- age width of presumptive daily growth increments was determined by counting a small number (me- ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service. NOAA. RALSTON and WILLIAMS: AGEING OF TROPICAL FISHES dian = 14, range = 5-22) of increments and measur- ing the axial length of the short segment in which they occurred. In addition, the curvilinear distance between the midpoint of each segment and the otolith focus was measured along the focus to postrostral growth axis. Up to 12 readings were made from each preparation. The focus was defined to be the most posterior of what typically were several primordia (e.g., Radtke 1987). The data were summarized by computing the ratio of segment length in micrometers to the included number of increments at each specific site examined, providing an estimate of the average increment width at some measured distance from the otolith focus. Under the assumption that one increment forms each day, these data can be used to estimate the instantaneous growth rate of the otolith (Ralston and Miyamoto 1981, 1983; Ralston 1985). To estimate age. a simple form of numerical in- tegration was employed. Starting at the focus, the data were subdivided into 500 jjm intervals of otolith length. For each interval, the arithmetic mean growth rate of the otolith was calculated based upon the number of readings falling therein. This aver- age growth rate was then divided into 500 fxm to estimate the number of days needed to complete growth through the intervals, which were sequen- tially accumulated away from the focus, and finally divided by 365.25 to convert age estimates to years. The size of the otolith upon completion of growth through each interval was used to predict the cor- responding FL of the fish after the natural loga- rithm of FL was regressed on the logarithm of total otolith length. These data (age [in years] and FL [mm]) were then fitted to the von Bertalanffy growth equation (Ricker 1979) by using a nonlinear regression routine (SAS Institute Inc. 1979, NLIN procedure). Monte Carlo simulation techniques (Naylor et al. 1966) were applied to this analytical procedure to evaluate the accuracy (i.e., bias) of the estimator and to study the precision of parameter estimates. The structure of the simulation model was such that von Bertalanffy growth was assumed by stipulating a decreasing linear relationship between somatic growth rate and length, i.e., d(FL)ldt = A'(L„ - FL). Likewise, the relationship between otolith length (OL) and FL was assumed to be governed by the power function, so that FL = oOL''. Otolith growth rate, rf(OL)/rff, was then obtained by form- ing the ratio of d{FL)ldt and rf(FL)/rf(OL). All parameters in the model were otherwise set equal to the estimates obtained from the otolith study. and the specific probability distributions invoked were similar to those encountered with the actual data. Length-Frequency Analysis As an independent means of verifying results ob- tained through the study of otoliths, the regression method of Wetherall et al. (1987) was used to esti- mate specific growth and mortality parameters characterizing the study population. The analysis was based on the combined length-frequency distri- bution (FL rounded to the nearest 10 mm) of all gin- dai sampled (see Ralston [in press a] for a discus- sion of the effects of pooling length data taken at different times throughout the year). Initially, this method requires determination of the least FL at which fish are fully represented in the catch ((c.min)- ^or this purpose, the first size class larger than the mode was assumed to be the small- est length category fully sampled (see, for example, Ricker 1975). Moreover, for this and any larger cut- off value (^f,,), we were able to compute the mean size of fully vulnerable fish in the catch (f ,), i.e., those fish greater than (',,,. As I,, was successive- ly advanced through the fully vulnerable size range, the mean and variance in size of larger fish were recalculated at each step, and a series of ordered pairs was developed. The actual estimation proce- dure involved regressing values of I , against suc- cessive values of f^ ,. The inverse of the standard error of (^ was used as a statistical weight for each point, leading to the best linear unbiased esti- mates of the slope (6) and intercept (C). With the resulting regression statistics, the formulae pro- vided in Wetherall et al. (1987) were used to obtain point estimates of the ratio of total instantaneous mortality rate to the von Bertalanffy growth coef- ficient (ZIK) and the von Bertalanffy asymptotic size parameter (LJ). In particular, they showed that ZIK = 6I{1 - 6) and L„ = ^/(l - 6). Likewise, error estimates for these statistics were calcuated as well. RESULTS Age Estimation from Increment Microstructure In all, 440 otoliths were extracted, and of these, 94 were sectioned and examined for daily incre- ments. As expected, there is a clear statistical basis for predicting FL from OL (Fig. I). The regression equation relating these variables is highly signifi- cant (P < 0.0001) and is given by FISHERY Bl'LLKTIN; VOL. «7, NO. 1 E E o 6.200 6.100 6.000 5.900 5.800 5.700 5.600 5.500 5.400 5.300 n O O . o A#% - o o 93.700 ^im. Sum of IVIean Source df squares square F P OL <3,700 (im Equality adjusted means 1 0.0015 0.0015 0.01 0.935 Zero slope 1 121.9917 121.9917 551.78 0.001 Error 316 69.8635 0.2219 Equality slopes 1 0.7158 0.7158 326 0.072 Error 315 69.1477 0.2195 OL >3,700 (im Equality adjusted means 1 0.1334 0.1334 1.14 0286 Zero slope 1 15.9656 15.9656 136.72 0.001 Error 389 45.4273 0.1168 Equality slopes 1 0.2202 0.2202 1.89 0.170 Error 388 45.2071 0.1165 FISHERY BULLETIN; VOL. 87. NO. 1 and the variance in growth rate o", within each of the i = 1,15 intervals of otolith length (Table 2) show that as otolith length increased both d{OL)ldt and o^ declined. The estimated age (in years) at the point of transition between each of the 15 otolith length intervals (i.e., upon completion of growth through interval k) was Aget = 1 365 k I i-i d(,OL)ldt, A(OL) where A(OL) is 500 fim in the application presented here. Otolith length upon completion of growrth through interval k was converted to the equivalent FL (see Figure 1), and the data fitted to the von Bertalanffy growth equation. Because this model poorly repre- sents growth during the early life history, only data representing otolith length intervals in excess of 3,000 i^m (i.e., ages >0.8 year) were used in the regression analysis (see Discussion). Table 2 also provides a statistical weight for each of the age esti- mates. Weighting was desirable because 1) the sam- ple size of each mean varied, 2) the o^, were heterogeneous (proportional to the square of the mean), and 3) compounding of error occurred be- cause of the additive property of the estimator. Weights were calculated as the reciprocal of the sum of standard errors of the means through interval k. The weighted least squares fit to the von Bertalanffy equation (Fig. 4) was FL = 442 (1 - exp(- 0.234 (Age + 0.892))), with 99.99% of the total variation in FL explained by the model, and with asymptotic standard errors for L„, K, and «„ equal to 14.85 mm, 0.0180 yr-\ and 0.078 year, respectively. The results of the Monte Carlo simulation indicate that the estimation procedure was unbiased. Follow- ing 50 computer replications of the same sampling procedures outlined above, there was no detectable bias in the estimation of either K or L^, even though the coefficients of variation for the standard errors of these statistics were both small (0.64 and 0.84%, respectively). Moreover, variance estimates derived from the approximately normal simulation sampling distributions of K and L„ provided a basis for placing confidence intervals on the point esti- mates as follows: P(0.213im) at annulus Age group 1 II III IV V 1 II - III 6 3,233 4,700 5,600 IV 13 3,054 4,400 5,408 6,315 V 10 3,130 4,440 5,610 6,530 7,260 Weighted mean (Mm) 3,117 4,476 5,517 6,409 7,260 Standard deviat ion 222 349 427 468 443 Fork length (mm) 128 189 237 278 318 FISHERY BULLETIN: VOL. 87, NO. 1 100-1 c O) 80- re 4) c 5, 60 X i « 40 o O o 20 o-\ 1 1 1 1 1 1 1 1 1 1 1 I JFMAMJJASOND Month Figure 6.— The seasonal occurrence of hyaline and opaque markings on the margins of gindai otoliths. parameters of the von Bertalanffy growth equation by means of a Walford plot (Ricker 1975), wherein results from a regression of FL at time t + 1 against FL at time t provide the basis for estimates of K = 0.156 yr-' and L„ = 537 mm FL. Length-Frequency Analysis The combined length-frequency distribution for all gindai sampled (Fig. 7) shows that the mean size was 368 mm FL (standard deviation = 48.1 mm). Fish ranged in size from 190 to 490 mm FL and the modal size was 380 mm FL. Thus, l^j„^„ was estimated to be 385 mm FL. There is evidence to show that, above this size, fish were equally vulnerable to the gear (Ralston 1982, unpubl. data), although smaller individuals were almost certainly underrepresented in the catch because of the selective sampling ac- tion of the fish hooks. As f^.i increased from^385 to 485 mm FL, the corresponding value of f, in- creased (Table 4). Due to a sample size of one, estimates of the variance and standard error of the mean could not be calculated when f ^, = 485 mm. Without a statistical weight, the point was excluded from the analysis. The regression of f, on f,, (Fig. 8) was highly significant (P « 0.0001), although there was an increasing lack of fit as t ^^ increased, especially beyond 435 mm FL. This result was due to the diminished statistical weights accorded these points (Table 4). Estimates of the slope and intercept of the regression were 6 = 0.7051 and i = 137.31, with standard errors of 0.0200 and 8.138, respectively. Thus, the mortality to growth ratio (ZIK) is esti- Table 4.— Length-frequency data fitted to the W/etherall et al. ( 1 987) regression model for estimating ZIK and L^ (fork length in mm). Class Relative fork Standard statistical length N ^c, l^ Variance error weight 390 362 385 409.5 284.4 0.438 22.82 400 314 395 415.9 212.1 0.435 22.94 410 301 405 422.1 157.3 0.441 22.65 420 214 415 429.3 110.7 469 21.32 430 165 425 436.3 79.4 0.523 19.10 440 83 435 444.6 62.1 0.705 14 18 450 30 445 453.8 58,3 1.178 849 460 10 455 463.3 78.7 2562 390 470 1 465 480.0 200.0 10.000 1.00 490 1 485 490.0 — — — 8 RALSTON and WILLIAMS: AGEING OF TROPICAL FISHES 400 o c V c V 1 50 200 250 300 350 400 450 500 Fork Length (mm) Figure 7.— Combined length-frequency distribution for all gindai sampled. CO E E en c 0) o Li_ c D 500 380 390 400 410 420 430 440 450 460 470 Full Vulnerability Cutoff Point (mm) Figure 8.-Wetherall et al. (1987) regression of I, on f ,, (see text for further discussion). mated to be 2.39 and L„ = 466 mm FL. Confi- dence intervals for these estimates are P(1.94 < ZIK < 2.62) = 0.95 and P(458 < L„ < 474) = 0.95. With the results presented earlier, it is possible to decompose the ZIK ratio and estimate total mortality rate (Z). Fori^T = 0.234 yr"' (increment microstructure), Z = 0.56 yr', and for ii" = 0.156 yr-^ (annual marks), Z = 0.37 yr-^ Spawning Season Gonadosomatic indexes for male and female gin- dai are summarized by month of capture in Figure 9. The relative size of gindai ovaries was consider- ably greater than the testes. More importantly, there was a distinct seasonal trend in the monthly mean gonadosomatic indexes of females, which FISHERY BULLETIN: VOL. 87. NO. 1 2.5%- 2.0%- ■o 1.5%- E o (0 % 1.0%- « c o a 0.5%- 0.0%-' 57 / 90 J^L--^-^ ^ -^- " -r- ~-r 1 ^ 1 J 1 F 1 M 1 A 1 M 1 J 1 J 1 1 1 1 1 A S O N D Month Figure 9.— The seasonal pattern of variation in mean gonadosomatic indexes of male and female gindai in the Mariana Archipelago. Note that mean index values are bracketed by sample standard deviations with the sample sizes given above. reached a peak in May and diminished as the sum- mer progressed. The same pattern was mirrored in the percent- age of ovaries classified to stages IV- VI (i.e., ad- vanced developing to ripe). During the January- March quarter, only 1.4% of the ovaries sampled were so classified. This statistic rose to 48.3% dur- ing the April-June period, but then dropped to 19.2% in the July- September quarter and to 16.7% over the last quarter of the year (October-Decem- ber). The similarity of these two patterns reinforces the interpretation of reproductive seasonality based on gonadosomatic indexes alone (but see deVlam- ing et al. 1982), and when taken together, these data indicate that peak spawning of gindai in the Mari- ana Archipelago occurs in late May and early June. DISCUSSION Other researchers also have measured the width of daily increments to study fish growth. Methot (1981) used the widths of the outermost three incre- ments in the otoliths of Engraulis mordax and Stenobrachius leucopsarus as a measure of recent somatic growth rate. Brothers and McFarland (1981) measured the thickness of daily increments in newly recruited Haemulon flavolineatum to dis- criminate life history transitions, as did Gutierrez and Morales-Nin (1986) in their study of Dicentrar- chus labrax. Moreover, integration of increment width data to estimate age has been reported, both analytically for Pristipomoides filaTnentosus (Ral- ston and Miyamoto 1981, 1983) and numerically for Merluccius angustimanus, Merluccius sp., Engrau- lis mordax, and Pristipomoides auricilla (Brothers et al. 1976; Methot 1983; Ralston 1985). Experimental work has revealed some of the fac- tors that affect the width of daily growth incre- ments. For example, decreased somatic growth in Oncorhynchus tshawytscha due to reduced temper- ature also results in reduced increment thickness (Neilson and Geen 1985). There is conflicting evi- dence, however, regarding the effect of food ration on daily increment width. Volk et al. (1984) experi- 10 RALSTON and WILLIAMS: AGEING OF TROPICAL FISHES mentally altered somatic growth rates of 0. keta juveniles with different experimental feeding regimes and showed a direct linear effect on the mean width of daily increments. SimOarly, Marshall and Parker (1982) presented data showing an in- crease in the relative size of otoliths of starved 0. keta compared with that of fed controls, even though starvation had no effect on the number of incre- ments. In contrast, Neilson and Geen (1985) found no effect due to ration alone on the thickness of in- crements in fry of 0. tshaunjtscha, although an inter- active effect due to ration level and water temper- ature was shovra. These authors also found that increased feeding frequency significantly reduced mean increment width. Lastly, Campana (1984) found that increments of larval (<10 days old) Porichthys notatus were more irregularly spaced than in juveniles, as were the increments of fish ex- posed to a constant photoperiod environment. From these results, it is apparent that the effect of food ration on the width of daily increments is complex and is at present not well understood. There is still some question concerning how close the coupling is between somatic and otolith growth rates (Brothers 1981; Bradford and Geen 1987). Over the entire lifespan, otolith length and FL typically are highly correlated (Templeman and Squires 1956; Blacker 1974). This situation could not arise were not the growth rates correlated over a similar scale. Still, in the most rigorous examina- tion of the extent of rate coupling to date, Bradford and Geen (1987) found no correlation between the observed growth rates of individual 0. tshawytscha fry and otolith increment widths over relatively short-term (7-15 d) intervals, although a good corre- lation over a 51 d interval was observed. These authors point to the relatively conservative char- acter of otolith growth (Casselman 1983; Gutierrez and Morales-Nin 1986) as the reason for short-term uncouplings between somatic and otolith growth rates. In our study, otolith microstructure typical of daily increments was observed in the sagittae of gindai (Fig. 2). Daily growth increments were previously described and illustrated for congeneric species by Ralston and Miyamoto (1981, 1983), Brouard et al. (1984), and Radtke (1987). Likewise, we observed annual hyaline and opaque zonations, which have been reported in the hard parts (otoliths and verte- brae) of other lutjanids (Loubens 1978; Chen et al. 1984; Edwards 1985; Manooch 1987; Samuel et al. 1987). Still, of the 11 deep slope species {Pristi- pomoides zonatus, P. auricilla, P. filamentosus, P. sieboldii, P.flavipinnis, Aphareus rutilans, Etelis comscans, E. carbunculus, Lutjanus kasmira, Caranx lugubris, and Selar crumenophthalmus) caught during the Marianas survey and whose oto- liths were examined in some detail (Ralston and Williams 1988), only gindai displayed hyaline and opaque zonations, even though all species exhibited microstructure typical of daily growth increments. The absence of annuli in the otoliths of these other species is difficult to explain because many are con- geners, most are confamilials, and all but one (S. crumenophthalmus) occupy the same general deep- water habitat where gindai are found. As a group, these fishes are exposed to virtually identical environmental conditions. Neither is the diet of gin- dai in the Marianas particularly distinctive (Parrish 1987). In contrast to the situation in the Marianas, studies by Loubens (1978) in New Caledonia and Samuel et al. (1987) in the Persian Gulf document distinctive hyaline and opaque annuli in a wide varie- ty of the taxa indigenous to these areas. Although the occurence of annuli in the otoliths of a variety of tropical and subtropical species is now well docu- mented (Manooch 1987), our understanding of when and how they form is quite limited (see below). Von Bertalanffy growth curves were developed for gindai by using both increment microstructure (Fig. 4) and annual marks. Likewise, the L„ param- eter of the von Bertalanffy growth equation was estimated by using the regression method of Wetherall et al. (1987). Moreover, the analysis based on annual markings was tentatively validated with an abbreviated form of marginal increment analysis, wherein the seasonal presence or absence of opaque margins was established for the various pooled ring groups. A preferred approach is to measure the marginal increment for each ring group separately (e.g., Chen et al. 1984; Matheson et al. 1986). Al- though the importance of this type of validation has been overlooked (e.g.. Beamish and McFarlane 1983), it is a very useful technique, especially in situations where capture is fatal. A comparison of von Bertalanffy parameter esti- mates obtained by the three wholly independent ap- proaches (increment microstructure, annuli, and length-frequency analysis) shows reasonable corre- spondence. The two estimates of growth coefficient (K) differed somewhat (0.234 versus 0.156 yr"'), although estimates of L„ were substantially closer (442, 537, and 466 mm FL, respectively). Given that the annual marks were only weakly expressed, these findings support the conclusion that the microstruc- ture observed in gindai otoliths (Fig. 2) results from the daily accretion of increments and, to the extent 11 FISHERY BULLETIN: VOL. 87, NO. 1 that annuli have been validated (Fig. 6), verifies the method of increment widths employed here. Like- wise, the Monte Carlo simulation demonstrated that from an analytical point of view the method is free of significant bias. The results obtained here were also compared to what we know of lutjanid growth by using the growth performance index developed by Munro and Pauly (1983) (see also Pauly and Munro 1983). For a specifically delimited taxon, this index empirical- ly quantifies the well-known inverse correlation be- tween K and L„ (Beverton and Holt 1959; Gushing 1968) and provides a simple basis for predicting K with an estimate of L„. Specifically, Manooch (1987) tabulated the results of growth studies cover- ing 46 snapper and 31 grouper (Epinephelinae) stocks and calculated the combined growth perfor- mance regression for these taxa (r^ = 0.57). With his equation, we predicted K by using each of our three estimates of L„ (see above). These calcula- tions resulted \n K = 0.228, 0.200, and 0.220 yr"' for maximum sizes derived from daily increment microstructure, annuli, and length-frequency anal- ysis, respectively. The estimates compare favorably with the value obtained solely from the study of otolith microstructure (K = 0.234 yr^'). indicating that our results are in close agreement with exist- ing information concerning lutjanid growth. Calculating the age at first annulus formation pro- vides additional evidence that the approach pre- sented here is valid. The data presented in Table 3 indicate that the first annulus occurs at an otolith length of 3,1 17 yxn. An estimate of age at this oto- lith length can be obtained from Table 2 by linear interpolation of the data falling in otolith length in- tervals 6 and 7; i.e., the otolith is 3,000 \ixa at age 0.6 and is 3,500 \im at age 0.8. This calculation in- dicates that the first annulus forms at an age of 0.65 year. Given that the opaque zone forms in January- February (Fig. 6), the predicted birth date by back- calculation is early June, in close agreement with observed spawning activity (Fig. 9). Moreover, the mean monthly sea surface temperature at Tanguis- son Point, Guam, reaches its annual minimum dur- ing January-March (data for the period 1963-72 from Eldredge (1983)), suggesting that temperature fluctuation may be responsible for the formation of the annuli, although this species is found below the thermocline throughout the year (Eldredge 1983) and other closely related sympatric species lack zonations. Some consideration of the underlying assump- tions, advantages, and disadvantages of the method presented here is required. Without doubt, the most important assumption of the approach is that incre- ments are deposited daily throughout the size range where increment width data are gathered. There is a substantial body of literature to show that inter- ruptions to the daily increment record can occur (e.g., Geffen 1982, 1986; McGurk 1984; Jones 1986), especially in larger and older individuals (e.g., Pan- nella 1971; Ralston and Miyamoto 1983). Likewise, we know that with light microscopy the resolution of increments much less than 1.0 \m\ in width is physically impossible (Campana and Neilson 1985). This problem therefore becomes increasingly acute among the largest fish (see Table 2 and Figure 3). Together these findings have led to the view that daily growth increments are of little use in ageing large, old fish (Beamish and McFarlane 1987). In this study, the deposition of daily increments became irregular at otolith lengths in excess of 7,500 /jm. Beyond this length, the increments were also difficult to resolve microscopically due to small size. Consequently, no increment width data were col- lected at otohth lengths >7,500 \im. This corre- sponds to a FL of 329 mm (Fig. 1), which, although of a size that is reproductively competent (S. Ralston, unpubl. data), is smaller than most of the gindai caught during the field surveys (Fig. 7). Thus, the estimated von Bertalanffy curve presented here is largely based on back-calculated data obtained from the younger stages of growth. Nonetheless, we believe that daily increments can be useful in developing growth curves for use in stock assess- ments, even if data representing the older stages are not included in the analysis. This is especially true if the L„ parameter is estimated from length- frequency data (Fig. 8, Wetherall et al. 1987), avoiding the extrapolation problem described by Hirschhorn (1974). Still, validation of the increment periodicity assumption remains an essential compo- nent for future applications of the method. Another assumption implicitly made is that no systematic bias was introduced into the estimation procedure by the manner in which sampling loca- tions were chosen for measuring increment widths. For example, readings were made at specific points along the postrostral growth axis, i.e., where it was possible to distinguish the characteristic bipartite structure of daily increments. However, we also observed broad transition areas lacking in visually conspicuous microstructural features. If these ill- defined regions were elicited by periods of either fast or slow growth, then our estimates of mean otolith growth rate would be biased. To counter this we tried to representatively sample all daily incre- ments (large and small) and we avoided measure- 12 RALSTON and WILLIAMS: AGEING OF TROPICAL FISHES ments beyond 7,500 /^m (see above). Still, we must assume that otolith growth rates calculated from regions where daily increments are visible are other- wise no different from regions where they are not. Numerical integration of otolith growth rates pro- vided a series of ordered pairs of age and otolith length. Otolith lengths were then converted to FL through regression analysis. The only FL data in- cluded in the nonlinear von Bertalanffy regression (Fig. 4), however, were based on otolith lengths in excess of 3,000 fxm. Note that the excluded data (in- tervals 1-6) represent the first year's growth, i.e., the early life history. Although the von Bertalanffy growth equation has historically been the model of choice in stock-assessment applications, including especially the Beverton and Holt (1957) dynamic pool model, it provides a poor description of growth during the early life history. Inflected growth typically characterizes this stage, which is better fit with a Gompertz-type curve (Zweifel and Lasker 1976). By excluding ages <0.8 years from the von Bertalanffy regression analysis, we constrain the data used to estimate the model to the domain over which meaningful predictions are made. Moreover, predictions of FL based on otolith length are also obtained from regression analysis (Fig. 1). Because the smallest otolith used in developing the regres- sion equation was 5,043 ^im (see Figure 1), applica- tion of the equation to predict the FL of a fish whose otolith is less than this size represents an unneces- sary extrapolation of the fitted model. One of the side effects of deleting points from the early life history is to diminish the importance of weighting. Note that the statistical weights of the data used in the regression (Table 2) are very similar (coefficient of variation = 0.78%). Thus, although it may be desirable from a theoretical perspective, weighting had a negligible effect on the parameter estimates. One of the principal advantages recommending this approach is an increase in efficiency and objec- tivity relative to studies that obtain complete counts of daily growth increments (Uchiyama and Struh- saker 1981; Brouard et al. 1984; Radtke 1987). Because all increments need not be visually con- spicuous for a particular preparation to provide useful information, as is true of studies relying on whole counts, the observer can utilize only those por- tions of the otolith where the microstructure is clear- ly expressed. Enumeration of ill-defined increments in poorly developed regions of the otolith is avoided. This feature also makes it possible to automate the procedure (Casselman 1983; McGowen et al. 1987) and ultimately to realize the goal of standardizing age determinations (Boehlert and Yoklavich 1984; Boehlert 1985). Powerful statistical tests of growth heterogeneity also are possible with the acquisition of increment width data (Table 1, Fig. 3). Evaluation of statistical differences in populations with respect to the param- eters of the von Bertalanffy growth equation is cumbersome at best (Gallucci and Quinn 1979; Ber- nard 1981; Kappenman 1981). Analysis of covari- ance of increment width data provides a convenient and widely available means of testing for growth heterogeneity among any statistical populations of interest. One of the principal disadvantages of the method outlined here is that growth variation among indivi- duals within the sampled population is lost through averaging of the data. The final growth curve given in Figure 4 describes the mean growth of the sam- pled population of gindai. Of course, length varia- tion at age is extremely important, and its descrip- tion is required for application of the more powerful and realistic stock-assessment models, especially in cohort or virtual population analysis (Ricker 1975). Nonetheless, given the difficult conditions surround- ing assessment work in tropical environments (Gulland 1982), the application of yield/recruit models is a significant step forward (Munro 1982; Pauly 1982). In conjunction with the analysis of length-frequency distributions, the method proposed here is well suited to help meet that need. ACKNOWLEDGMENTS This work 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, NOAA. Numerous people provided substantive comments as the research progressed and as the paper was variously revised. Some of these suggestions were invaluable to the development of the ideas presented here, and we gratefully acknowledge the efforts of those who spent some time thinking about them. LITERATURE CITED Allen, G. R. 1985. FAO species catalogue. Snappers of the world. An an- notated and illustrated catalogue of lutjanid species known to date. FAO Fish. Synop. (125) 6:1-208. 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. 1987. Current trends in age determination methodology. In R. C. Summerfelt and G. E. Hall (editors). The age and 13 FISHERY BULLETIN: VOL. 87. NO. 1 growth of fish, p. 15-42. Iowa State University Press, Ames, lA. Bernard, D. R. 1981. Multivariate analysis as a means of comparing growth in fish. Can. J. Fish. Aquat. Sci. 38:233-236. 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 19:1-533. 1959. A review of the lifespans and mortality rates of fish in nature, and their relation to growth and other physio- logical characteristics. In G. E. W. Wolstenholme and M. O'Connor (editors). The lifespan of animals, p. 142-177. CIBA Foundation Colloquium Ageing 5. Blacker, R. W. 1974. Recent advances in otolith studies. In F. R. Harden Jones (editor), Sea fisheries research, p. 67-90. John Wiley & Sons, New York, NY. Boehlert, G. W. 1985. Using objective criteria and multiple regression models for age determination in fishes. Fish. Bull., U.S. 83:103- 117. Boehlert, G. W., and M. M. Yoklavich. 1984. Variability in age estimates in Sebastes as a function of methodology, different readers, and different laboratories. Calif. Fish Game 70:210-224. Bradford, M. J., and G. H. Geen. 1987. Size and growth of juvenile chinook salmon back- calculated from otolith growth increments. In R. C. Sum- merfelt and G. E. Hall (editors). The age and growth of fish, p. 453-461. Iowa State University Press, Ames, lA. Brothers, E. B. 1981 . What can otolith microstructure tell us about daily and subdaily events in the early life history of fish? Rapp. P. -v. Rfeun. Cons. int. Explor. Mer 178:393-394. 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 [Haemukm Jlavolineatum (Desmarest), Haemulidae]. Rapp. P.-v. R6un. Cons. int. Explor. Mer 178:369-374. Brouard, F., R. Grandperrin, M. Kulbickl and J. Rivaton. 1984. Note on observations of daily rings on otoliths of deep- water snappers. ICLARM Transl. 3, 8 p. International Center for Living Aquatic Resources Management, Manila, Philippines. Campana, S. E. 1984. Interactive effects of age and environmental modifiers on the production of daily growth increments in otoliths of plainfin midshipman, Porichthys notatus. Fish. Bull, U.S. 82:165-177. Campana, S. E., and J. D. Neilson. 1985. Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci. 42:1014-1032. Casselman, J. M. 1983. Age and growth assessment of fish from their calcified structures— techniques and tools. In E. D. Prince (Con- vener) and L. M. Pulos (editor). Proceedings of the Interna- tional Workshop on Age Determination of Oceanic Pelagic Fishes: Tunas, Billfishes, and Sharks, p. 1-17. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 8. Chen, C.-Y, S.-Y, Yeh, and H.-C. Liu. 1984. Age and growth oiLutjanus malabaricus in the north- western shelf off Australia. Acta Oceanogr. Taiwan. 15: 154-164. Gushing, D. H. 1968. Fisheries biology, a study in population dynamics. University of Wisconsin Press, Madison, WI, 200 p. deVlaming, v., G. Grossman, and F. Chapman. 1982. On the use of the gonosomatic index. Comp. Biochem. Physiol. 73A:31-39. Dunkelberger. D. G., J. M. Dean, and N. Watabe. 1980. The ultrastructure of the otolithic membrane and oto- lith in the juvenile mummichog, Fundulus heteroclitiis. J. Morphol. 163:367-377. Edwards, R. R. C. 1985. Growth rates of Lutjanidae (snappers) in tropical Australian waters. J. Fish Biol. 26:1-4. Eldredge, L. G. 1983. Summary of environmental and fishing information on Guam and the Commonwealth of the Northern Mariana Islands: Historical background, description of the islands, and review of the climate, oceanography, and submarine topography. U.S. Dep. Commer., NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SWFC-40, 181 p. EVERSON, A. R. 1984. Spawning and gonadal matiu'ation of the ehu, Etelis carbunculus, in the Northwestern Hawaiian Islands. In R. W. Grigg and K. Y. Tanoue (editors). Proceedings of the Second Symposium on Resource Investigations in the North- western Hawaiian Islands, Vol. 2, May 25-27, 1983, Univer- sity of Hawaii, Honolulu, Hawaii, p. 128-148. UNIHI- SEAGRANT-MR-84-01. FouRNiER, D. A., and p. A. Breen. 1983. Estimation of abalone mortality rates with growth analysis. Trans. Am. Fish. Soc. 112:403-411. FOURNIER, D. A., and I. J. DOONAN. 1987. A length-based stock assessment method utilizing a generalized delay-difference model. Can. J. Fish. Aquat. Sci. 44:422-437. Gallucci, V. F., AND T. J. QUINN II. 1979. Reparameterizing, fitting, and testing a simple growth model. Trans. Am. Fish. Soc. 108:14-25. Geffen, a. J. 1982. Otolith ring deposition in relation to growth rate in her- ring (Clupea harengvs) and turbot (Scophthalmus nuMimus) larvae. Mar. Biol. (Beri.) 71:317-326. 1986. The growth of herring larvae, Clupea harengus L., in the Clyde: an assessment of the suitability of otolith ageing methods. J. Fish Biol. 28:279-288. GULLAND, J. A. 1982. The management of tropical multispecies fisheries. In D. Pauly and G. I. Murphy (editors). Theory and manage- ment of tropical fisheries, p. 287-298. ICLARM Con- ference Proceedings 9. International Center for Living Aquatic Resources Management, Manila, Philippines and Division of Fisheries Research Commonwealth Scientific and Industrial Research Organisation, Cronulla, Australia. 1987. Summary report. In D. Pauly and G. Morgan (editors). Length-based methods in fishery research, p. 137-146. ICLARM Conference Proceedings 13. International Center for Living Aquatic Resources Management. Manila, Philip- pines. GtrnfiRREZ, E., AND B. Morales-Nin. 1986. Time series analysis of daily growth in Dicentrarchus labrax L. otoliths. J. Exp. Mar. Biol. Ecol. 103:163-179. HiRSCHHORN, G. 1 974 . The effect of different age ranges on estimated Berta- lanffy growth parameters in three fishes and one mollusk of the northeastern Pacific Ocean. In T. B. Bagenal (editor). The ageing of fish, p. 192-199. Unwin Bros., Surrey, 14 RALSTON and WILLIAMS: AGEING OF TROPICAL FISHES England. Jones, C. 1986. Determining age of larval fish with the otolith incre- ment technique. Fish. Bull., U.S. 84:91-103. Jones, R. 1981. The use of length composition data in fish stock assess- ment (with notes on VPA and cohort analysis). FAO Fish. Circ. 734, 55 p. Kappenman, R. F. 1981. A method for growth curve comparisons. Fish. Bull., U.S. 79:95-101. LOUBENS, G. 1978. Biologie de quelques especes de Poisson du lagon nSo- Cal^donien. I. Determination de I'age (otolithomftrie). Cah. O.R.S.T.O.M., S^r. Octenogr. 16(3 and 4):263-283. Manooch, C. S., III. 1987. Age and growth of snappers and groupers. In J. J. Polovina and S. Ralston (editors), Tropical snappers and groupers: Biology and fisheries management, p. 329-373. Westview Press, Boulder. CO. Marshall, S. L., and S. S. Parker. 1982. Pattern identification in the microstructure of sockeye salmon (Onc.orkymkus nerka) otoliths. Can. J. Fish. Aquat. Sci. 39:542-547. Matheson, R. H., Ill, G. R. Huntsman, and C. S. Manooch III. 1986. Age, growth, mortality, food and reproduction of the scamp, Mycteroperca phenax, collected off North Carolina and South Carolina. Bull. Mar. Sci. 38:200-312. McGowAN, M. F., E. D. Prince, and D. W. Lee. 1987. An inexpensive microcomputer-based system for mak- ing rapid and precise counts and measurements of zonations in video displayed skeletal structures of fish. /« R. C. Sum- merfelt and G. E. Hall (editors). The age and growth of fish, p. 385-395. Iowa State Universtiy Press. Ames, lA. McGuRK, M. D. 1984. Ring deposition in the otoliths of larval Pacific herring, Clupea harengus pallasi. Fish. Bull., U.S. 82:113-120. Methot, R. D., Jr. 1981 . Spatial covariation of daily growth rates of larval north- ern anchovy, Engraulis mordajc, and northern lampfish, Stenobrafhius leufopsarus. Rapp. P. -v. RSun. Cons. int. Explor. Mer 178:424-431. 1983. Seasonal variation in survival of larval northern an- chovy, Engraulis mordai, estimated from the age distribu- tion of juveniles. Fish. Bull.. U.S. 81:741-750. Morgan, G. 1987. Incorporating age data in length-based stock assess- ment methods. In D. Pauly and G. Morgan (editors). Length-based methods in fishery research, p. 335-342. ICLARM Conference Proceedings 13. International Center for Living Aquatic Resource Management, Manila, Philip- pines. Min>iRO, J. L. 1982. Estimation of biological and fishery parameters in coral reef fisheries. In D. Pauly and G. I. Murphy (editors), Theory and management of tropical fisheries, p. 71-82, ICLARM Conference Proceedings 9. International Center for Living Aquatic Resources Management, Manila, Philip- pines and Division of Fisheries Research, Commonwealth Scientific and Industrial Research Organisation, Cronulla, Australia. MuNRO, J. L., and D. Pauly. 1983. A simple method for comparing the growth of fishes and invertebrates. Fishbyte (ICLARM) l(l):5-6. Naylor, T. H., J. L. Balintfy, D. S. Burdick, and K. Chu. 1966. Computer simulation techniques. John Wiley & Sons, New York, NY. Neilson, J. D., and G. H. Geen. 1985. Effects of feeding regimes and diel temperature cycles on otolith increment formation in juvenile chinook salmon, Oncorhyrwhus tshaurytscha. Fish. Bull., U.S. 83:91-101. Pannella, G. 1971. Fish otoliths; daily growth layers and periodical pat- terns. Science (Wash., DC) 173:1124-1127. Parrish, J. D. 1987. The trophic biology of snappers and groupers, /n J. J. Polovina and S. Ralston (editors). Tropical snappers and groupers: Biology and fisheries management, p. 405-463. Westview Press, Boulder CO. 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, p. 33-70. ICLARM Conference Proceedings 9. International Center for Living Aquatic Resources Management, Manila, Philip- pines and Division of Fisheries Research, Commonwealth Scientific and Industrial Research Organisation, Cronulla, Australia. 1987a. Application of information on age and growth of fish to fishery management. In R. C. Summerfelt and G. E. Hall (editors). The age and growth of fish, p. 495-506. Iowa State University Press. Ames. lA. 1987b. A review of the ELEFAN system for analysis of length-frequency data in fish and aquatic invertebrates. In D. Pauly and G. Morgan (editors). Length-based methods in fishery research, p. 7-34. ICLARM Conference Proceed- ings 13. International Center for Living Aquatic Resource Management, Manila, Philippines. Pauly, D., and J. L. Munro. 1983. Once more on the comparison of growth in fish and invertebrates. Fishbyte aCLARM) 2(1):21. Polovina, J. J. 1985. Variation in catch rates and species composition in handline catches of deepwater snappers and groupers in the Mariana Archipelago. Proc. Fifth Int. Coral Reef Congr., Tahiti 5:515-520. 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. Polovina, J. J., and S. Ralston. 1986. An approach to yield assessment for unexploited resources with application to the deep slope fishes of the Marianas. Fish. Bull., U.S. 84:759-770. Radtke, R. L. 1987. Age and growth information available from the otoUths of the Hawaiian snapper, Pristipomoides filamentosus. Coral Reefs 6:19-25. Ralston, S. 1982. Influence of hook size in the Hawaiian deep-sea handline fishery. Can. J. Fish. Aquat. Sci. 39:1297-1302. 1985. A novel approach to aging tropical fish. ICLARM Newsl. 8(1):14-15. International Center for Living Aquatic Resources Management, Manila, Philippines. In press a. Effect of seasonal recruitment on bias of the Beverton-Holt length-based mortality estimator. Am. Fish. Soc. Symp. Ser. In press b. Length-weight regressions and condition indices of lutjanids and other deep slope fishes from the Mariana Archipelago. Micronesica. Ralston, S., and G. T. Miyamoto. 1981. Estimation of the age of a tropical reef fish using the density of daily growth increments. In E. D. Gomez, C. E. 15 FISHERY BULLETIN; VOL. 87, NO. 1 Birkeland, R. W. Buddemeier, R. E. Johannes, J. A. Marsh, Jr., and R. T. Tsuda (editors), The reef and man, p. 83-88. Proceedings of the Fourth International Coral Reef Sym- posium Vol. 1. Marine Science Center, University of the Philippines, Quezon City, Philippines. 1983. Analyzing the width of daily otolith increments to age the Hawaiian snapper, Pristipomoidesfiiamentostts. Fish. Bull., U.S. 81:523-535. Ralston, S., and H. A. Williams. 1988. Depth distributions, growth, and mortality of deep slope fishes from the Mariana Archipelago. U.S. Dep. Com- mer., NOAA Tech. Memo. NMFS, NOAA-TM-NMFS- SWFC-113, 47 p. RiCKER, W. E. 1975. Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Board Can. 191, 382 p. 1979. Growth rates and models. In W. S. Hoar, D. J. Ran- dall, and J. R. Brett (editors). Fish physiology, Vol. VIII, Bioenergetics and growth, p. 677-743. Academic Press, New York, NY. SAS Institute Inc. 1979. SAS user's guide. SAS Institute Inc., Raleigh, NC, 494 p. Samuel, M., C. P. Mathews, and A. S. Bawazeer. 1987. Age and validation of age from otoliths for warm water fishes from the Arabian Gulf. In R. C. Summerfelt and G. E. Hall (editors). The age and growth offish, p. 253-265. Iowa State University Press, Ames, lA. SCHNUTE, J. 1987. A general fishery model for a size-structured fish popu- lation. Can. J. Fish. Aquat. Sci. 44:924-940. SCHNUTE, J., and D. FOURNIER. 1980. A new approach to length-frequency analysis: Growth structure. Can. J. Fish. Aquat. Sci. 37:1337-1351. 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. Templeman, W.. and H. J. Squires. 1956. Relationship of otolith lengths and weights in the had- dock Melanogrammus aeglefinus (L.) to the rate of growth of the fish. J. Fish. Res. Board Can. 13:467-487. UCHIYAMA, J. H., and P. STRUHSAKER. 1981 . Age and growth of skipjack tuna, Katsuwonus pelamis, and yellowfin tuna, Thunnus alha^ares, as indicated by daily growth increments of sagittae. Fish. Bull., U.S. 79:151- 162. 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 heterociittis (Linnaeus) and Tilapia nilotica (Linnaeus). J. Exp. Mar. Biol. Ecol. 58:127-134. Wetherall, J. A., J. J. Polovina, and S. Ralston. 1987. 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, p. 53-74. ICLARM Conference Proceedings 13. Inter- national Center for Living Aquatic Resource Management, Manila, Philippines. ZWEIFEL, J. R., AND R. LaSKER. 1976. Prehatch and posthatch growth of fishes— a general model. Fish. Bull., U.S. 74:609-621. 16 AGE AND GROWTH OF RED DRUM, SCIAENOPS OCELLATUS, FROM OFFSHORE WATERS OF THE NORTHERN GULF OF MEXICO' Daniel W. Beckman, Charles A. Wilson, and A. Louise Stanley^ ABSTRACT Otolith (sagitta) sections are used to accurately age red drum, Sciaejiops ocellatus. from the offshore northern Gulf of Mexico. Marginal increment analysis indicated that annuli were formed during winter and spring months. Ages of offshore schooling red drum ranged from 1 to 37 years. Age distributions indicated variability in relative abundances of year classes, with the majority of fish sampled being over 10 years of age. Male and female age distributions did not differ significantly. Growth differed significantly between males and females. The von Bertalanffy growth equation for males was L, = 909(1 - e ' "), and for females was L, = 1,013(1 - e"""**"*"'^'), where t is age (years) and L, is fork length (mm). The red drum, Sciaenops ocellatus, is a large sciae- nid that inhabits temperate and subtropical near- shore and estuarine waters from Massachusetts to northern Mexico. Juveniles are most abundant in estuarine waters and move from estuarine to near- shore waters as they near maturity (Pearson 1929). The primary spawning stock in the Gulf of Mexico is thought to spawn in nearshore open waters (Over- street 1983). The red drum is one of the most popular recrea- tional and commercial fish species in the northern Gulf of Mexico. Recent increase in demand for red drum has escalated the controversy concerning its management; however, little has been reported con- cerning its growth and population structure. Age and growth-rate estimates of red drum have only used immature fish from inshore estuarine waters. Pearson (1929) and Wakeman and Ramsey (1985) identified modes in length-frequency distribu- tions and performed scale analysis to determine age estimates. However, Wakeman and Ramsey (1985) reported that scale annuli were unsatisfactory for accurately estimating the age of red drum. Theil- ing and Loyacano (1976) reported age estimates of red drum from a South Carolina salt marsh im- poundment based on otolith examination. Growth rates of juveniles were reported by Roessler (1970), Bass and Avault (1975), and Simmons and Breuer (1962). 'Publication 88-06 of the Coastal Fisheries Institute, Louisiana State University, Baton Rouge, LA 70803-7503. ^Coastal Fisheries Institute. Center for Wetland Resources, Louisiana State University, Baton Rouge, LA 70803-7503. No age or growth rate estimates have been pub- lished for adult red drum from offshore waters. Ac- curate information on the age and growth of adult red drum is necessary for determining population dynamics and monitoring the population's response to fishing pressure. Due to the reduction in growth rate in larger individuals, which leads to size over- lap between age classes, age estimation by cohort analysis is not feasible. Otolith sections have provided valid age estimates for many large, long-lived fish species (Beamish and McFarlane 1987). The purposes of this study were to determine if otoliths (sagittae) could be used to obtain valid age estimates for red drum and to estimate grow^th rates and determine the age structure of the oceanic schooling population of red drum. MATERIALS AND METHODS Red drum (1,726 fish) were collected in Texas, Louisiana, Mississippi, and Alabama offshore coastal waters of the northern Gulf of Mexico from Sep- tember 1985 through October 1987 by purse seine (N = 1,428 from 67 sets) (Fig. 1), gill net (N = 134 from 9 sets), and hook and line (N = 164 from 12 dates). Samples captured by unknown gear from February 1985 through June 1987 (AT = 96) were included for marginal increment analysis only. After fish were randomly sampled from landings, they were measured (fork length) and weighed, and their sex was determined. Sex identifications were Manuscript accepted August 1988. FISHERY BULLETIN, U.S. 87:17-28. 17 FISHERY BULLETIN: VOL. 87, NO. 1 AUG N'E5 -29°N AP.IMAY N=62 ... <... I UAD.J GULF OF MEXICO JUNJUL m^fif^ OCT AUG N-34 N-40 FEBOCT N'413 •1986 1987 95°W _i 90'W _i Figure 1.— Purse seine sampling locations in the northern Gulf of Mexico. Points represent individual purse seine sets. N's refer to the total numbers of red drum sampled from sets indicated. Precise locations were not available for 6 sets {N = 161). unavailable for 182 individuals. Sagittae were re- moved, cleaned, and stored dry for later process- ing. Length-weight regressions were fit to the data using the model: weight = a FL*, where weight = body weight (g) and FL = fork length (mm). Re- gressions for male and female red drum were com- pared using analysis of covariance (Ott 1977). A Komolgorov-Smirnov two-sample test (Tate and Clelland 1957) was used to detect possible sampling bias by comparison of length-frequency distributions of fish caught by different sampling gears. Otoliths were processed for age analysis by em- bedding them in an epoxy resin (Spurr 1969) and sectioning transversely (0.7 mm thick) through the core of the left sagitta (or the right when the left sagitta was not available), using a Buehler Isomet^ low-speed saw. Sections were mounted on glass slides with thermoplastic cement (Crystalbond 509 adhesive), sanded on 600 grit wet sandpaper to remove saw marks, polished with alumina micro- polish (0.3 ixm), and then examined with a compound microscope (transmitted light at 40 x magnification). Opaque zones (annuli) were counted in sections from the core to the margin in the medial direction. Ap- pearance of the margin was recorded as either opaque or translucent. If the left sagitta was un- readable, the right sagitta, if available, was prepared and examined. Validation of age estimates was ac- complished and the timing of annulus formation determined by plotting percent occurrence of oto- liths with opaque margins by month. 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. Each otolith was aged by two readers, and the resulting age estimates were compared. The coef- ficient of variation was calculated for age estimates in order to test the reproducibility of age estimates independent of magnitude (Sokal and Rohlf 1981; Chang 1982). If readers' initial age estimates for an otolith did not agree, the section was reread. If the resulting age estimates did not agree, the fish's other sagitta was prepared and read. If the readers did not reach agreement on an age or sections from both otoliths were unreadable, the data for that fish were not used in analyses. All ageing was done without knowledge of the sample source or any previous age estimates. Year-of-birth was back-calculated from age esti- mates by subtracting estimated age from the year of capture and assuming that the first annulus formed in winter of year 2 (Beckman et al. in press). Age-frequency distributions were compared using a Komolgorov-Smirnov two-sample test (Tate and Clelland 1957). Von Bertalanffy (1938, 1957) growth curves were fit separately for males and females by nonlinear regression. The growth equation for length was L, = L„ [1 - e-*^*'-'oi] and for weight was W, = W^ [1 - e-K((-*o)]3, where L, and VF, are the esti- mated length and weight, L„ and W^ are the asymptotic length and weight, K is the growth coef- ficient, t is the age (years), and tf, is the hypothe- tical age when length or weight would be zero. A full model, in which separate parameters were fit for males and females, was compared with a reduced model in which sex was not considered. An F-test (Ott 1977) was used to test for differences in the models. 18 BECKMAN ET AL.: AGE AND GROWTH OF RED DRUM RESULTS Length-weight regressions for males and females were not significantly different (P = 0.842 for inter- cepts, P = 0.605 for slopes). The combined length- weight regression was Weight = 2.9 x 10-« FL^-^z N = 1,626. r2 = 0.91 The length-frequency distributions of red drum collected by purse seine (Fig. 2) were significantly different from those obtained by gill net (P < 0.01) and hook and line (P < 0.01). Therefore, to avoid gear selectivity bias, only purse seine samples were assumed to represent the age-frequency distribution of the offshore spawning population. Because the sagittae were extremely thick and opaque, they needed to be sectioned before they could be aged. Distinct opaque and translucent growth zones were observed in transverse sections. Annuli were most distinct and the most consistent growth patterns were observed in the region from the core to the proximal surface of the sagitta along the ventral margin of the sulcus acousticus. All counts were made in this region (Fig. 3). The percentage of sagittae with opaque margins was plotted by month to determine the timing of an- nulus formation. Opaque zones were deposited in the sagittae during winter and spring months in three successive years of sampling (Fig. 4A). As a consis- tent pattern of annulus formation was exhibited each year, data were combined for all years in order to compare annulus formation between size groups (Fig. 4B). Data were grouped according to matur- ity (Overstreet 1983) and growth patterns. Group- ings were chosen to include an adequate sample size within each group for analyses as follows: 0-4 an- nuli — immature and early maturity, rapid growth; 5-9 annuli — mature, rapid growth; 10-19 annuli — mature, reduced growth; and 20-36 annuli — maximum ages, reduced growth. A single peak per year in all plots indicates that one annulus was formed each year in all groups. Age in years for red drum was equal to the number of annuli observed in sections of sagittae. Age estimates were obtained o CNI ^ to CO o CN ^ CD CTl cn 0.20). Age distributions were grouped by year of capture (October through Sep- tember for 1985-86 and 1986-87) and compared (Fig. 5). Sufficient samples were not available for 1984-85 for comparisons. The 11-14 year age classes dominated the 1985-86 samples and 12-15 year old fish dominated in 1986-87. There was an apparent coherence between the age-frequency dis- tributions for the two sample years. Anomalies in the age distribution for 1985-86 lagged one year behind corresponding anomalies for 1986-87. Age distributions differed significantly between the two sample years (P < 0.01); however, there were no significant differences between year of birth distri- butions between sample years (P > 0.20). Therefore, samples were combined for all years to obtain year- 20 BECKMAN ET AL.; AGE AND GROWTH OF RED DRUM "T 1 r — -r 1 1 1 1 1 1 1 r— — 1 1 1 — -r 1 1— O »— CSI O -^ CM 1985 1986 1987 MONTH CO z o cr < LU O < Q. O I liJ o IT LU Q. JAN FEB MAR APR MAY JUN JUL AUG SEP OCT MONTH NOV DEC Figure 4.— Plot of percent occurrence of otoliths (sagittae) with opaque margins vs. month of capture for red drum A) by sample month and year and B) grouped by annulus counts, sample years combined. Sample size is indicated next to points. 21 FISHERY BULLETIN: VOL. 87, NO. 1 90 80 70 >-60 o gso D O40 LU CC U_30 20 50 40 O30 m ID o m2o U- 0- 1985-86 N=788 1986-87 N=540 AGE (years) Kzsz) UNKNOWN ekxs MALE bxxs FEMALE 22 BECKMAN ET AL.: AGE AND GROWTH OF RED DRUM Figure 5.— Age-frequency distributions for red drum captured October 1985-September 1986 and October 1986-September 1987 by purse seine from offshore northern Gulf of Mexico waters. and by weight: males: W, = 10,548(1 - e-«"'^('*8.69))3_ females: W, = 15,207(1 - e-o.o79«.ii.57))3 of-birth distributions (Fig. 6). Variability in year- class success is suggested by differences in relative numbers of individuals between year classes. The separation of sexes in growth models resulted in a significantly better fit by weight (P < 0.001) and length (P < 0.001) when compared with models in which sexes were combined. Separate von Berta- lanffy growth curves best described changes in length (Fig. 7A) and weight (Fig. 7B) of red drum. Equations by length were males: L, = 909(1 - e-o.i37((..7.74)) females: L, = 1,013(1 - e-o.o88<(.ii.29)) DISCUSSION Sampling Comparison of length-frequency distribution be- tween gear types demonstrated that gill net and hook and line were different from purse seine collections. Therefore, to provide a basis for docu- menting and comparing age structure in the off- shore schooling population only purse seine collec- tions were used. We assumed that purse seine samples would result in the smallest size selection bias (Nielson and Johnson 1983). We assumed that temporal and spatial bias was minimized because sets were made throughout the year and cvjrO'j-LOaar^ooaiO'— cNfO-^iniijr^aoaiO'— CNfO-^intor^QOoiO'-CNKi-* YEAR-OF-BIRTH csED UNKNOWN MALE FEMALE Figure 6.— Year-of-birth frequency distributions for red drum captured by purse seine from offshore northern Gulf of Mex- ico waters, September 1985-October 1987. 23 FISHERY BULLETIN: VOL. 87, NO. 1 O u < u. IJJ < PL s: I 1 3 ^^ i ^T 1 2 ^l 21 h ■' ^ ^ ^ f i 3 " +^ :i :i 2 ^ ^ ^ t - "^ 1=^ ^:^J =5 '^u. ^h:^ " 2 i^L^^ E «5\1^3 < 2 ^^!*^^^t- i 2 2 ll. s«-i B ii ti j.i. •y. 1 1 1 ^^ It ^i a s^ :f 2- 2 ^^ i#"- U. U. T|X*^flP^H «¥%^ ^ >- ^ ij^ ^^g^ 3 a V^H^ ^ " '^"^ffl^ ^3 '-'*- ^^ ™3 Lili. Lil^ i^g 2 1* ^ ^ ^2. [D ^QBhl 3 \ u. ^ \ j^ ^^ 2 '<^ u. 2 2 L. [D o CO O LO ID CNJ ^- a^ 00 r^ vD LD (lULU) H10N31 >|) 1H0I3M ■6 ^ CO la 1 LO ^ N ,„i^ c CO 1— u CO c 0) (D > 73 25 FISHERY BULLETIN: VOL. 87, NO. 1 across the coastline of the north-central Gulf of Mexico. Validation Periodicity of formation of ageing structures must be confirmed over all year classes to validate the use of that hardpart for ageing (Beamish and McFarlane 1983). Beckman et al. (in press) validated that the first two annuli were formed yearly in sagittae of immature red drum from estuarine waters. The use of marginal increment analysis in this study valid- ated that annuli continued to be deposited in red drum sagittae once per year in fish up to 37 years old. There was no significant variability in timing of annulus formation with stage of maturity or with change in growth rates with age. Precise, reproducible age estimates were obtained for red drum using transverse sections of sagittae. Almost 100% agreement between two readers was achieved by recounting otoliths or counting the fish's other sagitta when age estimates disagreed. Initial disagreements were usually resolved by recounting the otolith, suggesting initial miscounts or errors were due to recording and transcription. Unread- able otoliths were primarily those with inadequate sample preparation. Discarding difficult-to-age oto- liths, which are often from older fish, could bias age distributions as well as von Bertalanffy growth parameters (Hirschhorn 1974). Recounting otoliths for which age estimates did not initially agree and utilizing both sagittae to obtain a readable sample allowed us to minimize the number of unused sections. The same seasonal pattern of annulus formation reported in this study was observed in sagittae of red drum in inshore estuaries (Beckman et al. in press). This pattern is also similar to that observed in another sciaenid, the Atlantic croaker (Barger 1985). The formation of an opaque zone in red drum sagittae in winter and spring months may corre- spond to reduced growth rate during this period (Doerzbacher et al. 1988). In West African sciaenids an opaque zone was formed apparently in response to cold temperatures (Poinsard and Troadec 1966). Growth The von Bertalanffy growrth coefficients for other sciaenids (e.g., Barger 1985; Wakeman and Ramsey 1985, cited by Pauly 1980) were generally greater than those obtained for red drum in this study. Growth parameters reported herein differ from those obtained by Wakeman and Ramsey (1985) for red drum; however, their model was based only on young fish from inshore waters that have higher growth rates (Beckman et al. in press). The growth models reported in this study were derived primar- ily from mature slower growing fish. The negative values of t^ predicted suggests that our models do not adequately describe growth of young fish un- represented in our data. Separate models may be necessary to describe growth of immature red drum from inshore waters (Richard Condrey pers. com- mun.''). The large variation in size at age beyond year 5 makes it impossible to precisely predict age of red drum using length or weight. Our estimates of maximum red drum age are greater than those previously suggested. Pearson (1929), Simmons and Breuer (1962), and Wakeman and Ramsey (1985) used the scale method and re- ported a maximum age of 5, 3, and 4 years, respec- tively. The use of validated ageing techniques for red drum from otoliths more accurately estimates their ages and provides much improved manage- ment data bases. Female red drum attained significantly larger sizes than did males, with growth curves diverging with increasing age and maturity. Larger size in females has been postulated as a life history strategy in fish for increasing reproductive potential through increased egg production capability (Roff 1983). The similarities in age-class compositions between sexes indicated that the increased female size was attained through somewhat higher growth rates and not greater longevity. Age Structure Examination of the age composition of the off- shore population revealed that red drum begin to appear in the offshore population as early as year 2. Their appearance offshore coincides with their absence inshore by four or five years of age (Pear- son 1929; Simmons and Breuer 1962; Wakeman and Ramsey 1985). The 1973 year class was the most abundant, and earlier year classes demonstrated a decay pattern indicative of natural mortality. The year classes since 1973 were variable and could be interpreted variously to indicate several poor year classes, high mortality, or incomplete recruitment to offshore schooling populations, assuming no bias in the sampling procedures. Inadequate data are available to determine which are primary factors af- fecting age distributions. 'Richard Condry, Coastal Fisheries Institute. Louisiana State University, Baton Rouge, LA 70803, pers. commun. January 1988. 26 BECKMAN ET AL.: AGE AND GROWTH OF RED DRUM Comparison of age distributions between years provided two estimates of the population age-class structure, varying in time and areas sampled. The similarities in year-of-birth distributions in 1985-86 and 1986-87 suggest that the same population was sampled in both years and that distributions may reflect the true offshore schooling population of red drum, assuming no sampling selectivity. Recruit- ment into the population from one year to the next was evident only in the youngest age classes, possibly due to migration from inshore nursery areas. The relatively low numbers of individuals in age classes of less than 10 or 11 years suggests a possible delay or reduction in recruitment into the schooling population sampled. Other possible factors affecting abundance of younger age classes offshore are fishing pressure on inshore red drum, size specific fishing offshore, or other factors affecting survival. ACKNOWLEDGMENTS Sampling efforts were supported by the Louisiana Sea Grant College Program; the U.S. Department of Commerce, NOAA, Marine Fisheries Initiative Program (MARFIN); the Louisiana State Univer- sity, Coastal Fisheries Institute; the Louisiana Department of Wildlife and Fisheries; and the Na- tional Marine Fisheries Service. Sampling involved the cooperation of many recreational and commer- cial fishermen. We thank Gary Fitzhugh, Bruce Thompson, David Nieland, Robby Parker, David Stanley, and Tony Gaspard for their assistance in data collection and analysis, and James Geaghan for statistical assistance. LITERATURE CITED Barger, L. E. 1985. Age and growth of Atlantic croakers in the northern Gulf of Mexico, based on otolith sections. Trans. Am. Fish. Soc. 114:847-850. Bass, R. J., and J. W. Avault, Jr. 1975. Food habits, length-weight condition factor, and growth of juvenile red drum, Sciaenops ocellata, in Loui- siana. Trans. Am. Fish. Soc. 104:35-45. 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. 1987. Current trends in age determination methodology. In R. C. Summerfelt and G. E. Hall (editors). Age and growth of fish, p. 15-42. Iowa State Univ. Press. Beckman, D. W., G. R. Fitzhugh, and C. A. Wilson. In press. Growth rates and validation of age estimates of red drum, Sciaenops ocellatus, in a Louisiana salt marsh im- poundment. Contr. Mar. Sci. Chang, W. B. 1982. A statistical method for evaluating the reproducibil- ity of age determination. Can. J. Fish. Aquat. Sci. 39:1208- 1210. Ditty, J. G. 1986. Ichthyoplankton in neritic waters of the northern Gulf of Mexico off Louisiana: composition, relative abundance, and seasonality. Fish. Bull., U.S. 84:935-946. DOERZBACHER, J. W., A. W. GREEN, G. C. MaTLOCK, AND H. R. OSBORN. 1988. A temperature compensated von Bertalanffy growth model for tagged red drum and black drum in Texas bays. Fish. Res. 6:135-152. HiRSCHHORN, G. 1974. Effect of different age ranges on von Bertalanffy parameters in three fishes and one moUusk of the north- eastern Pacific Ocean. In T. Bagenal (editor). The aging of fish, p. 192-199. Unwin Brothers. Ltd., London. Nielsen, L. A., and D. L. Johnson. 1983. Fisheries techniques. American Fisheries Society, Bethesda, MD. 468 p. Ott, L. 1977. An introduction to statistical methods and data anal- ysis. Duxbury Press, North Scituate. MA. 730 p. OVERSTREET, R. M. 1983. Aspects of the biology of the red drum, Sciaenops ocellatiis, in Mississippi. Gulf Res. Rep. suppl. 1:45-68. Pauly, D. 1980. On the interrelationships between mortality, growth parameters and mean environmental temperature in 175 fish stocks. J. Cons. Int. Explor. Mer 39:175-192. Pearson, J. C. 1929. Natural history and conservation of redfish and other commercial sciaenids on the Texas coast. Bull. U.S. Bur. Fish. 4:129-214. Poinsard, F., and J. P. Troadec. 1966. Determination de I'age par la lecture des otolithes chez deux especes de Sciaenides Ouest-Africains. J. Cons. Int. Explor. Mer 30:291-307. ROESSLER, M. A. 1970. Checklist of fishes in Buttonwood Canal, Everglades National Park, Florida, and observations on the seasonal oc- currence and life histories of selected species. Bull. Mar. Sci. 20:860-893. ROFF, D. A. 1983. An allocation model of growth and reproduction in fish. Can. J. Fish. Aquat. Sci. 40:1395-1404. Simmons, E. G., and J. P. Breuer. 1962. A study of redfish, Sciaenops ocellata Linnaeus and black drum, Pogonias cromis Linnaeus. Contrib. Mar. Sci. 8:184-211. SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry. Freeman, San Francisco, p. 58-60. Spurr, A. R. 1969. A low-viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res. 26:31-43. Tate. M. W., and R. C. Clelland. 1957. Non-parametric and shortcut statistics in the social, biological, and medical sciences. Interstate Printers and Publishers, Inc., Danville, IL, p. 93-94. Theiling, D. L., and H. a. Loyacano, Jr. 1976. Age and growth of red drum from a saltwater marsh impoundment in South Carolina. Trans. Am. Fish. Soc. 105:41-44. VON Bertalanffy, L. 1938. A quantitative theory of organic growth. II. Inquiries 27 FISHERY BULLETIN: VOL. 87. NO. 1 on growth laws. Human Biol. 10:181-213. Wakeman, J. M., and P. R. Ramsey. 1957. Quantitative laws in metabolism and growth. Q. Rev. 1985. A survey of population characteristics for red drum and Biol. 32:217-231. spotted sea trout in Louisiana. Gulf Res. Rep. 8:1-8. 28 DISTRIBUTION, ADVECTION, AND GROWTH OF LARVAE OF THE SOUTHERN TEMPERATE GADOID, MACRURONUS NOVAEZELANDIAE (TELEOSTEI: MERLUCCIIDAE), IN AUSTRALIAN COASTAL WATERS R. E. Thresher,' B. D. Bruce,^ d. M. Furlani,' and J. S. GUNN' ABSTRACT Ichthyoplankton surveys in southern Australian coastal waters indicate that larvae of the temperate gadoid, Macruronus novaezelandiae, differed consistently in mean size and age between sample sites. These observations are consistent with the hypothesis that larvae are being passively advected by longshore currents from a spawning area on the west coast of Tasmania to habitats along the southeastern and eastern coasts. The ages of larvae at specific points along the advection route vary, which suggests there is considerable variation in rate of larval transport. Rates of larval growth increased exponentially for at least the first 50 days of planktonic life, though the slope of the growth curve varies both between years and between seasons. Growth rates also differ between sampling sites: early stage larvae (<15 d postfirst-feeding) grew more rapidly at sites close to the spawning area, whereas older larvae {>25 d postfirst-feeding) grew more rapidly the farther they were from the spawning area. Migration of M. novaezelandiae to a specific spawning area and the subsequent transport of larvae away from this area appears to be an adaptive response by the population to, on the one hand, regional differences in condi- tions for larval growth and, on the other, changing needs of the larvae at different stages of their development. Planktonic eggs and larvae of marine fishes are sub- ject to dispersion (= diffusion) and advection ( = transport or drift), topics of considerable theoretical and empirical interest to larval fish ecologists (Smith 1973; Wiedemann 1973; Talbot 1977; Okubo 1980; Naganuma 1982; Power 1986). The causes and con- sequences of diffusion, aggregation and patchiness of larvae are largely unknown due to problems of sampling at an appropriate scale (Hewitt 1981). Advection of larval fishes, however, has been fre- quently documented and has been studied in some detail (see Norcross and Shaw 1984). Temporal variability in advection can have considerable im- pact on rates of larval survival (Norcross and Shaw 1984) and has long been suggested to be a major determinant of year-class strength in populations subject to variable current regimes (Walford 1938; Harden Jones 1968; Nelson et al. 1977; Bailey 1981; Parrish et al. 1981). In at least some species, eggs ■CSIRO Marine Laboratories, GPO Box 1538, Hobart. Tasmania 7001, Australia. =CSIRO Marine Laboratories, GPO Box 1538, Hobart, Tasmania 7001, Australia; present address: South Australian Department of Fisheries, GPO Box 1625, Adelaide, South Australia 5001, Australia. 'CSIRO Marine Laboratories, GPO Box 1538, Hobart, Tasmania 7001, Australia; present address: Tasmania Department of Sea Fisheries, Crayfish Point, Taroona, Tasmania 7006, Australia. and larvae are placed in currents that transport them to larval and juvenile nursery areas (Parrish et al. 1981). Even within species, however, the ex- tent of adult migration and larval countermigration varies widely between populations, presumably in response to local hydrographic conditions (Gushing 1986). Eastern North Atlantic gadoid stocks, for ex- ample, provide some of the classic examples of adult migration to spawning grounds and subsequent passive drift of larvae to nursery areas (Harden Jones 1968); in contrast, larvae of at least some Western Atlantic stocks of the same species develop entirely in the immediate vicinity of the spawning grounds (O'Boyle et al. 1984, Sherman et al. 1984; Smith and Morse 1985). By comparison with their Northern Hemisphere relatives, little is known about the reproduction and larval ecology of southern temperate gadoids, despite the fact that several constitute major fish- eries. One species, the blue grenadier or hoki, Macruronus novaezelandiae, supports such a fishery in Australia and New Zealand, with combined an- nual landings of approximately 100,000 t (tonnes). Available data indicate that both the New Zealand and Australian populations migrate each winter to discrete spawning areas, located, respectively, on the west coasts of the New Zealand South Island Manuscript accepted; August 1988. FISHERY BULLETIN, U.S. 87:29-48. 29 FISHERY BULLETIN: VOL. 87, NO. 1 (Bladodyorov and Nosov 1978^; Patchell 1982; Kuo and Tanaka 1984a, b) and Tasmania (Wilson 1981, 1982). These migrations imply a countermigration by either larvae or juveniles back to adult habitats (e.g.. Harden Jones 1968; McKeown 1984). Patchell (1982) reported movement of eggs away from spawning areas in New Zealand, and subsequently collected juveniles in coastal habitats hundreds of km from the spawning area. Similarly small juvenile M. novaezelandiae have been collected in estuaries and on the coastal shelf along the southeastern and eastern coasts of Tasmania (Wilson 1981, 1982; Last et al. 1983; Bulman and Blaber 1986), over 200 km from the known spawning area. 'Bladodyorov, A. I., and E. V. Nosov. 1978. The biological basis of rational exploitation of Macruronus novaezelandiae. Unpubl. TINRO manuscr. English translation held by New Zea- land Ministry of Agriculture and Fisheries, Fisheries Research Division Library, 7 p. How juveniles move between the spawning area and these coastal habitats, or even whether this is a rare or common occurrence in the species is un- known. The present study investigated the distribu- tion, sizes, and ages of larval M. novaezelandiae, on the basis of which patterns of advection, larval growth, and the relationship between the two could be inferred. METHODS Sampling Procedures Ichthyoplankton samples were collected at ap- proximately two monthly intervals from April 1984 to September 1985. Samples were obtained at fixed stations along nine transects located roughly equi- distantly aroimd Tasmania (Fig. 1). Additional sam- ples were obtained in July and August 1985 along 39°S 40' 41' 42' 43' 44°- 143°E 144° 145° 146° 147° 148° 149° 150° Figure 1.— Location of ichthyoplankton sampling sites (solid circles) and release points (X's) of drift cards. The drogued buoy was released at the release point for drift cards south of the spawning area. Cross-hatched area indicates apparent principal spawning area of Macruronus novaezelandiae in Australian coastal waters. 30 THRESHER ET AL.: LARVAE OF GADOID. MACRURONUS NOVAEZELANDIAE the southern coast of mainland Australia (see Figure 3). Transects 1 through 8 consisted of 2-4 stations (4 on average), depending upon the vddth of the con- tinental shelf. These stations were designated "near- shore" (at a depth of 30-50 m), "midshelf" (70-100 m), "shelf edge" (immediately offshore of the shelf break and usually at a bottom depth of approximate- ly 200 m) and "offshore" (1 nmi offshore of the sur- face temperature/salinity front between inshore and offshore water masses or, if no front was evident, at 10 nmi offshore from the shelf edge station). In the second year of the study, occasional samples were collected at sites along the west coast between regular transect lines, in order to improve the spa- tial resolution of analyses and to increase sample sizes. Two samplers were used: a rectangular midwater trawl (RMT) 1 + 8 (see Baker et al. 1973 for de- scription) and aim diameter ring net fitted with a pivoting bridle system similar to the Tranter- George plankton net (Tranter and George 1972). Mesh sizes for the RMT-8 was 3 mm and 1 mm for the net and cod end, respectively, and 333 pm throughout for the RMT-1. The ring net consisted of 500 \im mesh with a 333 \^m cod end. Initially, all sampling was done with the RMT 1 -i- 8 in a fixed open mode. Because it was difficult to fish the net in rough seas and to calibrate its fishing character- istics (see Pommeranz et al. 1982), the RMT 1 -t- 8 system was replaced after three cruises (April- August 1984) with the more manageable ring net. The ring net was subsequently used on transects 1 through 8, while the RMT system was retained for study of the vertical distribution of larvae at tran- sect 9. Each station consisted of a stepped oblique tow made to a maximum depth of 200 m— bottom depth permitting— parallel to bottom contours. The net was fished at 10 m depth steps for three minutes each at a vessel speed of approximately 2 knots. Net depth was monitored continuously by a Simrad^ trawl eye. The volume of water filtered was calcu- lated using Rigosha B flowmeters, calibrated in a flume tank. Reported catch rates are standardized to numbers per 1,000 m^ of water filtered. Except where specified below, sampling was not standard- ized to time of day. Data on larval depth distributions were obtained with the 1 m ring net off the west coast of Tasmania. Sampling was conducted on 20 and 21 July 1986 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. between transects 5 and 6, over a bottom depth of 100-120 m. As this site is close to the spawning area of M. novaezelandiae, the catches consisted primar- ily of small larvae. On each tow, the net was sent to depth quickly, allowed to stabilize at the selected depth for 1-2 minutes, and then retrieved slowly on a continuous oblique path. Tows were made in the order of progressively deeper depths. As each tow integrated larval abundance to the maximum depth of the tow, it was assumed that differences in stan- dardized catch rates between adjacent strata re- flected larval abundance in the depth range added. Twenty-four tows were made, varying from 15 to 90 minutes and from 10 to 90 m depth. Tows were made in six sets, three during the day (0830-1330) and three at night (2300-0400). Sunrise and sunset were at 0730 and 1700 (Australian Eastern Stan- dard Time), respectively. Samples were divided by hand into two portions. One portion was fixed in a buffered 3.7% aqueous solution of formaldehyde and the other in 95% ethanol. The former were used to identify larvae; larvae in the ethanol-fixed samples were used for ageing and assessment of growth rates. Larval abundance data are based on both portions for each station. The ages of M. novaezelandiae larvae were determined by examination of otolith microstruc- ture, following procedures outlined in Brothers et al. (1976). Whole otoliths were extracted from the larvae and viewed under transmitted light at 720- 2500 X using a Leitz Orthoplan microscope and high resolution, closed-circuit television (Ikigami Model CTC-6000). Otolith features were measured with a sonic digitizer (Science Accessories Corporation Graf/Bar) supported by an Apple 2e microcomputer and a modified version of the Basic program DISBCAL (Erie 1982). Viewed laterally, the otolith measured (the lapillus) was virtually circular; all measurements reported are to the point on the perimeter farthest from the primordium (i.e., the axis of maximum growth). Rates of larval growth are uncorrected for shrinkage. Preliminary results suggest shrinkage (TL) due to alcohol preservation averages approximately 5% and is only weakly cor- related with larval size (regression of percent shrink- age against preshrinkage TL, slope = -0.005, R- = 0.18, n = 27). Shrinkage will affect estimates of absolute growth rates, but the available litera- ture (Theilacker 1980; Fowler and Smith 1983) suggests it should not bias comparisons between growth rates, provided the larvae being compared were collected and fixed in the same manner. Statistical analyses were done using Statview 512-1- , Vers. 1.1. 31 FISHERY BULLETIN: VOL. 87, NO. 1 Validation of Ageing Procedures Larvae of M. novaezelandiae were reared in cap- tivity to determine the age at which otoliths form. Fertile eggs were obtained by stripping running ripe males and females immediately after their capture by trawl on the spawning grounds. Eggs were incu- bated in 1 L plastic jars, which were filled with sea- water, and placed in a seawater bath. Initial incu- bation temperatures ranged from 14° to 18°C (sea-surface temperature was approximately 14°C). Upon return to the laboratory, eggs were trans- ferred to 2 L glass jars and placed in an aquarium maintained at a constant temperature of 14 °C (±0.2°C). Incubation jars were not aerated, and no attempt was made to feed the larvae. Under light microscopy, otoliths were first ap- parent in M. novaezelandiae embryos 10 hours prior to hatching. At hatching, the sagittae and lapilli were developed and conspicuous. The asterisci were first apparent 3-4 days after hatching. Otoliths from newly hatched larvae characteristically have a con- spicuous dark and broad band close to their edge, which is apparently laid down at hatching. Scanning electron microscopy indicated the otolith within this hatching mark consisted of a spherical primordium surrounded by an area with little conspicuous struc- ture. The radius of the hatching mark varied be- tween specimens but did not differ significantly between reared larvae (x = 6.9 txm, range = 6.7- 8.5 iim, n = 17) and wild-caught larvae (x = 7.5 \xm, range = 5.3-9.7 yon, w = 17) (P > 0.1, two tailed <-test). All wild-caught larvae had a second exceptional- ly dense and very conspicuous band. The radius of this band varied from 10.2 to 16.4 j^m (x = 13.2 fim, n = 17), i.e., approximately 5-7 \im outside the hatching mark. Although the otoliths of reared lar- vae reached sizes close to this (maximum radius = 13.1 ymi), this distinctive band was not evident in their otoliths. As the largest of these larvae had fully ossified jaws and well-developed guts and had all but exhausted their yolk reserves, the second major band in the otoliths may have formed close to or coincident with first feeding. The microstructure of the otolith differed markedly inside and outside of the "first-feeding band". Within its radius, there was little evidence of consistent structuring (other than the hatching mark); beyond the first-feeding mark, increments were unambiguous, increasing in width exponentially. As we saw no indication that any structure prior to the first-feeding mark formed daily, otolithic age for the larvae examined is de- fined as the number of increments external to this feeding mark. This age is used in analysis of growth and advection patterns, unless otherwise indicated. Based on the observed incubation time (55-60 hours) and the observed time required for reared larvae held under temperature conditions similar to those during the spawning season to develop to a stage where feeding was possible (6 days) (Bruce 1988), the total age of larval blue grenadier can be estimated as otolithic age + 6 days, with a probable error of about + 2 days. Hence, date at first-feed- ing for a particular larva was calculated as date of collection less otolithic age, and date of spawning was date of collection less total age. In general, the development of otolith structure prior to first-feed- ing of larvae in M. novaezelandiae is remarkably similar to that of other gadoids (Radtke and Wai- wood 1980; Bolz and Lough 1983; Dale 1984), as is the proposed time frame. The hypothesis that increments in the otolith are formed daily was tested by following cohorts of in- dividuals and determining whether the change in the number of increments matched the known sampling interval (Campana and Neilson 1985; Jones 1986). Larvae were sampled within 0.5 km of a drogue deployed near the spawning grounds (see descrip- tion of drogue below). Larvae from three plankton tows made near the onset of a 26 h period (0521- 0649) were compared with those from two tows made close to its end (0628-0701). The respective samples were pooled because the number of larvae caught in each tow was small. Mean sampling in- terval between the first and last set of tows was 24.6 hours (1.025 days). Size-frequency distributions of larvae collected are given in Figure 2A. Modal analysis (means and variances unconstrained) for the first sample set indicated the presence of two nor- mally distributed populations, with means at 3.61 and 4.66 mm SL (SE = 0.06 and 0.05, respective- ly); analysis of the size-frequency distributions of lar- vae collected approximately a day later also in- dicates two means, at 4.12 and 4.91 mm SL (SE = 0.29 and 0.07, respectively). The smaller of these two means is poorly defined statistically, however. Re-analysis with the added constraint that larvae grew at the same rate across the size range of the two means (which is a reasonable approximation for such small larvae— see below and Figure 9) indicated means for the first set of samples at 3.62 and 4.65 mm SL, and for the second at 3.86 (which is within one SE of the unconstrained mean) and 4.89 mm SL, which fitted closely observed distributions. The average difference in larval sizes between the first and second set of samples (i.e., mean growth for the 24.6 h period) was 0.24 mm SL. The number of 32 THRESHER ET AL.: LARVAE OF GADOID. MACRURONUS NOVAEZELANDIAE 40 C 30 20 . 10 40 30 20 10 S^^ 0628-0701 h 20 July 1985 n = 84 0521-0649 h 19 July 1985 n = 240 2.5 3.5 4.5 5.5 Size of Larvae (mm TL) 6.5 7.5 6 3 B R = 0.86 • • • • • • •^-•''•^ • • ^..<-'^« • • 6 • ^^.f'^m • • •^^'^^ • • 2 , f^^"^ r — 1 3.5 4 4.5 Size of Larvae (mm TL) 5.5 Figure 2.— A. Size-frequency distributions of larval Maeruronus rwvaezelandiae collected in tows near a drogue at the onset and end of a 26 h sampling period. Arrows indicate ap- parent progression of the modes. B. Correlation between larval total length and number of postfirst-feeding increments for pooled subsample of larvae drawn from populations in 2A. The correlation is significant at P« 0.01 and accounts for 86% of the variance in number of increments. growth increments in these larvae was estimated by drawing subsamples from both populations, pro- portional to the number of individuals in each 0.25 mm size class, and regressing increment nimiber against larval size. The relationship is linear and highly significant (Fig. 2B). Based on the least squares regression of increment number on SL, a change in mean larval size of 0.24 mm corresponded 33 FISHERY BULLETIN: VOL. 87, NO. 1 to a change in increment number of 1.102 incre- ments. This compared favorably with the samphng interval, 1.025 days, and is consistent with predic- tions based on daily increment formation. Physical Oceanography Current patterns in the spawning area were inves- tigated in June-August 1985 by 1) release of surface drift cards near the spawning grounds, 2) deployment of a surface drifter drogued at 50 m for 24 hours, and 3) examination of surface isotherms as indicated by a shipboard thermosalinograph. A total of 2,250 surface drift cards were released during the 1985 spawning season. Cards were re- leased in four lots, two at each of two points (Fig. 1), located immediately north and south of the spawning ground. A set of cards was released at each point on 21 and 22 July and again on 11 and 12 August. A drifter was deployed at 0800 h on 19 July south of the spawning area at a site at which large numbers of newly hatched larvae were collected Oat. 42°43.4'S, long. 145°04.0'E) (Fig. 1). The drifter consisted a 8.5 m parachute drogue suspended at a depth of 50 m below a large surface buoy fitted o°s- ^- 't^ ^ 41°- "^"^""-^—^ 'C^''"^^"^ [ •\ 42°- .\ ^ ^ r 43°- V ^m 1984 • ^ j^. I440E 145° 146° 147° 148° SB'S 40' 45' • 1-10 • 1 1-100 • 101-1000 • 1001-5000 • >5000 130°E 140° 150° Figure 3.— Total catches of larval Macruronus novaezelandiae, pooled by transect, for 1984 and 1985, standardized per 1,000 m^ of water filtered. 34 THRESHER ET AL.: LARVAE OF GADOID, MACRURONUS NOVAEZELANDIAE with a cross-shaped radar reflector and a flashing light. The buoy's position, determined by radar fixes on coastal features, was recorded at 3 h intervals. Surface temperature and salinity were recorded continuously during all cruises, using a Grundy thermosalinograph. The readings were calibrated against measurements taken during routine hydro- graphical sampling using depth-profiled CTD and Niskin bottle casts. RESULTS Distribution of Larvae In both 1984 and 1985, M. novaezelandiae larvae were caught almost entirely in the winter, peaking in abundance in July and August, and primarily along the western and southern coasts of Tasmania. The highest densities were caught off the midwest coast (Fig. 3). Larvae were collected in largest num- bers at nearshore and midshelf stations (Fig. 4), i.e., at bottom depths of 30-100 m and well inshore of the shelf break, a pattern consistent across all transects. During depth-stratified sampling, relatively few larvae were caught on tows made at depths <20 m (Fig. 5). Samples taken with the ring net suggest that larvae occurred predominantly between 20 m and 90 m (at a maximum depth of 100-120 m) and that the depth of peak abundance was greater at night (60 m and below) than during the day (approx- imately 40 m). In a two-factor analysis of variance. 11 a o o 2: >> o C 3 '^ oJun 84 DAug 84 70 . *S^^^ "Aug 85 ASept 85 50 . 30 10 ___^ U Nearshore Midshelf Shelf Edge Offshelf FiGUEE 4.— The proportion (percent of total for cruise) of larvae caught on each cruise dur- ing the spawning season at each of the four typical sampling positions across the continen- tal shelf. s 3 o o o o 4000 . o >— 1 u o • on 2000 , o • o 1 • m • • O o t • % • i • 20 40 60 Depth (m) 80 100 FiGtmE 5.— Number of larvae caught during day (open circles) and night (closed circles) periods by oblique tows made to ma.\imum depths ranging from 10 to 90 m. Numbers caught are standardized to 1,000 m' of water filtered. 35 FISHERY BULLETIN: VOL. 87, NO. 1 depth, time of day, and the interaction term were all highly significant (Fg ig = 22.6, Fj ,e = 13.5, and 7^3 le = 9.6, respectively, P < 0.01 in all cases). Dif- ferences between replicate samples were small, ac- counting for only 12.7% of the variance, despite the same patch of water not being sampled each time. Although larvae of M. novaezelandiae were col- lected at stations all along the western, southern, and southeastern coasts of Tasmania, the age and size-frequency distributions of these larvae differed conspicuously between collecting sites. In 1984, lar- vae younger than 5 d postfirst-feeding were caught only on transects 5 (9% of total) and 6 (91%). This is consistent with earlier suggestions (Wilson 1981, 1982) that the area along or on the continental shelf between Sandy Cape (transect 5) and Cape Sorell (transect 6) is the primary spawning area for M. novaezelandiae in Australian coastal waters. The ages of the larvae caught at transect 5 varied wide- ly. From transect 6 south and east along the coast, the ages of larvae collected increased consistently with increasing distance from the spawning area (Fig. 6A). Differences between transects in age distributions of larvae are highly significant (i^s no = 38.8, P « 0.01), as is the correlation between age and distance ( = number of transects, based on the equal spacing of transects along the coast) from transect 5 (r = 0.64, P « 0.01). The latter corre- lation was also significant for each of the 1984 spawning season cruises individually, except the last one (September), when all larvae collected were relatively old. Differences between transects in the sizes of larvae caught paralleled differences in ages (differences between transects, i^5 no = 27.9, P « 0.01), v/ith the largest larvae collected farthest along the coast from the spawning area (at transect 9) (cor- C ■3 60, 50. 40. 30. 20. 10 0. A 1984 • I 03 O a IE S B 1 1985 30, 20 ' 10 , 4 5 6 7 8 9 1 Transect Number Figure 6.— Ages of larvae caught at each transect, pooled across sampling periods, for 1984 and 1985. Vertical bars indicate means ± 1 SD; SE are in all cases <3 days. Dif- ferences in larval ages across transects are highly significant for both years, as are the correlations between age and distance from transect 5. 36 THRESHER ET AL.: LARVAE OF GADOID, MACRURONUS NOVAEZELANDIAE relation between distance from transect 5 and size, r = 0.63, P « 0.01). The relationships between sampling site and lar- val ages and sizes in 1985 were similar to those in 1984, though apparently complicated by several fac- tors. As in 1984, differences between transects were highly significant (F9355 = 14.2, P « 0.01, and J^9 355 = 12.38, P « 0.01, for age and size, respec- tively), as were the correlations between both vari- ables and distance from the midwest coast (r = 0.33, P < 0.01, and r = 0.36, P < 0.01, for age and size, respectively) (Fig. 6B). At transects 5 and 6, 95% of larvae aged <5 d postfirst-feeding were caught. However, some larvae <5 d postfirst-feeding were also collected at transects 4 and 7, and a few larvae <15 d postfirst-feeding were found off the north- eastern coast of Tasmania (near transect 1). The age of the latter is much less than would be expected based on northward advection from the known spawning area (Gunn et al. in press), which suggests strongly the presence of a second spawning area, involving few adults, of the northeastern coast. The occurrence of these larvae confounds a general relationship between distance from the west coast and the ages and sizes of larvae caught. Hence, although for each cruise in 1985 larvae consistent- ly increased in age and size with increasing distance from transects 5 and 6, there was a broad range of larval ages at each transect, relatively old larvae at several transects, and young larvae on the east coast. Larval Advection On the basis of the distribution of larvae of dif- ferent ages and sizes around Tasmania, we hypoth- esized that most larvae were being carried passive- ly by a longshore current southwards around the coast from the primary spawning area off the west coast. The drift card returns, the movement of the drogue deployed on the west coast, and the distribu- tion of surface isotherms are generally consistent with this hypothesis. Most drift card returns were from sites southeast along the coast from the release points, including all of those from the first series (July 1985) (Fig. 7A, B). The drogue, deployed at the southern point at the same time drift cards were released, also drifted longshore and to the south (straight-line distance of 11.8 km in 26 hours). Movement of the drogue was conspicuously related to wind speed and direction, varying from nil at slack winds (<9 km/h) (as measured by shipboard anemometer) to slightly >1 km/h for a 9 h period when wind speed averaged approximately 55 km/h from the northwest (350° magnetic). For the second release series, drift cards returned shortly after being released on 11 August 1985 at the northern site were predominantly from points inshore (east) and slightly north of the release point (Fig. 7C). Of the 43 cards returned from this release, only two were found south of the release point; four, found on mainland Australia, had been transported north more than 150 km. In contrast, southeasterly transport was indicated by the cards released at the southern point on 12 August; only three of 30 returns were from sites north of the release point (Fig. 7D). One of these cards was found on South Arm Beach (southeastern Tasmania) on 27 August 1985. It had drifted slightly over 350 km in 15 days. Additional returns from this release in- cluded three cards from New Zealand, one from Flinders Island (northeast of Tasmania), and one from the southeastern coast of mainland Australia. All of these were found several months after being released. The distribution of surface isotherms also suggests the presence of a southward flowing current along the west coast of Tasmania during the spawning period of M. novaezelandiae. In both years of the study, west coast temperature plots in late autxmin and early winter were dominated by a tongue of water, 1°-2°C warmer than the surrounding water, that extended southwards along the coast, becom- ing narrower and cooler to the south (Fig. 8). This tongue of warm water, oriented parallel to the coast, was observed on all winter cruises. Satellite imagery has since documented it to be a regular seasonal feature off the west coast of Tasmania (C. Nilsson, in prep.). Growth Otolithic age was determined for 116 larvae in 1984 and 365 larvae in 1985. Growth trajectories (length-at-otolithic age) for M. novaezelandiae lar- vae were log linear for both years (Fig. 9), account- ing for 96% of the variance in length at age in 1984 and 84% of the variance in 1985. Residuals exhibit no conspicuous systematic deviation from linearity in either year and no marked increase in variance with age. Hence pooled data indicate consistent ex- ponential growth through at least the first 50 days of larval life, with no indication that the rate of growth declined late in larval life. The slope of the semilog regression was steeper, albeit only slight- ly, in 1984 than in 1985 (0.043 vs. 0.039, respective- ly, ANCOVA Fi,47; = 2.56, P < 0.001), which sug- gests that growth was more rapid in 1984. 37 FISHERY BULLETIN: VOL. 87, NO. Figure 7.— Release and recovery points of surface drift cards for the first (20 and 22 July 1985, A and B) and second (11 and 12 August 1985, C and D). 38 THRESHER ET AL.: LARVAE OF GADOID, MACRURONUS NOVAEZELANDIAE Sea surface temperature (°C) 40°- 42° 440. 40°- 42°- 44" 143° 146° 149° Figure 8.— Distribution of surface isotherms during early winter of 1984 and 1985, based on shipboard thermosalinograph traces. 1984 R^= 0.96 n = 116 •01 C ^ o — I 1 1 1 1 1 1 — 10 20 30 40 50 60 C 3 2 10 20 30 40 50 Age (days) postfirst-feeding 60 Figure 9.— Regressions of In total length against age (days postfirst feeding) for 116 lar- vae of Mae-ruronus navaezelandiae collected in 1984 and 365 larvae collected in 1985. A semilog regression accounts for 96% of the variance in length at age in 1984 and 84% of the variance in 1985. Differences between years in the slopes of the regressions are signifi- cant at P < 0.01. 39 FISHERY BULLETIN: VOL, 87, NO. 1 Individual variability in rates of larval growth was assessed by examining the distribution of residuals around the mean exponential growth trajectory for each population each year; a positive residual in- dicates growth faster than average for the popula- tion and a negative one growth slower than average. Analysis of these residuals indicated that rates of larval growth varied seasonally in both years (Fig. 10). Although the variability of rates of larval growth was high within any given period, in both years growth residuals differed significantly for lar- vae hatched in different months (ANOVA F5 no = 6.72, P < 0.001, for 1984, and i^ggei = 50.86, P < 0.001, for 1985). In 1984, there was a weak, but con- sistent tendency for residuals to increase through- out the spawning season (correlation between resid- ual and hatching date, r = 0.36, P < 0.01). In 1985, deviations from population mean growth rates were generally negative early in the spawning season, reached a positive maximum during August, and then decHned in September. There was also evidence of a complex relationship between rates of larval growth and location. Over- all, the distributions of growth residuals differed significantly across transects (i^gsjs = 8.71, P < 0.01), with relative growth rates tending to be highest farthest from the west coast spawning area. The weakness of the correlation between growth rate and distance is due, in part, to two factors. First, there was a marked change in the relation- ship between location and growth rate with increas- ing age of the larvae examined. The older the lar- vae, the more positive the slope between distance from the spawning area and relative growth rate (Fig. 11). For larvae less than approximately 10 d postfirst- feeding, the slope was significantly < C s 15 3 en o 1 June 1 July 1 Aug 1 Sept 1 Oct 1 July 1 Aug Spawning Date 1 Sept Figure 10,— Temporal variation of residuals from the semilog regression of In total length against age for 1984 and 1985, Macruronus novaezelandiae spawning started approx- imately a month later in 1985 than 1984 (see Gunn et al. in press tor details). Differences in residuals for larvae pooled by month of spawning are significant at P < 0.01 for both years. 40 THRESHER ET AL.: LARVAE OF GADOID, MACRURONUS NOVAEZELANDIAE •a 3 T3 o -p '55 ? « o OJ u C o H t/) c 5 10 15 20 25 30 35 Mean Age of Larvae (days postfirst-feeding) 40 Figure 11.— Relationship between age olMaeruronus rwvaezelandiae larvae examined (pooled into 3 d increments) and the slope of the regression between relative growth rate (residual from semilog regression of In total length against age) and transect number. The correlation between slope of the regression line and age class of larvae is significant at P < 0.01. negative; relative growth rates were highest on the west coast, near the spawning area, and declined to the south and east (Fjioe = 20.25, P < 0.001) (Fig. 12). In contrast, for larvae older than approx- imately 25 d postfirst-feeding, the slope was significantly positive; relative growth rates were lowest on the west coast, increased towards the east coast (Fi 124 = 25.58, P < 0.001), and were highest farthest from the spawning area (Fig. 12). The tran- sition from a negative (west coast fastest) to a positive (east coast fastest) slope occurs at a larval age of approximately 17-22 d postfirst-feeding. The second factor confounding the correlation between distance from the spawning ground and relative growth rates is an apparent seasonal change in the strength of the correlation, particularly for older larvae. A correlation between distance from the west coast and relative growth rates of larvae accounts for 27% of the variance in growth residuals for larvae aged more than 25 d postfirst-feeding in the early, slow-growth portion of the 1985 spawn- ing season. By August, however, during the period when larval growth rates were uniformly high, the correlation accounts for only 10% of the variance in growth rates and, by the end of the spawning season, for larvae hatched after 25 August, the rela- tionship between location and growth rates for these older larvae disappears altogether (i?^ ■= 0.02). There are insufficient data for a comparable anal- ysis of seasonal changes in growth rates of older larvae in 1984. DISCUSSION The increase in mean age and size of larvae with increasing distance from the west coast, the pattern of drift card returns, and the distribution of surface isotherms on the west coast of Tasmania during winter all support the hypothesis that larval M. novaezelandiae are transported by longshore cur- rents from a spawning ground on the west coast to the southeastern and eastern coasts. This hypothe- sis is also supported by independent studies of the physical oceanography of the west coast. A south- ward flowing, longshore current off the west coast in winter was first suggested by Newell (1961); drift bottles he deployed off the coast moved in a similar pattern to our drift cards. Subsequently, Baines et al. (1983) inferred the presence of this current from a shelfward depression of isotherms and confirmed it by the drift pattern of a drogue released off the northwestern coast. Baines et al. (1983) reported the Zeehan Current, as they named it, to be relatively narrow (approximately 40 km wide) and restricted largely to the edge of the continental shelf. It moves southwards at a depth averaged flow in the order of about 20 km/d (C. Fandry, pers. commun."). This figure is reasonably consistent with our data on lar- val ages at different points along the advection route. The distance between the spawning ground 'C. Fandry, CSIRO Division of Oceanography, GPO Box 1538, Hobart, Tasmania 7001, Australia, pers, commun. June 1987. 41 FISHERY BULLETIN: VOL. 87, NO. 1 U S= -t-» 03 XI ■an S 6 13 'to .3 .2 .1 -.1 -.2 -.3 .6 .4 .2 -.2 -.4 -.6 Larvae < 10 d post-first feeding 8 • • t • t • • 2 — ,,^^» • • R = 0.26 t \ t • • • i- ^ • • • • • • • m^"--^^^^ • • • t • • • • • — I « — 1 • 4 5 6 7 8 9 Larvae > 25 d post-first feeding R^= 0.16 • • • • • • • • • • • • 1 1 ^^^_.....^— - • • • 1 • • • t • t 1 • • • • 1 1 1 • — 1 — • • • 5 6 7 8 Transect Number West coast East coast Figure 12.— Regressions between relative growth rates and transect number for Macruronus novaezelandiae larvae <10 d postfirst-feeding and those >25 d postfirst-feeding. Correlations are significant at P < 0.01 for both age classes of larvae. and transect 8, off the southeastern coast, for ex- ample, is about 400 km. Hence, the minimum time it should take a larva drifting passively in the main- stream of the current to reach transect 8 would be approximately 20 days. In fact, the shortest inter- val between release time and recovery of one of our drift cards on the southeastern coast was only 15 days suggesting that at least occasionally larvae could be transported around the southern end of Tasmania very quickly. Total ages of larvae collected at transect 8 varied from 22 to 41 days, averaging 31 days in 1984 and 32 days in 1985. As few larvae are likely to traverse a perfectly direct path between the spawning groimds and transect 8, the mismatch between predicted minimum and observed average ages is probably reasonable and the hypothesis that larval distributions are the result of passive advec- tion seems plausible. The range of ages of larvae at each point along the advection route appears to reflect, in part, spawnings by M. novaezelandiae at sites north and south of the primary spawning area, in part, the distribution of the larvae relative to the main axis of the Zeehan Current, the location of which is like- ly to vary with time, and, in part, variations in the strength and direction of that current. Baines et al. (1983) noted that the manifestation of the current may often be overridden by direct wind effects, which is supported by our observations. The drift rate and direction of our drogue varied as an imme- diate function of wind speed and direction. C. Fandry (fn. 6) suggested that wind affects move- 42 THRESHER ET AL.: LARVAE OF GADOID, MACRURONVS NOVAEZELANDIAE ment of the water column off the west coast to depths of at least 100 m, i.e., virtually the entire depth range occupied by larval M. novaezelandiae. Hence, it is likely that the direction and speed of larval transport vary, though still being predomi- nantly southwards. Such variability is indicated by our drift card data. Drift cards released at the mid- shelf station of transect 5 on 22 July 1985 were recovered inshore and south of the release point; cards released at the same location 19 days later, however, were mostly recovered north of the release point (Fig. 7). Given the depth of the wind-driven effects, it is likely that larval fishes present at that site on the two dates would also have been advected either south or north, depending on temporary con- ditions of wind and current. Indeed, some larvae apparently develop wholly off the west coast. In both years of the study, the range in sizes and ages of larvae at transect 5, just north of the spawning grounds, was nearly as wide as those at all other transects combined. On this basis, we suspect that some oceanographic feature on the mid-west coast of Tasmania results in significant retention of larvae in that area. One possibility is that, as larvae are most abundant near shore, some are trapped in relatively static pockets of water near the coast and not entrained in the general souther- ly current stream. Another possible retention mech- anism is a coastal gyre, as yet unreported, that perhaps forms in the winter off the west coast. In- deed, our sea-surface temperature data consistent- ly show a westward bend of surface isotherms im- mediately offshore of transect 5, which could in- dicate such a gyre. Whatever the retention mechanism, a conse- quence is that larvae vary widely in the location at which they undergo planktonic development. Such variability is not trivial in M. novaezelandiae. Ap- parent rates of larval growth in the species vary significantly both with time and location: faster in 1984 than in 1985, faster in some months than in others, and faster off the west coast for young lar- vae and off the east coast for older larvae. There are two ways these differences can be interpreted: either the differences are real and reflect variability in conditions that promote growth of larvae, or they are only apparent, deriving not from variations in grovrth rates, but from grovrth-dependent mortality that varies in intensity in time and space. Testing these hypotheses directly in the field is difficult. They can be tested indirectly, however, by examining the distributions of residuals around the population-mean grovi1;h trajectories. Consider three possibilities: first, local differences in growth are real and are determined wholly by food availability; second, local variation is real, but upper and lower limits to growth are determined by physiological constraints inherent in the metabolism of the lar- vae; and third, real growth rates do not vary local- ly, but appear to differ due to variably intense selec- tion (predation) against slower growing larvae. The first hypothesis (unconstrained growth) implies nor- mal distributions of growth rates around population means for both fast and slow growing populations; the variance may alter with the mean, but skewness should not. The second hypothesis (constrained growth), however, implies distributions of growth residuals will vary with mean growth rate: the dis- tribution will be negatively skewed (to the left) when mean growth rate is high (more individuals near the maximum growth rate) and positively skewed (to the right) when mean growth rate is low (more indivi- duals near the minimum growth rate). The third hypothesis (growth-dependent mortality) also im- plies a relationship between the distribution of growth residuals and apparent mean rates of growth, but the relationship is opposite that implied by the constrained grovHh hypothesis. If predators selectively remove slow growing larvae, such mor- tality will skew distributions of growth residuals to the right. The greater the intensity of growth- dependent predation (= the higher the apparent mean growth rate), the more positive the skew. Hence, the growrth-dependent mortality hypothesis implies that when apparent mean growth rate is low, the distribution of residuals should be normal or only weakly positively skewed; when apparent mean growth rate is high, the distribution should be skewed strongly to the right. These predictions can be applied to field data for M. novaezelandiae. The mean growth rate of larvae was higher in 1984 than in 1985 and, for older lar- vae, was higher on the south and east coast than on the west coast (too few young larvae were caught on the east coast to warrant a comparison for that age group). The distributions of residuals for 1984 and 1985 are depicted in Figure 13, and those for west and southeast coast populations of larvae older than 25 d postfirst-feeding are depicted in Figure 14. The data are throughout consistent with the constrained-growth-rate hypothesis. As predicted by this hypothesis, the distribution of growth residuals is skewed negatively, albeit weakly, in 1984 (^2 = -0.35, t = 1.58, P < 0.1), and skewed positively, also weakly, in 1985 {kz = 0.45, t = 1.36, P < 0.1). Similarly, growth residuals for the relatively fast- growing larvae caught off the south and east coasts are distributed normally (kz = 0.25, t = 0.61, NS), 43 FISHERY BULLETIN: VOL, 87, NO. 1 20 16 12 J C ID 1984 n = 116 , n , — 1 ^ — n , , , -.3 ,1 .1 60 40 . 20 . 1985 n = 365 -O- -.5 -.3 -.1 .1 .3 Residual from Ln- linear Figure 13.— The distribution of residuals about the semilog regression of In total length against age of Macruronus novaezelandiae for 1984 and 1985, based on growth trajec- tories pooled by year. whereas those for the slower growing larvae col- lected off the west coast exhibit a significant posi- tive skew (^2 = 0.68, t = 2.76, P < 0.01). We con- clude, therefore, that the data are consistent with the constrained growth hypothesis, that variations in rates of larval growth documented in this study are likely to be real, operating within whatever fac- tors constrain the limits of larval growth for the species, and that they reflect variations in environ- mental conditions that affect growth rates. Exactly what these environmental conditions are is still not clear, though it is likely they relate to water temperature and food availability. That lar- vae <15-20 d postfirst-feeding grew faster off the west coast of Tasmania than off the south and east coasts could, for example, reflect the presence of the relatively warm Zeehan Current off the west coast. Growth rates of gadoid larvae increase with water temperature (Lawrence 1978) and tempera- tures in this current near the spawning grounds were 1°-2°C warmer than off the south and south- eastern coasts. Circumstantial evidence suggests that regional differences in growth rates of older larvae, in turn, were related to differences in food availability. As noted above, larvae older than 25 d postfirst-feeding grew faster off the east coast than off the west coast early in the spawning sea- sons. This difference between coasts narrowed later in the season and disappeared altogether late in the spawning season (September). This pattern of spatial and temporal differences in growth was matched by variations in coastal productivity. Harris et al. (1987) reported that in winter (August, re- ferred to by them as "early spring"), autotrophic water column productivity was higher off the east coast in 1985 than off the west coast; reported values for shelf waters ranged from 1.71 to 4.5 mg C-m"^-h"^ for the east coast versus 0.06 to 0.84 for the west coast. In September, however, (Harris et al.'s "late spring"), differences in water column 44 THRESHER ET AL.: LARVAE OF GADOID. MACRURONUS NOVAEZELANDIAE 16 12 . >, o C -_n 12 . 4 . West Coast n = 96 -.3 -.1 .1 aa ■O-^ ri- SE and East Coast n = 106 -.3 -.1 .1 .3 .5 Seasonally Adjusted Growth Residual Figure 14.— The distribution of seasonally adjusted residuals from the semilog regres- sion of In total length against age for Macruronus novaezelandiae larvae ^25 d postfirst- feeding tor larvae collected off the west coast (transects 5 and 6) and the southeast and east coasts (transects 8 and 9). Residuals were adjusted for seasonal variations in mean rates of larval growth by fitting a polynomial to the seasonal patterns and extracting new, detrended residuals. productivity between the two coasts were less pro- nounced; measured values for two sites off the east coast were 1.51 and 2.89 mg C-m-^h'^ versus values for the west coast that ranged from 1.04 to 2.24. We suspect, therefore, that growth rates of these older larvae are driven by local differences in the abundance of copepods and other larger zoo- plankters that constitute their primary diet. Why such regional differences in productivity did not result in a parallel difference between coasts in growth rates of younger larvae is not known. It may be that the effects of food availability on growth rates of first-feeding larvae are overridden by those of water temperatures. Regardless of how location affected the growth rates of young and older larvae, summarized in Figure 11, the net effect remains that conditions favorable for early larval growth were not spatial- ly coincident with those favoring growth by older larvae. Specifically, growth rates of larvae aged <10-15 d postfirst-feeding were highest closest to the spawning area of M. novaezelandiae, whereas growth rates of larvae older than 25 d postfirst- feeding increased the farther away from the spawn- ing area the larvae were caught. Why M. novae- zelandiae aggregate to spawn off the west coast of Tasmania in the winter, rather than at any other site or time, cannot be known. Winter spawnings are not the norm in gadoids (Breder and Rosen 1966; Hislop 1984) nor, with the possible exception of a weak gyre off the coast, is there any conspicuous oceanographic feature or condition, such as a highly localized plankton bloom, yet documented that would uniquely characterize the site as a particularly good one for spawning. Nonetheless, the enhanced growth rates of early stage larvae at the site argue 45 FISHERY BULLETIN: VOL. 87, NO. 1 for a positive selective value for migrating to the west coast to spawn. At the same time, increased rates of growth by older larvae away from the spawning area suggest equally strong selection to ensure that, as they develop, larvae are transported away from the west coast. Larvae achieve the max- imum growth rate only by being at the right place at the right stage of their development. Hence, migration of M. novaezelandiae to a specific spawn- ing area and subsequent contra-natant migration of larvae away from that spawning area appears to be neither evolutionarily trivial nor solely the result of selection to place eggs and larvae upstream of some specific nursery habitat. Rather, it is an adaptive feature of the reproductive biology of the fish that relates directly to elements of its larval ecology. Further, if survival of larvae varies with growth rate, as has been widely suggested (Hunter 1981; Rosenberg and Haugen 1982; Folkvord and Hunter 1986), then spatial effects on rates of larval growth can provide a mechanism that links current vari- ability with year-class strength in M. novaezelan- diae. We have, as yet, no direct evidence for such a link in this species but such a hypothesis has been frequently proposed for marine fishes (Walford 1938; Sette 1943; Harden Jones 1968; Nelson et al. 1977; Parrish et al. 1981). In most cases, however, emphasis has been placed on the adverse effects of advection, in which inappropriate current patterns result in larvae being transported into oceanic habitats not well suited for their development. For example, Devonald (1983) and Theilacker (1986) pre- sented evidence that larval mackerel, Trachurus symmetricus, found well off the California coast feed less well and are in worse condition than those col- lected closer to shore, which is consistent with the adverse effects of offshore transport on year-class strength suggested by Parrish et al. (1981). In con- trast, advection is not a negative factor in M. novaezelandiae: larvae do better when advected away from the spawning area at the right stage of their development. Such a positive effect of advec- tion is impHcit in hypotheses involving spawning grounds, nursery areas, and adult habitats that are spatially separated (Harden Jones 1968; Shelton and Hutchings 1982). Data for most species, however, are still too sparse to determine the general signif- icance of a direct, positive effect of advection on rates of larval growth like that in M. novaezelandiae. ACKNOWLEDGMENTS We thank S. Blaber and G. Harris, and the other members of the CSIRO Temperate Program for their assistance and advice throughout this study; the captain and crew of the RV Soela for making possible the field sampling; G. Leigh and S. Wayte for assistance in statistical analyses; A. Gronell and K. Sainsbury for invaluable discussions of the results of the study as they developed; G. Cresswell, C. Fandry, and C. Nilssen for advice on the physical oceanography of Tasmanian coastal waters; F. Boland and D. McLaughlan for assistance with deploying the drogue; N. Elliott for help with the drift cards; G. Davis and A. Paul for assistance in the laboratory; R. Frie for providing a copy of his program DISBCAL; and S. Blaber, F. R. Harden Jones, G. Jenkins, V. Mawson, and P. Rothlisberg for reviewing the manuscript and offering helpful suggestions. This study was supported in part by grant number 1984/63 from the Fishing Industry Research Trust Account. LITERATURE CITED Bailey. K. M. 1981. Larval transport and recruitment of Pacific hake Merluccius productus. Mar. Ecol. Prog. Ser. 6:1-9. Baines, p. G., R. J. Edwards, and C. F. Fandry. 1983. Observations of a new baroclinic current along the western continental slope of Bass Strait. Aust. J. Mar. Freshwater Res. 34:155-157. Baker. A. deC, M. R. Clarke, and M. J. Harris. 1973. The N.I.O. combination net (RMT 1 + 8) and further developments of rectangular midwater trawls. J. Mar. Biol. Assoc. U.K. 53:167-184. Bolz, G. R., and R. G. Lough. 1983. Growth of larval Atlantic cod, Gadus morhua, and had- dock, Melanogrammus aeglefintis, on Georges Bank, spring 1981. Fish. Bull., U.S. 81:827-836. Breder, C. M., and D. E. Rosen. 1966. Modes of reproduction in fishes. Tropical Fish Hob- byist, Inc., Neptune, NJ. 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. Bruce, B. D. 1988. Larval development of blue grenadier, Macruronus novaezelandiae (Hector) from Tasmanian waters. Fish. Bull., U.S. 86:119-128. BULMAN, C. M., AND S. J. M. BLABER. 1986. Feeding ecology oi Macruronus novaezelandiae (Hec- tor) (Teleostei:Merlucciidae) in south-eastern Australia. Aust. J. Mar. Freshwater Res. 37:621-639. Campana. S. E.. and J. D. Neilson. 1985. Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci. 42:1014-1032, Gushing, D. H. 1986. The migration of larval and juvenile fish from spawn- ing ground to nursery ground. J. Cons. int. Explor. Mer 43:43-49. Dale, T. 1984. Embryogenesis and growth of otoliths in the cod (Gadus morhua L.). Flodevigen Rapp. 1:231-249. 46 THRESHER ET AL.: LARVAE OF GADOID. MACRURONUS NOVAEZELANDIAE 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., La Jolla, CA, 227 p. FOLKVORD, A., AND J. R. HUNTER. 1986. Size-specific vulnerability of northern anchovy, Engraulis mordai, larvae to predation by fishes. Fish. Bull., U.S. 84:859-869. Frie, R. V. 1982. Measurement of fish scales and back-calculation of body lengths using a digitizing pad and microcomputer. Fisheries 7:5-8. Fowler, G. M., and S. J. Smith. 1983. Length changes in silver hake (Merluccius bilinearis) larvae: effects of formalin, ethanol, and freezing. Can. J. Fish. Aquat. Sci. 40:866-870. GuNN, J. S., B. D. Bruce, D. M. Furlanl R. E. Thresher, and S. J. M. Blaber. In press. Timing and location of spawning of blue grenadier Maeruronus novaezelandiae (Teleostei: Merlucciidae) in Australian coastal waters. Aust. J. Mar. Freshwater Res. Harden Jones, F. R. 1968. Fish migration. Arnold, Lond. Harris, G., C. Nilsson, L. Clementson, and D. Thomas. 1987. The water masses of the east coast of Tasmania: seasonal and interannual variability and the influence on phytoplankton biomass and productivity. Aust. J. Mar. Freshwater Res. 38:569-590. Hewitt, R. 1981. The value of pattern in the distribution of young fish. Rapp. P.-v. RSun. Cons. int. Explor. Mer 178:229-236. HiSLOP, J. R. G. 1984. A comparison of the reproductive tactics and strategies of cod, haddock, whiting and Norway pout in the North Sea. In G. W. Potts and R. J. Wooton (editors). Fish reproduc- tion: Strategies and tactics, p. 312-329. Academic, Lond. 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 Program, Univ. Wash. Press, Seattle. Jones, C. 1986. Determining age of larval fish with the otolith incre- ment technique. Fish. Bull., U.S. 84:91-103. Kuo, C-L., AND S. Tanaka. 1984a. Distribution and migration of hoki Macruronus novae- zelandiae (Hector) in waters around New Zealand. Bull. Jpn. Soc. Sci. Fish. 50:391-396. 1984b. Maturation and spawning of hoki Macruronus novae- zelandiae (Hector) in waters around New Zealand. Bull. Jpn. Soc. Sci. Fish. 50:397-402. Last, P. R., E. 0. G. Scott, and F. H. Talbot. 1983. Fishes of Tasmania. Tasmanian Fish. Develop. Auth., Hobart, Tasmania. Australia. 563 p. Lawrence, G. C. 1978. Comparative growth, respiration and delayed feeding abilities of larval cod (Gadus morhiui) and haddock (Melano- grammus aeglefinus) as influenced by temperature during laboratory studies. Mar. Biol. (Berl.) 50:1-7. McKeov™. B. a. 1984. Fish migration. Croom Helm, Lond. Naganuma, K. 1982. Theoretical consideration on diffusing process of the larval Red Sea bream, Pagrus major, in the adjacent waters of the Noto Peninsula. Bull. Jpn. Sea Reg. Fish. Res. Lab. 33:1-29. Nelson, W. R., M. C. Ingham, and W. E. Schaap. 1977. Larval transport and year-class strength of Atlantic menhaden, Brevoartia tyrannus. Fish. Bull., U.S. 75:23- 41. Newell, B. S. 1961. Hydrology of south-east Australian waters: Bass Strait and New South Wales tuna fishing area. CSIRO Div. Fish. Oceanogr. Tech. Pap. 10, 22 p. NoRCROSS, B. L., and R. F. Shaw. 1984. Oceanic and estuarine transport of fish eggs and lar- vae: a review. Trans. Am. Fish. Soc. 113:153-165. O'Bo-iXE, R. N., M. Sinclair, R. J. Conover, K. H. Mann, and A. C. Kohler. 1984. Temporal and spatial distribution of ichthyoplankton communities of the Scotian Shelf in relation to biological, hydrological, and physiographic features. Rapp. P.-v. R^un. Cons. int. Explor. Mer 183:27-40. Okubo, A. 1980. Diffusion and ecological problems: mathematical models. Biomathematics 10, 254 p. 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. Patchell, G. J. 1982. The New Zealand hoki fisheries 1972-1982. DSIR Fish. Res. Div., Occas. Pap. (38):l-23. Pommeranz, T.. C. Herrmann, and A. Kuhn. 1982. Mouth angles of the rectangular midwater trawl (RMT l-f8) during paying out and hauling. Meerdw. 29:267- 274. Power, J. H. 1986. A model of the drift of northern anchovy, Engraulis mordax, larvae in the California Current. Fish. Bull., U.S. 84:585-603. 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. 929, 10 p. Rosenberg, A. A., and A. S. Haugen. 1982. Individual growth and size-selective mortality of lar- val turbot (Scophthalmus maxirmis) reared in enclosures. Mar. Biol. (Berl.) 72:73-77. Sette, 0. E. 1943. Biology of the Atlantic mackerel (Scomber scombrus) of North America. Part 1. Early life history, including the growth, drift, and mortality of the egg and larval popula- tions. U.S. Fish. Wildl. Serv., Fish. Bull. 50:149-237. Shelton, p. a., and L. Hutchings. 1982. Transport of anchovy, Engraulis capensis Gilchrist, eggs and early larvae by a frontal jet current. J. Cons. int. Explor. Mer 40:185-198. Sherman, K., W. Smith, W. Morse, M. Berman. J. Green, and L. Ejsymont. 1984. Spawning strategies of fishes in relation to circulation, phytoplankton production, and pulses in zooplankton off the northeastern United States. Mar. Ecol. Prog. Serv. 18: 1-19. Smith, P. E. 1973. The mortality and dispersal of sardine eggs and lar- vae. Rapp. P.-v. R«un. Cons. int. Explor. Mer 164:282- 292. Smith, W. G., and W. W. Morse. 1985. Retention of larval haddock Melanogrammus aeglefinus in the Georges Bank region, a gyre-influenced spawning area. Mar. Ecol. Prog. Ser. 24:1-13. 47 Talbot, J. W. 1977. The dispersal of plaice eggs and larvae in the Southern Bight of the North Sea. J. Cons. int. Explor. Mer 37:221- 248. Theilacker, G. H. 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. 1986. Starvation-induced mortality of young sea-caught jack mackerel. Trachurus symmetrieus. determined with histo- logical and morphological methods. Fish. Bull, U.S. 84:1- 17. Tranter, D. J., and J. George. 1972. Zooplankton abundance at Kavaratti and Kalpeni Atolls in the Laccadives. In C. Mukundan and C. S. Gopinadha FISHERY BULLETIN: VOL. 87, NO. 1 Pillai (editors), Proceedings of the First International Sym- posium on Coral Reefs, Vol. 1, p. 239-256. Marine Bio- logical Association of India, Ernakulum, India. Walford, L. a. 1938. Effects of currents on distribution and survival of the eggs and larvae of the haddock (Melanogrammus aeglefinus) on Georges Bank. U.S. Bur. Fish. Bull. 49:1-73. Weidemann, H. (editor) 1973. The ICES diffusion experiment RHENO 1965. Rapp. P.-v. Cons. int. Explor. Mer 163:1-111. Wilson, M. A. 1981. Blue grenadier spawning grounds. FINTAS 4:9-10. 1982. Spawning blue grenadier caught off Cape Sorell. FINTAS 4:13. 48 AGE AND GROWTH OF KING MACKEREL, SCOMBEROMORUS CAVALLA, FROM THE ATLANTIC COAST OF THE UNITED STATES' Mark R. Collins, David J. Schmidt, C. Wayne Waltz, and James L. Pickney^ ABSTRACT Whole sagittae from 683 and sectioned sagittae from 773 "adult" (age > ; 437-1,310 mm FL), and lapilli from 29 larval (2-7 mm SL) and 69 young-of-the-year (79-320 mm FL) king mackerel, were ex- amined. All fish were from waters off the Atlantic coast of the southeastern United States (Cape Canaveral, Florida to Cape Fear, North Carolina). Back-calculated lengths at ages and von Bertalanffy growth equations were calculated from both whole and sectioned sagittae. Ages determined from sec- tioned sagittae were significantly greater than ages determined from whole sagittae, and the magnitude of the difference increased with age (from sections). Rings on sectioned sagittae are considered to be true annual increments, forming during June-September. There was no clear pattern to ring formation on whole otoliths. The oldest fish examined was age 21. The daily nature of rings on lapilli of age king mackerel was not validated, but if the marks are formed daily they suggest growth rates of approx- imately 0.47 mm/d for early larvae and 2.9 mm/d for fish 1-3 months of age. The king mackerel, Scomberomorus cavalla, is a migratory pelagic scombrid occurring in coastal waters of the western Atlantic from Massachusetts to Brazil and throughout the Gulf of Mexico (Col- lette and Russo 1984). In the United States, this fish is highly sought by both commercial and recreational fishermen from North Carolina to Texas (Manooch 1979; Trent et al. 1983). Decreased abundance in part of its range has lead to the establishment of landings quotas and limits.^ Tagging studies indicate that king mackerel from the Atlantic coast and those from the Gulf of Mexico form separate migratory groups, with some overlap and mixing in the waters of southern Florida." Biological studies in each geographic area are essential due to the importance of the species, possible reproductive isolation of the groups, and the potential for group-specific life history traits. Considerable research effort has been directed toward king mackerel in the Gulf of Mex- ico, but fish from the Atlantic coast of the United States, especially north of Florida, have received little attention. Beaumariage (1973) utilized fish 'Contribution No. 265 of the South Carolina Wildlife and Marine Resources Department, Charleston, SC 29412. ^South Carolina Wildlife and Marine Resources Department, Marine Resources Research Institute, P.O. Box 12559. Charleston, SC 29412. 'South Atlantic Fishery Management Council, Charleston, SC. News release. 7 July 1987. 'Powers, J. E.,andP. Eldridge. 1983. Assessment of Gulf of Mexico and south Atlantic king mackerel. Unpubl. manuscr.. 24 p. Southeast Fisheries Center. National Marine Fisheries Service, NOAA, Miami, FL 33149. Manuscript accepted September 1988. Fishery Bulletin, U.S. 87:49-61. from both coasts of Florida, but the only sample he had from northeastern Florida was combined with the rest of his data. Similarly, Johnson et al. (1983) sampled fish from North Carolina and South Caro- lina, but they were pooled with larger samples from the Gulf of Mexico. A more recent study (Manooch et al. 1987) utilized only Gulf of Mexico fish. Thus, there are no previous studies of Atlantic group king mackerel on which to base management. Despite evidence that otolith sections may give more accurate ages than whole otoliths in long-lived species (Beamish 1979), major studies of king mack- erel age and growth have been based principally on data derived from whole otoliths (Beaumariage 1973; Johnson et al. 1983; Manooch et al. 1987). Ade- quate validation of the use of whole sagittae has apparently been achieved in at least one of these investigations (Manooch et al. 1987), but we en- countered difficulties in the interpretation of whole otoliths while using similar methods in the present study. This report describes age and growth of king mackerel from the Atlantic coast of the southeast United States, compares results from whole and sec- tioned otoliths, and describes presumed daily growth of larval and young-of-the-year (YOY) king mack- erel from the same geographic area. METHODS King mackerel were collected along the Atlantic coast of the southeastern United States (lat 29° to 49 FISHERY BULLETIN: VOL. 87, NO. 1 35°N) from May 1983 through January 1987. "Adult" (= age >0) fish were caught on hook and line in the recreational fishery, in the commercial fishery, and during research cruises aboard the RV Oregon and RV Lady Lisa. Most YOY kings were collected during research cruises aboard the RV Lady Lisa and RV Carolina Pride using trawls of various types, but some fish were taken with gill nets, seines, and from commercial shrimp trawling bycatch. Larvae were collected from the RV Oregon with bongo (505 j^m mesh) and neuston (505 or 947 ^im mesh) nets, and were preserved in 95% ethanol. Nonlarval king mackerel were weighed and mea- sured (total length [TL] and fork length [FL]), while larvae were measured to the nearest mm standard length (SL) using a dissecting microscope and ocular micrometer. Sagittae of adults were removed and stored dry, and gutted fish and gonads were weighed when possible. All otoliths were removed from larval and YOY fish. Larval otoliths were mounted on microscope slides, while otoliths from YOY fish were stored in 75% ethanol. The lapillus was the best structure from which to count presumed daily rings for both larval and YOY king mackerel.^ Larval lapilli were immersed in oil on a microscope slide and viewed with transmitted light at 623 x on a microscope equipped with a video camera. Two readers made three counts for each of 29 larvae (2-7 mm SL), and the mean of the six counts, rounded to the nearest integer, was used to estimate the number of presumed daily rings. Lapilli from 69 YOY fish (79-320 mm FL) were prepared by a series of polishings on a smooth whetstone, on 600 grit sandpaper, and on glass with a fine liquid abrasive (AO Scientific Instruments Cat. No. 938C^). Polishing continued until rings in the central por- tion of the lapillus became visible, and readings were made in the same manner as those for larvae. Some lapilli were also read from photomicrographs taken with a scanning electron microscope (SEM) to deter- mine differences in marginal increments (distance from the distal edge of the outer ring to the otolith margin) between fish caught at different times of day. Whole sagittae from 683 adult fish were ex- amined. Otoliths were placed in a dish of cedarwood ^Waltz, C. W. 1986. Evaluation of a technique for estimating age of young-of-the-year king (Scomheromorus cavalla) and Span- ish (S. maculatus) mackerels. Unpubl. manuscr. South Caro- lina Wildlife and Marine Resources Department, Marine Re- sources Research Institute, P.O. Box 12559, Charleston, SC 29412. ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. oil and viewed, concave side up, under a dissecting microscope (12 x) with reflected light. Measure- ments from the focus to the distal edge of each opaque ring, and from the distal edge of the last opaque ring to the otolith margin, were made with an ocular micrometer along an axis approximating the extension of the sulcus acousticus (Johnson et al. 1983). The marginal increment was zero when an opaque ring occurred at the otolith margin. Transverse sections (ca. 0.5 mm thick) of one sagit- ta from each of 773 fish, including otoliths also read whole, were made through the focus on a plane perpendicular to the long axis with a Buehler Isomet low speed saw. Sections were viewed at 50 x in the same manner as whole sagittae. The focus was not always definite on sections, so measurements were standardized by defining the focus as the midpoint of a line connecting the two most distant points of the first ring. This convention closely agreed with actual focus locations for sections in which the focus was apparent. Because the axis of sagittal growth changed after the first year, sections were measured in two parts: 1) from the focus to that point on the first ring, on the dorsal side of the sulcus acousticus, which minimized the length of the line without cross- ing the sulcus acousticus, and 2) from the first ring to the margin of the section, on a line perpendicular to the rings, along the recognizable major axis of sagittal growth after year 1. Additional sections were made of sagittae from 10 randomly chosen fish: one longitudinal section, and two sections at 45° perpendicular to each other. The purposes of these sections were to determine if there was evidence for splitting of rings and to ensure that the transverse section, described above, was the most legible prep- aration. All whole and sectioned sagittae were ex- amined by two readers, and the age was excluded from analyses if the readers did not agree. Sex was determined by gross examination and was verified histologically in subsamples. Regressions of fork length on otolith radius were performed for sexes separately and combined. Back-calculated sizes at age were computed for males, females, and sexes combined by the Fraser-Lee method (Carlander 1982; Poole 1961). The SAS NLIN procedure (SAS Institute 1982) was used to fit von Bertalanffy equa- tions to the weighted mean back-calculated lengths at age. RESULTS The astericus was not detected in any larvae, sug- gesting it forms at >7 mm SL. All larval lapilli had well-defined presumed daily rings that were easily 50 COLLINS ET AL.: AGE AND GROWTH OF KING MACKEREL counted with good agreement between readers. The regression of mean ring count (i?) on SL is R = 0.11 + 1.56 (SL); n = 29; r^ = 0.73 (siginificant at P < 0.001). That r- is not higher is attributed to coarse length measurements (nearest mm). If the rings are daily, the regression of SL on R (SL = 1.15 + 0A7(R)) indicates a growth rate of 0.47 mm SL/d for early larvae (Fig. 1). Presumed daily ring counts were obtained for 54 (78%) of 69 YOY king mackerel 79-320 mm FL. A strong correlation was found for the regression of mean ring count (R) on FL (R = 2.0 + 0.32(FL); n = 54; r- = 0.92, significant at P < 0.001). If these rings are actually daily, the regression of FL oni?(FL = 7.25 + 2.91(i?)) suggests that a growth rate of 2.9 mm/d occurs at 30-100 days of age (Fig. 2). Attempts to produce evidence for the daily nature of these rings by measuring diel variation in marginal increments using SEM were not success- ful, perhaps due to inadequate specimen prepara- tion. Rings were normally visible on portions of the lapilli, but we could not consistently read increments near the margin. Two readers agreed on annual ring counts for 77% of all whole sagittae and 70% for fish >850 mm FL, resulting in 15 age (= number of rings) classes. Ex- amination of sections made in the four planes veri- fied that sections perpendicular to the long axis of the sagitta were most legible, and no evidence for splitting of rings was found. Agreement on read- ing sections was greater than that for whole sagit- tae, with counts verified on 90% of all sections and 96% from fish >850 mm FL. The oldest fish aged from sections was age 21. Agreement between the two techniques was but 47% among fish on which both whole sagittae and sections were used, and the ages were significantly different (t test for paired observations: P < 0.001). Counts were very similar for the first three to five age classes, but sections from older fish commonly showed one or more rings not detected on whole sagittae and the difference increased with age. The two procedures differ at an earlier age for males than for females (Fig. 3). The correlations of fish length with otolith radius were significant {P < 0.001 for all) for whole and sectioned sagittae of males, females, and sexes com- bined (Table 1). Plots of focus-ring measurements from sections for successive age groups through age 5 show that the distribution was unimodal for each increment, that distances to the rings varied little with age, and that overlap increased with age (Fig. 4). The pattern for whole sagittae was not quite as well defined (Fig. 5). Back-calculated lengths at ages from whole and sectioned otoliths agree well with observed lengths, especially among (younger) age groups with large sample sizes (Tables 2-7). Annual growth increments from whole and sectioned oto- liths were generally higher for females than males, especially during the first few years of life. Lengths at age determined from whole otoliths were con- (0 9 o z U- 7 o m 13 Z 1 1 1 3 5 STANDARD LENGTH (mm) Figure 1.— Regression of number of presumed daily rings on standard length of larval king mackerel. 51 FISHERY BULLETIN: VOL. 87, NO. 1 10 o Z 70 a. Mi m 2 50 • Z -"-r ~r ~r so 100 150 200 250 300 FORK LENGTH (mm) Figure 2.— Regression of number of presumed daily rings on fork length of young-of-the-year king mackereL UJ " o z 111 UJ < CE UJ -1 u_ III u. -I n Q T UJ o < s z < UJ z o •? >- -3 O • Females ° Males 16 19 "2 .30 °6 012 "T 1 1 r- 1 1 13 "n 1 15 SECTION AGE (yrs) Figure 3.— Mean difference between whole and sectioned otolith ages for each sectioned age group, by sex. Sample size is indicated for each data point. Table 1 .—Least squares regression of fork length (FL, in mm) on otolith radius (OR, in ocular units) for sectioned and whole otoliths. Sectioned Whole N r= n r^ male female combined log,o FL = log,„ FL = log,o FL = 1.088 + 1.012 log,oOR 1.209 + 0.967 log,o OR 1 .350 + 0.884 log,o OR 204 448 704 0.90 0.83 0.80 log,o FL = log,o FL = log,o FL = 1.242 + 0.918 log,o OR 1.116 + 1.002 log,oOR 0.773 + 1.184 109,0 OR 172 409 632 0.80 0.77 0.83 52 COLLINS ET AL.: AGE AND GROWTH OF KING MACKEREL 1 RING H — 49 2 RINGS N = 37 4 RINGS N = 70 3 RINGS N = 31 11-15 21-25 31-35 A1--45 51 55 61-65 71 75 Bi B5 I6-2D 26-30 36-*0 A6-50 56-60 66-70 76-60 bo 5 A 5 RINGS 4.0 N = 75 30 1 / \ \ 2 i \ 20 10 I W ^ 11-15 21-25 31-35 41-45 51-55 61-65 71-75 81-85 !6-20 26-30 36-40 46-DO 56-60 66-70 76-SO MICROMETER UNITS Figure 4.— Distributions of focus-ring distances from otolith sections for age groups one through five. 53 FISHERY BULLETIN: VOL. 87, NO. 1 1 RING N = 46 2 RINGS N = 31 4 RINGS N = 69 11-15 21-25 51-35 41 45 51-55 61-65 71-75 B1- 16-20 26-30 36-40 46-50 56-60 66-70 76-80 bO 3 / \4 5 RINGS 40 / f \ N = 66 14 30 / \ ^/ X i \ 20 /x l\ \ 10 / \ ~-^ \^\^s 11-15 21-35 51-35 41-*5 5155 61-65 71-75 81-85 16-20 26-30 36-40 46-5G 56-60 66-70 76-ftO MICROMETER UNITS Figure 5.— Distributions of focus-ring distances from whole otoliths for age groups one through five. 54 COLLINS ET AL.: AGE AND GROWTH OF KING MACKEREL Table 2.— Mean fork lengths (mm) at capture and mean back-calculated fork lengths at ages from sectioned otoliths of male king mackerel. l»/1ean No. of speci- mens length at capture Mean back-calculated lengths at successive annull Age 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 19 511 433 2 7 716 479 649 3 6 758 457 615 712 4 16 791 465 601 690 754 5 17 808 440 585 670 730 778 6 14 825 441 580 652 712 759 805 7 19 838 417 566 647 698 743 784 819 8 11 884 455 600 666 716 757 792 826 862 9 16 882 434 569 643 692 734 769 803 833 864 10 13 882 404 546 612 670 709 744 773 805 836 867 11 8 912 419 545 614 663 702 742 774 806 833 865 895 12 10 909 387 532 597 641 681 718 748 780 810 838 863 892 13 11 918 366 516 585 635 673 706 739 768 798 826 851 878 907 14 7 954 375 505 581 630 673 706 742 773 805 833 857 884 911 938 15 5 930 383 507 581 623 657 687 718 743 770 797 822 844 871 894 919 16 3 948 383 475 547 607 646 677 709 737 758 787 815 839 864 892 912 937 Total number 182 163 156 150 134 117 103 84 73 57 44 36 26 15 8 3 Weighted mean 426 566 639 689 724 753 779 801 822 839 857 875 896 914 916 937 Growth increment 426 140 72 50 34 29 25 22 20 17 17 18 21 17 2 20 sistently greater than from sections, except for age 1 females (Fig. 6). The von Bertalanffy growth con- stants (K) from whole and sectioned otoliths are greater for males (Table 8), while females attain greater maximum length. Estimates of asymptotic length (L„) from both otolith preparations are con- servative for both sexes. The distribution of monthly percentages of sec- tioned otoliths with zero marginal increment was unimodal and reasonably normal, indicating annual ring formation that peaks in August- September (Fig. 7). Few section margins were opaque during October-May, though sample sizes were smaller then. Similar treatment of marginal increment data from whole sagittae produced completely different results: opaque margins seem to occur irregularly from March through November. This suggests either that readings of whole otoliths were often in error despite agreement between observers, or that rings were not true annuli. DISCUSSION The daily nature of rings on lapilli of larval and YOY king mackerel was not validated, although cor- relations between otolith radius and fish length were very strong. If the marks are daily, they imply a moderately high average growth rate for early lar- vae followed by very rapid growth (2.9 mm/d) for fish 1-3 months of age. Future studies should con- centrate on validation, possibly by chemical (tetra- cycline, calcein) labeling of otoliths or by describ- ing diel variations in marginal increments. Readability (percentage legible enough for ob- servers to agree on age) of sectioned otoliths was greater than that of whole otoliths, especially among fish >850 mm FL. The two techniques agreed only 47% of the time, primarily for smaller individuals. Why Johnson et al. (1983) and Manooch et al. (1987) found much higher agreement (96% and 87%, re- spectively) between whole and sectioned otoliths is not clear. The opacity and appearance of sagittae may differ between Gulf of Mexico and Atlantic king mackerel (pers. commun., S. P. Naughton'), and could account for differences in agreement. Beamish (1979) noted that readability and reliability of whole otoliths differed between stocks of Pacific hake, Merluccius productus, supporting this hypothesis. He reported a 47% agreement between whole and sectioned otolith ages and concluded that ages from sections were more reliable, especially in older age groups and for certain geographic areas. He also found even greater deviations that we found be- tween ages from whole and sectioned otoliths, but utilized all readings. If our procedures were liberal- ized in a like fashion, or if readings from a single observer were used, we feel that the deviations reported here would be much greater. 'S. p. Naughton. Southeast Fisheries Center Panama City Lab- oratory. National Marine Fisheries Service. NOAA, Panama City. FL 32407, pers. commun. 55 FISHERY BULLETIN: VOL. 87, NO. 1 E T3 CO -t- (M O c£ 'O CT *- .2 ^ C to Q. ^ 0) to — o ° O c o S o C\J CM ■»- r- ^ CO oj o 00 r^ O) CO IT) 1^ 1- ■«-■.- a> o o ^co^oocMcooocj>ojr^r^ COi-CD^CVjTfT-^CDCO^ OOOIOOOOOOOOIO i-oor^'i-'^eooocvioococDO cjiCT)r^coc\jo>oo>t-Troc\j OiOOCJ)OJ050010C001CT> incD^cocvjuocji-Oi-ojT-oj r^inco^ooocDcocDco*-cooi OCT)CT>0>CT)Q0CT)CT>0)CT>00C000 cor^a>c3iOJOcr>^^oooi-cDin iocot-c\j^r^incoLncviir)(j)mcD OlOlOlCT)CT)0OC0O>OiO^Oir^0O0O CT)r^cococor^i^'a)0>ocDtnco ^^CncDOir^COCMOlT-OSCNJCDCVJCO cj>a)cococooocooocoa)co(3iN.ooco cDcDcor-^i^ino^^oocooi-oinco 0OQ00O(DCOCOCO0Or^0OCOCO0Of^h-f^ cDoi^coin^CM^r^cocor^^-Ti-'d-^tco Lnc\jcoco^Oi-ocoLni-TtT-coo(DCD cocooocooococooor^r^cooocooor^r^h- ■^■^LDcoococjimLO'^coojr^cDincocor^ o>op^r^cj)comcDi/)C\j^f»-ocooocDcyoj t^cor^r^r^F^r^r^r^f^r^r^cor^r^cDh^f^ mr^cncocoT-CTJtDO'-tcooocooor^cO'^cD f^cMcO'»-T-coocj)^or^cDi~mococ\jr^co f^t^r^r^r^r^r^cDr-^r^cDtDr^r^r^r^cDcDcD C0i-Lr)O'^^o^r^'-r^f^c\jc£>cvjcoocO'^0(D'^^ CDOT-CJ)OOOCOO)r^h-COCOOOlOOr-OCO^CDCJ coo>oc\jGor^r^r^oioc\jcoir)r-mcy)coc\jm»-o ^■"~cjcor-o>cor^(3)OiocDcO'3-^oo'-r-h-mcM mh-OOCOOOOO(J)C31C»OOOOOi-^CMC\JO'-CM CO CO CO ^^C\JCT)OJCMCO ■.- comi/imtDco^ oj y- r~ y- 7- ^ CO '- '- eg O *- '- ^ oi O) eg CD O <- O 00 O) C\J OJ o o o in c\j r*- h- O IT) oocDOO^o in CO CDCOCTJCDinCO *- 00 O ^ T- O) o o o to N- to r*- *- 1- in in CO in CD N. ^ in o o »- '- o> o o COOJCOh-OlCDr^CJJ OCO^CVJCO'^^CVJ 0>00»-»-C3^00 O Oi (D CO o o o ^ in CM CO in^cocjir^f^r^co^ oi oi eo r-r--ococy)^csjcj)0 r- oo i- o>o^O(3iO'-a)cno o> ojf^r--Tj-cDr--r--cjo^ co in ■»- O^^COCOCDCOOh-CJ) '- h- ■»- 0010>0)CT)000101CT1 »— Ol CO o)*— ^ 0)010iOO_COa)CD cO'-^r^tDin^r^inoicooi CDh-^0)0)C0i-OC0inCMC0 OJO^OICOOOOIOJOOCOOO) T-c\j*-cO'-oo(?>0)r-^oo<3ico ^cOTj-^f^cDOcor^ococji'- a)CT)CT)0iC0C0O0001OC0C001 cocoO'^f^cvjocMt^r^ooinr^ ^oO'-oo'j-^oo(D^r^'-/^co OlOlOlOCOOOCOCOCOOOiOOOOOO co^^cDcoincocNj^a-r^ococDCDco cocor^cDr^in'^^incO'-'TCO'^^ COOOCOOOCOOOCOOOCOOOOiCJjr^CDCO T-r-^oo(Dinco-<-'-C7>cococDCMr^c«j in^in^co^ojcoco*-or^cncDT-cvj COOOCOOOCOOOCOI^I^OOCOCOCOt^COOO r^or^-^coDOtocooi^cDincDor^coin coT-o^ooioooinTj-cor»-cocDojcDa) oococococor»-cor^r^i^t^r^cooor^f^r^ CO'^COCMCO'— CO^tO-^Olf-^COCOCOCOCOCD cocTicocDr^cDincD^'-o^cooi'-co^in t-s.t-^r^r^r^r-.r^(^r^(^r^r^r^f^cocDi^h* o>h-coooocMcococot^inc\jo^r^r^coo5CD inc\jcoO'^c\toO''-oicDcDOoococoLno)i- r^r^f^t^h-r^r»-t^r^cotDcDr^cDi^t^cDtoi^ oooincMin->-ocDtooinc\jr^oiO'-cDOcoin cDr-^incocomcMco^cMocTiCM'<-cDr^cococo CDCDCDCDCOtOCDCDCDCDCOCOinCDtDCDCDincDCO »-inococoo'^ooinc\jc\jcj>^^'CO'^^CT)m ^inmininin^in^'j-in^^'t^'^inin'^^^ ooc\JOCoa)'^0) < o O CO CO ,_ CD in r- CO 1^ h*. ^ eo ^ CO K 1^ C7> o> -«*■ ^ C (0 E E (J C £- m ^ a> V ^ o O 57 FISHERY BULLETIN: VOL. 87, NO. 1 Table 5.— Mean fork lengths (mm) at capture and mean back-calculated fork lengths at ages from whole otoliths of male king mackerel. Mean No. of speci- mens length at capture Mean back-calculated lengths at successive annuli Age 1 2 3 4 5 6 7 8 9 1 18 505 402 2 8 689 511 670 3 11 758 468 654 731 4 20 794 488 655 726 772 5 17 821 451 642 721 764 802 6 5 827 420 629 688 737 772 805 7 6 847 417 642 705 745 776 806 823 8 _ _ — _ _ _ _ — — 9 5 896 452 625 699 741 773 797 831 853 871 Total number 90 72 64 53 33 16 11 5 5 Weighted mean 453 649 719 760 788 803 827 853 871 Growth increment 453 195 70 41 28 14 23 26 17 Table 6.— Mean fork lengths (mm) at capture and mean back-calculated fork lengths at ages from whole otoliths of female king mackerel. Mean No. Of speci- mens length at capture 1 »^ean back-calculated lengths at successive annuli Age 1 2 3 4 5 6 7 8 9 10 11 12 1 18 552 440 2 20 712 481 666 3 26 810 502 696 780 4 42 845 490 671 762 814 5 46 882 474 670 759 820 862 6 35 915 471 673 757 810 851 892 7 21 949 453 651 746 802 845 887 921 8 12 1,022 467 697 788 837 881 924 963 995 9 8 1,035 475 669 785 842 885 918 954 986 1,018 10 9 1,079 485 689 778 837 890 931 964 997 1,028 1,056 11 1 1,138 350 654 785 814 873 917 976 1,020 1,064 1,093 1,123 12 1 1,077 387 454 724 778 806 846 873 927 968 995 1,022 1,077 Total number 239 221 201 175 133 87 52 31 19 11 2 1 Weighted mean 475 673 764 817 861 901 943 992 1,022 1,054 1,073 1,077 Growth increment 475 197 91 52 44 40 42 48 30 31 18 3 Van Oosten (1929) listed assumptions involved in the use of hard parts to determine age of fish: 1) the structures used are constant in number and iden- tity throughout the life of the fish, 2) the ratio of structure size and fish size (length) remains constant with growth, and 3) marks (rings) are annual and form at about the same time each year. The first assumption is not in doubt for otoliths. Supporting the second assumption are the correlations between fish length and otolith radius, which were signifi- cant for whole and sectioned otoliths but stronger for the latter. It is in meeting the final assumption that the validity of ages from whole otoliths becomes doubtful. The distributions of focus-ring measure- ments were only slightly better for sectioned than for whole otoliths. However, the distribution of monthly percentages of whole otoliths with opaque margins was multimodal, indicating nonannual ring formation (or large and numerous reading errors), while that of sections was unimodal and fairly nor- mal, indicating annual ring formation peaking in August-September. Manooch et al. (1987) found a peak in ring formation during February-May, but they also found ring formation in September for some fish taken off northwest Florida and suggested that this difference may be due to separate spawn- ing groups within the Gulf of Mexico. We consider rings in otolith sections valid annuli, but our evidence for validation is indirect, as in previous studies of king mackerel. As pointed out 58 COLLINS ET AL.: AGE AND GROWTH OF KING MACKEREL • Whole X Sectioned E E ^ o c Mt -I 1400 Female ^ o LL 1200 1000 aoo eoo 400 200 • • • • • • - • • 1 3 6 7 a 11 13 15 17 19 21 Age (years) Figure 6.— Mean back-calculated lengths at age from whole and from sectioned otoliths for male and female king mackerel. by Van Oosten (1929) and restated by Beamish and McFarlane (1983), procedures that produce direct evidence and validate ages of all age groups include mark-recapture techniques (which wall involve chem- ical labeling if ages are to be determined from oto- liths) and capture of known-age fish. The only pre- vious study of king mackerel producing acceptable evidence for age validation (Manooch et al. 1987) generated very different life history characteristics from ours, including maximum ages of 11 and 14 for males and females, respectively, but was based on whole otoliths from Gulf of Mexico fish. Thus, whether the differing results are due to sepa- rate groups of king mackerel with different life history characteristics or to differences in tech- niques is not known. Regardless, we have demon- strated that dubious information from whole oto- liths can appear valid, and suggest that sectioned sagittae be used to age king mackerel in future studies. 59 FISHERY BULLETIN: VOL. 87. NO. 1 Table 7.— Me; in fork lengths (mm) at capture and mean back-calculated f ork lengths at ages from whole Jtoliths of king mackerel, sexes combined Mean No. of speci- mens length at capture Mean back-calculated lengths at successive annuli Age 1 2 3 4 5 6 7 8 9 10 11 12 1 46 532 396 2 30 706 453 662 3 47 793 439 656 755 4 69 827 437 635 734 792 5 64 866 412 626 727 794 843 6 44 898 410 629 721 780 827 872 7 27 926 385 606 704 764 812 858 894 8 13 1,014 404 644 749 803 853 902 947 983 9 14 985 406 602 711 770 816 853 895 927 960 10 9 1,079 421 636 734 800 861 907 945 983 1,019 1,052 11 1 1,138 285 593 736 768 833 883 950 ,000 1,051 1,086 1,120 12 1 1,077 322 390 675 735 765 811 841 903 950 981 1,013 1,077 Total number 365 319 289 242 173 109 65 38 25 11 2 1 Weighted mean 418 633 730 786 833 872 912 961 984 1,048 1,067 1,077 Growth increment 418 215 96 56 46 39 39 48 23 63 18 9 sectioned (S c "5) 0) N 0) o 0. Table 8.— von Bertalanffy growth parameters from whole and from sectioned otoliths of king mackerel. Param- eter Estimate Asymptotic 95% confidence interval Sex Lower Upper Sectioned otoliths Male L K 'o 942 0.1915 -2.5006 905 0.1471 -3.4139 979 0.2358 - 1 .5874 Female L K 'o 1,208 0.1239 - 3.7445 1,156 0.0978 -4.8442 1,260 0.1500 - 2.6448 Combined L K 'o 1,277 0.0872 - 5.6836 1,162 0.0572 - 7.7409 1,392 0.1172 - 3.6262 Whole otoliths Male K to 853 0.5170 -0.5266 816 03334 -1.1493 889 0.7006 -0.0960 Female L to 1,122 0.2278 -1.6572 1,051 0.1570 -2.5360 1,192 0.2987 - 0.7784 Combined L K to 1,127 0.2128 - 1 .4777 1,027 0.1304 - 2.5008 1,227 0.2951 - 0.4546 Figure 7.— Monthly percentages of zero marginal increments on whole and sectioned otoliths, with number of zero marginal in- crements over number in sample for each month. 60 COLLINS ET AL.: AGE AND GROWTH OF KING MACKEREL ACKNOWLEDGMENTS We thank P. Keener for her histological expertise, M. M. Dougherty and W. J. Dougherty for providing assistance with SEM, and M. J. Clise for assisting with computer programming. Specimens were pro- vided by Susan Shipman and Henry Ansley of the Georgia Department of Natural Resources, Capt. 0. C. Poke of Captain Robert, and Capt. W. Kaps of Bellatrix II. C. S. Manooch of the National Marine Fisheries Service at Beaufort, NC and C. A. Barans and G. R. Sedberry of the South Caro- lina Wildlife and Marine Resources Department at Charleston, SC provided valuable comments on the manuscript. This work was supported by the Na- tional Marine Fisheries Service (Southeast Fisher- ies Center) and the South Carolina Wildlife and Marine Resources Department. LITERATURE CITED Beamish, R. J. 1979. Differences in the age of Pacific hake (Merluccius pro- dtictus) using whole otoliths and sections of otoliths. J. Fish. Res. Board Can. 36:141-151. Beamish. R. J., and G. A. McFarlane. 1983. The forgotton requirement for age validation in fish- eries biology. Trans. Am. Fish. Soc. 112:735-743. Beaumarlage. D. S. 1973. Age. growth, and reproduction of king mackerel, Scam- beromorus cavalla, in Florida. Fla. Mar. Res. Publ. 1, 45 p. 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. COLLETTE, B. B., and J. L. RUSSO. 1984. Morphology, systematics, and biology of the Spanish mackerels (Sco7n,6(TomorTts, Scombridae). Fish. Bull., U.S. 82:545-692. Johnson, A. G.. W. A. Fable, Jr., M. L. Williams, and L. E. Barger. 1983. Age, growth, and mortality of king mackerel, Scom- heromorus cavalla, from the southeastern United States. Fish. Bull.. U.S. 81:97-106. Manooch, C. S., III. 1979. Recreational and commercial fisheries for king mack- erel, Scomberonurrus cavalla, in the South Atlantic Bight and Gulf of Mexico, U.S.A. In E.L. Nakamura and H. R. Bullis (editors). Proceedings of the mackerel colloquium, p. 33-41. Gulf States Mar. Fish. Comm., Brownsville, TX. Manooch, C. S., Ill, S. P. Naughton, C. B. Grimes, and L. Trent. 1987. Age and growth of king mackerel, Scamberonwrus cavalla, from the U.S. Gulf of Mexico. Mar. Fish. Rev. 49(2):102-108. 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. SAS Institute, Inc. 1982. SAS users' guide: statistic. SAS Institute, Gary, NC, 584 p. Trent, L., R. 0. Williams, R. G. Taylor, C. H. Saloman, and C. H. Manooch, III. 1983. Size, sex ratio, and recruitment in various fisheries of king mackerel. Scomberorrwrus cavalla, in the southeastern United States. Fish. Bull., U.S. 81:709-721. Van Oosten, J. 1929. Life history of the lake herring (Leucichthys artedii Le Suer) of Lake Huron, as revealed by its scales, with a critique of the scale method. Bull. U.S. Bur. Fish. 44:265-448. 61 EFFECTS OF FOOD CONCENTRATION AND TEMPERATURE ON DEVELOPMENT, GROWTH, AND SURVIVAL OF WHITE PERCH, MORONE AMERICANA, EGGS AND LARVAE Daniel Margulies' ABSTRACT Growth and mortality during the egg and early larval stages of white perch, Morone americana, were examined in relation to food concentration and temperature. Laboratory experiments were conducted utilizing variable food conditions (high, low, and initially delayed rotifer levels) and temperatures (13°, 17°, and 21°C). Egg and yolk-sac larva stage durations were inversely related to temperature, and op- timum hatch of eggs occurred at 17°C or lower. Larvae fed initially at low food levels for as little as 2 days exhibited significantly reduced survival and growth after 8 days of feeding at all temperatures. Survival rates of well-fed larvae after 8 days of feeding ranged from 43 to 55%. Feeding delays of 4-8 days resulted in markedly reduced survival at 17° and 21°C. Growth was slow under any food conditions at 13°C (<0.05 mm/d in length, <5%/d in dry weight). At 17° and 21 °C, well-fed larvae grew at significant- ly higher rates (>0.20 mm/d in length, >15%/d in dry weight). Based on these laboratory data and on seasonal abundance of food in Chesapeake tributaries, it was estimated that optimum temperatures for growth and survival of first-feeding white perch larvae are 15°-20°C. Results suggest that the estima- tion of variability in growth rates of larval white perch in Chesapeake tributaries would make a major contribution to our understanding of white perch recruitment. The white perch, Morone americana, is an impor- tant recreational and commercial fish species in the Chesapeake Bay drainage. Fluctuations in relative abundance of white perch are most likely related to survivorship during the early life history, yet surprisingly little is known about the effects of varying environmental factors on growth, devel- opment, and survival of white perch eggs and larvae. Past studies on the early life history of white perch have focused on distribution patterns (Man- sueti 1961), descriptions of egg and larval develop- ment (Mansueti 1964), electrophoretic (Morgan 1975), and biochemical (Sidell and Otto 1978) characterizations of larvae and temperature ef- fects on hatching (Morgan and Rasin 1982). The interacting effects of temperature and food on the development, growth, and survival of white perch eggs and larvae had not been studied pre- viously. Fecundity of white perch, which usually are 100- 250 mm SL, is high (50,000-300,000 ova per female). 'Chesapeake Biological Laboratory, Center for Environmental and Estuarine Studies, University of Maryland, Solomons, MD 20688; present address: Inter-American Tropical Tuna Com- mission, c/o Scripps Institution of Oceanography, La Jolla, CA 92093. Manuscript accepted September 1988. Fishery Bulletin, U.S. 87:63-72. thus larval mortality rates are expected to be high (Ware 1975). For most high-fecundity species, if large numbers of larvae are produced in a cohort, small changes in growth or mortality rates during the larval stages may produce large variations in recruitment (Houde 1987). White perch juveniles are large relative to reproductive size, with the greatest relative weight increases occurring in the larval stage. This growth pattern indicates a strong poten- tial for regulation of numbers through variable lar- val growth (Houde 1987). In this study, I examined the effects of two vari- able environmental factors, food concentration and temperature, on the development, growth, and sur- vival of first-feeding white perch larvae. White perch spawn in Chesapeake tributaries over a temperature range of 10°-20°C (Hardy 1978). Thus, first-feed- ing larvae can encounter a wide range of develop- mental temperattu-es. Microzooplankton, which forms the bulk of the diet for first-feeding white perch larvae, can fluctuate in Chesapeake tidal freshwaters during spring months from <50 to >1,000/L (Heinle and Flemer 1975; Lippson et al. 1980). By examining the interacting effects of temperature and food, the scope for growth and survival potential of white perch larvae were studied. 63 FISHERY BULLETIN: VOL. 87, NO. 1 METHODS Experimental Design White perch adults were collected by otter trawl from the Potomac and Patuxent Rivers, MD dur- ing April and May 1982. Eggs were collected from at least four females and milt from four males from each river system. Gametes were stripped into 8 L polycarbonate containers. Fertilized eggs were then transported to the laboratory and placed in well- aerated, 38 L tanks divided into three temperature groups: 13°, 17°, and 21 °C. Salinity was maintained at I'Voo. After hatching, yolk-sac larvae were trans- ferred to 38 L culture tanks. Feeding studies were initiated with larvae that had some yolk remaining and which had pigmented eyes, indicating readiness to initiate feeding (Blaxter 1969). All feeding experiments were conducted for a period of 8 days. Partial water changes of 25% were made in each feeding tank every other day to mini- mize buildup of metabolites. Fluorescent lighting provided constant 200-300 lux, with photoperiod maintained on a 13 h light: 11 h dark cycle. Tem- perature was controlled to the nearest 0.5°C by maintaining aquaria in water baths of ambient Patuxent River water and regulating individual tanks by aquarium heaters. Food for larvae consisted of rotifers {Brachionus plicatilis) cultured in the laboratory on the green alga Chlorella sp. Field studies demonstrated that Brachionus constitutes the bulk of the diet of first- feeding white perch larvae (Martin and Setzler- Hamilton 1983). Based on a size analysis of zoo- plankton prey consumed by Potomac River larvae, rotifers were graded as follows: Day 1 to day 3: 100-150 imi in breadth provided; and day 4 to day 8: all sizes (100-180 nm) provided. Food levels were measured in the feeding tanks by calculating the mean values of three 100 mL aliquots taken four times daily. Food concentrations subsequently were adjusted to nominal levels. Four food groups were established, representative of high, low, and delayed-high food conditions. Group 1 was a well-fed group maintained at 800 rotifers/L; group 2 was maintained at 50 rotifers/L concentrations for 2 days and then fed at 800 rotifers/L levels for 6 days; group 3 was fed at 50 rotifers/L levels for 4 days and then fed at 800 rotifers/L concentrations for 4 days; group 4 was maintained at low levels of 50 rotifers/L for the en- tire study period. The food levels of 800 and 50 rotifers/L were representative of high and low microzooplankton levels that typically occur in tidal freshwaters of the Chesapeake (Lippson et al. 1980). Each food group was tested at three tempera- tures: 13°, 17°, and 21°C. At each temperature, 10 eggs and 10 newly hatched larvae were sampled from the rearing tanks and fixed in 4% formalin to test for possible incubation temperature or paren- tal stock effects on egg and newly hatched larva sizes. Just prior to feeding, 10 larvae were removed from each temperature stock tank and preserved for initial length and dry weight measurements. At each temperature, 150 larvae were assigned to each of two replicates for each food group (with four food groups per temperature). Once feeding was initiated, at 2 d intervals, subsamples of 3 or 4 larvae were removed from each tank and pre- served in 4% formalin for growth analyses. Sample Analyses Mean egg diameter, larval hatching length, and length at first-feeding were measured and compared among temperatures. Yolk and oil globule dimen- sions of eggs and larvae were measured by ocular micrometer and converted to yolk and oil volumes (mm^); the stage-specific volumes were then com- pared among temperatures. Regressions also were developed predicting the duration of the egg and yolk-sac stages in relation to temperature. Expected mean survival after 8 days of feeding was calculated based on the relationship: N, = N^e ~^', where A^, = number of survivors at t days after first-feeding (8 days), A^o = initial number of larvae (150), t = number of days of feeding (8), and Z = instantaneous total mortality rate. Also, Z = F + M, where F = sampling mortality and M = natural mortality rate. The number of larvae pre- served for analyses was considered sampling mor- tality (F), and all other mortality was M. Thus, when Nf), N,, t, Z, and F were known, it was possible to solve for M, from which expected number of sur- vivors was calculated as N, (Expected) = A^o^ ~^' (Ricker 1975). Growth rates were calculated from the subsam- ples of preserved larvae. Lengths were measured after three weeks of preservation using a Wild^ dissecting microscope fitted with an ocular microm- eter. Lengths were recorded to the nearest 0.1 mm. Dry weight was obtained by drying larvae at 60°C for 48 hours, dessicating, and weighing to the near- est 0.1 ng on a Cahn electrobalance. Growth in length was estimated by linear regression: L; = a ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 64 MARGULIES: EFFECTS OF FOOD AND TEMPERATURE ON WHITE PERCH + bt, where L, is length at time "t" days and b is daily growth rate (mm per day). Growth in weight for the 8 d period was determined from the expo- nential regression of dry weight on days after first- feeding: Wi = W^e^', where Wt = dry weight at time "f days, G = instantaneous daily growth coef- ficient, and Wq = dry weight at first-feeding. Spe- cific growth rate (percent per day) was calculated as 100(e'" - 1). In addition, mean incremental growth coefficients (i.e., between sampling days 2, 4, 6, and 8) also were calculated (Ricker 1975). Data were analyzed by regression analysis, anal- ysis of variance (ANOVA), and analysis of covari- ance (ANCOVA) followed by the Student-Newman- Kuels (SNK) multiple comparison test (Sokal and Rohlf 1981). The probability level for rejecting null hypotheses was P = 0.05. RESULTS Development Preserved white perch eggs ranged from 0.76 to 1.03 mm in diameter. Mean diameters of eggs hatched at 13°, 17°, and 21°C were 0.91, 0.87, and Table 1. — Characteristics of fertilized wfiite perch eggs, newly hatched larvae and first-feeding larvae cultured at three temper- atures. Temperature (°C) 13 17 21 Parameter n : 10 10 10 Eggs X diameter (mm) SE, 0.91 0.02 0.87 0.02 0.92 0.03 X yolk vol (mm^) SEi 14 0.01 0.15 0.01 0.16 0.01 X oil glob, vol (mm^) SE, 0.0081 0.0005 0.0070 0.0004 0.0080 0.0004 Newly hatched larvae X SL (mm) SE; 2.52 0.03 2.46 0.03 2.53 0.02 X dry Wt (fig) SEi 35.2 1.3 34.0 1.1 36.4 1.2 First-feeding larvae X SL (mm) SE; 3.48 0.06 3.43 0.05 3.49 0.05 X dry wt l^g) SE, 18.8 0.5 18.4 0.4 19.4 0.6 X yolk vol (mm^) SE, 00026 0.0005 0.0018 0.0004 0.0014 0.0004 X % yolk remaining SE; 1.87 0.36 1.53 0.33 0.75 0.23 X oil glob, vol (mm'') SEi 0.0036 0.0007 0.0023 0.0003 0.0022 0.0006 X % oil remaining SE, 44.1 7.8 32.1 4.6 27.6 7.0 0.92 mm, respectively (Table 1). Each egg contained a large volume of yolk (0.14-0.16 mm^), and a prominent, amber-colored oil globule (0.0070- 0.0081 mm'). There were no significant differences (ANOVA, P > 0.10) among incubation temperatures for mean egg diameter, yolk volume, or oil globule volume (Table 1). Mean larval size at hatch ranged from 2.37 to 2.81 mm SL and averaged 2.50 mm (Table 1). Dry weight of hatchlings was approximately 35 j^g. Newly hatched larvae had unpigmented eyes, with the head deflected downward over the yolk sac. Mean lengths and weights of newly hatched larvae at the three temperatures did not differ significantly (ANOVA, P > 0.10). At the first-feeding stage, larvae averaged 3.45 mm SL and weighed approximately 19 ^ig (Table 1). First-feeding larvae had utOized at least 98% of their yolk reserves and from 55 to 75% of their oil. At the first-feeding stage no significant temperature effects were apparent for size of larvae, percentage of yolk remaining or percentage of oil volume re- maining (ANOVA, P > 0.10). Although not signifi- cant, there was a trend for larvae to retain more yolk and oil at lower temperatures. Temperature had a pronounced effect on the dura- tion of the egg and yolk-sac larval stages (Fig. 1). The relationships between the durations of these stages and temperature were best described by decreasing exponential functions. Although effect of temperature on hatching success was not mea- sured precisely, rough estimates based on removals of dead eggs were made. Percentage hatch was near 80% at 13°C, approximately 60% at 17° C, and near 20% at 21°C. Survival Expected survival after 8 days of feeding ranged from 4.0 to 55.0%, depending on temperature and food conditions (Fig. 2). As expected, survival at each temperature was highest for the well-fed lar- vae in food group 1. Larvae fed at low food levels for as little as 2 days (groups 2, 3, and 4) exhibited significantly reduced survival at all temperatures (ANOVA and SNK procedure, P < 0.05). In par- ticular, larvae fed at low food concentrations for 4-8 days (groups 3 and 4) displayed markedly reduced survival at 17° and 21°C. Growth At 13°C, grovrth in length was slow under all food conditions— all larvae were <4.0 mm SL after 8 days 65 FISHERY BULLETIN: VOL. 87. NO. 1 Figure 1.— The effect of temperature on egg and yolk-sac stage duration of white perch. Plotted values are means ± 2 SE. 125 100 75 o in ^ 50 25 200 c o o a 150 r$ ra en -Lq: 100 E n. o >, ■i: LU 50 o De ^ 686.9906"°^'*°^"'" (r2.0.99) where Dg = time in hours T = temperature (°C) ■'■•i Dy = 1745. 344e 0.1863T ,^2 f where Dy = - time in hours T = temperature •^ ^^>^ ^~--i 13 17 Temperature (°C) 21 60 - 50 40^ o 30 ~ 20 10 13 "C 17 21 ^C 2 3 4 Group Figure 2.— Mean expected survival of white perch larvae after eight days of feeding. Error bars are ±2 SE. 66 MARGULIES: EFFECTS OF FOOD AND TEMPERATURE ON WHITE PERCH (Fig. 3). At 17° and 21 °C, food level effects were clearly demonstrated and the final lengths attained by larvae in all food groups differed significantly from each other (SNK procedure, P < 0.05). The well-fed larvae in groups 1 and 2 were sigTiificantly longer after 8 days of feeding at 17° and 21° than they were at 13° (SNK procedure, P < 0.05). Depending on food and temperature conditions, larvae grew in length at rates ranging from 0.01 to 0.28 mm/d (Table 2). The larvae in group 1 exhibited the highest growth rate at all temperatures, grow- 5.6 5.2 4.8 4.4 4.0 3.6 T3 3.2 5.2 4.8 4.4 4.0 3.6 3.2 4.0 3.6 3.2 17 °C [• • group 1 ]• — -. group 2 •- -• group 3 • -group 4 13 °C 2 4 6 8 Days after First-Feeding Stage Figure 3.— Growth in standard length of white perch larvae tested under four food availability conditions and at three tem- peratures. Plotted values are means ±2 SE. ing at 0.05 mm/d at 13°C, 0.20 mm/d at 17°C, and 0.28 mm/d at 21°C. At 17° and 2rC, larvae in groups 1 and 2 grew significantly faster than those in groups 3 or 4 (ANCOVA and SNK procedure, P < 0.05). For either group 1 or 2, an increase in tem- perature resulted in a significantly higher growth rate compared to 13°C (SNK procedure, P < 0.05). The linear regressions gave good fits to the growth- in-length data, although there was some indication that growth at 17° and 21 °C for groups 1 and 2 was becoming more curvilinear after day 4 (Table 2, Fig. 3). Well-fed larvae (group 1) were significantly heavier at all three temperatures (ANOVA and SNK procedure, P < 0.05). At 17° and 2rC, final mean weights of larvae from all food groups differed significantly from each other (SNK procedure, P < 0.05). As temperature increased, weight increases were most pronounced for groups 1 and 2 (Fig. 4). The well-fed group 1 larvae had the highest Table 2.— Regression equations describing growth In length of white perch larvae tested under four food availability conditions and at three temperatures. Feeding duration was 8 days. In the regression equation, L Is standard length In mm, / equals days after first-feeding, b Is grovirth rate in mm. and a Is the y-lntercept. Results of ANCOVA and multiple comparison procedures (SNK) also are given. T Food Regression (°C) group n equation SE, r-" 13 1 40 — 3.546 + 0.054f 0,004 0.97 2 42 = 3.465 + 0.049( 0.005 0.96 3 41 = 3.511 + 0.029f 0,006 0.91 4 41 = 3.550 + 0.013( 0007 0.57 17 1 39 = 3.176 + 0.202/ 0,024 0.95 2 39 = 3.250 + 0.133/ 0.015 0.95 3 37 = 3.418 + 0.050/ 0.007 0.94 4 40 = 3.521 + 0.014/ 0.006 0.77 21 1 38 = 3,293 + 0.276/ 0.019 0.97 2 40 = 3.143 + 0,216/ 0.029 0.93 3 35 = 3.381 + 0.073/ 0.011 0.94 4 38 = 3.523 + 0.020/ 0.006 0.86 ANCOVA result: The growth rates differ significantly (P < 0.001). SNK summary (different superscript numbers on each line indicate significant differences among growth rates (P < 0.05)): Among food groups (FG): (°C) FG1 FG2 FG3 FG4 13 0.054' 0.049''2 0.0292'3 0.013^ 17 0.202' 0.1 33^ 0.050^ 0.014" 21 0.276' 0.216' 0.073^ 0.020^ Among temperatures: Food group 13° 0.054' 17° 21° 0.276^ 1 0.202^ 2 0.049' 0.1 33^ 0,216^ 3 0.029' 0.050''= 0.073^ 4 0.013' 0.014' 0.020' 67 FISHERY BULLETIN: VOL. 87, NO. 1 90 70 50 30 'Z 10 >, 50 a 30 10 30 10 21 °C " • group 1 • • group 2 I' • group 3 •- • group 4 « rf\—-- 17 °C 13°C 2 4 6 8 Days after First-Feeding Stage Figure 4.— Growth in dry weight of white perch larvae tested under four food availabihty conditions and at three temperatures. Plotted values are means ±2 SE. growth rates at all temperatures (Table 3). At 17° and 21 °C group 2 larvae, delayed only 2 days, had significantly reduced overall weight gains compared to group 1 larvae (ANCOVA and SNK procedure, P < 0.05). Weight gains after 8 days for groups 3 and 4 were significantly lower at all temperatures (SNK procedure, P < 0.05) (Table 3). The mean instantaneous growth rates attained by larvae at 2 d intervals showed several important pat- terns (Fig. 5). At all temperatures, feeding at 800 versus 50 rotifer/L food levels produced significantly different growth rates in 2 days or less (ANOVA, P < 0.05). At 13°C, growth differences among food groups were established after 2 days but became in- consistent, while food group differences became more pronounced at higher temperatures. At 17° and 21 °C, larvae that had 2 d delays before being Table 3. — Regression equations describing growth in weight of white perch larvae tested under four food availability conditions and at three temperatures. Feeding duration was 8 days. In the regression equation, W is dry weight in ng, t equals days after first- feeding, G is the instantaneous growth coefficient, and W^ is dry weight at time 0. Results of ANCOVA and multiple comparison pro- cedures (SNK) also are given. T Food (°C) group n Regression equation SEr. Percent gain (%cf-') 13 1 40 W = 19.029 6°°-^" 0.0062 0.96 4.9 2 42 W s 18.774 e"''^^^' 0.0020 0.98 3.3 3 41 W = 18.911 e°°'^^' 0.0076 0.68 2.0 4 41 W = 19.433 6""^^' 0.0054 0.24 0.6 17 1 39 W = 18.938 e°""' 0.0095 0.98 15.2 2 39 W = 16.910 e'"''^'" 0.0178 0.91 11.4 3 37 W = 18.099 e°°^"' 0.0068 0.88 3.2 4 40 W = 18.858 6°°°=^' 0.0037 0.66 0.9 21 1 38 w = 19.254 e<""3' 0.0101 099 21.8 2 40 w = 19.298 e°'"^' 0.0158 0.96 15.4 3 35 w = 18.216 e°°"°' 0.0106 0.89 5.5 4 38 w = 19.497 e""'"' 0.0050 0.73 1.5 ANCOVA result; The growth rates differ significantly (P < 0.001). SNK summary (different superscript numbers on each line indicate significant differences among growth rates (P < 0.05)): Among food groups (FG): CO FG1 FG2 FG3 FG4 13 0.0481 ' 0.0328'^ 0.0196=^ 0.0058=' 17 0.1413' 0.1084= 0.031 7^ 0.0089^ 21 0.1973' 0.1436= 0.0538^ 0.0147" Among temperatures: Food group 13° 0.0481' 17° 21 o 1 0.1413= 0.1973^ 2 0.0328' 0.1084= 0.1436= 3 0.0196' 0.0317' = 0.0538= 4 0.0058' 0.0089' 0.0147' offered the high food level (group 2) equalled group 1 growth rates after lag times of 2-4 days. Growth recoveries from 4 d delays were slower and incom- plete, but there were strong indications that group 3 larvae were initiating substantial growth during the last 2 days of feeding. Group 4 larvae lost weight from day 2 to day 4 and grew slowly throughout the study. Instantaneous growth coefficient also was re- gressed on temperature for each food group (Fig. 6). All four regression coefficients (slopes) differed significantly among food groups (ANCOVA with SNK procedure, P < 0.05). Growth rates of all food groups diverged at a faster rate in the upper half of the temperature range. Growth coefficients for groups 1 and 2 larvae increased by factors of 3.5-4.0 within the temperature range tested (13°- 21°C). 68 MARGULIES: EFFECTS OF FOOD AND TEMPERATURE ON WHITE PERCH O 'S o o o 21 °c , , ^|| 0.20 ' r'""^" ' 0.16 / / / • • group 1 0.12 / / • • group 2 . .group 3 ,■1 / ' — • group 4 0.08 / / / / / 0.04 / ' ^-f'"' 0.20 - 17 °c ^ 1 / / 0.16 / i /' / / 0.12 / / / / / 1 ^ "^ 0.08 .♦ 'T''" / .-' 1 0.04 0.08 13 °C y^ 0.04 "^^■^ ..-'f y^ ♦^■' "^ - -• 2 4 6 8 Days after First Feeding Stage Figure 5.— Mean instantaneous growth coefficients attained by white perch larvae at 2 d intervals. Error bars are ±2 SE. DISCUSSION White perch produce large numbers of eggs, hatch at small sizes (<3 mm) and undergo pronounced dif- ferences in development, growth, and survival in relation to variable food and temperature conditions. Egg stage duration decreased by a factor of three over the temperature range of 13°-21°C, but the reduced diu-ation at higher temperatures was off- set by a decline in percent hatch of nearly 60%. Morgan and Rasin (1982) reported that optimum hatch of white perch eggs in the laboratory occurred at 14°-16°C, and believed that greater percent hatch occurred in the estuary at these temperatures. Hardy (1978) reported that peak spawning activity for Chesapeake Bay white perch occurs at 12°- 16°C. My results indicate that optimum tempera- tures for hatch occur at <17°C. The effect of temperature on yolk-sac stage dura- tion may be important. Prolongation of this stage could have significant effects on cohort survival. Predation by planktivorous fishes on yolk-sac larvae is probably substantial in tidal freshwaters, based on the results of laboratory predation experiments (Margulies 1986). Results reported here indicate that a short-term decrease in temperature of 4°C dur- ing the spawning season, which is not unusual in tidal freshwaters (James et al. 1984^), could prolong the yolk-sac stage by at least 3 days, which could ^James, R. W, , R. H. Simmons, and B. F. Strain. 1984. Water resources data. Maryland and Delaware, Water Year 1983. U.S. Geol. Surv. Water-Data Rep. MD-DE-83-1. c 0} (U o o 5 o 5 3 o 0) c CO c (0 (0 c 0.20 0.16 0.12 0.08 0.04 • Group 1 Y = 0.0187T - 0.1882 {r^=.98) OGioup 2 V • 0.013?r - 0.14r,5 li^=.961 A Group 3 Y = 0.0043T - 0.0376 (r^=.97) AGroup 4 Y = O.OOllT - 0.0091 li^=.9^) - / o — a' 13 17 Temperature(°C) 21 FiGLTRE 6.— Relationship between temperature and instan- taneous growth coefficient for each food group. 69 FISHERY BULLETIN: VOL. 87, NO. 1 substantially increase larval mortality due to fish predation. Survival of white perch larvae is strongly depen- dent upon food availability and temperature. At 13°C, larvae were vulnerable to low food conditions, but survival differences after 8 days among mal- nourished and well-fed larvae were much more pro- nounced at higher temperatures. The metabolic demands of larvae are reduced at low temperatures, allowing relatively low caloric intake to sustain larvae. The food levels of 800 and 50 rotifers/L used in the study correspond to caloric values of 0.64 and 0.04 cal/L, respectively (Theilacker and McMaster 1971). It was apparent from survival data, par- ticularly at 17° and 21 °C, that 0.04 cal/L was in- adequate for white perch survival, and that critical levels for survival fall in the range of 0.04-0.64 cal/L. This estimate falls within the broad range of 0.01 to 10.0 cal/L that Houde (1978) summarized as reported critical caloric concentrations for marine fish larvae. White perch resemble many small marine and estuarine larvae (e.g., northern anchovy, Engraulis mordax; jack mackerel, Trachurus sym/metricus; spot, Leiostomus xanthurus) in having relatively low survival potential at low food levels (Theilacker and Dorsey 1980; Powell and Chester 1985). For exam- ple, at 17° and 21°C, 8 d survival values for white perch decreased by 60-80% with a 4 d delay in high food levels. At those same temperatures, an 8 d delay resulted in 80-90% decreases in survival. Larger larvae such as sand lance, Ammodytes americanus, (Buckley et al. 1984); Atlantic herring, Clupea harengus harengus, (Rosenthal and Hempel 1970; Kiorboe and Munk 1986); and striped bass, Morone saxatilis, (Houde and Lubbers 1986) are less vulnerable to starvation under low-food conditions. When food is scarce, smaller larvae such as white perch are often more vulnerable to starvation be- cause of low frequency of prey contact (Laurence 1982). However, comparisons among species should be done with caution because survival potential is species-specific. For example, sea bream, Archo- sargus rhomboidalis, (Houde 1978); plaice, Pleuro- netes platessa, (Blaxter and Staines 1971); and cod, Gadus morhua, (Ellertsen et al. 1981), all relative- ly small at first-feeding, are efficient feeders and exhibit significant survival at low prey levels (<50/L). For most species, larval growth variability and stage durations are important aspects of prerecruit survival (Gushing 1976; Houde 1987). Temperature variability resulted in more than fourfold differences in mean weights of white perch larvae after 8 days of feeding. Thus, the effect of temperature on feed- ing stage duration would be even more pronounced than its effects on yolk-sac stage duration. Under good feeding conditions, a drop in temperature of 2° (from 17° to 15°, for example) would result in a 30% reduction in growth after 8 days (see Figure 6). The magnitude of the prolongation of stage dura- tion would be similar. The effects of reduced food on stage duration would be even more pronounced. At 17° or 21°C, food levels need only be reduced for 2 days upon initiation of feeding to produce the same 30% reduction in growth after 8 days (Fig. 6). The growth potential of white perch is interme- diate between that reported for temperate and sub- tropical marine and estuarine species (Houde and Schekter 1981). White perch growth at 17°C and higher exceeded that reported for most temperate latitude species, which usually grow at rates of 10%/d or less (Houde and Schekter 1981). However, white perch growth rates were less than that of most subtropical species, such as bay anchovy, Anchoa mitchilli, (Florida populations); lined sole, Archirus lineatus; sea bream (Houde and Schekter 1981); and tidewater silverside, Menidia peninsulae, (McMullen and Middaugh 1985), which may grow at 3>20%/d. The specific growth rates of white perch larvae also appear to be slightly lower than those of the larger larvae of congeneric striped bass (Chesney 1986; Houde and Lubbers 1986). Springtime densities of microzooplankton in Chesapeake tidal freshwaters usually begin to in- crease when temperatures reach 14° and peak at 20°-22°C (Lippson et al. 1980; Martin and Setzler- Hamilton 1981). Temperature and food concentra- tion have important interacting effects on white perch during the first 2-3 weeks of life, with an ap- parent balance struck between hatching success, growth rate, and survival potential. Based on my results and historical patterns of zooplankton abun- dance, the optimum temperatures for white perch development and grov^^th are in the range 15°-20°C. Hatching success was optimal at <17°C. Larvae hatched at 13°C were not as vulnerable to starva- tion, but they grew at <5%/d regardless of food level. At temperatures above 17°C, larvae could grow at >20%/d if high food levels were available at first-feeding. However, at 21°C (and presumably at higher temperatures), hatching success declined and there was greater likeliliood of starvation under suboptimum food conditions. Ultimate survival of white perch larvae and poten- tial for recruitment will depend on environmental conditions in the estuary and how they effect subtle changes in growth and mortality rates of prerecruit 70 MARGULIES: EFFECTS OF FOOD AND TEMPERATURE ON WHITE PERCH stages. A study of larval growth patterns in Chesa- peake tributaries would be useful to understand early survivorship and establishment of year-class strength in white perch. Field estimates of growth rates of white perch larvae could be compared with indices of juvenile abundance that are now obtained in Chesapeake tributaries by the Maryland Depart- ment of Natural Resources. Results of the current study indicate that even short-term variations in food and temperature can result in significant changes in survival and growth of white perch eggs and larvae. ACKNOWLEDGMENTS I thank F. D. Martin, E. M. Setzler-Hamilton, and M. Beaven for providing adult white perch from Chesapeake tributaries. E. D. Houde provided many helpful suggestions concerning the study and re- viewed an earlier version of the manuscript. F. Younger and R. Allen prepared the figures. R. Jope typed the manuscript. Financial support was pro- vided by the Chesapeake Biological Laboratory and the Center for Environmental and Estuarine Studies, University of Maryland. Portions of this manuscript were taken from a thesis submitted to the Graduate School of the University of Maryland in partial fulfillment of the requirements for the Ph.D degree. LITERATURE CITED 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. Blaxter, J. H. S., and M. E. Staines. 1971. Food searching potential in marine fish larvae. In D. J. Crisp (editor). Fourth European Marine Biology Sym- posium, p. 467-485. Cambridge Univ. Press. Buckley, L. J., S. I. Turner, T. A. Halavik, A. S. Smigielski, S. M. Drew, and G. C. Laurence. 1984. Effects of temperature and food availability on growth, survival, and RNA-DNA ratio of larval sand lance (Ammo- dytes americanus). Mar. Ecol. Prog. Ser. 15:91-97. Chesney, E. J.. Jr. 1986. Multiple environmental factors as determinants of sur- vival and growth in larval striped bass, Morone scaatilis. Int. Counc. Explor. Sea C. M. 1986/M:26. Gushing, D. H. 1976. Biology of fishes in the pelagic community. In D. H. Gushing and J. H. Walsh (editors). The ecology of the seas, p. 317-340. W.B. Saunders Co., PhUadelphia, PA. Ellertsen, B., E. Moksness, P. Solemdal, S. Tilseth, T. Westgard, and V. Oiestad. 1981 . Growth and survival of cod larvae in an enclosure. Ex- periments and a mathematical model. Rapp. P. -v. Rfun. Cons. int. Explor. Mer 178:45-57. Hardy, J. D.. Jr. 1978. Development of fishes of the mid- Atlantic Bight, an atlas of egg, larval and juvenile stages. Vol. Ill, Aphredo- deridae through Rachycentridae. U.S. Fish Wildl. Serv., FWS/OBS-78/12, 394 p. Ann Arbor, MI. Heinle, D. R., and D. A. Flemer. 1975. Carbon requirements of a population of the estuarine copepod Eurytemora affinis. Mar. Biol. (Berl.) 31:235-247. Houde, E. D. 1978. Critical food concentrations for larvae of three species of subtropical marine fishes. Bull. Mar. Sci. 28:395-411. 1987. Fish early life dynamics and recruitment variability. Am. Fish. Soc, Symp. 2:17-29. Houde, E. D., and L. Lubbers III. 1986. Survival and growth of striped bass, Morone saxatilis, and Morone larvae: laboratory and pond enclosure experi- ments. Fish. Bull.. U.S. 84:904-914. Houde, E. D., and R. C. Schekter. 1981. Growth rates, rations and cohort consumption of marine fish larvae in relation to prey concentration. Rapp. P.-v. R^un Cons. int. Explor. Mer 178:441-453. KlORBOE. T.. .\ND P. MUNK. 1986. Feeding and growth of larval herring, Clupea harengus, in relation to density of copepod nauplii. Environ. Biol. Fishes 17:133-139. Laurence, G. G. 1982. Nutrition and trophodynamics of larval fish - review, concepts, strategic recommendations and opinions. In B. J. Rothschild and C. Rooth (convenors), Fish ecology III. A foundation for REX, a recruitment experiment. Univ. Miami Tech. Rep. No. 82008. LippsoN, A. J., M. S. Haire. A. F. Holland, F. Jacobs, J. Jensen, R. L. Moran-Johnson, T. T. Polgar, and W. A. RiCHKUS. 1980. Environmental atlas of the Potomac estuary. En- viron. Cent., Martin Marietta Corp.. Baltimore, 135 p. Mansueti, R. J. 1961. Movements, reproduction, and mortality of the white perch, Rooms anteruanus, in the Patuxent estuary, Mary- land. Chesapeake Sci. 2:142-204. 1964. Eggs, larvae and young of the white perch, Roceus americanus, with comments on its ecology in the estuary. Chesapeake Sci. 5:3-45. Margulies, D. 1986. Effects of food concentration and temperature on development, growth, survival, and susceptibility to preda- tion of larval white perch {Morone americana). Ph.D. Thesis. Univ. Maryland, College Park, 168 p. Martin, F. D., and E. M. Setzler-Hamilton. 1981. Assessment of larval striped bass stocks in the Potomac estuary for 1980. Annual report to the U.S. National Marine Fisheries Service. UMCEES Ref. No. 81-65-CBL, Chesapeake Biol. Lab., Solomons, MD. 1983. Assessment of larval striped bass stocks in the Potomac estuary for years 1980, 1981 , and 1982. Final report to the U.S. National Marine Fisheries Service. UMCEES Ref. No. 83-55-CBL. Chesapeake Biol. Lab., Solomons, MD. McMULLEN, D. M., AND D. P. MiDDAUGH. 1985. The effect of temperature and food density on survival and growth of Menidia peninsulae larvae (Pisces; Atheri- nidae). Estuaries 8:39-47. Morgan, R. P., II. 1975. Distinguishing larval white perch and striped bass by electrophoresis. Chesapeake Sci. 16:68-70. Morgan, R. P., II, and V. Rasin, Jr. 1982. Influence of temperature and salinity on development 71 of white perch eggs. Trans. Am. Fish. Soc. 111:396-398. Powell. A. B., and A. J. Chester. 1985. Morphometric indices of nutritional condition and sen- sitivity to starvation of spot larvae. Trans. Am. Fish. Soc. 114:338-347. RiCKER. W. E. 1975. Computation and interpretation of biological statistics of fish populations. Fish. Res. Board Can., Bull. 191, 382 P- Rosenthal, H., and G. Hempel. 1970. Experimental studies in feeding and food requirements of herring larvae {Clupea harengus L.) In J. H. Steele (editor), Marine food chains, p. 344-364. Univ. Calif. Press, Berkeley. SiDELL, B. D., AND R. G. OtTO. 1978. A biochemical method for distinction of striped bass and white perch larvae. Copeia 1978:340-342. FISHERY BULLETIN: VOL. 87, NO. 1 SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry. The principles and practice of statistics in biological research. W. H. Freeman and Co., San Franc, 776 p. Theilacker, G. H., and K. Dorsey. 1980. Larval fish diversity, a summary of laboratory and field research. In G. D. Sharp (rapporteur). Workshop on the ef- fects of environmental variation on the survival of larval pelagic fishes, p. 105-142. UNESCO, Int. Oceanogr. Comm. Workshop Rep. 28. Theilacker, G. H., and M. F. McMasteb. 1971. Mass culture of the rotifer Brachionus plicatilis and its evaluation as a food for larval anchovies. Mar. Biol. (Berl.) 10:183-188. Ware, D. M. 1975. Relation between egg size, growth and natural mor- tality of larval fish. J. Fish. Res. Board Can. 32:2503-2512. 72 ESTIMATING SOME EARLY LIFE HISTORY PARAMETERS IN A TROPICAL CLUPEID, HERKLOTSICHTHYS CASTELNAUI, FROM DAILY GROWTH INCREMENTS IN OTOLITHS' Simon R. Thorrold^ ABSTRACT Growth increments in otoliths were used to estimate the age of larval Herklotsichthys castehmui, a tropical clupeid. collected from Townsville, northeastern Australia, in spring/summer of 1987. Daily periodicity of increment formation was confirmed by treating larvae with tetracycline and examining otoliths after a known time period. Initial increments were assumed to form at hatching; ages were thus minimum estimates. Laird-Gompertz and von Bertalanffy growth models fitted the resultant length-at-age data equally well; therefore, only the Laird-Gompertz model is presented. Specific growth rates declined from 7.4% of standard length per day at 4-5 days old to 0.4% of standard length per day at metamorphosis, 45-50 days after hatching. Absolute growth rates also declined, from 0.6 mm per day at 4-5 days to 0.08 mm per day at 44-45 days. Initial absolute growth rates are as high as any reported for clupeid larvae in the field; after this initial burst, however, the growth trajectory appeared similar to those reported for herring and pilchard larvae in temperate waters. Spawning periodicity ofH. castelnaui during the sampling period was determined by e,xamining tem- poral distribution of birthdates from otolith-aged larvae. There was indication of semilunar peaks in spawn- ing activity, apparently associated with quarter moon phases. A central problem in fisheries research is under- standing mechanisms determining year-class strength. Evidence suggests that regulation of year classes occurs during the early life history of most fish species (Parrish 1973; Smith 1985), and at- tempts to account for recruitment variability have focussed on this period of the life cycle (e.g., Hjort 1914; Gushing 1975; Koslow et al. 1987). Growth has been established as a critical parameter in the sur- vival and subsequent recruitment of larval marine fishes (Houde 1987). Weight gains of orders of mag- nitude during larval life suggest a potential for ex- tremely variable growth trajectories which may be reflected in a concomitant variability in survivorship. Growth rates are intrinsically related to suscepti- bility to both starvation (Lasker 1981) and predation (Rothschild and Rooth 1982). Small changes in growth rate can also have a dramatic effect on recruitment by determining stage durations over which high mortality indices may operate (Houde 1987). Length-frequency methods have been used exten- 'Contribution No. 447 from the Australian Institute of Marine Science, Queensland, Australia. ^Australian Institute of Marine Science, P M B No. 3 Towns- ville M.S.O., Queensland 4810, Australia. Manuscript accepted October 1988. Fishery Bulletin, U.S. 87:73-83- sively to estimate growth in larval fishes, but growth curves generated by this technique may be biased by age- and cohort-specific changes in growth rates (Crecco et al. 1983). Protracted spawning seasons may further complicate growth estimates because of the difficulties associated with connecting length modes in polymodal length-frequency distributions (Lough et al. 1982). Modal progression can also only provide mean growth estimates for larval popula- tions. These estimates are often averaged over months or years, whereas the relevant temporal scale for critical life history events may be hours or days (Fortier and Leggett 1985). The accuracy and precision of growth estimates for larval fishes have been greatly enhanced by the discovery of daily incremental rings in the otoliths of some fishes (Pannella 1971; see Campana and Neilson 1985; Jones 1986 for recent reviews). Age- ing by counting otolith growth increments allows a direct measure of length-at-age for calculation of growth curves and may provide information on in- dividual age and growth rates. Growth estimates have been obtained from a variety of species in this manner (e.g., Struthsaker and Uchiyama 1976; Methot and Kramer 1979). Back-calculation of daily rings may reveal temporal distribution of birthdates (Townsend and Graham 1981; Methot 1983), and 73 has allowed both temporal and spatial variability in daily growth rates to be investigated (Graham and Townsend 1985; Thomas 1986; Leak and Houde 1987). Although temperate clupeid species have been the focus of considerable scientific attention (Blaxter and Hunter 1982), the diverse assemblage of heavily exploited clupeids of the tropical Indo-Pacific remain poorly understood (Longhurst 1971; Whitehead 1985). There is a limited selection of literature avaU- able on the Indian oil sardines {Sardinella aurita and S. longiceps (e.g., Nair 1959; Raja 1970)), but these species are not a conspicuous component of nearshore fish communities in tropical Australia (Whitehead 1985). Williams and Clarke (1983) have examined growth in juvenile and adult Herklot- sichthys quadrimaculatus from Hawaii using the otolith increment technique. A. I. Robertson (MS in prep.) has used length-frequency data to estimate growth in juvenile Herklotsichthys castelnaui and Sardinella albella from mangrove nursery areas in tropical northeastern Australia. Dayaratne and Gjosaeter (1986) analyzed age structure of juveniles and adults in four species of Sardinella from Sri Lanka using daily growth increments on otoliths. Relatively few studies have measured larval age and growth in field situations; these parameters have not been reported for any tropical clupeid species. In this paper, I examine daily increments in oto- liths to determine some early life history parameters of a common clupeid of tropical northeastern Aus- tralia. Herklotsichthys castelnaui (Harengula abbre- viata of many authors) is a coastal pelagic clupeid found along the eastern seaboard of Australia from Bloomfield (lat. 15°56'S) to Pambula (lat. 36°57'S; Whitehead 1985). Although little is known of the biology of H. castelnaui, it inhabits estuaries and inlets (Robertson and Duke 1987), spawning in simi- mer (January-March) in the southern parts of its range (Blackburn 1941), but probably earlier in the year in more northern areas (Robertson, MS in prep.). There is no information available on larval biology. The specific aims of this project were to 1) validate daily growth increments in the otoliths of larval H. castelnaui, 2) obtain estimates of daily growth for larvae in the field, 3) investigate relationships between otolith size, standard length, and age, and 4) determine the frequency distribution of larval birthdates during the spawning season. FISHERY BULLETIN: VOL. 87, NO. 1 METHODS Collection of Larvae Larvae were collected weekly from Breakwater Marina, Townsville, Australia (Fig. 1) during August to November 1987. The marina is some 5.2 hectares in area, with an average water depth of 5 m (mhw), and is connected to Cleveland Bay by a 30 m wide entrance. Water is flushed in and out of the marina during the normal tidal cycle. Cleveland Bay is shallow, approximately 25 km wide, and bounded by Magnetic Island on its eastern side (Fig. 1). Physical oceanographic parameters of the bay have been described by Walker (1981a, b). Sampling was conducted at night using three flu- orescent lamps sealed within a clear perspex tube and a 1 m X 250 /im mesh size plankton net. The lamps were switched on and the tube lowered into the water from a jetty to a depth of 1.5 m. The plankton net was then lowered approximately 3 m below the tube. The lamps were left on for 15 min- utes, then the plankton net was hauled rapidly up over the perspex tube to the surface. This sequence was repeated 4 times during a sampling night at hourly intervals commencing at 20:00. Almost all larvae were alive upon net retrieval and were transferred immediately into 98% ethanol for subsequent sorting and analysis. Handling speci- mens in this way minimized shrinkage (Theilacker 1980) and physical damage due to net capture (McGurk 1985). Two species of clupeid larvae were collected from samples taken in the Breakwater Marina. These species were identified as H. castelnaui and Escua- losa thoracata in a size series of specimens collected during the sampling period. Details of the number ofH. castelnaui larvae collected and numbers ana- lyzed for age and growth are given in Table 1. Table 1.— Summary of sampling dates in 1987, number of Herklotsichthys castelnaui collected and numbers sub- sequently used for otolith examination. Date No. collected No. analyzed 24 August 71 50 29 August 18 18 07 September 40 40 15 September 20 20 21 September 85 48 28 September 45 45 05 October 278 50 1 3 October 239 50 19 October 82 50 27 October 31 31 74 THORROLD: LIFE HISTORY PARAMETERS IN HERKLOTSICHTHYS CASTELNAUI Breakwater Marina Cape Cleveland y 10km Figure 1.— Map of Australia showing position of Breakwater Marina. Townsville, where sampling was conducted from July to December 1987. Hatching indicates mangrove areas. Otolith Preparation Standard length of larvae (tip of snout to hypural crease or tip of notochord in preflexion larvae) was measured under a stereo dissecting microscope with an ocular micrometer. Measurement was made to the nearest micrometer unit (0.135 mm at 10 x magnification). Specimens were placed in a drop of water on a microscope slide and otoliths were teased out with electrolytically sharpened tung- sten needles. The larva was removed from the slide, and the otoliths air dried. To ensure dehydration, a drop of 98% ethanol was added to the otoliths and allowed to evaporate. Otoliths were then mounted in immersion oil for microscopic examina- tion. Individuals of H. castelnaui have three pairs of otoliths: sagittae, asterisci, and lapUli. Sagittae were the only otoliths found to be deposited during the first days of larval life, and subsequently only sagit- tae were considered in the analysis. Growth incre- ments were visible in sagittae from larvae that ranged from 5 to 25 mm SL. These otoliths were viewed for counting under a compound microscope using polarized transmitted light. All counts were made at 1000 x magnification. An Ikegami^ high resolution video camera was mounted on the micro- scope, which was connected in turn to a video screen. Otolith increments were counted on the video screen as the increased contrast made rings easier to read. The system was interfaced with a Commodore Amiga personal computer for measure- ment of otolith radius and growth increment widths (Thorrold, MS in prep.). Otolith radius was measured from the center of the primordium to the outside 'Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 75 FISHERY BULLETIN: VOL. 87. NO. 1 edge of the otolith, through the longest axis. Three counts were made of each sagitta, and the mean in- crement count from a pair of sagittae was used in the analysis. Otoliths were rejected if incremental counts within or between pairs of sagittae differed by more than two. Validation of Ageing Technique To determine if the increments observed in the otoliths of H. castelnaui were deposited daily, lar- vae collected from the marina were treated with tetracycline. Tetracycline is an antibiotic that is in- corporated into calcium structures of fish during growth. This can be restricted to a single day's in- crement on the otolith (Tsukamoto 1985), and thus, the date of treatment can be accurately identified. This technique has become widely used in the vali- dation of ageing techniques (e.g., Campana and Neilson 1982; Schmitt 1984; Kingsford and Milicich 1987). Larvae were collected from Breakwater Marina on 14th of October 1987 and were transported to the laboratory at the Australian Institute of Marine Science (AIMS). The fish were kept in ambient photoperiod and temperature regimes for two days to allow time for acclimation. They were fed twice daily on wild zooplankton captured with a 15 fim mesh plankton net from Chunda Bay, adjacent to AIMS. Ten fish were kept overnight in a 4 L tank treated with a 0.25 g/L tetracycline hydrochloride solution (Schmitt 1984). Four larvae died during exposure to the tetracycline. The remaining six larvae were returned to a 120 L tank and fed as before for 10 nights and 11 days before being sacrificed. Sagit- tae were dissected out of the remaining larvae and viewed under fluorescent UV and natural light with a compound microscope. Under fluorescent light an ocular marker was aligned with the fluorescent band in the otolith. The otolith was then examined under natural light, and the number of increments between the marker and the otolith margin counted. Both sagittae for each fish were analyzed, and three counts were made of each otolith. Statistical Procedures Laird-Gompertz and von Bertalanffy growth models were fitted to the length-at-age data. Both models have been shown to provide adequate fits to length-age data of -i- fish in different situations (e.g., Ralston 1976; Laroche et al. 1982). Zweifel and Lasker (1976) presented a detailed discussion of the Laird-Gompertz function. The generalized equation of the model is L, = Lo exp[Aola,{l - e-")] where L, = length (mm) at age t; Lq = length at i = 0; 4q = specific growth rate at < = 0; and a = rate of exponential decay. Gallucci and Quinn's (1979) version of the von Bertalanffy equation was used, where the generalized equation of the model is L, = oj/A- {1 - exp[-fc(< - f,,)]) where A: = growth constant; L„ = maximum lar- val size obtained; oj = kL^; and ^q = x-axis intercept. The BMDP P3R'' nonlinear least-squares regres- sion program employing a modified Gauss-Newton algorithm was used to fit both models. A measure of goodness-of-fit was provided by calculating an r^ value from residual and explained sums of squares derived from the least-squares regression. Good- ness-of-fit can also be assessed by examination of standard errors and approximate 95% confidence intervals of parameter estimates. Spawning frequency of H. castelnaui during the sampling period was estimated by ageing larvae and then back-calculating birthdates from the time of capture. Periodicity in spawning was analyzed using the SYSTAT SERIES^ program, employing an autoregressive moving average (ARIMA) model (Box and Jenkins 1976). Autocorrelation of each value in a series with every other value will define relationships between all points in the series. A plot of partial autocorrelations will detect dependencies in the data, and identify the period of any depen- dency. RESULTS Otolith Morphology Growth increments were clearly visible in sagit- tae of larval H. castelnaui. No marked changes in increment morphology was evident, although in some otoliths a narrowing and subsequent widen- ing of increments occurred between increments 15 and 25 (Fig. 2). Counts of growth increments were obtained from 378 larvae ranging from 5.6 to 22.5 'BMDP 3R. 1983. BMDP statistical software. Univ. Calif. Press. Berkeley, CA 94720. 'SYSTAT SERIES. 1986. SYSTAT: The system for statis- tics. Evanston, IL 50201. 76 THORROLD: LIFE HISTORY PARAMETERS IN HERKLOTSICHTmS CASTELNAUI B m 0- 1 Figure 2.—Herklotsichthys castelnaui. Sagittae from larval herring showing daily growth increments of a 29 d old larva, 16.9 mm SL at (A) 250 X (scale bar = 0.1 mm) and (B) 1000 x (scale bar = 0.025 mm). 77 FISHERY BULLETIN: VOL. 87, NO. 1 mm SL (see Figure 4), and from estimated ages of 3-53 days old. The increments were easily read in most otoliths; only 24 larvae were rejected due to either the error in reading precision being greater than 2 for the sagittae of a larva (10, 2.5%) or be- cause otoliths could not be clearly read (14, 3.5%). The plot of standard length (SL) against sagittae maximum radius (OD) revealed a logarithmic rela- tionship (Fig. 3). Otolith diameter data were log,, transformed, and a regression equation fitted to the transformed data. This equation is described by SL = 5.61.log, OD - 10.56 (n = 378; F = 2156; P < 0.0001; r' = 0.85). Comparison of means from observed and predicted standard length values suggested that any bias caused by log,.-transforming otolith diameter was neglible. Validation of Ageing Technique Increments were deposited daily in the sagittae of the six larval H. castelnaui kept under ambient conditions in the laboratory (Table 2). When viewed under natural light, a mean of 10 increments were visible from the fluorescent band to the margin of the sagittae. This corresponded to the number of nights that the fish were held after the tetracycline treatment. Larval Age and Growth It was assumed that the first otolith increment was laid down at hatching (see Discussion); there- fore, the age ofH. castelnaui larvae was estimated directly from the number of growrth increments in the sagittae. Ages were thus minimum estimates for any given length. Descriptions of growth of larval H. castelnaui were based on age-at-length of 378 25.0 n t 20.0 X ^ 15.0 ^ Q < Q < I— 10.0 5.0 0.0 M I 11 M I I I I I I I I M I I I I I I I M I I I I I I I I M I I I I I I I I I I I I I I I I I I I M I I I 0.0 50.0 100.0 150.0 200.0 250.0 300.0 SAGITTA MAXIMUM RADIUS (um Figure 3.— Relationship between standard length and radius of sagittae for larval Herklotsichthys castelnaui, together with fitted logarithmic growth curve. 78 THORROLD: LIFE HISTORY PARAMETERS IN HERKLOTSICHTHYS CASTELNAUI specimens, 5.6-22.5 mm SL. Laird-Gompertz and von Bertalanffy models yielded good and nearly identical fits to the data (r^ = 0.74 for the Laird- Gompertz curve, r^ = 0.75 for the von Bertalanffy curve); therefore, only the Laird-Gompertz growth curve is presented (Table 3; Fig. 4). This relation- ship does, however, appear to underestimate growth at ages less than 10. By contrast, the von Berta- lanffy curve underestimated growth at ages greater than 38. Table 2.— Validation of ageing using the tetra- cycline tectinique. Fish were preserved 1 1 days after treatment with tetracycline. Table shows stan- dard length (SL), mean number of increments observed between the fluorescent band and the margin of both sagittae ( + standard error), and the range of increment counts on each of the six fish. Fish no. SL Mean + SE Range 1 20.6 9.5 ± 0.2 9-10 2 21.5 10 -f 0.5 9-11 3 21.0 10 -I- 10 4 22.6 10.3 -f 0.2 10-11 5 22.6 10 + 0.4 10-11 6 20.6 10 ± 10 Total 9.8 ± 0.1 9-11 Table 3.— Laird-Gompertz equation and estimated param- eters describing growth of 378 Herklotsichthys castelnaui larvae. The growth model was fitted using nonlinear least- squares regression. STDERR = asymptotic standard error of parameter estimates; C.L. = approximate con- fidence intervals of parameter estimates. Equation L, = 5.159 exp (0.1 04/0.075 (1 - e-""'^')] Parameters STDERR 95% C.L. Lo = 5.159 A„ = 0.104 a = 0.075 0.348 0.011 0.0054 I-, = 4.474, Lj = 5.843 0.082, /.2 = 0.125 0.064, Lj = 0.086 25.0 n 20.0 ^ 15.0 H UJ Q < 10.0 - Q < h- 00 5.0 - 0.0 n = 378 r^ = 0.74 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 10 20 30 40 50 60 ESTIMATED AGE (days) Figure 4.— Relationship between standard length and number of growth increments (increment number) on sagittae for larval Herklot- sichthys castelnaui, together with fitted Laird-Gompertz growth curve. 79 FISHERY BULLETIN; VOL. 87, NO. 1 Estimates of specific and absolute growth rates (Table 4) were calculated from length-at-age (L,) for various ages as predicted by the Laird-Gompertz equation 5.26 exp[0.104/0.075 (1 - e -0.076( )]• Specific growth rate declined from 7.4% to 1.1% per day SL at 30 days after hatching, before level- ling off at around 0.5% per day SL over the latter half of larval life. Absolute growth showed a similar pattern, with a rapid decline from a maximum value of 0.57 mm/d at day 5 to a value of approximately 0.1 mm/d approaching metamorphosis, 40-50 days after hatching. Spawning occurred over three months (mid-July to mid-October), peaking around the first week in September (Fig. 5). There are indications of lunar periodicity within those months; spawning peaks were separated by approximately two weeks, ap- parently associated with first and third quarter moon phases. Time series analysis also indicated a 14 d periodicity in spawning peaks. Autocorrelation coefficients larger than two standard errors of the mean autocorrelation coefficient are considered sig- nificant (fn. 5). Coefficients greater than this value Table 4.— Growth rates of Herklotsichthys castel- naui larvae predicted from the Lalrd-Gompertz growth equation at various times after hatching. Specific growth Absolute growth Age rate rate (days) (%/d SL) (mm/d) 4-5 7.4 0.57 9-10 5.1 0.53 14-15 3.5 0.45 19-20 2.4 0.36 24-25 1.7 0.27 30-31 1.1 0.20 34-35 0.8 0.15 40-41 0.5 0.10 44-45 0.4 0.08 (0.29) occurred on day (0.62) and day 14 (0.39), in- dicating the existence of a periodicity in the data of 14 days. DISCUSSION Clear growth increments consisting of alternating light and dark bands were visible in all three pairs of otoliths in Herklotsichthys castelnaui. Asterisci and lapilli were formed, however, some time after the sagittae. This suggested that the number of F R E Q U E N C Y 3b o 30 - 25 - 20 - 15 - 10 - 5 1 1 1 1 n 378 7/11 7/19 7/27 8/4 8/12 8/20 8/28 9/5 9/13 9/21 9/29 10/7 10/15 DATE OF INITIAL INCREMENT FORMATION Figure 5.— Spawning periodicity of Herklotsichthys castelnaui based on back-calculated dates of initial increment formation. Frequency values are 2 d class means, two increments being the maximum accepted error in increment counts. 80 THORROLD: LIFE HISTORY PARAMETERS IN HERKLOTSICHTHYS CASTELNAUI growth increments on the sagitta provided the closest estimate of age in larval H. castelnaui. Age at initial increment deposition was not deter- mined in this study. Initial growth increments have been shown to be deposited prior to egg hatching, at hatching, just after hatching, and at onset of exogenous feeding (Brothers et al. 1976; McGurk 1984; Kingsford and Milicich 1987). All temperate clupeids studied have initiated ring formation at yolk-sac absorption (Geffen 1982; Lough et al. 1982; McGurk 1984; Re 1984) from 3 to 5 days after hatch- ing. Although no work has been published on oto- lith formation in tropical clupeids, a comparatively high water temperature, and hence a rapid devel- opmental rate, suggests that endogenous reserves would be quickly exhausted (Houde 1974). It was thus assumed here that the first otolith increment is laid down at hatching, and otolith counts were assumed to be a direct measure of age. Violation of this assumption will have led to biased estimates of Lq, the size at hatching, and ^o. the specific growth rate at hatching, in the Laird-Gompertz model. The magnitude of absolute and specific growth rates re- main valid. The age at which the growth rates were calculated will, however, have a systematic error corresponding to the time from hatching to initial increment formation. Standard length increased as a logarithmic func- tion of otolith radius. Linear (e.g., Riceetal. 1985), logarithmic (Nishimura and Yamada 1984; Tsuji and Aoyama 1982), and some combination of the two functions (Jenkins 1987) have been reported in the literature. A close correlation between standard length and otolith growth at a daily level implies that the width of any growth increment is a measure of instantaneous growth (Campana and Neilson 1985). The smoothly monotonic relationship between stan- dard length and otolith radius presented here sug- gests that it may be valid to reconstruct individual growth histories by examination of growth incre- ment spacings in sagittae of H. castelnaui. Both Laird-Gompertz and von Bertalanffy growth curves adequately fitted the length-at-age data. The growth trajectory of larval H. castelnaui indicates that growth is rapid for the first two to three weeks, but slows after this period. Growth may become asymptotic after this point, as predicted by the single cycle Laird-Gompertz model, or alternative- ly, enter a new growth stanza during juvenile life, as has been reported for Herklotsichthys quadri- maculatus (Williams and Clarke 1983). Data presented here for larval H. castelnaui af- fords good comparison with some temperate clupeid species, where growth rates have also been elu- cidated using the otolith increment technique. Ini- tial growth rates of 0.5-0.6 mm/d in H. castelnaui are as high as any recorded for clupeid larvae in the field. Similar growth estimates have been reported off South West Africa, where Sardinops ocellatus larvae grow linearly at rates of approximately 0.7 mm/d (Thomas 1986). Growth estimates of 0.2-0.4 mm/d after this initial burst are closer to those presented for Clupea harengus from the northern Atlantic (Townsend and Graham 1981; Lough et al. 1982; Henderson et al. 1984). Growth rates of lar- val H. castelnaui may reflect higher ambient water temperatures, as both S. ocellatus and C. harengus have a higher L„ and hence higher predicted growth rates (Ricker 1975). Spawning periodicity in H. castelnaui was ap- parently correlated with the quarter moon phases. Lunar-synchronized spawning has been reported in salmoniform, atheriniform, tetraodontiform, and perciform fishes (Taylor 1984). Most fish species with lunar-spawning rhythms spawn on or around the new or full moon (e.g., Lobel 1978; Middaugh et al. 1984), although spawning in French grunts, Haemulonjlavolineatum, also appears to be coupled with quarter moons (MacFarland et al. 1985). It should be noted that results presented here may be subject to some systematic error in ageing. If, for example, initial increment formation occurs some time after hatching, then birth dates will have been consistently underestimated. MacFarland et al. (1985) hypothesized that currents favorable for set- tlement may account for fertilization and recruit- ment events peaking on the quarter moon. My results suggest that spawning occurs with some semilunar periodicity, but the time of initial incre- ment formation needs to be determined before relating spawning events to moon phases and possi- ble tidal influences on egg and larval distributions. The most significant advantage of using otolith ageing techniques is the ability to produce individual rather than population statistics. Although it has been possible to fit a growth equation to the length- age data presented here, there is also an amount of variability surrounding the curve. This variabil- ity may, at least in part, be a sampling artifact caused by methodological problems. Inaccurate age determinations may be caused by nondaily deposi- tion of rings under some conditions (e.g., Geffen 1982), or failure to detect all rings within an otolith due to the resolution problems of light microscopy (Campana et al. 1987). Conversely, if the data are accurate, variable growth rates on small spatial (tens of meters) and temporal (days) scales are detectable by otolith analysis. It is often tacitly assumed in lar- 81 FISHERY BULLETIN: VOL. 87, NO. 1 val studies that a fast growth rate will be reflected in lower mortality rates, although the implications of fast or slow growth to subsequent survivorship have yet to be addressed. Otolith analysis empha- sizing individual rather than population growth parameters may provide a tool for approaching such questions in field situations. ACKNOWLEDGMENTS This study was conducted while the author was the holder of the Monkman Fellowship, Department of Marine Biology, James Cook University. Addi- tional financial support was provided by the Austra- lian Institute of Marine Science and a Great Bar- rier Reef Marine Park Authority augmentative grant. J. H. Choat and D. McB. Williams supervised this study and provided constructive criticisms on the manuscript. I thank A. Robertson, T. Fowler, and two anonymous referees for critically review- ing, and considerably improving, this paper. LITERATURE CITED Blackburn, M. 1941. The economic biology of some Australian clupeoid fish. Counc. Sci. Ind. Res. Aust. Bull. No. 138, 135 p. Blaxter, J. H. S., AND J. R. Hunter. 1982. The biology of the clupeoid fishes. Adv. Mar. Biol. 20:1-223. Box, G. E. P., AND G. M. Jenkins. 1976. Time series analysis, forecasting and control. Holden Day, San Franc, 575 p. Brothers, E. B., and W. N. McFarland. 1981. Correlations between otolith microstructure, growth, and life history transitions in newly recruited French grunts (Haemulon Jlavolineatum. Desmarst, Haemulidae). Rapp. P.-v. R^un. Cons. int. Explor. Mer 178:369-374. Campana, S. E., and J. D. Neilson. 1982. Daily growth increments in otoliths of starry flounder {Platicthys stellatus) and influence of some environmental variables in their production. Can. J. Fish. Aquat. Sci. 39: 937-942. 1985. Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci. 42:1014-1032. Campana, S. E., J. A. Gagne, and J. Munro. 1987. Otolith microstructure of larval herring (Clupea haren- gus): Image or reality? Can. J. Fish. Aquat. Sci. 44:1922- 1929. Crecco, v., T. Savoy, and L. Gunn. 1983. Daily mortality rates of larval American shad (Alosa sapidissima) in the Connecticut River with changes in year- class. Can. J. Fish. Aquat. Sci. 40:1719-1728. Gushing, D. H. 1975. Marine ecology and fisheries. Cambridge Univ. Press, Cambridge, England, 278 p. Dayaratne, p., and J. Gjosaeter. 1986. Age and growth of four Sardinella species from Sri Lanka. Fish. Res. 4:1-33. Fortier, L., and W. C. Leggett. 1985. A drift study of larval fish survival. Mar. Ecol. Prog. Ser. 25:245-257. Gallucci, V. F., and T. J. Quinn. 1979. Reparameterizing, fitting, and testing a simple growth model. Trans. Am. Fish. Soc. 108:14-25. Geffen, a. J. 1982. Otolith ring deposition in relation to growth rate in her- ring (Clupea hareiifftis) and turbot {Scophthalmiis maximus) larvae. Mar. Biol. (Beri.) 71:317-326. Graham, J. J., and D. W. Townsend. 1985. Mortality, growth, and transport of larval Atlantic her- ring Clupea harengus in Maine coastal waters. Trans. Am. Fish. Soc. 114:490-498. Henderson, P. A., J. W. Whitehouse, and G. H. Cartwright. 1984. The growth and mortality of larval herring, Clupea harengus L., in the River Blackwood Estuary, 1978-1980. J. Fish Biol. 24:613-622. Hjort, J. 1914. Fluctuations in the great fisheries of Northern Europe viewed in light of biological research. Rapp. P.-v. R^un. Cons. perm. int. Explor. Mer 20:1-228. HOUDE, E. D. 1974. Effects of temperature and delayed feeding on growth and survival of larvae of three species of subtropical marine fishes. Mar. Biol. (Berl.) 26:271-285. 1987. Fish early life dynamics and recruitment variability. Am. Fish. Soc. Symp. 2:17-29. Jenkins, G. P. 1987. Age and growth of co-occurring larvae of two flounder species, Rhomhosolea tapirina and Ammotretis rostratus. Mar. Biol. (Beri.) 95:157-166. Jones, C. 1986. Determining age of larval fish with the otolith incre- ment technique. Fish. Bull., U.S. 84:91-102. KiNGSFORD, M. J., and M. J. Milicich. 1987. Presettlement phase of Parika scaber (Pisces:Mono- canthidae): a temperate reef fish. Mar. Ecol. Prog. Ser. 36:65-79. Koslow, j. a., K. R. Thompson, and W. Silvert. 1987. Recruitment to northwest Atlantic cod (Gadus rrwrhua) and haddock (Melanogrammus aeglefinus) stocks: Influence of stock size and climate. Can. J. Fish. Aquat. Sci. 44:26- 39. 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. 1981. The role of a stable ocean in larval fish survival and subsequent recruitment. In R. Lasker (editor), Marine fish larvae: morphology, ecology, and relation to fisheries, p. 80-87. Univ. Wash. Press, Seattle. Leak, J. C, and E. D. Houde. 1987. Cohort growth and survival of bay anchovy Anchoa mitchilli larvae in Biscayne Bay, Florida. Mar. Ecol. Prog. Ser. 37:109-122. Lobel, p. S. 1978. Diel, lunar, and seasonal periodicity in the reproduc- tive behavior of the pomacanthid fish, Centropyge potteri, and other reef fishes in Hawaii. Pac. Sci. 32:193-207. Longhurst, a. R. 1971. The clupeoid resources of tropical seas. Oceanogr. Mar. Biol. Annu. Rev. 9:349-385. 82 THORROLD: LIFE HISTORY PARAMETERS IN HERKLOTSICHTHYS CASTELNAUI Lough, R. G., M. Pennington, G. R. Bolz, and A. A. Rosenberg. 1982. Age and growth of larval Atlantic herring, Clupea harerufus L., in the Gulf of Maine-Georges Bank region based on otolith growth increments. Fish. Bull., U.S. 80:187-199. McFarland, W. N., E. B. Brothers, J. C. Ogden, M. J. Shulman, E. L. Bermingham, and N. M. Kotchian-Prentiss. 1985. Recruitment patterns in young French grunts, Hae- mulon flavolineatum (Family Haemulidae), at St. Croix, Virgin Islands. Fish. Bull., U.S. 83:413-426. McGuRK, M. D. 1984. Ring deposition in the otoliths of larval Pacific herring, Clupea harengus pallasi. Fish. Bull., U.S. 82:113-120. 1985. Effects of net capture on the postpreservation morphometry, dry weight, and condition factor of Pacific herring larvae. Trans. Am. Fish. Soc. 114:348-355. Methot, R. D., Jr. 1983. Seasonal variation in survival of larval northern anchovy, Engraulis mordax. estimated from the age distri- bution of juveniles. Fish. Bull.. U.S. 81:741-750. 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. Middaugh, D. P., R. G. Domey, and G. I. Scott. 1984. Reproductive rhythmicity of the Atlantic silverside. Trans. Am. Fish. Soc. 113:472-478. Nair, R. V. 1959. Notes on the spawning habits and early life history of the oil sardine, Sardinella longiceps Cuv. and Val. Indian J. Fish. 6:342-359. Nishimura, A., and J. Yamada. 1984. Age and growth of larval and juvenile walleye pollock, Theragra chalcogramma (Pallas), as determined by otolith daily growth increments. J. exp. mar. Biol. Ecol. 82:191- 205. Pannella, G. 1971. Fish otoliths: daily growth layers and periodical pat- terns. Science (Wash., D.C.) 173:1124-1127. Parrish, B. B. 1973. Foreward, fish stocks and recruitment. Rapp. P. -v. Reun. Cons. int. Explor. Mer 164:1-3. Raja, B. T. A. 1970. Estimation of age and growth of the Indian oil-sardine, Sardinella longiceps Val. Indian J. Fish. 17:26-42. Ralston, S. 1976. Age determination of a tropical butterflyfish utilizing growth rates of otoliths. Fish. Bull., U.S. 74:990-994. Re, p. 1984. Evidence of daily and hourly growth in pilchard larvae based on otolith growth increments, Sardina pikhardus (Walbaum, 1792). Cybium 8:33-38. Rice, J. A., L. B. Crowder, and F. P. Binkowski. 1985. Evaluating otolith analysis for bloater Coregonus hoyi: do otoliths ring true? Trans. Am. Fish. Soc. 114:532-539. Richer, W. E. 1975. Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Board Can. 191, 382 p. Robertson, A. I., and N. C. Duke. 1987. Mangroves as nursery sites: comparisons of the abun- dance and species composition of fish and crustaceans in mangroves and other nearshore habitats in tropical Austra- lia. Mar. Biol. 96:103-206, Rothschild, B. J., and C. G. H. Rooth (editors) 1982. Fish ecology III - a foundation for REX, a recruitment experiment. Univ. Miami Tech. Rep. 82008. Schmitt, P. D. 1984. Marking growth increments in otoliths of larval and juvenile fish by immersion in tetracycline to examine the rate of increment formation. Fish. Bull., U.S. 82:237-242. Smith, P. E. 1985. Year-class strength and survival of 0-group clupeoids. Can. J. Fish. Aquat. Sci. 42(Suppl. l):69-82. Struthsaker, 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. Taylor, M. H. 1984. Lunar synchronization of fish reproduction. Trans. Am. Fish. Soc. 113:484-493. Theilacker, G. H. 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. Thomas, R. M. 1986. Growth of larval pelagic fish in the south-east Atlantic from daily otolith rings in 1982/83 and 1983/84. S. Afr. J. Mar. Sci. 4:61-77. Tow-nsend, 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 otoUths of the lar- vae of Japanese sea bream Pagurus vfiajor. Bull. Jpn. Soc. Sci. Fish. 48:1559-1562. TSUKAMOTO, K. 1985. Mass-marking of ayu eggs and larvae by tetracycline- tagging of otohths. Bull. Jpn. Soc. Sci. Fish. 51:903-911. Walker, T. A. 1981a. Seasonal salinity variations in Cleveland Bay, north- ern Queensland. Aust. J. Mar. Freshwater Res. 32:143- 149. 1981b. Annual temperature cycle in Cleveland Bay, Great Barrier Reef province. Aust. J. Mar. Freshwater Res. 32: 987-991. Whitehead, P. J. P. 1985. FAO species catalogue. Vol. 7. Clupeoid fishes of the world. An annotated and illustrated catalogue of the herrings, sardines, pilchards, sprats, anchovies and wolf- herrings. Part 1 - Chirocentridae, Clupeidae and Pristigas- tridae. FAO Fish Synop. 7(Pt. l):l-303. Williams, V. R., and T. A. Clarke. 1983. Reproduction, growth, and other aspects of the biology of the gold spot herring, Herklotsichthys quadrimaculatus (Clupeidae), a recent introduction to Hawaii. Fish. Bull., U.S. 81:587-597. Zweifel, J. R., and R. Lasker. 1976. Prehatch and posthatch growth of fishes— a general model. Fish. Bull., U.S. 74:609-621. 83 ENERGETIC AND BEHAVIORAL EFFECTS OF NET ENTANGLEMENT ON JUVENILE NORTHERN FUR SEALS, CALLORHINUS URSINUS Steven D. Feldkamp,' Daniel P. Costa, ^ and Gregory K. DeKrey^ ABSTRACT The energetic costs and behavioral changes associated with net entanglement were studied in three cap- tive juvenile male northern fur seals, Calhrhinus ursirms. Rates of energy expenditure were highly depen- dent upon swim velocity and size of the net fragment. At a speed of 1 . 1 m/s, northern fur seals expended a mean ( ± SD) of 6.5 ( + 0.7) W/kg before entanglement, 9.7 ( ± 3.8) W/kg when entangled in 100 g nets, and 13.8 W/kg with 200 g nets. These results showed that a free-ranging animal entangled in a net frag- ment of 200 g or larger will experience considerable difficulty swimming. The northern fur seals' average daily metabolic rates (ADMR) were measured with doubly labeled water over 6 day periods before and during entanglement in 225 g net fragments. Concurrent behavioral observations revealed a 75% reduction in time spent swimming and a 138% increase in time spent resting due to entanglement. Nevertheless, the northern fur seals' mean ADMR rose from 8.0 (±0.4) W/kg to 9.3 (±1.9) W/kg. While this increase was primarily due to one animal's performance, it suggests that entanglement may also elevate the costs of resting and grooming. At 17 months of age, the northern fur seals had averaged head diameters ( ± SD) of 14.7 ( ± 0.2) cm, making them most susceptible to entanglement in nets with stretched mesh sizes of 23 cm or more. Obser- vations showed that these juvenile fur seals were naturally inquisitive and rapidly became entangled upon their first encounter with a floating net. Subsequent entanglements depended more upon each animal's behavior than upon net fragment size. Captive animals were unable to free themselves from the entangling fragments. Since the mid-1950's, the Pribilof Island population of northern fur seals, Callorhinus ursinus, has undergone several declines. The initial reduction in population size can be attributed to a harvest of adult females, conducted from 1957 through 1968 (York and Hartley 1981). However, from 1974 until 1981, the number of pups born each year continued to decline (Fowler 1985; York and Kozloff 1987). As a result, the present northern fur seal population numbers 800,000 animals, down from an estimated 1.2 miUion in 1976. In the mid-1960's, the percentage of young male northern fur seals found entangled in synthetic trawl net fragments and other marine debris began to rise, reaching a peak of about 0.7% in 1975 (Fowler 1987). Since 1976, the entanglement rate has remained roughly stable at 0.4% of the subadult male population. The northern fur seal population declines, concurrent with the rising entanglement rate, have led some authors to speculate that en- 'Long Marine Laboratory, Institute of Marine Sciences, Univer- sity of California, Santa Cruz, CA 95064; present address: P.O. Box 524, Roseburg, OR 97470. '^Long Marine Laboratory, Institute of Marine Sciences, Univer- sity of California, Santa Cruz, CA 95064. Manuscript accepted October 1988. Fishery Bulletin. U.S. 87:85-94. tanglement may be one contributing factor (Fowler 1985, 1987). Using available data on entanglement rates, net size distribution, and assumed mortality rates, Fowler (1982) derived and demonstrated a model that entanglement induced mortality could account for the current population trends. Although based on several unverified assumptions, it none- theless points to the potential seriousness of net entanglement. Several lines of indirect evidence suggest that en- tanglement related mortality has its greatest impact on younger age classes (less than 2-3 years old). Since 1965, the at-sea survival rate of 0-2 yr old northern fur seals has declined relative to the sur- vival rate of nursing pups on land (Fowler 1985). Prior to 1965, these parameters were positively correlated. Furthermore, this decline in the ex- pected survival rate is correlated with the increased incidence of observed entanglements (Fowler 1985). Working with captive animals, Yoshida and Baba (1985) have also demonstrated that younger animals entangle themselves more frequently than older ones. The impact of entanglement would be more severe on these smaller animals; because of their size, smaller animals will suffer relatively higher drag and greater power requirements during swim- 85 FISHERY BULLETIN: VOL. 87, NO. 1 ming than would larger animals entangled in a similar-sized fragment (Feldkamp 1985). Many questions remain unanswered concerning the impact of marine debris on the demographics of northern fur seals. While it is virtually impossi- ble to directly measure mortality arising from en- tanglement in different age and sex classes, mea- surements can be made of the behavioral changes and energetic costs associated with entanglement, the susceptibility of different age classes to en- tanglement, and the effects of net size on these parameters. In this study we examine the energetic and behavioral costs associated with entanglement. The swimming metabolic rate of three juvenile male northern fur seals entangled in various-sized nets was measured. Changes in behavior and average daily energy expenditure during extended periods of entanglement were quantified. The northern fur seals' responses to floating debris, the likelihood of entanglement, and their ability to free themselves after entanglement were also examined to provide a better understanding of the biological conse- quences of net entanglement in these animals. MATERIALS AND METHODS Three newly weaned male northern fur seal pups (age = 4 months, based on estimated birth date of July 1; Gentry 1981) were captured on St. Paul Island, AK in November 1985. They were trans- ported to the Marine Laboratory, University of California at Santa Cruz and placed in a large hold- ing tank supplied with filtered seawater. Twice per day, the animals were fed a ration of herring sup- plemented with vitamins. These three northern fur seal pups were weighed weekly. Measurements of standard length (nose to tip of tail) and of girth around the head (at ears), neck, and shoulder region were made at several month intervals. Girths were converted to diameters by assuming a circular circumference. Net fragments used in this study were all cut from polypropylene trawl nets found on St. Paul Island, AK. Each fragment had a stretched mesh size of 23 cm (9 in). The twine had a diameter of 3 mm (Yg in). Swimming Energetics The energetic cost of swimming, before and dur- ing entanglement, was measured by placing the northern fur seals in a water flume constructed in- side of a circular tank, 7.6 m in diameter and 2.7 m deep. A wooden ring (4.9 m in diameter and 1.2 m in height) was placed in the tank, forming a 1.3 m wide channel between it and the tank wall. A water current was generated inside this channel with two pumps. The first, a 15 hp pump, was sub- merged to a depth of 48 cm and produced a flow of 0.75 m/s. The second, a 10 hp nonsubmersible pump, was located above the tank with its intake and outlet hoses fixed in the channel; this pump could generate flows of 0.6 m/s. Run simultaneously, the two pumps created flows of 1.1 m/s. The fur seals swam inside of a metabolic test sec- tion (2.2 m in length, 1.1m wide, 0.9 m deep) con- structed in the channel. The walls of the tank and inner ring formed its sides. Front and back ends were framed with wood and covered with 8 cm x 13 cm mesh wire screen. A sheet of plywood cov- ered the top. A plastic dome (0.9 m x 0.6 m x 0.3 m) was set into this plywood and served as an open circuit metabolic chamber. Animals in the test sec- tion could only surface to breathe inside of the dome. To minimize turbulence, the test section was located approximately 7.5 m away from the outflow of the pumps, along the tank's circumference. Water velocity was measured with a General Oceanics Model 2035 MKIIP flow meter, accurate to ±3%. All flow measurements were made in the test sec- tion, 50 cm from the floor and 90 cm from the front screen. Air was drawn through the metabolic dome at a rate of 20 L/min. Oxygen content of the air was measured with an Ametek oxygen analyzer cali- brated using the methods of Fedak et al. (1981). A computer monitored the analyzer's output each sec- ond and produced a 1 min average of the percent 0, concentration. Oxygen consumption (VO2) was calculated using equation 11 of Fedak et al. (1981). Every 10 minutes, these minute readings were averaged to provide single data points at each swim- ming speed. All values were corrected to STPD, and VO2 (in mL Oo • min "' • kg"') was converted to W/kg by assuming a caloric equivalent of 20. 1 J/mL O2 (Bartholomew 1977). Prior to the actual measurements, the three north- ern fur seals were trained for several weeks to swim in the flume. Training was considered complete when consistent values for VO2 were obtained at each speed. During experiments, 12 h fasted animals were placed in the flume and allowed to rest, groom, or swim at their own speed for approximately 15 minutes while VO2 was monitored. The first water pump was then turned on and water velocity main- tained at 0.75 m/s. Ten minutes were allowed for ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA, 86 FELDKAMP ET AL.: NET ENTANGLEMENT ON NORTHERN FUR SEALS animals to reach a steady state and then VOg mea- surements were made for 20-30 minutes. The sec- ond pump was then activated, creating a water speed of 1.1 m/s, and VO2 was monitored as de- scribed above. Only those trials where steady swim- ming occurred were used in the final calculations of metabolism. Experiments that reversed the order of swimming speeds revealed no differences in VO2 that could be attributed to the effects of order- ing. Each fur seal was run once per day over a 2 wk period until baseline values were established. Fur seals were then entangled in small net fragments and the measurements were repeated. All three ani- mals were run daily, and trials with one net size were completed before beginning the next size. For each net size, experimental trials lasted approx- imately 2 weeks. Nets were attached by placing the fur seal's head through a wire ring, 15 cm in diameter, sewn into the center of each fragment. Three sizes of net were used in the following order: 61, 100, and 200 g with dimensions of 4 x 4 meshes (0.6 m x 0.6 m), 7 x 4 meshes (1.2 m x 0.6 m), and 7x7 meshes (1.2 m X 1.2 m), respectively. Nets were folded over once (100 g), or twice (200 g) to prevent fouling of the foreflippers, and were removed after each session. Experiments were run both in the winter and spring. Winter experiments were conducted at a single flume speed of 0.75 m/s using 61 g and 100 g nets. During the spring, 100 g and 200 g nets were used at flume speeds of 0.75 m/s and 1.1 m/s. At the completion of the spring trials, baseline values were again established through daily runs conducted without nets. Behavioral and Energetic Changes Associated with Entanglement In this study, the effect of entanglement on the northern fur seals' behavior and average daily meta- bolic rate (ADMR) was measured. The fur seals were 14 months old and weighed an average ( ± SD) of 19.1 (±1.0) kg. They were kept in a circular hold- ing tank, 7.6 m in diameter and 1 m deep with no haul-out provided. Two experiments were undertaken. First, free- swimming fur seals were monitored over a 6 d period. They were then entangled in 225 g net frag- ments and the measurements were taken for another 6 days. A nylon dog collar, sewn into the middle of each fragment, was used to fasten the net around the animal's neck. The ADMR of each fur seal was determined before and during entanglement using isotopic tracers (Nagy 1980; Schoeller and van Santen 1982; Costa and Gentry 1986). Prior to each experiment, blood samples were taken, and then the animals were in- jected interperitoneally with 5.5 mL of 0.66 mCi tritium (HTO) per mL and 2.5 g of Hji^O at 95 atoms percent. After equilibration (3 hours), a 10 cc blood sample was taken from a flipper vein and the animal was released into the tank. After the 6 d measurement period, fur seals were removed from the holding tank, and final blood samples were taken. Tritium specific-activity in water that was vaccum distilled from blood samples was determined by liquid scintillation spectrometry. Oxygen-18 levels were measured by isotope mass ratio spectrometry in a commercial laboratory (Global Geochemistry, Canoga Park, CA). Rates of CO, production (VCO2) were calculated using equation 2 in Nagy (1980), and water flux rates determined using equation 4 in Nagy and Costa (1980). We assumed an RQ of 0.80 to calculate energy consumption. The northern fur seals' behavior over the course of each study period was quantified using a discon- tinuous time sampling method (Tyler 1979). Every hour, from 0800 to 2000, the fur seals were observed for 10 minutes. At exactly 1 min intervals during this period, the behavior displayed by each fur seal was noted. Behaviors were broken into four cate- gories: swimming, grooming, resting, and other ac- tivities. Animals were considered to be swimming when they were actively stroking or gliding between strokes. Grooming was defined as scratching, rub- bing the fur, or shaking the head. Animals at rest were lying quietly, often holding their flippers out of the water. Activities such as rolling, nuzzling one another, or other slow movements were placed in the "other" category. Entanglement Observations The reactions of northern fur seals to the pres- ence of floating nets, their ability to free themselves after entanglement, and the likelihood of entangle- ment in net fragments of various sizes were inves- tigated. Two fur seals were placed in a 7.6 m diam- eter holding tank, 1 m deep, along with floating net fragments of various sizes, and were denied access to haul-out areas during this time. The time from net presentation to entanglement was recorded and correlated with fragment size. Once entangled, nets were left on for periods ranging from several hours to several days. 87 FISHERY BULLETIN: VOL. 87, NO. I RESULTS Measurements of the northern fur seals' length, mass, and body diameters as a function of age are presented in Table 1. Table 1.— Age', weight, length, and body diameters of the three northern fur seals used in this study. Means ( + SD) are given. Age Mass Length Diameter ( ;cm) at (mo) (kg) (cm) ears neck shoulders 4 14.4 87.7 13.5 14.9 21.1 (2.02) (1.56) (0.15) (0.40) (1.02) 12 19.6 102.3 14.3 15.8 21.4 (1.21) (1 16) (0.00) (0.46) (1.25) 17 21.3 107.7 14.7 16.8 22.9 (0.70) (2.08) (0.23) (0.51) (0.30) 'Age based on estimated birth date of July 1 Energetic Measurements During the winter swimming trials, mean water temperature ( ± SD) was 14.9 ( ± 0.2) °C. Mean body mass of the three northern fur seals was 14. 1 ( + 0.6) kg. At zero water flow, fur seals expended a mean of 6.95 (±1.02) W/kg; at 0.75 m/s it was slightly, though not significantly, lower and averaged 6.89 ( ± 0.45) W/kg (Table 2). The greater metabolic rate at zero flow was due to uncontrolled activity (swim- ming and grooming) inside the chamber. At 0.75 m/s the animals swam steadily and did not groom. At zero water flow, the presence of a net also slightly increased each fur seal's metabolic rate (Fig. 1). Although there was no evidence of any behav- ioral change, the net may have caused slight stress and led to an elevated 0^ consumption. Additional- ly, there may have been a loss of air from the pelage in the region of the net allowing the infiltration of cold water. This rise was noted both in the winter and spring experiments (Table 2). At the relatively slow speed of 0.75 m/s, small net fragments did not significantly elevate mean meta- bolic rates (Fig. lb). In fact, metabolism was 3.5% lower with a 61 g net at 0.75 m/s than at zero flow (Table 2). While a slight elevation did occur with the 100 g net at 0.75 m/s, this was not significantly greater than at zero flow with a 100 g net or than at 0.75 m/s with no net (Fig. lb). During the spring experiments, water tempera- ture had increased to a mean of 16.6 ( + 0.5) °C, and the fur seals had increased in weight to 16.4 ( + 0.5) kg (Table 2). As a result, each animal's routine energy consumption, determined at zero flow and without a net, had declined to a mean of 5.22 ( + 0.21) W/kg. This decline was also evident when the results of the winter and spring experiments at 0.75 m/s and with 100 g net fragments were compared (Table 2). As noted during the winter trials, animals with 100 g nets did not expend greater amounts of energy at 0.75 m/s than during unentangled swimming at this speed. At 1.1 m/s, however, a 100 g net caused a significant 40% increase in metabolic rate (^test; P < 0.05) (Fig. Ic). The 200 g fragments caused significant metabolic increases at both flume speeds. When entangled in a fragment of this size, fur seals expended 66% more energy at 0.75 m/s than at zero flow speed. At 1.1 m/s, mean metabolic rate was elevated 2.1 times that measured for unentangled Table 2.— Mean ( + SD) rate of energy utilization (W//kg) for the three northern fur seals during the winter and spring swimming experiments. Means were determined by combining data from all animals. The sample size for each animal (FS1. FS2. FS3) is given in parentheses (n,, n^, and n^. respec- tively). Body mass and water temperature averaged 14.1 ( + 0.55) and 14.9 ( + 0.2) °C in the winter, and 16.4 (±0.5) kg and 16.6 (±0.5) °C in the spring. Winter Spring No net Net size No net Net size Flume speed 61 g 100 g 100 g 200 g m/s 6.95 8.26 7.78 5.22 610 5.90 + 1.02 0.91 0.71 0.21 0.85 0.78 (n,, nj. nj) (8, 7, 6) (6, 5, 5) (5, 6, 5) (6, 6, 5) (4, 4, 4) (6, 7, 6) 0.75 m/s 6.89 7.98 8.40 5.21 6.89 8.63 + 0.45 1.87 1.65 0.32 0.42 1.52 (n,, n2, nj) (8, 7, 6) (5, 6, 5) (6, 5, 5) (5, 5, 5) (4, 4, 4) (6, 7, 6) 1.1 m/s 6.54 9.68 13.83 -t- 0.72 3.79 1.27 (n,, nj, nj) (5. 5, 5) (6, 4, 4) (6, 7, 6) 88 FELDKAMP ET AL.: NET ENTANGLEMENT ON NORTHERN FUR SEALS 1a 15 -- 100 NET SIZE (q) -r 15.0 -- 10.0 -- 5.0 200 Figure 1.— VOo (mL 0, min"' kg"') and energy expenditure (W/kg) of three northern fur seals plotted as a function of net fragment size, a) Measurements at zero flow speed, b) measurements at 0.75 m/s. and c) measurements at 1.1 m/s. Open symbols and dashed lines are measurements conducted during the winter; solid symbols and lines are from the spring. 89 FISHERY BULLETIN: VOL. 87, NO. 1 swimming at this speed, and 2.7 times that mea- sured at zero flow (Fig. Ic). All animals with 200 g nets struggled against the flow at 1.1 m/s. Behavioral and Energetic Changes Associated with Entanglement Before entanglement, swimming was the predom- inate behavior of the northern fur seals (Table 3). Grooming was the next most frequent activity, while resting accounted for 18% of the total activity. After entanglement in a 225 g fragment, the fur seals' behavior was substantially altered. Time spent swimming declined by roughly 75% from control measurements, while the percent time spent resting increased by a factor of 2.4. Time spent grooming, however, was not significantly altered; fur seals con- tinued to spend approximately 1/3 of their daylight hours engaged in this activity (Table 3). Table 3.— Percentage of time spent by three northern tur seals at swimming, resting, grooming, or in other activities over 6 d intervals before and during entanglement in a 225 g net. Before entanglement During entanglement Swim Rest Groom Other Swim Rest Groom Other Animal FS1 FS2 FS3 Mean ( + SD) 36,2 49.2 44.3 43.3 (6.6) 23.4 33.4 15.0 29.4 15.7 28.5 18.0 30.4 (4.7) (2.6) 7.0 6.4 11.5 8.3 (2.8) 6.1 21.0 5.2 10.8 (8.9) 55.0 36.1 37.6 42.9 (10.5) 28.4 38.4 306 32.5 (5.3) 10.5 4.5 26.6 13.8 (11.4) ment was slightly greater than in the control experi- ment, although this difference was not significant. While mean ADMR increased by 16% after en- tanglement (Table 4), this increase was primarily due to one animal (FS2). FS2 swam more than either of the other two animals both before and during en- tanglement (Table 3), and this undoubtedly led to a higher ADMR. No consistent trend was observed for FSl and FS3 with respect to altered energetic expenditures associated with entanglement. FSl's metabolic rate increased 1 W/kg above the control measurement. FS3 showed a decrease of roughly the same amount (Table 4). FSl exhibited an increased ADMR even though it spent less time swimming and grooming during the entanglement period. While FS3 showed similar behavioral trends, its ADMR decreased dur- ing entanglement. Table 4.— CO^ Production fVCOj in L COj h ' kg " ') and rate of three northern fur seals' energy expenditure' (W/kg) determined by the doubly labeled water method. Measurements were made over 6 d intervals before and during entanglement in a 225 g net. Before During Animal Mass VCOj W/kg Mass VCOj W/kg change FSl 19.9 1.12 7.79 20.5 1.28 8.90 + 14 FS2 17.8 1.12 7.81 18.6 1.63 11,34 + 45 FS3 18.9 1.22 8.51 18,9 1.08 7.50 -12 Mean 18.7 1.15 8.04 19.3 1.33 9.25 ( + SD) (1.05) (0.06) (0.41) (1.02) (0.28) (1.94) 'CalculatecJ using an RQ of 80 The nature of the resting and grooming behaviors of fur seals appeared to change as a result of en- tanglement. Unentangled fur seals typically rested quietly on the surface in a "jug-handle" position with one foreflipper and both rear flippers held out of the water. Entangled fur seals most often rested with their ventral surfaces down and both fore- and rear flippers submerged. In this position, they had to lift their heads to breathe. Resting was also not com- pletely quiet. Unlike before, fur seals changed posi- tions frequently and were always alert in the pres- ence of observers. When grooming, fur seals spent more time vio- lently shaking their heads and scratching at the net vdth both their rear and foreflippers than before en- tanglement. Grooming was less vigorous in the absence of a net. More time was spent slowly roll- ing around in the water, or passively moving about (Table 3). The rate of energy expenditure during entangle- Entanglement Observations Individual variations in behavior were the most important factors influencing whether entanglement occurred (Table 5). For example, FSl became en- tangled 1 hour and 45 minutes after it was first presented with a 6 x 4 mesh net (0.9 m x 0.6 m). From presentation to entanglement, the net was the focus of the animal's attention. FSl bit and pulled at it, laid beneath it, and often rested on top of it. Once entangled, FSl became quite agitated and swam around vigorously while violently shaking its head. The holding tank was drained after 1 hour and the net was removed from its neck. Another net (4 X 4 meshes; 0.6 m x 0.6 m) was then introduced into the holding tank and FSl ignored it complete- ly throughout the rest of the day. By the following morning, FSl had become entangled again. This net was left on the animal for 2 days. At no time did it appear that FSl could free itself, and the net was subsequently removed. Another net was then 90 FELDKAMP ET AL.: NET ENTANGLEMENT ON NORTHERN FUR SEALS introduced, and this time it was completely ig- nored. After 8 hours, the net was removed from the tank. In contrast to FSl's apparent increased wariness to floating nets after its first entanglement, FS3 became entangled almost immediately on every net presentation (Table 5). Upon encountering the first net, FS3 played with it constantly, exhibiting similar behaviors as shown by FSl. Within 20 minutes FS3 had entangled itself. This net was left on for 2 days without the animal freeing itself and was then removed by hand. Approximately one week later, FSB was presented with a second, smaller net, and it quickly became entangled. Rather than remove this net, another small net was introduced into the tank. Within 30 minutes, FS3 became entangled in this as well. The tank was subsequently drained and both nets were Table 5- — Entanglement observations of two captive northern fur seals held In a 7,6 m diameter, 1 m deep circular holding tank. The net used had a 23 cm stretched mesh size. Net size Animal Date (meshes) Time to entanglement FSl 2/15 6x4 1 h 45 min' 2/15 4x4 <20 h^ 2/17 6x4 No entanglement in 8 h 2/20 6x4 2 h 15 min 2/20 6x4 <6 h FS3 1/29 6x4 23 min' 2/11 4x4 14 min^ 2/11 4x4 28 min 2/11 6x4 8 min^ 2/11 4x4 37 min (1)^ (two nets) 30 min (2) 2/12 6x4 36 min 'First presentation of a net, ^Net left floating in water over nigfit 3Net left on animal when next one was introduced into the tank removed. The tank was then refilled and a net im- mediately placed in the water. Within 10 minutes the animal became entangled. Two additional nets were then placed in the water. Within 1 hour, FS3 had become entangled in these as well. All three nets were then removed from the animal's neck. From these and from subsequent observations (Table 5), it is unclear whether net fragment size influences the probability of entanglement. FS3 became entangled almost immediately following every net presentation, regardless of the net's size. FSl, however, seemed to be wary of nets following its first entanglement. This wariness appeared to subside after several days without encountering a net. Video recordings made of FS3 documented the behaviors preceding and following an entanglement. While playing with the net, FS3 approached from below and began to inspect it. After several seconds, FS3 pushed its nose up through a mesh opening in an apparent attempt to breathe. Immediately upon sensing the net around its muzzle, FS3 attempted to free itself by quickly shaking its head from side to side. This served to pull the net further down its head and, within several seconds, meshes were tightly wrapped around the animal's neck. During this time, and immediately following, FS3 became extremely agitated. It continued its quick and vio- lent headshaking, but also began to swim rapidly around the tank, porpoising frequently. This action undoubtedly caused the net to be pulled even fur- ther down its neck. After approximately 2 minutes, swimming slowed, but FS3 continued to stop and shake its head violently. Approximately 5 minutes later, FS3 had several meshes looped over its head and neck. FS3's eating ability was not impaired by the net, and so it was left on for 2 days. There was never any indication that the animal would be able to free itself and the net was finally removed. Although the net was so tightly wrapped around the animal's neck that it had to be cut off, there was no evidence of abrasions or lacerations. DISCUSSION In a recent review. Fowler (1987) suggested that younger northern fur seals are more prone to en- tanglement related mortality than are older animals. Results from the present study help to shed light on possible reasons for these apparent age related discrepancies. Small physical size and the inquisi- tive nature of juvenile animals are likely to be the two important factors leading to a higher mortal- ity from entanglement. Naive animals may become entangled with greater frequency than older, per- haps less inquisitive animals, and smaller animals have the potential to become entangled in a greater range of mesh sizes. Moreover, once entangled, relative swimming costs will be higher for smaller animals (Feldkamp 1985). The majority of nets found on two Alaskan islands (St. George and Amchitka) had stretched mesh sizes of 20 cm or less, with a mode of 10-15 cm (Fowler 1987). If this is representative of material adrift at sea, then most fragments have a small mesh size. It seems reasonable to conclude, therefore, that relative to larger animals, a greater number of frag- ments exist that are potentially hazardous to smaller 91 FISHERY BULLETIN: VOL. 87. NO. 1 animals. Young fur seals are most often found in trawl net fragments having a stretched mesh size of 20 cm or more, with 23 cm mesh observed most frequently (Scordino 1985). The diameter of a 23 cm mesh is 14.6 cm, almost exactly the average head diameter of the captive 17 mo old northern fur seals (Table 1). Similarly, a 20 cm mesh net has a circular diameter of 12.7 cm. Although this is slightly smaller than the head diameter of captive 4 mo old fur seals, it may well pose an entanglement threat for smaller animals. At 17 months of age, the captive northern fur seals had average shoulder diameters of 23 cm. A 23 cm mesh net would therefore lodge tightly around the neck region but would not slip further down the body. Based on these dimensions, a net would have to have a stretched mesh size of 73 cm or more before a fur seal of this age could pass through a single mesh opening. Scordino (1985) has shown that most webbing found on young seals weighs less than 150 g. He sug- gested that the high incidence of small debris en- tanglement may be due to the seals "playing" with small pieces of debris, as they do with kelp. This sug- gestion is supported by our observations of the fur seals' investigative nature when presented with net fragments. Prior to their first entanglement, all animals showed an immediate interest when they encountered a floating net and played with it almost continuously until they became entangled. While it is difficult to draw conclusions about the behavior of northern fur seals at sea from studies of a small number of captive animals, our observations none- theless suggest that young fur seals are naturally inquisitive. Interestingly, however, these captive animals appeared indifferent to other floating ob- jects (plastic bats, frisbees) that were occasionally placed in their tank. Scordino's (1985) observations may also reflect a high incidence of at-sea mortality caused by en- tanglement in fragments larger than 150 g. Starva- tion, resulting from an increased energy demand during swimming, may be one consequence of en- tanglement in larger fragments. Previous studies have shown that entangled animals experience greater drag during swimming and that this drag increases exponentially with swim velocity and with greater net size (Feldkamp 1985). Because swim- ming energy requirements increase in relationship to drag, it was expected that metabolic increases would parallel increases in drag. Results from the swimming experiments support these predictions (Fig. 1; Table 2). At slow speeds, and with small (61 g and 100 g) nets, metabolism did not differ signif- icantly from that measured at zero flow. At the higher speed of 1.1 m/s, metabolism was signifi- cantly elevated by both the 100 g and 200 g nets. With a 200 g net, animals visibly struggled against the 1.1 m/s flow. On several occasions, the experi- ment had to be stopped because of the fear of in- jury to the animal. Metabolic rates at zero flow speeds and at 0.75 m/s were also higher during the winter experiments. This may be accounted for by differences in water temperature and body size. Miller (1978) has shown that the metabolic rate of northern fur seals in- creases linearly with decreasing water temperature. In 15°C water, animals in Miller's study had a meta- bolic rate of about 6.8 W/kg, close to the 6.95 W/kg (Table 2; no net, zero flow) measured for animals in this study. Under similar conditions during the spring, when water temperature had increased by 1.7°C, our measurements showed a 25% reduction in metabolic rate (Table 2). This reduction in metabolic rate during the spring experiments was also observed for swimming and entangled animals. At 0.75 m/s in the winter, a 100 g net resulted in an average metabolic rate of 8.4 W/kg (Table 2). Under similar conditions during the spring, metabolism had dropped by 18%. Although the reasons for these metabolic changes are diffi- cult to interpret, given the small sample size and changes in body mass, they do suggest, as did Miller (1978), that water temperature is an important fac- tor influencing the energetic demands of swimming juvenile northern fur seals. Metabolic rate measurements suggest that if juve- nile northern fur seals become entangled in nets of 200 g or more, they will experience considerable dif- ficulties in swimming and likely suffer a greater mortality than unentangled animals. Although our measurements were conducted over relatively slow swimming speeds, they do provide a basis of esti- mating the impact of entanglement on the energetic requirements of animals at sea. If animals with 200 g net fragments maintained an average speed of 1.1 m/s over the course of a day, they would need to consume 284 kcal of fish/kg body mass to maintain body weight. Using data on the caloric density of pollock (1.4 kcal/g) and on fur seal assimilation effi- ciencies (Miller 1978), this energetic requirement equals roughly 5 kg of pollock per day, compared with 1.9 kg for an unentangled animal. While it is likely that entangled fur seals would not swim con- stantly at sea, they may have to reach swim speeds higher than 1.1 m/s in order to catch prey, thereby increasing their metabolic expenditures. Moreover, water temperature of the Bering Sea is consider- 92 FELDKAMP ET AL.: NET ENTANGLEMENT ON NORTHERN FUR SEALS ably colder than that during the swimming trials. For these reasons, this value should be viewed as a minimum estimate of the energy required for sur- vival by a juvenile fur seal entangled in a 200 g net. The elevated swimming costs associated with en- tanglement and the resultant rise in food require- ments suggest that northern fur seals enter a vicious cycle when entangled in larger fragments. As swim- ming costs increase, so will food demands. The need to capture more prey requires more swimming. Greater drag and perhaps reduced aquatic agility will undoubtedly lower capture success. Under these conditions, starvation would be a likely outcome. The observation that northern fur seals virtually stopped swimming when entangled in 225 g nets is consistent with this scenario. By reducing the time spent swimming, fur seals should lower their ener- getic expenditures and hence their energy require- ments. However, ADMR measurements before and during entanglement showed no significant differ- ences. Since swimming activity declined by 75%, it is possible that the costs of resting and grooming were elevated by entanglement. A larger sample size would be needed to verify these findings. None- theless, grooming appeared to be much more vig- orous and resting was not completely quiet. The fur seals often rested with both foreflippers submerged in the water, which may have elevated heat loss to the environment and led to greater energy require- ments. Fur seals were unable to free themselves from en- tanglement during the 2-3 d periods (Table 5). These results, however, must be interpreted with caution. Because fur seals were confined to a round holding tank with no haul-out areas provided, there were no objects present that might have caught the net and might have been used to remove it. From our observations, it is doubtful that animals could have freed themselves. However, Scordino (1985) has documented several instances where wild fur seals have lost their nets. It is possible that under natural conditions, fur seals might snag the encumbering fragment on rocks or other objects and be able to pull free. The results of the present study show that juve- nile northern fur seals are susceptible to, and adversely impacted by entanglement. Our captive fur seals were highly inquisitive and usually inves- tigated and played with floating nets. Measurements of their head, neck, and shoulder diameters indicated that they were most susceptible to entanglement in nets with mesh sizes of 23 cm or more. Observations of actual entanglements substantiated this finding. Once entangled, northern fur seals virtually stopped swimming and spent considerably more time rest- ing. However, energy expenditure did not drop ac- cordingly, suggesting that the energy expended for grooming or resting may have been elevated by the presence of a net. Direct measurements also showed that at zero swimming speed, oxygen consumption was slightly, though not significantly, elevated because of the net. This elevation increased both with the size of the net and with increasing swim- ming speed. It is evident from these findings that net fragments of 200 g or more can lead to signif- icant behavioral changes in captive northern fin- seals and greatly influence their energy require- ments during swimming. ACKNOWLEDGMENTS We would like to thank B. Fadely, A. C. Huntley, and D. Murnane for help in all phases of this work. We are especially grateful to S. Beltramy, J. Burger, D. Busch, J. Lajala, and P. Wolfe for their assistance in feeding and maintaining the northern fur seals. We also thank R. L. DeLong for help in the capture and transport of the animals, R. L. Gentry and 2 anonymous reviewers for useful comments on the manuscript, and D. Sims for technical assistance. This work was supported by NOAA contract 85- ABC-00185 from the National Marine Mammal Laboratory. LITERATURE CITED Bartholomew, G. A. 19T7. Energj' metabolism. In M. S. Gordon (editor). Animal physiology: Principles and adaptations, p. 57-110. Macmil- lan Publ. Co., N.Y. Costa, D. P., and R. L. Gentry. 1986. Free-ranging energetics of northern fur seals, hi R. L. Gentry and G. L. Kooyman (editors). Fur seals. Mater- nal strategies on land and at sea, p. 79-101. Princeton Univ. Press, Princeton, NJ. Fedak, M. a., L. Rome, and H. J. Seeherman. 1981. One-step N2-dilution technique for calibrating open- circuit VOj measuring systems. J. Appl. Physiol.: Respir. Environ. E.xercise Physiol, 51:772-776. Feldkamp, S. D. 1985. The effects of net entanglement on the drag and power output of a California sea lion. Zalopkus califomianus. Fish. Bull., U.S. 83:692-695. Fowler. C. W, 1982. Interactions of northern fur seals and commercial fish- eries. Trans. N. Am. Wildl. Nat. Resour, Conf. 47:278-292. 1985. An evaluation of the role of entanglement in the popu- lation dynamics of northern fur seals on the Pribilof Islands. In R. S. Shomura and H. 0. Yoshida (editors). Proceedings of the workshop on the fate and impact of marine debris, 27-29 November 1984, Honolulu, Hawaii, p. 291-307. U.S. Dep. Commer., NOAA Tech. Memo. NMFS. NOAA-TM- 93 FISHERY BULLETIN: VOL. 87. NO. 1 NMFS-SWFC 54. 1987. Marine debris and northern fur seals: a case study. Mar. Pollut. Bull. 18:326-335. Gentry, R. L. 1981. Northern fur seai-Callorhinus ursxnus. In S. H. Ridgway and R. J. Harrison (editors), Handbook of marine mammals. Volume 1 , The walrus, sea lions, fur seals and sea otter, p. 143-160. Acad. Press, Lond. 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.. Wash., D.C., Rep. MMC-75/08, 27 p. Nagy, K. 1980. CO2 production in animals: Analysis of potential errors in the doubly labeled water method. Am. J. Physiol. 238: R466-R473. Nagy, K., and D. P. Costa. 1980. Water flux in animals: Analysis of potential errors in the tritiated water technique. Am. J. Physiol. 238:R454- R465. SCHOELLER, D. A., AND E. VAN SaNTEN. 1982. Measurement of energy expenditure in humans by the doubly labeled water method. J. Appl. Physiol. 53:955-959. SCORDINO, J. 1985. Studies on fur seal entanglement, 1981-84, St. Paul Island, AK. In R. S. Shomura and H. 0. Yoshida (editors). Proceedings of the workshop on the fate and impact of marine debris, 27-29 November 1984, Honolulu. Hawaii, p. 278-290. U.S. Dep. Commer., NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SWFC-54. Tyler, S. 1979. Time-sampling: A matter of convention. Anim. Behav. 27:801-810. 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. York, A. E., and P. Kozloff. 1987. On the estimation of numbers of fur seal pups born on St. Paul Island, 1980-85. Fish. Bull., U.S. 85(2):367-375. YOSHIDA, K., AND N. BaBA. 1985. The problem of fur seal entanglement in marine debris. In R. S. Shomura and H. 0. Yoshida (editors). Proceedings of the workshop on the fate and impact of marine debris, 27-29 November 1984, Honolulu, Hawaii, p. 448-452. U.S. Dep. Commer., NOAA Tech. Memo. NMFS, NOAA-TM- NMFS-SWFC-54. 94 AN EXPERIMENTAL TRANSPLANT OF NORTHERN ABALONE, HALIOTIS KAMTSCHATKANA, IN BARKLEY SOUND, BRITISH COLUMBIA B. Emmett' and G. S. Jamieson^ ABSTRACT The biological and economic feasibilities of transplanting northern abalone, Haliotis kamtschcdkanH Jonas 1845, from exposed beds to two sites in sheltered, productive abalone habitat were investigated. After nine months, 39% and 72% of transplanted abalone were recovered at the two replicate sites. Recovery of tagged abalone at a control site, situated in the exposed source area, was 32%. Growth in shell length of transplanted abalone over the nine months averaged 7.8% whereas the average growth of non- transplanted controls was 3.7%, significantly less. There was little emigration of abalone from the transplant sites. The study concludes that it is feasible to transplant 50-100 mm H. kamtschatkaiM in order to enhance growth. The economic feasibility of transplants is dependent on site-specific recovery rates and the costs of harvesting seed abalone. The population dynamics of abalone in exposed beds and the long-term poten- tial for enhancing abalone settlement by introducing broodstock to depleted areas are two aspects which now require investigation. The northern or pinto abalone, Haliotis kamtschat- kana Jonas 1845, ranges from San Diego, CA to Sitka, AK (Mottet 1978); is most abundant in British Columbia and southeast Alaska; and is the only species of abalone found in British Columbia. Al- though present in the low intertidal zone in the northern part of its range, northern abalone are nor- mally foimd subtidally to depths of 15 m (Cox 1962). In British Columbia the species is common in rocky habitats associated with surface kelps (Macroqjstis integrifolia and Nereocystis lv£tkeana) at depths of 3-7 m. In 1976 a market for Canadian abalone developed in Japan, and annual landings increased from less than 50 1 (Farlinger and Bates 1985) to 425 1 by 1978 (Breen 1980). Attempts were made to reduce the catch through effort control and the imposition of annual catch quotas. Despite these harvest restric- tions, the northern abalone population in British Columbia has been extensively depleted and recruit- ment of legal-sized (>100 mm) abalone to the fishery is low (Breen 1980; Boutillier et al. 1984, 1985). Although unharvested beds of legal size northern abalone are now uncommon, sublegal size abalone are often abundant in exposed habitats adjacent to once-productive commercial grounds. These smaller 'Archipelago Marine Research 11, 1140 Fort Street, Victoria, British Columbia V8V 3K8, Canada. ^Department of Fisheries and Oceans, Fisheries Research Branch, Pacific Biological Station, Nanaimo, British Columbia, V9R 5K6. Canada. northern abalone are referred to as "surf" abalone by fishermen. They most often occur in beds of Pterygophora califomica or under Laminaria set- chellii cover. Breen (1980) estimated mean popula- tion densities of 9.5 abalone m"- in seven beds of Pterygophora and 1.1 abalone m'- in 20 beds of canopy-forming Afacroci/siis. However, only 3% of the abalone in the Pterygophora habitat were of legal size as compared with 46% in Macrocystis habitat. In exposed areas, northern abalone may be slow- growing and never reach legal size because of food limitation. Alternatively, these northern abalone may grow at normal rates but experience high rates of mortality, or emigrate to other habitats. Breen (1986) transplanted 617 sublegal size H. kamtschatkana from exposed habitat in the Queen Charlotte Islands to a more sheltered Macrocystis community. Recovery after one year was 10%, and the author concluded that growth of these "surf" abalone was enhanced when transplanted to more favorable habitat. The present study examines the feasibility of transplanting large numbers of sublegal size northern abalone from an exposed area to more sheltered habitats. Specific goals were 1) to deter- mine the growth of transplanted individuals relative to nontransplanted controls, 2) to monitor the recovery of northern abalone in transplant and con- trol areas after approximately one year, and 3) to assess the economic feasibility of transplanting sublegal size northern abalone for subsequent com- mercial harvest. Manuscript accepted August 1988. Fishery Bulletin, U.S. 87:95-104. 95 FISHERY BULLETIN; VOL. 87, NO. 1 MATERIALS AND METHODS Study Sites Study sites were located in Barkley Sound on the west coast of Vancouver Island (Fig. 1). Sublegal size northern abalone were removed from 5 km of exposed shoreline at the entrance to Barkley Sound (source area), and these abalone were transplanted to site A on Fleming Island and site B on Tzartus Island, 10-12 km towards the head of Barkley Sound from the source area. For the purposes of this study, sublegal size northern abalone are defined as 50-99 mm in length. These individuals should recruit to the fishery within 0-3 years, given suitable habitat. An exposed rocky pinnacle (site C) within the source area was designated as a control site to measure growth and recovery of nontransplanted northern abalone. The three study sites were iso- lated by natural features (e.g., sand) from nearby abalone habitat to minimize immigration or emigra- tion. The source area consisted of a series of rocky headlands and bays. The habitat of the headlands and control site was typical of exposed rocky out- crops on the west coast of Vancouver Island. At the control site, a 2 m band of vegetation {Lessoniopsis littoralis and Laminaria groenlandica) formed the lower intertidal zone, and northern abalone and sea urchins, Strongylocentrotiisfranciscamis, occurred below this zone on a rocky reef dominated by en- crusting coralline algae. Bays in the source area were sloped less steeply and contained beds of Nereocystis luetkeana and Pterygophora califomica. Transplanted northern abalone were collected from both headland and bay habitats in the source area. Both transplant sites were located in and direct- ly below beds of Macrocystis integrifolia situated on isolated rocky reefs. Sites were defined by mark- ing 60 m wide x 8 m deep sections of these beds with a weighted line at each lateral boundary. The Macrocystis bed at site A was 2-5 m wide, bordered by a deeper 3 m wide band of brown algae, Des- marestia ligulata. The substrate at this site was steeply sloped bedrock. Large boulders, covered by P. califomica and encrusting coralline algae, oc- curred at the base of the bedrock slope. Sea urchins (S. frandscanus and S. purpuratus) occurred below the vegetation zone to a depth of 8 m. At deeper depths the bottom was composed of sand, isolated cobbles, and boulders. At site B, the Macrocystis zone was 6-8 m wide and bounded at the lower edge by kelps (Laminaria satcharina, P. califomica, and Agarum fimbria- tum). Desmarestia ligulata, although present, did not form a distinct zone as at site A. The bedrock substrate was sloped less steeply than at site A and was overlain with loose cobbles. At deeper depths, sand was the primary substrate. As at site A, sea urchins were present below the vegetation zone. Figure 1.— Location of the study sites in Barkley Sound on the west coast of Vancouver Island. 96 EMMETT and JAMIESON: TRANSPLANT OF NORTHERN ABALONE Removal, Tagging, and Transplanting Prior to transplanting any northern abalone, divers searched for and removed indigenous abalone from the transplant sites. The divers then collected the abalone from the source area using a dull knife or the arm of the sea star, Pycnopodia heliantJuyides. The arms elicit an escape response, which allows the abalone to be hand-picked without damaging the foot. Tagging and transplanting were conducted from mid-June to July 1984. Approximately 20% of the northern abalone, selected haphazardly, were tagged with individually numbered, stainless steel washers immediately after capture. A loop of stain- less steel wire was inserted through the last two respiratory pores of the abalone shell, a washer tag was added to one end of the loop, and the wire was then twisted with a pair of pliers to anchor the tag against the abalone shell. Length, width, and sex of all tagged northern abalone were recorded. Shell lengths of a subsample (10%) of untagged abalone were measured for comparison with the tagged samples. Tagged abalone were placed between layers of moist kelp so that the tag wires did not damage overlying abalone. Abalone were then trans- ported to transplant sites within 3-4 hours of har- vest and placed by divers in or immediately below the Macroeystis zone, the preferred habitat for adult abalone in sheltered locations (Breen 1986). A total of 502 tagged abalone were placed at site A and 506 at site B. Abalone (w = 438) were also removed from the control site (C), tagged and replaced onto the site within 2-3 hours. Divers searched the transplant sites within 48 hours of completing the transplant, and monthly from August 1984 to February 1985. All tagged and untagged abalone shells found in the study sites were collected. Recovery of Transplanted Abalone In March 1985, divers harvested tagged and un- tagged northern abalone at the transplant sites and tagged abalone at the control site. To maximize recovery, divers divided each site into a series of 5 m sections using cinder blocks and cord. The area of each section was measured and divers then searched repetitively for abalone within each sec- tion. Harvesting was terminated when repetitive searches in each area recovered less than 5% of the abalone harvested in the initial search. Divers also searched areas adjacent to the study site for tagged abalone to establish the magnitude and distance of emigration. Length, width, and sex of recovered. tagged abalone were recorded, along with the lengths of all tagged abalone. RESULTS Abalone Transplants A total of 2,737 northern abalone were trans- planted to site A and 2,677 abalone were trans- planted to site B (Table 1). The mean length of tagged abalone transplanted to site A was 88.7 mm and to site B, 90.2 mm. The mean length of abalone tagged at the control site (site C) was 78.7 mm. The differences in mean length between sites were all significant (P < 0.05). The mean length (±SD) of subsamples of untagged abalone transplanted to sites A and B were 84.6 -i- 12.8 mm (n = 204) and 8.8 + 11.5 mm (w = 257), respectively. Table 1— Summary of number and mean length of abalone at eacfi site. Density for site C is estimated from random quadrat surveys conducted prior to tagging (±SD). Site Study area (m^) Number trans- planted Density (abalone m"^ Tagged abalone X SD n (mm) (mm) A 550 2,737 4,98 502 88.7 11,5 (transplant) B 590 2,677 4.54 509 90.2 9,8 (transplant) C 1,800 0.56 ± 0.91 438 78.7 11.0 (control) Although divers carefully placed each trans- planted northern abalone foot down in suitable rock crevices or loose rock within the study areas, many abalone subsequently moved outside the lower boundary of the sites to depths of 7-10 m. This movement made the effective area of the transplant sites about three times larger than the original 60 m X 8 m dimensions. Little lateral movement of northern abalone beyond the boundaries of the study sites was observed. At sites B and C, the recovery of tagged shells by monthly diving inspections was highest in Aug- ust, one month after the transplant (Fig. 2). Recov- ery of both tagged and untagged shells at all sites in early July suggests that mortality one week after transplanting was less than 2%. The relatively high recovery of tagged shells in March 1985 was prob- ably a consequence of the more intensive searching effort during the final harvest. Cumulative recover- ies of tagged shells were 10.5% at site A, 18.0% at 97 FISHERY BULLETIN: VOL. 87, NO. 1 r D J F 1984/1985 M Figure 2.— Monthly recovery of tagged shells and tagged shell fragments from dead abalone. Results expressed as the percent- age of the original number of tagged abalone. site B, and 7.2% at site C (expressed as a percent- age of the original tagged number). Predation and Seasonal Variation in Habitat Shells recovered during monthly inspections were categorized by the nature of breakage. Shells pres- ent on the sites prior to the transplant were not categorized. These older shells were identifiable by degree of shell deterioration and epiphytization. Most shells were recovered whole: 68% at site A, 51% at site B, and 79% at site C. Recoveries of loose tags owing to tag loss or predation ranged from 0% of the initial tagged population at site C to 1.3% at site B. The relative proportion of loose tags recovered at each site cor- responded to the proportion of broken shells at each site, suggesting that shell breakage due to preda- tion may be the main cause of loose tags. One to three octopus dens were present at each site. Abalone shells found outside the dens were un- broken and not drilled. Over the course of this study, four octopuses were removed from site A, two from site B, and one from site C. Dens were often re- occupied three to four months after removal. Red rock crab. Cancer productus, were also a numeri- cally important prey item of octopus at the study sites. In July and August, the sunflower star, Pycno- podia helianthoides, preyed intensively on abalone at the transplant sites. This species was seen to prey on weakened or stressed abalone, and starfish were observed actively feeding on abalone immediately following the transplant. As with octopus, shells of abalone eaten by P. helianthoid£s were always recovered unbroken, either under actively feeding stars or entirely within the stomach. In contrast, broken or chipped shells were pre- sumed to be due to predation by red rock crabs, wolfeels (Anarrhichthys ocellatus), or cabezons (Scorpaenichthys marmoratus). Red rock crabs were abundant during the spring and summer at all sites except site C, but were rarer in the fall and winter. A few abalone shells were recovered outside a wolf- eel den at site A, and one or two cabezon were observed at all sites throughout the study period. Considerable seasonal variation in the marine plant community at the transplant sites was ob- served. Annuals, such as Desmarestia ligulata, died back in October and were completely gone by November. The Macrocystis canopy was also re- duced in fall and winter as a result of storm dam- age. Plants at site A were stripped of most fronds over the winter while losses were lower at site B, the more sheltered of the two transplant sites. Holdfasts remained intact and growth was renewed by March. Recovery of Abalone After nine months, 72% of the transplanted north- ern abalone were recovered from site A, and 39% from site B (Table 2). When shells from dead north- ern abalone collected during the 9-mo period were included, 88% of abalone at site A and 55% of abalone at site B could be accounted for. At the con- trol site (C), 31% of the tagged abalone were re- covered live and 40% of the original tagged abalone could be accounted for by including tagged shells recovered over the study period. The difference in percent recovery between the two transplant sites suggests that either abalone survival, abalone move- ment, or the ability of divers to find abalone differed between the sites. The recovery of tagged northern abalone was 6% less than recovery of untagged abalone at both transplant sites (Table 2), and the ratio of tagged to untagged abalone at recovery (0.20) was less than the initial ratio of 0.23. This difference is not significant (jc analysis, P < 0.05), indicating that losses due to the tagging procedure were minimal. The number of tagged shells recovered by divers over the 9-mo study allowed estimation of minimal instantaneous natural mortality (M„i„) (Ricker 1975). This calculation assumes that divers recov- 98 EMMETT and JAMIESON: TRANSPLANT OF NORTHERN ABALONE Table 2.— Recovery of live abalone in March 1985, recovery of abalone shells over the study period, and estimation of M^^„ (from tagged shell recovery) and M^^ (from live tagged abalone recovery). Number recovered was after / # tagged survivors \ 12 # initially tagged j 9 Recovery 9 months, so M = - In Initial number tagged shells live untagged shells live % 1 ive recovery M Site tagged untagged tagged untagged Total Mmln Mmax A 502 2,235 53 330 396 1,628 66 73 72 0.15 0.56 (transplant) B 509 2,168 92 175 353 861 34 40 39 0.27 1.42 (transplant) C 438 33 141 31 _ 31 0.10 1.51 (control) ered the shells of all tagged northern abalone that had died. The recovery of live tagged abalone allows estimation of maximal instantaneous natural mor- tality (Afmax). assuming that divers recovered all the living abalone. Values for M^^^ ranged from 0.10 to 0.27 for the three study sites; M^^^ ranged from 0.56 to 1.51 (Table 2). The movement of northern abalone from the lower edge of the Macrocystis bed to deeper (5-11 m) water shortly after the transplant (described above) effectively increased the area of each transplant site by threefold. Approximately 30% of recovered northern abalone were found below the vegetation zone at both transplant sites. This movement, coupled with abalone losses, resulted in final abalone densities of 1.27 m"- at site A and 0.73 m"- at site B. Prior to final harvesting, divers searched beyond the expanded boundaries of the sites for tagged northern abalone. Little lateral movement of north- ern abalone along the shoreline had occurred. At site A, 11 tagged abalone (2.2% of the original tagged number) were found outside the study area. Max- imum distance from the site was 37 m, and one tagged abalone was found at a depth of 18 m. At site B, three tagged abalone were found outside the site area, all in deeper water. The most extensive movement was recorded at the control site. Ten tagged abalone (2.2% of the tagged population) were recovered outside the site boundaries; one abalone had moved 125 m; and three abalone had crossed a 50 m wide sandy channel. At sites B and C, a considerable proportion of the transplanted or tagged population could not be accounted for. The low recovery of tagged abalone outside the boundaries of the sites suggests that emigration is not the sole explanation. However, searches conducted outside the site boundaries were less intensive than those conducted within. Growth Figure 3 gives the length frequencies of tagged northern abalone at each site at the initiation of the study in June 1984 and upon recovery in March 1985. Differences between mean initial and final lengths at each site were, in all cases, significantly different (paired t-test, P < 0,05). The mean growth of tagged northern abalone after 9 months was 7.1 mm at site A, 7.2 mm at site B, and 2.9 mm at site C, the control site. Mean growth of untagged north- ern abalone was 9.4 mm at site A and 9,9 mm at site B, Growth rates of northern abalone were analyzed by Walford plots, in which the initial length of in- dividual tagged abalone (/(,) are plotted against the length of the same individual at recovery in March 1985, 9 months later (l^). The numbers of data pairs were 306, 167, and 126 at sites A, B, and C, respec- tively. Table 3 summarizes parameters of the regres- sion lines of Walford plots for each site as well as the annual Brody coefficient and asymptotic length calculated from these regression parameters (Ricker 1975). The annual Brody coefficient varied from 0.178 to 0.440, and was lowest at the control site. Values for asymptotic length varied from 104 to 112 mm, also being lowest at the control site. Table 3.— Parameters for linear regression, annual Brody coefficient (K). and asymptotic length (/_) as calculated from Walford plots. Linear regression / Site W a b r^ K (mm) A 306 30.8 0.719 0.64 0.440 110 (transplant) B 167 24.8 0.779 0.76 0.333 112 (transplant) C 126 13.0 0.875 0.87 0.178 104 (control) 99 s? 10- SITE A X = 88 7 1 1 1 5 n=502 50 x = 95.8^8.0 n= 306 UL 80 90 100 no SHELL LENGTH (mm) ~~r" 50 — r- 60 .III 'O 80 90 100 SHELL LENGTH (mm) li. = FISHERY BULLETIN: VOL. 87, NO. 1 SITE B X = 902*98 n = 509 Uli -1 — ^ 50 60 70 80 90 100 110 120 SHELL LENGTH (mn) j< = 97.4t 7.5 n=l67 lii_ 70 80 90 100 110 120 SHELL LENGTH (mm) Figure 3.— Length frequencies of tagged abalone at each site at the beginning (June 1984 - T(,) and at end Growth of northern abalone at the transplant and control sites was compared by using the Walford regression to estimate final lengths and associated confidence intervals for abalone of initial lengths equal to the lower (75 mm) and upper (100 mm) size range of abalone placed at the two transplant sites (Table 4). As confidence intervals are narrowest at the mean value of Iq, between site comparisons using these more extreme values are more rigorous than using l^ values which fall between 75 and 100 mm. The hypothesis that predicted i, values at Iq = 75 and 100 mm for transplanted northern abalone are greater than the corresponding /j value for con- trol abalone was then tested (one tailed ^test). All differences were significant at P < 0.05 (Table 4), indicating that northern abalone growth was significantly greater for abalone transplanted to sites A and B as compared with nontransplanied abalone at the control site. Economic Feasibility The economic feasibility of transplanting wild northern abalone seed for subsequent commercial harvest depends primarily on three factors: 1) the cost to collect and transplant stock, 2) the rate of recovery of legal-sized abalone after a suitable grow- Table 4.— Estimates of growth calculated from Walford plots for abalone for the lower (75 mm) and upper (100 mm) size range of transplanted abalone. Values are expressed + 95% confidence interval. /(, = length at initiation of study, /, = length after nine months. Values in parentheses are / - statistic and degrees of freedom comparing the /, values at each transplant site with the corresponding /, values at the control site. P < 0.05 in all cases. Site Length after 9 months /, (/,) /„ = 100 mm A (transplant) B (transplant) C (control) 84.7 ± 0.7 (16.22, 305) 83.3 + 0.9 (10.02, 166) 78.6 ± 0.4 102.7 ± 0.5 (7.97, 305) 102.8 ± 0.5 (8.85, 166) 100.5 ± 1.0 ing period, and 3) the price of abalone. The first fac- tor depends on abalone density in the source area and the distance to the transplant sites. The pres- ent study shows that the second factor (recovery rate) can vary greatly between sites. In this study 6 diver-days were required to col- lect 5,000 sublegal-sized northern abalone at the source area and move them to the transplant sites. This variable cost was estimated to be $1,500, at a rate of $250 diver-day ' for wages and fuel costs. 100 EMMETT and JAMIESON: TRANSPLANT OF NORTHERN ABALONE SITE C x=78 7t|IO n = 438 iLo^ ^ 50 60 70 80 90 100 110 120 SHELL LENGTH (mm) x = 81.6* 9.3 n= 126 50 60 70 80 90 100 110 120 SHELL LENGTH (mm) Table 5 summarizes these economic parameters for values of M^^ ranging from 0.10 to 1.00. These data indicate that a reasonable value for the inter- nal rate of return (i.e., >20%) would be obtained at •^max values of 0.80 or less. Transplants to site A but not site B would show a reasonable rate of return. This model can be generalized to estimate eco- nomic returns for variable abalone seed costs in the case of transplanting hatchery- reared seed to the wild. Figure 4 summarizes internal rates of return for 20 mm hatchery seed of variable cost, a 4.5 yr growth period, planting costs of $0.20 per abalone, and harvest costs of $0.40 per kg. All other assumptions are the same as the model given above. Under these price assumptions, M^^ values must be less than 0.6 to show a reasonable rate of return if seed costs are ^$0.10 per abalone. At M^^ values greater than 0.8, transplanted abalone seed will not yield a reasonable rate of return unless seed costs are extremely low (<$0.02 per abalone). Figure Z.-Continued-(U[a.rch 1985 Tj) of the study. Harvest costs were similarly estimated at $1,575. Fixed costs were not included and were assumed to be zero. This information is used in the following simple economic model which examines the rate of economic return as a function of M^^ (instantane- ous natural mortality estimated from recovery of live abalone). Assuming that 5,000, 80 mm abalone are trans- planted, they reach legal size (100 mm and 340 g) in two years, and can be sold at a price of $11 kg"^ Then gross return = $11 kg ' x 0.34 kg abalone"' x 5,000 abalone x e"''^, where t = 2, M = M^ gross return-harvest costs, where harvest costs = $1,575 net return-initial costs, where ini- tial costs = $1,500 1 net return profit discounted profit net return 1 -I- d initial internal rate of retiu-n (IRR) costs, where d = discount rate = 10% gretun, rate _ ;l_o, whorO rotum rate = (In (net return/initial costs))/i. Table 5, — Summary of economic returns from transplanting abalone, assuming 2 years to recov- ery. Calculated from ttie economic model given in text. IRR = Internal rate of return; f = 2 years. Discounted IRR M,„ax Profit profit (x 100%) 010 $12,235 $9,865 203 0.20 9,460 7,553 171 0.40 5,328 4,139 113 0.60 2,557 1,851 648 0.80 700 318 21.4 0.90 16 -250 0.8 1.00 -544 -710 — DISCUSSION In this study northern abalone transplanted from exposed areas to more sheltered habitat grew faster than nontransplanted controls. These results cor- roborate the observations of Breen (1986) that "surf" abalone retain the potential to grow well when placed in more productive habitat. These ob- servations suggest that abalone are, to some degree, food limited in exposed habitats which have little to no canopy-forming algae. Because northern abalone varied in size in differ- ent sections of the source area, the initial size of non- transplanted abalone at the control site was signif- icantly less than that of the transplanted abalone (Fig. 3). This bias would be expected to reduce the difference in growth rate between the abalone at 101 FISHERY BULLETIN: VOL. 87, NO. 1 z o \ 20% RETURN \ ~\ I I I I I T" 3 4 5 6 07 8 0.9 Figure 4.— Isoprofit lines drawn from internal rates of return calculated at varying seed cost and natural mortality values. Solid line = breakeven point, dotted line = 20% internal rate of return. the control and transplant sites, because growth rate is inversely related to body size. Therefore the growth rate differences observed in this study are likely smaller than would have been observed if the mean length of transplanted abalone had been equal to the mean length of the control group at the ini- tiation of the study. Analyzing growth rates by Walford plots diminishes this bias because the anal- ysis compares the length of individual abalone at the beginning and end of the study period and does not use pooled data to compare growth rates among sites. In most transplant experiments, recovery follow- ing transplanting depends on both abalone size and source and is greatest with larger abalone collected from the wild. In the present study wild-harvested northern abalone of 50-100 mm length were trans- planted, and recovery was 72% and 39% at the two sites 9 months after the transplant. Saito (1984) reported 18% recovery 9 months after transplant- ing 25 mm hatchery-reared Haliotis discus hannai in Japan. The author also stated that commercial recapture rates are 5-10% for hatchery-reared seed and 20-25% for wild seed. Recovery of 45-71 mm, hatchery-reared Haliotis rufescens in California was less than 1% one year after transplanting (Tegner and Butler 1985). Inoue (1976) reported increased survival with increasing seed size up to 70 mm. Tegner (pers. commun.)^ estimated an annual mor- tality rate of 9.1% for mature, native green abalone, Haliotis fulgens, one year after being transplanted in California. The use of large, wild-harvested north- ern abalone likely contributed to the relatively high recovery rates observed in the present study. The markedly different rates of recovery between the two transplant sites seemed independent of handling, tagging, and transplant procedures. Shells collected within two weeks of release indicated that immediate posthandling mortality was similar (<2%) 'M. J. Tegner, Scripps Institution of Oceanography. La Jolla, CA 92092, pers. commun. January 1987. 102 EMMETT and JAMIESON: TRANSPLANT OF NORTHERN ABALONE at all sites. The similar ratio of tagged to untagged northern abalone at the initiation and end of the study demonstrated that both tagged and untagged animals had similar survival rates. Breen (1986) calculated M, from population size structure and growth rate estimates, to be 0.05- 0.24 for H. kamtschatkana at eight sites in British Columbia. These values are consistent with esti- mates of M, derived from a variety of techniques, of 0.05-0.40 for abalone populations in Australia and New Zealand (Shepherd et al. 1982; Sainsbury 1982). In California, estimates of Af (partly based on dead shell recovery) are higher, ranging from 0.36 to o° for four native species of Haliotis (Tutshulte 1976; Hines and Pearse 1982). The highest estimates are from areas that experience sea otter, Enhydra lutris, predation. Estimates of M calculated from data given by Tegner and Butler (1985) are 0.40 and 0.55 for two transplanted populations of red aba- lone, H. rufescens. In the present study, estimates of M based on recovered, tagged shells {M^,„) are similar to values determined for abalone populations from similar latitudes in British Columbia (Breen 1986) and southern Australia (Shepherd et al. 1982). Values of M determined from the recovery rate of live aba- lone (M^ax) are higher than most values of M re- ported in the literature. It is likely that M^^y^ esti- mates of instantaneous natural mortality are high because some abalone probably emigrated or re- mained hidden within the sites. However, while un- recovered abalone would still be able to contribute to population reproduction, they would not likely be available for harvest; the after-harvest population density would be too low to encourage the return of fishermen, and the animals might remain well hid- den. Effectively, these abalone can be considered removed from the harvestable biomass, and since there are only two categories, available and unavail- able animals, in most cost-benefit and/or exploita- tion models, unrecoverable abalone should be con- sidered unavailable abalone. For this reason M^^^ is an appropriate term for use in models assessing the economic feasibility of abalone transplants and in other situations where animals are established in an area for the purpose of future exploitation. A considerable proportion of tagged and/or trans- planted northern abalone were unaccounted for at sites B and C. The difference in percent recovery of live abalone at the two transplant sites (72% versus 39%) was due primarily to these abalone, as approximately the same number of shells were col- lected at each site. There are several explanations: 1) difficulty in locating abalone due to complex bot- tom topography, 2) physical removal of abalone from the site by mobile predators such as octopus and sea stars, 3) the destruction of shells by pred- ators such as crabs, 4) emigration, and 5) transport of shells from the site by waves or currents. In California, Tegner and Butler (1985) attributed abalone loss during transplant experiments to both predation and emigration, citing the recovery of shells in all directions outside the study site as evidence of random dispersal of live animals. In the present study, searches outside the sites at the ter- mination of the study suggested little emigration of tagged abalone, except at the control site. Although no studies have been done on the natural movement of Haliotis kamtschatkana, the mean distance moved in a year by tagged ormers (Haliotis tuber- culata) in France was only 6.7 m for the 68% of the population that showed any evidence of movement (Clavier and Richard 1984). That study also showed that smaller abalone tended to be less mobile. Hines and Pearse (1982) reported that marked abalone shells drifted 2-3 m in three months. The degree of shell drift due to wind or current action is ob- viously site specific and probably only occurred at the more exposed control site in the present study. Three fundamental questions concerning the feasibility and benefit of transplanting abalone from exposed areas remain: 1) the number and extent of abalone in exposed coastal areas has not been established, 2) the population dynamics and the reproductive contribution from such populations to the total coastal stock remain unknown, and 3) the potential of transplanted abalone to enhance popula- tion reproduction and ultimately recruitment at specific transplant sites has to be determined on a site-by-site basis. ACKNOWLEDGMENTS This study was carried out in conjunction with the West Coast Abalone Harvesters Association (W.C.A.H.A.). Eric Wickham, Bob Harrington, and Guy Whyte of the W.C.A.H.A. and Thomas Shields of Archipelago Marine Research were responsible for much of the initial planning of this project. We thank Dave Johnstone and Mark Bath of the W.C.A.H.A. for their participation in all phases of the field work; their experience as abalone divers and their firsthand knowledge of the study area made the field program run smoothly and efficient- ly. Paul Breen of the New Zealand Minister of Agriculture and Fisheries offered many useful com- ments throughout the study and provided the equa- tions for the economic feasibility model. Howard 103 FISHERY BULLETIN: VOL. 87. NO. 1 McElderry provided advice on data analysis. Valu- able comments on the manuscript were provided by three anonymous reviewers. This project was finan- cially supported by the Department of Fisheries and Oceans unsolicited proposal program, D.S.S. Con- tract No. 08SB.FP 597-4-0145. LITERATURE CITED BouTiLLiER, J. A., W. Carolsfeld, P. A. Breen, and K. Bates. 1984. Abalone survey in the Estevan Group and Aristazabal Island May, 1983. Can. MS Rep. Fish. Aquat. Sci. 1747, 60 p. BoirriLLiER, J. A., W. Carolsfeld. P. A. Breen, S. Farlinger, AND K. Bates. 1985. Abalone resurvey in the southeast Queen Charlotte Islands, July 1984. Can. MS Rep. Fish. Aquat. Sci. 1818, 87 p. Breen, P. A. 1980. Measuring fishing intensity and annual production in the abalone fishery of British Columbia. Can. Tech. Rep. Fish. Aquat. Sci. 947, 49 p. 1986. Management of British Columbia fishery for northern abalone (Haliotis kamtschatkana). In G. S. Jamieson and N. Bourne (editors). Proceedings of the North Pacific Work- shop on Stock Assessment and Management of Inverte- brates, p. 300-312. Can. Spec. Publ. Fish. Aquat. Sci. 92. Clavier. J., and 0. Richard. 1984. Etude experimentale de depalcement de I'ormeau (Haliotis turberculata) dans le milieu natural. Rev. Trav. Inst. Peches Marit. 46:315-326. Cox, K. W. 1962. California abalones, family Haliotidae. Calif. Fish. Game, Fish Bull. 118, 133 p. Farlinger, S., and K. T. Bates. 1985. Review of shellfish fisheries in northern British Colum- bia to 1984. Can. MS Rep. Fish. Aquat. Sci. 1841. 35 p. Hines, a. J., AND J. S. Pearse. 1982. Abalone, shells and sea otters: dynamics of prey popula- tions in central California. Ecology 63:1547-1560. Indue, M. 1976. Awabi [Abalone]. In Suisan Zoyoshoku Deeta Bukku (Fisheries propagation data book), p. 19-60. Published by Suisan Shuppan. Translation by M. Mottet, Dep. Fish., State of Wash. Mottet, M. G. 1978. A review of the fishery biology of abalones. Wash. Dep. Fish. Tech. Rep. 37, 81 p. Ricker. W. E. 1975. Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Board Can. 191, 382 p. Sainsbury, K. J. 1982. Population dynamics and fishery management of the paua, Haliotis iris I. Population structure, growth, reproduc- tion, and mortality. N.Z. J. Mar. Freshwater Res. 16, 147 p. Saito, K. 1984. Ocean ranching of abalones and scallops in northern Japan. Aquaculture 39:361-373. Shepherd, S. A., G. P. Kirkwood, and R. L. Sandland. 1982. Studies on southern Australian abalone (genus Halio- tis). III. Mortality of two exploited species. Aust. J. Mar. Freshwater Res. 33. 265 p. Tegner. M. J., and R. A. Butler. 1985. The survival and mortality of seeded and native red abalone, Haliotis rufescens, on the Palos Verdes peninsula. Calif Fish Game 71:150-163. TUTSHULTE, T. C. 1976. The comparative ecology of three sympatric abalone species. Ph.D. Thesis, Univ. California, San Diego. 104 COMPARISON OF SOME GENERA AND SPECIES OF BOX CRABS (BRACHYURA: CALAPPIDAE), SOUTHWESTERN NORTH ATLANTIC, WITH DESCRIPTION OF A NEW GENUS AND SPECIES Austin B. Williams' and C. Allan Child^ ABSTRACT Five species of calappid crabs from the southwestern Atlantic that belong to the genera Calappa, Cydozo- dion new genus, and Paracyclois are analyzed on the basis of morphology, morphometries, geographic, and bathymetric range. Calappa tortugae, new rank, known in the past as C. angusta in the broad sense, is restricted and compared with its eastern Pacific twin species, C. satissurei. Two small species placed in Cyclozodion were until now unrecognized and partly included in Calappa angusta, broad sense. Cyclozo- ditm angnstum. a relatively smooth form, is the tj-pe species of the new genus, and C. tuheratum, a rough form superficially resembling Calappa tortugae, is described as new. Species of both Paracyclois and the Early Tertiary genus Calappilia in which it was subsumed are reviewed, the former is revalidated, and its only two species, western Atlantic P. atlantis and western Indo-Pacific P. milneedwardsii, are rediagnosed. Diagnoses and discriminations are accompanied by illustrations. Keys to caJappid genera in the Western Atlantic, and for identification of Cyclozodion and Paracyclois species are given. Holthuis (1958) revised five species of West Indian box crabs, Calappa cinerea Holthuis 1958, C.flam- mea (Herbst 1794), C. nitida Holthuis 1958, C. ocellata Holthuis 1958, and C. sulcata. Rathbun 1898, but a species from that region known until now as C. angusta A. Milne Edwards 1880 was not included in his paper because the collection he studied in- cluded no representatives of that form. We find that this latter species is not at all well defined, and the purpose of this paper is to clarify its status and that of similar species in related genera. Samples of decapod crustaceans from exploratory trawling by the Bureau of Commercial Fisheries RV Pelican, U.S. Fish and Wildlife Service RV Combat, National Marine Fisheries Service RV Silver Bay, RV Oregon, and RV Oregoji II deposited in the crustacean collection of the National Museum of Natural History (USNM), Smithsonian Institution, contain specimens of a seldom reported calappid crab, Paracyclois atlantis Chace 1939, 1940 from the Caribbean region of the western North Atlan- tic, and representatives of a genus not previously recognized. Two small calappid species in the cata- logued USNM crustacean collection have been at- tributed to Calappa angusta A. Milne Edwards 1880 by Rathbun (1937) and other authors (see Williams 'Systematics Laboratory, National Marine Fisheries Service, NOAA, National Museum of Natural History, Washington, DC 20560. ^Department of Invertebrate Zoology, National Museum of Natural History, Washington, DC 20560. Manuscript accepted August 1988. Fishery Bulletin, U.S. 87:105-121. 1984) on the basis of what were thought to be juve- nile characters exhibited by the carapace of that species. Review of the material in the USNM shows this concept to be in error. Moreover, representa- tives of the extant type series of C. angusta in the Museum of Comparative Zoology (MCZ), Harvard University, consist of very small juveniles, a holo- type and four paratypes in which definitive char- acters are poorly developed, that surprisingly belong not to one but three calappid species. "Calappa angusta" as presently understood is in reality a com- plex of species belonging in Calappa Weber 1895 and the previously unrecognized genus. Only two species of Paracyclois Miers 1886 have been described, the above mentioned, and the type species, P. milneedwardsii Miers 1886, from the western Indo-Pacific. Glaessner (1969) synon- ymized Paraclyclois with Calappilia A. Milne Edwards 1873, considered until that time to in- clude only species of Middle Eocene to Upper Oligocene ages in North America, Europe, and the East Indian region, but did not discuss reasons for his action. Because our determinations involved generic placement of material from trawl samples, we reviewed literature concerned with both of these genera and studied specimens of selected species of Calappilia in the fossil crustacean collection of the USNM. Austin B. Williams devel- oped the text, C. Allan Child rendered the draw- ings, and both of us identified and cross-checked material. 105 FISHERY BULLETIN: VOL. 87. NO. 1 Key to Recent genera of Calappidae in the western Atlantic Ocean 1. Chelae essentially symmetrical, no unusually enlarged teeth or protuberances, Subfamily Matutinae 2 Chelae dissimilar, major chela with large tooth on dactyl and pair of protuberances on pro- podus. Subfamily Calappinae 3 2. Carapace considerably broader than long, regularly convex above Hepatus Latreille 1802 Carapace nearly as long as broad, dorsal surface uneven Osachila Stimpson 1871 3. Posterolateral region of carapace not ex- panded into dentate, winglike projection. . . .4 Posterolateral region of carapace expanded into dentate, winglike projection 5 4. Merus of cheliped bispinous on distal outer surface, with lower spine strong and greatly extended laterally Acanthocarpus Stimpson 1871 Merus of cheliped not bispinous on distal outer surface, carapace subcircular, small spine at lateral angle Cycloes De Haan 1837 5. Pereopod 5 with articles spineless 6 Pereopod 5 with row of spines on flexor sur- face of ischium-merus Paracydois Miers 1886 6. Greatest span of winglike posterolateral pro- jections less than maximal span between anterolateral margins; outer proximolateral corner of palm bearing short, flattened, smoothly crested ridge Cyclozodion new genus Greatest span of winglike posterolateral pro- jections exceeding maximal span between anterolateral margins; outer proximolateral corner of palm bearing flattened acute spine or subrectangular ridge Calappa Weber 1795 Calappa tortugae Rathbun 1933, new rank Figures 1, 2 Calappa saussurei tortugae Rathbun 1933:183. Calappa ang%ista.—(Pa.rt, not selected juveniles.) A. Milne Edwards 1880:18. -Hay and Shore 1918: 421, pi. 31, fig. 7.-Rathbun 1937:210, pi. 64, figs. 4-6.-Chace 1956:18 (list). -Williams 1965:154, fig. 134; 1984:273, fig. 203.-Pequegnat 1970: 177.-Powers 1977:30. Material studied.Specimen lots in USNM re- corded by Rathbun (1937) under C. angusta and C. saussurei tortugae (catalog numbers only) plus material added since that time. North Carolina: USNM 68530.-101676. 1 a, 19 Cjuv.); 34°18'N, 75°58'W, SE off Cape Lookout, 137 m; Combat stn. 405, 21 June 1957.-101675. 1 cr; 34°19'N, 75°54'W, SE off Cape Lookout, 183 m; Combat stn. 402, 21 June 1957.-202745. 1 a, 2 9; 33°48'48"N, 76°34'24"W, 46 m; BLM, 4 Mar. 1981.-202746. 1 9; 33°48'12"N, 76°34'24"W, 116 m; Duke Univ. for BLM, 14 May 1981.-202747. 19 (ovig.); 33°47'36"N, 76°34'24"W, 116 m; Duke Univ. for BLM. 14 May 1981.-202748. 1 o-; 33°48'06"N, 76°34'24"W, 105 m; Duke Univ. for BLM, 14 May 1981.-202749. lo-; 33°48'42"N, 76°34'12"W, 102 m; Duke Univ. for BLM, 14 May 1981.-202750. 1 9 (juv.); 33°48'42"N, 76°34'30"W, 99 m; Duke Univ. for BLM, 14 May 1981.-220962. 1 o-, 3 9; 33°48'36"N, 76°34'06"W, 69 m; Duke Univ. for MMS, 4 Mar. 1981. South Carolina: 188682. 2 o-, 1 9; 32°18'30"N, 79°00'30"W, 84 m; Dolphin 577096, 3/4 Yankee trawl, MARMAP, 9 Mar. 1977.-188677. 1 o-; 33° 17'N, 77°08'42"W, 155 m; Dolphin 573426, 3/4 Yankee trawl, MARMAP, 15 Nov. 1973.-Silver Bay stn. 2263. 2 ct; E of Charleston, 33°04'N, 78° 12'W, 29 m; trawl, 28 July 1960. Georgia: 155583. 1 o-; 30°50'30"N, 80°01'W, 93 m; M. Gray 209, 7 May 1963.-155582. 3 a; 30° 55'30"N, 79°57'W, 91-119 m; M. Gray, 12 June 1963.-188680. 1 undet.; 31°43'30"N, 79°38'30"W, 64 m; Dolphin 576078, 3/4 Yankee trawl, MAR- MAP, 5 May 1976. Florida: 66382. C. saussurei tortugae holotype, o"; Tortugas, about 12 mi S Red no 2 Buoy, 110 m, W. L. Schmitt, stn. 33-31, 22 July 1931.-66381. 1 9; same.-234461. 1 o-, 5 9; same.-68506, 68507, 68508, 68509, 68515, 71369.-101413. 6 o-, 3 9; off Jacksonville, 30°11'N, 80°17'W, 59 m; Combat stn. 72, 31 Aug. 1956.-101414. 2 o-, 1 9; SE Cape Canaveral, 28°32'N, 80°05'W, 119 m; Combat stn. 90, 3 Sept. 1956.-91137. 1 cc, 1 9; W Cape Romano, 25°35'N, 83°42'W, 110 m; Oregon stn. 35, 26 June 1950.- 97487. 1 cr; SW Sarasota, 27°07'N, 83°19' W, 42 m; Oregon stn. 963, 4 Apr. 1954.-101678. 1 cr; S Cape San Bias, 29°10'N, 85°48'W, 101-130 m; Silver Bay stn. 100, 26 July 1957 .Silver Bay 106 WILLIAMS and CHILD: COMPARISONS OF SOME BOX CRABS Figure l.—Calappa tortugae Rathbun. cr holotype, USNM 66382: a, carapace, eyes, and part of left cheliped; b, orbital region in frontal view; c, right chela and part of carpus; d, abdomen; e-f, first and second pleopods. 9, USNM 202747: g, abdomen. 107 FISHERY BULLETIN: VOL. 87, NO. 1 80 75 70 65 60 - 55 50 - - 45 ^ 40 35 - 30 - 25 1 L Figure 2.— Interorbital width expressed as percent maximum span across posterolateral projections for samples of six species of Calappidae: vertical lines = ranges of percentage; horizontal lines = means; open rectangles = standard deviations. A - Calappa tortugae, B - Calappa saiissurei, juveniles in sample excluded; C - C. saussurei, juveniles in sample included; D • Cydozodion angustum, juveniles in type series excluded; E - same, juveniles in type series included; F C. tuberatum; G - Paracylois atlan- tis; H - P. milneedwardsii. stn. 2263. 1 9; off St. Augustine, 29°40'N, 80°14'W, 64-87 m; dredge, 7 Oct. 1962.-3438. 1 ct; off Or- mond Beach, 29°34'N, 80°15'W, 73-74 m; dredge, 24 Sept. 1961.-3171. 3 o-, 2 9 (ovig.); same, 29° 30'N, 80°15'W, 71-73 m; dredge, 10 May 1961.-3519. 3 cc, 4 9; Straits of Florida, 24°59'N, 80°14'W, 183 m; dredge, 9 Nov. 1961.-2362. 2 cr, 1 9; off Key Largo, 24°56'N, 80°22'W, 84 m; dredge, 25 Oct. 1960.-2416. 1 9; Straits of Florida, 24°18'N, 81°29'W, 229 m; dredge, 28 Oct. 1960.-50. 1 9; S of Apalachicola Bay, 28°58'N, 85°20'W, 70-80 m; dredge, 15 July 1957. Caribbean: Oregon stn.— 6715. 1 O"; W Anguilla I., 18°36'N, 63°27'W, 201-238 m; dredge 30 May 1967.-4400. 1 9; Venezuela, off Los Mongos Is., 12°37'N, 70°45'W, 97 m; dredge, 26 Sept. 1963.-5036. 1 o"; Venezuela, off Peninsula de Paria, 11°36'N, 62°54'W, 183 m; dredge, 24 Sept. 1964. MCZ 6654. 1 juv.; off Sombrero [Is.], 99 m; labeled as Calappa angusta A.M.E. para type. Dia^wosis.- Carapace convex longitudinally and from side to side; mean length 0.9 times mean width {N = 66); surface elevated in median tract and bran- chial regions, separated by well-marked furrow at each side running from orbit to level of cardiac region but thereafter becoming obsolescent; covered by prominent, densely and minutely granular pro- tuberances of varied size more or less symmetrical- ly arranged, with more widely scattered and larger granules between them; arcuate anterolateral mar- gins finely granulate, with larger granules at inter- vals; winglike extension with teeth largest at posterolateral angle preceded by up to 4 teeth pro- gressively diminishing in size anteriorly, and fol- lowed posteriorly by 2 or 3 smaller teeth successively diminishing in size, all with beaded edges; mean maximal span between tips of posterolateral teeth slightly greater (1.02) than mean maximal span be- tween anterolateral margins; axis of largest tooth on winglike protuberance diverging from midsagit- tal line at angle of 20-25°. Front trilobed, downturned, slightly broader than orbits; large central lobe with rather narrowly rounded tip barely visible in dorsal view, smaller lateral lobes directed anteriorly to accommodate narrowly oblique folded antennular peduncles; or- bits noticeably raised above surrounding surface; interorbital width relatively narrow, its span relative to maximal span between posterolateral winglike extentions rather narrow (see Figures 1 and 2). Palms of chelipeds with external surface bearing irregular ornamentation moderately reminiscent of that on carapace; a lower zone of closely crowded coarse granules adjacent to beaded ventral margin, larger widely scattered irregular protuberances in central region becoming stronger and more closely arranged near base of dorsal "cockscomb" (crest of teeth), widely spaced irregular granules between these varying from obsolescent to well formed; short obliquely curved ridge rising from proximolateral corner to end anteriorly in subrectangular angle, crest minutely crenulate and in line with subdistal crest of 4 similar, narrowly separated broad teeth on merus. Abdomen of each sex broadest at segment 3; lat- ter fused with narrower segments 4 and 5 in male, segments in female relatively broader but essentially linear and free; segment 2 somewhat trilobed and bearing sparsely scattered low granules clustered laterally, segment 3 with much lower relief and low 108 WILLIAMS and CHILD: COMPARISONS OF SOME BOX CRABS granules clustered laterally; telson subtriang^lar. Male pleopod 1 rather stout, slightly curved and con- ically elongate, tapering to narrow distal opening with nearby cluster of minute horny spinules; pleopod 2 with slender stylet divided into 2 parts, gently curved proximal part stronger than distal part curved mesially upon itself as a crook, distal half of crook extending beyond tip of pleopod 1 . Measurements in -mm.— Carapace: smallest o* length 14.4, maximum anterior width 15.8, max- imum width across winglike projections 15.4; largest 0-, same 35.1, 42.3, 44.8; smallest 9, same 10.7, 11.5, 11.4; largest 9, same 29.7, 34.5, 35.7. Known range.— North Carolina to Florida, around Gulf of Mexico, Leeward Islands to off Venezuela, 13-238 m (see Powers 1977 in part). Remarks.— Milne Edwards's Calappa angusta 1880 has been generically misplaced. The next avail- able name for the species is Calappa saussurei tor- tugae Rathbun 1933, raised to full specific rank. The young of C. tortugae have long been regarded as having the greatest carapace width anterior to the winglike posterolateral projections. That is con- firmed by measurements of young individuals noted above, but measurement of a series ranging from juvenile to adult indicates that the winglike postero- lateral projections quickly become the widest part of the carapace as growth progresses, as is true of Calappa in general. Another way of expressing this width is to compare it wdth the interorbital distance. Interorbital distance expressed as a percent of max- imum span across the posterolateral winglike pro- jections is plotted for measured samples in Figure 2A {N = 71, X = 0.347, SD = 0.039). The eyes of C. tortugae are relatively smaller and the orbits more elevated than are those of species belonging to either Paraeyclois or Cyclozodion new genus, and it is clear that the indicated ratio lies largely beyond that for these species, although it is comparable to that computed for a sample of C. saussurei Rathbun 1898 available in the USNM (Fig. 2B, TV = 14, x = 0.297, SD = 0.030, juveniles excluded). That sam- ple contains a disproportionate number of very small juveniles; therefore it is useful to compute two ratios for that species, one that excludes the juveniles and one that includes them (Fig. 2C, N = 21, x = 0.297, SD = 0.131). These two species oi Calappa are similar enough to be regarded as a geminate pair from either side of the Central American land mass, as implied by Rathbun's descriptions. The chief dif- ference is that C. saussurei has a much more coarse- ly and uniformly tuberculate extensor face on the palms of the chelae than does C. tortugae. Cyclozodion new genus Diagnosis.— Carapace slightly wider than long and moderately convex; front narrow and trilobate; me- dian lobe rounded and much broader than lateral lobes; without lateral epibranchial spine or tooth; anterolateral margins regularly arcuate and entire or lightly crenulate, broadest span anterior to junc- ture w\t\\ posterolateral margin; each posterolateral margin bearing strongly spiniferous winglike pro- jection, width between principal spines on latter less than greatest width of carapace, axis of principal spine on lobe diverging from midsagittal line at angle of about 40°. Eyes large, peduncles short, robust, closely en- cased in oval orbits scarcely raised above surround- ing area; interorbital distance 0.40-0.70 (0.80 in smallest juveniles) of span between tips of principal spines on posterolateral margin. Antennules folding obliquely; antennae with quadrate basal article not reaching frontal margin, flagellum very short. Outer maxillipeds with ischium longer than broad, longer than distally truncate merus with its anterointernal angle distinctly notched. Pereopods 2-5 spineless. Type species.— Cyclozodion angustum (A. Milne Edwards 1880). Etymology.— From the Greek "cyclo", round, and "zodion", a small carved figure, for the shape of the carapace. The gender is neuter. Remarks.— Two small species fit between Calap- pa and Paraeyclois. These species have the orbital characteristics of Paraeyclois. They have postero- lateral spines that cover a narrower span than do those of Calappa, but in general shape they resem- ble some juveniles of that genus. The two small species could almost be cited as examples of brachyuran neoteny, for they seemingly maintain a juvenile Calappa-like carapace facies while attain- ing sexual maturity. We are faced with the pros- pect of further splitting the family by introducing a new genus to contain these two species, or broad- ening the concept of Paraeyclois to contain them. However, lack of any spines on the pereopods and shape of the proximolateral ridge on the extensor face of the cheliped palms, to point out only two features, clearly set them apart from Paraeyclois. Rathbun (1937) and others perhaps unconsciously took the alternate route of accommodating them in 109 FISHERY BULLETIN: VOL. 87. NO. 1 what she called Calappa angusta, saying that the narrow span across the posterolateral winglike pro- jections of the young of that species broadened with age into a full Caiappa-like form. Analysis of mea- surements on a large series of specimens does not support this viewpoint (see Figure 2), and we there- fore choose to erect the new genus for reception of these two small species. Key to species of Cyclozodion 1 . Carapace smooth to slightly tuberculate; front with central lobe shallowly concave, margin smooth; chelipeds with upper surface of carpus smooth C. angustum Carapace definitely tuberculate; front with broadly concave central lobe sharply granular near tip and on margins continuous with mesial margin of lateral lobe; chelipeds with upper surface of carpus tuberculate C. tuberatum Cyclozodion angustum (A. Milne Edwards 1880) Figures 2, 3 Calappa angusta A. Milne Edwards 1880:18 (part).-A. Milne Edwards and Bouvier 1902:123, pi. 24, figs. 5-8; pi. 25, figs. 1-3; p. 125, fixed type locality. -Rathbun 1937:210 (part, selected juveniles).- Williams 1965:154; 1984:273 (part, selected juveniles). Material studied.— MCZ 6653. Juvenile holotype; off Barbados, 183 m; Hassler, 27-30 Dec. 1871.- MCZ 2702. 1 o- (juv.) paratype; off Barbados, 188 m; Blake stn. 273, 1878-79.-MCZ 2917. 1 juv. para- type; N Yucatan, Mexico, 23°13'N, 89°16'W, 154 m; Blake stn. 86, 1877-78. Florida: USNM 101419. 1 9; off Cape Canaveral, 27°30'N, 78°52'W, 421 m; Combat stn. 238, 3 Feb. 1957.-Silver Bay stn. 2480. 1 o-, 2 9; 26°06'N, 79°10'W, 223-229 m; dredge, 9 Nov. 1960.-2445. 1 o- (juv.); Straits of Florida, 24°08'N, 80°08'W, 252 m; dredge, 3 Nov. 1960.-2452. 4 ct, 4 9, 3 9 ovig.; same, 23°30'N, 79°04'W, 228-238 m; dredge, 5 Nov. 1960. Silver Bay stn. 3467. 1 juv.; off Great Bahama Bank, 27°27'N, 79°00'W, 229-274 m; dredge, 25 Oct. 1961.-3502. 1 juv.; S Great Inagua I., 20°54'N, 73°37'W, 137-183 m; dredge, 5 Nov. 1961.-3496. 1 9; same, 20°53'N, 73°42'W, 183 m; dredge, 4 Nov. 1961.-5193. 1 0-, 1 9 (ovig.); Puerto Rico, W Mayaguez, 18°16'N, 67°22'W, 274 m; trawl, 18 Oct. 1963.— Oregroti stn. 2643. 1 juv.; off Virgin Gorda, B.W.I., 18°03'N, 64°27'W, 274-329 m; trawl, 5 Oct. 1959.-6715. 2 cr, 1 9; W Anguilla I., 18°36'N, 63°27'W. 201-238 m; dredge, 30 May 1967.-5015. 2 9 (juv.); off Barbados, 13°02'N, 59°34'W, 201- 247 m; dredge, 20 Sept. 1964.-USNM 110230. 1 9; same, 91-336 m; J. B. Lewis, NR4-2, date un- known.-USNM 110231. 1 juv.; same, NR8-2.- USNM 110232. 1 o-(juv.); same, NR12-4. -Oregon stn. 4932. 1 9; Honduras Banks off Thunder Knoll, 16°06'N, 81°10'W, 165 m; dredge, 9 June 1964.-4928. 1 o-, 1 9 (juvs.); Colombia off Isla Pro- videncia, 14°05'N, 81°21'W, 183 m; dredge, 8 June l9U.-Oregon 11 stn. 10190. 1 9; Nicaragua, off Mosquito Coast, 14°42'N, 81°38'W, 141 m; dredge, 19 Nov. 1968.-10515. 1 9 (ovig.); Guyana, N New Amsterdam, 07°47'N, 57°12'W, 95 m; trawl, 28 Apr. 1969. Description.— Ca.ra.pa.ce convex, slightly more arched in longitudinal than in transverse profile, length 0.94 width; surface densely but smoothly and uniformly covered with closely crowded granules; obsolescent raised tubercles in median longitudinal row on gastric and cardiac regions and in more or less concentric arcs on branchial regions; raised median tract separated from branchial regions by well-defined longitudinal depression at either side extending from protogastric to intestinal region; anterolateral margin regularly convex, minutely granulate; posterolateral margin extended into winglike prolongation bearing 1 large spine pre- ceded by 3 or 4 much smaller spines, and succeeded by a single obsolescent spine and imperceptibly curved sector converging toward obscurely trilobed posterior margin. Front trilobed, broader than orbits; broad central lobe concave in dorsal view, downturned, rounded tip not visible; narrower lateral lobes slightly diver- gent, partly enveloping curved antennular peduncles folded obliquely at slightly less than 45° angle to each other; orbits raised above surrounding region but not markedly so, a single obscure dorsal fissure; mean maximal interorbital distance 0.60 mean max- imal span between principal spines on posterolateral winglike extensions. Chelipeds with ornamentation on extensor surface not well divided into horizontal zones typical of many calappid species; lower margin with almost uniform- ly crowded obsolescent granules merging into a field of similar granules extending over lower 1/2 of sur- face; horizontal row of 3-5 low tubercles subparallel to lower margin; 4 or 5 similar scattered tubercles tending to arrangement in diagonal rows in central 110 WILLIAMS and CHILD: COMPARISONS OF SOME BOX CRABS Figure Z.—Cyclozodion angustum (A. Milne Edwards), 9 ovigerous, Silver Bay stn. 5193: a, carapace, eyes, and part of left cheliped; b, orbital region in frontal view; c, right chela and part of carpus; d, fifth pereopod; h. abdomen, cf. Silver Bay stn. 2452: e, abdomen; f-g, first and second pleopods. Ill FISHERY BULLETIN: VOL. 87, NO. 1 area, and 6-10 more obscure tubercles dorsally near "cockscomb"; a low flattened smooth ridge prox- imolaterally in line with tubercles subparallel to lower margin and with subdistal crest of broad flat- tened teeth on merus, anterior tooth of latter with subrectangular tip, second biconcave acute, third and fourth obsolescent and slightly crenulate. Pereo- pods 2-5 spineless. Abdomen of each sex broadest at segment 3; lat- ter fused with narrower segments 4 and 5 in male, segments in female relatively broader but essentially linear and free; segment 2 of male somewhat tri- lobed, that of female less strongly so, each with scat- tering of obsolescent granules on these members; telson subtriangular. Male pleopod 1 stout, slightly curved and conically elongate, tapering to narrow distal opening with nearby cluster of minute horny spinules; pleopod 2 with slender stylet divided into 2 parts, gently curved proximal part stronger than distal part diverging obliquely mesad, tip only slight- ly exceeding that of pleopod 1 . Measurements in mm.— Carapace: holotype o" length 7.3, maximum anterior width 7.9, maximum span across winglike posterolateral projections 6.5; nontypes, same, smallest o" 17.3, 19.0, 15.1; largest o- 21.5, 22.9, 18.2; smallest 9 19.8, 18.5, 16.1; ovigerous 9 26.4, 24.5, 20.8. CoZor.— Preserved specimens display a sprinkling of tiny pale orange spots on posterior 2/3 of carapace and upper exposed parts of chelipeds. Known rawpe.— Florida off Cape Canaveral to Colombia, off Isla Providencia, and Guayana, 95-421 m. Remarks.— Cyclozodion angustum was originally based on juvenile specimens of quite small size and placed in the genus Calappa. Subsequent authors have followed this lead, attributing the narrowed span across the posterolateral winglike projections in all stages from juvenile to adult to youthful allo- metric phases seen in Calappa. Broadening of the winglike span in C. tortugae actually becomes estab- lished at very early stages, as pointed out above in the discussion of that species. The eyes are relatively larger than in C. tortugae, the orbits less protuberant, and in frontal view the orbits are less elevated above the plane of the beaded anterolateral margin than in that species. Inter- orbital width expressed as a percent of maximum span across the posterolateral winglike projections is significantly higher in Cyclozodion angustum than in Calappa tortugae, another indication of the differ- ential in size of orbits and carapace shape in these two species (Fig. 2A, D), although there is minimal overlap in this ratio for a few specimens. Two ver- sions of this ratio are given for Cyclozodion angus- tum: one for the bulk of material measured and ana- lyzed (Fig. 2D, N=27,x = 0.581, SD = 0.040) and one that includes the very small individuals in the type series (Fig. 2E, N = 30, x = 0.595, SD = 0.060). Except for the range of percentages, in- dicating the relatively larger eyes of the types, there is no difference between the two sets of data. Other features that distinguish C. angustum and Calappa tortugae are found on the chelipeds. The exposed carpal surface is smooth in the former, rough in the latter, and the proximoventral corner of the extensor surface on the palm bears a low rounded crest in the former but an anteriorly sub- rectangular crest in the latter. Cyclozodion tuberatum new species Figures 2, 4 Calappa angusta A. Milne Edwards 1880:18 (part, selected juveniles).— A. Milne Edwards and Bou- vier 1902: 123 (part, selected juveniles). Material siwrfied.— Specimen lots in USNM re- corded by Rathbun (1937) under Calappa angusta (catalog numbers only) plus material added since that time. Bahamas: USNM 234462. Holotype o-; N Little Bahama Bank, 27°55'N, 79°05'W, 183 m; Silver Bay stn. 3466, dredge, 25 Oct. 1961. -USNM 234463. Allotype 9; same.-USNM 234464. Para- type 0-; same, 27°26'N, 78°57'W, 137 m; stn. 3468, dredge, 25 Oct. 1961.-USNM 234465. Paratype ct; Straits of Florida off Great Bahama Bank, 26°06'N, 79°10'W, 223-229 m; stn. 2480, dredge, 9 Nov. 1960. North Carolina: USNM 51070.-101676. 1 9; off Cape Lookout, 137 m.— Silver Bay stn. 3333. 1 o-; off Cape Fear, 33°48'N, 76°34'W, 73 m; trawl, 14 Aug. 1961. Florida: USNM 20028, 68505, 68515, 71370, 71371.-169921. 2 unsexed; off Sebastian Inlet, 80 m.-101415. 1 CT, 1 9 (juv.); Florida Straits, 119 m.-77291. 2 cr; off Key West.-101420. 1 9; same, 73-91 m.-101677. 1 ct; Gulf off W Fla., 31-35 m.— 91140. 2 CT, 1 juv.; same 113 m. Oregon stn. 6040. 1 9; off St. Augustine, 29°47'N, 80°33.5'W, 35 m; dredge, 24 Apr. 1966.-Silver Bay stn. 3704. 1 9; off Cape Canaveral, 28°30'N, 112 WILLIAMS and CHILD: COMPARISONS OF SOME BOX CRABS Figure A.—Cyclozodimi tuberatum new species, c holotype, USNM 234462: a. carapace, eyes, and part of left cheliped; b, orbital region in frontal view; c, right chela and part of carpus; d, fifth pereopod; e, abdomen; f-g, first and second pleopods. 9 allotype, USNM 234463: h, abdomen. 113 FISHERY BULLETIN: VOL. 87, NO. 1 80°02'W, 68-75 m; dredge, 25 Jan. 19e2.-Triton. 1 0-, 1 9 (ovig.); off Palm Beach, 183-229 m; Thomp- son & McGinity, no date.— Same. 1 damaged; off Palm Beach, 55-73 m; 20 Apr. 1950. -Same. 3 juv.; SW Sombrero Lt., 165-183 m; 6 June 1950.- Schmitt stn. 207. 1 o-; Tortugas, 17.7 km (11 mi) S Loggerhead Key, 68 m; dredge, 10 June 1925.— Oregon stn. 4084. 1 9; Gulf of Mexico W Tampa, 27°45'N, 84°27'W, 91 m; dredge, 4 Dec. 1962.- Pelican stn. 143-2. 1 o-; SW Panama City, 29°49.5' N, 86°23'W, 70 m; try net, 5 Mar. 1939. -Silver Bay stn. 2455. 1 9 (ovig.); S Great Bahama Bank, 23°34'N, 79°03'W, 165-188 m; dredge, 5 Nov. 1960.-3502. 1 0-, 1 9; S Great Inagua I., 20°54'N, 73°37'W, 137-183 m; dredge, 5 Nov. 1961.-Orec/ow stn. 4297. 1 9; Surinam off Nieuw Amsterdam, 07°46'N, 54°17'W, 640 m; trawl, 22 Mar. 1963. Description.— Carapace convex, slightly more arched in longitudinal than in transverse profile, length 0.92 width; low tubercles of varied sizes scat- tered more or less symmetrically, much bolder on gastric, cardiac, and anterior branchial regions than on posterior 1/3 and tract within perimeter, similar raised ornamentation on extensor surfaces of chelipeds; tubercles covered with low, smooth, tight- ly packed granules, but surface between elevations more coarsely and less thickly granular; raised me- dian tract on gastric and cardiac region separated from branchial regions by well-defined but shallow depression to either side extending from postorbital to intestinal regions; anterolateral margins regularly convex, rather evenly and closely granular but 2 or 3 remote slightly larger granules along hepatic margins and tendency to development of broad ob- solescent teeth near juncture with posterolateral margin; posterolateral margin extended into wing- like prolongation bearing large spine preceded by 3 or 4 much smaller and increasingly diminished spines, and succeeded by small spine, a rudimentary tubercle, and flared arch over coxa of pereopod 5; posterior margin obscurely trilobed, lateral lobes ex- tended ventrally to flank base of abdomen, intestinal region adjacent to median lobe coarsely granulate. Front trilobed, broader than orbit; central lobe broadly concave, dowTiturned, narrowly rounded tip not visible in dorsal view, sharply granular near tip and on raised margins continuous with mesial margin of lateral lobes, latter directed almost straight forward; slightly curved basal article of antennular peduncles folded at less than 45° to each other. Orbits raised above surrounding region; a single obscure dorsal suture; mean maximal inter- orbital distance 0.52 mean maximal span between principal spines on posterolateral winglike exten- sions. Palm of chelipeds with ornamentation on exten- sor surface not well divided into horizontal zones typical of calappid species; lower margin with almost uniformly crowded, well-formed granules merging into a horizontal field of similar granules extending over lower part of palm and bounded by almost horizontal row of 5 or 6 low tubercles; surface above this covered thickly with obsolescent granules and a scattering of widely spaced low tubercles of varied sizes tending to diagonal arrangement, crowded more closely at base of "cockscomb"; a low, flat- tened, smoothly arched ridge, obliquely situated and sometimes dorsally cupped, at posterolateral corner in line with flattened subdistal crest on merus, lat- ter divided into anterior rectangulo-acute tooth, followed by a biconcave tooth and 2 more lower teeth, all slightly crenulate on margins; field above this crest coarsely granulate; exposed surface of car- pus tuberculate and granulate like palm. Abdomen of each sex broadest at segment 3; lat- ter fused with narrower segments 4 and 5 in male, segments in female relatively broader but essentially linear and free; segment 2 somewhat trilobed and bearing scattered obsolescent granules, segment 3 with much lower relief and low granules scattered laterally; telson subtriangular. Pleopod 1 stout, slightly curved and conically elongate, tapering to narrow distal opening with nearby cluster of minute horny spinules; pleopod 2 with slender stylet divided into 2 parts, gently curved proximal part stronger than distal part diverging obliquely mesad, tip only slightly exceeding that of pleopod 1. Measurements in mm.— Carapace: holotype c length 20.6, maximum anterior width 23.2, max- imum span across winglike posterolateral projec- tions 20.7; nontypes, same, smallest o* 16.0, 16.1, 14.9; smallest 9 12.0, 12.6, 10.9; allotype 9 21.1, 23.1, 21.7. Color.— 'No evidence of persistent minute spots of color as on preserved specimens of Calappa angusta. Known raw^e.— North Carolina off Cape Lookout through Bahamas, eastern Gulf of Mexico, Surinam; 31-188, rarely 640 m. Etymology.— The name is from the Latin "tubera- tus", covered with knobs or bosses. Remarks.— Cyclozodion tuberatum has been con- fused with Calappa tortugae because of the similar- 114 WILLIAMS and CHILD: COMPABISONS OF SOME BOX CRABS ity in ornamentation. However, body proportions of the two species differ, as exemplified by the rela- tionship of interorbital width to maximum span between posterolateral projections of the carapace (Fig. 2A, F, iV = 40, X = 0.519, SD = 0.057). Other differences include shape of the proximo- ventral crest on the extensor face of the cheliped palm, rounded in the former, ending anteriorly in a subrectangular point in the latter, and in shape of the male pleopod 1 (see Figures 1 and 4). Cyclo- zodion tuberatum most closely resembles Calappa angustum, although there are superficial similarities to fossil Calappilia scopuli Quayle and Collins as pointed out below. Paracyclois Miers 1886 Paracyclms Miers 1886:288. Type species, P. milne- edwardsii Miers 1886:288.-Glaessner 1969:R494 (part, not Calappilia). Diagnosis paraphrased and emended. —Carapace about as long as broad, and moderately convex; front narrow and trilobate; median lobe rounded and much broader than lateral lobes; without lateral epibranchial spine or tooth; anterolateral margins regularly arcuate, broadest span anterior to junc- ture with posterolateral margin; each posterolateral margin bearing strongly spiniferous lobe or wing- like projection, width between principal spines on lobes less than greatest width of carapace (postero- lateral winglike prolongations more fully developed in Calappa); axis of principal spine on winglike pro- jection diverging from midsagittal line at angle of 25-40°. Subhepatic regions of carapace concave; channel thus formed communicating with antennary region (and thereby with buccal cavity) by a notch situated between it and inferior wall of orbit. Pos- terior abdominal segments distinct. Eyes large, peduncles short, robust, closely en- cased in oval orbits scarcely raised above surround- ing area; interorbital distance at least 0.40 and usually 0.45-0.60 or more of span between tips of principal spines on posterolateral margin. Anten- nules folding obliquely; antennae with quadrate basal article not reaching frontal margin, flagellum very short. Outer maxillipeds with ischium longer than broad, longer than distally truncate merus with its anterointernal angle distinctly notched. Pereo- pods 2-5 with row of spines on flexor surface of ischium-merus. Remarks.— Miers (1886, emended) considered Paracyclois to be an apparent connecting link be- tween Calappa, Cycloes De Haan 1837, and Platy- mera H. Milne Edwards 1837 in which the merus of the outer maxilliped is distally truncate and bears the next article at its anterointernal angle, which is prolonged in the form of a lobe or tooth; but Paracyclois is distinguished from the first two of the above-mentioned genera by the absence of any lateral spine on the margin of the carapace and the broader basal antennal article, and from Calappa by both the reduced winglike prolongations of the carapace which bear strong spines, and by presence of spines on the flexor margin of the ischium and merus of pereopods 2-5. Key to species of Paracyclois 1. Carapace with 3 obviously projecting lobulate spines on posterior margin P. milneedwardsii Carapace with posterior margin only faintly trilobed P. atlantis Paracyclois atlantis Chace 1939 Figures 2, 5 Paracyclois atlantis Chace 1939:51.-1940:27, figs. 11, 12. Material studied.— Silver Bay stn.— 3467. 1 a; off Grand Bahama Bank, 27°27'N, 79°00'W, 228-274 m; dredge, 25 Oct. 1961.-3510. 2 a; Santaren Channel, 22°55'N, 78°36'W, 273 m; dredge, 7 Nov. 1961. -USNM 81986. 1 9; off Punta Alegre, Cuba, 22°46.5'N, 79°W, 329 m; Atlantis stn. 3419, 30 Apr. 1939.-Oregon stn. 2603. 3 ct, 1 9 (ovig.); Puerto Rico, E San Juan, 18°30'N, 65°55'W, 256-292 m; trawl, 25 Sept. 1959.-5914. 1 o-, 2 9, Leeward Is., W Anguilla I., 18°13'N, 63°19'W, 201 m; dredge, 25 Feb. 1966.-6700. 3 o-; S Barbuda I., 17°27'N, 62°04'W, 248-285 m; trawl, 19 May 1967.-3636. 1 o-; Belize, 17n7'N, 87°59'W, 228 m; trawl, 10 June 1962.-4445. 1 juv.; Netherlands Antilles, S Bonaire, 10°50'N, 68°00'W, 183 m; trawl, 10 Oct. 1963.-4856. 1 9; Colombia, off Barranquilla Is., 11°08'N, 74°23.8'W, 183 m; trawl, 19 May 1964.- Oregon stn. 3585. 1 O"; Panama, Gulf of Mosquitos, 09°12'N, 81°30'W, 247-256 m; trawl, 25 May 1962.-3587. 1 o-; Panama, Canal Zone, 09°18'N, 80°25'W, 137 m; trawl, 29 May 1962.-1983. 1 9; Venezuela off Orinoco R. mouths, 09°53'N, 59°53'W, 228 m; trawl, 3 Nov. 1957.-2294. 1 9; Surinam E of Paramaribo, 07°25'N, 54°08'W, 192-210 m; trawl, 9 Sept. 1958. 115 FISHERY BULLETIN; VOL. 87, NO. 1 Figure i.—Paraeyclois atlantis Chace, o- Silver Bay stn. 3510: a, carapace, eyes, and part of right cheliped; b, orbital region in frontal view; c, right chela and part of carpus; d, fifth pereopod; e, abdomen; f-g, first and second pleo- pods. 9, Oregon stn. 5914: h, abdomen. 116 WILLIAMS and CHILD: COMPARISONS OF SOME BOX CRABS Diagnosis. —Carapace convex longitudinally and from side to side except where posterolateral wing- like extensions occur; surface uneven, elevations roughly falling into 5 longitudinal rows, pair of fur- rows bordering median elevation deepest by far; minutely granular and coarsely punctate except on extreme posterior part where punctations disappear and granules become larger; small posterolateral winglike projections bearing 4 large and 1 or 2 rudimentary spines, 1 rudimentary spine often pres- ent between 2 anterior larger ones, spine next to posteriormost always largest and very much so in juveniles, somewhat curved anteriorly, with tenden- cy for anterior curvature in others as well; posterior margin trilobate in dorsal view, lateral lobes pro- longed ventrally on either side of abdomen. Front deflexed, tip invisible in dorsal view, very slightly wider than greatest diameter of orbit and trilobate, median lobe rounded triangular and lateral lobes very narrow and traversed by notch separating front from orbits; orbital margin very slightly raised above surrounding region; mean maximal inter- orbital distance 0.50 mean maximal span between major posterolateral spines (see Figure 2G). Palm of chelipeds with extensor surface orna- mented in horizontal zones, well defined in lower 1/4 but obscure in upper 3/4; lower margin beaded with sharp granules, progressively raised, spiniform and remote proximally, flanked by narrow band of moderately crowded granules; lower half bearing low scattered protuberances, partly interspersed in granular zone and tending to horizontal arrange- ment, but becoming more widely and somewhat diagonally scattered in upper 1/2; proximolateral corner bearing short oblique obsolescent ridge sur- mounted by 3 or more acute to crenulate spines, most prominent distally; in line with subdistal crest of larger, uneven, ragged spines on merus. Pereo- pods 2-5 with row of almost uniform spines on flexor surface of merus, extensor surface of carpus entire. Abdomen of each sex broadest at segment 3, lat- ter fused with narrower segments 4 and 5 in male though nonfunctional articulations sometimes ap- parent, segments in female relatively broader but essentially linear and free; segment 2 trilobed and rather sharply granular, segment 3 with lower relief and bearing obsolescent granules clustered lateral- ly; telson subtriangular. Male pleopod 1 rather stout, slightly curved and conically elongate, tapering to distal opening; pleopod 2 with slender stylet divided into 2 parts, gently curved proximal part stronger than distal part curved mesially upon itself as a crook, distal half of crook extending beyond tip of pleopod 1 and recurved near tip. Known range.— Grand Bahama Bank to Panama and Surinam, 137-365 m. Measurements in mm.— Carapace: smallest o- length 19.8, maximum anterior width 20.7, max- imum span across posterolateral winglike projec- tions 17.2; same, largest cr, 57.1, 62.2, 50.5; small- est 9, 20.6, 22.2, 18.3; largest 9, 53.2, 58.5, 48.7. Remarks.— See next species. Paracyclois milneedwardsii Miers 1886 Figures 2, 6 Paracyclois milneedwardsii Miers 1886:289, pi. 24, fig. l.-Sakai 1976:134 (Engl, text), 85 (Jpn. text), pi. 41, fig. 2. Calappilia milne-edwardsi.—G\aessner 1969:R494. Material studied. -VS'NM 233655. 2 o-, 2 9; Japan, Shikoku I., Tosa Bay; K. Sakai.-233654. 1 9; Philippines, Balayan Bay, southern Luzon, 13° 47'20"N, 120°43'30"E, 329 m; Albatross stn. 536, trawl, 20 Feb. 1909. -Same. 1 9; S Balayan Town, 141-195 m; trawl, 21 June 1966.-Same. Ice; S Sapating, 270-305 m; trawl, 29 July 1966.-Alba- tross stn. 5453. 2 9; E coast Luzon, San Bernardino Str., NE Legaspi Light, 13°12'N, 123°49'18"E, 267 m; trawl, 7 June 1909.-5242. 2 o- (juv.). Mindanao near Vanivan Is., 06°51'53"N, 126°14'10"E, 349 m; trawl, 14 May 1908. Diagnosis. —Carapace irregularly orbiculate, broadest at a point anterior to midlength of antero- lateral margins, latter sweeping in regular curve to winglike protuberance of posterolateral margins bearing 4 unequal spines, anterior one longest; posterior margin bearing 3 strong flattened lobular spines ornamented with coarse tubercles extending onto adjacent intestinal region; margins behind anterior 1/4 of length tending to be rimmed by nar- rowly upturned, granular lip; median tract separ- ated from branchial regions by rather prominent groove at either side extending from gastric to in- testinal regions; surface granular and ornamented with low, smooth rounded tubercles tending to ar- rangement in concentric arcs diminishing in size toward lateral, posterolateral, and intestinal areas. Front slightly narrower than orbit, trilobed, broadly rounded central lobe with downturned tip not visible in dorsal view, 3 low peripheral lobes on its upper surface; lateral lobes much narrower and slightly divergent to accommodate folded anten- 117 FISHERY BULLETIN: VOL. 87, NO. 1 Figure 6.—Paracyclois milneedwardsii Miers, ct, USNM 233655: a, carapace, eyes, and part of right cheliped; b. orbital region in frontal view, c, right chela and part of carpus; d, fifth pereopod, e, abdomen; f-g, first and second pleopods. ?, USNM 233654: h, abdomen. 118 WILLIAMS and CHILD; COMPARISONS OF SOME BOX CRABS nular peduncles; ocular peduncles short and thick, granulated above; orbital margins slightly raised above surrounding region, mean maximal inter- orbital distance 0.50 mean maximal span between major posterolateral spines (Fig. 2H, TV = 7, SD = 0.051). Palm of chelipeds with ornamentation on exten- sor surface obscurely arranged in horizontal zones; lower margin granulate, sharply so in proximal 3/4; lower 1/3 of surface coarsely granular, becoming less sharply so as it merges into central zone; upper 2/3 bearing obscure diagonal rows of obsolescent tuber- cles in central portion, but stronger and less regu- larly arranged tubercles near base of dorsal "cocks- comb"; a spine near proximoventral corner in line with subdistal row of ragged, forward trending spines on merus. Pereopods with flexor surface of ischium and merus strongly but irregularly spinose; carpus bearing biserial row of smaller spines on ex- tensor surface. Abdomen of each sex broadest at segment 3; lat- ter fused with narrower segments 4 and 5 in male, segments in female relatively broader but essentially linear and free; segment 2 trilobed, less so in female than in male and bearing obsolescent granules close- ly clustered or fused on lobes, segment 3 with much lower relief and obsolescent granules clustered mainly on lobes; telson subtriangular. Male pleopod 1 rather stout, slightly curved and conically elon- gate, tapering to narrow distal opening; pleopod 2 with slender stylet divided into 2 parts, gently curved proximal part stronger than distal part curved mesially upon itself as a rather closed crook, distal half of crook extending beyond tip of pleopod and recurved near tip. Known range.— Japan, Philippines, the type local- ity north of Admiralty Islands (Sakai 1976), 141-349 m for specimens studied. Measurements in mm.— Carapace: smallest o" length 18.2, maximum anterior width 17.3, max- imum span across posterolateral winglike projec- tions 15.4; same, largest c, 53.2, 53.3, 47.8; small- est 9, 21.3, 19.9, 17.7; largest 9, 45.6, 44.8, 40.4. Remarks.— The two species oiParacyclois, basic- ally similar in carapace outline, have relatively larger eyes and orbits than the two species of Calap- pa discussed above (Fig. 2), and the orbits in fron- tal view are less elevated above the plane of the anterolateral margin. Interorbital width expressed as percent of maximum span across the postero- lateral projections is virtually the same in samples of the two species (P. atlantis, N = 20, x = 0.494, SD = 0.044, Fig. 2G; P. milneedwardsii, N = 10, X = 0.496, SD = 0.051, Fig. 2H). Spination of the posterolateral projections is much more slender and remote than in either Calappa or Cyclozodion, and well-developed spination on the chelipeds and ven- tral margin of the ischium-merus of the fifth legs clearly sets them apart from species of these genera. Distribution in two well-separated centers, western Indo-Pacific and Caribbean, seems to reflect an an- cient Tethyan track. Calappilia A. Milne Edwards 1873 Calappilia A. Milne Edwards 1873:434.— Rathbun 1930:7. -Glaessner 1969:R494 (part, not Paracy- clois). Ross and Scolaro (1964) summarized scattered references to fossil species of Calappilia known up to that time, Glaessner (1929) compiled a listing and an overview (1969), and Quayle and Collins (1981) gave notes along with description of an additional species. We reviewed all references to these species, and examined selected species (*) in the paleon- tological crustacean collection of the USNM in order to compare features of Calappilia with those of other genera treated herein. Five species of Calappilia are known from the western hemisphere: *C. hondoensis Rathbun 1930, Upper Eocene, Calif.; C. bonairensis Van Straelen 1933, Upper Eocene, Bonaire, Netherlands, West Indies; *C. diglypta Stenzel 1934, Middle Eocene, Tex.; C. sp.? Roberts 1956, Lower Eocene, N.J.; *C. brooksi Ross and Scolaro 1964, Upper Eocene, Fla. Seven species and one variety are known from Europe: C. verrucosa A. Milne Edwards 1873, the type species, and C. sexdentata A. Milne Edwards 1876, Middle Oligocene, SW France; C. perlata Noelting 1885, Lower Oligocene, Germany; C. in- cisa Bittner 1886, Middle Eocene, Italy; C. dacica Bittner 1886, Middle-Upper Eocene, Hungary; C. dacica var. lyrata Lorenthy and Beurlen 1929, Upper Eocene, Hungary; C. vicetina Fabiani 1910, Upper Eocene, Italy; C. scopuli Quayle and Collins 1981, Upper Eocene, England. Two species are known from the East Indies: C. bomeoensis Van Straelen 1923, Middle Eocene, Borneo; C. bohmi Glaessner 1929, Upper Eocene, Java. Diagnosis.— For purposes of comparing Calap- pilia with Calappa, Cyclozodion, and Paracyclois, 119 FISHERY BULLETIN: VOL. 87, NO. 1 we paraphrase essential features of A. Milne Ed- wards's original description. Near Calappa and Mursia; distinguished from former because carapace not extended above ambu- latory legs (Fig. 7) in manner of a shield, and from latter by absence of large spines laterally prolonged beyond cephalothoracic shield; front very narrow and ornamented with 2 small slightly divergent points very similar to those of Calappa; [orbital] border cut by two narrow fissures. Figure l.—Calappilia brooksi Ross and Scolaro, carapace and left eyestalk; USNM 648599, Upper Eocene, Fla. Carapace very convex, recalling that of Calappa or certain representative Leucosiidae; gastric and cardiac regions separated in lateral portions by deep grooves; hepatic region not clearly delimited; bran- chial region very inflated in anterior part but much narrowed posteriorly, surface covered with coarse tubercles in anterior part; posterior branchial lobe extended, constituting a prominence directed laterally and a little posteriorly; posterior border bearing a tubercle much less developed than bran- chial prominence at level of branchiocardiac groove. Ambulatory legs missing; fragment of chela with very compressed dactyl bearing granular crest, armed at base with large tubercle recalling that developed in Calappa; palm covered with large tubercles analogous to those ornamenting cara- pace, and their size notable compared to those on body. Measurements of selected species in mm. — Cara- 120 pace: C. brooksi length 18.8, width 21.5; C. dacica length 32, width 37; C. hondoensis length 19, width 18.7. Remarks. — The features of Calappilia mentioned by A. Milne Edwards suggest much closer similar- ity to Calappa than to Paracyclois, and the brief diagnosis by Rathbun (1930) confirms this in broad outline. All of the species of Calappilia are small, comparing favorably with the range of sizes shown by the two species of Cyclozodion described here. There is considerable diversity in ornamentation of the carapace among species of Calappilia, with a tendency to development of coarse tubercules dor- sally and along the margins, especially postero- laterally, but minimal development of posterolateral winglike projections, with some exceptions. Lobular tubercles along this margin are usually similar in size, although in C. scopuli (Quayle and Collins 1981:740, pi. 104, fig. 8) there is a developed pos- terolateral spine and, except for the problematic frontoorbital region, a marked similarity to Cyclo- zodion in outline of the carapace. The holotype of Calappilia hondoensis (USNM 371094) has an ob- scure posterolateral spine rather wider than long. Rathbun (1930) pointed out that Milne Edwards's (1873) figure of C. verrucosa is longer than wide whereas the measurements given show it wider than long. The left eyestalk of C. brooksi (USNM 648599, Fig. 7), fossilized projecting forward in its orbit, seems relatively slender compared with eyestalks of both Cyclozodion and Paracyclois, although only a remnant of it may be preserved. On the basis of size, shape, and ornamentation of the carapace, relative thickness of eyestalks, and age, we regard Early Tertiary Calappilia and Recent Paracyclois as distinct. Calappilia scopuli and perhaps C. hondoensis seem to form closer links with Recent Cyclozodion than with Calappa, emphasizing similarities among the latter three genera. ACKNOWLEDGMENTS We thank A. B. Johnson of MCZ for loan of material, R. E. Gibbons for drafting the graph, and G. A. Bishop, B. B. CoUette, R. B. Manning, and N.N. Rabalais for critical review of the manuscript. LITERATURE CITED Chace, F. a., Jr. 1939. Reports on the scientific results of the first Atlantis Expedition to the West Indies, under the joint auspices of WILLIAMS and CHILD: COMPARISONS OF SOME BOX CRABS the University of Havana and Harvard University. Prelim- inary descriptions of one new genus and seventeen new species of decapod and stomatopod Crustacea. Mem. Soc. Cubana Hist. Nat. 13(l):31-54. 1940. Reports on the scientific results of the Atlantis Expe- ditions to the West Indies, under the joint auspices of the University of Havana and Harvard University. The brachyu- ran crabs. Torreia, Havana 4:1-67. 1956. List of mysidacean, amphipod, euphausiacean, decapod, and stomatopod crustaceans. In S. Springer and H. R. Bullis, Collections by the Oregon in the Gulf of Mexico. List of crustaceans, mollusks, and fishes identified from collec- tions made by the exploratory fishing vessel Oregon in the Gulf of Mexico and adjacent seas 1950 through 1955, p. 5-23. U.S. Fish. Wildl. Serv., Spec. Sci. Rep.-Fish. 196. Glaessner, M. F. 1929. Crustacea Decapoda: Fossilium Catalogus. 1. Animalia. W. Junk, Berlin, Pt. 41:1-464. 1969. Decapoda. In R. C. Moore (ed.). Treatise on inver- tebrate paleontology, part R, Arthropoda 4, Vol. 2, p. R399-R533, R626-R628. Univ. Kansas and Geol. Soc. Am., Inc. Hay, W. p., and C. A. Shore. 1918. The decapod crustaceans of Beaufort, N.C, and the surrounding region. Bull. U.S. Bur. Fish. 55 (for 1915 and 1916):369-475, pis. 25-39. HOLTHUIS. L. B. 1958. West Indian crabs of the genus Calappa, with a descrip- tion of three new species. Studies on the Fauna of Curasao and other Caribbean Islands. 8(7). In Uitg. Natuurwet. Studierkring Suriname Ned. Antillen 17:146-186. MiERS, E. J. 1886. Part 49. Report on the Brachyura. Rep. Sci. Res. Voyage of H.M.S. Challenger during the years 1873-76. Zoology 17:i-l, 1-362, pis. 1-29. Milne Edwards, A. 1873. Pages 8-9, pi. 4. In M. Le Comte R. de BouillS, Paleontologie de Biarritz et de quelques autres localities des Basses-Pyr^n^es. C. R. Trav. Congr. Sci. Fr., sess. 39e, Pau, t. 4, f. 3, pt. 1. 1880. Reports on the results of dredging, under the super- vision of Alexander Agassiz, in the Gulf of Mexico and in the Caribbean Sea, 1877, '78, '79. by the United States Coast Survey Steamer "Blake" . . . .VIII. Etudes preliminaires sur les Crustacfe. Bull. Mus. Comp. Zool. Harv. Coll. 8:1-68, 2 pis. MiLNE Edwards, A., and E. L. Bolivier. 1902. Reports of the results of dredging, under the super- vision of Alexander Agassiz, in the Gulf of Mexico (1877-78), in the Caribbean Sea (1878-79), and along the Atlantic coast of the United States (1880), by the U.S. Coast Survey Steamer "Blake" .... XXXIX. Les Dromiacfe et Oxystomes. Mem. Mus. Comp. Zool. Harv. Coll. 27:1-127, 25 pis. Pequegnat, W. E. 1970. Deep-water brachyuran crabs. In W. E. Pequegnat and F. A. Chace, Jr. (eds.). Contributions on the biology of the Gulf of Mexico, p. 171-204. Texas A&M Univ. Oceanogr. Stud. 1. Powers, L. W. 1977. A catalogue and bibliography to the crabs (Brachyura) of the Gulf of Mexico. Contrib. Mar. Sci. 20(suppl.): 1-190. QUAYLE, W. J., and J. S. H. Collins. 1981. New Eocene crabs from the Hampshire Basin. Pale- ontology 24:733-758, pis. 104-105. Rathblin, M. J. 1930. Fossil decapod crustaceans from Mexico. Proc. U.S. Nat. Mus. 78(8):1-10, pis. 1-6. 1933. Preliminary descriptions of nine new species of ox- ystomatous and allied crabs. Proc. Biol. Soc. Wash. 46: 183-186. 1937. The oxystomatous and allied crabs of America. U.S. Nat. Mus. Bull. 166:1-278, 86 pis. Ross, A., and R. J. Scolaro. 1964. A new crab from the Eocene of Florida. Q. J. Fla. Acad. Sci. 27:97-106. Sakai, T. 1976. Crabs of Japan and the adjacent seas. Kodansha Ltd., Tokyo, 773 p. (Engl, text), 461 p. (Jpn. text), 251 pis. (many colored), as 3 separate volumes. Williams, A. B. 1965. Marine decapod crustaceans of the Carolinas. U.S. Fish Wildl. Serv., Fish. Bull. 65:i-vi, 1-298. 1984. Shrimps, lobsters, and crabs of the Atlantic coast of the eastern United States, Maine to Florida. Smithson. Inst. Press, i-xviii, 550 p. 121 YELLOWFIN TUNA, THUNNUS ALBACARES, CATCH RATES IN THE WESTERN PACIFIC Tom Polacheck' ABSTRACT The surface fishery for yellowfin tuna, Thunnus albacares. in the western Pacific has increased drama- tically since 1978. Catch and effort statistics from the Japanese purse seine and longline fisheries are examined in terms of changes in catch rates and the interaction between these two fisheries. In spite of a 10-fold increase in surface catches to around 100,000 metric tons per year, purse seine catch rates have remained relatively constant. Longline catch rates since 1980 have been declining, with the excep- tion of high rates in 1983. Comparison of purse seine and longline catch rates within the same area and time period indicated no relation between them and suggests that the yellowfin tuna stocks are not homogeneous with respect to the two gears. In addition, observed changes in longline catch rates within areas of the western Pacific appear not to be related to the magnitude of the purse seine catches within these areas. The results provide no direct evidence for any interactions between the two gears, but whether purse seine catches are contributing to the possible, overall decline in longline catch rates remains an open question. Purse seine catches of yellowfin tuna, TMmnus alba- cares, in the western Pacific have increased from 8,000 to 10,000 1 (metric tons) in 1978 (Habib 1984^) to estimates of around 100,000 t in 1984. Prior to the advent of purse seining, the main vessels har- vesting yellowfin tuna in this region were the Japa- nese, Korean, and Taiwanese longliners. Longliners still continue to harvest significant amounts of yellowfin tuna (an estimated 60,000 t in 1984). The effect of this 10-fold increase in purse seine catches since 1978, both on the overall stocks of yellowfin tuna in the western Pacific and the effect of the purse seine catches on the longline fisheries, is un- known, but the status of yellowfin tuna stocks is a critical question for a number of reasons. Yellowfin tuna represent the second largest fishery resoiu-ce for the tropical western Pacific area. Yet, there is no adequate assessment of the magnitude of the harvestable catch for the region, while yellowfin tuna stocks in other regions appear vulnerable to overexploitation by purse seiners (lATTC 1979, 1980, 1981, 1982; Fonteneau and Diouf 1983; Au 1983). In addition, about two-thirds of the yellowfin tuna longline catch is targeted for the Japanese Sashimi market and, as such, has an economic value 'Tuna and Billfish Assessment Programme, South Pacific Com- mission, B.P. D5, Noumea Cedex, New Caledonia; present address: Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. ^Habib, G. 1984. An overview of tlie purpose seine tuna fish- ery in the central/western Pacific and development opportunities for island states. Workshop on National Fishing Operations, Tarawa, Kiribati, 28 May-4 June 1984. Manuscript accepted August 1988. Fishery Bulletin, U.S. 87:123-144. exceeding that of the purse seine-caught fish. Long- liners harvest older and larger fish than purse seiners (Cole 1980). In the present paper, the most recent data available on the catch and effort for yellowfin tuna are examined for information on the current yellowfin tuna stocks and on the interaction between longline and purse seine fisheries. METHODS Data The data available for examining catch rates come from records of daily catch and effort supplied by vessels to individual island states in the western Pacific as part of access arrangements which allow vessels to fish within the 200-mile EEZ's (Exclusive Economic Zone) of these states. These catch records have been subsequently transmitted to the Tuna and BOlfish Assessment Programme of the South Pacific Commission (SPC), and have formed the core of the regional statistical data base. Data are only supplied as a requirement of access for fishing within EEZ's. While some vessels include activity in international waters in their reports, the available data are rela- tively incomplete for these waters. Also, for some states in past years, adequate data reporting was not included in the access ag'-eements. In addition, prior to 1984 almost no data are available from United States and some other eastern Pacific purse seiners. Because of incompleteness and limitations in the 123 FISHERY BULLETIN: VOL. 87, NO. 1 available data, the analyses in this present paper are based on catch rates for Japanese vessels. These data form the most extensive and complete set of data currently available to the SPC. Records of daily fishing activity go back as far as the second half of 1978 for the longline fishery and 1979 for the purse seine fishery. However, these earliest data are not complete and need to be interpreted with caution. The stock or subpopulation structure and their geographic limits for yellowfin tuna in the Pacific are unknown despite considerable tagging and gene- tic research (Cole 1980). However, a single stock spanning the entire Pacific is considered unlikely. For the present paper, the geographic boundaries used for the western Pacific are from long. 130°E to 180°E and from lat. 10°S to 15°N. For the Japa- nese purse seine fishery, this area encompasses vir- tually all of the reported catch and effort data. For the Japanese longline fishery, this represents an area in which the fishery has been relatively con- sistent and its reporting fairly complete. Catch Rates Catch rates or catch per unit effort are calculated below as a measure of relative abundance. An ex- tensive literature exists on the use of catch rates as abundance indices (Gulland 1956a; Beverton and Holt 1957; Paloheimo and Dickie 1964; Allen and Punsley 1984). However, the question of the rela- tion between catch rates and abundance for these yellowrfin tuna fisheries needs further research (see Discussion). For longlining, the effort measure used here is the number of hooks set (in thousands). Catch is reported as the number of fish caught. For purse seining, the effort measure used is the number of days in which vessels made a set or were actively searching for schools of tuna. The catch is recorded in metric tons. In the earliest purse seine data, there may be an underestimation of effort, as it is not clear whether days in which vessels were searching for fish, but did not catch any, were accurately reported. The average catch rates and their variances with- in any statistical stratum were calculated as the weighted mean of the observed catch rates for all cruises within the stratum. Thus, an individual cruise's catch rate within a stratum constitutes the primary sampling unit or replicate in the analyses below. The weights used were equal to a vessel's fishing effort. For the estimates of the mean catch rate, this is equivalent to the sum of the total catch divided by the sum of the total effort within a stratum. Various temporal and areal stratifications of the data have been considered. Monthly, quarterly, and annual stratifications are examined. When the data were stratified by area, geographic strata were defined as rectangular areas of 2.5° of latitude and 10° of longitude. These strata were chosen because preliminary analyses indicated that there was much greater variation both in effort and catch rates lati- tudinally than longitudinally. If smaller areas are selected, there tends to be too little data in many of the strata for meaningful analysis. There are two statistical reasons for stratifying data: 1) to eliminate biases due to unequal distribu- tion of sampling effort in strata with different means, and 2) to reduce the variance associated with the estimate of the mean. The first reason is a primary concern in calculating catch rates from fisheries data since the distribution of fishing effort both spatially and temporally is likely to be related to catch rates (i.e., fishermen probably concentrate on when and where the fishing is best). In order to estimate an average catch rate for time periods and areas of interest, the estimates of the catch rates in the various strata need to be com- bined. For stratified data, an estimate of the aver- age catch rate across strata is the weighted mean of the average catch rate within each stratum, where the weights are proportional to the magni- tude of a stratum (Snedecor and Cochran 1967). The geographical and temporal stratifications presented below were considered to be equal in area and time. (This is not strictly true both because of land masses and differences in the length of a degree of longi- tude at different latitudes. For two of the geograph- ical strata, the amount of land area of Papua New Guinea is large and these two strata should prob- ably be given smaller weight in any extensions or refinements to the estimates presented below.) When all strata are of equal magnitude, the aver- age catch rate across strata is the simple average of the within-strata estimates. Similarly, in this situation, an estimate of the variance is the average of the variance estimates for each stratum (Snede- cor and Cochran 1967). Because catch and effort statistics are not derived from a well-designed and controlled sampling ex- periment, there is not an a priori single best esti- mate for the average catch rate covering large areas and time periods. Thus, when considering estimates of the annual average catch rates, a set of different estimates based on various areal and temporal strat- ifications are presented. Comparison of the esti- 124 POLACHECK: YELLOWFIN TUNA CATCH RATES mates for different stratifications of the data may indicate possible sources of bias and can provide some indication of the robustness of any temporal trends suggested by any single set of estimates. Another approach for dealing with possible biases due to unequal distribution of fishing effort is to calculate standardized catch rates using a general linear model (Gulland 1956b; Robson 1966; Allen and Punsly 1984). The advantage of this approach is that well-developed, standard statistical procedures can be employed to test for significant differences in catch rates over time where the effect of other fac- tors on catch rates have been taken into account. Disadvantages of this approach include: 1) the data may be nonnormal even when transformed, 2) ef- fects may not be simply additive (or multiplicative if a logarithmic transformation is used), and 3) the design matrix is almost always unbalanced and incomplete. Extensive attempts were made to fit a general linear model to the catch rate data presented here. While the model was successful in greatly reducing the total sums of squares (e.g., an i?'^ as high as 0.80), in all cases, the models included significant and large interaction effects between year and area, and between year and season. Such interactions are an indication of changes in availability and distri- bution between years and are not surprising given the large El Nino of 1983. When large interaction terms exist in a model, particularly when it is un- balanced and incomplete, direct interpretation of the main effects (in this case year) is problematical. An alternative to estimating the main effects in this situation is to develop fjy -models (Searle 1971) to compare directly the average effect between those combination of cells which are of interest. Concep- tually this approach is similar to the stratified means approach developed above, but the calculation of the variance for the stratified means makes no assump- tion about the equality of the variance between cells. Because of the similarity of these two approaches and the problems with traditional general linear model estimates for unbalanced and incomplete data, the results of the general linear model have not been included in the present paper. that yellowfin tuna are a homogeneous stock with respect to the two fisheries. For this analysis, it is important that relatively fine scale temporal and area strata be used in order that differences in abun- dance between areas and time do not mask any rela- tionship. Comparison of quarterly longline and purse seine catch rates are made for each individual 2.5° X 10° rectangular area in which there were at least five quarters with a reasonable amount of effort by both gears (i.e., 5 days of purse seine effort and 20,000 longline hooks). The second approach involves the comparison of changes in longline catch rates in different areas to the purse seine catches that have occurred within these areas. This approach is a direct test of whether any reduction in longline catch rates can be detected as a result of the large catches by purse seiners. A fundamental assumption of this approach is that the stocks of yellowfin tuna within the areas being com- pared are largely spatially distinct or mixing only slowly. If the stock being fished is a homogeneous mixture, then no purse seine-induced differences between areas would occur. For this second approach the percentage change in the average 1984-85 longline catch rate, relative to the average 1979-81 catch rate, are calculated for each of the 2.5° x 10° rectangular areas. The average catch rates within an area for the periods 1979-81 and 1984-85 were calculated as the sim- ple average of the quarterly rates for an area. The percentage changes between these two periods are then examined in relation to past purse seine catches that have occurred in these areas. These two time periods were chosen for this comparison in order to see whether there has been differential and consis- tent long-term changes in abundances, and if so, whether these changes can be related to the distri- bution of purse seine catches. It should be noted that these two approaches are meant to test for specific, possible localized inter- actions (either temporal or spatial). They are not meant as an exhaustive examination of the inter- actions between these two gears, but as feasible analyses given the short time series and limits of the current data. Interactions The relationship between the longline and purse seine fisheries is considered in detail from two dif- ferent approaches. In the first, catch rates of purse seiners and longliners operating in the same area during the same time period are compared. In this case a strong positive relationship would suggest RESULTS Purse Seine Catch and Effort Effort by Japanese purse seiners increased steadi- ly through the first half of 1982 to around 450 days per month (Fig. lA). Since 1982, levels of effort have remained relatively steady and have fluctuated 125 FISHERY BULLETIN: VOL. 87. NO. 1 O Q 600 500- 400 ■^ 300 O »- 200- 100- EFFORT BY JAPANESE PURSE SEINERS liiii|iiii I I t r I I IT I I I I I I T I I I M I I I I I I M I 1979 1980 1981 1982 1983 1984 1985 1986 5000 1 4000 3000- u 2000 2 1000- YELLOWFIN CATCH - JAPANESE PURSE SEINERS t M t 1 i* TTTfrTT1 n T|rTTTTTTTII1III M IIIIIll|ITII I TTTT T TT T T I TT T T I T H T T T I T T I I T 1 1 1 [ I I I I T 1 I I I T T I 1 1 T I I 1 T I T 11 1979 1980 1981 1982 1983 1984 1985 1986 Figure 1.— Monthly catch of yellowfin tuna and effort statistics for Japanese purse seiners in the western Pacific based on data currently reported to the South Pacific Commission. around this level. (Note that the apparent drop in effort for 1986 is an artifact due to time lags in receiving catch reports.) The total catch of yellowfin by Japanese purse seiners roughly parallels the temporal distribution of effort (Fig. lA, B). Overall, the corresponding catch rates have remained fairly constant with the lowest rates observed in 1983 (Fig. 2). Table 1 presents a range of estimates of the an- nual catch rates for the various areal and temporal stratifications of the data. There are no consistent differences among the different stratifications 126 POLACHECK: YELLOWFIN TUNA CATCH RATES 35- 30 25 H >- < Q q: 111 20 15 10 5- II I I I I I J I I I I I M ! T 1 I I I I I I ITTTTTTJT1 I I T 1 I • I I I J 1 11 1 I I I I I I r I rl I T I T 1 1 T I T I I T T 1 T T 1 I ri I I T I 1 1 I I I T 1 I T T TTTTTTTTTTI 1979 1980 1981 1982 1983 1984 1985 1986 Figure 2.— Estimates of the monthly catch rates of yellowfin tuna for Japanese purse seiners (metric tons per day of effort) in the western Pacific. Error bars represent the estimates of one standard error. Table 1 .—Comparison of annual estimates of the overall average catch rate of yellowfin tuna (metric tons per day) by Japanese purse seiners in the western Pacific based on various areal and temporal stratifications of the data. Values in parentheses are estimates of standard error and n is the number of strata contained within each estimate. Stratum with less than five days of effort are not included. Year No areal or temporal stratification Stratified by month Stratified by quarter Stratified by area Stratified by quarter and area 1979 5.50 5.39 7.35 5.26 5.48 (0.59) n = 1 (0,46) n = 6 (1.59) n = 4 (0.53) n = 6 (0.85) n = 9 1980 5-02 4.80 5.21 5.18 4.66 (0.31) n = 1 (0.42) n = 11 (0.38) n = 4 (1.45) n = 7 (0.30) n = 13 1981 5.98 6.35 6.18 4.15 5.34 (0.32) n = 1 (0.38) n = 12 (0.32) n = 4 (0.39) n = 15 (0.51) r? = 29 1982 4.82 4.74 4.97 4.69 4.91 (0.23) n = 1 (0.22) n = 12 (0.23) n = 4 (0.60) n = 11 (0.34) n = 24 1983 3.83 3.89 3.74 3.24 3.48 (0.20) n = 1 (0.21) n = 12 (0.20) n = A (0.29) n = 14 (0.27) n = 25 1984 4.77 4.75 4.85 3.70 3.39 (0.23) n = 1 (0.16) n = 12 (0.20) n = A (0.20) n = 13 (0.17) n = 36 1985 5.04 5.09 5.00 4.61 5.36 (0.25) n = 1 (0.20) n = 12 (0.22) n = A (0.24) n = 18 (0.26) f> = 41 1986 6.26 6.26 6.19 4.84 5.26 (0.32) n = 1 (0.30) fJ = 7 (0.32) n = 3 (0.29) n = 15 (0.27) n = 21 127 FISHERY BULLETIN: VOL. 87, NO. 1 within a year. The larger differences that do exist tend to include stratification by area. If a normal distribution is assumed, the only significant dif- ferences at a 0.05 probability level among the stratifications within a year (i.e., 1981, 1984, and 1986) would be in stratifications which include area. Sooon 4000- 3000- o o o 1 2000- 1000 0* o %° o °ooo ^"^^ o o o ° 9, o o O o o o o °o o ^ ° ^8s O o 100 200 300 400 500 Effort — days fished plus searched o 5000-1 4000- 3000 •t 2000 « 1000 100 200 300 400 500 Effort — days fished plus searched Figure 3.— The relationship between monthly yellowfin tuna catch and effort by Japanese purse seiners. The points have been connected in the temporal sequence in which they occur. 128 POLACHECK: YELLOWFIN TUNA CATCH RATES However, for estimates stratified by both area and season, only the 1984 estimate would be significantly different from any of the other stratified means within a given year. A lack of consistent differences among the various stratifications within years does not mean that siginificant area and seasonal differences may not exist, but only that whatever effects may exist tend to balance in the present data. Among the various annual estimates in Table 1, the estimates for 1983 tend to be the lowest (perhaps reflecting the large El Nifio of that year), while those for 1979 and 1986 tend to be the highest. While the length of the time series is short, there is no indica- tion within any of the stratifications of an overall temporal trend in the annual estimates. Relationship Between Purse Seine Catch and Effort A production plot of total monthly catch versus total monthly effort suggests that monthly catch rates can be highly variable and that months with the highest effort tend to have lower catch rates (Fig. 3). Thus, the catch rates in the 6 months in which the total effort exceeded 500 days of effort are all below the overall mean catch rate (Fig. 4) 12 11 - 10 9 8 > o ■D 7 |- - 5000- 1000 2000 3000 4000 Effort — days fished plus searched 5000 Figure 6.— The relationship between Japanese yellowfin tuna catch and purse seine effort based on annual statistics. The points have been connected in their temporal sequence. 130 POLACHECK: YELLOWFIN TUNA CATCH RATES 10000^ 8000- 6000 EFFORT BY JAPANESE LONGLINERS A 'l"""""'l"""""'l"' 'I' "" I' n i n | n iiiiiiiii| nnn |i 1979 1980 1981 1982 1983 1984 1985 1986 150000 100000 X '1 o m O o o 50000 CATCH BY JAPANESE LONGLINERS B ^^^ [ n I I I 1 1 1 i r I I 1 1 I I n 1 1 I I I [ 1 1 I I I I I I I I I I 1 1 I I I | i i i i i | i M I r i i 1 1 i 1979 1980 1981 1982 1983 1984 1985 1986 Figure 7.— Monthly catch of yellowfin tuna and effort statistics for Japanese longliners in the western Pacific based on data currently reported to the South Pacific Commission. 131 FISHERY BULLETIN: VOL. 87, NO. 1 O o X o o o o a o o >- 30 25 20 15 10 5- TTTTTTTTTTTTr TJTTTTTTTTTTTJTTrrr TTTTTTm 1979 1980 1981 1982 1983 1984 1985 1986 Figure 8.— Estimates of the monthly catch rates for Japanese longliners (number of yellowfin tuna per 1,000 hooks) in the western Pacific. Error bars represent estimates of one standard error. Table 2.— Comparison of annual estimates of the overall average catch rate of yellowfin tuna (number/1 ,000 hool ° o o OOOO OoOOOOOo oQo o o o oOOOO" » ooo o oooOO° ° o ooooOOooo o oooOoooo oQQQQoooQoOOooo ooo oOOOCOOOOOo o o o » o o = )00* oQo o ° O O O O o O o o o o o o o oooooo-oo o 00«0000» • oOOo • ooooOOOQOOOOOOOOoo • ooo o ooOOQQQOOOOOo o o o o o o o o OO o o o OO00^^°^^*-*° OOoQooooo oo o oOOOOo o oOo o o o oOOo o o o O t 1 - 80 tonnes o 81 - 250 tonnes o 251 - 700 tonnes O More than 700 tonnes Figure 12.— The distribution of Japanese yellowfin tuna purse seine catches by one degree square: A) 1979 to 1983; B) 1979 to 1985. 138 POLACHECK: YELLOWFIN TUNA CATCH RATES evidence that surface yellowfin tuna stocks have declined. However, the time sequence of changes in catches in relation to changes in effort are not those that would be expected if these catch curves were a reflection of the overall population dynamics. Thus, when the temporal sequence of changes in catch and effort are considered by connecting the points in Figure 3, the resulting pattern suggests that during a time interval of one month, a large change in effort results in correspondingly large changes in the catch rates. If these changes in catch rates reflected changes in the overall yellowfin tuna abundance, it would mean that catches of 3,000- 5,000 t represented a very significant proportion of the total yellowfin tuna stock and that a very rapid recovery of the yellowfin tuna stocks (i.e., during the course of a month) can occur with reductions in effort. Neither of these conclusions seem reasonable. Also, the fact that there is no evidence that catch rates are lower at the highest effort levels so far experienced when the data are combined into an- nual statistics, further suggests that the catch curves based on monthly and quarterly statistics do not reflect the overall population dynamics. The apparent reduction in catch rates at the high- est effort levels based on the monthly or quarterly stratification is an interesting phenomenon warrant- ing further investigation. The reduction in catch rates at these highest effort levels does not appear to be the result of increased handling time at higher effort levels. The number of sets per day has re- mained relatively constant and unrelated to the total number of days fished. Two possible explanations for the decline in monthly catch rates with higher effort are localized depletions and interactions with skipjack tuna catches. In this regard, it is interesting to note that monthly or quarterly catch curves for skipjack tuna, Katsuwonus pelamis, from this same fishery do not show this apparent decline in catch rates at highest effort. The lack of decline in the catch rate for skipjack tuna is another indication that the decline observed for yellowfin tuna is not due to handling time. Longline Longline catch rates in 1984 and 1985 are substan- tially lower than those in 1979. Whether this de- crease represents a general long-term decline is not possible to determine without a longer time series of data. Interpretation of the temporal trend de- pends partially upon whether the observed rates in 1983 are attributable to the large El Nino of 1983 or whether they are a measure of the random vari- ability in the fishing process. The magnitude of the increase observed in 1983 is much larger than might be expected given the observed variability both be- tween and within months (the latter is indicated by the error bars in Figure 8). While it is tempting and even reasonable to attribute the high rates in 1983 as an El Nino effect, the length of the current time series and available information on the effects of El Niiio on yellowfin tuna are insufficient to objectively resolve whether the high 1983 rates are the results of El Nifio. Caution in interpreting longline catch rates as directly reflecting changes in population abundances is also warranted. While the operational procedures in tuna longlining would appear not to be very sus- ceptible to inducing a nonlinear relationship between abundance and catch rates (i.e., handling time is not a major factor and the length of a single longline insures that effort can not be highly concentrated in space). However, concerns have been raised about potential hook competition at higher densities (Rothschild 1967; Au 1985). More importantly, long- liners target different depths depending upon local conditions, market factors and the relative abun- dance of different species. In addition, the fact that surface catches in the Atlantic were able to greatly exceed previous catches of large yellowfin tuna by longliners despite the fact that longline catch rates had declined steeply suggests that the rela- tionship between availability to the different gears versus overall abundance is not simple (Fonteneau 1981). In order to gain a broader temporal perspective to compare the current catch rates, longline hook- ing rates from 1962 to 1980 for the same area con- sidered in this paper are plotted in Figure 14 based on published data by the Fisheries Agency of Japan (1962-80). Longline hooking rates were generally declining through the mid-1970s and then appear to have entered a period of recovery. Because of the commencement of the purse seine fishery in 1980, interpretation of the overall long-term temporal trend is confounded and depends upon whether the apparent increase in the 1970s was a true recovery or a reflection of the variability that can be expected in this fishery. Interaction The results presented in this paper suggest that the relation between longline and purse seine fish- eries is complex. The above discussion indicates that the current data is insufficient to determine whether a general decline is occurring in longline catch rates. 139 FISHERY BULLETIN: VOL. 87, NO. 1 January March Figure 13.— Examples of monthly prospective block drawings showing the distribu- tion of fishing effort by one degree square for Japanese purse seiners. The 140 POLACHECK: YELLOWFIN TUNA CATCH RATES February April Figure 13.— Co»i(tnj«'d.— figures presented are for the first four months of 1984. The boundaries of the area are from lat. 10°S to 15°N and long. 130°E to 180°E. 141 FISHERY BULLETIN: VOL, 87, NO. 1 25 n o o x: o o o o « 20 15- 10 — 1 — 1 — I — I — 1 — I— 1 — 62 64 66 68 — I — I — I — r — I I I — I — I — I — I — I — I — I — I — I — I — 70 72 74 76 78 80 82 84 86 Figure 14.— Estimates of annual catch rates for Japanese longliners (number of yellowfin tuna per 1.000 hooks) in the western Pacific. The solid line represents stratified estimates based on five degree square areas from published data by the Fisheries Agency of Japan (1962-80). The dotted line represents the estimates stratified by area from Table 5 based on data held by the Tuna and Billfish Assessment Programme of the South Pacific Commission. Even if a general decline is occurring, it would not be possible to evaluate whether the purse seine catch is a likely cause of the decline without either more detailed information on the age structure of the catches or a much longer time series of data. Based on the comparison of catch rates within the same area and time period, yellowfin tuna do not appear to be a homogeneous stock with respect to purse seining and longlining. The lack of any rela- tionship at a fine spatial and temporal scale could be due to 1 . factors affecting vulnerability to surface gear are unrelated to factors affecting vulnerability to longline gear, or 2. those portions of the yellowfin tuna population being exploited by the purse seine fishery (i.e., primarily 2-3 year old fish) have a spatial-tem- poral distribution which does not coincide with that for the older and larger yellowfin tuna be- ing harvested by longliners. In reality, probability both of these factors, plus ran- dom elements in the fishing process are contributing to the apparent lack of any relationship. The fact that the observed changes in longline catch rates within areas appear not to be related to the purse seine catches taken from that area may be due to any number of factors. Some possible hypotheses include 1. The level of exploitation by purse seiners within any of the areas considered has been insufficient to affect a significant decline in longline catch rates. 2. Given the difference in the size and age of the fish exploited by the two fisheries, a time lag would be expected before any effect could be ob- served and the presently available time series may be too short to detect the effects. 3. There is a large amount of movement of yellow- fin tuna so that the yellowfin being harvested by longliners are not merely the escapement from the purse seine fishery within that area. 4. There are two independent stocks or substocks of yellowfin tuna— a deep and a surface one- each of which is primarily vulnerable to only one gear type. 5. The available purse seine catch statistics are in- complete and areas in which the greatest decline in longline catch rates have occurred may in fact be areas where large, unreported purse seine catches have occurred. 6. The main areas in which the largest decline in longline catch rates have occurred border the EEZ's of the Philippines and Indonesia. The 142 POLACHECK: YELLOWFIN TUNA CATCH RATES Philippine surface tuna fishery has increased dramatically and there is a suggestion that over- fishing has occurred there (Floyd and Pauly 1984). It is not possible with existing knowledge to dis- tinguish between these hypotheses while available data suggest that all of the above may be contrib- uting to the observed results. Thus, for example, very limited tagging data from the western Pacific suggest that the yellowfin tuna stocks may be very large and that yellowfin tuna caught by longliners in the Pacific can travel long distances from their initial place of captxire. Ten yellowfin tuna tagged by the SPC were recaptured by longliner and trav- eled an average distance of 1,280 miles from their point of release (unpubl. data). Tag experiments from the Atlantic yielded no returns by longliners which suggests that yellowfin tuna cannot be con- sidered as a single homogeneous stock in that ocean with respect to the different gears (Fonteneau 1981). Yet, the fact that surface tagged fish have been recaptured by longliners in the Pacific means that they are not totally distinct. A better under- standing of the interactions between longline and purse seine fisheries is dependent upon both a more complete set of catch and effort statistics and a longer time series of data, plus biological informa- tion from other sources. The present low longline catch rate and the importance of longline fisheries in the South Pacific make this a question of imme- diate concern. ACKNOWLEDGMENTS The catch and effort data base used in the anal- yses in this paper would not exist without the coop- eration and help of the fisheries officers from the individual island states of the western Pacific. Their efforts are gratefully acknowledged. In addition, present and past staff of the Tuna and Billfish Assessment Programme of the South Pacific Com- mission were instrumental in the creation and main- tenance of this data base and also provided useful reviews and comments on drafts of this manuscript. I also wish to thank Veronica van Kouwen for her help in the preparation of this manuscript. LITERATURE CITED Allen, R.. and R. Punsly. 1984. Catch rates as indices of abundance of yellowfin tuna, Thunnus albacares, in the eastern tropical Pacific. lATTC Bull. 18:303-379. Au. D. 1983. Production model analysis of the Atlantic yellowfin tuna {Thunnus albamres) fishery. ICCAT, Coll. Vol. Sci. Pap. 18:177-196. 1985. Interpretation of longline hook rates. ICCAT, Coll. Vol. Sci. Pap. 25:377-385. 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- Cole, J. S. 1980. Synopsis of biological data on the yellowfin tuna, Thun- nus alhacares (Bonnaterre, 1788), in the Pacific Ocean. In W. H. Bayliff (editor). Synopses of biological data on eight species of Scombrids, p. 71-150. lATTC Spec. Rep. 2. Inter-American Tropical Tuna Commission, La Jolla, CA, U.S.A. Farman, R. S. 1987. Report on observer activities on board a Japanese group purse seining operation (24 March-20 April 1984). Tuna Billfish Assess. Program. Tech. Rep. No. 19. South Pacific Commission. Noumea, New Caledonia. Fisheries Agency of Japan. 1962 to 1980. Annual report of effort and catch statistics by area on Japanese tuna longline fishery. Fish. Agency Jpn., var. pagination. Floyd, J. M.. and D. Pauly. 1984. Smaller size tuna around the Philippines can fish aggregating devices be blamed? Infofish Mark. Dig. 5/84: 25-27. Fonteneau, A. 1981. Dynamique de la population d'albacore (Thunnus alba- cares) de rOcean Atlantique. These doctorat, univ. Paris, 324 p. Fonteneau, A., and T. Diouf. 1983. Analyse de I'etat des stocks d'albacore de I'Atlantique au 30 September 1982. ICCAT, Coll. Vol. Sci. Pap. 18: 197-204. Gillett, R. D. 1986. Observations on two Japanese purse seining operations in the equatorial Pacific. Tuna Billfish Assess. Program. Tech. Rep. No. 16, 20 p. South Pacific Commission, Nou- mea, New Caledonia. GULLAND, J. A. 1955. Estimation of growth and mortality in commercial fish populations. Fish. Invest. Minist. Agric. Fish. Food (G.B.). Ser. II, Vol. 18, 46 p. 1956a. The study of fish populations by the analysis of com- mercial catches. A statistical review. Rapp. P. -v. R6un. Cons. int. Explor. Mer 140:21-27. 1956b. On the fishing effort in English demersal fisheries. Fish. Invest. Minist. Agric. Fish. Food (G.B.), Ser. II, Vol. 20, 41 p. 1983. Fish stock assessment. John Wiley, N.Y., 223 p. LATTC. 1979. Annual report of the Inter-American Tropical Tuna Commission - 1978. lATTC, La Jolla, CA. U.S.A., 163 p. 1980. Annual report of the Inter-American Tropical Tuna Commission - 1979. lATTC. La Jolla, CA, U.S.A., 227 p. 1981. Annual report of the Inter-American Tropical Tuna Commission - 1980. lATTC. La Jolla, CA, U.S.A.. 234 p. 1982. Annual report of the Inter-American Tropical Tuna Commission 1981. lATTC, La Jolla, CA, U.S.A., 303 p. Mangel. M. 1982. Search effort and catch rates in fisheries. Eur. J. 143 FISHERY BULLETIN: VOL. 87, NO. 1 Oper. Res. 11:36P366. Robson. D. S. Neyman, J. 1966. Estimation of the relative fishing power of individual 1949. On the problem of estimating the number of schools ships. Int. Comm. Northwest Atl. Fish. Res. Bull. 3:5-14. offish. Univ. Calif. Publ. Stat. 1:21-36. Rothschild, B. J. Paloheimo, J. E., AND L. M. Dickie. 1967. Competition for gear in a multiple-species fishery. J. 1964. Abundance and fishing success. Rapp. P. -v. R^un. Cons. 31:102-110. 155:152-163. Searle, S. R. QuiNN. T. J. 1971. Linear models. John Wiley. N.Y., 532 p. 1980. Sampling for the abundance of schooling populations Snedecor, G. W., and W. C. Cochran. with line-transect, mark-recapture and catch-effort methods. 1967. Statistical methods. 6th ed. Iowa State Univ. Press, Ph.D. Thesis, Univ. Washington, Seattle, 373 p. Ames, lA, 593 p. 144 SEASONAL SPAWNING CYCLE, SPAWNING FREQUENCY, AND BATCH FECUNDITY OF THE CABEZON, SCORPAENJCHTHYS MARMORATUS, IN PUGET SOUND, WASHINGTON' Robert R. Lauth^ ABSTRACT The seasonal spawning cycle, spawning frequency, and batch fecundity -of the cabezon, Scorpaenichthys maTmoratus, were studied in Puget Sound, Washington, USA between September 1984 and October 1985 using scuba techniques. Seasonal embryo mass abundance and ovarian condition indicated that the spawning season started in November and continued 10 months through the following September while peak spawning activity occurred during March and April. Three factors revealed in this study indicated that females may spawn more than once during a single spawning season: 1) the presence of an inter- mediate mode of yolked oocytes, 2) a low wet gonadosomatic weight index, and 3) a protracted spawn- ing season. Batch fecundities predicted from regressions on weight and length ranged between 66,000 and 152,000 eggs for females from 2.5 kg to 10.5 kg and between 57,000 and 137,000 eggs for females from 500 mm to 775 mm. Out of approximately 300 cottid species worldwide (Nelson 1984), the cabezon, Scorpaenichthys mar- moratus, is perhaps the largest (Jordan and Ever- man 1898) and can attain a length of 990 mm and a weight of 11.4 kg (Feder et al, 1974). Cabezon range from Pt. Abrejos, Baja California (Miller and Lea 1972) to Samsing Cove near Sitka, AK (Quast 1968). Their depth range in California is from near- shore tidepools to 76 m (Feder et al. 1974). Cabezon are demersal and solitary and are usually associated with reefs, boulders, or beds of kelp, algae, or eelgrass. A small recreational fishery exists for cabezon. For divers who spearfish, cabezon are prime targets because of their trophy size, desirable food qualities, and general vulnerability in shallow nearshore habitats. Knowledgeable anglers also enjoy catching and eating cabezon even though they are not generally targeted (Olander 1984). Although cabezon are not targeted by a commer- cial fishery at present, they are incidental in com- mercial catches and they do occasionally appear in fish markets along the west coast (Ayres 1854; O'Connell 1953; personal observations in fish mar- kets in Seattle, WA). There is little published information on cabezon 'Contribution Number 786, School of Fisheries, University of Washington, Seattle, WA 98195. ^School of Fisheries, WH-10. University of Washington, Seattle, WA 98195; present address: Inter-American Tropical Tuna Com- mission, cyo Scripps Institution of Oceanography, La JoUa, CA 92093. reproductive biology aside from a life history study in Monterey, CA done by O'Connell (1953), studies of cabezon roe toxicity (Fuhrman et al. 1969, 1970; Hashimoto et al. 1976; Hubbs and Wick 1951; Pills- bury 1957), and diving observations of cabezon nest- ing behavior in a California kelp forest (Feder et al. 1974). Spawning season, spawning frequency, and batch fecundity of cabezon north of California have to date, not been studied. Thus, it seems prudent that we learn about the reproductive biology of cabezon in other areas, especially because of their value as a fishery resource. The objective of this study was to examine the spawning ecology of cabe- zon in Puget Sound, WA and to make a geograph- ical comparison with data for cabezon in California. METHODS AND MATERIALS Study Sites Sampling consisted of transect and collection dives. Sampling began in September 1984 and ended in October 1985 and was done using scuba tech- niques. Edmonds Underwater Park and the Ed- monds Marina breakwater, both located in Ed- monds, WA, USA Oat. 47°48'N, long. 122°22'W; Fig. 1), were chosen as study sites because they had been previously identified by the author as spawn- ing areas for cabezon. Two transects, each cover- ing 250 m', were established along a scuttled dry dock at Edmonds Underwater Park. Transect 1 was the remains of the northern bulkhead of the drydock Manuscript accepted September 1988. Fishery Bulletin, U.S. 87:145-154. 145 FISHERY BULLETIN: VOL. 87, NO. 1 Bellingham Everett Edmonds Seattle PACIFIC OCEAN Olympia 48 °N Figure 1.— Map of Puget Sound, Washington, USA showing the general location of scuba sampling sites (X). and was 65.5 m long and 3.8 m wide. The southern half of the drydock, designated Transect 2, was 30 m south of Transect 1 and was 55 m long by 4.5 m wide. The eastern and western ends of both Tran- sects 1 and 2 were in 6.0 m and 9.0 m of water (MLLW, i.e., Mean Lower Low Water), respec- tively. Transect 3 was a portion of the Edmonds Marina breakwater parallel to a Washington Department of Fisheries' fishing pier. The transect was 150 m in length by 5 m in width and covered a total area of 750 m^. The breakwater consisted of large basalt boulders that extended from 3 to 5 m (MLLW) below the surface of the water. In addition to the break- water, the transect included a sandy area with inter- spersed boulders to a depth of 7 m (MLLW). Transect and Collection Dive Sites and Procedures Each transect was sampled at least once a month. Dives were made more frequently when spawning activity increased. Fifty transect dives totalling 46 hours of bottom time were made. Physical data col- lected on each dive included water temperature and depth. Biological data gathered included number of cabezon and number and depth of embryo masses. Dives in spawning areas were made during all hours of daylight. Collection of specimens for biological data was by pneumatic speargun. Forty-eight col- lection dives totalling 36 hours of bottom time were made. Fifty female cabezon were collected through- out Puget Sound, including areas in the Strait of 146 LAUTH: CABEZON IN PUGET SOUND, WA Juan de Fuca and San Juan Islands (Fig. 1). All specimens captured were weighed to the nearest 0.1 kg and total length measured to the nearest Processing of Ovaries From 1 to 10 females were sampled each month so that the progression of ovarian development could be followed throughout the study period and spawning frequency could be determined. Entire ovaries were excised, weighed to the nearest 0.1 g, put in gauze bags, and placed in modified Gilson's solution to harden the eggs and to break down ovarian tissue (Simpson 1951). After 5 to 6 months in Gilson's solution, the eggs from each ovary were separated from the ovarian tissue using a mild jet of water while passing them through a series of Tyler^ brass sieves with open- ings of 1.651 mm, 0.295 mm, 0.180 mm, and 0.075 mm. Most eggs with diameters less than 0.075 mm passed through the smallest screen and were dis- carded. Loose eggs were retained by the sieves and stored in jars with 5% formalin. Ova Diameter Frequencies Eggs and water (2.5 L) were homogeneously mixed in a 4 L beaker with magnetic stirrer. A ran- dom 5 mL subsample was drawn with a pipette and the eggs were measured with a calibrated ocular micrometer. At least 200 eggs were measured from each ovary. Ova diameters were grouped using 0.05 mm incre- ments as midpoints. Based on ova diameter frequen- cy histograms, ovaries were grouped into eight stages (I-VIII). Ranges, means, medians, and stan- dard deviations were calculated for the apparent modes within each stage. I calculated a wet gonadosomatic weight index (WGSI) for each female using the formula of Gunderson and Dygert (1988). The WGSFs for each stage of ovarian development were averaged and used as a measure of relative gonadal investment of females. The Number of Eggs to be Spawned The subsampling procedure for estimating the number of eggs to be spawned was identical to the one for measuring ova diameters. Subsamples were ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. enumerated using a dissecting microscope, a gridded petri dish, and a laboratory counter. When modes in an ovary overlapped, the number of eggs from the largest mode were counted twice and averaged. At least three subsamples were taken for each ovary and the mean total number of eggs to be spawned was calculated, using a simple volumetric propor- tion. If the coefficient of variation was greater than 10%, additional subsamples were made until it dropped below 10%. Unweighted least-squares linear regression was used to predict the total number of eggs to be spawned using lengths and weights of females as independent variables. All regression analyses were done with a personal computer according to methods described by Kleinbaum and Kupper (1978). RESULTS Seasonal Embryo Mass Abundance The beginning and end of the spawning season were defined as the dates when embryo masses were first and last seen. During 19 collection and tran- sect dives throughout Puget Sound from mid- September until the end of November, no cabezon embryo masses were observed. The first embryo mass observed in Puget Sound was on 6 December 1984. A sample of eggs from this embryo mass was placed in a 5 gal bucket filled with seawater and most hatched within 30 minutes. The eggs were ap- parently near the end of their incubation period since little of the yolk sac was remaining. Based on work from this study, incubation time until hatch- ing is several weeks, thus the embryo mass was most likely deposited sometime in middle or late Novem- ber 1984. Along Transects 1, 2, and 3, embryo masses were first observed on 2 January 1985 and 7 and 21 December 1984, respectively. After the first embryo masses appeared, there was a steady increase in abundance with some fluctuation (Fig. 2). The peak number of embryo masses at all three transects oc- curred during March and April 1985, after which there was a general decline. By 30 August 1985, em- bryo masses were totally absent from Transects 1 and 2 and by mid-September 1985, none were found at Transect 3 or elsewhere in Puget Sound. Between November 1984 and September 1985, there were 35 embryo masses observed on Transect 2, 23 embryo masses on Transect 1, and 15 embryo masses on Transect 3. It is possible that some em- bryo masses may have hatched or disappeared (e.g., predation, cannibalism, dislodged by physical dis- 147 FISHERY BULLETIN: VOL. 87, NO. 1 Transect 1 Transect 2 Transect 3 OCT NOV DEC JAN FEB MAR APRIL MAY JUNE JULY AUG SEPT MONTH Figure 2.— Temporal fluctuation in the abundance of cabezon embryo masses at Transects 1, 2, and 3. turbance) between dives; thus the totals may have been underestimated. It was assumed that incubation time of eggs till hatching was the time between when an embryo mass was first observed and when hatching was first noted. For 13 nests that were monitored from 21 January to 26 May 1985, incubation time ranged between 25 and 49 days and averaged 34 days with a standard deviation of 6.8 days. Water temperature varied between 8° and 10°C during this period. Embryo masses were found in the intertidal to depths of 17 m and were deposited on hard sub- strates including wood pilings and logs, rocks, and steel. Embryo masses were always observed on ex- posed surfaces rather than underneath structures or inside crevices. Spawning Frequency Ova diameter frequency plots were used, in part, to determine the frequency of spawning. Eight stages of ovarian development, designated I to VIII, were delineated based on the modal configurations of ova diameter frequency plots (Fig. 3). Seven ovaries were in Stage I, characterized by relatively small resting oogonia with diameters <0.40 mm (Fig. 31). The bulk of the eggs from Stage I were translucent, devoid of yolk, and had diameters <0.20 mm. Eggs of this size and with these characteris- tics were present in all eight stages. Stage II ovaries were found in six cabezon (Fig. 311). In addition to the large reserve of resting oogonia, there was another mode of opaque eggs which averaged 0.46 mm and ranged from 0.35 to 0.65 mm. Stages III to VII represented two basic types of female spawners: those which were going to spawn for the first time (Stages III to V) and those which had already spawned once and had the potential for spawning again (Stages VI and VII). For both groups (spawned and unspawned), there were two groups of yolked oocytes. There were seven female cabezon with F ^e III ovaries (Fig. 3III). Besides the resting oogonia, there was an intermediate mode (average 0.47 mm, range 0.35 to 0.65 mm) which represented a reserve group of immature oocytes for future spawoiing, and a larger mode (average 0.84 mm, range 0.70 to 1.10 mm) which consisted of maturing ova destined to be spawned within the current spawning season. In Stage IV, egg hydration was beginning and the largest mode was more distinct than in Stage III ovaries (Fig. 3IV). The largest mode of yolked oocytes averaged 1.23 mm and ova diameters ranged from 1.00 to 1.45 mm in Stage IV ovaries. For females with Stage V ovaries, spawning was imminent and there was no evidence of prior spawm- ing (Fig. 3V). The modal configuration of an ova diameter frequency plot of a female captured while actually spawning was Stage V. For all Stage V ovaries combined, the average diameter of the largest mode (hydrated eggs) was 1.48 mm and the range was from 1.35 to 1.65 mm. Eggs from the in- termediate mode had an average diameter of 0.55 mm and ranged from 0.35 to 1.0 mm. Stage VI ovaries were characteristic of recently 148 LAUTH: CABEZON IN PUGET SOUND. WA spawned females (Fig. 3VI). Three females with Stage VI ovaries were collected in the immediate vicinity of freshly deposited embryo masses, pre- sumably after spawning them. Within these ovaries, an incipient mode, which ranged in size from 0.70 to 0.95 mm and had a mean size of 0.79 mm, was apparent. A relatively small number of large diam- eter eggs (~1.5 mm) were scattered within all Stage VI, Stage VII, and Stage VIII ovaries (Fig. 3VI- VIII) and were presumably remnants of the recent spawning event. In Stage V females the largest mode would mask evidence of residual eggs so it was not possible to ascertain whether they had already spawned in the 1984-85 spawning season. As the ovaries progressed to Stage VII, the eggs of the incipient mode were larger and distinct from the intermediate mode. These yolked eggs appeared to be hydrating in preparation for another spawn- ing. Females exhibiting this condition had spawned previously and since eggs of an intermediate size were still present, these females were capable of spawning again. The incipient mode for Stage VII ovaries ranged from 0.95 to 1.50 mm with a mean of 1.16 mm. The intermediate mode of Stage VII ovaries ranged from 0.35 to 0.90 mm and averaged 0.51 mm. Stage VIII ovaries were similar to Stage I ovaries. There was a single and small mode of eggs which were deteriorating noticeably. Irregularly shaped ova were translucent or transparent and devoid of yolk. Unlike Stage I, Stage VIII ovaries had rem- nant eggs (~1.5 mm in diameter) from at least one previous spawning event (Fig. 3VIII). When the eight stages were plotted against the date when females were captured, a general pro- gression of ovarian development was seen (Fig. 4). Stage III to V ovaries were only seen in females caught between December and May. Stages VI and VII were found from February through August. Stage VIII females were caught both before and after Stage III through VII females. The early Stage VIII's were probably carry-overs from the previous spawning season. Females in Stages I and II were found prior to all other stages. The WGSI values for the eight stages of ovarian development were in agreement with what might be expected in a multiple spawner (Fig. 5). For ovaries in the resting condition (Stage I), the WGSI was at its lowest point. The WGSI gradually in- creased to a maximum in Stage V when eggs were hydrated and females were in spawning condition. After the eggs were released there was an obvious drop in the weight of the ovaries relative to the body weight. The WGSI slightly increased in Stage VII and then fell in Stage VIII. Without the aid of histological techniques, it was not possible to dis- tinguish an intermediate stage between VII and VIII; this stage would have been virtually identical to Stage V, and had there been such a stage, it is conceivable the WGSI would have reached another maximum before finally declining in Stage VIII. Relation Between Batch Fecundity and Weight and Total Length From ova diameter frequency plots, two basic types of female spawners were evident: 1) those which were going to spawn for the first time (un- spawned; Stages III to V), and 2) those which had already spawned at least once and had potential for spawning another batch (spawned; Stages VI and VII). The number of eggs to be spawned during each spawning event (batch fecundity) was deter- mined by estimating the number of eggs in the largest mode for females possessing ovaries in stages III to VII. Data for spawned and unspawned females was pooled for regression analysis because the range of values for comparable fish weights and lengths was similar, and because separate regressions for spawn- ed and unspawned females were not statistically dif- ferent (P > 0.05). Furthermore, pooling spawned and unspawned data provided analysis over a broader size range of fish and considerably increased the sample size. The resulting regressions of batch fecundity on length and weight were significant at P < 0.001, and the correlation coefficients were 0.69 and 0.73, respectively (Fig. 6). The regression of batch fecundity on length predicted that females from 500 mm to 775 mm would release between 57,000 and 137,000 eggs during a spawning event, and the regression of batch fecundity on weight predicted that females from 2.5 kg to 10.5 kg would release between 66,000 and 152,000 eggs during each spawning event. DISCUSSION Along the western U.S. coast, the length of the spawning season for marine cottids varies from 1 month to year-round (Atkinson 1939; Jones 1962; Marliave 1975; Tasto 1975; DeMartini 1978; DeMar- tini and Patten 1979; Goldberg 1980; Garrison and Miller 1982). Based on temporal embryo mass abundance and ovarian condition, cabezon spawn- ing in Puget Sound commences in late November and lasts 10 months through early September of the following year while peak spawning occurs from 149 FISHERY BULLETIN; VOL. 87, NO. 1 0.5 T 0.4 3 0.2 0,1 0.0 ova diameters < 0.40 mm I n=7 ^M^^-, I I I I I I I I I 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 > o z: LU Z) O LU DC Li. LU > I- 3 LU DC 0.5 4 3 2 1 0.0 II n=6 0.46 ± 0.08 mm h-l I I I I I I I I I I t I I > I I t t I I t I I I I I I 0.2 T 0.1 0.0 47 ± 0.09 mm 0.84 ±0.08 mm lullUliiiulu III n=7 ii I I I I I I t I I I t I I t t I I I I I t 0.2 0.1 0.0 54 i 0.14 mn IV n=7 luluM 1.23 ±0.12 mm Ij m , , I ,.,i, I ,I, M ,I,I^^ , , , , , 0.2 0.4 0.6 08 1.0 12 1.4 1.6 1.8 2.0 OVA DIAMETER (mm) Figure 3.— Eight stages of ovarian development (Stages I to VIII) based on the modal configuration of ova diameter frequency plots. 150 LAUTH: CABEZON IN PUGET SOUND. WA u.z V n=5 0.1 0.55 ± 0.14 mm n n 1. llllllllllll. 1.48 ± 0.08 mm , , , , lllllp^, , , . t * t 1 o LL! Z> o LU 0.4 3 0,2 0,0 VI 0.46 i 0.09 mm 0.79 ±0.06 mm Remnant ova from previous spawning Ii^IipL^^pIm 'i-i-t l^t\ti^^^^^t^ii^**t *■■■< LU > ! I 3 LU 0.3 2 1 0.0 VII n=6 51 ± 0.13 mm Remnant ova from prgvious spawning lilul 1.16 ±.0 13 mm ' PP ^ P^^i " ! I iWi*iM(Mi*i"t-i"^i I I I I ( I t t I I I I 0.4 VIII 0.3 n=8 0.2 1 Remnant ova from previous spawning ; llll«*l. 02 04 0.6 0.3 1.0 1.2 1.4 1.6 1,8 2.0 OVA DIAMETER (mm) Figure 3.— Continued.— Moda\ average ± 1 standard deviation are listed above each mode. 151 FISHERY BULLETIN: VOL. 87, NO. 1 Figure 4.— Stage of ovarian development versus time of capture. Vlli • ♦* t • VII t • • *♦ VI- s t * * T V ♦• : > :•: » • E III • ••• 11 - • • • • • 1 * H -H 1 1- • 1 h • \- 1 1 h- 1 1- 1 1 SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT MONTH Spawned Unspawned Figure 5.— Average wet gonad-somatic weight index (WGSI) X 100 versus stage of ovarian development. Ver- tical bars represent the 95% confidence limits. March to April. The spawning season reported for S. marmoratus in California is half as long (Novem- ber to March) and peak spawning occurs 3 to 4 months earlier (O'Connell 1953) than found in this study. This is contrary to what one might expect based on general patterns (Qasim 1956). Teleosts in high latitudes generally spawn once and have relatively short spawning seasons during the winter and early spring. On the other hand, most fishes at lower latitudes have protracted spawning seasons and spawn more than once. Seasonal fluctuations in production cycles (food supply) are less defined in lower latitudes, hence females are able to feed more or less continuously to build sufficient energy reserves for a longer spawning season consisting of multiple batches. Spawning time and duration usual- ly synchronize with production cycles so that larvae have a better chance for survival (Nikolsky 1963; Gushing 1982). ■o c >- Q Z) o X o < CD #eggs . 29TL - 87.54 r = 0.69 P<0.001 450 500 550 600 650 700 750 800 TOTAL LENGTH (mm) » eggs - 10-8W + 39.25 r = 0.73 P<0,001 4 5 6 7 8 9 BODY WEIGHT (KG) 10 11 Figure 6.— Plots and regressions of batch fecundity (thousands of eggs) versus total length (mm) and weight (kg) for female cabezon from Puget Sound. Female cabezon are probably similar to other species of marine sculpins along the west coast which spawn multiple batches of eggs during a single spawning season (Atkinson 1939; DeMartini 1978; Goldberg 1980). O'Connell (1953) suspected 152 LAUTH: CABEZON IN PUGET SOUND, WA that cabezon in California spawned more than once. Evidence from this study also strongly suggests that sexually mature female cabezon spawn more than once during a single spawning season. Species with protracted spawning seasons characteristically spawn more than once per season compared with species with relatively short spawning seasons which spawn only once (Qasim 1956). Another strong indication of multiple spawning is the presence of two distinct modes of yolked oocytes in ovaries. A second mode was present both in females about to spawn, and in females which had spawned at least once. An intermediate mode con- sisting of vitellogenic oocytes suggests that females are capable of spawning more than once in a single season (Goldberg 1981). It does not appear that the intermediate generation of yolked eggs are retained for the following season, because cabezon ovaries undergo resorption in the fall (Stage VIII) prior to beginning another cycle the following season (Stage I; Fig. 31). Interestingly, after March 1985, all females cap- tured had spawned at least once. The ovaries of all females captured between March and September were either in the process of bringing another batch of eggs to maturity, or in the process of resorption. Multiple spawning is a possible explanation for the absence of females with ovaries in the unspawned condition during the March to September period. Multiple spawning is also a logical explanation for the relatively low WGSI value for cabezon ovaries with yolked and unhydrated eggs (Stage III). Gunderson and Dygert (1988) showed a relation between "reproductive effort" (WGSI) and natural mortality (M) in numerous species of marine fish (r^ = 0.81). The higher the natural mortality (M), the shorter the longevity (<(, fn ) of the species and thus the greater the "reproductive effort" invested in any given year. The two extremes cited were the northern anchovy (M = 0.92, io.oi = 6 years, WGSI = 0.65) and dogfish (M = 0.09, 0.5 High > 0.3 Moderate EX?i > 0.1 Low I I < 0.1 Very Low w =) o cc o CO UJ o LU a. CO € I ^^^H ^H ^^^H A B c ■•■•■•■•■•■•■•■' ■■■■■■• WipWWII^BW W^ D iVi\'.'.'.*.|.*. E ;.; •:•:•:•:• •:•:•:•:•:•:■:■: \*X*> >>>>><<>>> ^yyy^y^y^•^ ■:■:■:•:•:•:■:■:■:•:•:■:■:■:•:•: v F Vf'-'tViVi *','','','',' ':•:•:•:■:■:■:•:•:•:• •:•:•:•:•:•:■:■:■:■ :■:•:•:•; '.'.•.•.•.•.•.•.•.*.• i'l'i'iVi'i" >:•:■:■: '■'■'■"•* ■:■:•:•:•:•:•:•:•:■ ■:•:•:•:■:•:•:•:•:•:•:•:•:•:•:•: •:■ ■:■:■:•:■ 1 2 3 0.0 - OC -0.4 — < w -0.8 ■1.2 —I SITE GROUPS Figure 3.— Inverse and normal classification hierarchies and nodal diagram showing constancy and fidelity of site— species group May to August but are known to occur from Feb- ruary to November in other portions of its range (Dudley and Judy 1961). In the present study, 185 of the 187 ovigerous females collected were taken in spring. Male P. gihhesii were slightly larger (x CW = 45.2 mm) than females (x CW = 43.2 mm) but were less numerous. Analysis of sex ratios indicated significantly more female crabs than males were col- lected each season (Table 6). Squilla etnpusa This stomatopod is widely distributed in the west- ern Atlantic, occurring from Maine to South Ameri- ca as far south as Surinam (Manning 1969; Gore and Becker 1976). Camp (1973) found S. empiisa to be most abundant at 18 m depths on the central west Florida shelf. This species was the most abundant stomatopod collected and ranked second among the total catch of decapod and stomatopod species. It was frequent- ly collected throughout the study area, occurring in 78% of the 803 trawl tows made. In terms of bio- mass, S. em-pusa constituted 11% of the total catch, being outranked only by the blue crab, Callinectes sa'pidMS (Table 1). The stratified mean catch per tow was highest in spring (40 individuals/tow) and sum- mer (42 individuals/tow) (Table 5). The number of individuals per tow differed between the areas with more S. empusa collected off Florida during every season (Fig. 4). 164 WENNER AND WENNER: CRUSTACEANS FROM COASTAL HABITATS FIDELITY SIMILARITY -0.8 -0.4 0-0 I I I I I I CO CL o o CO UJ o 111 Q. CO ■c 0.0 - tr -0.4 -\ < CO -0.8 — -1.2 —I > 4 Very High > 3 High > 2 Moderate Esa 2 ' Low I I < 1 Very Low — :•:•:•:•: ,'.','.'.'.'.'.•. .;.;.;.;.;.;.;.;.;.;.;.;.;.;.;. yy.'i ■■■... .^TT- ;i;i;i;i;i,'i,-iyJ;i';i\\';»';^« J^x-x-xxJivilixivX:::::::::::::: ^H :j:i:;:i:|:|:|:i:i:i:;:;:;:i:;:j i:-:!^:-:-: ;>: ■X'Xv:-:-:-:-:-:-:-:-:-;-: 1 2 3 SITE GROUPS Figure 3.— Confma^rf— coincidence from pooling collections dur- ing a season within each stratum. Ovalipes stephensoni This portunid crab, which was the third most abundant species collected in this study (12% of total catch), occurs off Virginia (Musick and McEachran 1972) to near Biscayne Bay, FL (Park 1969). Adult 0. stephensoni are found farther from shore than adults of its congener 0. ocellatus; however, young of both species occur nearshore (Williams 1984). Wenner and Read (1982) found 0. stephensoni to reach maximum abundance from 9 to 18 m between Cape Fear, NC and Cape Canaveral, FL. In the study area, 0. stephensoni was most numerous in strata off South Carolina and North Carolina where 43 individuals/tow were collected. Catches decreased off Georgia to 29 individuals/tow and were lowest in strata off Florida (<1 individual/tow). Stratified mean catch per tow was highest in spring (120 in- dividuals/tow) and declined markedly during other seasons (Table 5, Fig. 4). Analysis of size composition indicated mean cara- pace width differed between strata and season. Largest crabs were found off Florida (x CW = 45 mm, n = 63), while average sizes off Georgia (n = 713) and South CarolinaTNorth Carolina {n = 2,349) were 30 mm and 34 mm, respectively. Ovalipes stephensoni collected in fall were larger (x CW = 42 mm, n = 273) than those collected during other seasons (winter: x CW = 38 mm, n = 199; spring: z CW = 30 mm, n = 1,591; summer: S CW = 36 mm, n = 1,062). Analysis of sex ratios by season indicated signif- 165 FISHERY BULLETIN: VOL. 87, NO. 1 icantly more female than male 0. stephensoni were collected (Table 6). Male crabs (S CW = 34 mm, n = 1,278) were comparable in size to females (x CW = 33 mm, n = 1,840); however, most of the crabs collected were immature, being smaller than sizes (short carapace width) at full sexual maturity of 61 mm for males and 51 mm for females given by Haefner (1985). This supports previous observa- tions (Williams 1984) that there is a positive size- depth relationship for the species. Callinectes similis The lesser blue crab occurs in the oceanic littoral zone where it is commonly associated with the blue crab. Along the east coast of the United States, C. similis ranges from Delaware Bay to Key West, but is primarily a Carolinean species. Northern occur- rence of the species occurs seasonally during years with favorable annual temperature (Williams 1984). Callinectes similis ranked fourth in terms of num- ber of individuals among all species collected in this study and occurred in 64% of the trawl tows made. The species constituted 10% of the total biomass, which was considerably less than C. sapidus (Table 1). The stratified mean catch per tow for numbers and weight was highest in summer and fall (Table 5). Tagatz (1967), who did not distinguish between Portunus gibbesii Ul Q. CO —I < 9 > Q O CC LU m < LU 120 100 80- 60- 40- 20 180 160H 140 120- 100- 80 60i 40 20-1 5/20 10/15 4/15 ^m '°^^'> I 1^1 I 6/15 " Ovalipes stephensoni 29/38 1/25 6/25 p-| i^^i_HMJ L WINTER SPRING Figure 4.— Seasonal catch rates of the SUMMER FALL dominant decapod and stomatopod species. 166 WENNER AND WENNER: CRUSTACEANS FROM COASTAL HABITATS Table 5. — Stratified mean catch per tow of 15 most numerous species caught during the coastal study. Stratified mean catch per tow Winter Spring Summer No. Weight Fall Species No. Weight No. Weight No. Weight Portunus gibbesii 26 0.240 54 0.473 41 0.156 79 0.588 Squilla empusa 28 0.335 40 0.580 42 0.635 31 0.515 Ovalipes slephensoni 4 0.096 120 0.578 14 0.151 4 0.075 Callinectes similis <1 0.016 4 0.091 34 0836 35 0.571 Penaeus setiferus 15 0.289 2 0.063 12 0.260 30 0.791 Penaeus aziecus <1 _ <1 0.005 44 0.722 2 0.048 Portunus spinimanus 3 0096 13 0.214 14 0.207 11 0.269 Ovalipes ocellatus 1 0.063 17 0.344 12 0.389 4 0.164 Hepalus epheliticus <1 0.007 5 0.138 12 0.357 10 0.352 Squilla neglecia <1 0.002 9 0.120 7 0.065 4 0.059 Callinectes sapidus 2 0.361 6 1.316 8 1.202 1 0.214 Arenaeus cribrarius <1 0.010 2 0.076 9 0.348 5 0.210 Trachypenaeus constrictus 5 0.014 2 0.005 1 0.001 6 0.019 Libinia emarginata 6 0.390 6 0.203 2 0.085 2 0.1 45 Penaeus duorarum <1 0.005 9 0.152 2 0.020 3 0.031 o cc LU Q. CO -I < 9 > Q u. O cc UJ QQ 3 Z Z < LU 100 80- 60- 40- 20- Squilla empusa 1/15 531 ti 11/15 17/24 19/20 13/15 I 26/31 I 21/25 Florida H Georgia 1 I South Carolina/ North Carolina 60- 40- 20- Callinectes sim ills JS3J 25/2 5 ^H 32/38 WINTER SPRING SUMMER Figure i.— Continued. FALL 167 FISHERY BULLETIN; VOL. 87, NO. 1 C. similis and C. omatus, noted increased abun- dance from May to November with the largest catches in May, June, and October. Comparison of mean catch per tow between areas showed that C. similis was most abundant in strata off Florida and North Carolina/South Carolina during summer and in strata off Georgia during fall (Fig. 4). For all seasons, however, the mean catch per tow was high- est for strata off North Carolina/South Carolina where an average of 24 individuals and 0.49 kg were taken per tow. Size composition of C. similis differed between seasons with average size smallest in fall (x CW = 61 mm, n = 1,685). Average sizes during other sea- sons were spring (68 mm CW, w = 167); summer (67 mm CW, n = 2,558); and winter (64 mm CW, n = 36). The average size of individuals collected from strata off Florida (iCW = 72 mm, w = 1,025) was larger than those from Georgia (x CW = 59 mm, n = 1,111) and South Carolina/North Carolina (5 CW = 65 mm, n = 2,310). Sex ratios of C. similis indicated a significant 60 40H 20 Penaeus setiferus 10/15 ^1 rn '/I5'^ ,/24 J I II ^ I I 24/25 31/36 ^1 31/38 125/45^1 JIL. cc LU Q. CO _i < 9 > Q 40 20H Portunus spinimanus 13/15 1^1 40- 20- Hepatus epheliticus 5/15 3/i5 4/24 I I ^^ I I _^_ 17/24 pX^l 8/20 36/5 TT^ 29/36 i 20/25 l^^5 CC LU m 40-1 20- Callinectes sapidus 10/15 14/24 '^,6/24 '^ LU CO < LU 60-1 40- 20 40- 20- 40 20-1 40 20-1 Penaeus aztecus 1/15 0/15 0/15 ^/'5 '/I5 7/24 Ovalipes ocellatus 21/24 1/15 ' 7/24 1/15 40-1 Squilla neglecta 20- ^/I5 „,,, 5/24 f/15 1 Arenaeus cribrarius 1/15 4/15 3/24 Libinia emarginata Ml 9/15 I WINTER SPRING 6/20 1 29/5€ 1 n F^ °B f^' 13/20 18/36 17/51 I I ^^ I I SUMMER 4/i= 7 2= ?2/?» 27/38 15/25 13/25 23/38 I I ^^ I I FALL Florida | Georgia Q South Carolina/ North Carolina Figure 4.— Continued. 169 FISHERY BULLETIN: VOL, 87, NO. 1 Table 6— Frequency of males and females for spec ies by season. ' reflects sign lificant deviation of M:F from 1:1 (P < 0.05) as determined by i^ analysis. Species Winter Spring Summer Fall Portunus gibbesii • • • • Male 239 563 890 1,144 Female 369 739 1,019 1,908 Total 608 1,302 1,909 3,052 Ovalipes Stephenson! • * • Male 84 694 398 102 Female 114 895 661 171 Total 198 1,589 1,059 273 Callinectes similis Male 18 77 1,166 874 Female 18 90 1,378 814 Total 36 167 2,544 1,688 Portunus spinimanus • Male 74 254 450 395 Female 69 335 443 383 Total 143 589 893 778 Ovalipes ocellatus Male 28 289 457 149 Female 30 269 502 208 Total 58 558 959 357 Hepatus ephelilicus Male 14 100 320 231 Female 16 206 718 541 Total 30 306 1,038 772 Callinectes sapidus • • • • Male 7 8 23 12 Female 145 306 863 116 Ovigerous 214 405 2 Total 152 314 886 128 Arenaeus cribrahus • Male 9 61 446 205 Female 4 36 348 157 Total 13 97 794 362 Libinia emarginata * Male 161 158 92 107 Female 120 153 86 89 Total 281 311 178 196 (South Atlantic Fishery Management Council 1981). Penaeus setiferus was the most abundant penaeid collected in this survey, and it constituted 9% of the total catch of decapod crustaceans (Table 1). This species accounted for 10% of the total biomass of stomatopods and decapods and occurred in 63% of the 303 collections made. The stratified mean catch per tow differed among seasons, with abundance greatest in fall (30 individuals/tow) and lowest in spring (2 individuals/tow) (Table 5). This seasonal difference in abundance of P. setiferus in the near- shore coastal zone is explained by movement of white shrimp from estuaries to offshore waters in fall. This emigration is associated with declining water temperatures (Lindner and Anderson 1956; Pullen and Trent 1970). White shrimp enter the estuaries as postlarvae in May, grow rapidly in the estuarine nursery grounds, and move seaward through late summer and fall (Weymouth et al. 1933). Within the three areas sampled, white shrimp were most abundant in strata off Georgia during every season except spring (Fig. 4). This may result from a predominantly southward movement of white shrimp during fall, winter, and summer as discussed by Shipman (1980). The mean total length of white shrimp differed by season with largest individuals occurring in spring (x = 162 mm, n = 93). The larger average size at this time was probably influenced by occur- rence of female roe shrimp that move to nearshore coastal waters from estuaries during the spring (Lindner and Anderson 1956; Joyce and Eldred 1966; Harris 1974; Music 1979; Farmer et al. 1978). The mean size of shrimp collected was largest in strata off North Carolina/South Carolina (x = 152 mm, n = 1,439), while those from strata off Georgia and Florida averaged 145 mm (w = 2,269) and 142 mm (n = 1,121), respectively. Penaeus aztecus Brown shrimp occur from Martha's Vineyard, MA to the Florida Keys and into the Gulf of Mexico where they occur on the Sanibel grounds, in Appa- lachicola Bay, and to northwestern Yucatan (Wil- liams 1984). Along the Atlantic coast of the United States, P. aztecus is most abundant off North and South Carolina (Cook and Lindner 1970). Brown shrimp were collected in 51% of the trawl tows made during the study and were the second most abundant Penaeus collected (Table 1). The stratified mean catch per tow was much greater in summer (44 individuals/tow) than during the other seasons when <2 individuals/tow were collected (Table 5). Brown shrimp usually occupy the estua- rine nursery grounds from March through July before emigrating to coastal waters in August. Emi- gration, however, may be delayed if cooler than nor- mal temperatures occur in spring (South Atlantic Fishery Management Council 1981). The summer cruise occurred from mid-July into September, which overlapped the period when brown shrimp emigrated from the estuary. During the summer sampling period, brown shrimp abundance was greatest in strata sampled off North Carolina/ South Carolina (Fig. 4). Abundance during other seasons was too low to assess any difference be- tween areas. 170 WENNER AND WENNER: CRUSTACEANS FROM COASTAL HABITATS The mean total length of brown shrimp was great- est in summer (x = 126 mm, n = 2,718) and fall (x = 126 mm, n = 214) following emigration from the estuaries. Size at emigration for brown shrimp has been reported to be 100-105 mm (Joyce 1965) and 60-103 mm (Trent 1967). Off North Carolina, P. aztecus enters the commercial fishery in June at a size of 100 mm (South Atlantic Fishery Manage- ment Council 1981). Mean size of brown shrimp was largest off Georgia (x = 138 mm, n = 871) and Florida (i = 127 mm, n = 493), while those from strata off North Carolina and South Carolina aver- aged 119 mm (w = 1,600). The capture of larger shrimp further south may result from migration of individuals to waters off Georgia and Florida. These shrimp are probably supplied by the sounds and estuaries of North and South Carolina. Tagging studies off North Carolina indicate that brown shrimp move to the south and to deeper, nontrawl- able waters once they leave the sounds (Purvis and McCoy 1974). Shipman (1980) noted few recaptures of brown shrimp tagged off Georgia and suggested that lower return rates may indicate offshore move- ment of brown shrimp, out of the nearshore trawl- ing grounds. Portunus spinimanus This common portunid of the inner continental shelf ranges from New Jersey to southern Florida (Powers 1977), often co-occurring with P. gibbesii. Camp et al. (1977) found P. spinimanus to be one of the commonest decapod crustaceans in samples from nearshore east Florida waters with salinity of 32-39"/oo and temperature ranging from 19.2° to 32°C. In the present study, P. spinimanus was the 7th most numerous species (4.1% of the total catch) and the 11th most important by weight (4%) (Table 1). This species occurred at 73% of the trawl stations from all seasons. The stratified mean catch per tow was lowest in winter (3 individuals/tow). Catch per tow values by stratum were higher in spring (13 in- dividuals/tow) and summer (14 individuals/tow), but decreased to 11 individuals/tow in fall (Table 5). Catch of P. spinimamis per tow differed among areas with most individuals collected in trawl tows off Florida (15 individuals/tow). Increased abun- dance in strata off Florida was observed for every season except summer (Fig. 4). The size composition of P. spinimanus in trawl catches differed between strata and seasons. Mean carapace width was largest for individuals collected off Florida (x CW = 54 mm, n = 974) and Georgia (x CW = 53 mm, n = 343), while those collected in combined strata off South Carolina and North Carolina averaged 46 mm CW {n = 1,129). Largest individuals were collected in winter (i CW = 58 mm, n = 144) and fall (i CW = 56 mm, n = 782). The smaller average size of in- dividuals collected in spring (x CW = 48 mm, n = 601) and summer x CW = 46 mm, n = 919) may reflect an influx of small crabs from the previous fall hatch. Analysis of sex ratio by season indicated no sig- nificant deviation from unity except during spring when female P. spinimanus outnumbered males (Table 6). Average size was similar for males (i CW = 51.4 mm, n = 1,172) and females (x CW = 49.8 mm, n = 1,230). Ovalipes ocellatus This portunid crab has a broad geographic range from Canada to Georgia (Williams 1984). Abun- dance decreases in southern latitudes, apparently in response to lessened tolerance to warm-water temperatures (Vernberg and Vernberg 1970). In the present study, 0. ocellatus occurred more frequently than 0. stephensoni, but was not as numerous (Table 1). Abundance of this crab de- creased from the northern to southern area, with most individuals collected in trawl catches off North Carolina/South Carolina (14 individuals/tow) (Fig. 4). Abundance in strata off Georgia was 6 in- dividuals/tow, while <1 individual/tow was caught off Florida. The stratified mean catch per tow dif- fered among season with most individuals collected in spring (Table 5). Average size of 0. ocellatus differed among areas. The average size of individuals collected in strata off Florida was larger (x CW = 65 mm, n =11) than that from other areas (Georgia: x CW = 53 mm, n = 588; South Carolina/North Carolina: x CW = 51 mm, n = 1,340). Seasonal differences in size composition were noted as well, with average carapace width smallest in spring (x C W = 48 mm, n = 558). This may reflect occurrence of juveniles from a fall-winter hatch (Dudley and Judy 1971). Average size of individuals during other seasons was winter (x CW = 59 mm, n = 58), summer (x CW = 52 mm, n = 966), and fall (i CW = 57 mm, n = 357). No significant seasonal difference in sex ratio was noted, with the exception of fall when females were more numerous than males (Table 6). Male 0. ocellatus (x CW = 54 mm, n = 923) were larger than females (i CW = 50 mm, n = 1,009). 171 FISHERY BULLETIN: VOL. 87, NO- 1 Hepatus epheliticus The known range for this crab extends from Chesapeake Bay to southern Florida where it is a common inhabitant of nearshore waters. Evidence suggests it buries in sandy substrate (Wilhams 1984) and may be nocturnally active (Powers 1977). Hepatus epheliticus occurred throughout the study area and was present in 65% of the collections made during all seasons (Table 1). Abundance differed among seasons with the stratified mean catch per tow being highest in fall (10 individuals/tow) and summer (12 individuals/tow) (Table 5). Number of individuals per tow also differed between areas with highest catches noted from strata off South Caro- lina and North Carolina (10 individuals/tow) (Fig. 4). Larger crabs were noted in this region with a mean carapace width (x CW) of 58 mm {n = 1,176). The mean size oiH. epheliticus from Georgia coastal waters was 58 mm (n = 526), while those from strata off Florida averaged 54 mm (n = 456). There was a noticeable decrease in size and number of crabs collected in winter (x CW = 38 mm, n = 30) compared with sizes noted for other seasons (spring: i CW = 52 mm, n = 306; summer: x CW = 57 mm, n = 1,050; fall: x CW = 60 mm, n = 772). This may reflect movement of larger crabs further offshore during the winter. Female H. epheliticus significantly outnumbered male crabs during every season except winter (Table 6). Carapace width was similar among the sexes (male x CW = 58 mm, n = 663) (female x CW = = 57 mm, n = 1,479). Squilla neglecta This stomatopod species has a more disjunct dis- tribution than its congener, S. empusa, and occurs from North Carolina to Florida, the Gulf of Mexico from western Florida to Texas, and southwest to Brazil (Gore and Becker 1976). Squilla neglecta was found by Camp (1973) to co-occur with S. empusa on the central west Florida Shelf where both were most abundant at 18 m depths. Squilla neglecta occurred in 49% of the trawl tows and was most abundant in spring (9 individuals/tow) (Table 5). The number of individuals per tow was highest in strata off Georgia during every season except winter when none occurred there (Fig. 4). Callinectes sapidus The blue crab occurs along the western Atlantic 172 coastline from Maine to northern Argentina, with the main commercial fishery in Chesapeake Bay (Williams 1984). Blue crabs occur on a variety of bot- tom types and are mainly abundant out to depths of 35 m. Callinectes sapidus ranked first in terms of bio- mass, making up about 19% of the entire catch of decapods and stomatopods (Table 1). Blue crabs oc- curred in 173 of the 303 trawl tows made during the survey. The stratified mean catch per tow for number and weight was greatest in the coastal zone during spring and summer (Table 5). Comparison of catches between areas showed abundance was comparable for strata off Georgia and North Carolina/South Carolina during all seasons (Fig. 4). Size composition of blue crabs differed between seasons with the average carapace width being greatest in winter and spring (Fig. 5). Mean cara- pace wadth was similar between areas, however, with those collected off Florida averaging 137 mm {n = 164) and those from strata off Georgia {n = 485) and North Carolina/South Carolina {n = 835) averaging 139 mm and 138 mm, respec- tively. Sex ratios were overwhelmingly dominant in terms of female C. sapidus for each season (Table 6). No ovigerous female crabs were collected in winter and only two individuals were found in fall collections. During spring and summer, however, the number of ovigerous females constituted 70% and 47% of the catch of female crabs, respectively. Among non-ovigerous females (n = 809), 95% of the blue crabs were mature. Greater numbers of females in the coastal zone are expected in view of the life history of the blue crab. With the exception of the breeding season, when females migrate into lower salinity waters of the estuary, they are usually found near the mouths of estuaries where the eggs are spawned and hatch. Most spawning occurs in spring and early summer, with the season becoming progressively shorter from Florida to North Carolina (Norse 1977). Males, however, remain in the middle to upper reaches of estuaries as juveniles and adults (Gunter 1950; Hildebrand 1954). Arenaeus cribrarius This portunid is a common inhabitant of the shal- low coastal zone along beaches (Hoese 1972; Wil- liams 1984). The known geographic range extends from Massachusetts to Brazil. It occurs abundant- ly in the penaeid shrimp grounds of the Gulf of Mex- WENNER AND WENNER: CRUSTACEANS FROM COASTAL HABITATS ico (Hildebrand 1954). Anderson et al. (1977) seined 422 specimens in the surf at Folly Beach, SC and A. cribrarius was the most abundant macroinver- tebrate collected in the same area by DeLancey (1984). In the present study, A. cribrarius consti- tuted only 2% of the total catch of decapods and stomatopods but occurred in 43% of the total col- lections (Table 1). Mean catch per tow increased from northern to southern sampling areas, with highest catches (10 individuals/tow) off Florida. Catches decreased to 5 individuals/tow off Georgia to 2 individuals/tow off North Carolina/South Caro- lina. The stratified mean catch per tow showed a seasonal trend with highest catch occurring in sum- mer and fall (Table 5). This corresponds with obser- vations reported by Anderson et al. (1977) who found a positive correlation of number of crabs with water temperature. Average carapace width was greatest for indivi- duals collected off South Carolina/North Carolina (x CW = 82 mm, n = 286). Those from strata off Georgia (n = 436) and Florida {n = 544) averaged 78 mm. Size differences were noted between sea- sons, as well; however, the small number of indivi- duals collected in winter (n = 13) did not provide adequate information on size composition for that season. During spring (x CW = 81 mm, n = 97), summer (x CW = 77 mm, n = 794), and fall (i CW = 83 mm, n = 362), the size composition of the catch was similar. The M:F ratio was significant for every season ex- cept winter (Table 6). Male A. cribrarius were larger (S CW = 82 mm, n = 721) than females (i CW = 76 mm, n = 545). Trachypenaeus constrictus This penaeid shrimp is caught incidentally in the commercial shrimp fishery along the southeastern and Gulf coasts. Eldred (1959) reported that T. con- strictus, along with T. similis, constituted 7% of the annual catch in the Tortugas area of Florida. In the South Atlantic Bight, T. constrictus was most abun- dant in the 9-18 m depth zone sampled by Wenner and Read (1981, 1982). This species was seasonally abundant in collections from the coastal zone, with stratified mean catch per tow highest in fall (6 individuals) and winter (5 individuals) (Table 5). Increased abundance of the species during fall and winter was previously noted by Wenner and Read (1981, 1982) and is probably due to recruitment from spawning in spring and late summer (Williams 1969; Anderson 1970; Subrah- manyam 1971). The number of individuals per tow did not noticeably differ between the areas sampled (Fig. 4). Libinia emarginata The common spider crab ranges from Nova Scotia to south Florida where it occurs mostly on mud and mud-sand bottom in shallow water (Powers 1977). This species was reported by Hildebrand (1954) to be the most common large spider crab on the west- ern Gulf of Mexico shrimping grounds. Winget et al. (1974) found L. emarginata seasonally most abundant in spring and summer in Delaware Bay where it was common in mud of sloughs. This spe- cies ranked 14th in overall abundance in the current study and occurred in 56% of the trawl collections (Table 1). Abundance of L. emarginata was nearly equal between the three areas: 4 individuals/tow off Florida, 3 individuals/tow off Georgia, and 3 indivi- duals/tow off North Carolina/South Carolina. The stratified mean catch per tow differed among sea- sons with abundance highest in winter and spring (Table 6). Carapace length was similar between areas, with largest individuals collected off Georgia (S CL = 54 mm, n = 232), while those from Florida and North Carolina/South Carolina waters averaged 52 mm (n = 283) and 50 mm (n = 451), respectively. Analysis of size frequencies by season (not shown) indicated a broad range of sizes. Small individuals, reportedly associated with the coelenterate Stomo- lophus meleagris (Hildebrand 1954), occurred in low numbers during every season. Average size of the sampled individuals was lowest in spring (5 CL = 47 mm, n = 312) and summer (x CL = 50 mm, n = 179), while those collected in fall (i CL = 57 mm, n = 193) and winter (5 CL = 54 mm, n = 282) were slighter larger. Sex ratios were significantly different from unity in winter, when males dominated (Table 6). Winget et al. (1974) also noted dominance by male L. emarginata in winter. Carapace length differed be- tween the sexes, with males slightly larger (x CL = 53 mm, n = 514) than females (x CL = 50 mm, n = 447). Penaeus duorarum Pink shrimp occur from southern Chesapeake Bay to the Florida Keys, along the coast of the Gulf of Mexico to the southern Yucatan Peninsula (Williams 1984). In the southern United States, P. duorarum occurs in commercial quantities only off North Carolina. Pink shrimp reach maximum abun- 173 FISHERY BULLETIN: VOL. 87, NO. 1 dance in the coastal zone at depths from 11 to 37 m (South Atlantic Fishery Management Council 1981). Pink shrimp were the least abundant Penaeus col- lected in this study (Table 1). They were collected in 101 of the 303 trawl tows made. The stratified mean catch per tow was highest for collections in spring when 0.15 individuals per tow were collected (Table 5). During spring, catches were highest in strata off Florida (Fig. 4). The average size of pink shrimp was greatest in spring (J TL = 121 mm, n = 502). Mean sizes during other seasons were winter (x TL = 113 mm, n = 11), summer (x TL = 95 mm, n = 101), and fall (ic TL = 106 mm, n = 209). Average total length decreased from northern to southern areas as follows: North Caro- lina/South Carolina (x = 119 mm, n = 256), Georgia (i = 114 mm, n = 108), Florida (i = 111 mm, n = 459). The increased abundance and size of shrimp in spring probably relates to their movement to the nearshore zone then. In North Carolina, pink shrimp emigrate from the estuaries in May and June, at which time spawning takes place (Williams 1984). Kennedy and Barber (1981) reported that movement offshore of Cape Canaveral begins in April and May. The larger average size of pink shrimp in spring probably reflects the presence of roe-bearing females in the coastal zone at that time. ACKNOWLEDGMENTS We sincerely thank P. Richards, M. Schwarz, and J. LaRoche of the South Carolina Wildlife and Ma- rine Resource Department RV Lady Lisa and RV Atlantic Sun. Scientific personnel whose assistance was invaluable included K. Thornley, W. Waltz, W. Roumillat, D. Stubbs, C. Barans, and others from the S.C. MARMAP project. Technical assistance was provided by M. J. Clise, K. Swanson, and M. Lentz. This work is a result of research sponsored by the National Marine Fisheries Service (MAR- MAP program) under contract no. 6-35147. LITERATURE CITED Anderson, W. D., Jr., J. D. Dias, R. K. Dias, D. M. Cupka, AND N. A. Chamberlain. 1977. The macrofauna of the surf zone off Folly Beach, South Carolina. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-704, 23 p. Anderson, W. W. 1956. January to April distribution of the common shrimp on the South Atlantic continental shelf. U.S. Fish Wildl. Serv., Spec. Sci. Rep., Fish. 171, 14 p. 1970. Contributions to the life histories of several penaeid shrimps (Penaeidae) along the South Atlantic coast of the United States. U.S. Fish Wildl. Serv., Spec. Sci. Rep.- Fish., 605 p. Blanton, J. 0., AND L. P. Atkinson. 1978. Physical transfer processes between Georgia tidal in- lets and nearshore waters. In J. Wiley (editor). Estuarine interactions, p. 515-532. Acad. Press, N.Y. BOESCH, D. F. 1977. Application of numerical classification in ecological in- vestigations of water pollution. Va. Inst. Mar. Sci., Spec. Sci. Rep. No. 77, 114 p. Gloucester Pt., VA. Briggs, J. C. 1974. Marine zoogeography. McGraw-Hill Book Co., N.Y., 475 p. Buchanan, C. C. 1973. Effects of an artificial habitat on the marine sport fishery and economy of Murrells Inlet, South Carolina. Mar. Fish. Rev. 35(9):15-22. Bumpus, D. F. 1973. A description of the circulation on the continental shelf of the east coast of the United States. In B. A. Warren (editor). Progress in oceanography. Vol. 6, p. 111-157. Camp, D. K. 1973. Stomatopod Crustacea. Mem. Hourglass Cruises 3(pt. 2);1-100. Camp, D. K.. N. H. Whiting, and R. E. Martin. 1977. Nearshore marine ecology at Hutchinson Island, Flo- rida: 1971-1974. V. Arthropods. Fla. Mar. Res. Publ. 25:1-63. CERAME-VrVAS, M. J., AND I. E. GRAY. 1966. The distributional pattern of the benthic invertebrates of the continental shelf off North Carolina. Ecology 47:260-270. Clifford, H. T., and W. Stephenson. 1975. An introduction to numerical classification. Acad. Press, N.Y., 229 p. Cochran, W. G. 1977. Samphng techniques. 3rd ed. John Wiley, N.Y. , 428 P- COELHO, P. A., AND M. DE A. RaMOS. 1972. A constitui(;ao e a distribucicjao de fauna de decApodos do litoral leste da America do Sul entre as latitudes de 5°N e 39°S. Trabalhos Inst. Oceanogr., Univ. Fed., Pemanbueo, Recife 13:133-236. Cook, H. L., and M. J. Lindner. 1970. Synopsis of biological data on the brown shrimp Penaeus azteeus Ives, 1891. FAO Fish. Rep. 57:1471-1497. DeLancey, L. B. 1984. An ecological study of the surf zone at Folly Beach, South Carolina. M.S. Thesis, Coll. Charleston, Charleston, SC, 131 p. Dudley, D. L., and M. H. Judy. 1971. Occurrence of larval, juvenile, and mature crabs in the vicinity of Beaufort Inlet, North Carolina. U.S. Dep. Com- mer., NOAA Tech. Rep. NMFS SSRF 637, 10 p. Eldred, B. 1959. A report on the shrimps (Penaeidae) collected from the Tortugas controlled area. Fla. State Board Conserv.. Spec. Sci. Rep. 2, 6 p. Farmer, C. H., J. D. Whitaker, and N. L. Chipley. 1978. Pilot study to determine the overwintering patterns of white shrimp. Final report for NMFS Contract 03-6-042- 35105, 93 p. Gore, R. H., and L. J. Becker. 1976. Studies on stomatopod Crustacea from the Indian River 174 WENNER AND WENNER: CRUSTACEANS FROM COASTAL HABITATS region of Florida. II. An annotated check list of the mantis shrimps of the central eastern Florida coast. Proc. Biol. Soc. Wash. 89(10):147-184. Grosslein, M. D. 1969. Groundfish survey program of BCF Woods Hole. Commer. Fish. Rev. 31(8-9):22-30. GUNTER. G. 1950. Seasonal population changes and distributions as related to salinity, of certain invertebrates of the Texas coast, including the commercial shrimp. Publ. Inst. Mar. Sci., Univ. Tex. 1(2):7-51. Haefner, p. a., Jr. 1985. Morphometry, reproduction, diet and epizoites of Ovalipes stephensoni Williams, 1976 (Decapoda, Brachyura). J. Crust. Biol. 5(4):658-672. Harris, C. Duane. 1974. Observations on the white shrimp {Penaeus setiferus) in Georgia. Ga. Dep. Nat. Res., Contrib. Ser. No. 27, 54 p. Hay, W. p., and C. A. Shore. 1918. The decapod crustaceans of Beaufort, N.C., and the surrounding region. Bull. [U.S.] Bur. Fish. 35:369-475. HiLDEBRAND, H. H. 1954. A study of the fauna of the brown shrimp {Penaeus aztecus Ives) grounds in the western Gulf of Mexico. Publ. Inst. Mar. Sci., Univ. Tex. 3:233-366. HOESE. H. D. 1972. Invertebrates. In H. D. Hoese and J. M. Valentine, Jr. (editors). University of Southwestern Louisiana Studies on the Chandeleur Islands, p. 33-45. Res. Ser. No. 10, Biology. 1973. A trawl study of nearshore fishes and invertebrates of the Georgia coast. Contrib. Mar. Sci. 17:63-98. Johnson, A. S., H. 0. Hillestad, S. F. Shanholtzer, and G. F. Shanholtzer. 1974. An ecological survey of the coastal region of Georgia. Natl. Park Sen-. Monogr. Ser. No. 3, 233 p. Joyce, E. A. 1965. The commercial shrimps of the northeast coast of Flo- rida. Fla. Board Conserv. Mar. Lab., Prof. Pap. Ser., No. 6, 224 p. Joyce, E. A., and B. Eldred. 1966. The Florida shrimping industry. Fla. Board Conserv., Educ. Ser. No. 15, 47 p. Reiser, R. K., Jr. 1976. Species composition, magnitude and utilization of the incidental catch of the South Carolina shrimp fishery. S.C. Mar. Resour. Cent. Tech. Rep. 16, 94 p. 1977. The incidental catch from commercial shrimp trawlers of the South Atlantic states. S.C. Mar. Resour. Cent. Tech. Rep. 26, 38 p. Kennedy, F. S., Jr., and D. G. Barber. 1981. Spawning and recruitment of pink shrimp, Penaeus duorarum. off eastern Florida. J. Crust. Biol. 1:474-484. Lambert, J. M., and W. T. Williams. 1962. Multivariate methods in plant ecology; Nodal analysis. J. Ecol. 50:775-802. Lance, G. N., and W. T. Williams. 1967. A general theory of classificatory sorting strategies. I. Hierarchical systems. Comput. J. 9:373-380. Leber, K. M. 1982. Seasonality of macroinvertebrates on a temperate, high wave energy beach. Bull. Mar. Sci. 32:86-98. Lee, T. N., and L. J. Pietrafesa. 1987. Summer upwelling on the southeastern continental shelf of the U.S. during 1981 : Circulation. Prog. Oceanogr. 19:267-312. Llndner, M. J., AND W. W. Anderson. 1956. Growth, migrations, spawning and size distribution of shrimp, Penaeus setiferus. Fish. Bull.. U.S. 56:555-645. Manning, R. B. 1969. Stomatopod Crustacea of the western Atlantic. Stud. Trop. Oceanogr. (Miami) No. 8, 380 p. Mathews, T. D., and 0. Pashuk. 1984. Shelfwater response to the cold winters of 1977 and 1978 in the South Atlantic Bight. Litoralia 1:41-58. Music, J. L. 1979. Assessment of Georgia's shrimp and crab resources. Ga. Dep. Nat. Res., Coastal Res. Div., Contrib. Ser. No. 30, 75 p. MusicK, J. A., AND J. D. McEachran. 1972. Autumn and winter occurrence of decapod crustaceans in Chesapeake Bight, USA. Crustaceana 22:190-200. Norse, E. A. 1977. Aspects of the zoogeographic distribution of CaUinectes (Brachyura:Portunidae). Bull. Mar. Sci. 27:440-447. Park, J. R. 1969. A preliminary study of portunid crabs in Biscayne Bay. Q. J. Fla. Acad. Sci. 32(l):12-20. 1978. A study of the distribution and ecology of the tropical swimming crabs of the western Atlantic. MS Thesis, Univ. Miami, Coral Gables, FL. 88 p. Pearse, a. S., H. J. HuMM, AND G. W. Wharton. 1942. Ecology of sand beaches at Beaufort, N.C. Ecol. Monogr. 12:35-190. Poole, R. W. 1974. An introduction to quantitative ecology. McGraw-Hill, Inc., N.Y., 532 p. Powers, L. W. 1977. A catalogue and bibliography to the crabs (Brachyura) of the Gulf of Mexico. Contrib. Mar. Sci. 20(supple.):l-190. PuLLEN, E. J., and W. L. Trent. 1970. White shrimp emigration in relation to size, sex, tem- perature and salinity. FAO Fish. Rep. 57:1001-1014. Purvis, E. E., and E. G. McCoy. 1974. Population dynamics of brown shrimp in Pamlico Sound, N.C. Dep. Nat. Econ. Res., Spec. Sci. Rep. No. 5, 26 p. Roe, R. B. 1969. Distribution of royal-red shrimp, Hymenopenaeus robustus, on three potential commercial grounds off the southeastern United States. U.S. Fish. Wildl. Serv., Fish. Ind. Res. 5:161-174. Rouse, W. L. 1970. Littoral Crustacea from southwest Florida. Q. J. Fla. Acad. Sci. 18:129-176. SaNDIFER, P. A., J. V. MiGLARESE, D. R. Calder, J. J. Manzi, and L. A. Barclay. 1980. Ecological characterization of the Sea Island coastal region of South Carolina and Georgia. Vol. Ill: Biological features of the characterization area. U.S. Fish Wildl. Serv., Off. Biol. Serv., Wash., D.C.. FW5/0B5-79/42, 620 p. Shipman, S. 1980. Penaeid shrimp migration and growth along the Georgia coast. Annual report October 1979-September 1980. Ga. Dep. Nat. Res., Coast. Res. Div., 46 p. Smith, N. P. 1983. Temporal and spatial characteristics of summer up- welling along Florida's Atlantic shelf. J. Phys. Oceanogr. 13:1709-1715. South Atlantic Fishery Management CoirNciL. 1981. Profile of the penaeid shrimp fishery in the South Atlantic. S. Atl. Fish. Manage. Counc, Charleston, S.C. 175 FISHERY BULLETIN: VOL. 87, NO. 1 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. SUBRAHMANYAM, C. B. 1971. The relative abtmdance and distribution of penaeid shrimp larvae off the Mississippi coast. Gulf Res. Rep., 3: 291-345. Tagatz, M. E. 1967. Noncommercial crabs of the genus CaUinectes in St. Johns River, Florida. Chesapeake Sci. 8:202-203. Trent, L. 1967. Size of brown shrimp and time of emigration from the Galveston Bay system, Texas. Proc. Gulf Caribb. Fish. Inst. 19:7-16. Vernberg, F. J., AND W. B. Vernberg. 1970. Lethal limits and the zoogeography of the faunal assem- blages of coastal Carolina waters. Mar. Biol. (N.Y.) 6:26- 32. Van Dolah, R. F., D. M. Knott, E. L. Wenner, T. D. Mathews, AND M. P. Katuna. 1984. Benthic and sedimentological studies of the George- town Ocean Dredged Material Disposal Site. S.C. Mar. Resour. Cent. Tech. Rep. 59, 97 p. Wass, M. L. 1955. The decapod crustaceans of Alligator harbor and adja- cent inshore areas of northwestern Florida. Q. J. Fla. Acad. Sci. 18:129-176. Wenner, E. L., W. P. Coon III, M. H. Shealy, Jr., and P. A. Sandifer. 1984. A five-year study of seasonal distribution and abun- dance of fishes and decapod crustaceans in the Cooper River and Charleston Harbor, S.C. prior to diversion. U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-782, 16 p. Wenner, E. L., D. M. Knott, R. F. Van Dolah, and V. G. BURREL, Jr. 1983. Invertebrate communities associated with hard bottom habitats in the South Atlantic Bight. Estuarine Coastal Shelf Sci. 17:143-158. Wenner, E. L., and T. H. Read. 1981. Distribution and assemblages of decapod crustaceans from the continental shelf of the South Atlantic Bight: 1977-79 MARMAP investigations. S.C . Mar. Resour. Cent. Tech. Rep. 49, 41 p. 1982. Seasonal composition and abundance of decapod crus- tacean assemblages from the South Atlantic Bight, USA. Bull. Mar. Sci. 32:181-206. Weinstein, M. P. 1979. Shallow marsh habitats as primary nurseries for fishes and shellfish. Cape Fear River, North Carolina. Fish. Bull., U.S. 77:339-357. Weymouth, F. W., M. J. Lindner, and W. W. Anderson. 1933. Preliminary report on the life history of the common shrimp Penaeus setiferus (Linn.). Bull. [U.S.] Bur. Fish. 48:11-26. Williams, A. B. 1969. A ten-year study of meroplankton in North Carolina estuaries: Cycles of occurrence among penaeidean shrimps. Chesapeake Sci. 10:36-47. 1984. Shrimps, lobsters, and crabs of the Atlantic coast of the eastern United States, Maine to Florida. Smithson. Inst. Press, 550 p. Williams. W. T., and G. M. Lambert. 1961. Multivariate methods in plant ecology. III. Inverse association analysis. J. Ecol. 49:717-729. WiNGET, R. R., D. Maurer, and H. Seymour. 1974. Occurrence, size composition and sex ratio of the rock crab, Cancer irroratus Say, and the spider crab, Libinia emarginata Leach, in Delaware Bay. J. Nat. Hist., 8:199- 205. Zar, J. H. 1984. Biostatistical analysis. 2nd ed. Prentice-Hall, Inc., Englewood Cliffs, NJ, 718 p. 176 STATUS OF THE TILEFISH, LOPHOLATILUS CHAMAELEONTICEPS, FISHERY OFF SOUTH CAROLINA AND GEORGIA AND RECOMMENDATIONS FOR MANAGEMENT Joseph E. Hightower' and Gary D. Grossman^ ABSTRACT We used a sex- and age-structured model and CPUE data from commercial and research vessels to assess the current status of the tilefish, Lopholutilus chamaeleont-keps, substock off South Carolina and Georgia. Based on commercial CPUE data and assumed natural mortality (M) rates of 0.10-0.25, we estimated that adult popiJation density prior to fishing ranged from 603 to 950 per km" and stock biomass ranged from 1.130 to 1,570 tonnes (t). Our estimates of the recommended fishing mortality rate ranged from 0.10 {M = 0.10) to 0.48 (M = 0.25), resulting in sustainable yields of 40 (A/ = 0.10) to 82 t (M = 0.25) per year. We obtained higher estimates of virgin population density (883-1,710 per km") when research CPUE data were used. Sustained yield estimates also were higher, ranging from 55 (M = 0.10) to 148 t{U = 0.25) per year. Average estimates of recommended yield from commercial and research CPUE data were 58 and 95 1, respectively. Observed yields m the developing fishery exceeded 100 1 in 1981-84 and in 1986; however, current observations indicate that fishing effort has declined to a low level in response to reduced catches. Based on the assumption that commercial CPUE data better reflect popula- tion trends, we recommend that the annual harvest not exceed about 50 t, which should result in a stock biomass of about 400-800 t. Apparent limitations on sustainable yield from the fishery probably can be attributed to the long lifespan, slow growth rate, and sedentary natiire of tilefish. The tilefish, Lofholatilus chamaeleonticeps, is a large demersal species found along the outer con- tinental shelf of North America from Nova Scotia to Key West, FL, along the Gulf coast to Campeche Bank, and off South America from Venezuela to Surinam (Freeman and Turner 1977). Tilefish are long-lived and have relatively slow growth rates (Harris and Grossman 1985). They are most com- mon at depths of 100-400 m and water tempera- tures of about 9°-14°C (Freeman and Turner 1977). Abundance is greatest in areas where substrates are suitable for burrow construction (Able et al. 1982; Grossman et al. 1985) or afford other shelter such as scour depressions around boulders (Valentine et al. 1980) or rubble piles (Low and Ulrich 1983). Katz et al. (1983) described two genetically distinct tilefish stocks through use of morphometric and elec- trophoretic data: one composed of tilefish from the Middle Atlantic Bight (MAB), and one composed of tilefish from the South Atlantic Bight (SAB) and Gulf of Mexico. Larval transport from the Gulf of 'School of Forest Resources, University of Georgia, Athens, GA 30602; present address: Southwest Fisheries Center Tiburon Laboratory, National Marine Fisheries Service, NOAA, 3150 Paradise Drive, Tiburon, CA 94920. ^School of Forest Resources, University of Georgia, Athens, GA 30602. Manuscript accepted October 1988. Fishery Bulletin, U.S. 87:177-188. Mexico to the SAB may be responsible for similar- ities in electrophoretic results for these two areas (Katz et al. 1983). Katz et al, (1983) suggested, how- ever, that it may be necessary to manage these sub- stocks separately because of their wide geographic separation. If true, this should be done with the understanding that Gulf of Mexico populations may serve as a source of recruits to SAB populations (Katz et al. 1983). Tilefish have been harvested commercially in the MAB since 1915, with annual landings ranging from <1 tonne (t) to 4,500 t (Turner et al. 1983). Land- ings from the SAB and Gulf of Mexico were small prior to 1980 (Low et al. 1983). A limited number of tilefish were caught incidentally in the deepwater grouper fishery off South Carolina (Low and Ulrich 1982). Recreational catches were small because of the depth at which tilefish occur (Low and Ulrich 1982). Commercial fisheries have since developed in both the SAB and Gulf of Mexico, due in part to an interest in diversification within the shrimp in- dustry (Low et al. 1983). For the segment of the SAB tilefish fishery oper- ating off South Carolina and Georgia, increased fishing effort has resulted in a substantial increase in tilefish landings since 1978 (Table 1). In addition, a considerable nimiber of tilefish caught off Georgia 177 FISHERY BULLETIN: VOL. 87, NO. 1 Table 1 .—South Carolina and Georgia tilefish, Lopholatilus cham- aeleonticeps. landings from 1978 to 1986. Georgia landings of tilefisfi and blueline tilefish, Caulolatilus microps, were not reported separately routinely prior to 1985. Landings by gear type were available for 1980-85, however, so rod and reel and electric reel catches of unclassified tilefish (which are almost exclusively blue- line tilefish) were excluded to yield a more accurate estimate of Georgia tilefish landings. and Grossman 1986) to estimate maturity at age by using a logistic model: 1 Landings (metric tons) Year South Carolina ' Georgia^ Total 1978 2.59 ^0.10 2.69 1979 10.08 '0.02 10.10 1980 43.89 (confidential) >43.89 1981 103.74 0.35 104.09 1982 163.78 7.99 171.77 1983 252.55 6.67 259.22 1984 190.44 39.49 229.93 1985 70.49 13.04 83.53 1986 142.96 (confidential) >1 42.96 lAndy Jennings, Fisheries Statistics Office. South Carolina Wildlife and fvlarine Resources Department. Cfiarleston SC. pers. commun 1987, 2S Gordon Rogers, Fisheries Statistics Coordinator, Georgia Department of Natural Resources, Brunswicl^, GA, pers commun, August 1987. ^Georgia Department of Natural Resources, 1983 Georgia landings an- nual summary 1980. Ga Dep Nat. Resour,, Coastal Resour. Div., Bruns- wick, GA. from about 1982 to 1984 were landed in Florida; therefore, landings in Table 1 are a conservative estimate of removals from these grounds. Low et al. (1983) provided a preliminary estimate of maximum sustained yield (MSY = 162 t) for the developing fishery. Their estimate was derived from Gulland's (1971) model MSY = O.SMBq, where M was the natural mortality rate and Sq was an esti- mate of virgin biomass. Our objective was to pro- vide updated yield estimates for use in managing the tilefish fishery off South Carolina and Georgia. We base these estimates on the 9-yr sequence of catches from the developing fishery as well as re- cently obtained information on the growth, mortal- ity, and reproductive biology of tilefish from this area. METHODS We used a deterministic sex- and age-structured model to simulate the developing tilefish fishery and to calculate sustainable yields. The model was based on the following assumptions and data sources. We assumed that the natural mortality rate (M) ranged from 0.10 to 0.25 (Harris and Grossman 1985). We used sex-specific estimates of weight-at-age because the von Bertalanffy growth curves and length- weight relationships used to calculate weight-at-age differed significantly by sex (Harris and Grossman 1985). We used maturity-at-length data (Erickson Ps.a = 1/(1 + exp{-bMa,{L,a ' L50Mn,))) (1) where p^^ was the proportion of sex-s, age-a fish that were sexually mature, 6ji^„, was a parameter affecting the steepness of the curve, L^^ was the standard length (SL) of sex-s fish at age a, and L50ji^„, was a parameter representing the length at which 50% of the fish were sexually mature. We assumed that the total biomass of sexually mature females (Sf) was an adequate measure of spawning potential. We used a logistic model to relate selectivity to length: set 1/(1 + expi- bseliLs,, - LbOsel))) (2) where sel^^ was the proportion of sex-s, age-a fish that were vulnerable to fishing, 6^^, was a param- eter affecting the steepness of the curve, and LbOggi was a parameter representing the length at which 50% of the fish were vulnerable to fishing. Based on length-frequency data (Harris and Gross- man 1985), we assumed that female and male tile- fish reached 50% vulnerability at about 475-500 and 500-525 mm SL, respectively. We used a slope parameter (65^,) of 0.05 so that selectivity-at-age in the simulated fishery ranged from about at age 5 to 1.0 at age 11 (Harris and Grossmar '985). Parameter estimates used in the model are sum- marized in Table 2. We assumed that the relationship between spawn- ing stock size and subsequent recruitment was weak or nonexistent because 1) tilefish produce pelagic larvae (Fahay and Berrien 1981); 2) there may be substantia! egg or larval transport between the Gulf of Mexico and SAB (Katz et al. 1983); and 3) tile- fish are dependent on the availability of shelter (Valentine et al. 1980; Able et al. 1982; Low and Ulrich 1983; Grossman et al. 1985). To represent the stock-recruitment relationship, we used a Beverton- Holt curve of the following form (Kimura 1988): N,.S + 6] = 0.5Afe[0]g/[<]«/[0] 1 - A(l - Sa«]/S,[0]) (3) where N^f^lt + 6] was the number of sex-s, age-6 recruits in year t + 6, N(,[0] was the virgin recruit- ment level for both sexes combined, and S,[t] and S,[0] were the biomass levels for spawning females in year t and prior to fishing, respectively. The pa- 178 HIGHTOWER and GROSSMAN: TILEFISH FISHERY rameter A controlled the degree of density-depen- dence. We assumed that recruitment was either con- stant (A = 1.000) or decreasing by 10% when the spawning stock was reduced by 50% (A = 0.889). Few studies have shown a statistically significant relationship between spawning stock and recruit- ment (Hennemuth 1979); nevertheless, recruitment would be expected to decline at high Fs. For that reason, we used the latter assumption to explore the effect of the stock-recruitment relationship on the form of the yield curve. Other investigators have used this approach to obtain conservative estimates of equilibrium yield when information on the stock- recruitment relationship was unavailable (Lenarz and Hightower 1985; Henry 1986; Hightower and Lenarz 1986). We simulated the fishery using virgin recruitment levels of 10,000-200,000 6-yr-old fish. This range would result in virgin population sizes of 45,000- 2.1 million fish, depending on the assumed level of natural mortality. Assuming that the area inhabi- tated by tilefish off South Carolina and Georgia is about 476 km- (Low et al. 1983), these population sizes correspond to adult densities of 95-4,400 per km-. This appeared to be an adequate range of den- sities, given that estimates of tilefish burrow den- sity in the Hudson and Veatch Canyons off southern New England ranged from 119 to 2,434 per km- in 1980 (Grimes et al. 1986). As Low et al. (1983) noted, the 1974-78 catch rates off southern New England (0.49-0.93 kg/hook; Grimes et al. 1980) were sim- ilar to the 1981-82 catch rate in the expanding fish- ery off South Carolina and Georgia (0.86 kg/hook). Table 2— Parameter estimates for the sex- and age-struc- tured model of tfie tilefish fishery off South Carolina and Georgia. Female Male von Bertalanffy' i-. =792 922 k 0.090 0086 r(zero) - 1 .774 - 0.920 Length-weight' b(1) 2.28572E-8 7.92693E-9 to (2) 2.974 3.141 tvlaturity-at-age ^Mal 0.030 0018 L50«„ 495 458 Selectivity-at-age bsei 0.05 0.05 L50s„ 475. 500 500, 525 'Harris and Grossman (1985) Lengtti-weight reiationstiip: w = bllji."!'! 2The estimate of L„ in Harris and Grossman (1985) {895 mm) was incorrect. Because tilefish catches were negligible prior to 1978, the starting (1978) number-at-age vector at each recruitment level was assumed to be the equi- librium vector obtained at an F of 0. We assumed that our estimates of total landings were much more reliable than our estimates of fishing effort. For that reason, we solved iteratively for the se- quence of fishing mortality rates that would produce the observed 1978-86 catches (Methot in press). For example, we began by solving for the 1978 F that would produce the 1978 catch biomass, and then used that F to project the number-at-age vector re- maining in 1979. [We assumed that the final (1986) F should not exceed 2.0 (an exploitation rate of 80-84%), in order to rule out those cases where the 1986 harvest was attained by removing essentially all remaining tilefish.] Using this approach to esti- mate F, the observed and simulated catch biomass levels match exactly, although the observed and simulated age distributions may be different. Note that if we had a similar degree of confidence in our estimates of catch and fishing effort, it might be more appropriate to minimize differences between observed levels and model estimates of both catch and effort (see for example, Deriso et al. 1983), rather than forcing the model to reproduce the catches exactly. At each virgin recruitment level, we calculated the correlation between the estimated 1978-86 Fs and estimates of total effort based on CPUE data. We used two sources of CPUE data: 1) commercial snapper reel CPUE from 1980 to 1982 South Carolina vessels (Low and Ulrich 1983); and 2) mean longline CPUE from 1982 to 1985 research cruises aboard the RV Georgia Bulldog. Based on commercial snapper reel kg/landing (figure 13 in Low and Ulrich 1983), we estimated that observed annual landings would have required more than 22 trips in 1980, 89 in 1981, and 445 in 1982 (Table 3). Using research cruise estimates of longline kg/hook, we estimated that observed annual lan- dings would have required 351,000 hooks fished in 1982, 1.9 million in 1983, 1.6 million in 1984, and 380,000 in 1985 (Table 3). The research catches were made using standard commercial longline gear (Har- ris and Grossman 1985). We also obtained a com- posite 1980-85 effort series using the ratio of hooks fished to trips in 1982 (Table 3), but our results were the same as when only research CPUE data were used. At each level of natural mortality, we selected the virgin recruitment level that maximized the corre- lation between estimates oiF and fishing effort. The selected recruitment level was used in the equilib- 179 FISHERY BULLETIN: VOL. 87, NO. 1 Table 3.— Estimates of total effort based on commercial and research! catch per unit effort (CPUE) and total commercial landings. Estimates of 1980-82 commercial CPUE (kg/landing) were based on landings by snapper reel vessels (figure 13, Low and Ulricti 1983). Researcfi CPUE estimates (kg/tiook) were based on longline sets aboard tfie RV Georgia Bulldog, each estimate is an average of seasonal averages from spring, summer, and fall cruises. Estimates of hooks fished in 1980 and 1981 were based on the commercial CPUE data, using the ratio of trips to hooks fished in 1982. CPUE Total Commercial Research landings Estimated Estimated Year (kg/landing) (kg/hook) (kg) trips hooks fished 1980 1,950 >43,890 >22 (>17,331) 1981 1,174 _ 104,090 89 (70,110) 1982 386 0.490 171,770 445 350,551 1983 0.137 259,220 1,892,117 1984 0.140 229,930 1,642,357 1985 — 0.220 83,530 379,682 rium yield calculations. Our approach for selecting virgin recruitment levels was based on the "tuning" process used in cohort analysis (Mohn 1983; Rivard 1983). In that approach, auxiliary information is used to "adjust" or "fine-tune" the estimates itera- tively so that the output from cohort analysis "matches" some series of observations (Rivard 1983). The level of agreement between the obser- vations and model predictions can be measured using correlation or regression techniques (Mohn 1983). We obtained estimates of equilibrium yield by ex- pressing the number of sex-s fish in each age class {N,„,a = 6, . . .,n, where n refers to fish ages 30 and older) as a function of the number of age-6 female fish (iV^g)- Following Getz (1980), we as- sumed that N, .1 = [n exp{-Z,^j)]Nfe, 1 = 6 6, .... w - 2 (4) JV,„ = [n exp(-Z,,)/(l - exp(-Z,,„)]7V,;6 (5) where Z, ^ was the total mortality rate for sex-s, age-j fish. Using Equations (4) and (5), female spawning stock can be redefined as a function of N, ffi- Sf= 2 Af,-.„ w^,, p/.. (6) n-l I a = 7 NffilWffi Pf.6 + 2!^ W/,a Pf.a ^.n^ exp(-Z,j) + Wf, Vfn "n' exv(-ZfJ(\ - exp(-Z,;„)] (7) Nf, \(F) (8) where i^(F) is the bracketed expression in Equation (7) for a specified F. We then substituted (iV/g 'j>(i^)) for Sf and solved Equation (3) for the equilibrium recruitment level as a function of F: iV,6 = (0.5 iVelO] W) - S/[0] + A5^[0]/(A \{F)). (9) The virgin spawning stock S,[0] was calculated from Equations (4) to (6) for the specified level of M and virgin recruitment. We used Equations (4)- (9) to calculate the equilibrium number-at-age vec- tor and associated yield for F& from 0.0 to 0.5. Following Francis (1986), we defined the target fishing mortality rate as Fq.i for the constant re- cruitment case and i^^sy for the density-dependent case. F„ 1 was the fishing mortality rate at which the slope of the yield curve was one-tenth the slope of the curve at the origin (Gulland and Boerma 1973). Compared to managing for maximum sus- tained yield, the i^o i policy usually results in greater economic efficiency when constant recruit- ment is assumed (Gulland and Boerma 1973; Sissen- wine 1981; Francis 1986). An additional advantage is that a larger spawning stock would be maintained (Sissenwine 1978). The less conservative i^^sy policy was assumed to be appropriate for the more conservative density-dependent case. The recom- mended yields for the constant recruitment and density-dependent cases were the equilibrium yields at FffA and F^^^ respectively. 180 HIGHTOWER and GROSSMAN: TILEFISH FISHERY RESULTS AND DISCUSSION Our approach for estimating virgin recruitment level was similar to stock reduction analysis (SRA) (Kimura et al. 1984; Kimura 1985) except that we used a more general model to represent stock dynamics. In both approaches, a model is fully spe- cified and the sequence of J^'s used to drive the model are those that would have produced the observed sequence of catches. A range of solutions can be ob- tained corresponding to a range of virgin recruit- ment levels, but the solution set can be restricted by comparing model predictions to auxiliary data. We obtained similar 9-yr patterns of F at differ- ent levels of virgin recruitment, particularly at higher recruitment levels where Fs were low in all years (Fig. 1). For that reason, correlation coeffi- cients were similar over a wide range of recruitment levels (Fig. 2). Stronger conclusions about the true level of virgin recruitment may be possible once additional years of catch and CPUE data become available. Based on commercial snapper reel CPUE data, the virgin recruitment level that maximized the cor- relation between estimates of F and fishing effort ranged from 30,000 to 100,000, depending on the assumed selectivity parameters and level of natural mortality (Table 4, Fig. 2). Correlations were high at all virgin recruitment levels, with maximum values obtained at the lowest recruitment levels capable of sustaining the 1978-86 observed catches (in order to match the sharp decline in CPUE from 1980 to 1982). Estimates based on research cruise CPUE data were higher, ranging from 40,000 to 180,000 (Table 5, Fig. 2). In both cases, the results were much more sensitive to M than to the LSOg^; parameter of the selectivity function (Tables 4, 5). Based on these estimates of the virgin recruitment level, the adult population prior to fishing would have ranged from 287,000 to 452,000 fish (commer- cial CPUE) or 420,000 to 814,000 fish (research CPUE). Assuming 476 km^ of tilefish habitat off South Carolina and Georgia (Low et al. 1983), the estimated density prior to fishing would have been 603-950 (commercial CPUE) or 883-1,710 (research CPUE) per km^. We are not aware of other esti- mates of tilefish density prior to fishing. Submer- sible dives were made on the South Carolina tile- fish grounds after the period of (assumed) heavy exploitation; unfortunately, no density estimates are currently available. Comparisons with the exploited MAB stock are of some interest because MAB catch rates in the late 1970s were similar to initial catch rates off South Carolina and Georgia (Low et al. 0.35 1978 1979 1980 1981 1982 1983 1984 1985 1986 2000 1500 ■^ 1 000 500 1978 1979 1980 1981 1982 1983 1984 1985 1985 Year Figure 1.— Upper panel: estimated fishing mortality rates (F) from 1978 to 1986 at four (arbitrarily selected) levels of virgin recruitment (30,000-120,000 age-6 tilefish), assuming a natural mortality rate (M) of 0.10. Year-to-year changes in F were similar at other levels of M. Lower panel: estimates of fishing effort (hooks fished) based on commercial (1980-81) and research (1982-85) CPUE data. 1983). Our density estimates were similar to the bur- row density estimates from the MAB. Grimes et al. (1986) reported that biu*row density in Hudson Can- yon ranged from 1,815 per km' in 1980 to 1,132 in 1982. Estimates for Veatch Canyon ranged from 772 per km^ in 1981 to 1,531 in 1984. Tilefish den- sity in the MAB may be lower than these estimates because not all burrows may be occupied (Able et al. 1982). In addition, burrow density is highly vari- able (Able et al. 1987) and some burrows may be in- habited only during certain seasons, depending on water temperature (Grimes et al. 1986). Neverthe- less, these comparisons suggest that the density estimates we generated from catch data were reasonable. 181 FISHERY BULLETIN: VOL. 87, NO. 1 Table 4. — Estimates of virgin levels of recruitment, adult population density, and biomass; recommended levels of fishing mortality, biomass, and yield; and 1987 biomass. Estimates were obtained from commercial CPUE data, using tvi/o sets of selectivity (LSOs^,) parameters, four levels of natural mortality (M), and two assumptions about the stock-recruitment relationship (A = 0.889 - recruitment dependent on spawning stock, A = 1 .000 - recruitment constant). Female/male selectivity parameter LSOj^, M: Virgin recruitment level (thousands) Virgin population density (#/km^) Virgin biomass (t) Recommended F (A = 0.889) Recommended F (A = 1 .000) Recommended biomass (t) (A = 0.889) Recommended biomass (t) (A = 1 .000) Recommended yield (t) (A = 0.889) Recommended yield (t) (A = 1.000) Estimated 1987 biomass (t) 475/500 mm SL 500/525 mm SL 0.10 0.15 0.20 0.25 0.10 0.15 0.20 0.25 30 40 70 90 30 40 70 100 662 603 811 855 662 603 811 950 1,574 1,128 1,266 1,160 1,574 1,128 1,266 1,288 0.10 0.15 0.23 0.33 0.13 020 0.30 0.48 0.10 0.16 0.23 0.32 0.11 0.17 0.25 0.34 553 437 511 485 509 408 504 535 678 524 636 628 700 551 680 753 41 40 57 62 42 42 58 70 51 51 72 78 52 51 70 82 624 247 468 435 636 261 484 584 c o Commercial CPUE 0.970 - \ \ 0.9B5 - 0.960 _ ■. 0.955 -\ "•.. 0.950 - ^^_;; „ 0.945 \ 30 50 70 90 110 1 30 1 50 170 1 90 0.10 0.15 0.20 0.25 o O 30 50 70 90 110 130 150 170 190 Virgin recruitment (thousands) Figure 2.— Correlation between fishing mortality rates in the simulated fishery and estimates of fishing effort from commercial (1980-82) and research (1982-85) CPUE data, at natural mortal- ity rates of 0.10-0.25 and virgin recruitment levels of 30,000- 200,000 age-6 tilefish. Recruitment levels without correlation coef- ficients were inadequate to sustain the observed catches. Results for the alternative set of selectivity parameters (L50^,,) were similar and are not shown here. 182 HIGHTOWER and GROSSMAN: TILEFISH FISHERY Table 5.— Estimates of virgin levels of recruitment, adult population density, and biomass; recommended levels of fishing mortality, biomass, and yield; and 1987 biomass. Estimates were obtained from research CPUE data, using two sets of selectivity (LSOg^,) parameters, four levels of natural mortality (M). and two assumptions about the stock-recruitment relationship (A = 0.889 - recruitment dependent on spawning stock, A = 1.000 - recruitment constant). Female/male selectivity parameter LSOj^, M^ Virgin recruitment level (thousands) Virgin population density (#/km^) Virgin biomass (t) Recommended F (A = 0.889) Recommended F (A = 1 .000) Recommended biomass (t) (A = 0.889) Recommended biomass (t) (A = 1 .000) Recommended yield (t) (A = 0.889) Recommended yield (t) (A = 1 .000) Estimated 1987 biomass (t) 475/500 mm SL 500/525 mm SL 0.10 0.15 0.20 0.25 0.10 0.15 0.20 0.25 40 70 110 160 40 70 120 180 883 1,056 1,275 1,520 883 1,056 1,391 1.710 2,098 1,974 1,989 2,062 2,098 1,974 2,170 2,319 0.10 0.15 0.23 0.33 0.13 0.20 0.30 0.47 0.10 0.16 0.23 0.32 0.11 0.17 0.25 0.34 738 765 802 863 679 713 864 962 904 917 999 1,117 932 965 1,165 1,355 55 71 89 109 56 73 100 126 69 90 114 139 70 89 120 148 1,152 1,107 1,203 1,354 1,164 1,121 1,400 1,627 The two sets of CPUE estimates resulted in differ- ent conclusions about the current status of the stock. Based on commercial CPUE data, 1987 stock bio- mass was 22-45% of virgin biomass and 51-105% of the recommended level (Table 4). Estimated fish- ing mortality rates increased from about 0.1 in 1981 to a range of 0.3-1.4 in 1986 (Fig. 3). Based on research CPUE data, 1987 stock biomass was 55- 70% of virgin biomass and 132-145% of the recom- mended level (Table 5). Estimated fishing mortality rates were much lower than from commercial CPUE data, increasing from about 0.08 in 1981 to 0.2 in 1986 (Fig. 3). We believe that the results obtained from com- mercial CPUE data are far more likely, given recent reported declines in directed fishing. The decrease in landings observed in 1985 was attributed in part to reduced fishing pressure. A large group of boats from the Port Canaveral, FL area left the fishery, and a number of Georgia vessels began fishing fur- ther north (M. V. Rawson^). As of April 1988, most South Carolina longline vessels had switched to other fisheries and little directed tilefish fishing was occurring (R. LoW). The difference in results for commercial and re- search CPUE estimates may be due to differences in areas fished. The RV Georgia Bulldog cruises were exploratory in nature, and catches were ob- tained primarily in the southern section of tilefish habitat off the Georgia coast (Harris and Grossman 'M. V. Rawson. University of Georgia Marine Extension Ser- vice, Brunswick, GA 31523, pers. commun. March 1987. 'R. Low, South Carolina Wildlife and Marine Resources Depart- ment, Charleston, SC 29412, pers. commun. April 1988. 1985). Early commercial effort was concentrated in the more northerly part of the tilefish habitat (Low et al. 1983). We recognize the commercial catch data provide a biased measure of abundance. Neverthe- less, because South Carolina landings predominated in the developing fishery, we believe that commer- cial catch data from the primary fishing grounds will be a better overall measure of changes in abundance. For that reason, we restrict our remaining com- ments to results from the commercial CPUE data. Declines in commercial CPUE may underrepresent actual dechnes in abundance because fishermen presumably would change tactics over time in order to maintain high catch rates. If so, our use of com- mercial catch data may result in an optimistic esti- mate of current abundance. We obtained equivalent estimates of 1987 biomass for the constant recruitment and density-dependent cases because of the short length of the data series, relative to the 6-yr lag between a reduction in spawning stock and subsequent lower recruitment to the fishery. Equilibrium yield curves differed substantially for the two recruitment assumptions (Fig. 4). Despite differences in equilibrium yield, recommended Fs, were very similar for the two cases because of the different criteria used to develop F and yield recommendations (Table 4). Recommended yield was moderately higher under the optimistic constant recruitment assumption (Table 4). Recom- mended F was higher when LSO^,,, was increased; however, differences in recommended yield were negligible. Estimated Fs differed substantially for the four levels of natural mortality (upper panel, Fig. 3). The large differences in 1984-86 Fs probably were due 183 FISHERY BULLETIN: VOL. 87, NO. 1 0.10 - 0.15 0.20 --- 0,25 1978 1979 1980 1981 1982 1983 1984 1985 1985 1978 1979 1980 1981 1982 1983 1984 1985 1986 Year Figure 3.— Estimated fishing mortality rates (F) from 1978 to 1986 at natural mortality rates of 0.10-0.25. At each level of natural mortality, the virgin recruitment level was determined based on the correlation between F and estimated fishing effort, where estimates of effort were based on commercial vessel or research vessel CPUE. Results for the alternative set of selectiv- ity parameters (LSOg^,) were similar and are not shown here. to the lack of CPUE estimates for those years. Virgin biomass levels were similar at different levels of M, ranging from 1,130 t atM = 0.15 to 1,570 t at M = 0.10 (Table 4). Estimates of 1987 biomass were somewhat similar and relatively low, ranging from247tatM = 0.15 to 636 t atM = 0.10(Table 4). Equilibrium yield curves differed in a predictable way at different levels of M (Fig. 4). Except at low Fs (<0.10), equilibrium yield increased as M in- creased, due to the higher estimates of virgin re- cruitment at higher levels of M. Recommended F was higher at higher levels of Af, ranging from 0.10 atM = 0.10 to 0.48 atM = 0.25 (Table 4). Recom- mended yield increased from 41-52 t at M = 0.10 to 62-82 t at M = 0.25. These results demonstrate that an important source of uncertainty in assessing tilefish stock status is the estimate of M. Harris and Grossman (1985) obtained 1982-83 catch curve estimates of Z equal to 0.25 for both female and male tilefish. Because the areas sampled by the RV Georgia Bull- dog were thought to have received little fishing pressure, M could be as high as 0.25. Catch curve estimates of Z would be biased, however, if vulner- ability to fishing increased with size. Turner et al. (1983) reported a decline in size at recruitment in the expanding MAB fishery and suggested that when larger tilefish were present, smaller ones either were less vulnerable to the gear or were avoided by fishermen. Z (and M) could be under- estimated if vulnerability to fishing increased with size. Alternatively, Z (and M) could be overestimated if significant catches of predominantly older fish were made in the areas sampled by the RV Georgia Bulldog. 184 HIGHTOWER and GROSSMAN: TILEFISH FISHERY c o — A= 1 .000, 475/500 ... A= 0.889, 475/500 A= 1 .000, 500/525 -.- A= 0.889, 500/525 0.1 0.2 0.3 0.4 Fishing mortaiity rote Figure 4.— Estimated equilibrium yield at four rates of natural mortality (Af ), two assumptions regarding the stock-recruitment relationship (A = 0.889 - recruitment dependent on spawning stock, A = 1.000 - recruitment constant), and two values for the female and male (F/M) selectivity parameter LSO^,, . Evidence that M is less than 0.25 was provided by predictive models used to estimate M from the growth rate {k) and maximum age (Alverson and Carney 1975) or from k, L„, and mean water tem- perature (Pauly 1980). Estimates from the Alverson- Carney model were 0.107 (female) and 0.118 (male), whereas estimates from Pauly's method were 0.175 (females) and 0.163 (males) (Harris and Grossman 1985). Furthermore, Hoenig (1983) provided a model that predicts total mortality rate (Z) as a function of maximum age. Maximum observed age from May 1982 to August 1983 samples was 32 for female and 33 for male tilefish (Harris and Grossman 1985). Using Hoenig' s model, Z would be 0.13 for each sex; therefore, that would be a maximum estimate of M. A potential source of error in this assessment is the assumption that the selectivity pattern is con- stant over time. Recent observations (R. Low, fn. 4) indicate that the size at first vulnerability to fish- ing has decreased from about 1 kg (Harris and Grossman 1985) to about 0.45 kg. The decreasing size at recruitment increases the Hkelihood of re- cruitment overfishing because fish are being har- vested well before the size at maturity (about 2-3 kg for females). Thus, the current model probably underestimates the impact of fishing. Higher Fs could be sustained if small tilefish were not vulner- able to fishing; unfortunately, it is difficult to regu- late age at entry for hook-and-line gear (Myhre 1974) and discard mortality under a minimum size regula- 185 FISHERY BULLETIN: VOL. 87, NO. 1 tion would likely be substantial (Huntsman and Manooch 1978a). A second potential source of error is the use of a deterministic model to represent recruitment. Changes in population size and CPUE are due not only to the impact of fishing but also to fluctuations in year-class strength. Because the estimates of virgin recruitment were based on changes in CPUE, fluctuations in recruitment that increase (decrease) the decline in CPUE would result in a lower (higher) estimate of virgin recruitment. Based on size-fre- quency data from the MAB tilefish fishery, Turner et al. (1983) suggested that fluctuations in tilefish year-class strength may be substantial. Using size- and age-frequency data collected off Georgia, Har- ris and Grossman (1985) found little evidence for strong fluctuations in year-class strength. Tilefish are somewhat difficult to age, however, so differ- ences in year-class strength could be hidden by age- ing errors. A third source of error is the unknown number of fish caught off South Carolina or Georgia, but landed in Florida. The impact on the assessment would depend on the magnitude of the catches and the years in which the catches occurred. If we ar- bitrarily assume that actual annual removals were 25% higher than combined South Carolina-Georgia landings, recommended Fs would be unchanged, whereas estimates of virgin recruitment and recom- mended yield would increase by about 25%. Thus, if Florida removals could be accounted for, estimates of stock size would be more accurate, but the in- crease in recommended yield would be offset by the increased catches, and would not result in increased overall landings. CONCLUSIONS The results of this study provide estimates of the relationship between yield and fishing mortality and of the recommended level for F. Results obtained using commercial CPUE estimates indicate that sus- tainable harvests from the fishery are quite low, and would be obtained at Fs considerably lower than observed in the developing fishery. We obtained higher estimates of population size and sustained yield using research CPUE data; however, we be- lieve that the commercial CPUE data better reflect population trends. We estimate that current stock size is about 200-600 t, compared with a recom- mended level of 400-800 t. If the stock could be rebuilt to the recommended level, it should support an annual harvest of about 50 t. A rebuilding stra- tegy is feasible for tilefish because catches are low except when directed fishing occurs (G. Ulrich^). At present, however, there are no restrictions on the tilefish fishery and despite reductions in effort, catches are probably large enough to prevent the stock from rebuilding (G. Ulrich fn. 5). Apparent limitations on sustainable yield of the tilefish fishery probably can be attributed to the demographic characteristics of the stock. In a typical fishery for a long-lived, slow-growing species, a few years of high catches are followed by a sharp decline and a subsequent period of low yield (Huntsman and Manooch 1978b; Leaman and Beamish 1984; Fran- cis 1986). Long-lived, sedentary species, such as reef fishes, may be particularly vulnerable to overfish- ing, even though fishing intensity may be low or the method inefficient (Huntsman and Manooch 1978b). Because tilefish are long-lived, slow-growing, and sedentary (due to their dependence on the availabil- ity of shelter), a similar pattern of exploitation can be expected for the tilefish fishery off South Caro- lina and Georgia. Leaman and Beamish (1984) recommended that conservative harvest strategies be developed for long-lived species untO the evolutionary implications of longevity are better understood. They suggested that extreme longevity (>50 years) may be an adap- tive response to ensure population persistence under reproductive uncertainty. For example, a long re- productive life might enable a species to inhabit deeper water (200-1,000 m) where few competitors or predators are found, even though recruitment into such areas may be highly variable (Leaman and Beamish 1984). If variability in recruitment has a significant effect on tilefish stocks, a conservative management strategy emphasizing maintenance of a range of age classes may be appropriate. ACKNOWLEDGMENTS Funding for this project was provided through contract NA8DAA-D-000918 from the Georgia Sea Grant Program, whose support is gratefully ac- knowledged. We thank the crew of the RV Georgia Bulldog for support in obtaining CPUE data, as well as C. Barans, M. J. Harris, R. A. Low, Jr., M. V. Rawson, S. G. Rogers, G. F. Ulrich, and two anony- mous reviewers for comments on an earlier version of the manuscript. We thank R. D. Methot for pro- viding a sex- and age-structured model upon which our model was based. We also thank C. Barans, R. A. Low, Jr., M. V. Rawson, and G. F. Ulrich for 'G. Ulrich, South Carolina Wildlife and Marine Resources Department, Charleston, SC 29412, pers. commun. July 1987, 186 HIGHTOWER and GROSSMAN: TILEFISH FISHERY their observations on the current status of the fishery. LITERATURE CITED Able, K. W., C. B. Grimes, R. A. Cooper, and J. R. Uzmann. 1982. Burrow construction and behavior of tilefish, Lophola- tilus chamaeleonticeps, in Hudson Submarine Canyon. En- viron. Biol. Fishes 7:199-205. Able, K. W., D. C. Twiche, C. B. Grimes, and R. S. Jones. 1987. Sidescan sonar as a tool for detection of demersal fish habitats. Fish. Bull., U.S. 85:725-736. Alverson, D. L., and M. J. Carney. 1975. A graphic review of the growth and decay of popula- tion cohorts. J. Cons. int. Explor. Mer 36:133-143. Deriso, R. B., T. J. Quinn II, AND P. R. Neal. 1985. Catch-age analysis with auxiliary information. Can. J. Fish. Aquat. Sci. 42:815-824. Erickson, D. L., and G. D. Grossman. 1986. Reproductive demography of tilefish from the South Atlantic Bight with a test for the presence of protogynous hermaphroditism. Trans. Am. Fish. Soc. 115:279-285. Fahay. M. p., and p. Berrien. 1981. Preliminary description of larval tilefish (Lopholatilzis chamaeleonticeps). Rapp. P. -v. RSun., Cons. int. Explor. Mer 178:600-602. Francis, R. C. 1986. Two fisheries biology problems in west coast ground- fish management. North Am. J. Fish. Manage. 6:453-462. Freeman, B. L., and S. C. Turner. 1977. Biological and fisheries data on tilefish. Lopholatilus chamaeleonticeps (Goode and Bean). U.S. Dep. Commer., Natl. Mar. Fish. Serv., NOAA, Tech. Serv. Rep. No. 5, 41 p. Getz. W. M. 1980. Harvesting models and stock recruitment curves in fisheries management. In W. M. Getz (editor). Mathemat- ical modelling in biologj' and ecology, p. 284-304. Springer- Verlag, Berlin, Germany. Grimes, C. B., K. W. Able, and R. S. Jones. 1986. Tilefish, Lopholatilus chamaeleonticeps. habitat, be- havior and community structure in Mid-Atlantic and south- ern New England waters. Environ. Biol. Fishes 15:273- 292. Grimes, C. B., K. W. Able., and S. C. Turner. 1980. A preliminary analysis of the tilefish, Lopholatilus chamaeleonticeps, fishery in the Mid-Atlantic Bight. Mar. Fish. Rev. 42(11):13-18. Grossman, G. D., M. J. Harris, and J. E. Hightower. 1985. The relationship between tilefish, Lopholatihus chamae- leorUie^s. abundance and sediment composition off Georgia. Fish. Bull., U.S. 83:443-447. GULLAND, J. A. 1971. Science and fishery management. J. Cons., Cons. int. Explor. Mer 33:471-477. GULLAND, J. A., AND L. K. BOEREMA. 1973. Scientific advice on catch levels. Fish. Bull., U.S. 71:325-335. Harris, M. J., and G. D. Grossman. 1985. Growth, mortality, and age composition of a lightly ex- ploited tilefish (Lopholatilus chamaeleonticeps) substock off Georgia. Trans. Am. Fish. Soc. 114:827-836. Hennemuth, R. C. 1979. Man as predator. In G. P. Patil and M. Rosenzweig (editors), Contemporary quantitative ecology and related econometrics, p. 507-532. Int. Stat. Ecol. Program, Stat. Ecol. Series. Vol. 12. Henry, F. D., Jr. 1986. Status of the coastwide chilipepper (Sebastes goodei) fisherj'. In Status of the Pacific coast groundfish fishery and recommendations for management in 1987. Pacific Fishery Management Council. Portland, OR. Hightower, J. E., and W. H. Lenarz. 1986. Status of the widow rockfish fishery, /ri Status of the Pacific coast groundfish fisherj' and recommendations for management in 1987. Pacific Fishery Management Coun- cil. Portland, OR. HOENIG, J. M. 1983. Empirical use of longevity data to estimate mortality rates. Fish. Bull., U.S. 81:898-903. Hlintsman, G. R., and C. S. Manooch III. 1978a. Minimum size limits for reef fishes. Proc. Annu. Conf. Southeast. Assoc. Fish Wildl. Agencies 32:509-513. 1978b. Coastal pelagic and reef fishes in the South Atlantic Bight. In Henry Clepper (editor). Marine recreational fish- eries. No. 3, p. 97-106. Proceedings of the Second Annual Marine Recreational Fisheries Symposium; Sport Fishing In- stitute, Washington. D.C. Katz, S. J., C. B. Grimes, and K. W. Able. 1983. Delineation of tilefish, Lopholatilus chamaeleonticeps, stocks along the United States east coast and in the Gulf of Mexico. Fish. Bull., U.S. 81:41-50. KiMURA, D. K. 1985. Changes to stock reduction analysis indicated by Schnute's general theory. Can. J. Fish. Aquat. Sci. 42: 2059-2060. 1988. Stock-recruitment curves as used in the stock-reduc- tion analysis model. J. Cons. int. Explor. Mer 44:253-258. KiMURA, D. K., J. W. Balsiger, and D. H. Ito. 1984. Generalized stock reduction analysis. Can. J. Fish. Aquat. Sci. 41:1325-1333. Leaman, B. M., and R. J. Beamish. 1984. Ecological and management implications of longevity in some northeast Pacific groundfishes. Int. North Pac. Fish. Comm., Bull. 42, p. 85-97. Lenarz, W. H., and J. E. Hightower. 1985. Stattis of the widow rockfish fishery. In Status of the Pacific coast groundfish fishery and recommendations for management in 1986. Pacific Fisherj' Management Coim- cU, Portland, OR. Low, R. a., and G. F. Ulrich. 1982. Reef fishes and associated management issues in South Carolina. S.C. Wildl. Mar. Resour. Dep. Educ. Rep. No. 14, 49 p. 1983. Deep-water demersal finfish resources and fisheries off South Carolina. S.C. Mar. Resour. Cent. Tech. Rep. No. 57, 24 p. Low, R. A., Jr., G. F. Ulrich, and F. Blum. 1983. Tilefish off South Carolina and Georgia. Mar. Fish. Rev. 45(4-6):16-26. Methot, R. D. In press. Synthetic estimates of historical abundance and mortality for nothern anchovy. In E. F. Edwards and B. A. Megrey (editors). Mathematical analysis of fish stock dynamics. Am. Fish. Soc. Symp. 6. MOHN, R. K. 1983. Effects of errors in catch and effort data on tuning cohort analysis, with a postscript on logistic production models. In W. G. Doubleday and D. Rivard (editors). Sam- pling commercial catches of marine fish and invertebrates, p. 141-150. Can. Spec. Publ. Fish. Aquat. Sci. 66. 187 FISHERY BULLETIN: VOL. 87. NO. 1 Myhre, R. J. 1974. Minimum size and optimum age at entry for Pacific halibut. Int. Pac. Halibut Comm., Sci. Rep. 55, 15 p. Pauly, D. 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. RiVARD, D. 1983. Effects of systematic, analytical, and sampling errors on catch estimates: a sensitivity analysis. In W. G. Double- day and D. Rivard (editors). Sampling commercial catches of marine fish and invertebrates, p. 114-129. Can. Spec. Publ. Fish. Aquat. Sci. 66. SiSSENWINE, M. P. 1978. Is MSY an adequate foundation for optimum yield? Fisheries 3:22-24, 37-42. 1981 . An overview of some methods of fish stock assessment. Fisheries 6:31-35. Turner. S. C, C. B. Grimes, and K. W. Able. 1983. Growth, mortality, and age/size structure of the fish- eries for tilefish, Lopholatilixs charmieleonticeps, in the Mid- dle Atlantic-Southern New England region. Fish. Bull., U.S. 81:751-763. Valentine, P. C, J. R. Uzmann, and R. A. Cooper. 1980. Geology and biology of Oceanographer Submarine Canyon. Mar. Geol. 38:283-312. 188 HABITAT VALUE OF NATURAL VERSUS RECENTLY TRANSPLANTED EELGRASS, ZOSTERA MARINA, FOR THE BAY SCALLOP, ARGOPECTEN IRRADIANS Ingrid Smith,' Mark S. Fonseca,^ Jose A. Rivera,^ and Keith A. Rittmaster'' ABSTRACT Bay scallops, Argopecten irradians, were used in a mark and recapture experiment to determine the habitat value of recently transplanted eelgrass. Zostera marina, meadows for fishery restoration and enhancement through stocking. The study site, adjacent to an island formed from dredge material, con- sisted of natural and transplanted eelgrass and of unplanted areas. Seventy-five marked bay scallops were placed in plots at a density of 2.2 scallops per m" on 20 February 1986. A month later, only 18 marked scallops were recovered; of these. 15 were found in the natural eelgrass beds. On the study site, 94% of 207 unmarked naturally occurring bay scallops were found in the natural eelgrass beds. Recovery of marked adult bay scallops was not affected by the distance from the dredge island; rather densities of natural scallop populations increased with distance from the island. A second, modified survey (30 March to 7 April 1986) was conducted specifically to examine the recovery of marked bay scallops; this survey again showed a high rate of loss both in the transplanted and unplanted areas. The two surveys showed that recently transplanted eelgrass meadows do not provide the same habitat functions as natural meadows for bay scallops. Stocking of adult scallops in early stage eelgrass transplants to enhance or restore that fishery does not appear to be feasible. A protracted period of time may pass before habitat function is returned for the bay scallops in transplanted eelgrass meadows. Results from these surveys also illustrate the need for careful consideration in the placement of dredge material in the coastal environment. Seagrass meadows form an essential habitat for a variety of marine organisms (Thayer et al. 1975, 1984; Kenworthy et al. 1988). These highly produc- tive ecosystems provide refuge, food resources, and nursery grounds for a number of commercially and recreationally harvested species. Recent concerns about loss of seagrass habitat in general (Thayer et al. 1984, 1985; Fonseca et al. 1985, 1987, 1988) have prompted research into ways in which that loss can be reduced. Since mitigation measures often require the creation of new seagrass meadows to replace damaged ones, it is critical that this trade-off provide a persistent habitat that is the functional equivalent of the one that is lost. Given our approach to creating seagrass beds by install- ing widely spaced planting imits that coalesce in 1-2 'Duke University Marine Laboratory, Beaufort, NC 28516. ^Southeast Fisheries Center Beaufort Laboratory, National Marine Fisheries Service, NOAA, Beaufort. NC 28516. [To whom correspondence and reprint requests should be addressed.] 'NOAA Corps, on assignment to Southeast Fisheries Center Beaufort Laboratory, National Marine Fisheries Service. NOAA, Beaufort, NC 28516. 'Southwest Fisheries Center La Jolla Laboratory, National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038. years, it is possible that an artificially propagated bed will require a certain time interval before it will attain natural meadow functions. If these created beds do not provide similar functional values as natural ones or if they require a very long time to do so, then the entire concept of seagrass bed miti- gation will have to be reexamined. These are critical questions, especially when seagrass restoration pro- jects have not produced more acreage than was lost (Fonseca et al. 1988). In the temperate zone, the dominant seagrass spe- cies is eelgrass, Zostera marina. Eelgrass has been utilized in many seagrass restorations (Fonseca et al. 1988). Recent losses of eelgrass and scallops in Long Island Sound due to a "brown tide" (Chris Smith pers. commun.^), and losses of scallops in Bogue and Back Sounds, Carteret County, NC, apparently due to a Ptychodiscus bloom, have prompted questions regarding seagrass and scallop restoration. Given the paucity of information on faunal recovery in restored or created seagrass beds, Manuscript accepted October 1988. Fishery Bulletin, U.S. 87:189-196. ^Chris Smith, Cooperative Extension Association of Suffolk County, Sea Grant Program, 39 Sound Avenue, Riverhead, NY 11901, pers. commun. June 1987. 189 FISHERY BULLETIN: VOL. 87, NO. 1 we sought to evaluate whether bay scallop, Argo- pecten irradians, stocking could be conducted simul- taneously with eelgrass bed creation. Under this approach, stocked adult bay scallops would be used as a source of spat settlement in the maturing eelgrass bed. Bay scallops often utilize eel- grass meadows throughout their life cycle (Outsell 1930; Kirby-Smith 1970; Thayer and Stuart 1974). During the postveliger stage of development, a bay scallop attaches itself to submerged substrates such as vegetation (eelgrass blades), shells, rocks, animal tubes, or macroalgae. At approximately 10 mm in shell width, the scallop detaches and settles onto the bottom sediments to complete its life cycle; adult sizes range between 5 and 7 cm (Outsell 1930; Kirby- Smith 1970; Thayer and Stuart 1974). During its lifespan (1.5-2 years), bay scallops feed upon phyto- plankton (Peirson 1983) and detritus (Kirby-Smith and Barber 1974), which are plentiful in eelgrass systems (Thayer et al. 1975). If scallop stocking could not be done concomitantly with bed creation, natural recovery of the scallop population could be substantially delayed. Our study was embedded in a larger, long-term study of eelgrass restoration and faunal recovery. In that study, eelgrass was transplanted onto sub- tidal dredge material and monitored to determine the rate at which these propagated areas attain functional characteristics of adjacent, natural meadows. In preparing for the bay scallop stocking study, we observed in an independent scallop dredging survey that scallop densities near the study site declined from 2.0 to nearly 0/m- between Novem- ber and January 1985. During this period, laughing gulls, Larus atricilla, were seen dropping live scal- lops, a common feeding activity for these birds (Pearson et al. 1959), onto a dredge material island adjacent to our study area. This suggests that the gulls were at least partially responsible for the ob- served decline in scallop densities. Because this por- tion of the study was designed to include an evalua- tion of developing eelgrass meadows as scallop habitat, the close proximity of the dredge island and the increased likelihood of high predation on the scallops by gulls had to be considered in the assess- ment. Oiven the decline in the natural scallop popu- lation, possibly exacerbated by gull predation, we utilized a mark and recapture technique to assess stocking feasibility. The general objectives of the study were to com- pare the capabilities of natural eelgrass, trans- planted eelgrass, and unplanted areas in support- ing a stocked adult bay scallop population as a means of enhancing recovery of a local fishery. Specifical- ly, we sought to 1) examine the feasibility of seed- ing adult scallops in newly transplanted eelgrass beds; 2) relate scallop density in experimental plots to a) the proximity of these plots to the dredge is- land and b) any preferential migration from the transplanted or unplanted areas to the adjacent, natural eelgrass beds; and 3) control for adult scallop recruitment by comparing the densities of naturally occurring scallops in natural and transplanted eelgrass beds of two spatial arrange- ments, as well as in unplanted plots within the study site. METHODS The study site (long. 76°32'W, lat. 34°40'N, Fig. 1) was located at the southern end of Core Sound and northwest of Cape Lookout, NC. Specifically, the experiment was conducted off the southwest side of a dredge material island in relatively shallow waters (0.15 m at low tide and 1.0 m at high tide). The island was originally created 10 years before the study with maintenance dredging deposits added every 2-3 years. The overall study site covered 4,556 m^, which was divided into five separate blocks ex- tending out from the island (Fig. 2). For this study on the scallops, only blocks 1, 3, and 5 were utilized. Each block contained five different experimental units which were 7.5 m on a side (56.25 m^). An ex- perimental unit was separated from adjace'-* units by a 7.5 m corridor. The five treatments for each experimental unit were as follows: 1) natural in- terior eelgrass (NI, >15 from unvegetated sub- strate), 2) natural eelgrass bordering unvegetated substrate (NE), 3) low perimeter to area (LPA) eelgrass transplant arrangement (see below), 4) high perimeter to area (HPA) eelgrass transplant ar- rangement (see below), and 5) bare (B), unplanted dredge material. Although positioning of the two natural treatments were fixed, the other three treatments were randomly assigned to the remain- ing three experimental units within each block. Each experimental unit contained eight plots (2.25 m-), which were consecutively located around the perimeter of the experimental unit (Fig. 2). These eight plots were designated to accommodate eight faunal sampling periods for the parallel study of fishery habitat establishment. The two transplant arrangements had different perimeter to area ratios in order to examine the refuge value of large, un- broken seagrass cover versus patchy cover. The LPA treatments had eelgrass planting units throughout the 7.5 m x 7.5 m area, whereas HPA 190 SMITH ET AL.: TRANSPLANTED EELGRASS FOR BAY SCALLOPS ONSLOW BAY Cape Lookout Figure 1.— An aerial view of Back Sound, North Carolina. Location of the study site and dredge island is long. 76°32'W. lat. 34°40'N. treatments had 16 planting units on 0.5 m centers only within the eight 2.25 m^ sampling plots. A planting unit consisted of 15 shoots of eelgrass tied together and anchored in the substrate with metal pins (Fonseca et al. 1985). Eelgrass transplantation was performed in Sep- tember 1985. Eelgrass cover, shoot addition, and seedling recruitment were monitored periodically in the transplanted treatments. Shoot density and cover in the natural meadow were monitored simul- taneously with eelgrass seedling recruitment into unplanted areas by surveying randomly chosen plots within each treatment type. A 1.5 m x 1.5 m frame subdivided with cords into 36, 0.25 x 0.25 m (0.063 m^) sections was laid down, marking the perimeter of the plots. In the natural meadow and unplanted areas, three randomly selected 0.063 m- sections were surveyed within each plot for the number of eelgrass shoots and seedlings. In transplanted plots, the intersections of alternate cords fell on the 16 eelgrass planting units per plot. Three of these were randomly chosen and the number of shoots and area of bottom covered were recorded for each planting unit. To obtain the coverage estimate, a smaller grid with cords on 5 cm intervals was placed over the planting, and the number of squares (0.0025 m^) and half squares with eelgrass shoots were summed as area covered by the planting unit. Bay scallops were collected from eelgrass beds to the southwest of the study site using a commercial scallop dredge and were held in tanks supplied with continuously flowing seawater. Seventy-five scallops (size range from umbo to lip, 6.0-7.5 cm) were marked with waterproof pens to denote the number of the individual and its block assignment. Addition- ally, we cut small notches in the shell ridges with 191 FISHERY BULLETIN: VOL. 87, NO, 1 Block 1 Nl MM MM M^' M^' Natural Seagrass 7.5 m HPA B B HPA LPA Unvegetated Sand LPA HPA LPA B HPA B LPA HPA LPA B TREATMENTS (6 (^ S (^ PA= LOW PERIMETER/AREA RATIO /^ CT~\ f"^ /-^ (all planted) <^ K±J KJJ i^J ch & (£> &) (§) S (3) (^ Nl= NATURAL INTERIOR NE= NATURAL EDGE L HPA^HIGH PERIMETER /AREA RATIO (plots planted) B = BARE, UNPLANTED PLANTING UNIT Figure 2.— Map view of the study site with the dredge island to the northeast. Each treatment is 7.5 m on a side and each plot is 1.5 m on a side. a fine-tooth hacksaw blade to indicate the type of treatment. The first placement of bay scallops was done dur- ing high tide on 20 February 1986. Based on natural scallop density surveys conducted in November 1985 showing densities of ~2.0 scallops/m-, we stocked 5 bay scallops/2.25 m- plot. The 5 scallops were in- dividually placed next to 5 randomly selected plant- ing units out of the 16 in the plot. Thirty-four days after deployment, scallop surveys were conducted over four days from 26 to 30 March 1986 (survey I). The survey was conducted by placing a 7.5 m x 7.5 m grid made of Ys" nylon line, subdivided into 25 sections (2.25 m^ each) over an experimental unit. Bay scallops were located by systematically searching the substrate and grasses by sight and touch while snorkeling. Within each 2.25 m^ sec- tion, the efficiency of this method in recovering bay scallops <15 mm was untested, but recovery of bay scallops in the size range that was marked was 100% in three separate field trials. All unmarked bay scallops were measured to the nearest 0.1 cm on site, recorded by the section in which they were found, and replaced after measuring. Marked bay scallops were identified and recorded in the same manner. Due to low recovery of marked bay scallops from the transplant and bare areas over the 34-38 d period, a second survey (survey II) was initiated which excluded the natural seagrass beds. This sec- ond set of scallops was identified with waterproof pens and notching, but both shells were notched in the event the shells became separated after death. Forty-five bay scallops, five in each of nine plots, were released in LPA, HPA, and B treatments in blocks 1,3, and 5, but not in any natural treat- ments, on 30 March 1986 and surveyed 8 days later. RESULTS General observations during February through April revealed a variety of shorebirds, especially laughing gulls, brown pelicans {Pelecanus occiden- talis), and cormorants {Phalacrocorax olivacev^), frequenting the dredge island. Seagulls dropped 192 SMITH ET AL.: TRANSPLANTED EELGRASS FOR BAY SCALLOPS mollusc shells onto the island repeatedly to fracture the shell and feed on the contents. A marked shell from the HPA treatment (survey II, block 1) was found in the intertidal zone of the island, and one from a natural edge plot in survey I was found at a high, central point on the island during a random- ized search of the island. Other potential predators, blue crabs {Callinectes sp.. A'' = 3) and whelks (Busycon sp., A'' = 12), also were observed in the grassbeds during the surveys. Eelgrass cover and density in the natural meadow remained relatively constant throughout the study period. Natural bed experimental units had a con- sistent 77% cover, while shoot densities ranged between 441 and 1,148 shoots/m', with an average of 635 shoots/m- over the time between 20 Feb- ruary and 7 June 1986. Seedlings of eelgrass were observed among the natural and transplanted eelgrass in late March and early April. No eel- grass seedlings were recorded in the randomly chosen unplanted plots, although some were ob- served nearby. Throughout this time, transplanted treatments generally increased in number of shoots and area covered. By early June 1986, planting units averaged 0.02 m^, or approximately 15 cm in diameter with an average of 25 shoots/planting unit. After 34-38 days (survey I), 18 of 75 marked bay scallops (24%) were recovered (Fig. 3) and all were located in the plot in which they had been deployed. Fifteen of these 18 bay scallops were recovered in the natural grassbeds, with 9 located in the natural interior (NI) treatments and 6 in the natural edge (NE) treatments. Of the three remaining scallops, two were found in HPA treatments and one in a B, unplanted treatment. Three scallops were recovered from block 1 (farthest from the dredge island), 8 from block 3 (intermediate), and 7 from block 5 (closest). A total of 207 unmarked, naturally occurring bay scallops were counted and measured during survey I (Fig. 4). There were 77 from the natural interior, 119 from the natural edge, 3 from LPA, 6 from HPA, and 2 from B, unplanted area treatments. One hundred and twenty-five bay scallops were found in block 1 (farthest from land), 50 in block 3, and 32 in block 5. Our second, shorter survey recovered 10 out of the 45 (22%) bay scallops deployed in the trans- planted grassbeds and B, unplanted areas (Fig. 3). Five of those recovered were located in LPA areas, 4 in HPA, and 1 in a B treatment. Five scallops were found in block 1, 2 in block 3, and 3 in block 5. DISCUSSION The greater recovery of marked as well as un- marked, naturally occurring bay scallops from the natural beds as compared to the transplanted and bare areas (Figs. 3, 4) indicated that natural bed treatments provided a more suitable habitat for adult bay scallops. Bay scallops in the transplanted areas apparently suffered a higher mortality than occurred in denser, natural vegetation as suggested by the low recovery of marked scallops and our ob- servations of seabird predation. None of the bay scallops deployed in the transplants or bare areas were found in the natural beds, although in some instances the natural bed was only a few meters dis- tant. The few scallops recovered from these trans- plant and bare treatments were found in the plot of their deployment. Either there was little move- ment of the deployed bay scallops, and they were preyed upon, or the ones that moved were preyed upon. Whichever the mechanism of loss, it was ap- parent that few survived the 34 d deployment in these treatments. Neither treatment (LPA, HPA) of 5-6 mo old transplanted areas or bare areas provided the same habitat resource as adjacent, natural grassbeds (survey I); transplants did, however, provide a slight- ly better habitat for adult bay scallops than bare, unplanted areas over a short time (results from survey II). Twenty-two percent of the marked bay scallops were recovered from the transplant and bare treatments in survey II (8 day) deployment as opposed to 7% over the same area in survey I (34 days), suggesting a steady decline in numbers as a function of time. The extensive dense vegetation of the natural beds likely provides better refuge from predators such as gulls or blue crabs, along with in- creased protection from physically disruptive fac- tors such as wave action. Recovery of marked bay scallops from the treat- ment areas could not be attributed to the distance from the dredge island (Fig. 3). In survey I, the number of marked bay scallops recovered decreased with distance from the island, while in survey II, the opposite was observed. Distances from the island may not have been great enough to record a notic- able difference in seabird predation upon adult bay scallops as a function of distance. The natural scallop population, however, did demonstrate a fivefold in- crease in numbers with increasing distance from the dredge island (Fig. 4). There is no bottom elevation gradient across this distance. Tidal flow and wave energy patterns around dredge island conceivably could interfere with recruitment of water-borne 193 FISHERY BULLETIN: VOL. 87, NO. 1 SURVEY I Block 1 MARKED SCALLOPS 3 ' y^y Natural Seagrass SURVEY I % Recovered Block Dredge Island 1 3 5 Nl 40 60 80 NE 100 20 LPA HPA 20 20 B 20 MARKED SCALLOPS SURVEY II LPA SURVEY II % Recovered Dredge Island 1 DIOCK 3 5 LPA 40 60 HPA 60 20 B 20 Figure 3.— Distribution of recovered scallops as numbers per experimental unit (survey 56.25 m^). Survey I deployed 20 February 1986, and surveyed 30 March 1986. Survey II deployed 30 March 1986 and surveyed 7 April 1986. Five scallops were originally deployed in each plot. Treatment types: NI = Natural Interior, NE = Natural Edge, HPA = High Perimeter to Area, LPA = Low Perimeter to Area, B = Bare. scallop larvae or diminish food sources closer to the island, making recruitment and feeding, not preda- tion, a more likely factor influencing the existing natural scallop distribution. Natural seeding of the eelgrass, together with the transplanted treatments, should gradually provide more protection for adult bay scallops and greater amounts of vegetative cover for postveliger scallop attachment, but this coverage will not occur within the first year using transplanted eelgrass (Fonseca et al. 1985). Since eelgrass must be transplanted in the fall in North Carolina (Fonseca et al. 1985), the eelgrass transplants during the first year, will not be of a size to provide habitat functions equivalent to natural beds when bay scallop larvae settle in the late winter. There is, therefore, a substantial time interval in which eelgrass transplants in this area do not have scallop habitat value equivalent to natural beds. The creation of islands with dredge material in coastal waters may result in a reduction of bay scallop recruitment or survival within the area, as well as increasing bird predation by providing them with a substrate for dropping and opening scallops. 194 SMITH ET AL . TRANSPLANTED EELGRASS FOR BAY SCALLOPS NATURAL SCALLOPS SURVEY I Block 1 94.7% Natural Abundance ^S- Unvegelaled Sand Lp/^ ^^Q b[T] LPA 2 8 1 LPA hpaQ bQ Dredge Island Treatment * ( Scallops Nl 77 NE 119 LPA 3 HPA 6 B 2 Block of Scallops 125 50 32 Figure 4.— Distribution of natural scallops as number per experimental unit survey (56.25 m^) on 30 March 1986 (survey I). Treatment types: NI = Natural Interior: NE = Natural Edge, HPA = High Perimeter to Area, LPA = Low Perimeter to Area. B. = Bare. Due to enhanced seabird predation, restoring eel- grass beds adjacent to these islands will likely not provide a suitable area for bay scallop stocking until the bed matures and coalesces. These results may not be widely applicable because our study focused on a single eelgrass-dredge island system over one scallop settlement season. However, it is apparent that the location and manner of dredge material disposal should be examined closely. Although shore- bird and seabird habitat was certainly enhanced by the creation of the dredge material island, there may be local environmental and economic impacts on the scallop population and its fishery, as well as other existing, soft bottom communities, even without direct destruction of the adjoining seagrass bed itself as evidenced by the gradient of scallop abundance away from the island. There are two major conclusions to be drawn from this study. First, if natural eelgrass meadows are destroyed and transplants are used as replacements for the lost habitat, it is essential to recognize that the transplants will not immediately function as the natural bed it replaced. The delay or lack of habitat replacement could permanently reduce the produc- tion of economically valuable fauna in the area if proper measures are not taken to insure that any removed or destroyed eelgrass is properly balanced with a functionally equivalent habitat replacement. Second, this study has shown that natural eelgrass beds at this site provided a substantially more suit- able habitat for scallops than the transplanted treat- ments, within the first 5-6 months after planting. Stocking of recently transplanted eelgrass beds with scallops as a means of restoring or enhancing that fishery cannot be supported by these data. ACKNOWLEDGMENTS We would like to thank David Colby, Ford Cross, David Meyer, Jud Kenworthy, Gordon Thayer, and two anonymous reviewers for their helpful review comments. Carolyn Currin, Deborah Shulman, and Vicky Thayer assisted with field collections, while Herb Gordy prepared the graphics and Jean Fulford typed the manuscript and revisions. LITERATURE CITED FONSECA, M. S., W. J. Kenworthy, and G. W. Thayer. 1988. Restoration and management of seagrass systems, a review. In D. D. Hook et al. (editors). The ecology and management of wetlands. Vol. II: Management, use and value of wetlands, p. 353-368. Timber Press, Portland, OR. Fonseca, M. S., W. J. Kenworthy, G. W. Thayer, D. Y. Heller, and K. M. Cheap. 1985. Transplanting of the seagrasses, Zostera marina and Halodule wrightii, for sediment stabilization and habitat development on the east coast of the United States. Army Engineers Waterways Experiment Station, Vicksburg, MS, Tech. Rep. EL-85-9, 49 p. Fonseca, M. S., G. W. Thayer, and W. J. Kenworthy. 1987. The use of ecological data in the implementation and management of seagrass restorations. In M. Durako. R. C. Phillips, and R. R. Lewis (editors). Proceedings of the 36th Annual Conference of the American Institute of Biological Sciences, p. 175-187. Fla. Mar. Res. Publ. No. 42. 195 GUTSELL, J. S. 1930. Natural history of the bay scallop. Bull. U.S. Bur. Fish. 45:569-632. Kenworthy, W. J.. G. W. Thayer, and M. S. Fonseca. 1988. The utilization of seagrass meadows by fishery organ- isms. In D. D. Hook et al. (editors), The ecology and man- agement of wetlands. Vol. I, Ecology of wetlands, p. 548- 560. Portland, OR. Kirby-Smith, W. E. 1970. Growth of the scallops, Argopecten irradians concen- triciis (Say) and Argopecten gihbus (Linne), as influenced by food and temperature. Ph.D. Thesis, Duke Univ., Durham, NO, 126 p. Kirby-Smith. W. W., and R. T. Barber. 1974. Suspension-feeding aquaculture systems: effects of phytoplankton concentration and temperature on the growth of the bay scallop. Aquaculture 3:135-145. Pearson. T. G., C. S. Brimley, and H. H. Brimley. 1959. Birds of North Carolina. Byrum Print. Co., Raleigh, 434 p. Peirson, W. M. 1983. Utilization of eight algal species by the bay scallop. FISHERY BULLETIN: VOL. 87, NO. 1 Argopecten irradians cmicentricus (Say). J. exp. mar. Biol. Ecol. 68:1-11. Thayer, G. W., S. M. Adams, and M. W. LaCroix. 1975. Structural and functional aspects of a recently estab- lished Zostera marina community, /n L. E. Cronin (editor), Estuarine research, vol. 1, p. 517-540. Acad. Press, N.Y. Thayer, G. W., M. S. Fonseca, and W. J. Kenworthy. 1985. Wetland mitigation and restoration in the southeast United States and two lessons from seagrass mitigation. In Estuarine management practices. Proceedings of the 2nd National Estuarine Research Symposium, Baton Rouge, 1985, p. 95-117. Louisiana Sea Grant, Baton Rouge. Thayer, G. W., W. J. Kenworthy, and M. S. Fonseca. 1984. The ecology of eelgrass meadows of the Atlantic coast: a community profile. U.S. Fish Wildl. Serv. FW5/OB5-84/ 02. 147 p. Thayer, G. W.. and H. H. Stuart. 1974. The bay scallop makes its bed of eelgrass. Mar. Fish. Rev. 36(7):27-39. 196 THE FOOD HABITS OF FIVE CRAB SPECIES AT PETTAQUAMSCUTT RIVER, RHODE ISLAND John W. Ropes^ ABSTRACT The stomach contents of five crab species— green crab, Carcinus maenas; blue crab, Callinectes sapidity; lady crab, Ovalipes oceilatus; mud crab, Neopanope texana; and spider crab, Libinia emarginata—viere examined from collections made in the Pettaquamscutt River, Rhode Island, during 1955-57. A car- nivorous food habit characterized all species, although spider crabs contained plant foods more often than animal foods. Mollusks (especially pelecypods) and arthropods were frequent dietary components of the green, blue, and lady crabs. Intense predation on small, recently set pelecypods was indicated. The three species of portunid crabs (green, blue, and lady) appeared to have similar food habits, sug- gestive of potential interspecific competition for food. Crab remains were most frequently encountered in blue crab stomachs; lady crabs contained this food more often than green crabs. Small Crustacea and plant foods occurred more often than hard-shelled foods and with equal frequency in the stomachs of small green crabs (<20 mm carapace width). Predation by crabs has been identified as a serious threat to successful management of commercial bivalve resources (Carriker 1967; R. N. Hanks 1963; R. E. Hanks 1969). Many studies have concentrated on the green crab, Carcinus maenas, because of its abundance, its extensive distribution in the coastal zone of northeastern United States and Canada and Europe, and its predation on bivalves, especially the soft-shelled clam, Mya arenaria. Ropes (1968) and Welch (1968) have provided extensive reviews of the U.S. literature on this species; Davies (1966) and Kitching et al. (1959) have reported on its effects on European or blue mussel, Mytiltcs edulis, culture. Blue crabs, Callinectes sapidus; lady crabs, Ovalipes oceilatus; and mud crabs, Neopanopeus texana, have also been found to be predators of bivalves (Ryder 1884; Hay 1905; Fowler 1911; Belding 1930; Anon. 1941; Lunz 1947; Turner 1948; Bulter 1954; Landers 1954; Dunnington 1956; Darnell 1958, 1959; McDer- mott 1960; Galtsoff 1964; Loosanoff 1965). Many of these studies described the relationship between a particular predator and prey. After completing collection and examination of green crab stomachs from Plum Island Sound, MA, I foimd that four of the species mentioned above and the spider crab, Libinia emarginata, could be col- lected from a fairly restricted area at the mouth of the Pettaquamscutt River, RI. This was an oppor- 'Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA, Woods Hole, MA 02543. [Deceased September 1988.] Manuscript accepted October 1988. Fishery Bulletin 87:197-204. tunity to examine possible inter- and intraspecific feeding habits by sympatric decapod crustaceans. The taxonomic relationship and morphological dif- ferences of the three portunid crabs (blue, green, and lady crabs) suggested making comparisons of stomach contents with each other and the two other crab species to determine the potential for preda- tion on bivalves and to observe possible similarities and differences in their diets. The impact of such predation on bivalves of commercial importance has practical implications for resource management. METHODS From 1955 through 1957, crabs were collected during daylight hours from three subtidal areas of Pettaquamscutt River, RI, (Fig. 1) by towing a scal- lop dredge from a 12 ft aluminum boat powered by an 18 hp outboard motor (see Ropes [1968] for a description of the dredge). Intertidal areas were limited by the sharply sloping marsh banks which could not be sampled by the dredge. Thus, tows were made subtidally over shoal bars, along the edges of bars in the channel, and near the banks of the river. All samples were taken during ebb tide and before low water because experience at Plum Island Sound had shown that green crabs were actively moving about at that time. In 1955, five collection trips were made in July, September, and October; in 1956, six trips were made from May through August; and in 1957, nine trips were made in August through Oc- tober (Table 1). At the laboratory, the species and 197 FISHERY BULLETIN: VOL. 87. NO. 1 7^28' 71'26' Figure 1.— Areas dredged (black areas) for crabs in Pettaquamscutt River, RI, 1955-57. Shoal bars were exten- sive near dredging areas. The islands were marshy (stippled areas) and were contiguous with mainland areas (dashed lines). An insert (bottom, right comer) shows Pettaquamscutt River in relation to the West Passage to Narragansett Bay. sex of each crab were determined, and the carapace width (in mm) of each crab was measured with ver- nier calipers. Stomachs were then removed and preserved in 10% formalin. Food items were iden- tified using a stereoscopic microscope, and the fre- quency of occurrence of each item was recorded. For some stomachs it was possible to count individual bivalves. Descriptions of the amount of food in the stomachs were as follows: 1 ) full stomachs, containing tissues and hard parts of foods, 2) nearly empty stomachs, containing only a few fragments of hard parts of foods, and 3) empty stomachs. The stomachs in the first category were tabulated as the percentage of all crabs in a sample, and those in the second cate- gory as a percentage of the stomachs containing 198 ROPES: FOOD HABITS OF FIVE CRAB SPECIES Table 1. — Numbers of crabs caught in dredge tows at Pettaquamscutt River, Rl, 1955-57. Number of Callinectes Carcinus Ovalipes Neopanopeus Libinia Date tows sapidus maenas ocellatus texana emarginata 1955 3 July 11 12 33 6 'A 6 Sept. 5 97 27 8 Sept. 9 14 26 14 8 3 Oct. 4 6 12 17 4 Oct. 4 18 1956 2 May 4 16 4 June 7 25 2 '5 5 June 6 22 1 1 3 July 7 5 1 7 30 July 8 37 3 '6 29 Aug 7 58 1957 5 Aug. 3 14 M 6 Aug. 3 30 '1 8 Aug. 3 19 15 Aug. 6 10 2 23 Aug. 5 62 24 1 3 4 Sept. 2 '2 20 M 25 Sept. 4 19 27 5 7 14 Oct. 3 '2 29 31 Oct. 3 65 'Stomachs not examined. food. Empty stomachs were omitted from all calcu- lations. Since most food consisted of crushed or frag- mented remains, hard structures were relied upon for food identification. To assign a food to a definite species category was not always possible, but it could usually be included in a general taxonomic group. Thus, not all major taxonomic groups are the sum of specific items. Mollusk shells, and especial- ly the hinge structure of pelecypods, could be most readily recognized. Annelids were identified by their jaws (whole worms were rarely found). Arthropods could rarely be identified to species, but were separated into three general groups: 1) crabs, which consisted of heavy pigmented exoskeletal remains, 2) small crustaceans, which consisted of translucent exoskeletal remains of amphipods, isopods, and small shrimp, and 3) barnacles (shells and bodies). Many stomachs contained food re- mains that were too fragmented or digested for identification. These were classified as unidentified remains. Analysis of food habits by sex, molting condition, egg bearing, and mating were not possible due to the small numbers of each species collected. In a study of nearly 4,000 green crabs, Ropes (1968) found that feeding habits by sex were inconsistent and that feeding habits by premolt and very soft- shelled crabs (not included in the present study) were arrested. Similarly, o\'igerous green crabs ex- amined by Ropes (1968) tended to feed less than nonovigerous crabs, and the stomachs of mated pairs were nearly empty. Thus, analyses in the pres- ent study focused only on general food habits of the five species. RESULTS Although a few blue crabs were caught in 1955, none were taken in 1956 (Table 1). They were most numerous in the 1957 samples. The mean number per tow varied from to 12.4. Green crabs were caught during every collection trip, except 8 August 1957, and were usually more numerous than the other crab species. The 6 September 1955 collection was unusual: 90 green crabs <10 mm CW (carapace width), 6 crabs <20 mm, and 1 crab 58 mm were caught entangled in decaying algae. It was the only collection containing such a large number of small- sized crabs. The mean number of green crabs caught per tow varied widely (0 to 21.7). Other crab species were caught infrequently and then in relatively low numbers. Blue crabs ranged from 20 to 160 mm CW (Fig. 2). Although none exceeded 105 mm in 1955, 35% were larger in 1957. No size group was dominant; most ranged from 40 to 109 mm. Green crabs ranged from 4 to 70 mm, with most between 20 199 FISHERY BULLETIN: VOL. 87, NO. 1 c allinecte s sapidi js N = 190 20 1 1 1 1 r 10 f f 1 i 120 1 - Carcinus maenus N=5 1 8 80 I 12 8 . Neopanopeus texana N=56 ^ < 4 [— ^ or o 40 , r^ ^ o u 2 4 6 8 10 IZ NUMBER O o Ovalipes ocellat us N-3 1 L— ^ , , A Libinia emargina ta N=30 2 _ I 1 1 n 1 . , 1 20 40 60 80 100 120 140 160 CARAPACE WIDTH (mm) Figure 2.— Frequency distribution of carapace widths of five crab species collected from Pettaquamscutt River, RI, 1955-57. and 59 mm. Lady crabs ranged from 27 to 83 mm CW, with most <50 mm. Spider crabs ranged from 13 to 93 mm and mud crabs from 3 to 12 mm. Blue Crab Animal foods were found in 91% of the blue crab stomachs and plant foods in 19% (Table 2). This in- dicated a predominantly carnivorous food habit. Of the animal foods, arthropods were more frequently (68%) encountered than moUusks (55%) or annelids (17%). Crabs (40%) and small crustaceans (33%) eaten by blue crabs were examined. Pelecypods (49%) were found more often than gastropods (12%), while the gem clam, Gemma gemma, (28%) and blue mussel (16%) occurred most often. The number of gem clams in 43 stomachs ranged from 2 to 821 (with an average of 66.8 per stomach). Few whole mussels had been eaten (<4 per stomach), but most of them occurred as shell fragments. Other pelecy- pods found infrequently were soft-shelled and hard- shelled clams, and glossy shell (possibly Te.llina sp.) and ribbed shell (possibly Argopeden sp.) fragments. Gastropods were found infrequently and, except for Hydrobia sp., were incomplete, broken shells or operculi. The jaws of Nereis sp. found in 13% of the stomachs indicated that it was the most frequently eaten annelid. Fish remains were found in 18% of the stomachs. Of the plant foods, Spartina occurred in 12% of the stomachs and algae in 8%. Uniden- tified tissues were in 67% of the stomachs. Although 75% of the blue crab stomachs contained food, 24% of these were nearly empty. 200 ROPES: FOOD HABITS OF FIVE CRAB SPECIES Table 2.- -Occurrence and (percent frequency of occurrence) of foods eaten by five species of crabs collected in Pettaquamscutt River, Rl, 1955-57. Callinectes Carcinus Ovalipes Neopanopeus Libinia Food items sapidus maenas ocellatus texana emarginata Animal 126(91) 240 (80) 17(94) 16 (42) 5(42) Annelids 24(17) 48 (16) 2(11) 1 (3) Nereis sp. 18 (13) 37 (12) 2(11) Ottier 1 (1) 3 (1) IVIollusks 77 (55) 130 (43) 14 (78) 1 (3) 3(25) Pelecypods 68 (49) 92(31) 13 (72) 1 (3) 3(25) Mytilus edulis 22 (16) 72 (24) 10 (56) 1 (3) 2(17) Gemma gemma 39 (28) 4 (1) 3(17) Mercenaria mercenaria 3 (2) Mya arenaria 1 (1) Ottier 12 (4) 4(22) 1 (8) Gastropods 16(12) 48 (16) 1 (6) 1 (8) Hydrobia sp. 6 (4) 1 (1) Ottier 9 (7) 34(11) 1 (6) 1 (8) Arttiropods 94 (68) 106 (35) 9(50) 13 (34) 1 (8) Crabs 55 (40) 11 (4) 5(28) 1 (3) 1 (8) Small crustaceans 46 (33) 75 (25) 3(17) 12 (32) 1 (8) Barnacles 1 (1) Foraminifera 2 (1) 20 (7) Fisti 25 (18) 1 (1) 1 (3) Otfier 1 (8) Plant 26(19) 118 (39) 8 (21) 9(75) Algae 11 (8) 92(31) 6 (16) 8(67) Spartina sp. 17(12) 55(18) 3 (8) 3(25) Unidentified 93 (67) 250 (83) 4(22) 38 (100) 7(58) Stomactis witti food 139(75) 301 (71) 18 (58) 38 (69) 12 (67) Stomactis nearly empty 33 (24) 50(17) 1 (6) 3 (8) 3(25) Total number of crab stomactis examined 186 426 31 55 18 Green Crab Animal foods were found in 80% of the green crab stomachs; plant foods occurred in 39% (Table 2), This indicated a predominantly carnivorous food habit with herbivorous tendencies. Of the animal foods, moUusks were found more frequently (43%) than arthropods (35%) or annelids (16%). Pelecypods (31%) occurred more often than gastropods (16%). Of the pelecypods, blue mussels were found most often (24%); gem clams and other pelecypods were infrequent food items. Other gastropods occurred most often (1 1%) as unidentifiable broken shells or operculi. The jaws of Nereis were found in 12% of the stomachs. Of the plant foods, Spartina was found in 31% of the stomachs, and algae in 18%. Unidentified tissues occurred in 83% of the stom- achs. Although 71% of the green crab stomachs con- tained food, 17% of these were nearly empty. Stomach analyses of small green crabs (<20 mm CW) caught on 6 September 1955 were sufficiently different to warrant separate representation (Table 3). Animal and plant foods occurred with equal fre- quency (69%) in the stomachs. Of the animal foods, Table 3.— Occurrence and (percent frequency of occurrence) of foods eaten by small (<20 mm CW) green crabs caugtit on 6 September 1955. Carapace width (mm) Combined Food item <10 10-19 (Percent) Animals 47 (70) 3(60) 70 (69) Annelids Nereis sp. Ottier 1 (2) 1 (2) 1 (2) 1 (2) fvlollusks Pelecypods Mytilus edulis 1(2) 1(2) 1(2) 1 (2) 1 (2) 1 (2) Arttiropods Crabs Small crustaceans 43 (64) 1 (2) 41 (61) 3(60) 3(60) 46 (64) 1 (2) 44(61) Foraminifera 1 (2) (2) Plants Algae Spartina sp. 46 (69) 44(66) 2(3) 4(80) 4(80) 1 (20) 50 (69) 48 (67) 3 (4) Unidentified 25(37) 3(60) 28 (39) Stomachs with food 67 (74) 5(83) 72 (75) Stomachs nearly empty 6(8) 6 (6) Total number of crab stomachs examined 90 6 96 201 FISHERY BULLETIN: VOL. 87, NO. 1 arthropods were found most frequently (64%) and were predominantly (61%) small crustacean re- mains. The occurrence of the blue mussels, crab remains, and foraminifera in 2% of the stomachs in- dicated that these were minor dietary components. Algae was the dominant (69%) plant food; Spartina occurred in only 4% of the stomachs. Unidentified remains were found in 39% of the stomachs. Most stomachs (75%) contained food, although 6% of these were nearly empty. Lady Crab Animal foods were found in 94% of the lady crab stomachs; none contained plant foods (Table 2). This indicated a strictly carnivorous food habit. Of the animal foods, moUusks were encountered more often (78%) than arthropods (50%) or annelids (11%). Pele- cypods occurred much more often (72%) than gas- tropods (6%). Blue mussels found in 56% of the stomachs were usually shell fragments; gem clams were found in 17% of the stomachs, and one stom- ach contained as many as 17. Other pelecypods found in 22% of the stomachs were glossy-white fragments (possibly Tellina sp.). Arthropod foods encountered were crab (28%) and small crustacean (17%) remains. The jaws of Nereis were found in 11% of the stomachs. Unidentified tissues occurred in 22% of the stomachs. Although 58% of the lady crab stomachs contained food, 6% of these were nearly empty. Mud Crab Animal foods were found in 42% of the mud crab stomachs; plant foods were in 21% (Table 2). This indicated a predominantly carnivorous food habit with herbivorous tendencies. Of the animal foods, arthropods were more frequently (34%) encountered than mollusks (3%) or annelids (3%). Small crusta- ceans were found in 32% of the stomachs, and crabs in 3%. Blue mussels were found in only 3% of the stomachs and none contained gastropods. Fish re- mains were found in only 3% of the stomachs. Of the plant foods, algae was found in 16% of the stomachs and Spartina in 8%. All of the stomachs with food contained unidentified tissues. Although 69% of the mud crab stomachs contained food, only 8% of these were nearly empty. Spider Crab Animal foods were found in 42% of the spider crab stomachs and plant foods in 75% (Table 2). This in- dicated a predominantly herbivorous food habit with carnivorous tendencies. Of the animal foods, mol- lusks occurred more often (25%) than arthropods (8%). Pelecypods were found more often (25%) than gastropods (8%). Blue mussels occurred in 17% of the stomachs. None of the stomachs con- tained annelids. Of the plant foods, algae was found in 67%, and Spartina was found in 25% of the stomachs. Unidentified tissues were found in 58% of the stomachs. Although 67% of the spider crab stomachs contained food, 25% of these were near- ly empty. DISCUSSION Food habits of the five crab species were generally similar, a probable reflection of prey availability. Blue, green, and mud crabs tended to be carnivorous and spider crabs tended to be herbivorous, while lady crabs were observed to be exclusively carniv- orous. However, neither of the latter two species was well represented (Table 2). Mollusks and arthro- pods were frequent dietary components of the blue and lady crab specimens (50% to 78%); such foods were in 43% and 35% of the green crab stomachs (Table 2). Many of the lady crabs (72%) and blue crabs (49%) contained pelecypods, but only 31% of the green crabs had eaten this food. Green crabs are the sole portunid in the decapod fauna of Plum Island Sound, and Ropes (1968) found mollusks in 75% of the stomachs examined, with pelecypods (68%) the most frequent type of moUusk eaten. At Pettaquamscutt River, the low occurrence of pelecypods in green crab stomachs suggested that interaction between portunid species may have been affecting their feeding habits. Blue and lady crabs are adept at swimming and may use this ability in obtaining food and avoiding conflicts over food items; green crabs may be at a disadvantage in com- peting for food by their relatively poor swimming abOities. Many blue crabs were larger than the green or lady crabs taken, and their powerful claws may have been of positive advantage in encounters with other crabs. The high occurrence (40%) of crab re- mains in blue crab stomachs suggests inter- or intra- specific predation occurred, although the frag- mented remains did not allow identification to species. Lady crabs may also have exerted predator pressure because 28% of their stomachs contained crab remains. Crab predation by green crabs was the lowest (4%) for the portunids examined and was lower than reported by Ropes (1968) at Plum Island Sound (13%). Small green crabs caught on 6 September 1955 202 ROPES: FOOD HABITS OF FIVE CRAB SPECIES were almost equally carnivorous and herbivorous (Table 3). Relatively soft-shelled, small crustaceans (61%) and algae (67%) were found in the stomachs of green crabs; while only 2% of hard-shelled foods, such as pelecj'pods and crabs, were found. The mat of algae that entangled the crabs probably provided the foods seen in the stomachs and shelter from large predatory crabs and fishes. Ropes (1968) observed finger-sized holes, mats of algae, and Spar- tina high on the marsh banks of Plum Island Sound that were a refuge for small green crabs. This may be a means of circumventing cannibalism, because large green crabs typically occurred on the lower level clam flats during high tide and migrated to the subtidal zone during low tide. Ropes (1968) also found that soft foods, such as Spartina and algae, were important dietary components of small crabs. The omnivorous food habit and existence of small crabs in ecological niches apart from large predators has probable survival value. Stomach analyses of the five crab species indicate that their food habits have probable important im- pact on the macrobenthic fauna of Pettaquamscutt River; these results support similar findings of other investigators of the food habits of crabs. The omniv- orous food habit of the crabs has survival value, minimizing their dependency on particular food items. The carnivorous habit of blue, green, and lady crabs, which have a tendency to include many small pelecypods in their diet, suggests that recently settled pelecypods may be rapidly eliminated or severely reduced in numbers by predation. Clearly, a management scheme that minimizes the effects of crab predation on a bivalve fishery has a better potential for success. ACKNOWLEDGMENTS I am indebted to helpful review of the manu- script by Mark L. Botton, Fordham University, N.Y. and Ray E. Bowman and William L. Michaels of the Northeast Fisheries Center Woods Hole Laboratory, National Marine Fisheries Service, NOAA. LITERATURE CITED Anonymous. 1941. Atlantic Coast blue crab an important enemy of oysters. Oyster Inst. N. Am., Trade Rep, 44, 2 p. Belding, D. L. 1930. The soft-shell clam fishery of Massachusetts. Com- monw. Mass.. Dep. Conserv., Div. Fish Game, Mar. Fish. Ser. No. 1, 65 p. Butler, P. A, 1954. A summary of our knowledge of the oyster in the Gulf of Mexico. In P. S. Galtstoff (editor), Gulf of Mexico, its origin, waters, and marine life, p, 479-489. U.S. Fish Wild]. Serv,, Fish, Bull, 55 [No. 89]. Carriker. M, R. 1967. Ecology of estuarine benthic invertebrates: a perspec- tive. In G. A. Lauff (editor). Estuaries, p. 442-487. Am. Assoc. Adv. Sci., Wash,. D.C, Publ, No. 83. Darnell, R. M. 1958. Food habits of fishes and larger invertebrates of Lake Pontchartrain, Louisiana an estuarine community. Inst. Mar, Sci., Univ. Texas 5:353-416. 1959. Studies of the life history of the blue crab (Callinectes sapidm Rathbun) in Louisiana waters. Trans, Am, Fish. See. 88:294-304. Davies, G. 1966, Shore crabs, Carcinus maenas, and mussel culture. Rep, Challenger Soc, 3:31, Dunnington, E. A. 1956. Blue crabs observed to dig soft shell clams for food. Md, Tidewater News 12(12):1, 4. Fowler. H, W, 1911, The Crustacea of New Jersey. Ann. Rep. N.J. State Mas. p. 29-650. Galtsoff. P, S, 1964. The American oyster Crassostrea virginica Gmelin. U.S. Fish WUdl. Serv,, Fish. Bull, 64:1-480. Hanks, R. N, 1963. The soft-shell clam. U.S. Fish Wildl. Serv,, Cir. 162. 16 p. Hanks, R, E, 1969. Soft-shell clams. In F. E, Firth (editor), The en- cyclopedia of marine resources, p, 112-119, Van Nostrand Reinhold Co,. N,Y„ Hay, W. P, 1905. The life history of the blue crab. Callinectes sapidus. Bull [U.S.] Bur. Fish. 24:395-413. Doc. 580. Kitching, J. A,, J. F. Sloane, and F. J. Ebling. 1959. The ecology of Lough Ine, 8. Mussels and their predators. J. Anim. Ecol. 28:331-341. Landers, W, S, 1954. Notes on the predation of the hard clam, Venvs mercenaria, by the mud crab, Neopanope texana. Ecology 35:442, Loosanoff, V, L, 1965. The American or Eastern oyster. U.S. Fish Wildl, Serv., Cir. 205, 36 p, LUNZ, G. R. 1947. Callinectes versus Ostrea. J. Elisha Mitchell Soc. 63:81. McDermott. J, J. 1960. The predation of oysters and barnacles by crabs of the family Xanthidae. Proc. Pa. Acad. Sci. 34:199- 211. Ropes, J. W. 1968. The feeding habits of the green crab. Carcinus mae- nas (L.). U.S. Fish Wildl. Serv., Fish. Bull. 67:183- 203. Ryder, J. A. 1884. A contribution to the life-history of the oyster. In G. B. Goode (editor), The fisheries and fishery industries of the United States, Sec. 1. 4(10):711-758. Turner, H, J,, Jr. 1948. Appendix I. The soft-shell clam industry of the east 203 FISHERY BULLETIN: VOL. 87, NO. 1 coast of the United States. In H. J. Turner, Jr. (editor), Welch, W. R. Report on investigations of the propagation of the soft-shell 1968. Changes in abundance of the green crab, Careinus c\?Lms,Myaarenaria,p. 11-39. Div. Mar. Fish., Dep. Con- maenas (L.), in relation to recent temperature changes, serv., Commonw. Mass. U.S. Fish Wildl. Serv., Fish. Bull. 67:337-345. 204 STOCK IDENTIFICATION OF WEAKFISH, CYNOSCION REGALIS, IN THE MIDDLE ATLANTIC REGION Maurice K. Crawford/ Churchill B. Grimes, ^ and Norman E. Buroker' ABSTRACT The hypothesis that a single stock of weakfish. Cynoseiun regalis, existed in the Middle Atlantic was tested. Using starch gel electrophoresis we identified two polymorphic loci (6-phosphogluconate dehy- drogenase and malate dehydrogenase) out of a total of 25 protein loci surveyed. Statistical analysis of allelic frequencies revealed that the populations were statistically indistinguishable. Weakfish, Cynoscion regalis, commonly reach sizes between 70 and 80 cm in length and 3.0 and 4.5 kg in weight. They occur from Cape Cod, MA to Florida but are most common in the Middle Atlantic region (Wilk 1979). Weakfish participate in a spring spawn- ing migration into bays and estuaries. In the fall, the migrations reverse and fish move either offshore or to more southern waters to overwinter (Welsh and Breder 1923; Bigelow and Schroeder 1953; Wilk 1979). Spawning occurs from May to mid-July in northern estuaries (e.g., Delaware Bay and Gardi- ners Bay, NY; Shepherd and Grimes 1984) and from March to September in more southern waters (e.g.. North Carolina; Merriner 1976). Weakfish are an important commercial and rec- reational species and historically landings have fluc- tuated widely. From 1940 to 1949, commercial land- ings averaged 8,800 metric tons (t), with a high of 18,800 t in 1945. Between 1950 and 1969, annual catches declined to an average of 2,600 t, but a resurgence occurred when the catches rose to a 7,700 t average between 1970 and 1979 (Wilk 1981). Recreational landings of weakfish have been sim- ilarly variable, and in some years have been esti- mated to be as large as the commercial landings 'Department of Horticulture and Forestry, Cook College, Rut- gers University, New Brunswick, NJ 08903; present address: School of Forest Resources, University of Georgia, Athen. GA 30602. ^Department of Horticulture and Forestry, Cook College, Rut- gers University, New Brunswick, NJ 08903; present address: Southeast Fisheries Center Panama City Laboratory, National Marine Fisheries Service, NOAA, 3500 Delwood Beach Road. Panama City, FL 32408. Reprint requests should be sent to the second author (CBG) at the present address. 'Department of Zoology, Rutgers University. New Brunswick, NJ 08903; present address: Department of Pediatrics RR20. School of Medicine, University of Washington, Seattle, WA 98195. (Murawski 1977). In 1965, catches only amounted to 1,000 t but increased to 7,100 t in 1970 (Wilk 1981). In 1974, recreational landings were approx- imately 9,100 t, or about 60% of the estimated total catch (Murawski 1977). Landings dropped to 5,000 t in 1979, and Middle Atlantic states accounted for 95% of the catch (Wilk 1981). Several studies (Nesbit 1954; Perlmutter et al. 1956; Seguin 1960) have concluded that there were multiple stocks of weakfish in the Middle Atlantic region based upon mark recapture, scale circuli spacing, and morphological data, respectively. More recent studies have shown geographic differences in growth and reproduction of weakfish between Cape Cod, MA and Cape Hatteras, NC (i.e., north- ern fish lived longer, grew larger, and had a lower relative fecundity than southern fish; Shepherd and Grimes 1983, 1984). These life history differences could be due to environmental effects or could be indicative of discrete stocks (Shepherd and Grimes 1983). We hypothesized that a single panmictic pop- ulation of weakfish exists in the Middle Atlantic region. In order to test this hypothesis, starch gel electrophoresis was used to identify protein varia- tion for two polymorphic structural loci (malate dehydrogenase-2 and 6-phosphogluconate dehydrog- enase) found among weakfish in this region. MATERIALS AND METHODS We sampled adult and juvenile (young-of-year) weakfish along the east coast of the United States from Buzzards Bay, MA to Cape Hatteras, NC (Fig. 1). Adult fish were caught in the fall of 1982 and summer of 1983 by the National Marine Fisheries Service bottom trawl survey cruises (Grosslein 1969). We also purchased adults in some locations Manuscript accepted October 1988. Fisherj- Bulletin, U.S. 87:205-211. 205 FISHERY BULLETIN: VOL. 87, NO. 1 Buzzards Bay- Middle Atlantic Region • -Juveniles -Adults Cape Cod 700 km Figure 1.— Sampling localities of adult and juvenile weakfish sampled in the Middle Atlantic Region. 206 CRAWFORD ET AL.: STOCK IDENTIFICATION OF WEAKFISH (i.e., Belford, NJ, the York River, VA and Cape Hat- teras, NC) during the spring of 1982 and 1983. We collected juvenile fish in their natal estuaries to minimize possible effects due to mixing after the fish migrated. Fish were trawled in Delaware Bay, NJ, Chesapeake Bay, VA, and Pamlico Sound, NC from August to October 1981 and 1982. We mea- sured fish to the nearest mm total length (TL) and determined sex (when possible). Extracts of eye, liver, and skeletal muscle tissue were separately placed in centrifuge tubes and stored on dry ice. When this was not possible, we froze whole fish on ice and later removed the tissues in the laboratory. All tissue samples were stored at -8°C until anal- ysis (for electrophoretic details see Crawford 1984). Electromorph banding patterns were interpreted based upon the protein's subunit structure and pre- vious studies with homologous enzymes. (Utter et al. 1974; Harris and Hopkinson 1976). We numbered loci from the anode to the cathode in ascending order. Allozymes were measured in millimeters relative to the most common homomeric electro- morph which was designated 100 and numbered ac- cordingly. The bands exhibited v/ere consistent with reported information on their molecular structure (Manwell and Baker 1970). Allelic frequencies of polymorphic systems were calculated and examined for conformance to Hardy- Weinberg expectations (HWE) using Levene's (1949) method for small sam- ple sizes {N < 100). For polymorphic loci that had null alleles the statistical procedures of (Speiss 1977) were followed to test for HWE. Sampling localities were compared by using a chi-square contingency test for haploid frequencies (Speiss 1977). We tested juvenile and adult allelic frequencies of polymorphic loci (excluding rare and null alleles) to determine whether significant differences in gene frequencies existed among 1) geographic location, 2) size/age (i.e., adults vs. juveniles) groups, and 3) sexes. Sampling locations were tested within regions, and if there were no significant differences among sam- pling locations, they were pooled. Pooled samples were then compared with all other regions to deter- mine if there were regional differences. We calcu- lated (averaged over the polymorphic loci) the genetic variation among the samples (F^i) (Hartl 1980). We obtained the percent polymorphic loci (common allele (p) < 0.950) and genetic distances (Nei 1972) using a BASIC computer program by Green (1979). RESULTS Weakfish from four populations (Long Island Sound, NY, Delaware Bay, NJ, York River, VA, and Cape Hatteras, NC) were initially screened by starch gel electrophoresis using 15 protein staining systems (Table 1) to identify polymorphic loci and calculate genetic distances. In samples obtained from other collecting localities only the polymorphic loci were evaluated. The activity of the 15 enzyme (protein) staining systems was interpreted to reflect 25 structural loci of which only two (8%) were polymorphic, 6-phos- Table 1.— Proteins and gel buffer-tissue combinations that provided the best resolution. #of Protein systems E.C.' Tissue Buffer loci Alcohol dehydrogenase 1.1.1.1 Liver 1 1 Aspartate aminotransferase 2.6.1.1 Muscle IV 2 Esterase 3.1.1.1 Muscle IV 2 o-Glycerophosphate dehydrogenase 1.1,1.8 Muscle II 1 Glycerate dehydrogenase 1.1.1.29 Muscle II 2 Isocitrate dehydrogenase 1.1.1.42 Muscle 1 1 Lactate dehydrogenase 1.1.1.27 Eye II 3 Malate dehydrogenase 1.1.1.37 Muscle 1 3 Malic enzyme 1.1.1.40 Muscle 1 1 Muscle protein (nonspecific) Muscle II 3 Phosphoglucomutase 2.7.5.1 Eye III 1 6-phosphogluconate dehydrogenase 1.1.1.44 Muscle 1 1 Phosphoglucose isomerase 5.3.1.9 Eye III 2 Sorbitol dehydrogenase 1.1.1.14 Muscle II 1 Xanthine dehydrogenase 1.2.3.2 Liver IV 1 'Enzyme Commission number I . Aminopropyl, pH 6.0 (Clayton and Tretiak 1972) II - Tris Citrate. pH 6,8 (Shaw and Prasad 1970), III - Tris Versane Borate, pH 8 (Shaw and Prasad 1970) IV - Ridgway, pH 8 5 (Ridgway et ai 1970) 207 FISHERY BULLETIN: VOL. 87, NO. 1 phogluconate dehydrogenase (Pgd) and malate dehy- drogenase (Mdh-2). In addition, three other loci— aspartate aminotransferase (Aat), phosphoglucose isomerase (Pgi), and xanthine dehydrogenase (Xdh)— exhibited rare alleles (i.e., common allele (p) > 0.950). SLx-phosphogluconate dehydrogenase pro- duced a single zone of allozyme activity on the starch gel that we interpreted as the product of a single gene locus. The heterozygotes displayed three bands which is typical of this molecule's dimeric structure (Manwell and Baker 1970). In weakfish, Pgd ex- hibited three alleles designated as 100, 98, and 96, a rare allele. Both juveniles and adults had similar frequencies of the most common allele (Table 2). At the Mdh-2 locus, a dimeric protein product formed heteropolymers with the products of other Mdh loci. These heteropolymers occurred between the products of Mdh-1 and Mdh-2, and Mdh-1 and Mdh-3. The Mdh-2 locus was associated with liver, and the Mdh-3 locus is thought to be expressed in mitochondria (Thorne et al. 1963). This enzyme system also displayed a fourth isozyme band that migrated cathodally. This Mdh isozyme band is not reported in other similar studies and we do not know what protein loci it represented (Fig. 2). The Mdh-2 locus was polymorphic and exhibited four alleles: 103, 100, 97 (a rare allele) and a fourth null allele {Mdh-2 {N)). Two fish homozygous for Mdh-2 (N) were found. One was in a sample of juvenile fish from Spencer's Bay, NC and the other in an adult from Chesapeake Bay. The frequencies for the most common allele are found in Table 2. The uneven sam- ple sizes (Table 2) occurred because of protein de- naturation. The denatured samples indicated by streaks in the gels were excluded from analysis. Allelic frequencies of three samples differed sig- nificantly from HWE for Mdh-2 (Table 2). This deviation reflected a deficiency in the number of heterozygotes which may have been due to the pres- ence of the null allele (Selander 1970; Speiss 1977). We estimated null allele frequencies from the square root of the phenotype for Spencer's Bay (0.151) and Chesapeake Bay (0.146). Using the mean of the two values to estimate the null allele frequency for Table 2.— Allelic frequencies of juvenile and adult fish for Pgd and Mdh including sample size, frequency of the most common allele and the standard error. All samples were collected in 1982 unless otherwise noted. Pgd{100) Mdh-2(100) Location N Frequency (SE) W Frequency (SE) Juveniles' Northern Region Buzzard's Bay and Long Island. NY 125 0580(0.083) 98 0.576(0.041) Sandy Hook, NJ 53 0.632(0.047) 47 0.617(0.050) Delaware Bay, NJ (1981) 49 0.541(0.050) 46 0.511(0.052)** Delaware Bay, NJ 91 0.527(0.037) 101 0.584(0.035) Chesapeake Bay Region James R., VA 38 0.605(0.056) 48 0.510(0.051)"* York R., VA(1981) — 47 0.543(0.073) York R., VA 89 0.584(0.037) 67 0.612(0.060) North Carolina Region Far Creek, NC (1981) 28 0536(0.066) 38 0.645(0.055)* Far Creek, NC 94 0.500(0.036) 98 0.602(0.035) Spencer's Bay, NC 43 0.430(0.053) 45 0.547(0.052) Wysocking Bay and Far Creek, NC 64 0.428(0.044) 63 0.571(0.044) Adults^ Northern Region Belford, NJ (1983) 11 0.636(0.103) 29 0.500(0.066) Chesapeake Bay Region Virginia 121 0.620(0.044) 138 0.554(0.038) Southern Region North Carolina 57 0.596(0.065) 59 0.636(0.063) •HWE x^O.05 = 3.841, df = 1. ••HW/E x^O.O! = 6.635. df = 1. •••HWE x^ 0,001 = 10.828, df = 1. 'Geographic comparisons among regions of allelic frequencies for juveniles; x^ = 1 942, P > 0.05, df = 2 (Pgd); x^ = 2 268, P > 05, df = 2 (Mdh) ^Geograpfiic comparisons among regions of allelic frequencies for adults; x^ = 0.566, P > 0,05, df = 2 (Pgd), x' = 3 020, P > 0.05. df = 2 (Mdh). 'Represents a pooled sample, where samples <10 were combined. 208 CRAWFORD ET AL : STOCK IDENTIFICATION OF WEAKFISH Malate dehydrogenase Mdh-1 Mdh-2 Mdh-3 Origin <^ ^ ^ ^' CJ OL <3 a o ^ 0.05, df = 1; Mdh-2: x^ = 0.001, P > 0.05, df = 1). Comparisons between male and female fish in- dicated significantly different allozyme frequencies for Mdh-2 but not for Pgd. The frequency of the common allele (100) at the Mdh-2 locus was 0.518 + 0.033 (SE) for 114 males and 0.637 ± 0.035 (SE) for 95 females, and these frequencies were signifi- cantly different (x' = 6.024, P < 0.05, df = 1). No differences were found at the Pgd locus (x" = 1.785, P > 0.10, df = 1). DISCUSSION Our study of allelic frequencies from populations of C. regalis along the east coast between Cape Cod, MA and Cape Hatteras, NC identified no statistical- ly distinguishable differences. Nei's (1972) genetic distances are quite small and the F^f value (0.046) is low; both indicate little genetic variation among the populations. Nonsignificant allelic frequency comparisons among geographic locations and size/ age classes were consistent with population homo- geneity. A comparison of allelic frequencies at the Mdh-2 locus showed a significant difference between sexes. We are unable to explain this difference and cannot discount sex linkage or sexual selection as possible causes. Alternatively, with the numerous chi-square tests used in the analyses a Type II statistical error may have occurred. Sample populations at several locations showed a heterozygote deficiency at Mdh-2 causing devia- tions from Hardy-Weinberg equilibrium. Several factors may cause heterozygote deficiencies (e.g., inbreeding, Wahlund effect, selection against het- erozygotes, scoring biases, and null alleles; Speiss 209 FISHERY BULLETIN: VOL. 87, NO. 1 1977), but the presence of null heterozygotes seems the most likely explanation. The low number of in- dividuals observed to be homozygous for the null allele suggests that it may be lethal for these in- dividuals (Speiss 1977). Previous investigators have suggested that two or three distinct stocks of weakfish occur in the Mid- dle Atlantic region (Nesbit 1954; Perlmutter et al. 1956; Seguin 1960). Nesbit (1954) examined dis- tances between circuli on scales and conducted a marking study using celluloid belly tags. He tagged 5,789 fish and 7.5% were returned when the fish were eviscerated. Thirty-six percent of the returned tags were from retail dealers and consumers pro- viding little information regarding actual recapture location. Nesbit concluded that the fishery consisted of two stocks. Perlmutter et al. (1956) examined intercirculi distances, fin rays, age, and growth data as well as Nesbit's (1954) data and concluded that there were northern and southern spawning weak- fish populations. Seguin (1960) performed a univariate analysis of morphometric and meristic data on juvenile weak- fish and separated Middle Atlantic weakfish into three segments: 1) New York, 2) Delaware (and possibly Virginia), and 3) North Carolina. She re- ported "a north-south trend in regression coeffici- ents" which may have been associated with environ- mental gradients (e.g., temperature) and clinal variation in the characters. Meristic characters, however, may be influenced by temperature (Barlow 1961) and intercirculi distances are related to growth rates that can vary geographically (Lux 1972; Shepherd and Grimes 1983; Harris and Gross- man 1985). Because growth is affected by many en- vironmental factors (e.g., temperature and food availability), it may not be indicative of genetic dis- continuity (Joseph 1972). Our results suggest that weakfish populations in the Middle Atlantic are not sufficiently distinct, genetically, to be considered as separate stocks (i.e., reproductively isolated). Weakfish perform exten- sive spring and fall migrations that could permit am- ple gene flow between populations. There are no ob- vious isolating mechanisms and only a small number of migrants would be needed to cause allelic fre- quencies to converge and make the population homogenous (Hartl 1980). In conclusion, the results of this investigation do not support the findings of earlier studies that distinct stocks of weakfish are present in the Mid- dle Atlantic. Even though there do not appear to be genetically discrete weakfish populations, there are variations in the population parameters (Shep- herd and Grimes 1983, 1984). The ability of a pop- ulation to sustain a harvest is largely dependent upon its growth, mortality, and fecundity. These life history parameters are used in fishery assessments (e.g., dynamic pool and stock-recruitment models). Use of northern weakfish growth parameters would predict overly optimistic yields for southern fish- eries, and an incorrect stock-recruitment relation- ship. Therefore, as a practical matter it is probably best to manage weakfish as discrete northern and southern units. These units may not be reproduc- tively independent, and the effects of fishing (par- ticularly recruitment overfishing) are likely to be imposed upon the entire population. ACKNOWLEDGMENTS The authors thank all the people who helped out in sampling including Jess Hawkins at the North Carolina Department of Natural Resources, Division of Marine Fisheries, the scientists at the Virginia Institute of Marine Sciences, Ichthyological Asso- ciates, the personnel at NMFS, Woods Hole, MA, and many commercial fishermen. We are grateful to Bob Vrijenhoek for his advice and use of his laboratory for the electrophoresis. Special thanks to Ken Able of Rutgers University and Brad Brown and Ambrose Jearald, Jr. at NMFS, Miami, FL and Woods Hole, MA respectively. We also thank two anonymous reviewers for their comments. We will not forget the support and encouragement of Steve Turner, Russ Schenk, Deb Shalders, Bob Palmer, Peter Hood and Rhett Lewis. We thank G. Shepherd, G. Grossman, M. Freeman, M. Flood, J. Barrett, D. Stouder, and J. Hill for their com- ments on the manuscript. Funding for this project was provided by NMFS (contract No. NA-79-FAC- 00041). LITERATURE CITED Barlow, G. W. 1961. Causes and significance of morphological variation in fishes. Syst. Zool. 10:105-117. BiGELOW. H., AND W. SCHROEDER, 19.53. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv. Fish. Bull. 53:1-577. Crawford, M. K. 1 984 . Stock identification of the weakfish (Cynoscion regalis) in the Middle Atlantic region. M.S. Thesis, Rutgers Univer- sity, New Brunswick, NJ, 60 p. Clayton, J. W., and D. W. Tretiak. 1972. Aminecitrate buffers for pH control in starch gel elec- trophoresis. J. Fish. Res. Board Can. 29:1160-1172. Green, D. M. 1979. A BASIC program for calculating indices of genetic 210 CRAWFORD ET AL.: STOCK IDENTIFICATION OF WEAKFISH distance and similarity. J. Hered. 70:429-430. Grosslein, M. D. 1969. Groundfish survey programs of BCF Woods Hole. Commer. Fish. Rev. 31(8-9):22-35. Harris, H., and D. A. Hopkinson. 1976. Handbook of enzyme electrophoresis in human genet- ics. American Elsevier, New York, NY, USA. Harris, M. J., and G. D. Grossman. 1985. Growth, mortality and age composition of a lightly ex- ploited tilefish substock off Georgia. Trans. Am. Fish. Soc. 114:837-846. Hartl, D. L. 1980. Principles of population genetics. Sinauer Associates, Sunderland, MA, USA. Joseph, E. B. 1972. The status of the Seiaenid stocks of the Middle Atlan- tic coast. Chesapeake Sci. 13:87-100. Levene, H. 1949. On a matching problem arising in genetics. Annu. Math. Stat. 20:91-94. Lux, F. E. 1972. Age and growth of the winter flounder Pneudopleuro- nectes americanus on Georges Bank. Fish. Bull., U.S. 71: 505-512. Manwell, C, and C. M. a. Baker. 1970. Molecular biology and the origin of species. Univ. Wash. Press, Seattle, WA, USA. Merriner, J. V. 1976. Aspects of the reproductive biology of the weakfish, Cynoscion regalis (Sciaenidae), in North Carolina. Fish. Bull., U.S. 74:18-26. MURAWSKI, S. A. 1977. A preliminary assessment of weakfish in the Middle Atlantic Bight. Natl. Mar. Fish. Serv., Woods Hole Lab. Ref. 77-26. Nei, M. 1972. Genetic distance between populations. Am. Nat. 106: 283-292. Nesbit, R. a. 1954. Weakfish migration in relation to its conservation. U.S. Fish Wildl. Serv., Spec. Sci. Rep.-Fish. 115, 81 p. Perlmutter, a., S. W. Miller, and J. C. Poole. 1956. The weakfish (Cynoscion regalis) in New York waters. N.Y. Fish Game J. 3:1-43. Ridgway, G. J., S. V. Sherburne, and R. D. Lewis. 1970. Polymorphism in the esterases of Atlantic herring. Trans. Am. Fish. Soc. 99:147-153. Seguin, R. T. 1960. Variation in the Middle Atlantic coast population of the grey squetague, Cynosciov regalis (Bloch and Schneider). 1801. Ph.D. Thesis, Univ. Delaware, Newark, DE, 70 p. Selander, R. K. 1970. Behavior and genetic variation in natural populations (Mus musculus). Am. Zool. 10:53-66. Shaw, C. R., and R. Prasad. 1970. Starch gel electrophoresis a compilation of recipes. Biochem. Genet. 4:297-320. Shepherd, G. R., and C. B. Grimes. 1983. Geographic and historic variations in growth of weak- fish, Cynoscion regalis, in the Middle Atlantic Bight. Fish. Bull., U.S. 81:803-813. 1984. Reproduction of weakfish, Cynoscion regalis, in the New York Bight and evidence for geographically specific life history characteristics. Fish. Bull., U.S. 82:501-511. Speiss, E. B. 1977. Genes in populations. John Wiley and Sons, New York, NY, USA. Thorne, C. J. R., L. I. Grossman, and N. Kaplan. 1963. Starch-gel electrophoresis of Malate Dehydrogenase. Biochim. Biophys. ACTA 73:193-203. Utter, F. M., H. 0. Hodgins, and F. W. Allendorf. 1974. Biochemical genetic studies of fishes: potentialities and limitations. In D. C. Malins and J. R. Sargent (editors). Biochemical and biophysical perspectives in marine biology. Vol. 1, p. 213-238. Acad. Press, San Franc, CA, USA. Welsh, W. W., and C. M. Breder. 1923. Contributions to the life histories of Sciaenidae of the eastern United States coast. Bull. U.S. Bur. Fish. 39:141- 201. WiLK. S. J. 1979. Biological and fisheries data of weakfish. Cynoscion regalis (Bloch and Schneider). Natl. Oceanic Atmos. Adm. Tech. Ser. Rep. No. 21. Sandy Hook Laboratory. High- lands. NJ 07732. 1981. A review of the fisheries for Atlantic croaker spot, and weakfish. 1940-1979. In H. Clepper (editor). Marine recrea- tional fisheries symposium VI, p. 15-28. Sportfishing In- stitute, Washington, D.C., USA. 211 ESCAPEMENT BY FISHES FROM MIDWATER TRAWLS: A CASE STUDY USING LANTERNFISHES (PISCES: MYCTOPHIDAE) John V. Gartner, Jr.,' Walter J. Conley,^ and Thomas L. Hopkins' ABSTRACT Escapement of fishes through the meshes of a trawl is a recognized but unquantified problem in estimating size of mesopelagic fish populations. This paper provides estimates of net escapement by midwater fishes, using the lanternfishes as an example. Comparison of overall catches for Tucker trawls of 1.6 mm and 4.0 mm mesh show highly significant differences. The small mesh size outcaught the large by a factor of 2.7 for fishes smaller than 30 mm SL, while for larger fishes, the small mesh catches averaged 90% of the larger mesh. Among six ranking abundant species, three patterns of escapement were observed, based on significant differences in cross-sectional fish dimensions and morphological characters: 1) The entire size range of the species was significantly underestimated {Benthosenm suborbitale and Notolychnits valdiviae); 2) only size ranges below those of sexually mature adults were significantly underestimated (Lampanyctus alatus and Lepidophanes guentheri); 3) only juveniles <30 mm SL were significantly underestimated (Ceratoscopelu^ townsendi and Diaphus dwnerilii). "Conventional" midwater trawl meshes of 4 to 6 mm diameter mesh provide adequate data for general distributional surveys and also for some quantitative estimations such as overall biomass. Determinations of juvenile biomass, spavra- ing period, trophic impact, and relative species abundances based on conventional mesh collections may be prone to substantial error depending on species size. It is suggested that a net mesh of <2 mm be used in conjunction with larger mesh trawls if quantitative life history data on smaller size classes and species are required. Requisite to studies of the roles of mesopelagic fishes in deep-sea ecological processes are accurate determinations of species composition and the ver- tical and horizontal structure of populations. Al- though these are now well documented for many groups in many regions of the world ocean (see Mar- shall 1980), accurate abundance estimates, particu- larly over the entire size range of a species, are often not possible because of sampling biases related to net construction and trawling methods (Stein 1985). Two factors responsible for much of the difficul- ty in estimating abundance of midwater fishes are net avoidance by large size classes and escapement through the net meshes during capture by small fishes of slender body shapes (Harrisson 1967). Both result in underestimates of species abundance, which can apply to either particular size classes, or, in the case of diminutive species, an entire popula- tion. While some studies show that net avoidance may be reduced through the use of trawls with large 'University of South Florida, Department of Marine Science, 140 Seventh Avenue S.E., St. Petersburg, FL 33701. ^Florida Department of Natural Resources, Florida Marine Re- search Institute, 100 Eighth Avenue S.E.. St. Petersburg, FL 33701. Manuscript accepted October 1988. Fishery BuUetin. U.S. 87:213-222. mouth areas, there are a number of attendant diffi- culties including enhanced escapement due to in- creased mesh size (Stein 1985). Of the two problems, net avoidance remains the most difficult to quan- tify. Escapement is more easily calculated, but little quantitative research has been directed towards this problem in studies of midwater fishes (Harrisson 1967; Clarke 1983a). In this study we quantify escapement through net meshes of midwater trawls using the lanternfishes (family Myctophidae) as an example. The ecological implications of net escapement are discussed. MATERIALS AND METHODS Myctophids were collected during eight cruises aboard the RV Suncoaster from an area centered at lat. 27°N, long. 86°W. The cruises covered four seasons over a period of 30 months. Sampling months were September (1984), November (1985), January (1986, 1987), March (1985, 1987), May (1986), and July (1985). Station data are presented in Table 1. All samples were taken using modified Tucker trawls fished open in an oblique "V" sweep from the surface to 200 m at night. This depth range en- compasses the peak nighttime abundance of all 213 FISHERY BULLETIN: VOL. 87, NO. 1 numerically dominant lanternfish species in the east- ern Gulf of Mexico (Gartner et al. 1987). All nets were fished at 1.5 to 2.5 knots with a total fishing duration for each sample of approximately 1 hour. Sampling usually began about 1 hour after sunset and ended 1 hour before sunrise. The bottom trawl bar was weighted to incline the net mouth about 30° from the vertical at these towdng speeds (determined from observations using scuba). Two trawl con- figurations were used: a 5.3 m^ (effective fishing area) net of 4 mm bar mesh in the body and 1 mm mesh in the funnel, and a 2.6 m' net of 1.6 mm mesh. Both nets had cod ends lined with 505 i^m mesh. Trawl depths were recorded by mechanical time- depth recorder (TDR) and monitored with an elec- tronic deck readout linked to a transducer mounted on the trawl frame. The volume of water filtered during each net haul was calculated from flow meters mounted on the trawl frame. Myctophids were fixed in 10% (v:v) formalin and preserved in 50% isopropanol. During all cruises ex- cept January and March 1987, a large number of postlarval specimens from the dominant species were removed from the catches for use in life history studies. These were blotted to remove excess mois- ture and measured to the nearest millimeter stan- dard length (mm SL). The remaining myctophids were measured in the lab after preservation. Be- cause of shrinkage of preserved specimens, the lengths of freshly measured individuals were de- creased by 12% (shrinkage factor determined from Gartner, unpub. data; K. J. Sulak pers. commun.^). All myctophids were identified to the lowest possi- ble taxon, with species identifications made using Nafpaktitis et al. (1977). The effect of using nets of differing mouth areas was minimized by calculating the abundance of in- dividuals per 10"* m^ for each net over the entire size range, which was then divided into 5 mm SL size classes. Kolmogorov-Smirnov (K-S) two-sample tests (Siege! 1956) were used for overall internet comparisons of capture over the size range by size classes and by net mesh for size groups smaller than 30 mm SL and larger than 30 mm SL. The K-S tests were appHed to similar comparisons for each of the ranking myctophid species. Except where noted, the significance level for all tests was P < 0.01. Rank- ing species for all cruises were defined as the most abundant species which combined comprised 75% or more of the total number of specimens captured (Gartner et al. 1987). To evaluate if escapement was related to body morphology as well as size, measurements of the greatest cross-sectional dimensions were made on a series of preserved specimens of each of the rank- ing species. Measurements were made to the near- est 0.01 mm using dial calipers on a series of ran- domly selected individuals which encompassed the postlarval size range of each species. Assuming that myctophids are elliptical in cross section, areas were calculated for each specimen using the formula nab, where a and b are the radii of the short and long axes of the ellipse. Cross-sectional areas were re- gressed against the square of length and tested for significance (P < 0.01) among species using a Stu- dent's i-test (Sokal and Rohlf 1981). RESULTS Collection Data The 4 mm mesh net was used at 78 stations, from which 7,861 myctophids were collected with a total volume filtered of 1.65 x 10« m^ (Table 1). The 1.6 mm mesh net was also used at 78 stations, with totals of 7,494 individuals captured and 8.97 x 10^ m^ filtered (Table 1). The mean ratio of volume filtered for the larger to smaller nets was 1.84:1 (range for all cruises was 1.72:1 to 2.07:1). Table 1 .—Collection data. Number of samples (Volume filtered 10" m' ') Cruise 3.2 m= ' 1.6 mm 6.5 m^ 4 mm September 1984 3 (3.84) 3 (6.42) March 1985 14 (16.94) 11 (24.31) July 1985 9 (10.28) 8 (17.56) November 1985 4 (4-21) 17 (30.92) January 1986 10 (11.85) 11 (26.13) May 1986 13 (13.91) 10 (20.19) January 1987 16 (18.89) 7 (16,32) March 1987 9 (9.82) 11 (22.74) Totals 78 (89.74) 78 (164.59) 'K. J. Sulak, Atlantic Reference Centre, Huntsman Marine Lab- oratory, St. Andrews, New Brunswicl<, Canada EOG 2X0, pers. commun. May 1988. Abundances by Size Class The numbers of individuals collected per 10^ m^ for both nets are shown in Figure 1. Data for fishes larger than 80 mm SL were not included because only 16 specimens were collected. Catch differences were highly significant between the two mesh sizes (P < 0.001). In both nets, the 16 to 20 mm SL size class was most abundant, but the 1.6 mm mesh net 214 GARTNER ET AL.: FISH ESCAPEMENTS FROM MIDWATER TRAWLS Q 30 LLI DC fiJ 25 CO _i < cr ziuJ ii 08 z o cr UJ CL 20 15 10 1,6mm MESH 4.0mm MESH SIZE GROUP (mmSL) Figure 1.— Overall numbers of myctophids collected per 10''m^ water filtered for the 1.6 mm mesh and 4.0 mm mesh nets. collected over twice as many specimens as the 4 mm mesh net. Overall, the 1.6 mm mesh net was signif- icantly more effective in collecting individuals of smaller than 30 mm SL with a mean calculated abundance ratio for the 6 to 30 mm size groups of 2.7:1 between the 1.6 mm and 4 mm meshes. The abundance ratios for the small to large mesh sizes is highest for the smallest size group considered (4.4:1 for the 6 to 10 mm SL group). Although the 4 mm mesh captured more fishes at sizes >30 mm SL, the differences, while significant, were not pronounced and never approached the ratios noted for the smaller size groups. The mean ratio for the 1.6 mm to 4 mm meshes for size groups 31 to 65 mm SL was 0.9:1 (range 0.8:1 to 1.0:1). At sizes larger than 65 mm SL, the ratios were variable owing to small sample sizes. Abundances, Cross-Sectional Dimensions and Morphologies of Ranking Species The same five species made up the ranking mycto- phids from both nets, although the order of abun- dance differed (Table 2). A sixth species, Cerato- swpelus townsendi (formerly C. warmingii, see Badcock and Araujo 1988), was also a dominant myctophid in the 4 mm net catches. Comparisons of internet abundances of ranking species for each size group revealed three basic patterns: 1) Virtual- ly the entire size range was underestimated by the 4 mm mesh net {Benthosema suborbitale and Noto- lychnus valdiviae, Fig. 2a, b); 2) only size groups up to sexually mature adults (ca. 40 mm SL) were underestimated by the 4 mm mesh net {Lampanyc- tus alatus and Lepidophanes guentheri Fig. 2c, d); and 3) only juveniles smaller than 26 to 30 mm were underestimated by the 4 mm mesh net {Ceratosco- pelus townsendi and Diaphus dumerilii Fig. 2e, f). Of the ranking species, only these last two species were collected in greater numbers by the 4 mm mesh net at sizes larger than 30 mm SL. The patterns of net capture vs. size ranges were directly related to the general body dimensions and morphologies of the ranking species. Maximum cross-sectional depths and widths were measured on the body at the pectoral fin base in Benthosema suborbitale, Lampanyctus alatics, Lepidophanes guentheri, and Notolychnus valdiviae, while for Ceratoscopelu^ townsendi and Diaphus dumerilii, the maxima were on the head anterior to the oper- cular openings. Head profiles also differed among species, with the first four species having pointed or wedge shaped outlines, while the latter two had blunt, rounded heads. Both A'^. valdiviae and B. sub- orbitale (Pattern 1) are diminutive species not ex- ceeding 22 mm and 33 mm, respectively, in the eastern Gulf, while the other four species grow much larger (Gartner et al. 1987; Gartner, unpub. data). Mean cross-sectional measurements (Table 3) show that in relation to body length, B. suborbitale is deep bodied, while A^. valdiviae, Lepidophanes guentheri, and Lampanyctus alatus (Pattern 2) are all slender. When compared with the previous species at equi- valent lengths, C. townsendi and D. dumerilii (Pat- tern 3) have generally thick cross-sections. 215 FISHERY BULLETIN: VOL. 87, NO. 1 Table 2.— Ranking species of myctophids collected, by net. 1.6 mm mesh 4.0 mm mesh % of %of No. total No. total Species Rank captured captured Nc/IO* m^ Rank captured captured No./IO" m^ Notolychnus valdlviae 1 1.752 23.40 19.52 2 1,285 16.30 7.81 Diaphus dumerilii 2 1,317 17.60 1468 1 1,572 20.00 9.55 Lampanyclus alalus 3 895 11.90 9.98 4 783 10.00 4.76 Lepidophanes guenthen 4 877 11.70 9.78 3 1,039 13.20 6.31 Benlhosema suborbilale 5 778 10.40 8.67 6 597 7.60 3.63 Ceraloscopelus townsendi 5 645 8.20 3.92 SIZE GROUP (mmSL) Figure 2. a-f.— Numbers of ranking species of myctophids collected per 10'' m'' water filtered for the 1.6 mm mesh and 4.0 mm mesh nets. 216 GARTNER ET AL.: FISH ESCAPEMENTS FROM MIDWATER TRAWLS Table 3. — Mean cross-sectional dimensions (mm) and body morphologies for ranking myctophid species by size group. Underline indicates dimensions where number of individuals per 10" m'' filtered are approximately equal between two net meshes (crossover point on Figure 2a-f). D = body depth (mm); W = body width (mm). Spec es Benthosema suborbitale Pointed Deep D W Ceratoscopelus townsendi Diaphus dumerilii Lampanyctus alatus Lepidophanes guentheri Pointed Slender Notolychnus valdiviae Head profile: Body morphology: Blunt Thick Blunt Thick Pointed Slender Pointed Slender Size class D w D W D W D w D W 8-10 1.72 X 1.14 11-15 3.05 X 1.43 2.69 X 1.49 2.06 X 1.42 16-20 4.52 X 2.30 3.51 X 1.87 3.58 X 1.83 3.09 X 1.56 2.76 X 1.44 2.80 X 1.74 21-25 5.70 X 3.02 4.75 X 2.60 5.03 X 2.54 3.90 X 2.14 3.85 X 1.77 3.26 X 2.10 26-30 6.41 X 3.34 5.74 X 3.34 5.95 X 3.03 4.85 X 2.45 4.76 X 2.36 31-35 7.40 X 4.02 6.43 X 3.82 6.78 X 3.66 5.89 X 3.01 5.88 X 3.08 36-40 7.56 X 4.41 7.61 X 4.24 6.77 X 3.43 6.87 X 3.64 41-45 8.31 X 5.12 8.59 X 4.92 7.82 X 4.00 7.97 X 4.30 46-50 9.21 X 6.14 9,59 X 5.71 8.35 X 4.18 8.27 X 4.43 51-55 11.33 X 6.35 11.42 X 7.00 9.38 X 5.07 56-60 11.75 X 7.46 12.57 X 7.88 10.24 X 5.21 61-65 10.97 X 5.77 The smallest size at which the 4 mm mesh net showed comparable catches to the 1.6 mm mesh for any ranking species was 23 mm SL (Fig. 2). Regres- sion of cross-sectional areas vs. length for sizes larger than 23 mm clearly grouped the ranking species according to the catch patterns (Fig. 3). Dif- ferences between groups were highly significant (P < 0.001). ■♦OO 800 1200 1600 2000 2400 2800 3200 3600 4000 LENGTH (mm2) Figure 3.— Regressions of body cross-sectnonal area (mm^) on the square of length for ranking species of myctophids {Notolychnus valdiviae excluded). 217 FISHERY BULLETIN: VOL. 87, NO. 1 DISCUSSION Net Escapement: Considerations In the present study, the extent of bias due to net avoidance is unknown, but our trawling program was designed to minimize the problem as much as possible. All of our samples were taken at night when net avoidance is supposedly mitigated (e.g., Pearcy and Laurs 1966). Observations from submer- sibles suggest that small fishes (<60 to 70 mm) are not as successful in avoiding nets as are larger ones (B. H. Robison pers. commun.""). The myctophid faunas in low latitude ecosystems (upwelling regions excepted) tend towards the small end of the size range (Clarke 1973; Gartner et al. 1987). Also, the detection of nets by mechanical stimuli and the ef- fective mouth area of a net have been suggested to enhance avoidance by midwater fishes (Harrisson 1967; Clarke 1973; Pearcy 1980; Stein 1985). Based on these factors, a smaller catch per unit volume would have been predicted for the 1.6 mm mesh net, which was white and had one-half the effective fish- ing area of the dark green 4 mm net. However, at size ranges smaller than 30 mm SL, it outcaught the larger mesh net by a factor of 2.7, whereas among larger fishes, it collected 80% to 97% (x = 90%) of the number of specimens of the large mesh. Since the introduction of the Isaacs-Kidd mid- water trawl (Isaacs and Kidd 1953), most midwater fish surveys have used gear with a net mesh size of 4 to 6 mm bar length (e.g., Badcock 1970; Hulley 1972; Clarke 1973; Gartner et al. 1987; Karnella 1987). While such gear is necessary for general surveys, our data suggest that they may be inade- quate for obtaining certain quantitative estimates for small micronekton because of net escapement. It is clear that underestimates of myctophid abun- dance are marked in the 4 mm mesh (Figs. 1, 2). These underestimates may apply only to certain size ranges of the population, as in D. dumerilii or L. guentheri, or may include the entire size spectrum of a species, as in N. valdimae. Similar trends of escapement have been noted among dominant spe- cies of sergestid shrimps examined from the same collections used in the present study (M. E. Flock pers. commun.^). Many myctophid species, especially strong vertical *B. H. Robison. Monterey Bay Aquarium Research Institute, 160 Central Avenue, Pacific Grove, CA 93940, pers. commun. June 1988. 'M. E. Flock, Department of Marine Science, University of South Florida, 140 Seventh Avenue S.E., St. Petersburg, FL 33701, pers. commun. June 1988. migrators, are generally muscular and slender and possess very small teeth. They present a relatively small cross section that does not appear to be ef- fectively retained by the large net meshes until some critical threshold of body thickness is reached. Among species with relatively pointed heads, i.e., those whose maximum body dimensions lie behind the head, the ability of large mesh nets to hold in- dividuals is as much a function of the lateral thick- ness as it is of the dorsoventral dimension of the body. Even though dorsoventral thickness may greatly exceed mesh size, lateral measurements must be close to or exceed the mesh bar length in order for the 4 mm mesh net to sample these spe- cies as efficiently as the 1.6 mm mesh net (Table 3, Fig. 3). It appears that until both body axes are equal or greater in size than the mesh diameter, if a "pointed head" fish succeeds in getting its head through the mesh, it can readily escape. In contrast, species with maximum body dimensions on the head (blunt heads) show reduced escapement from the large mesh when the dorsoventral aspect alone reaches a critical threshold, i.e., they are not able to push their head through the mesh. Implications of Escapement for Ecological Data Collections of mesopelagic fishes from many re- gions are now extensive enough to provide a good representation of species composition and distri- bution, especially with respect to the families Myc- tophidae and Gonostomatidae (Gjosaeter and Kawaguchi 1980; Hulley 1981; Gartner et al. 1987, Karnella 1987). There has been increasing emphasis on quantitative assessment of various aspects of myctophid ecology, such as population dynamics (J. Gjosaeter 1973a, 1981; Clarke 1983b, 1984; Lin- kowski 1985; H. Gjosaeter 1987), trophodynamics (Clarke 1978, 1980; Baird and Hopkins 1981a, b; Hopkins and Baird 1981, 1985) and fishery poten- tials for midwater fish species (Gjosaeter and Kawa- guchi 1980). In virtually all of these studies, data from mesh sizes of 4 mm or greater were used. Our findings indicate that escapement among size classes and species smaller than 30 mm SL is pro- nounced and that midwater trawls with mesh of <2 mm diameter should be used in order to obtain ac- curate estimates of fishes in this size range. It is not enough to assume that catch efficiencies are propor- tional over all size ranges and that some factor can be applied to catches with larger mesh nets to ac- count for escapement. It is clear that some species do appear to have proportional catch rates between 218 GARTNER ET AL.: FISH ESCAPEMENTS FROM MIDWATER TRAWLS the two mesh sizes, e.g., Notolychnus valdiviae, while others, e.g., Lampanyctus alatics, show very different catch rates depending on size group and mesh size (Fig. 2). Overall biomass values for myctophids from the two mesh sizes are very similar, 19.61 g/10'* m^ (1.6 mm mesh) and 19.22 g/lO* m'' (4.0 mm mesh). This suggests that the larger mesh sizes provide a fairly accurate estimate of overall standing stock of myc- tophids and allow for interregional data comparisons (Maynard et al. 1975; Hopkins and Lancraft 1984). However, estimates of standing stock for certain size classes would be erroneous, especially for juve- niles smaller than 30 mm SL (Fig. 4). Net escapement by small size classes can also bias quantitative determinations of relative species abun- dance, spawning period, juvenile recruitment, and trophodynamic impact. Benthosema suborbitale, Ceratoscopelus townsendi and Notolychmis valdiviae are pan-oceanic in tropical-subtropical latitudes and are among the most abundant species throughout their zoogeographic range (Gartner et al. 1987). Based on our calculations, it is likely that B. sub- orbitale and A'', valdiviae are of much greater nu- merical importance than previous data sets have suggested (e.g., Clarke 1973; Backus et al. 1977; Hulley 1981; Gartner et al. 1987). Net escapement can also affect assessment of spawning period. Abundance comparisons for the two nets by cruise for D. duvxmlii show that if one were to attempt to determine periods of larval re- cruitment for this species using length frequencies from the 4.0 mm mesh net, there are only sugges- tions of a spring-early summer spawning period (May 1986, July 1985), whereas the 1.6 mm net catches clearly indicate an early spring through fall influx of newly metamorphosed juveniles (Fig. 5). Using the small mesh net, the birthdays of juveniles for age and growth studies can more readily be fixed. In trophodynamic studies that have considered the impact of midwater fishes on zooplankton prey, con- siderable predation pressure has been shown on cer- tain size classes and taxa of zooplankters (Gjosaeter 1973b; Merrett and Roe 1974; Clarke 1978, 1980; Hopkins and Baird 1981, 1985; T. M. Lancraft pers. commun^; Hopkins and Gartner unpub. data). It is well documented that ontogenetic changes in prey taxa and size selection occur among myctophids. Our data suggest that predation pressure would be much higher from small size classes or species of mycto- phids <30 mm SL than could be calculated from studies using larger mesh collection gear. CONCLUSIONS As Harrisson (1967) remarked, no single midwater net will adequately sample all species or size ranges. At high latitudes and in the lower mesopelagic zone where many midwater fish species may attain sizes >100 mm SL, it has been observed that myctophids readily avoid even large midwater trawls (Pearcy 'T. M. Lancraft, Department of Marine Science, University of South Florida, 140 Seventh Avenue S.E., St. Petersburg, FL 33701, pers. commun. June 1988. O O O 1- nj LU C/5 UJ o CD 1.6mm MESH 4.0mm MESH SIZE GROUP (mmSL) Figure 4.— Overall biomass of myctophids per 10^ m^ water filtered for the 1.6 mm mesh and 4.0 mm mesh nets. 219 FISHERY BULLETIN: VOL. 87. NO. 1 MAR '85 12 10 JUL '85 8 \ 6 \ 4 ^^\ 2 n 1 1 J 1 , "^^^-^ 1 1 1 20 18 Figure 5.— Size frequencies oiDia-phus dume- rilii among collection periods. SIZE GROUP (mmSL) 220 GARTNER ET AL.: FISH ESCAPEMENTS FROM MIDWATER TRAWLS and Laurs 1966; B. H. Roblson, unpub. data from submersible observations). Our data and those of Clarke (1973) from conventional midwater trawls suggest that avoidance is not a primary concern in lower latitudes which are dominated by small (<70 mm SL) species. In all regions, however, escapement of small size groups and species through meshes is a real problem which until now has not been well quantified. Future quantitative ecological research on post- larval midwater fishes should use a midwater trawl of mesh size <2 mm in conjunction with larger mesh in order to correct for the effects of net escapement. This ancillary net should ideally be mounted on an identical frame design (although not necessarily identical mouth area) as the large mesh and fished in a similar manner in order to readily facilitate internet comparisons. Should logistic considerations restrict gear use to a single type, our data indicate that the small mesh gear would be more efficient overall for collection of size groups from 10 to ap- proximately 70 mm SL. ACKNOWLEDGMENTS We wish to thank the crew of the RV Suncoaster for their ongoing assistance and professionalism during the many long cruises to Standard Station. Shiptime was funded by NSF research grant OCE #8410787 to T. L. Hopkins. We also thank D. F. Williams, University of South Florida, Department of Marine Science (USFMS) for preparation of graphics and R. R. Wilson, K. L. Carder, and T. M. Lancraft (USFMS), T. Bailey of Harbor Branch Oceanographic Institution, and B. H. Robison of the Monterey Bay Aquarium Research Institute for their critical reviews of the manuscript. M. D. Murphy and R. G. Muller, Florida Department of Natural Resources, Florida Marine Research Insti- tute provided valuable assistance with statistical analyses. The senior and second authors are especially grateful to the Houston Underwater Club, Houston, TX for financial assistance provided through their SEASPACE Scholarship program. This paper forms part of doctoral dissertation research in progress by the senior author at USFMS. LITERATURE CITED Backus, R. H.. J. E. Craddock. R. L. Haedrich. and B. H. Robison. 1977. Atlantic mesopelagic zoogeography. In R. H. Gibbs, Jr. (editor), Fishes of the Western North Atlantic. Sears Found. Mar. Res. Res.. Yale Univ., New Haven, CT, Mem. 1, Pt. 7, p. 266-287. Badcock. J. 1970. The vertical distribution of mesopelagic fishes collected on the SOND cruise. J. mar. biol. Assoc. U.K. oO;1001- 1044. Badcock. J., and T. M. H. AraOjo. 1988. On the significance of variation in a warm water cosmo- politan species, nominally Ceratoscopetus warmingii (Pisces, Myctophidae). Bull. Mar. Sci. 42:16-43. Baird, R. C. and T. L. Hopkins. 1981a. Trophodynamics of the fish Valemiennellus tripunc- tulatus. II. Selectivity, grazing rates and resource utiliza- tion. Mar. Ecol. Prog. Ser. 5:11-19. 1981b. Trophodynamics of the fish Valenciennellus tripunc- tulatus. III. Energetics, resources and feeding strategy. Mar. Ecol. Prog. Ser. 5:21-28. Clarke, T. A. 1973. Some aspects of the ecology of lanternfishes (Mycto- phidae) in the Pacific Ocean near Hawaii. Fish. Bull., U.S. 71:401-434. 1978. Diel feeding patterns of 16 species of mesopelagic fishes from Hawaiian waters. Fish. Bull. U.S. 76:495- 513. 1980. Diets of fourteen species of vertically migrating meso- pelagic fishes in Hawaiian waters. Fish. Bull.. U.S. 78: 287-304. 1983a. 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(2-3-4): 311-340. 1983b. Sex ratios and sexual differences in size among meso- pelagic fishes from the central Pacific Ocean. Mar. Biol. 73:203-209. 1984. Fecundity and other aspects of reproductive effort in mesopelagic fishes from the North central and equatorial Pacific. Biol. Oceanogr. 3(2):147-165. Gartner, J. V., Jr., T. L. Hopkins, R. C. Baird, and D. M. Milliken. 1987. The lanternfishes (Pisces: Myctophidae) of the eastern Gulf of Mexico. Fish. Bull., U.S. 85:81-98. GjOsaeter. H. 1987. Primary growth increments in otoliths of six tropical myctophid species. Biol. Oceanogr. 4:359-382. GjOsaeter. J. 1973a. Age, growth and mortality of the myctophid fish, Benthosema glaciale (Reinhardt), from western Norway. Sarsia 52:1-14. 1973b. The food of the myctophid fish, Benthosema glaciale (Reinhardt), from western Norway. Sarsia 52:53-58. 1981. Growth, production and reproduction of the myctophid fish Benthosema glaciale from western Norway and adjacent seas. Fiskeridir. Skr. Ser. Havunders. 17:79-108. GjOsaeter, J., and K. Kawaguchi. 1980. A review of the world resources of mesopelagic fish. FAO Fish. Tech. Pap. No. 193, 151 p. Harrisson, C. M. H. 1967. On methods for sampling mesopelagic fishes. Symp. Zool. Soc. Lond. 19:71-126. Hopkins. T. L.. and R. C. Baird. 1981. Trophodynamics of the fish Valenciennellus tripunc- tulatus. I. Vertical distribution, diet and feeding chronol- ogy. Mar. Ecol. Prog. Ser. 5:1-10. 1985. Aspects of the trophic ecology of the mesopelagic fish Lampanyctus alatus (Family Myctophidae) in the eastern Gulf of Mexico. Biol. Oceanogr. 3:285-313. 221 FISHERY BULLETIN: VOL. 87, NO. 1 Hopkins, T. L., and T. M. Lancraft. 1984. The composition and standing stock of mesopelagic micronekton at 27°N, 86°W in the eastern Gulf of Mexico. Contrib. Mar. Sci. 27:143-158. HULLEY, P. A. 1972. A report on the mesopelagic fishes collected during the deep-sea cruises of R.S. ' Africana IT, 1961-1966. Ann. S. Afr. Mus. 60(6):197-236. 1981. Results of the research cruise of F.R.V. "Walther Her- wig" to South America. LVHI. Family Myctophidae (Osteichthyes, Myctophiformes). Arch. Fischereiwiss. 31(1): 1-300. Isaacs, J. E., and L. W. Kidd. 1953. Isaacs-Kidd midwater trawl. Rep. Scripps Inst. Oceanogr. 1, 18 p. Karnella, C. 1987. Family Myctophidae, lanternfishes. In R. H. Gibbs, Jr. and W. H. Krueger (editors). Biology of midwater fishes of the Bermuda Ocean Acre. Smithson. Contrib. Zool. 452: 51-168. LlNKOWSKI, T. B. 1985. Population biology of the myctophid fish Gymnoscope- lus nicholsi (Gillbert, 1911) from the western South Atlan- tic. J. Fish. Biol. 27:683-698. Marshall, N. B. 1980. Deep-sea biology. Developments and perspectives. Gariand STPM Press, New York, NY, 566 p. Maynard, S. D., F. V. RiGGs, and J. F. Walters. 1975. Mesopelagic micronekton in Hawaiian waters: faunal composition, standing stock and diel vertical migration. Fish. Bull., U.S. 73:726-736. Merrett, N. R., and H. S. J. Roe. 1974. Patterns and selectivity in the feeding of certain meso- pelagic fishes. Mar. Biol. 28:115-126. Nafpaktitis, B. G., R. H. Backus, J. E. Craddock, R. L. Haedrich, B. H. Robison, and C. Karnella. 1977. Family Myctophidae. In R. H. Gibbs, Jr. (editor). Fishes of the Western North Atlantic. Sears Found. Mar. Res., Yale Univ., New Haven, CT, Mem. 1, Pt. 7, p. 13- 265. Pearcy, W. G. 1980. A large, opening-closing midwater trawl for sampling oceanic nekton, and comparison of catches with an Isaacs- Kidd midwater trawl. Fish. Bull., U.S. 78:529-534. Pearcy, W. G., and R. M. Laurs. 1966. Vertical migration and distribution of mesopelagic fishes off Oregon. Deep-Sea Res. 13:153-165. SlEGEL, S. 1956. Nonparametric statistics for the behavioral sciences. McGraw-Hill Kogakusha, Tokyo, Japan. 312 p. SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry. 2nd ed. W. H. Freeman and Co., N.Y., 859 p. Stein, D. L. 1985. Towing large nets by single warp at abyssal depths: methods and biological results. Deep-Sea Res. 32:183- 200. 222 THE ECONOMIC VALUE OF FISHING SUCCESS: AN APPLICATION OF SOCIOECONOMIC SURVEY DATA Richard J. Agnello' ABSTRACT This paper focuses on an economic frameworl( for analyzing some of the elements in the management of marine recreational fisheries. In addition, estimates are provided for valuing fishing success to marine anglers targeting on three Atlantic coast species: bluefish. summer flounder, and weakfish. Demand functions for sport fishing are estimated with cross-section data using the travel cost method. Fishing trips per season are related to travel cost, fishing success, and income for individual fishermen. The marginal value of fishing success is determined using alternative models and estimation techniques. The data come from a one-time socioeconomic survey conducted by the National Marine Fisheries Service in 1981. The findings show that marginal valuations for fishing success as measured by the number of fish kept by fishermen vary considerably among target species. In addition, these marginal values are quite sensitive to the empirical formulation of the model. The findings provide managers with some objective basis for evaluating policies affecting marine recreational fisheries. The wide range of values computed from the same data set, however, should caution us, and indicates the need for more theoretical and applied economic research in this area. In order to efficiently manage marine recreational fisheries, information on economic valuations is re- quired. Since recreational fisheries are typically in the nonmarket sector, traditional markets do not provide much direct information on recreational values in total or at the margin. As a result, man- agement is hampered for recreational fisheries especially when attempting to evaluate activities which have potential effects on these fisheries. In recent years, many studies have been per- formed to determine economic valuations of changes in several dimensions of recreational experiences. Examples from a variety of areas include water quality (Bouwes and Schneider 1979; Desvousges et al. 1983), congestion levels on beaches (McConnell 1977), and harvest rates for himting (Miller and Hay 1981). For recreational fisheries, most studies tra- ditionally have focused on freshwater sports fish- ing where the data base is generally stronger. Ex- amples of empirical studies focusing on valuation of freshwater recreational fishing with emphasis on the importance of fishing success include Stevens (1966), Vaughan and Russell (1982), and Samples and Bishop (1985). In recent years, more attention has been directed towards saltwater recreational fish- eries (examples include McConnell and Strand 1981 and Thompson and Huppert 1987). 'Department of Economics, University of Delaware, Newark, DE 19716. Manuscript accepted October 1988. Fishery BuUetin, U.S. 87:223-232. Marine recreational fishing is particularly impor- tant because of its size and interactions with other sectors. It is estimated that more than 17 million marine anglers catch over 717 miUion pounds of fish and contribute over $7.5 billion dollars to the U.S. economy (U.S. Department of Commerce 1985). Al- though commercial marine harvests are consider- ably larger (6.3 billion pounds in 1985), conflicts between the two sectors are increasing and provide additional rationale for investigation into marine recreational valuation (Bishop and Samples 1980). In this paper we focus on an economic framework for analyzing some of the crucial elements in man- aging marine recreational fisheries. In addition, findings are presented which provide an empirical basis for valuing fishing trips and fishing success to marine anglers targeting on three Atlantic coast species: bluefish, Pomatomus saltatrix; summer flounder, Paralichthys dentatus; and weakfish, Cynosdon regalis. THEORETICAL BACKGROUND The management of recreational fisheries would be enhanced if the value of the fishing experience and the impact of fishing effort on the resource base (and, hence, the future value of the fishing experi- ence) were known. The former consideration in- volves measurement of economic demand which, for recreational fishing, can be complicated since mar- 223 FISHERY BULLETIN: VOL. 87, NO. 1 ket prices and quantities are generally not available. Even less well known is the impact that fishing ef- fort (both past and present) has on the resource base and, hence, the current and future value of the fish- ing experience (e.g., quantity and average size of catch, crowding, etc.). Effects of stock externalities have been studied extensively for commercial fish- eries and, although these externalities may exist for sport fisheries, little empirical evidence is available. ^ We present an economic methodology for valuing recreational fishing assuming no stock externalities. Of particular interest is to separate the value of the quantity of fishing (e.g., the number of trips) from the value of the quality or success of the fishing ex- perience (e.g., catch rate). Economic value can be derived from a demand relationship where the level -The stock externality results when increased fishing effort by individual participants affects the fish stock such that catch per day or average size of catch are adversely affected, and, hence, the value of a recreational fishing day for all participants is dimin- ished. (Anderson 1983.) or quantity (Q) demanded is related to price (P), in- come (I), and a vector of other relevant variables (S) including quality measures such as fishing suc- cess. The demand relationship is given as Q = f(P, I, S), (1) where P, I, and S are treated as exogenous in the individual's demand or consumption level decision. For recreational fishing, Q is usually measured as the number of fishing trips; P may reflect an entry price but more often is measured in terms of trip related costs; I reflects angler income (e.g., annual salary or hourly wage); and S reflects such things as fishing success and prices of substitute and com- plementary goods. Fishing success may be measured in terms of number and size of fish caught and/or kept. The model is graphically presented in Figure 1 with quantity (Q) and price (P) on the horizontal and vertical axes respectively. The relationship between P (e.g., distance travelled per trip) f(P, l,S2) = Dj f(P.I,S,) = D, Q (e y,, trips per season) Figure 1.— Demand model relating travel frequency (Q) and cost (P). 224 AGNELLO: ECONOMIC VALUE OF FISHING SUCCESS Q and P is embodied in the slope of the curve. Rela- tionships between Q and income (I) as well as other variables (S) can be shown as shifts in the demand curve. For simplicity and without loss of general- ity, let S represent single fishing success variable. For example, an increase in a relevant variable such as fishing success (S) from S, to So is shown as a shift in the demand curve from Dj to D, (i.e., from f(P, I, S,) to f(P, I, S2). The value of an improvement in site quality, such as an increase in fishing success, can be measured in various ways. A common approach is to compare the areas under each demand curve and evaluate an increase in fishing success as a difference in the area over some quantity range (Freeman 1979). For ex- ample, let us assume that in a particular year or season an individual consumes Qi units at a price of Pi when the level of fishing success is Sj (i.e., reflected by demand curve Dj). Suppose the level of fishing success increases to S, (e.g., during the next year or season). Given the demand shift to Do and the old price of Pj, the individual would now consume Q2. The economic valuation for the im- provement in success or site quality totaled for the fishing season or year is approximately measured as the sum of areas A, B, and C. These areas repre- sent an increase in consumer surplus for the fisher- man experiencing an increase in fishing success and, thereby, increasing the fishing level from Qi to Qo. An alternative approach to valuation of fishing success is to measure the instantaneous (or mar- ginal) change in welfare when fishing success changes (i.e., on a per-visit basis) rather than the accumulated gain over an entire season. This is the primary focus of our paper and can be accomplished in various ways. One approach is to convert the con- sumer surplus over an entire season into that of a single trip by dividing areas (A, B, C) by the num- ber of trips per season. A more direct approach can be accomplished by first solving Equation (1) for P. For the moment, let us assume that Equation (1) is deterministic (i.e., nonstochastic) in nature and can be inverted mathematically. Thus we can solve for P as P = g(Q, I, S). (2) Equation (2) is often referred to as the inverse de- mand function. The marginal value of fishing suc- cess a P/3 S can be measured as ag/9 S from Equa- tion (2) where did represents the partial derivative operator. In Figure 1, this may be viewed as the distance (P2 - Pi) when the number of fishing trips is Qi. This second approach will be the primary focus of the empirical analysis. It is more direct, has the advantage of less extrapolation from typical values of P and Q, and avoids any potential difficul- ty with an unbounded measure for area A arising with certain functional forms. DATA Since fishing trips and success are not commodi- ties bought and sold in the marketplace, data are not readily available on P, Q, and S. As a result, survey methods are usually used to generate data on the number (Q) and price (P) of fishing trips and fishing success (S). The two most common survey approaches for relating Q, P, and S for individual fishermen have been 1) to directly ask marine anglers for valuation estimates of hypothetical changes in fishing trip frequency and success, or 2) to impute implicit valuation or trade-offs based on the various cost and activity level responses of a cross section of marine anglers. The first approach is usually referred to as contingent valuation and has been employed in fisheries valuation. Recent studies using contingent valuation surveys which attempt to incorporate catch rate and site informa- tion include Cameron and Huppert (in press) and Cameron and James (1987). The second approach using the travel cost method focusing on individual marine anglers will be used in this study. The travel cost method, although not without pitfalls, has been widely accepted as a means for valuing recreational resources when distance for fishing trips is well defined. An early implementation of the travel cost method can be found in Clawson (1959). For a re- cent summary of the travel cost method and its com- plexities, see Kealy and Bishop (1986). The individual travel cost approach to evaluation relates travel cost and visitation frequency to rec- reational sites for individuals. This relationship pro- vides an indirect way of observing how individual visitation frequency might respond to changes in an entry or purchase price as in a traditional economic demand relationship. Thus, behavior of marine ang- lers with respect to travel cost, travel frequency, and site quality (e.g., fishing success) provides the basis for estimating a demand equation for marine recreational fishing. The parameters of Equation (1) and/or (2) can be estimated using cross-section data on individual anglers. In this study we are able to measure travel cost, travel frequency, success, and income variation for individuals from the Socioeconomic Survey con- ducted as a part of the Marine Recreational Fishery Statistics Survey (MRFSS) by the National Marine 225 FISHERY BULLETIN: VOL. 87. NO. 1 Fishery Service (U.S. Department of Commerce 1981). Since in this study we wish to investigate the value of success by fish species, we chose samples of fishermen who preferred one of the three species of fish (bluefish, summer flounder, or weakfish). These species are of considerable importance to managers developing plans for fisheries along the Atlantic coast and are fairly similar with respect to mode, sites, and season. Since the number of obser- vations for a given fishing site is generally quite low, and reduced further by focusing on specific fish species, pooling individual observations over sites was necessary in order to have enough interviews for statistical validity. Our sample sizes of anglers for bluefish, summer flounder, and weakfish are 270, 161, and 57 respectively and comprise sites from the Florida east coast to New York State. These data are pooled within a covariance statistical framework (i.e., with intercept and slope dummy variables) thus allowing the testing for differences across target species. Although the survey contains a large and useful set of economic information on marine recreational fishing, the data provided are by no means ideal for an application of the travel cost method. Certain enhancements to the travel cost method could not be performed due to lack of data.^ In addition, adjustments to travel distance and income were needed given the nature of the survey instrument.'' The actual survey questions providing the data base can be found in Table 1. A final point about the data base concerns the fish- ing success measure. Since trip frequency repre- sents activity over the past year, ideally one would like a measure of fishing success to be reflective of the last year and thus reflect ex ante or expected fishing success. Unfortunately, the survey provides no longitudinal information on individual anglers. The measure of success is only for the day of the interview and may not have been typical and, there- fore, inconsistent with the fisherman's past be- havior.'' We are forced to assume that ex post fishing success is a proxy variable for ex ante (expected) suc- cess. Travel frequency, distance, and fishing success thus reflect long-run equilibrium adjustment by the fishermen.* The empirical significance of fishing suc- cess reflects on both the importance of success to fishermen and the closeness of success realizations versus expectations. EMPIRICAL MODEL Trip demand for the ith fisherman is specified as a long-linear equation of either of the following forms: In Qi = bo -t- bi In P; -i- bg In Sj + bg In Ij -(- bZ -I- ej (3) 'Two refinements that are noteworthy, but could not be incor- porated into the analysis due to the lack of data, include time costs and multiple site substitutions. It has been argued that time spent travelling as well as time spent at the recreational site reflects op- portunity costs and should be included as part of the price of the fishing trip (Wilman 1980). The survey provides no information on travel nor visitation time. Multiple fishing sites can provide an opportunity to construct prices for recreational substitutes and, thus, include these vari- ables in the statistical estimation of the demand curve. See Samples and Bishop (1985) and Vaughn and Russell (1982). Unfortunately, no information on the angler's point of origin (e.g., ZIP code or area code) was available on the tabulated survey available to us so as to construct accurate distance (and price) measures for sub- stitute sites. 'Since travel distance is a proxy for travel costs associated vrith fishing, travel distance from a permanent home to the fishing site might overstate travel costs for those individuals who were part- year residents of the area, vacationers, or on business. For part- time residents and those on business, the distance from last night's accommodation rather than home was used as the appropriate measure of travel distance. For vacationers, who comprised around one-sixth of the sample, one-half of the distance from home was used as their fishing travel cost. Adjustments for the income variable included 1) assigning midrange values since respondents were asked for their income category rather as an actual dollar amount and 2) dealing with miss- ing data since the income question appeared on a follow-up tele- phone survey for which the response rate was approximately half that of the field survey. Missing observations were handled by the zero-order approach whereby means replace missing values (Mad- dala 1977). Since income is an exogenous control variable and not central to the valuation calculations, these procedures were felt to be acceptable. In Pj = ao -t- aj In Qj -i- a.^ In S; -I- a3 In Ij -(- aZ -I- V; (4) where P, Q, S, I > 0; and Qi = the number of site-specific fishing trips (including the day of the survey) made in the last 12 months (Table 1, question 16), Pj = round-trip cost in dollars to the site from either home or last night's accom- modation (Table 1, question 18, as mod- ^An attempt was made to improve the success measure by focus- ing only on fishermen for whom the fishing success on the day of survey could be considered normal. This was done by utilizing a satisfaction level variable (Table 1, question 23) and eliminating those observations whose satisfaction was very high or very low. By eliminating those individuals with extreme satisfaction, it was felt that those individuals for whom the day's catch was not nor- mal (or what was expected), would be eliminated from the sam- ple, (jnfortunately, the filter did not distinguish perfectly, and, in addition, reduced the sample to unacceptably low levels in part because satisfaction is measured on the follow-up telephone survey which had a lower response rate. The statistical results using this filter were less significant and, thus, the approach was abandoned. "The implication of these potential errors in measurement is that the coefficient of success will be underestimated to a degree de- pending on the ratio of the variance of the error in measuring true success over the variance of observed success. 226 AGNELLO: ECONOMIC VALUE OF FISHING SUCCESS Table 1 . — Survey questions used in the estimations. From intercept survey 16. Including today's trip, about how many times would you say you have fished from [this (specify exact mode) in the last 12 months?/a (specify exact boat mode) leaving from this launching area in the last 12 months?] 18. To the nearest highway mile, about how far is it from your home to this fishing location? 29. May I look at the fish that you caught that you're taking with you? Enter species codes and number kept. Did you land any (specify common name) that you're not taking with you? 30. How many additional (specify common name) did you land? From telephone survey 23. How satisfied were you with your fishing trip on ( Month/Day )? Would you say you were Very satisfied Somewhat satisfied Not too satisfied Not at all satisfied (1) (2) (3) (4) 28. Finally, how much do you estimate that you personally earned in 1980 before taxes? Would that be Less than $5,000 1 $5,000 to $10,000 2 $10,000 to $15,000 3 $15,000 to $25,000 4 $25,000 to $35,000 5 More than $35,000 6 (rescaled to 1987 dollars using the GNP price deflator). bZ, aZ Ci, Vi ified in the above discussion,' fishing success measured by the total number of fish Icept (Table 1, question 29), previous year's income of the respond- ent (Table 1, question 28), and vector products of additive and multi- plicative dummy variables and param- eters allowing pooling across species to be tested using a covariance model (Kmenta 1986), independent, identically distributed random errors. The log linear specification is used since it provides a better fit over linear and semilog models in terms of t-statistics and the equation F-statistics. Recent studies estimating travel cost models have also found that log models provide better fits to the data. The choice of functional form has received much at- tention in the literature. Discussions of some of the issues including utility consistency, benefit sensi- ti'vity, and transformed parameter biases can be 'Dollar valuations are obtained by assuming a driving cost of $0.16 per mile. This figure reflects a rescaling to 1987 dollars of estimates appearing in "Cost of Owning and Operating Automo- biles and Vans 1984," U.S. Department of Transportation, and in- cludes only variable driving costs averaged over several vehicle types. found in Bockstael et al. (1986), Stynes et al. (1986), and Ziemer et al. (1980). Whether Equation (3) or Equation (4) is the appro- priate model depends on the individual angler's choice process. If we assume that trip frequency (Q) is chosen after the site and thus travel cost (i.e., distance) is specified. Equation (3) is appropriate. If, on the other hand, anglers choose travel distance or cost (P) by choosing a recreational site after the frequency of visitation (i.e., the number of trips per year, Q) is determined, then Equation (4) is appro- priate. Most likely both Q and P are endogenous to an individual angler so that ideally a multiequation model should be estimated that would include many competing sites as well as determinants of residen- tial location choice. Unfortunately our data do not allow us to employ such a model.* In our empirical analysis Equations (3) and (4) are estimated as single equation models and compared. Although Equation (3) is standard in the literature 'Fishing success (S) also could be treated as an endogenous vari- able related to angler skill, experience, equipment, and the fish stock. An additional equation would be added to the model if one wished to "explain" S. The empirical approach would be affected depending on whether the model were simultaneous or recursive in nature. To the extent that fishing success (S) is related to travel frequency, Q (a proxy for experience), and travel cost, P, the model should be estimated as a simultaneous equation system. Unfor- tunately, additional variables required to adequately identify such a system are not available. 227 FISHERY BULLETIN: VOL. 87, NO. 1 (e.g., see Kealy and Bishop 1986), our focus on mar- ginal success valuations (i.e., 3P/3S) makes Equa- tion (4) more appropriate since no parameter trans- formations are necessary.' Equations (3) and (4) were estimated first by or- dinary least squares (OLS). Because the data are cross sectional on individual marine anglers, large variations in travel frequencies and cost exist which could lead to errors with unequal distributions. Vari- 'Two statistical issues are relevant in the context of choice of dependent the variable: 1 ) The choice of dependent variable (e.g. , In Q or In P) affects the regression slope unless the correlation (e.g., between In Q and In P) is perfect. Thus, estimating the In Q rela- tionship and solving for In P generally yields a different estimate for 3 In P/3 In Q than estimating the In P relationship directly. For a clear treatment of this point, see Wonnacott and Wonna- cott (1979). 2) In addition, we note that parameter unbiasedness generally does not hold under nonlinear transformation although consistency does. Thus, partial effects on P using Equation (4) are potentially both unbiased and consistent whereas when using Equa- tion (3) unbiasedness is lost for partial effects on P. ous tests for heteroscedasticity were performed on the OLS residuals including Park, Glejser, and Bruesch-Pagan tests. The results were mixed with some tests indicating insignificant heteroscedas- ticity and some indicating significant (0.05 level, two-tailed) relationships between OLS residuals and travel cost (In P) or travel frequency (In Q) in Equa- tions (3) and (4) respectively. Since the Glejser tests indicated the strongest relationship between the ab- solute OLS residual and the square root of In P or In Q in Equations (3) and (4) respectively, weighted least squares (WLS) was performed using l/\/X (i.e., where X is In P or In Q in Equations (3) and (4) respectively). The results are found in Tables 2 and 3 for both OLS and WLS applied to the demand frequency (Q endogenous) and demand price (P endogenous models). The variables trip frequency (Q), trip cost (P), fishing success (S), and income (I) were defined Table 2. — Log-linear demand frequency regressions (Equation 3). OLS = ordinary least squares; WLS = weighted least squares. Estimated coefficients (absolute t-values in parenth esis) OLS WLS Exogenous variable (1) (2) (3) (1) (2) (3) Constant 1.930 1.970 1.792 2.383 2.421 2.833 (2.17) (2.22) (1.59) (2.54) (2.57) (2.21) Log travel cost (P) -0.173 -0.171 -0,181 -0.096 -0.096 -0.149 (4.85) (4.74) (4.08) (1.87) (1.87) (2.02) Log fish kept (S) 0.050 0.048 0.074 0.034 0.034 0.055 (3.13) (3.05) (3.53) (2.28) (2.25) (2.72) Log income (1) -0.000 -0.005 0.020 -0.072 - 0,075 -0.100 (0.00) (0.06) (0.17) (0.78) (0.82) (0.78) Flounder (F) - 0.062 -0.124 -0.069 -1.799 (0.44) (0.06) (0.51) (0.91) Weakfish (W) 0.290 1.966 0.182 0.622 (1.29) (0.49) (0.85) (0.16) Interactions F and P (FP) 0.089 (1.00) 0.182 (1.66) F and S (FS) -0.079 (2.22) -0.071 (2.15) F and 1 (Fl) -0.031 (0.16) 0.112 (0.58) W and P (WP) -0.100 (0.85) -0.113 (0.62) W and S (WS) -0.020 (0.36) 0.010 (0.20) W and 1 (Wl) -0.154 (0.39) -0.012 (0.03) R^ 0.057 0.061 0.074 0.016 0.019 0,035 F (model) 9.71 6.28 3.45 2.66 1,86 1,56 F (species)' 1.03 1.08 0.74 1,31 n 488 488 488 488 488 488 ^Computed from the formula (Afl^) {n-k-1)/(1 - R^) (r) where r, R^, and (n-k-1) represent the number of restrictions, coefficient of determination, and degrees of freedom of the unrestricted model in hier- archial order (1), (2). and (3) respectively See Wonnacott and Wonnacott 1979 228 AGNELLO: ECONOMIC VALUE OF FISHING SUCCESS previously. The variables in the Z vector are defined in Tables 2 and 3. These variables reflect the addi- tive and interactive (multiplicative) dummy variables which allow us to test for parameter differences across target species. Since the control group in all regressions is bluefish (i.e., anglers indicating blue- fish as the species preference), qualitative (0,1) vari- ables for flounder (F) and weakfish (W), along with their interactions with other exogenous variables are included in each regression. F tests (noted as F (species) in Table 2 and 3) were performed on the interaction and additive dummy variable terms. For the demand frequency regressions (Table 2), since the F (species) statistics for both the additive and multiplicative terms are insignificant, the data can be combined across target species. Thus, model (1) for both OLS and WLS are most appropriate when using Table 2). In the demand price regressions (Table 3), the species terms have significant F- statistics (to at least the 0.05 level) indicating that intercept and slope coefficients are different across species. Thus, models OLS (3) and WLS (3) are most appropriate from Table 3. The empirical findings for the demand price model (Table 3) are stronger than for the demand frequen- cy model (Table 2) although both have significant equation F-statistics (probability values < 0.05). WLS increases the significance of the results in Table 3 but lowers significance levels in Table 2. The parameter estimates for the travel cost and frequen- cy coefficients (bj and aj < 0) as well as the success coefficients (bg and a, > 0) generally confirm theo- retical expectations. Travel cost and frequency are significantly inversely related, and fishing success as measured by the number of fish kept is general- ly a significant determinant of both fishing fre- quency and travel distance. Various measures and combinations of fishing success were investigated. Table 3.— Log-linear demand price regressions (Equation 4). OLS = ordinary least squares; WLS = weighted least squares. Estimated coefficients (absolute t-values in parenthesis) OLS WLS Exogenous variable (1) (2) (3) (1) (2) (3) Constant -0.310 (0.28) -0.442 (0.40) - 1 .892 (1,36) 0.372 (0.33) 0.225 (0.20) -1.314 (0.96) Log trip frequency (0) - 0.268 (4.85) -0.261 (4.74) 0.289 (4.20) -0.413 (5.59) -0.393 (5.38) - 0.433 (4.54) Log fish kept (S) 0.087 (4.45) 0089 (4.59) 0.113 (4.45) 0.088 (4.32) 0.095 (4.72) 0.135 (5.23) Log income (1) 0.260 (2.37) 0.253 (2.32) 0.408 (2.97) 0.228 (2.04) 0.215 (1.95) 0.388 (2.87) Flounder (F) 0.501 (2.86) 3.891 (1.64) 0.718 (3.84) 4.051 (1.66) Weakfish (W) 0.234 (0.84) 6.634 (1.33) 0,104 (0,38) 11,79 (2,28) Interactions F and P (FP) 0.182 (1.47) 0.243 (1.55) F and S (FS) -0.027 (0.63) -0.056 (1.25) F and 1 (Fl) -0.371 (1.58) -0.404 (1.68) W and P (WP) -0.316 (1.51) -0.632 (2.33) W and S (WS) -0.108 (1.57) -0.144 (2.10) W and 1 (Wl) -0.600 (1.23) -1 02 (2.07) 0= F (model) F (species)' n 0.087 15.42 488 0.103 11.02 4.30 488 0.126 6.24 2.13 488 0.097 17.41 488 0.125 13.74 7.71 488 0.164 8.49 3.70 488 'Computed from the formula (Afl^) (n-k-1)/(1 - H^) (r) where r, R^. and (n-k-1) represent the number of restrictions, coefficient of determination, and degrees of freedom of the unrestricted model in hier- archial order (1), (2), and (3) 229 including the number of fish caught as well as kept. These numbers were available in total as well as by species. Since the total number of fish kept con- sistently provided the best statistical fit, we report these results only.'" Our findings on income are mixed and appear to depend on the equation specification. While an im- portant theoretical variable in most demand func- tions, we find that income is a significant positive determinant of travel cost but not travel frequen- cy. Thus anglers with higher incomes travel greater distances but do not fish with greater frequency. This result is perhaps not surprising given the higher time opportunity cost for anglers with higher in- come. Our results for the lack of significant income effects on demand frequency are similar to findings in other studies (e.g., Vaughan and Russell 1982). The coefficients for travel cost (P), frequency (Q), and success (S) in Tables 2 and 3 provide the basis for valuing fishing success. The valuation algorithm is outlined below using the instantaneous (marginal) approach discussed in the paper. Of particular in- terest is the measurement of the marginal value of fishing success (3P/3S) shown as (P, - Pi) in Fig- ure 1)." We illustrate these calculations for sum- mer flounder using the WLS model (3) results from Table 3. Since the regression slope coefficients reflect log derivatives (sometimes referred to as elasticities or price flexibilities), we begin by noting that 3 In P 3 In S 3 P S 3 S P' (5) Solving this equation for 3 P/3 S provides a basis for valuing fishing success (S) using a log-linear model. 3P as 3 In P 3 In S P S' (6) For summer flounder 3 In P/3 In S = (0.135 - 0.056) = 0.079 which reflects the combination of the fish kept (S) term and the flounder and fish kepi (FS) interaction term. Evaluating P and S at their sam- '»The design of the survey may in part be responsible for the better fit with fish kept versus fish caught. Fishermen were asl P < 0.05), except that Polychaeta were consumed in higher numbers at Blueberry Pool. More Cumacea and Copepoda were consumed by fish in West Side Pool, but this trend at West Side Pool may be a reflection of fish size and the presence of Z. marina— excellent mlcrohabltat for Copepoda and Cumacea. These organisms were significantly more Important in the diet of fish <15 mm (x" = 32.0; P < 0.05), while Amphlpoda, Isopoda, and Polychaeta were more significant (x^ = 51.0; P < 0.01) In the diet of juve- niles >15 mm TL than in the diet of smaller fish. Algal Associations Juvenile lumpfish observed were from 6 to 50 mm long and included primarily fish of age 0, but also age 1, as judged by length-frequency graphs and other work of Cox and Anderson (1922) and Dabom and Gregory (1983). One juvenile, 80 mm TL (ap- parently from an older year class), was also col- lected. From 328 observations of algal attachments of lumpflsh during daylight hours, definite patterns emerged. Zoster a marina and 18 species of marine algae were identified from the three pools during a survey in July 1986. Juvenile lumpflsh were also found attached to an additional species, Rhodymenia Table 1.— Percent occurrence and percent of total weight of food items in tfie diet of juvenile Cycloplerus lumpus, and foods by pool location and size (percent occur- rence). West Pond data were not sufficient for inclusion in site comparisons, but are used in length comparisons; n = 1 50 for all three pools. Seven stomachs were empty (4.7%) and are not included in food data. Fish size range was 9-50 mm, with a length frequency peak of 15 mm. Total length of Occur- rence Pool fish (mm) Weight Blueberry West Side <15 >15 Item % % {n = 78) {n = 62) {n = 60) (n = 83) Amphlpoda 68 68 71 61 45 84 Copepoda 53 6 49 61 73 40 Isopoda 38 14 41 32 20 51 Cumacea 35 3 27 47 42 30 Acarina 20 1 18 23 20 19 Polychaeta 15 3 24 2 5 23 Mysidacea 3 1 3 5 3 4 Other items '<3 4 — — — — 'Other items, in order of decreasing percentage occurrence; Cladocera. fish eggs. Diptera, Littorina spp (Gastropoda), Mytilus edulis, algae, fish larvae, Caprellidea, dinoflagellates. cypris larval stages, megalops larval stages. 234 palmata, in a different intertidal locale on the Maine coast. Fish were found attached to 12 of these species of algae, Zostera marina, and 1 inverte- brate—the blue mussel, Mytiliis edulis. Only two lumpfish were encountered free-swimming, without a substrate or algal association, during these day- light observations. In 39% of the observations, juveniles were primarily associated with one of three species of Laminaria. In areas without Laminaria, but with Zostera marina, however, fish were fre- quently associated with Z. marina (Table 2). Associations with Laminaria were significantly higher (x^ = 251.4; P < 0.01) than with Z. marina but, because algal species composition varies with locale, associations were also analyzed by location (Table 2). In West Side Pool, which contained almost no Laminaria spp., 76% of the associations were with Zostera marina. No one algal species was domi- nant in West Pond Pool. In Blueberry Pool, Lami- naria is much more abundant (but <50% of algal surface area), and 67% of the associations were with that genus. As juvenile lumpfish increase in size, fewer were associated with Z. marina and more with Asco- phyllum nodosum (Table 2). The difference was significantly in favor of attachments to Z. marina for fish <19 mm, but significantly in favor of at- tachments to A. nodosum for lumpfish over 26 mm (P < 0.05, paired comparison t tests and chi-square tests). Areas containing Z. marina may thus be ex- tremely important to juveniles <20 mm long, but the protective function of the plant decreases as fish size increases. Discussion Juvenile lumpfish in Maine appear to use inter- tidal areas seasonally during more than one year of life. An array of sizes of C. lumjtus can be taken within a single tidepool (e.g., lengths of 9-49 mm from a single pool in August). Although most juve- niles in intertidal areas were age 0, fish of age 1 were not rare, and one fish collected was prob- ably age 2. An adult was also observed guarding a nest in a deep tidepool near Blueberry Pool in 1982. The food of juveniles is less diverse than that of adults (as reported by others), probably because the younger fish have smaller mouths and less ability to capture prey. The availability of larger prey items may also be limited in tidepools; ctenophores and coelenterates are generally uncommon in such waters. However, the consumption of copepods and amphipods by more than half the juveniles examined in this study coincides well with the studies of Daborn and Gregory (1983) of juvenile lumpfish in surface waters offshore. Although it is commonly believed that adult lumpfish feed only during winter (Cox and Anderson 1922; Collins 1976), the juveniles assuredly feed in summer: <5% of the stomachs that I examined were empty. The information presented here dealing with juve- nile lumpfish and algae are field observations of in situ associations. Given a choice between several genera or species of algae, the algal preference might be different. However, data from Blueberry Pool, where Z. marina and at least 13 species of algae were present, showed that 67% of the juvenile lumpfish were encountered with Laminaria spp., even though those three species made up less than one half of the submerged algal surface area (visual estimation). Because juvenile lumpfish are typically observed attached to marine algae or to Z. marina, the ques- tion remains why associations are with specific algae? There may be several possible explanations, including functional morphology of the fish species, coloration, hydraulics, and adhesion. Table 2— Algal and plant associations by Cyclopterus lumpus (%) by pool and total length. Taxon Laminaria Site or spp Fucus Ascophyllum Agarum Zostera size (n) (3 species) vesiculosus nodosum cribrosum marina Others Pool West Side (76) 13 5 76 6 West Pond (17) 29 29 24 2 16 Blueberry (150) 67 12 10 6 1 4 Total length (mm) <12(72) 34 13 4 5 33 11 13-18(84) 47 11 8 4 27 3 19-24 (54) 43 15 11 6 20 5 >25 (38) 49 13 18 8 12 235 Marine algae serve as arractants for invertebrates (Hicks 1986). Juvenile lumpfish are not rapid or ef- ficient swimmers, and thus cannot effectively pur- sue active prey. It would seem advantageous for such fish to live in concentrations of algae near con- centrations of invertebrates. Second, unlike some species of tidepool fishes (e.g., Oligocottus snyderi and Xererpes fucorum of the Pacific coast), juvenile lumpfish show only limited variations in color. The brown-orange color- ation of juveniles may explain why they prefer algal genera and species of similar coloration, such as Laminaria. This explanation does not hold for the large number of juveniles associating with Z. marina, which is green. However, the strong asso- ciation with Z. marina apparently holds only for small lumpfish which feed heavily on small crusta- ceans (Tables 1, 2). Brown (1986) recently found that small juveniles (about 10 mm long) spent more time attached to structures than was spent swimming. Algae of any type or color may thus be especially important to the smallest fish, particularly if avail- ability of brown algae is reduced in a particular locale (e.g.. West Side Pool, where 60% of the fish were 15 mm, compared with 41% in Blueberry Pool). As lumpfish size increases, there appear to be increasing associations with brown algae and decreasing associations with green-colored Z. marina, even when both types are present (Table 2). Third, Laminaria spp. can provide some protec- tion for fish from direct wave action, perhaps more than from other genera or algal species present (for general concepts, see Wieser 1952, O'Connor et al. 1979, and Seed 1986). Laminaria often occurs in clusters, resulting in a diffusing of wave action that would otherwise displace fishes. The distribution of L. saccharina. has been shown to be independent of exposure (Sze 1982). Lumpfish associated with this functional type of alga may be effectively protected at flood tides from full wave action, and at ebb tides from avian or terrestrial predators. Finally, the ability of lumpfish to attach to objects has been well documented; juveniles of the sizes col- lected in the tidepools of Schoodic Peninsula, ME can adhere to objects and withstand water speeds of up to 170 cm/s (Gibson 1969). Lumpfish use this attachment ability to avoid the adverse impacts of wave action (Alexander 1967). Suction efficiency would be improved by adherence to a flat, somewhat rigid surface, though several species of marine algae have smooth, nonrippled fronds, species of Lami- naria provide the most surface area of this type in the pools examined. Acknowledgments This research was funded by the National Geo- graphic Society, Grant 3145-85, The Nature Con- servancy (Maine Chapter), and the University of Maine through a Faculty Research Award. John H. Dearborn, Department of Zoology, University of Maine, and David A. Misitano, National Marine Fisheries Service, kindly reviewed the manuscript. Literature Cited Able, K. W., and W. Irion. 1985. Distribution and reproductive seasonality of snailfishes and lumpfishes in the St. Lawrence River estuary and the Gulf of St. Lawrence. Can. J. Zool. 63:1622-1628. Alexander, R. McN. 1967. Functional design in fishes. Hutchinson Univ. Libra- ry. Hutchinson and Co., London, 160 p. BENFEY, T. J., AND D. A. Methven. 1986. Pilot-scale rearing of larval and juvenile lumpfish (Cyclopterus lum-fms L.), with some notes on early develop- ment. Aquaculture 56:301-306. BiGELOW, H. B., AND W. C. ScHROEDER. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53:1-577. Blacker, R. W. 1983. Pelagic records of the lumpsucker, Cyclopterus lum- pus L. J. Fish Biol. 23:405-417. Brown, J. A. 1986. The development of feeding behaviour in the lumpfish, Cyclopterus lumpus. J. Fish Biol. 29 (Suppl. A):171-178. Collins, M. A. J. 1976. The lumpfish (Cyclopterus lumpus L.) in Newfoundland waters. Can. Field-Nat. 90:64-67. Cox, p., and M. Anderson. 1922. A study of the lumpfish (Cyclopterus lumpus L.). Con- trib. Can. Biol. 1:1-20. Daborn, G. R., AND R. S. Gregory. 1983. Occurrence, distribution, and feeding habits of juvenile lumpfish, Cyclopterus lumpus L. in the Bay of Fundy. Can. J. Zool. 61:797-801. FORSMAN. B. 1970. Smadjur i fast och drivande tang. Zool. Revy 32:3-8. Fulton, T. W. 1907. On the spawning of the lumpsucker (Cyclopterus lum- pus) and the paternal guardianship of the eggs. Ann. Rep. Fish. Board Scotl., Part 111:169-178. Gibson, R. N. 1969. Powers of adhesion in Liparis rrumtagui (Donovan) and other shore fish. J. Exp. Mar. Biol. Ecol. 3:179-190. Goulet, D., J. M. Green, and T. H. Shears. 1986. Courtship, spawning, and parental care behavior of the lumpfish, Cyclopterus lumpus L., in Newfoundland. Can. J. Zool. 64:1320-1325. Gregory, R. S., and G. R. Daborn. 1982. Notes on adult lumpfish Cyclopterus lumpus L. from the Bay of Fundy. Proc. Nova Scotia Inst. Sci. 32:321- 326. Hicks, G. R. F. 1986. Meiofauna associated with rocky shore algae. In P. G. Moore and R. Seed (editors), The ecology of rocky shores, p. 36-56. Columbia Univ. Press, N.Y. 236 MORING, J. R. 1985. Intertidal areas of northern New England: nursery habitat for coastal fishes. (Abstr.) Proc. Gulf Maine Work- shop, Assoc. Res. Gulf Maine, Portland, ME, p. 27. O'Connor, R. J., R. Seed, and P. J. S. Broaden. 1979. Effects of environment and plant characteristics on the distribution of Bryozoa in a Fiicus serratus L. community. J. Exp. Mar. Biol. Ecol. 38:151-178. PiNKAS, L., M. S. Oliphant, and I. L. R. Iverson. 1971. Food habits of albacore, bluefin tuna, and bonito in California waters. Calif. Dep. Fish Game, Fish Bull. 152, 139 p. Procter, W. 1933. Biological survey of the Mount Desert region. Part V. Marine fauna. Wistar Inst. Anat. Biol., Philadelphia, 402 p. Seed, R. 1986. Ecological pattern in the epifaunal communities of coastal macroalgae. In P. G. Moore and R. Seed (editors), The ecology of rocky coasts, p. 22-35. Columbia Univ. Press, N.Y. SZE, P. 1982. Distributions of macroalgae in tidepools on the New England coast (USA). Bot. Mar. 25:269-276. Wieser, W. 1952. Investigations on the microfauna inhabiting seaweeds on rocky coasts. IV. Studies on the vertical distribution of the fauna inhabiting seaweeds below the Plymouth labora- tory. J. Mar. Biol. Assoc. U.K. 31:145-173. Yarrell, W. 1841. A history of British fishes. 2nded. John Van Voorst, London, Vol. II, 628 p. John R. Moring U.S. Fish and Wildlife Service Maine Cooperative Fish and Wildlife Research Unit ' Department of Zoology, University of Maine Orono, ME 0Ue9 'Cooperators are the University of Maine, Maine Department of Inland Fisheries and Wildlife. U.S. Fish and Wildlife Service, and the Wildlife Management Institute. 237 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/her affiliation, and mailing address, including 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 footnotes should be tyi)ed 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 need- ed 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 arable 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 Data Base. {Chemical Abstrax;ts also uses this system, which was developed by the American Standards Association.) Common abbreviations and symbols, such as mm, m, g, L, mL, mg, °C (for Celsius), %, Voo, and so forth, should be used. Abbreviate units of measure only when used with numerals. Periods are only rarely used with abbreviations. 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 arable numeral and heading provided). LIST OF FIGURES (Entire figure legends) FIGURES (Each figure should be numbered with an arable numeral; legends are desired) ADDITIONAL INFORMATION Send the ribbon copy and two duplicated or carbon copies of the manuscript to: Dr. Andrew E. Dizon, Scientific Editor Fishery Bulletin Southwest Fisheries Center La Jolla Laboratory National Marine Fisheries Service, NOAA P.O. Box 271 La Jolla, CA 92038 Fifty separates will be supplied to an author free of charge and 50 supplied to his organization. No covers will be supplied. _„^....»ias^---»»i;- (Contents— ConHmied) ,.;... .if! ..■^\K^^ \ ^HIGHTOWER, JOSEPH E, and GARY D. GROSSMAN. Status of the tilefish, Lopho- ^ latilus chamaelepnticepes, fishery off South Carohna and Georgia and recommenda- tiong for management 177 SMITH,; INGRID.'iMARK S. FONSECA, JOSE A. RIVERA, and KEITH A. RITT- MASTER. Habitat value of natural versus recently transplanted eelgrass, Zostera marina, for the '.bay scallop, Argopecten irradians 189 ROPES, JOHN W. ; The food habits of five crab species at Pettaquamscutt River, Rhode '.., island. ; .:rr::: 197 CRAWFORD, MAURICE K, CHURCHILL B. GRIMES, and NORMAN E. BUROKER. / Stock identification of weakfish, Cynoscion regalis, in the Middle Atlantic region . . . 205 GARTNER, JOHN V., JR., WALTER J. CONLEY, and THOMAS L. HOPKINS. Escape- ment by fishes from midwater trawlers: a case study using lanternfishes (Pisces: Myctophidae) 213 AGNELLO, RICHARD J. The economic value of fishing success: an application of socioeconomic survey data 223 Notes MORING, JOHN R. Food habits and algal associations of juvenile lumpfish, Cyclopterus lumpfus L., in intertidal waters 233 • GPO 791-008 .,<°>^, ^'Atcs of •■ Fishery Bulletin Vol. 87, No. 2 April 1989 UTTER, R, G. MILNER, G. STAHL, and D. TEEL. Genetic population structure of Chinook salmon, Oncorhynchus tshawytsclia, in the Pacific Northwest 239 JACKSON, GEORGE DAVID. The use of statolith microstructures to analyze life- history events in the small tropical cephalopod Idiosepius pycmaeus 265 / BAILEY, KEVIN. Description and surface distribution of juvenile Peruvian jack mackerel, Trachurus murphyi, Nichols from the Subtropical Convergence Zone of the central South Pacific 273 TUCKER, JOHN., JR. Energy utiUzation in bay anchovy, Anchoa mitchilli, and black sea bass, Centropristis striata striata, eggs and larvae 279 SHERIDAN, PETER P., REFUGIO G. CASTRO M., FRANK J. PATELLA, JR., and GILBERT ZAMORA, JR. Factors influencing recapture patterns of tagged penaeid shrimp in the western Gulf of Mexico 295 TEGNER, MIA J., PAUL A. BREEN, and CLERIDY E. LENNERT Population biology of red abalones, Haliotis rufescens, in southern California and management of the red and pink, H. corrugata, abalone fisheries 313 SQUIRES, DALE, SAMUEL R HERRICK, JR., and JAMES HASTIE. Integration of Japanese and United States sablefish markets 341 MULLEN, ASHLEY J. Aggregation of fish through variable diffusivity 353 NOTES SZELISTOWSKI, WILLIAM A. and JUAN GARITA. Mass mortality of sciaenid fishes in the Gulf of Nicoya, Costa Rica 363 BARSHAW, DIANA E. Growth and survival of early juvenile American lobsters, Homarus americanus, on a diet of plankton 366 VIDAL, OMAR, and GARY PECHTER. Behavioral observations on fin whale, Balaenoptera physaliis, in the presence of killer whale, Oreimis orca 370 Notice Marine Biological Laboratory r LIBRARY ' OCT 1 8 1989 Woods Hole, Mass. 374 Seattle, Washington U.S. DEPARTMENT OF COMMERCE Robert Mosbacher, Secretary NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION William E. Evans, Under Secretary for Oceans and Atmosphere NATIONAL MARINE FISHERIES SERVICE James W. Brennan, Assistant Administrator for Fisheries Fishery Bulletin The Fisiiery Bidletin 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 quarter- ly. 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. Andrew E. Dizon Southwest Fisheries Center La JoUa Laboratory National Marine Fisheries Service, NOAA P.O. Box 271 La Jolla, CA 92038 Editorial Committee Dr. Jay Barlow National Marine Fisheries Service Dr. William H. Bayliff Inter-American Tropical Tuna Commission Dr. George W. Boehlert National Marine Fisheries Service Dr. Bruce B. CoUette National Marine Fisheries Service Dr. Robert C. Francis University of Washington Dr. James R. Kitchell University of Wisconsin Dr. William J. Richards National Marine Fisheries Service Dr. Tim D. Smith National Marine Fisheries Service Mary S. Fukuyama, Managing Editor The Fwhery 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 Office, Washington, DC 20402. Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated. The Secretary of Commerce has determined that the publication of this periodical is necessary in the transaction of the public business required by law of this Department. Use of funds for printing of this periodical has been approved by the Director of the Office of Management and Budget. Fishery Bulletin CONTENTS Vol. 87, No. 2 April 1989 UTTER, F., G. MILNER, G. STAHL, and D. TEEL. Genetic population structure of Chinook salmon, Oncorhynchus tshaurytscha, in the Pacific Northwest 239 JACKSON, GEORGE DAVID. The use of statolith microstructures to analyze life- history events in the small tropical cephalopod Idiosepius pycmaeus 265 BAILEY, KEVIN. Description and surface distribution of juvenile Peruvian jack mackerel, Trachurus murphyi, Nichols from the Subtropical Convergence Zone of the central South Pacific 273 TUCKER, JOHN., JR. Energy utilization in bay anchovy, Anchoa mitchilli, and black sea bass, Centropristis striata striata, eggs and larvae 279 SHERIDAN, PETER R, REFUGIO G. CASTRO M., FRANK J. PATELLA, JR., and GILBERT ZAMORA, JR. Factors influencing recapture patterns of tagged penaeid shrimp in the western Gulf of Mexico 295 TEGNER, MIA J., PAUL A. BREEN, and CLERIDY E. LENNERT Population biology of red abalones, Haliotis rufescens, in southern California and management of the red and pink, H. corrugata, abalone fisheries 313 SQUIRES, DALE, SAMUEL F HERRICK, JR., and JAMES HASTIE. Integration of Japanese and United States sablefish markets 341 MULLEN, ASHLEY J. Aggregation of fish through variable diffusivity 353 NOTES SZELISTOWSKI, WILLIAM A. and JUAN GARITA. Mass mortality of sciaenid fishes in the Gulf of Nicoya, Costa Rica 363 BARSHAW, DIANA E. Growth and survival of early juvenile American lobsters, Homarus americanus, on a diet of plankton 366 VIDAL, OMAR, and GARY PECHTER. Behavioral observations on fin whale, Balaenoptera physalus, in the presence of killer whale, Orcinus orca 370 Notice 374 Seattle, Washington 1989 Marine Biological Laboratory LIBRARY OCT 1 8 1989 For sale by the Superintendent of Documents. U.S. Government Printing Ofrae. Washington DC 20402— Subscription price per year: $16.00 domestic and $20.00 foreign. Cost per single issue: $9.00 domestic and $11.25 foreign. | Woods Hole, Mass. 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. GENETIC POPULATION STRUCTURE OF CHINOOK SALMON, ONCORHYNCHUS TSHAWYTSCHA, IN THE PACIFIC NORTHWEST F. Utter,' G. Milner.i G. StIhl,^ and D. Teel' ABSTRACT Variation at 25 polymorphic protein coding loci was examined for 86 populations of chinook salmon, Oncorhynchus tshaun/tscha, ranging from the Babine River in British Columbia to the Sacramento River in California. Substantial differences in allele frequencies identified patterns of genetic variability over the geographic range of the study. The following nine major genetically defined regions were formu- lated: 1) the Fraser River tributaries east of the Cascade Crest (no downstream drainages were sam- pled). 2) Georgia Strait, 3) Puget Sound, 4) a broad coastal region ranging from the west coast of Vancouver Island southward through northern California, 5) the Columbia River below The Dalles Dam, 6) the Columbia River above The Dalles Dam, 7) the Snake River. 8) the Klamath River, and 9) the Sacramento River. Populations sampled within a region tended to be genetically distinct from each other although they exhibited the general patterns of variability that defined the region. Within a region there was little distinction among populations returning to spawn at different times. The persistence of these geographic patterns in the face of natural opportunities for introgression, and sometimes massive transplantations, suggests that genetically adapted groups within regions have resisted large-scale introgression from other regions. Repopulation of deglaciated areas in the Fraser River, Georgia Strait, and Puget Sound ap- parently occurred from multiple sources; most likely sources included Columbia River populations and northern refuges rather than from the large coastal group of populations. Patterns of genetic distribu- tion of chinook salmon differed from those of other anadromous salmonids studied within this region. A conservative policy for stock transfers was suggested based on distinct genetic differences observed both between and within regions. Population studies of chinook salmon, Oncorhynchus tshawytscha, based on electrophoretically detected genetic variation have been carried out since the late 1960s. As data have accumulated, an increasingly clear picture of the breeding structure of this species has emerged. While early investigations based on only a few polymorphic loci identified differences among populations, they failed to identify any geo- graphic trends (e.g.. Utter et al. 1973; Kristiansson and Mclntyre 1976). Differences within and among drainages became apparent as additional polymor- phic loci were found and a more comprehensive sampling of populations was made (Utter et al. 1976, 1980; Gharrett et al. 1987). This paper outlines the genetic structure of chinook salmon in the Pacific Northwest using allele frequency data collected for the purpose of esti- mating the stock composition of ocean caught Chinook salmon (Milner et al. 19813; 19834. MiUej. et al. 1983; Utter et al. 1987). Our purpose is to ex- amine these data in the light of other relevant bio- logical and historical information 1) to understand genetic relationships among chinook salmon popu- lations better and 2) to provide biologists with new insights to assist in the preservation and manage- ment of this important biological resource. MATERIALS AND METHODS Our data were obtained from samples of juvenile or adult fish collected at 86 locations ranging from British Columbia through California (Table 1, Fig. 1). These data include allele frequencies from 25 protein-coding loci with sample sizes between 38 and 200 individuals. Data were accumulated between 1980 and 1984 and were reported in part in Milner et al. (fn. 3, 4). Electrophoretic procedures followed those de- 'Coastal Zone and Estuarine Studies Division. Northwest and Alaska Fisheries Center, National Marine Fisheries Service, NOAA, 2725 Montlake Boulevard East, Seattle, WA 98112. ^Department of Genetics. Stockholm University, S-10691, Stock- holm, Sweden. 'Milner, G. B.. D. J. Teel. F. M. Utter, and C. L. Burley. 1981. Columbia River stock identification study: Validation of genetic Manuscript accepted January 1989. Fishery Bulletin. U.S. 87:239-264. method. Report to Bonneville Power Administration under con- tract DE-A179-80BP18488, 51 p. Available Bonneville Power Ad- ministration, P.O. Box 3621, Portland, OR 97208. 'Milner, G. B., D. J. Teel, and F. M. Utter. 1983. Genetic stock identification study. Report to Bonneville Power Administration under contract DE-A179-82BP28044M001, 95 p. Available Bonne- ville Power Administration, P.O. Box 3621, Portland, OR 97208. 239 lecw I50°W FISHERY BULLETIN: VOL. 87, NO. 2, 1989 ISCW I20°W - ecN - 50° N - 40° N 30' N Figure 1.— Locations of sample collections based on map code of Table 1. 1. Total range of sampling identifying general locations or drainage systems. 2. Oregon (OR) coast. 3. California (CA). 4. Columbia River. 5. Georgia Strait, British Columbia (BC) coast, and Fraser River. 6. Washington (WA) coast and Puget Sound. 240 UTTER ET AL.: GENETIC POPULATION STRUCTURE OF CHINOOK SALMON Table 1 —Chinook salmon collections made from British Columbia through California. Map codes refer to Figure 1. Samples representing a hatchery stock are marked by $. Locations followed by (a) represent adult samples; all other samples were from juvenile fish. Season of return Iden- tifies the time of entry by adults Into freshwater. Pooled samples are Indicated by hyphens. Season Map Area or Sample of code Location of samples drainage system Region size return A1 Bablne (a) Skeena R. Inland 39 Summer B1 Tete Jaune (a) Fraser R Inland 38 Summer B2 Clearwater (a) Fraser R. Inland 45 Summer B3 Chlico Fraser R. Inland 49 Summer B4 Stuart-Nechako (a) Fraser R. Inland 105 Summer CI $Blg Quallcum Georgia Strait Coastal 85 Fall C2 SPuntledge (a) Georgia Strait Coastal 100 Fall C3 $Qulnsam (a) Georgia Strait Coastal 97 Fall C4 $San Juan Georgia Strait Coastal 50 Fall C5 SCapilano Georgia Strait Coastal 99 Fall C6 Nooksack, south fork Georgia Strait Coastal 50 Spring C7 Nooksack, north fork Georgia Strait Coastal 50 Spring D1 SRobertson Ck. (a) British Columbia coast Coastal 100 Fall E1 swells Dam Upper Columbia R. Inland 50 Summer E2 $Carson-$Leavenworth' Upper Columbia R Inland 148 Spring E3 SWInthrop Upper Columbia R Inland 129 Spring E4 $Prlest Rapids Upper Columbia R. Inland 100 Fall F1 $Soleduck Washington coast Coastal 100 Summer F2 SSoleduck Washington coast Coastal 100 Spring F3 SNaselle Washington coast Coastal 99 Fall F4 SHumptulips Washington coast Coastal 50 Fall F5 SQulnault Washington coast Coastal 100 Fall F6 Queets Washington coast Coastal 120 Fall F7 Hoh Washington coast Coastal 100 Fall F8 SSoleduck Washington coast Coastal 50 Fall F9 SEIwha Washington coast Coastal 100 Fall G1 SSkykomlsh Puget Sound Coastal 100 Summer G2 SSkagIt Puget Sound Coastal 100 Summer G3 SHood Canal Puget Sound Coastal 98 Fall G4 SDeschutes Puget Sound Coastal 150 Fall 05 SGreen R.-SSammlsh Puget Sound Coastal 149 Fall HI SCowlitz-SKalama Lower Columbia R. Coastal 100 Spring H2 SLewls R. Lower Columbia R. Coastal 50 Spnng H3 SCowlltz (a)-SKalama Lower Columbia R. Coastal 149 Fall H4 SLewls R. Lower Columbia R. Coastal 50 Fall H5 SWashougal R. Lower Columbia R. Coastal 50 Fall H6 SEagle Ck.-$McKenzie R. Lower Columbia R. (Willamette) Coastal 88 Spring 11 SKIIckltat R. MId-ColumbIa R. Inland 50 Spring 12 SSpring Ck.-SBIg Ck.^ MId-ColumbIa R. Inland 150 Fall 13 SWarm Sprlngs- SRound Butte (a) MId-ColumbIa R. Inland 109 Spring 14 Deschutes (a) MId-ColumbIa R. Inland 49 Fall J1 Ice Harbor (a) Snake R. Inland 200 Fall J2 McCall-Johnson Ck. Snake R. Inland 106 Summer J3 SRapId R. -Valley Ck.=' Snake R. Inland 165 Spring K1 SCole R -Hoot Owl Ck. Oregon coast Coastal 163 Spring K2 SRock Ck. Oregon coast Coastal 100 Spring K3 SCedar Ck. Oregon coast Coastal 99 Spring K4 STrask R. Oregon coast Coastal 100 Spring K5 Chetco Oregon coast Coastal 100 Fall K6 Lobster Ck. Oregon coast Coastal 50 Fall K7 SEIk R. Oregon coast Coastal 100 Fall K8 Sixes R estuary Oregon coast Coastal 100 Fall K9 Coqullle R. estuary Oregon coast Coastal 115 Fall K10 Sluslaw Bay Oregon coast Coastal 82 Fall K11 SSalmon R, Oregon coast Coastal 99 Fall K12 SNestucca R.-SAIsea R.'' Oregon coast Coastal 346 Fall K13 SCedar Ck. Oregon coast Coastal 100 Fall K14 STrask R. -Tillamook Bay Oregon coast Coastal 188 Fall K15 Nehalem estuary Oregon coast Coastal 141 Fall 'Includes Little White Salmon. ^Includes Little White Salmon ^Includes Sawtooth and Red River. ^Includes Siletz Estuary and $Fall Creek. 241 FISHERY BULLETIN: VOL. 87, NO Table a.— Continued. Season Map Area or Sample of code Location of samples drainage system Region size return LI $Feather R. Sacramento R. Coastal 50 Spring L2 $Coleman-$Nimbus Sacramento R. Coastal 300 Fall L3 SFeather R.-$Mokelumne Sacramento R. Coastal 200 Fall M1 STrlnity R. Klamath R. Inland 50 Spring M2 $lron Gate Klamath R. Inland 99 Fall M3 $Trinity R. Klamath R. Inland 100 Fall scribed in Aebersold et al. (1987). Buffer systems included the following: 1) a Tris-boric acid, EDTA system (pH 8.5) (Boyer et al. 1963); 2) an amine (3-aminopropyl morpholine) citrate system (pH 6.5) (Clayton and Tretiak 1972); and 3) a discontinuous Tris-citric acid (gel pH 8.15), lithium hydroxide, boric acid (electrode pH 8.0) system (Ridgway et al. 1970). Methods for visualizing enzyme activity followed Siciliano and Shaw (1976) and Harris and Hopkin- son (1976). A system of nomenclature suggested by Allendorf and Utter (1979) was used to designate loci and alleles. The 25 polymorphic loci (Table 2) were selected from a larger set of loci known to be variable in Chinook salmon. Variable loci were excluded when data were unavailable for one or more of the sam- pling locations listed in Table 1. Much of the de- scriptive data for the loci and alleles were previ- ously reported (Utter et al. 1980; Milner et al. fn. 4). Two previously unreported polymorphic en- zymes in Chinook salmon, Gr and Gpi-l(H), were used for population studies and are described in the appendix. Allele frequencies were calculated directly from phenotypic classes for 14 nonduplicated loci. Tests for departures of genotypes from the expected binomial distribution (Hardy-Weinberg equilibrium) were made using a G statistic (Sokal and Rohlf 1969) with degrees of freedom equaling the number of ex- pected genotypes minus the number of alleles. The isoloci Aat-1,2; Idh-3,4; Mdh-1,2; Mdh-3,4; and Pgm- 1,2 (see Allendorf and Thorgaard 1984) were ex- cluded from such tests because every individual was scored on the basis of four allelic doses from two loci. Combined allele frequencies of both loci were calculated directly from phenotypic expressions and were assumed to be the same at both loci for statis- tical calculations. The data for the Gpi-2 locus and the Gpi-l(H) allele were also excluded from Hardy- Weinberg calculations because common homo- zygotes and heterozygotes could not be reliably distinguished, and allele frequency estimates were Table 2.— Background information on Chinook salmon tissue samples for protein coding loci. Buffer Refer- Protein name and enzyme number Locus Tissue' system ence^ Aconitate hydratase (4.2.1.3) Adenosine deaminase (3.5.4.4) Aspartate aminotransferase (2.6.1.1) Dipeptidase (3.4.13.11) Glucose-6-phosphate isomerase (5 3 19) Glutathione reductase (16.4.2) Isocitrate dehydrogenase (1.1.1.42) Lactate dehydrogenase (1.1.1.27) Malate dehydrogenase (1.1.1 .37) ru1annose-6-phosphate isomerase (5.3.1.8) Phosphoglucomutase (2.7.5.1) Phosphoglycerate kinase (2.7.2.3) Superoxide dismutase (1.15.1.1) Tripeptide aminopeptidase (3.4.11.4) Ah L 2 1 Ada-1 E,H,M 1 1 Aat-1,2 H,M 1 1 Aat-3 E 1 1 Dpep-1 E,H,M 1.3 1 Dpep-2 E 1,3 1 Gpi-1 M 3 2 Gpi-2 M 3 1 Gpi-3 M 3 1 Gr E,M 1,3 2 Idh-3,4 E,L.H.M 2 1 Ldh-4 E.L,M 1 1 Ldh-5 E 1 1 Mdh-1,2 L,H,M 2 1 Mdh-3,4 E,H,M 2 1 Mpi e.L.H.M 1 1 Pgm-1,2 E,L,H,M 2 1 Pgk-2 E,L,M 2 1 Sod L 1 1 Tapep-1 E.H.M 3 1 21 liver, E = eye, H = heart. M = muscle. Milner et al 1983. 2 = variation described in this study 242 UTTER ET AL.: GENETIC POPULATION STRUCTURE OF CHINOOK SALMON based on the frequency of homozygotes for the respective variant alleles. Expected heterozygosities were calculated for polymorphic loci. Pairwise com- parisons were made for all loci between all popu- lations by a contingency table analysis using a G statistic. Two or more sample collections lacking significant allele frequency differences for any poly- morphic locus were considered a single population. All subsequent analyses were performed on the resulting 65 individual and pooled populations. A critical value of 1% was used (for both the Hardy- Weinberg and the pairwise population comparisons) to reduce the erroneous rejection of the null hypoth- esis when using multiple tests. Nei's (1975) measure of genetic distance (D) was used to compare pair- wise levels of genetic divergence between individual or pooled populations. A dendrogram based on a matrix of these comparisons was constructed by the unweighted pair group method (UPGM) (Sneath and Sokal 1973). Principal component analysis of the allele frequency data followed procedures outlined in Sneath and Sokal (1973). A nested gene diversity analysis followed procedures described by Nei (1973) and Chakraborty (1980) and was performed through the NEGST computer program described by Cha- kraborty et al. (1982). RESULTS AND DISCUSSION Tests for Hardy-Weinberg Equilibrium Tests for significant deviations from Hardy-Wein- berg proportions were made on each of the 86 data sets for 14 loci including Ah, Ada-1, Aat-3, Dpep-1, Dpep-2, Gpi-1 (excluding the subsequently described Gpi-l(H) allele affecting heterodimer formation), Gpi-3, Gr, Ldh-4, Ldh-5, Mpi, Pgk-2, Sod-1, and Tapep-1. Six deviations were observed (Table 3). These deviations probably were random errors ex- pected from the 1,204 independent calculations at the 1% level of significance. The direction of the deviations indicates both excesses and deficits of heterozygotes in both instances where the same Table 3.— Populations and loci with significant (o = 0.01) depar- tures from expected Hardy-Weinberg proportions. Population Level of Excess Vdeficit- (map code) Locus significance of heterozygotes Queets (F6) Dpep-1 0.005 -f Humptulips (F4) Mpi 0,01 - Washougal (H5) Mpi 0.005 + Lobster Creek (K6) Pgk-2 0,005 + Eagle Creek (H6) Sod-1 0.005 + Stuart (B4) Sod-1 0.0001 - locus is involved (Mpi and Sod-1). Two of the popu- lations having significant deviations. Eagle Creek and Stuart, were combined for subsequent analysis with other populations having Hardy-Weinberg pro- portions; combinations were based on overall non- significant differences of allele frequencies. The high significance of the Stuart sample for Sod-1 is in- flated through an expected value less than unity for the homozygous genotype of the (-260) allele. Description of Allelic Distribution The allele frequencies observed for all 25 polymor- phic loci over the geographic range of this study (Appendix) indicate considerable heterogeneity among loci with regard to levels of variation and geographic distribution. This variation is summar- ized from three perspectives— heterozygosity, fre- quency range for common allele, and index of gene diversity (G^, ) (Table 4). Heterozygosity measures within-population variation. Those loci having higher heterozygosities have greater potential for diver- gence of allele frequencies among populations. Mean heterozygosity over all loci was 0.102, and five loci (Ah, Mpi, Pgk-2, Sod-1, Tapep-1) exceeded 0.200. The range of allele frequencies and the index of gene diversity reflect the actual divergences ob- served among populations. The range is a simple identification of allele frequency extremes. The in- dex of gene diversity is a quantitative measure of genotype deviations of the overall data set from those expected in a single panmictic population. Seven of the eight most heterozygous loci (Pgk-2, Mpi, Sod-1, Ah, Tapep-1, Gpi-2, Dpep-1) were among the eight loci having either the greatest range in frequency or the highest index of gene diversity, indicating considerable genetic differences among the populations samples. Typically, adjacent popula- tions tended to have allele frequencies more similar to one another than to those from other areas (see Appendix). Notable examples include the follow- ing: 1) restriction of Gpi-2 variation largely to coastal populations from Vancouver Island through Oregon, 2) the highest frequency of the Gpi-l(H) allele in populations from the Sacramento River, 3) Aat-3 variation that is largely restricted to popula- tions from Georgia Strait and western Vancouver Island, 4) low frequencies of variant alleles for most loci in all Klamath River populations and in spring and summer run populations from the Snake River, and 5) high frequencies of Tapep-1 variants in Puget Sound populations. Two procedures for graphic analysis (a dendro- gram [Fig. 2] based on pairwise genetic distance 243 FISHERY BULLETIN: VOL. 87. NO. 2, 1989 Table 4.— Outline of frequency range for common alleles, heterozygosity, and diversity for 25 polymorphic loci of Chinook salmon sampled from British Columbia through Cali- fornia. Single entries for isoloci assume identical allele frequencies for individual loci. Loca- tions and areas are based on map codes of Table 1 and Figure 1 . Only areas are iden- tified when one or more populations of an area have a maximum value of 1.000. Both locations and areas are identified for maximum values less than 1 .000. Frequency range for common allele (location and area) Heterozy- Diversity Locus Minimum Maximum gosity (Gst) Ah 0.366 (C3) 1,000 (l,J,M) 0.232 0.091 Ada-1 0.865 (G1) 1,000 (C-F,H,I,K-M) 0.044 0,045 Aat-1.2 0.888 (G3) 1,000 (A-C,E,F,H-J,L,M) 0.030 0,035 Aat-3 0.735 (C2) 1,000 (B,E-M) 0.030 0.143 Dpep-1 0.652 (K3.K9) 1,000 (B,D,E,J,M) 0.164 0.116 Dpep-2 0.939 (B3) 1 ,000 (A-M) 0.004 0.045 Gpl-1 0.576 (L1) 1,000 (A-K,M) 0.040 0,245 Gpi-2 0.432 (K5) 1,000 (A-C.E-I,K-M) 0.169 0,310 Gpi-3 0.875 (B3) 1.000 (C,E-M) 0.022 0,060 Gr 0.420 (H6) 1.000 (CD, F,G,I,K-M) 0.068 0,215 ldh-3,4 0.862 (E2) 1.000 (B.CK.M) 0.080 0,040 Ldh-4 0.933 (B2) 1 .000 (A-M) 0.009 0,037 Ldh-5 0.964 (E4) 1 .000 (A-M) 0.008 0,017 IVIdh-1,2 0.945 (K8) 1 .000 (A-M) 0.003 0,023 l^dh-3,4 0.843 (C3) 1.000 (A-C,H,K,M) 0.040 0,025 Mpi 0.386 (H4) 0.990 (M3) 0.401 0,089 Pgm-1,2 0.903 (K3) 1 ,000 (A-M) 0.031 0,041 Pgk-2 0.062 ;J2) 0.931 (H6) 0.420 0,153 Sod-1 0.530 (12) 0.990 (M2) 0.345 0,086 Tapep-1 0,483 (G5) 1.000 (B.M) 0.226 0,134 Average 0.724 0.996 0.102 0,123 measures, and plots of principal component scores) assist in identifying patterns of allelic variability. The approximate location of each population is iden- tified in Figure 2 on the basis of its inclusion in one of eight clusters (diverging beyond a genetic dis- tance of 0.01) or major subgroupings (below a genetic distance of 0.01). A notable feature of Figure 2 is the geographic basis for much of the aggrega- tion. For instance, clusters 1 and 2 represent down- stream populations of the Columbia River, cluster 3 contains the two northernmost populations of Georgia Strait, and cluster 4 is comprised of coastal populations from Vancouver Island southward through Oregon. The nine population units shown in Figure 2 are explained in the following section and represent a synthesis of possible relationships among these 65 populations. The two plots of principal components (Fig. 3) pro- vide an alternative picture of the allelic variation based on different perspectives of the total variance in a multidimensional space. The first four principal components (PC), which account for almost 80% of the total genetic variation, also project a geographic picture of this variation in these plottings. Six of nine population groupings (described in the next sec- tion) are essentially resolved by PCI and PC2. Two of the remaining units are resolved by PCS and PC4. We used three different hierarchies in the gene diversity analysis to give a more detailed examina- tion beyond the data on gene diversity presented in Table 4 (Table 5). The hierarchies based on geo- graphic and temporal clusters are discussed at this point; the hierarchy based on population unit clusters is discussed following the synthesis of these units. The geographic hierarchy was based on the locations of the samples using two regions (inland and coastal) with six areas within the inland region and seven areas wdthin the coastal region (see Table 1). The within-population component of gene diver- sity (i.e., the mean average heterozygosity) in each hierarchy was 87.7% of the total diversity (i.e., the expected heterozygosity based on the mean allele frequencies). The remaining 12.3% of the total diver- sity was the index of gene diversity, G(st) resulting from population subdivision (see also Table 5). Most of the gene diversity in the geographic hierarchy was due to genetic differences between populations within areas (4.6%) and areas within regions (6.2%). The regional component contrasting inland popula- tions of major drainages with populations from downstream tributaries and coastal drainages contributed only 1.5% of the total diversity. By far the largest portion of subdivision in the temporal 244 UTTER ET AL.: GENETIC POPULATION STRUCTURE OF CHINOOK SALMON 12 3 4 5A 5B 6 7aA8B 8C8D 8E MaiOf dendrogram clusters ¥ Figure 2.— Dendrogram and nine population units formulated from allele frequency data of this study. Popula- tions are approximately located by numbered squares which identify membership in clusters on the superimposed genetic distance dendrogram. An exception is the most northern location of Unit I (Babine River) which lies beyond map range. Dotted line represents maximum glaciation during late Pleistocene (McPhail and Lindsey 1986). 245 FISHERY BULLETIN: VOL. 87, NO. 2, 1989 0.8 0.6 0.4 0.2 -0.2 -0.4 - -0.6 -0.8 Unit VII Unit VIII Unit I 13 H2 G2 Gl B1B2 B4)C1e3 '\ 06.,^,." El K6 F2C1 £4 G5 Jl C5 G3/I2\ ,, ,, 14 ^ \UnitV Unit IX _L _L _L _L -1.0 -0.8 -0.6 -0.4 -0.2 PC 1 (33%) 0.2 0.4 0.6 0,( 0.6 p 0.4 - 0.2 - _ ^ -0.2 o -0.4 -0.6 -0.8 -1,0 E3 11 '" " J2 F2 "' M2 i3 J1 F8 Ml E4 15? 12 M3 E3 K1 _L _L -L. J -1.0 -0.8 -0.6 -0.4 -0,2 0,2 0,4 0,6 0,8 PC 3 (14%) Figure 3.— Plots of scores for principal components 1 c& 2 and 3 & 4 derived from allele frequencies in the Appendix. Major contributing loci include PCI - Mpi, Pgk-2. Sod, and Tapep-1; PC2 - Ah, Dpep-1, Gpi-2, and Pgm-1,2; PC3 - Aat-3, Gr, and Tapep-1; PC4 - Mpi. 246 UTTER ET AL : GENETIC POPULATION STRUCTURE OF CHINOOK SALMON Table 5— Summary of distribution of relative gene diversity of chinook salmon in geographical and temporal fiierarcfiies based on 65 individual or pooled populations and 25 polymorphic loci. Areas, regions, and seasons are given for each population in Table 1 . Absolute values of gene diversity include mean average hetero- zygosity (Hs) - 0.1018 and total diversity (Ht) - 0.1161. Population unit Geographic clusters % Temporal clusters % clusters % Within populations 87.7 Within populations 87.7 Within populations 87.7 Between populations, Between populations. Between populations, within areas 4.6 within seasons 11.4 within units 4.4 Between areas. within regions 6.2 Between seasons 0.9 Between units 7.9 Between regions 1.5 Total 100.0 100.0 100.0 hierarchy resulted from differences between popula- tions within seasons (11.6%), with only 0.7% of the total diversity being due to differences between seasons. Interpretation of Observed Variation We interpreted the overall data set primarily as a reflection of patterns and levels of gene exchange among populations. This interpretation does not ex- clude the possibility of some selective forces influ- encing the frequencies of some alleles and genotypes in some environments (e.g., Powers et al. 1983; Mork et al. 1984). However, empirical data from diverse animal species justify an assumption of predominant neutrality (Ihssen et al. 1981; Chakra- borty et al. 1980; Eanes 1987). This assumption is strengthened when many polymorphic loci are ex- amined and is particularly pertinent in anadromous salmonids where restricted population sizes accentu- ate the influence of genetic drift (Utter et al. 1980). The data presented here indicate that chinook salmon consist of a genetically complex network of populations throughout the geographic range of this study. This information yields some clear conclu- sions and suggests a number of additional possibil- ities that must await confirmation or rejection from additional studies. One conclusion is that the time of return (i.e., season) is not a major factor in establishing relation- ships of stocks among areas. Both the geographic clustering in Figures 2 and 3 and the small between- seasons component of the temporal gene diversity analysis point away from the concept of a recent common ancestry of fish returning at the same time in different areas. This finding comes as no surprise based on published data of other anadromous salmonids (e.g., rainbow trout, Allendorf and Utter 1979). However, it is still commonly accepted that the chinook salmon is separated into temporally distinct "races" (e.g., McClane 1978). Although a strong genetic component for the time of return has been clearly demonstrated in anadromous salmonids (e.g., Helle 1981), and this is not debated here, it appears that genetic divergence into temporally distinct units tends to occur within an area from a common ancestral stock of chinook salmon. In contrast to the lack of evidence for genetic structuring of time of return, a geographic basis for genetic structuring is apparent. The relatively large area component of gene diversity (over half of the between-population diversity in column 1 of Table 5) coupled with the predominantly geographic clusterings warrant an attempt to define different geographically discrete population units. Most imits (Figure 2 and Appendix) incorporate one or more of the areas or drainage systems listed in Table 1. Inevitably, overlap occurs between these formulated population units and the a priori groupings of areas or drainages. The Fraser River grouping (unit I) is necessarily limited to the upstream areas because no down- stream populations were sampled. The single sam- ple from the Babine River, tributary to the Skeena River and adjacent to drainages of the upper Fraser River, is also tentatively included in Unit I. The Babine population aggregates with those of the Fraser River in the dendrogram (cluster 5A) and the plots of PCI and PC2. Most populations of Unit I (including the Babine) are distinguished by the pres- ence of the Gr-110 allele at a mean frequency of 0.05. This allele was not included in the Appendix or in most analyses because of the incomplete data sets from some coastal populations. The Gr-(llO) allele was not observed in other populations that aggre- gate in the dendrogram and PC plots with those of unit I; these populations include the San Juan River (southern Vancouver Island), the spring- and sum- mer-run fish of the Snake River, and the Klamath River. 247 FISHERY BULLETIN: VOL. 87, NO. 2. 1989 The population unit of Georgia Strait (unit 11) com- prises populations forming clusters 3 and 6 in the dendrogram, plus the San Juan River population. These six populations aggregate adjacently in the plottings of PCS and PC4. Populations of Unit II typically have relatively high allele frequencies of Aat-3 (90), Pkg-2 (90), and Tapep-1 (130), although exceptions occur at each locus. Carl and Healey (1984) reported similar high frequencies for allelic variations of Aat-3 and Tapep in a study of chinook salmon populations of the Nanaimo River which flows from Vancouver Island into Georgia Strait. Populations in the Puget Sound unit (Unit III), bounded to the north by the population from the South Fork of the Nooksack River, aggregate fair- ly clearly in both the dendrogram (clusters 8D and 8E) and the plots of PC3 and PC4. The cohesiveness among the fall-run populations vary likely reflects both genetic isolation and present (or very recent) gene flow through transfers among hatcheries. Like unit II, populations of unit III also have high allele frequencies for Tapep-1 130; in fact, it has the high- est mean frequencies for this allele among the nine population units that were formulated. However, the mean frequencies of the common (i.e., 100) alleles for Aat-3 and Pgk-2 are much higher in unit III than in unit II. No influence of reported transfers of lower Columbia River fish to Puget Sound hatcheries (e.g., Ricker 1972) is apparent from the graphic projec- tions or the allelic data. An extended grouping of coastal populations (unit IV) ranges from northern California (see Utter et al. 1980) to Robertson Creek on the west coast of Vancouver Island. Populations of the Columbia, Klamath, and Sacramento Rivers are excluded from unit IV. This unit is distinguished by high frequen- cies of the Gpi-S (60) allele and (in most instances) some Pgm variation. Most populations appear either in clusters 4 or 8B of the dendrogram and aggregate distinctly in the plottings of PCI and PC2. Two populations are retained in Unit IV for geographic consistency which do not congregate with other populations of this unit; the spring run returning to the Soleduck River on the Washington coast, and the Lobster Creek population returning to the upper Rogue River on the Oregon coast. The outlying of the Soleduck spring-run population appears to be related to its heterogeneous origins. Records in- dicate that this run originated from crosses of fish from the Cowlitz River (lower Columbia River) and Umpqua River (Oregon coast) with some contribu- tion from the spring run of the Dungeness River, a drainage entering the Strait of Juan de Fuca (C. Johnson^). An explanation for the outlying of the Lobster Creek population is less apparent and re- quires further investigation. Two individual and four paired hatchery popula- tions sampled from the lower Columbia River form a geographically and genetically discrete unit (unit V). This group represents the most divergent pair of clusters (1 and 2) on the dendrogram and general- ly aggregates distinctly in the plotting of PCI and PC2. Populations of unit V are particularly distin- guished by high allele frequencies of (yr (85) and Mpi (109). Unit V is bounded upstream by the U.S. Fish and Wildlife Service Spring Creek Hatchery popu- lation (Spring Creek Hatchery is located on the pool impounded by Bonneville Dam). The pairing of four of the six populations is consistent with high levels of gene flow resulting from an extensive history of transplantation among the populations of the lower Columbia River (Simon 1972; Howell et al. 1985). This group's distinctness from other groups is also consistent with a minimal impact of transplantations of these populations beyond the lower Columbia River on indigenous populations in other areas (e.g., Cowlitz spring-run fish to the Snake River, C. Burley^; Kalama fall-run fish to Puget Sound, men- tioned above). The upper Columbia River unit (unit VI)— more than any of the other groupings— is composed of genetically diverse elements placed together more on the basis of geographic convenience rather than genetic unity. Unit VI is somewhat loosely bounded downstream by populations of the Klickitat and Deschutes Rivers; both rivers enter the Columbia River near The Dalles Dam. Unit VI's component populations include individuals of mixed ancestral origins, along with others of presumably pure line- age. Two populations known to have mixed ances- tral origin are those of the U.S. Fish and Wildlife Service Carson and Leavenworth Hatcheries. The Carson Hatchery population (located on the Wind River which drains into the Bonneville pool) was derived from interceptions of spring-run fish destined for areas of the upper Columbia and Snake Rivers. The Leavenworth population (combined with Carson in the analyses) has been largely maintained by continued infusions from fish of the Carson Hatchery (Howell et al. fn. 5). The Ice Harbor pop- ulation—another group of mixed ancestral origins- is composed of fall-run fish destined for different areas within the Snake River that were intercepted at Ice Harbor Dam near the mouth of the Snake '^C. Johnson, Washington Department of Fisheries, General Ad- ministration Bldg., Olympia, WA 98504, pers. commun. May 1985. «C. Burley, U.S. Fish and Wildhfe Service, 9317 Highway 99, Vancouver, WA 98665, pers. commun. May 1985. 248 OTTER ET AL.: GENETIC POPULATION STRUCTURE OF CHINOOK SALMON River. This population is included in unit VI because of its geographic proximity and genetic similarity to populations of unit VI contrasted with its dis- tinctness from spring- and summer-run populations of the upper Snake River. Populations of purer lineage within unit VI aggre- gate within cluster 8 of the dendrogram. The spring- run population returning to the Lewis River lies geographically within unit V, entering the Colum- bia River below Bonneville Dam. This population is included in unit VI because it is genetically distinct from other downstream populations and more typical of certain spring- and fall-run fish within Unit VI (i.e., Klickitat, Deschutes, and Winthrop popu- lations) with which it closely aggregates on the den- drogram (cluster 8C) and the plots of PCI and PC2. The similarity of the populations from Wells Dam and Priest Rapids Dam in unit VI is presumably a reflection of the two groups being different temporal segments of the same major run. All fish migrating past Priest Rapids Dam prior to 13 August are per- mitted to pass upstream and sequentially constitute the spring- and summer-runs of the upper Colum- bia River. The latter segment of this migration ar- riving at Wells Dam is captured and spawned for hatchery production. Most arrivals at Priest Rapids Dam later than 14 August are intercepted and spawned there (Chris Carlson'). This process inevi- tably results in considerable gene flow between these two artificially maintained populations. The Snake River unit (unit VII) contains the two combined populations of McCall Hatchery-Johnson Creek and Rapid River Hatchery-Valley Creek- Sawtooth-Red River, all managed by the Idaho Department of Fish and Game; all populations are from the Salmon River drainage of central Idaho. This unit is distinguished by very low average heterozygosities (see Winans in press) and by high frequencies of the Pgk-2 (90) allele. The Klamath River populations (unit VIII) are geographically isolated from, but genetically similar to those of the Snake River. However, populations of unit VIII lack variation of Idh-3,4 contrasted with a mean frequency of 0.925 for the Idk-S.J, (100) allele in unit VII. Klamath River populations, like those of unit VII, are characterized by very low aver- age heterozygosities. This characteristic contrasts sharply with most adjacent coastal populations for which the highest heterozygosities among all popu- lations are observed. Allele frequency data from the Shasta and Scott river populations, two wild pop- 'Chris Carlson, Grant County Public Utility District, P.O. Box 878, Ephrata, WA 98823, pers. commun. March 1986. ulations of the Klamath River are statistically iden- tical with frequencies in the Iron Gate Hatchery sample; these data were recently collected which precluded their use in most of the analyses of this study. Thus the low heterozygosity of Klamath River populations cannot be attributed to effects of hatch- ery management (see Allendorf and Ryman 1987). The three samples from the Sacramento River drainage form a distinct geographic and genetic unit (unit IX). These samples cluster together in the den- drogram (cluster 7) and in PCI and PC2. As men- tioned above, these populations are distinguished by high frequencies of the Gpi-l(H) allele. An analysis of gene diversity within and between the nine proposed population units (Table 5, column 3), provides further support for the reality of these genetic subdivisions. It is appropriate that almost two-thirds of the total gene diversity due to popula- tion structuring (7.9/12.3 = 64.2%) occurred be- tween the population units. Furthermore, the diversity between populations within the units was smaller than the diversity between populations within areas (Table 5, column 1) calculated prior to the synthesis of the units. Relationships and Origins of Population Units The common genetic and geographic attributes of populations within units have been stressed, but relationships between units also require considera- tion. The geographic areas of the Fraser River, Georgia Strait, and Puget Sound (units I, II, and III) were completely glaciated during the late Pleisto- cene, and therefore must have been entirely re- populated within roughly the last 15,000 years (McPhail and Lindsey 1986). Those areas of the Columbia River sampled in this study were outside of the ice sheet, although the upper third of the drainage was glaciated. However, downstream pop- ulations (units V and VI) were doubtlessly affected by massive runoffs and temporary impoundments resulting from sudden releases of glacial Lake Mis- soula initially occurring some 18,000 years ago (Bunker 1982); most of the Snake River drainage (unit VII), entering the mid-Columbia River from the south, was presumably unaffected by these events above its lower reaches. The coastal region (Unit IV) from the Chehalis River (Washington) southward, and the entire Sacramento-San Joaquin River drainage (unit HI), were likewise free of glaciation during the late Pleistocene. Much of the presently observed genetic diversity almost certainly existed during the Pleistocene. The 249 FISHERY BULLETIN: VOL, 87, NO, 2, 1989 broad geographic range and high heterozygosity of the coastal populations support the long-term exist- ence of unit IV in which cohesiveness among popu- lations appears to have been maintained through some degree of gene flow (Soule 1976; Campton and Utter 1987). Ecological as well as geographic bar- riers to extensive gene flow from the coastal area apparently existed in the Columbia, Klamath, and Sacramento drainages. However, the presence of the Grpi-2(60) allele— typical of coastal populations— in some populations of units V, VI, VIII, and IX suggests some degree of introgression from coastal populations. Natural obstructions of the mid-Columbia River such as Cascade Falls and CelUo Falls (presently obscured by Bonneville and The Dalles Dams, respectively) may have restricted migration between populations of the lower Colum- bia River and those of the upper Columbia and Snake Rivers. The relationship of the Snake River populations of unit VII to other groups within and beyond the Columbia River is unclear. Its most distinguishing feature is its very low average heterozygosity (H = 0.04), an attribute shared with the Klamath River populations (H = 0.029) (unit VIII) with which it also aggregates in the dendrogram and the principal component projections. In spite of this similarity, we favor an explanation that both Snake River and Klamath River populations had independent origins. The high frequencies of common (i.e., 100) alleles over the present sampling of loci are interpreted as reflecting loss of variation through genetic drift ac- centuated by periodic bottlenecks and restricted gene flow (see also Winans in press). This explana- tion is consistent with the inland locations of both drainages. In addition, both drainages continued to flow within their present courses during the Pleis- tocene. Thus, similarity is presently interpreted as an artifact based on minimal allelic variation de- tected over most of the loci sampled. However, drift coupled with isolation should lead to divergent fre- quencies of some alternate alleles with an adequate sampling of variable loci. If such differences are not observed as additional genetic marks continue to be detected in chinook salmon, then a zoogeographical explanation based on gene flow or recent ancestry must be pursued for Snake River and Klamath River populations. Following glacial regression, the newly habitable regions appear to have been repopulated from diverse sources. Origins of the northern portions of the coastal unit can be readily explained by immi- grations from more southern coastal streams. How- ever, populations of units I, II, and III apparently arose from other sources based on their virtual absence of Gpi-2 variation. Seeding of the Fraser River from sources including the upper Columbia River and Snake River units, and of Georgia Strait and Puget Sound drainages from the lower Colum- bia River or Alaska, are possibilities that seem more likely. The Aat-3 (85) allele is recorded in most Alas- kan populations studied by Gharrett et al. (1987) at frequencies up to 0.32. The highest frequencies of this allele occur in populations from Vancouver Island suggesting immigration from northern refugia. Comparisons with Sympatric Salmonid Species It is of interest to compare the present data set with similar information from other anadromous salmonid species within the same geographic range. These species presently share habitats and have been subjected to the same geological processes throughout their periods of common habitation. Thus, some common patterns of genetic population structuring may be anticipated. However, differ- ences among species in life histories and long-term distributions may likewise result in unique popula- tion structures. Similar data sets have been collected from four species within this range: rainbow trout, Salmo gairdneri; coastal cutthroat trout, S. clarki; chum salmon, 0. keta; and sockeye salmon, 0. nerka. Investigations of rainbow trout include both anad- romous (i.e., steelhead) and nonmigratory popula- tions (Huzyk and Tsuyuki 1974; Allendorf 1975; Allendorf and Utter 1979; Allendorf et al. 1980; Busack et al. 1980; Chilcote et al. 1980; Parkinson 1984; Wishard et al. 1984). A geographic basis for population structure is also apparent in this species and allelic similarities persist among indigenous populations of a particular region regardless of migratory tendencies, times of migration, or local environments. Apparent population units for chinook salmon and rainbow trout differ, however. A single major population unit of rainbow trout com- prising the upper Fraser River, the upper Colum- bia River, and the Snake River contrasts with at least three distinct groupings for chinook salmon. A clear distinction between coastal streams of Washington and Oregon from those of the lower Columbia River, Puget Sound, and Georgia Strait is also not apparent in rainbow trout as it is in chinook salmon. Distribution of sockeye salmon over the geograph- ic range of this study is less continuous than that of chinook salmon because of the more stringent 250 UTTER ET AL.: GENETIC POPULATION STRUCTURE OF CHINOOK SALMON ecological requirements of sockeye salmon during their freshwater life history. This irregular distribu- tion is accompanied by greater geographic hetero- geneity of allelic distributions, perhaps reflecting severe founder events and restricted gene flow (Utter et al. 1984). One population of sockeye salmon on the Quinault River (Washington coast) deviated strongly from all other groups sampled, but the possibility of a coastal unit of sockeye salmon, anal- ogous to that of Chinook salmon (i.e., unit IV), appears unlikely. Allele frequencies from Lake Ozette on the Washington coast (W. K. Hershber- ger*) were typical of noncoastal populations. Popula- tions north of the Skeena River (approximately the position of "A" in Figure 1) are distinguished by the presence of Ldh-4 variation which is virtually ab- sent from more southern groups (Utter et al. 1980; Withler 1985), presumably reflecting postglacial repopulation from a more northern refuge. Studies of population groups of chum salmon and coastal cutthroat trout within Puget Sound and Georgia Stait suggest similar genetic structures to that observed in chinook salmon. Populations of chum salmon from south Puget Sound were distin- guishable from those of north Puget Sound and Georgia Strait (Okazaki 1981). Populations of Georgia Strait and the lower Fraser River were likewise distinguishable from populations immedi- ately north of Georgia Strait (Beacham et al. 1985). Intensive subsampling of cutthroat trout within Hood Canal and north Puget Sound indicated strong and consistent differences between these regions (Campton and Utter 1987). More comprehensive comparisons will be possible as data accumulate on these and other species of anadromous salmonids. Both the similarities and the differences observed are of considerable interest in gaining further insights into the determinants of allele frequency variation, zoogeography, behavior, and management of these species. Effects of Hatchery Operations Further consideration of the effects of hatchery operations is also warranted. Hatchery operations and transplanted hatchery fish do not appear to have drastically altered the geographic distributions of protein coding alleles among the major population units. There is presently little question that hatchery operations have homogenized allele frequencies among many fall chinook hatcheries of the lower Columbia River (Simon 1972). However, the tem- porally isolated spring and fall populations of this region retain a greater similarity to one another than to populations of other regions. Thus it seems probable that the allele frequencies of unit V approx- imate those existing prior to the present century in spite of this region's large predominance of hatch- ery fish. Hatchery populations established from (and still reflecting) exotic origins (e.g., Carson and Leavenworth Hatcheries) have not noticeably per- turbed the allelic distributions of adjacent popula- tions having indigenous origins (Utter et al. 1987^). Where they exist (e.g., unit IV), indigenous wild and hatchery populations within a unit are generally separated by small genetic distances, reflected by close aggregations in the dendrogram and principal component clusters. Infrequent alleles do not strongly affect genetic distance or heterozygosity, but their loss in hatch- ery stocks relative to comparable wild populations is a good indication of an inadequate number of spawning individuals used to establish or maintain a hatchery stock (Allendorf and Ryman 1987). A comparison was therefore made of the average number of alleles per locus and heterozygosity between seven hatchery and six wild samples from the Oregon coast, the most extensive collection of hatchery and wild samples within a restricted geo- graphic range made in this study (two statistically indistinguishable combined populations each involv- ing a hatchery and a wild sample were excluded). The mean values were very similar (heterozygos- ity— hatchery 0.137, wild 0.132; alleles per locus- hatchery 1.74, wild 1.68) and were not significant- ly different. Presumably, sufficient numbers of breeders have been used in Oregon coastal hatch- eries to prevent losses of heterozygosity or alleles. However, the data provide no information concern- ing possible losses of genetically distinct geographic or temporal segments as a result of hatchery prac- tices along the Oregon coast. The present data set also pertains to additional aspects of hatchery management. Evidence con- tinues to accumulate from numerous sources that individual populations of anadromous salmonids represent gene pools that are uniquely adapted to a particular location and spawning time (see Ricker 1972). Stocks transferred to areas beyond those to which they are locally adapted perform poorly »W. K. Hershberger, Univ. of Washington, Seattle, WA 98195. pers. commun. December 1985. 'Utter, F., P. Aebersold, M. Griswold, G. Milner, N. Putas, J. Szeles. D. Teel, and G. Winans. 1987. Biochemical genetic vari- ation of chinook salmon stocks of the mid-Columbia River. Pro- cessed Report 87-19, 22 p. Northwest and Alaska Fisheries Center, Seattle, WA 98112. 251 FISHERY BULLETIN: VOL. 87, NO. 2. 1989 relative to indigenous populations (Withler 1982; Altukhov and Salmenkova 1987; Reisenbichler 1988). Transfers from maladapted populations not only waste effort and resource, but also carry the risk of disrupting locally adapted genomes through interbreedings (Reisenbichler and Mclntyre 1977; Shields 1982). Sets of data such as those reported here are valuable in outlining at least the maximum distribution of locally adapted gene pools and there- by provide guidelines for stock transfers. In the absence of any other data, it would be inadvisable to translocate populations between sites such as the lower Columbia River and the Washington or Ore- gon coasts. Stock transfers within major genetic units should also be performed with caution. Each of the indivi- dual or pooled populations within the nine units is also genetically distinct for some loci sampled in this study from other populations within the unit; they are therefore divergent from such populations at a much larger number of additional loci throughout the genome. It is pertinent to recall that a consid- erable amount of the total gene diversity results from population subdivision (4.4/12.3 = 35.8%) resided within the population units (Table 5, column 3). Likewise, slight or no divergence between two populations based on samplings of polymorphic pro- tein-coding loci does not necessarily mean these populations are identically adapted (discussed in Utter 1981). For example, two groups of rainbow trout in the Snake River drainage having similar allele frequencies at five polymorphic loci are adapted to drastically different local environments and life history patterns (Wishard et al. 1984). CONCLUDING OBSERVATIONS Three points require emphasis following this ini- tial outline of population units. First, it warrants restating that each of the nine units represents a genetically heterogeneous grouping. It is important that this heterogeneity be recognized and main- tained within the respective units. Second, these units are based on limited data within the range of sampling and, in some instances, on arbitrary decisions; the units are intended to be modified as more information accumulates and therefore to serve as guidelines for further inves- tigation. For purposes of clarification, allelic data beyond those listed in the Appendix have been intro- duced at various places in the text. Additional alleles and polymorphic protein-coding loci are continual- ly being identified through ongoing investigations, and further clarification is inevitable as these data accumulate. Genetic data other than from protein- coding loci are accumulating on chinook salmon populations within the geographic range of this study. Such genetic data show differences among populations in mitochondrial DNA (E. Berming- ham'"), and life history variables (Nicholas and Hankin 1988; Schreck et al. 1986), and provide com- plementary insights that will ultimately result in a much more detailed understanding of genetic struc- turing of these chinook salmon populations. Third, numerous distinct population units exist in North America beyond the sampling area of this study (e.g., Gharrett et al. 1987) and nothing is known of Asiatic populations. The nine units pre- sented here, then, are viewed as a necessary part of a much more complete picture of the genetic structure of chinook salmon that will ultimately emerge. ACKNOWLEDGMENTS Assistance in sampling was provided by person- nel of agencies including the Canadian Department of Fisheries and Oceans, California Department of Fish and Game, Oregon Department of Fisheries and Wildlife, and Washington Department of Fish- eries. Valuable technical assistance was provided by P. Aebersold. Valuable reviews were provided by C. Mahnken, G. Winans, A. Gharrett, Northwest and Alaska Fisheries Center and three anonymous reviewers. LITERATURE CITED Aebersold. P. B., G. A. Winans. D. J. Teel. G. B. Milner, AND F. M. Utter. 1987. Manual for starch gel electrophoresis: A method for the detection of genetic variation. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 61, 19 p. Allendorf, F. W. 1975. Genetic variability in a species possessing extensive gene duplication: Genetic interpretation of duplicate loci and examination of genetic variation in populations of rainbow trout. Ph.D. Thesis, Univ. Washington, Seattle, 98 p. Allendorf, F. W., D. M. Espeland, D. T. Scow, and S. Phelps. 1980. Coexistence of native and introduced rainbow trout in the Kootenai River drainage. Proc. Mont. Acad. Sci. 39: 28-36. Allendorf, F. W., and N. Ryman. 1987. Genetic management of hatchery stocks. In N. Ryman and F. M. Utter (editors). Population genetics and fishery management, p. 141-159. Univ. Wash. Press, Seattle. Allendorf, F. W.. and G. H. Thorgaard. 1984. Tetraploidy and the evolution of salmonid fishes. In B. Turner (editor), Evolutionary genetics of fishes, p. 1-53. '"E. Bermingham, NMFS, 2725 Montlake Boulevard East, Seattle, WA 98112, pers. commun. November 1987. 252 UTTER ET AL-: GENETIC POPULATION STRUCTURE OF CHINOOK SALMON Plenum Press, N.Y. Allendorf, F. W., and F. M. Utter. 1979. Population genetics. In W. S. Hoar, D. J. Randall, and R. Brett (editors). Fish physiology. Vol. 8, p. 407-454. Acad. Press. N.Y. Altukhov, Y. p., and E. a. Salmenkova. 1987. Stock transfer relative to natural organization, man- agement, and conservation of fish populations. In N. Ryman and F. M. Utter (editors). Population genetics and fishery management, p. 33-343. Univ. Wash. Press, Seattle. Beacham, T., R. Withler, and a. Gould. 1985. Biochemical genetic stock identification of chum salmon (Oneorhynchus keta) in southern British Columbia. Can. J. Fish. Aquat. Sci. 42:437-448. BoYER, S. H., D. C. Fainer, and E. J. Watson-Williams. 1963. Lactate dehydrogenase variant from human blood; Evidence for molecular subunits. Science 141:642-643. Bunker, R. C. 1982. Evidence of multiple late- Wisconsin floods from glacial Lake Missoula in Badger Coulee, Washington. Q. Res. (NY) 18:17-31. BusACK, C. A., G. H. Thorgaard, M. P. Bannon, and G. A. E. Gall. 1980. An electrophoretic, karyotypic and meristic character- ization of the Eagle Lake trout, Salmo gairdneri aquilarum. Copeia 1980:418-424. Campton, D. C, and F. M. Utter. 1987. Genetic structure of anadromous cutthroat trout (Salmo clarki clarki) populations in two Puget Sound regions: Evidence for restricted gene flow. Can. J. Fish. Aquat. Sci. 44:573-582. Carl, L. M., and M. C. Healey. 1984. Differences in enzyme frequency and body morphology among three juvenile life history types of chinook salmon (Oneorhynchus tshawytscha) in the Nanaimo River, British Columbia. Can. J. Fish. Aquat. Sci. 41:1070-1077. Chakraborty, R. 1980. Gene diversity analysis in nested subdivided popula- tions. Genetics 96:721-726. Chakraborty', R., P. A. Fuerst, and M. Nei. 1980. Statistical studies on protein polymorphism in natural populations III. Distribution of allele frequencies and the number of alleles per locus. Genetics 94:1039-1063. Chakraborty', R., M. Hagg, N. Ryman, and G. Stahl. 1982. Hierarchical gene diversity analysis and its appUcation to brown trout population data. Hereditas 97:17-21. Chilcote, M. W., B. a. Crawford, and S. A. Leider. 1980. A genetic comparison of sympatric populations of sum- mer and winter steelheads. Trans. Am. Fish. Soc. 109: 203-206. Clayton, J. W., and D. N. Tretuk. 1972. Amine-citrate buffers of pH control in starch gel elec- trophoresis. J. Fish. Res. Board Can. 29:1169-1172. Eanes. W. F. 1987. Allozymes and fitness: Evolution of aproblem. Trends Evol. Ecol. 2(2):44-48. Gharrett, a. J., S. M. Shirley, and G. R. Tromble. 1987. Genetic relationships among populations of Alaskan Chinook salmon (Oneorhynchus tshawytscha). Can. J. Fish. Aquat. Sci. 44:765-774. Harris, H., and D. A. Hopkinson. 1976. Handbook of enzyme electrophoresis in human gene- tics. Am. Elsevier, N.Y. Helle, J. H. 1981. Significance of the stock concept in artificial propa- gation of salmonids in Alaska. Can. J. Fish. Aquat. Sci. 38:1665-1671. Howell, P., K. Jones, D. Scarnecchia, L. LaVay, W. Kendra, AND D. Ortmann. 1985. Stock assessment of Columbia River anadromous salmonids. I. chinook, coho, chum and sockeye salmon stock summaries. Final report to Bonneville Power Administra- tion on contract No. DE-A1179-84BP12737, 558 p. Avail- able from Bonneville Power Administration, P.O. Box 3621, Portland, OR 97208. Huzy'K, L., and H. Tsuyuki. 1974. Distribution of Ldh-B gene in resident and anadromous rainbow trout (Salmo gairdneri) from streams in British Col- umbia. J. Fish. Res. Board Can. 31:106-108. 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. Kristunsson, a. C, and J. D. McIntyre. 1976. Genetic variation in chinook salmon (Oneorhynchus tshawytscha) from the Columbia River and three Oregon coastal rivers. Trans. Am. Fish. Soc. 105:620-623. McClane, a. J. 1978. Field guide to freshwater fishes of North America. H. Holt and Co., N.Y., 232 p. McPhail, J. D., AND C. C. Lindsey. 1986. Zoogeography of the freshwater fishes of Cascadia (the Columbia system and rivers north to the Stikine). In C. H. Hocutt and E. 0. Wiley (editors). The zoogeography of North American freshwater fishes, p. 615-637. J. Wiley and Sons, N.Y. Miller. M., P. Pattillo, G. B. Milner, and D. J. Teel. 1983. Analysis of chinook stock composition in the May 1982 troll fishery off the Washington coast: An application of genetic stock identification method. Wash. Dep. Fish., Tech. Rep. 74, 27 p. MORK, J., R. GiSKEODEGARD, AND G. SUNDNES. 1984. The haemoglobin polymorphism in Atlantic cod (Gadus morhua L.): genotypic differences in somatic growth and in maturing age in natural populations. Flodevigen rapp. 1: 721-732. Nei, M. 1973. Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Sci. USA 70:3321-3323. 1975. Molecular population genetics and evolution. Am. Elsevier, N.Y., 288 p. Nicholas, J. W., and D. G. Hankin. 1988. Chinook salmon populations in Oregon coastal river basins: description of life histories and assessment of recent trends in run strengths. Oreg. Dep. Fish. Wildl., Inf. Rep. (Fish.) 88-1, 359 p. Portland, OR. Okazaki, T. 1981. Geographical distribution of allelic variations of en- zymes in chum salmon, Oneorhynchus keta. populations of North America. Bull. Jpn. Soc. Sci. Fish. 47:507-514. Parkinson, E. A. 1984. Genetic variation in populations of steelhead trout (Salmon gairdneri) in British Columbia. Can. J. Fish. Aquat. Sci. 41:1412-1420. Powers, D. A., L. DiMichele, and A. R. Place. 1983. The use of enzj'me kinetics to predict differences in cellular metabolism, developmental rate, and swimming per- formance between Ldh-B genetypes of the fish, Fundulus heteroclitus. Genet. Evol. 10:147-170. Reisenbichler, R. R. 1988. Relation between distance transferred from natal 253 FISHERY BULLETIN: VOL. 87, NO. 2, 1989 stream and recovery rate for hatchery echo salmon. North Am. J. Fish. Manage. 8:172-174. Reisenbichler, R. R., and J. D. McIntyre. 1977. Genetic differences in growth and survival of juvenile hatchery and wild steelhead trout, Salmo gairdneri. J. Fish. Res. Board Can. 34:123-128. Richer, W. E. 1972. Hereditary and environmental factors affecting certain salmonid populations. h> R. C. Simon, and P. A. Larkin (editors), The stock concept in Pacific salmon, p. 19-160. H. R. MacMillan Lectures in Fisheries, University of British Columbia, Vancouver, B.C. RiDGWAY, G. J., S. W. Sherburne, and R. D. Lewis. 1970. Polymorphism in the esterases of Atlantic herring. Trans. Am. Fish. Soc. 99:147-151. ScHRECK, C. B., H. W. Li, R. C. Hjort, and C. M. Sharpe. 1986. Stock identification of Columbia River chinook salmon and steelhead trout. Final Report for Agreement No. DE- A179-83BP13499, 184 p. BPA from Oregon Coop. Fish. Res. Unit. Available from Bonneville Power Administra- tion, P.O. Box 3621, Portland, OR 97208. Shields, W. M. 1982. Philopatry, inbreeding and the evolution of sex. State Univ. New York Press, Albany, 245 p. SiciLiANO, M. J., and C. R. Shaw. 1976. Separation and visualization of enzymes on gels. In I. Smith (editor). Chromatographic and electrophoretic tech- niques. 4th ed., p. 185-209. W. Heinemann, Lond. Simon, R. C. 1972. Gene frequency and the stock problem. In R. C. Simon and P. A. Larkin (editors). The stock concept in Pacific salmon, p. 161-172. H. R. MacMillan Lectures in Fisheries, University of British Columbia, Vancouver, B.C. Sneath, p. H., and R. R. Sokal. 1973. Numerical taxonomy. Freeman, San Franc, 573 p. SoKAL, R. R.. and F. J. Rohlf. 1969. Biometry: the principles and practice of statistics in biological research. Freeman, San Franc. 859 p. Soule, M. 1976. Allozyme variation: its determinants in space and time. In F. J. Ayala (editor). Molecular evolution, p. 60-77. Sinauer Associates, Inc., Sunderland, MA. Thorgaard, G. H., F. W. Allendorf, and K. L. Knudsen. 1983. Gene-centromere mapping in rainbow trout: High inter- ference over long map distances. Genetics 103:771-783. Utter, F. M. 1981. Biological criteria for definition of species and distinct intraspecific populations of anadromous salmonids under the U.S. Endangered Species Act of 1973. Can. J. Fish. Aquat. Sci. 38:1626-1635. Utter, F. M., P. Aebersold, J. Helle, and G. Winans. 1984. Genetic characterization of populations in the south- eastern range of sockeye salmon. In J. Walton and D. Houston (editors), Proceedings of the Olympic Wild Fish Conference, p. 17-32. Fisheries Technology Program, Peninsula College and Olympic National Park. Utter, F. M., F. W. Allendorf, and B. P. May. 1976. The use of protein variation in the management of salmonid populations. Trans. 41st North Am. Wildl. Nat. Resour. Conf., p. 373-384. Utter, F. M., D. C. Campton, W. S. Grant, G. B. Milner, J. E. Seeb, and L. N. Wishard. 1980. Population structures of indigenous salmonid species of the Pacific Northwest, /n W. J. McNeil and D. C. Hims- worth (editors), Salmonid ecosystems of the North Pacific, p. 285-304. Oregon State Univ. Press, Corvallis. Utter, F. M., H. 0. Hodgins, F. W. Allendorf, A. G. Johnson, AND J. Mighell. 1973. Biochemical variants in Pacific salmon and rainbow trout: Their inheritance and application in population studies. In J. H. Schroder (editor). Genetics and muta- genesis of fish, p. 329-339. Springer-Verlag, Berl. Utter, F. M., D. J. Teel, G. B. Milner, and D. McIsaac. 1987. Genetic estimates of stock compositions of 1983 chinook salmon harvests off the Washington coast and the Colum- bia River. Fish. Bull., U.S. 85:13-23. Winans, G. A. In press. Genetic variability in chinook salmon stocks from the Columbia River basin. North Am. J. Fish Manage. Wishard, L., J. Seeb, F. Utter, and D. Stefan. 1984. A genetic investigation of suspected redband trout populations. Copeia 1984:120-132. WiTHLER, F. C. 1982. Transplanting Pacific salmon. Can. Tech. Rep. Fish. Aquat. Sci. 1079, 27 p. WiTHLER, R. E. 1985. Ldh-4 allozyme variability in North American sockeye salmon {Otworhynchus nerka) populations. Can. J. Zool. 63:2924-2932. Wright, J. E., J. R. Heckman, and L. M. Atherton. 1975. Genetic and developmental analyses of Ldh isozymes in trout. In C. L. Market (editor). Isozymes III: Develop- mental biology, p. 375-401. Acad. Press, N.Y. APPENDIX A Allele frequencies and average heterozygosities for 65 individual and pooled populations of naturally reproducing and hatchery stocks of chinook salmon. Hatchery stocks are identified by ($). The map code refers to Figure 1. 254 UTTER ET AL, .: GENETIC POPULATION STRl JCTURE F CHINOC IK SALMO N Append lix A.— Continued. Population Locus and alleles Map code Ah Ada-1 Aat-1,2 100 86 116 108 69 100 83 100 85 A1 Babine 0.986 0.014 0986 0.014 1.000 B1 Tele Jaune 0.882 0.118 0.986 0.014 1.000 B2 Clearwater 0.786 0.214 0.900 0.100 1.000 B3 Chiico 0.922 0.078 0.969 0.031 1.000 B4 Stuart-Nechako 0.958 0042 0.894 0.106 1 000 C1 $Blg Qualicum 0.838 0.162 0953 0.047 1.000 C2 SPuntledge 0.610 0.390 0.975 0.025 0990 0.010 C3 SQuinsam 0.366 0.629 0.005 0,995 0.005 0.997 0.003 C4 $San Juan 0.820 0.160 0.020 1.000 0.987 0.013 C5 SCapilano 0.763 0.237 0.909 0.091 0.967 0.033 C6 Nooksack SF 0.780 0.220 0.870 0.130 0.995 0.005 C7 Nooksack NF 0.810 0.190 0.927 0.073 0.922 0078 D1 SRobertson Ck. 0.806 0.194 1 000 0.981 0.019 E1 swells Dam 0.800 0.200 1 000 1.000 E2 $Carson-$Leavenworth 0.987 0.010 0.003 0.969 0.031 1.000 E3 SWinttirop 0.920 0.070 0.010 0.973 0.027 1.000 E4 $Priest Rapids 0.825 0.175 0,985 0.015 1.000 F1 SSoleduck (sum) 0.959 0.036 0.005 1,000 0929 0.071 F2 SSoleduck (spr) 0.848 0.152 0.995 0.005 0.998 0.003 F3 SNaselle 0.908 0.092 0.980 0.020 0.965 0.035 F4 SHumptulips 0920 0080 1.000 0.975 0025 F5 SQuinault 0.920 0.080 0.985 0.015 0.975 0025 F6 Queets 0.959 0.032 0.009 0.985 0,015 0.994 0.006 F7 Hoh 0.930 0.040 0.030 1.000 0.994 0.006 F8 SSoleduck (f) 0.837 0.133 0.031 1-000 1.000 F9 SEIwha 0.920 0.080 0.980 0.020 1.000 G1 SSkykomish 0.860 0.135 0.005 0.865 0.135 0.980 0.020 G2 SSkagit 0.838 0.162 0.959 0.041 0.985 0.015 G3 SHood Canal 0.918 0.077 0.005 0.903 0.097 0.888 0.112 G4 SDeschutes 0.842 0.158 0.953 0,047 0.913 0.088 G5 SGreen R.-SSammish 0.903 0.097 0.973 0,027 0.966 0.034 H1 SCowlitz-SKalama 0.845 0.149 0,006 0.975 0,025 1.000 H2 SLewis R. (spr) 0.910 0.080 0010 0980 0,020 0.995 0.005 H3 SCowlitz-SKalama 0.855 0.131 0.014 0.993 0,007 1.000 H4 SLewis R. (f) 0.800 0200 0.980 0,020 1.000 H5 SWashougal R. 0.850 0.120 0,030 1.000 0.995 0.005 H6 SEagle Ck.-SMcKenzie R. 0.782 0.190 0.029 1.000 1.000 11 SKIickitat R. 0.930 0.070 0.980 0,020 0.995 0.005 12 SSpring Ck.-$Big Ck. 0.990 0.010 1.000 1.000 13 SWarm Spr.-SRound Butte 1.000 1.000 0.996 0.004 14 Deschutes 0.867 0.102 0.031 0.990 0.010 1.000 J1 Ice Harbor 0.874 0.111 0.003 0.013 0998 0.003 1.000 J2 McCall-Johnson Ck. 1.000 0.953 0.047 0981 0019 J3 SRapid R. -Valley Ck. 0.994 0.006 969 0.031 0.997 0.003 K1 SCole R.-Hoot Owl Ck. 0.957 0.043 1.000 0,998 0.002 K2 SRock Ck. 0.890 0.105 0.005 1.000 0,990 0.010 K3 SCedar Ck. 0.760 0.087 0.036 0.010 0.107 1.000 0.987 0.013 K4 STrask R. (spr) 0.735 0.110 0.020 0.020 0.115 1.000 0.978 0023 K5 Chetco 0.890 0.110 0.990 0,010 0990 0.010 K6 Lobster Ck. 0.930 0.070 1 000 0.975 0.025 K7 SEIk R. 0.800 0.185 0.015 0.950 0050 0.950 0.050 K8 Sixes R. estu. 0.850 0.105 0.035 0.010 0.985 0,015 0.968 0032 K9 Coquille R. estu. 0.883 0.113 0.004 0.965 0,035 0.949 0.051 K10 Siuslaw Bay 0.790 0,136 0049 0.006 0.019 0.976 0,024 0.959 0.041 K11 SSalmon R. 0.737 0-076 0.152 0.035 1.000 0.990 0.010 K12 SNestucca R.-$Alsea R. 0.811 064 0.113 0.001 0.012 0.969 0.031 0.976 0024 K13 SCedar Ck. 0.610 0.215 0.120 0.055 0.995 0.005 0.944 0.056 K14 STrask R. -Tillamook Bay 0.730 0.141 0.113 0.003 0.013 0968 0.032 0.991 0.009 K15 Nehalem estu. 0.685 0.236 0.079 0.984 0.016 0990 0.010 L1 SFeather R. 0.720 0.240 0.040 1.000 1.000 L2 SColeman-SNimbus 0.815 0.173 0.007 0.005 1.000 0.992 0008 L3 SFeather R.-SMokelumne 0.797 0.195 0.005 0.002 1.000 0.999 0.001 Ml STrinity R. 1.000 1 000 1.000 M2 SIron Gate 0.995 0.005 1.000 0997 0.003 M3 STrinity (f) 1.000 1,000 1.000 255 FISHERY BULLETIN: VOL. 87, NO. 2, 1989 Appendix /K.— Continued. Population Locus and alleles Map code Aat-3 Dpep-1 Dpep-2 Gpi-1 100 90 100 90 100 105 100 60 H A1 Babine 0.957 0.043 0.972 0.028 1.000 1.000 B1 Tete Jaune 1.000 0.987 0.013 1.000 1.000 B2 Clearwater 1.000 1.000 1.000 1.000 B3 Chiico 1.000 0.979 0.021 0.939 0.061 1.000 84 Stuart-Nechako 1.000 0.963 0.037 1.000 1.000 CI $Blg Qualicum 0.829 0.171 0.935 0.065 1.000 1.000 C2 SPuntledge 0.735 0.265 0.995 0.005 1 000 1.000 C3 SQuinsam 0.831 0.169 0.974 0.026 1.000 1.000 C4 $San Juan 0.990 0.010 0.970 0.030 1.000 1.000 C5 $Capilano 0.803 0.197 0.985 0.015 1.000 1.000 C6 Nooksack SF 0.990 0.010 0.918 0.082 1.000 1.000 C7 Nooksack NF 1.000 0.980 0.020 1.000 1.000 D1 $Robertson Ck. 0.911 0.089 1.000 1.000 1.000 El $Wells Dam 1.000 0.980 0.020 1.000 0.859 0.141 E2 $Carson-$Leavenworlh 1.000 0.993 0.007 1.000 1.000 E3 SWinttirop 1.000 1.000 1.000 1.000 E4 $Priest Rapids 1.000 0995 0.005 1.000 0.824 0.176 F1 $Soleduck (sum) 0.995 0.005 0.745 0.255 1.000 1.000 F2 SSoleduck (spr) 0.995 0.005 0.970 0.030 1.000 1.000 F3 $Naselle 1.000 0.843 0.157 0.995 0.005 1.000 F4 $Humptulips 1.000 0840 0.160 1.000 1.000 F5 $Quinault 1.000 0.890 0110 1.000 0.995 0.005 F6 Queets 1.000 0.833 0.167 0.954 0.046 0.996 0004 F7 Hoh 1.000 0.905 0.095 0.995 0.005 0.980 0.020 F8 SSoleduck (f) 1.000 0.740 0.260 1.000 1.000 F9 $Elwha 0.979 0.021 0.890 0.110 1 000 1.000 G1 SSkykomish 0.967 0.033 0.980 0.020 1.000 1.000 G2 SSkagit 1.000 0.925 0.075 1.000 1.000 G3 SHood Canal 0.995 0.005 0.923 0.077 1.000 1.000 G4 SDesctiutes 1.000 0.893 0.107 1.000 1.000 G5 $Green R.-$Sammish 1.000 0.876 0.124 1.000 1.000 HI SCowlitz-SKalama 1.000 0.949 0.051 1 000 0.895 0.105 H2 $Lewis R. (spr) 1.000 1.000 1.000 1.000 H3 $Cowlitz-$Kalama 1.000 0.913 0.087 1.000 1.000 H4 SLewis R. (f) 1.000 0.830 0.170 1.000 1.000 H5 $Washougal R. 1.000 0.850 0.150 0990 0.010 1 000 H6 $Eagle Ck.-$McKenzie R. 1,000 1.000 1.000 1.000 11 SKIickitat R. 1,000 0.990 0.010 1.000 1.000 12 SSpring Ck.-$Big Ck. 1.000 0.987 0.013 1.000 1.000 13 $Warm Spr,-$Round Butte 1.000 0.972 0.028 1.000 1,000 14 Deschutes 0989 0.011 0898 0.102 1 000 1.000 J1 Ice Harbor 0.996 0.004 0.967 0.033 1.000 0.842 0.158 J2 McCall-Johnson Ck. 1.000 1.000 1.000 1.000 J3 $Rapid R.-Valley Ck. 1.000 0.994 0.006 1.000 1 000 K1 SCole R.-Hoot Owl Ck. 0.972 0.028 0908 0.092 1.000 0.890 0.110 K2 SRock Ck. 0.955 0.045 0925 0075 1 000 1.000 K3 $Cedar Ck. 0.995 0.005 0.652 0.348 1.000 1.000 K4 $Trask R. (spr) 0995 0.005 0.783 0.217 1.000 1.000 K5 Ctietco 1.000 0855 0.145 1.000 1.000 K6 Lobster Ck. 1.000 0850 0.150 1.000 1.000 K7 $Elk R. 1.000 0.732 0.268 1.000 1.000 K8 Sixes R. estu. 1.000 0.655 0345 1.000 1.000 K9 Coquille R. estu. 1.000 0.652 0348 1.000 1.000 K10 Siuslaw Bay 1.000 0.701 0.299 1.000 1.000 K11 SSalmon R. 0.995 0.005 0.783 0.217 1.000 1.000 K12 SNestucca R.-$Alsea R. 0999 0.001 0.708 0.292 1.000 1.000 K13 $Cedar Ck. 0.995 0.005 0.700 0.300 1.000 1.000 K14 STrask R. -Tillamook Bay 1 000 0.704 0296 1.000 1.000 K15 Nehalem estu. 0.980 0.020 0.770 0.230 1.000 1.000 L1 SFeattier R. 1.000 0.890 110 1.000 0.576 0.424 L2 $Coleman-$Nimbus 1.000 0.869 0.131 1.000 0.705 0.295 L3 SFeather R -SMokelumne 1.000 0.905 0.095 1.000 0689 0.311 Ml STrinity R. 1.000 0990 0010 1 000 1 000 M2 Siron Gate 1.000 0.995 0.005 1.000 1.000 M3 STrinity (f) 1.000 1.000 1.000 1 000 256 UTTER ET AL,: GENETIC POPULATION STRUCTURE OF CHINOOK SALMON Appendix A.— Continued. Population Locus and alleles Map code Gpi -2 Gpi-3 Gr 100 60 100 105 93 100 85 A1 Sabine 1.000 0.917 0.083 0.973 0027 B1 Tete Jaune 1.000 0.932 0.068 0.958 0.042 B2 Clearwater 1.000 0.989 0.011 0.856 0.144 B3 Chiico 1.000 0.875 0.125 0.939 0.061 84 Stuart-Nechako 1.000 0933 0.067 0860 0.140 CI $Big Qualicum 1.000 1.000 1.000 C2 SPuntledge 1.000 0.995 0.005 1.000 C3 SQuinsam 1.000 1.000 0995 0.005 C4 $San Juan 1.000 1.000 0.970 0.030 C5 SCapilano 1.000 1.000 1.000 C6 Nooksack SF 1.000 0.880 0.120 0.990 0.010 C7 Nooksack NF 1.000 0.950 0.050 0.908 0.092 D1 $Roberlson Ck. 0.755 0.245 1.000 1.000 El swells Dam 1.000 0.970 0.030 0.980 0.020 E2 SCarson-SLeavenworth 1.000 1.000 0.993 0.007 E3 $Winthrop 1.000 0.984 0.016 0.918 0.082 E4 SPriest Rapids 1.000 1.000 0.975 0025 F1 SSoleduck (sum) 0.613 0.387 1.000 1.000 F2 SSoleduck (spr) 1.000 1.000 1.000 F3 SNaselle 0.788 0.212 1.000 1.000 F4 SHumptulips 0.755 0.245 0.970 0.030 0990 0010 F5 SQuinault 0.700 0.300 0.980 0.020 1.000 F6 Queets 0.671 0.329 0.996 0.004 1.000 F7 Hoh 0.542 0.458 0.995 0.005 1 000 F8 SSoleduck (f) 0.553 0.447 0980 0.020 1.000 F9 SEIwha 0.684 0.316 0.975 0.025 0.995 0.005 G1 SSkykomish 1.000 0.955 0.045 0.995 0.005 G2 SSkagit 1.000 0.995 0.005 1.000 G3 SHood Canal 1.000 1.000 1.000 G4 SDeschutes 0.916 0.084 0.990 0.010 1.000 G5 SGreen R.-$Sammish 1.000 0.993 0.003 0.003 0,990 0.010 H1 $Cowlitz-$Kalama 1.000 1.000 0,822 0.179 H2 SLewis R^ (spr) 1.000 1.000 0908 0.092 H3 SCowlitz-SKalama 0.916 0.084 1.000 0.795 0.205 H4 SLewis R. (f) 1.000 1.000 0.820 180 H5 SWashougal R. 1.000 1.000 0800 0.200 H6 SEagle Ck.-SMcKenzie R. 1.000 1.000 0.420 0580 11 SKIickitat R. 1.000 1.000 0.760 0.240 12 SSpnng Ck.-SBig Ck, 1.000 1.000 0.663 0.337 13 SWarm Spr -SRound Butte 1.000 1.000 1.000 14 Deschutes 1.000 0.990 0.010 0949 0.051 J1 Ice Harbor 0.929 0.071 1.000 1.000 J2 McCall-Johnson Ck. 1,000 1.000 1.000 J3 SRapid R. -Valley Ck. 1.000 1.000 1.000 K1 SCole R.-Hoot Owl Ck. 0842 0.158 1.000 0.997 0.003 K2 SRock Ck. 0755 0.245 1.000 0.925 0.075 K3 SCedar Ck. 0.698 0.302 1.000 1.000 K4 STrask R. (spr) 0.553 0.447 1.000 0.961 0.039 K5 Chetco 0.827 0.173 1.000 1.000 K6 Lobster Ck. 1.000 1.000 0.960 0.040 K7 $Elk R. 0.520 0.480 1.000 1.000 K8 Sixes R. estu. 0.434 0.566 1.000 1.000 K9 Coquille R. estu. 0.441 0.559 0.991 0.009 1.000 K10 Siuslaw Bay 0.545 0.455 1.000 1.000 K11 SSalmon R. 0.682 0.318 1.000 1.000 K12 SNestucca R.-SAIsea R. 0.525 0.475 0.997 0.003 1.000 K13 SCedar Ck. 0.432 0.568 1.000 1.000 K14 STrask R. -Tillamook Bay 0.435 0.565 1.000 1.000 K15 Nehalem estu. 783 0.217 1.000 1 000 LI SFeather R. 1 000 1.000 1.000 L2 SColeman-SNimbus 0.945 0.055 1.000 1.000 L3 SFeather R.-$Mokelumne 0.900 0.100 1.000 1.000 Ml STrinity R. 0.859 0.141 1.000 1.000 M2 SIron Gate 1.000 1.000 0.995 0.005 M3 STrinity (f) 1.000 1.000 1.000 257 FISHERY BULLETIN: VOL. 87, NO. 2, 1989 Appendix A. — Continued. Locus and alleles Map ldh-3,4 Ldh-4 code Population 100 127 74 142 100 112 134 71 A1 Babine 0.947 0.007 0046 1.000 B1 Tete Jaune 1.000 1.000 B2 Clearwater 0.967 0.006 0.011 0.017 0.933 0.067 B3 Ctiiico 0.985 0.005 0.010 1.000 84 Stuart-Nechako 0.976 0.009 0.002 0.012 1.000 CI $Big Qualicum 1.000 1.000 C2 SPuntledge 0.995 0.005 1.000 C3 SQuinsam 1.000 1.000 C4 $San Juan 1.000 1.000 C5 SCapilano 0.970 0.013 0018 1.000 C6 Nooksack SF 0.984 0.016 1.000 C7 Nooksack NF 1.000 1.000 D1 $Robertson Ck. 0.980 0.020 1.000 El swells Dam 0.875 0.125 1.000 E2 SCarson-SLeavenworth 0.862 0.002 0.136 0.973 0.027 E3 SWinttirop 0965 0010 0.025 0.996 0004 E4 SPriest Rapids 0908 0.090 0.003 1.000 F1 SSoleduck (sum) 0.874 0.111 0.003 0.013 1.000 F2 $Soleduck (spr) 0.958 0.037 0.005 1.000 F3 $Naselle 0987 0.010 0.003 1 000 F4 $Humptulips 0.985 0.010 0.005 1 000 F5 SQuinault 0.903 0.090 0.003 0.005 1.000 F6 Queets 0.892 0.108 1.000 F7 Hoh 0.908 0.093 1 000 F8 SSoleduck (f) 0.990 0.010 1.000 F9 SEIwha 0.898 0.095 0.003 0.005 1 000 G1 SSkykomish 0.958 0.008 0.035 1.000 G2 SSkagit 0960 0.010 0.010 0020 1.000 G3 SHood Canal 0.957 0003 0.005 0.036 1.000 G4 SDeschutes 0.942 0.055 0.003 1.000 G5 SGreen R.-$Sammish 0.968 0.009 0.002 0022 1.000 HI SCowlitz-SKalama 0.915 0.055 0030 1 000 H2 SLewis R. (spr) 0.925 0.005 0.070 0.980 0.020 H3 SCowlitz-SKalama 0.971 0.012 0.017 1.000 H4 SLewis R. (f) 0.933 0.022 0.044 1.000 H5 SWastiougal R. 0.955 0.015 0.030 1.000 H6 SEagle Ck.-$McKenzie R. 0.868 0.126 0.006 1 000 11 SKIickitat R. 0.900 0.070 0.030 1.000 12 SSpring Ck.-SBig Ck. 0.990 0.008 0.002 1.000 13 SWarm Spr.-$Round Butte 0.865 135 1.000 14 Desctiutes 0.969 0.031 1.000 J1 Ice Harbor 0.977 0.023 1.000 J2 McCall-Johnson Ck. 0.913 0.087 1 000 J3 SRapid R.-Valley Ck. 0.937 0.006 0.057 0972 0.028 K1 SCole R.-Hoot Owl Ck. 0.962 0.038 0994 0.003 0.003 K2 SRock Ck. 0.977 0.023 1 000 K3 SCedar Ck. 0.995 0.003 0.003 1.000 K4 STrask R. (spr) 0995 0.005 1 000 K5 Chetco 0.985 0.015 1.000 K6 Lobster Ck. 0.978 0.022 0.940 0.060 K7 SEIk R. 0.973 0.027 0.990 0.010 K8 Sixes R. estu. 0.972 0.028 0970 0005 0.010 0.015 K9 Coquille R. estu. 1.000 0.961 0.003 0.009 K10 Siuslaw Bay 0.994 0.003 0003 1.000 K11 SSalmon R. 0.975 0.025 1.000 K12 SNestucca R.-$Alsea R. 0981 0.016 0.001 0.001 0.999 0.001 K13 SCedar Ck. 0.947 0.053 1.000 K14 STrask R. -Tillamook Bay 0963 0037 1 000 K15 Nehalem estu. 0.947 0053 1 000 LI SFeattier R. 0.940 0060 1 000 L2 SColeman-SNimbus 0.950 0.050 1.000 L3 SFeather R.-SMokelumne 0.945 0.055 1.000 Ml STrinity R. 1.000 1 000 M2 SIron Gate 1.000 1.000 M3 STrinity (f) 1 000 1.000 258 UTTER ET AL.: GENETIC POPULATION STRUCTURE OF CHINOOK SALMON Appendix a.— Continued. Population Locus and alleles Map code Ldh-5 Mdh-1,2 Mdh-3,4 100 90 70 100 120 27 100 121 70 83 A1 Babine 1.000 1.000 1.000 B1 Tete Jaune 1 000 1.000 1 000 B2 Clearwater 1.000 1.000 0.989 0.006 0006 B3 Chlico 1.000 1.000 0.990 0.010 84 Stuart-Nechako 1.000 1.000 0.936 0.064 CI $Big Qualicum 1.000 1.000 1.000 C2 SPuntledge 1.000 1.000 0.948 0.037 0015 C3 SQulnsam 0.966 0.034 1.000 0.843 0.103 0.054 C4 $San Juan 1.000 1.000 0.990 0.010 C5 $Capllano 0.995 0.005 1.000 1.000 C6 Nooksack SF 1.000 1.000 0.965 0.030 0.005 C7 Nooksack NF 1.000 1.000 0.965 0.035 D1 SRobertson Ck. 1.000 1.000 0.993 0.008 El swells Dam 0.980 0.020 1.000 0,970 0.010 0.020 E2 $Carson-$Leavenworth 1.000 1.000 0978 0.022 E3 SWinthrop 1.000 1.000 0,990 0.010 E4 SPriest Rapids 0.964 0.015 0020 1.000 0,955 0.030 0.015 F1 SSoleduck (sum) 1.000 1.000 0,963 0.003 0035 F2 SSoleduck (spr) 1.000 1.000 0,975 0.015 0.010 F3 SNaselle 1.000 1.000 0.944 0.040 0.015 F4 SHumptulips 1.000 1.000 0,985 0.015 F5 SQuinault 1.000 1.000 988 0.013 F6 Queets 1.000 1.000 0,963 0.037 F7 Hoh 1.000 1.000 0993 0.003 0.003 0.003 F8 $Soleduck (f) 1.000 1.000 0,985 0,015 F9 SEIwha 1.000 1.000 0,968 0,015 0.017 G1 SSkykomish 0,980 0020 1.000 0,990 0.010 G2 $Skagit 0990 0.010 1.000 0,993 0.008 G3 $Hood Canal 1.000 1.000 0.967 0.033 G4 $Deschutes 1.000 1.000 0.992 0.008 G5 SGreen R.-$Sammish 0.990 0.010 1,000 0.991 0,003 0.005 HI $Cowlitz-$Kalama 1.000 1.000 0988 0,013 H2 $Lewis R. (spr) 0.990 0.010 1.000 0.965 0,035 H3 $Cowlitz-$Kalama 0.997 0.003 1.000 0.983 0,017 H4 SLewis R. (f) 1.000 1.000 1.000 H5 SWashougal R. 1.000 1,000 0.990 0,010 H6 $Eagle Ck.-$McKenzie R. 1.000 1,000 0.963 0.037 11 SKIickitat R. 1.000 1.000 0.970 0.030 12 SSpring Ck.-$Big Ck. 1.000 1.000 0.945 0.055 13 $Warm Spr.-$Round Butte 1.000 0.995 0.005 1.000 14 Desctiules 1.000 1.000 0985 0.010 0,005 J1 Ice Harbor 0.995 0.003 0.003 1.000 0.985 0.005 0,010 J2 McCall-Jotinson Ck. 0.976 0.024 1.000 0.998 0.002 J3 SRapid R. -Valley Ck. 1.000 1.000 0.995 0.005 K1 $Cole R.-Hoot Owl Ck. 0988 0012 0.989 0.011 0,992 0.008 K2 SRock Ck. 1.000 1.000 0,968 0.030 0,003 K3 SCedar Ck. 0.975 0025 1.000 0.995 0.005 K4 $Trask R. (spr) 0.985 0.015 1.000 0.990 0.010 K5 Chetco 1.000 0.998 0.003 0.955 0.045 K6 Lobster Ck. 1.000 1.000 0.975 0.025 K7 $Elk R. 1.000 0.998 0.003 0.983 0.017 K8 Sixes R. estu. 1.000 0.945 0.027 0.027 0.993 0.008 K9 Coquille R. estu. 0.996 0.004 0987 0.011 0.002 0996 0.004 K10 Siuslaw Bay 0.982 0,018 0.997 0.003 0.982 0,018 K11 SSalmon R. 1.000 1.000 1.000 K12 SNestucca R.-$Alsea R. 0.999 0,001 0.999 0.001 1.000 K13 SCedar Ck 1.000 1.000 1.000 K14 STrask R. -Tillamook Bay 1.000 0.999 0.001 0.985 0.015 K15 Nehalem estu. 1,000 1.000 1.000 L1 SFeather R. 1.000 1.000 0.945 0.055 L2 SColeman-SNimbus 1.000 1.000 0.968 0.032 L3 SFeather R.-SMokelumne 1.000 1.000 0.977 0.023 Ml Slrinity R. 1.000 1.000 1.000 M2 SIron Gate 1,000 1.000 0.997 0.003 M3 STrinity (f) 1,000 1,000 1.000 259 FISHERY BULLETIN: VOL. 87, NO. 2, 1989 Appendix A.— Continued. Population Locus and alleles Map code Mpi Pgm-1 ,2 Pgk-2 100 109 95 113 -100 -70 -84 100 90 A1 Sabine 0.730 0.270 1.000 0.095 0.905 B1 Tete Jaune 0.689 0.311 1.000 0.421 0.579 B2 Clearwater 0.535 0.465 1.000 0.178 0.822 B3 Chiico 0.633 0.367 1.000 0.194 0.806 84 Stuart-Nechal> n a. JZ o 3 1- b o m =! 0) D o _ o 0) ro c 2 JD a 1- 03 ■D u —> N -□ h o 5 <" I 276 BAILEY: DESCRIPTION AND SURFACE DISTRIBUTION OF JACK MACKEREL Table 3— Comparison of meristic and morphometric characters of Trachurus murphyi (numbers sfiown are range and mean). Distinguishing Present study Hildebrand (1946) Berry and Cohen (1974) Shaboneyi Kotlyar ( ev and 1979) character' juvenil les juveniles adults juveniles & adults adults Size range (SL mm) 46-83 60-117 485-497 94-552 233-475 (FL) Number examined 40 — 7 17 250 Scales and scutes number Curv 48-58 (52.8) — — 51-56 (53.0) 46-55 (50.9) number Str 43-53 (48.4) — — 41-50 (44.8) 44-56 (49.4) number total 96-1 1 1 (101.2) — 93-104 94-106 (97.8) 92-107 (100.3) Ht Curv (%SL) 3.2- 4.7 (4.1) — — 4.6-5.6 3.1- 5.1 (4.1) Ht Str (%SL) 3.9- 5.4 (4.8) — — 4.8-5.7 4.1- 5.9 (5.3) Ht Curv (%HL) 10.0-17.4 (14.3) 15.9-17.5 17.0-20.0 — 11.0-18.6 (14.4) Ht Str (%HL) 12.2-19.6 (16.8) 15.9-21.3 16.1-18.9 — 13.8-21.7 (17.3) Endpoint LL Ace 1-5 (1.9) — — 1-5 ^2-7 (4.2) Gill rakers upper limb 14-18 (15.5) — 15-17 15-18 (16.5) 14-23 (17.2) lower limb 40-45 (41.9) _ 45-48 42-45 (43.9) 39-49 (44.7) total 54-61 (57.4) — — 58-63 (60.4) 57-68 (61.9) Softrays dorsal 30-35 (33.2) — 30-33 30-36 (33.6) ^32-38 (34.0) anal 26-31 (28.5) — 25-27 27-31 (28.9) ^26-33 (29.5) 'See text page 274 regarding abbreviations. 2|ncluded in these counts is the first spine of the second dorsal fin and anal fin. pletely formed in the juveniles, for example, Santander and de Castillo (1971) noted that fin for- mation was complete in T. murphyi of 13 mm SL. In other Trachurus species lateral line scales and scutes are well developed by 35 mm (Ahlstrom and Ball 1954; Haigh 1972; Stephenson and Robertson 1977). There are significant differences in the numbers of scales and scutes along the curved and straight sections of the lateral line between the present spe- cimens and those described by Berry and Cohen (1974) (Student's t test, P < 0.05, df = 51), Sha- boneyev and Kotlyar (1979) {Z test, P < 0.05), and Kotlyar (1976) (Z test, P < 0.05). These differences may be due to how the dividing line between the two sections is defined. When the curved and straight sections are combined into single counts along the entire lateral line, the present specimens only dif- fer significantly from those of Berry and Cohen (1974). Evseenko (1987) suggested that Trachurus mur- phyi has a spawning area centered on the Sub- tropical Convergence Zone extending from Chile to between 150°W and 160°W. He based his sugges- tion on an average transport figure for the area, growth data of T. symmetricus and the occurrence of one juvenile and two larvae in the central South Pacific. Results from the present study verifies his suggestion in the central South Pacific and, by using a similar approach, extends the probable spawning area westward to include the Chatham Rise. It is apparent that Trachurus murphyi is found and likely to spawn across the South Pacific from New Zealand to Chile. The abundance of juveniles in the Subtropical Convergence Zone between 127°W and 165°W further suggests that a large commercial resource may exist in the central and western parts of the South Pacific. ACKNOWLEDGMENTS I thank W. Smith-Vaniz for confirming the iden- tification of Trachurus murphyi. The assistance of the captains and crew of RV's Townsend Cromwell, Coriolis, and Kaharoa as well as scientific person- nel from the National Marine Fisheries Service laboratories in Honolulu, HI and La JoUa, CA and the Noumea, New Caledonia Centre of the Office de Recherche Scientifique et Technique Otre-Mer is gratefully acknowledged. Thanks are also due to J. A. Wetheral, T. E. Murray, J. B. Jones, R. M. Laurs, P. J. McMillan, B. B. Collette, and an anony- mous reviewer for criticism of the manuscript. LITERATURE CITED Ahlstrom, E. H., and 0. P. Ball. 1954. Description of eggs and larvae of jack mackerel [Tra- churus symmetricus) and distribution and abundance of larvae in 1950 and 1951. U.S. Fish Wildl. Serv., Fish. Bull. 56:209-245. Berry. F. H. 1968. A new species of carangid fish (Decapturas tabl ) from 277 FISHERY BULLETIN: VOL. 87, NO. 2, 1989 the western Atlantic. Contrib. Mar. Sci. 13:145-167. Berry, F. H., and L. Cohen. 1974. Synopsis of the species of Trachurus (Pisces: Caran- gidae). Q. J. Fla. Acad. Sci. 35:177-211. EVSEENKO, S. A. 1987. Reproduction of Peruvian jaclt mackerel, Trachurus symnwtricus murphyi, in the southern Pacific. J. Ichthyol. 27(3):151-160. Gutierrez, A. T. 1986. Migraciones de Trachurus murphyi en el norte de Chile. Invest. Pesq. (Chile) 33:99-103. Haigh, E. H. 1972. Development of Trachurus trachurus (Carangidae), the South African Maasbander. Ann. S. Afr. Mus. 59(8): 139- 150. Heath, R. A. 1985. A review of the physical oceanography of the seas around New Zealand - 1982. N.Z. J. Mar. Freshwater Res. 19(1):79-124. Hildebrand, S. F. 1946. A descriptive catalogue of the shore fishes of Peru. U.S. Natl. Mus. Bull. 189, 530 p. HUBBS, C. L., AND K. F. Lagler. 1957. Fishes of the Great Lakes Region. Univ. Mich. Press, Ann Arbor, 213 p. Kawahara, S., Y. Uozumi, and H Yamada. 1988. First record of a carangid fish, TrcLchurus murphyi from New Zealand. Jpn. J. Ichthyol. 35:212-214. Kotlyar, a. N. 1976. A morphological description of the Peruvian jack mackerel, Trachurus symmetrieus murphyi. J. Ichthyol. 16(l):45-55. Roberts, P. E. 1980. Surface distribution of albacore tuna, Thunnus ala- lunga Bonnaterre, in relation to the Subtropical Conver- gence Zone east of New Zealand. N.Z.J. Mar. Freshwater Res. 14:373-380. Santander, H., and 0. S. de Castillo. 1971. Desarrollo y distribucion de huevos y larvas de "jurel" Trachurus symmetrieus murphyi (Nichols) en la costa Peru- ana. Inst. Mar Peru (Callao) Inf. 36, 23 p. Shaboneyev, I. Y. 1980. Systematics, morpho-ecological characteristics and origin of carangids of the genus Trachurus. 3. Ichthyol. 20(6):15-24. Shaboneyev, I. Y., and A. N. Kotlyar. 1979. A comparative morphoecological analysis of the east- ern Pacific forms of Trachurus symmetrieus and the Atlan- tic oceanic horse mackerel, Traehurus picturatus picturatus. J. Ichthyol. 19(2):24-29. Stephenson, A. B., and D. A. Robertson. 1977. The New Zealand species of Trachurus (Pisces: Caran- gidae). J. R. Soc. N.Z. 7(2):243-253. 278 ENERGY UTILIZATION IN BAY ANCHOVY, ANCHOA MITCHILLI, AND BLACK SEA BASS, CENTROPRISTIS STRIATA STRIATA, EGGS AND LARVAE' John W. Tucker, Jr.^ ABSTRACT Bay anchovy, Anchoa mitchilli. and black sea bass, Centropristis striata striata, both produce abun- dant, small, planktonic eggs and larvae, but these appear to have contrasting nutritional strategies. Developmental changes and energy utilization in eggs, unfed larvae, and fed larvae of the two species suggest that black sea bass are better able to resist fluctuations in food availability (survive and grow at lower prey densities). Black sea bass have more time to find food and develop feeding skills— 47 hours between first feeding and yolk exhaustion vs. 8 hours for bay anchovies. Sea bass feed more efficiently than anchovies. Over the first 96 hours after first feeding, capture success averaged 85% for sea bass and 60% for anchovies. Gross growth efficiency of sea bass (13%) was more than twice that of anchovies (5%). Sea bass may also be more resistant to starvation because their yolk lasts longer (180 hours vs. 80 hours after hatching) and because, during starvation, their metabolism is lower and they lose body calories at a lower rate. An important determinant of survival of larval fishes is their ability to fulfill nutritional require- ments after yolk energy is exhausted. The manner in which energy is used by fish eggs and larvae may indicate adaptability of early stages relative to food composition or abundance. Differences in energy utilization among species might result from different feeding strategies or from adaptation to different feeding conditions (Hunter 1980). The bay anchovy, Anchoa mitchilli, a clupeiform planktivore, is a major food item for predaceous fishes along the U.S. Gulf and Atlantic coasts. Adults are pelagic and live in shallow coastal waters from the Gulf of Maine to Yucatan, Mexico (Hilde- brand 1963). In North Carolina, spawning by large schools occurs just after sunset in estuaries and coastal waters from late April to early September and peaks during late June to early August (Kuntz 1914; Hildebrand and Cable 1930; pers. obs.). Eggs (which lack oil globules) and larvae are planktonic and occur in estuaries and bays and just offshore. Spawning might occur over a wide temperature range (Dovel 1971), but larval growth is best in the mid to high twenties (Houde 1974). Early juveniles are abundant in brackish water and also enter fresh water. 'Contribution 673, Harbor Branch Oceanographic Institution, Fort Pierce, FL. ^Harbor Branch Oceanographic Institution, 5600 Old Dixie Highway, Fort Pierce, FL 34946. Manuscript accepted November 1988. Fishery Bulletin, U.S. 78:279-293. The black sea bass, Centropristis striata striata, a perciform piscivore generally found offshore, sup- ports important commercial and sport fisheries along the U.S. Atlantic coast. It is distributed over the continental shelf and in bays from Cape Cod, MA to Cape Canaveral, FL and occasionally to the Gulf of Maine or Florida Keys (Miller 1959; Musick and Mercer 1977). Adults are demersal, and south of Cape Hatteras they are found on rough bottom over the inner shelf. Spawning takes place over the inner shelf, mostly in spring or summer, depending on latitude (Musick and Mercer 1977). Off North Carolina, peak spawning is from March to early June. Eggs (with a single oil globule) and larvae are planktonic and occur in shelf waters of 15-51 m depth (Kendall 1972). Juveniles are often found in high salinity estuaries and bays but move into deep- er water as they grow. Several aspects of the feeding ecology of bay an- chovy larvae have been investigated, but little is known about black sea bass larvae. Houde and Schekter (1981, 1983) compared growth and ener- getics of bay anchovy; sea bream, Archosargus rhomboidalis; and lined sole, Achirtis lineatus, lar- vae. No studies of black sea bass larval ecology have been published, but the southern sea bass, C. striata melana, has been reared under experimental mari- culture conditions in Florida (Hoff 1970; Roberts et al. 1976; Harpster et al. 1977). This paper presents information on developmen- tal events and energy utilization for bay anchovy and 279 FISHERY BULLETIN: VOL. 87, NO. 2. 1989 black sea bass from just after fertilization through the eighth day of feeding. Results are used to infer differences in early survival and growth capabilities in nature. Particularly important are differences during the first 96 hours of feeding, which probably arise from adaptations necessary for exploiting dif- ferent food supplies. Prey densities tend to be lower in the larval black sea bass's habitat (Theilacker and Dorsey 1980). MATERIALS AND METHODS The study was conducted with eggs, unfed larvae, and larvae fed for 8 days. The timing of the follow- ing developmental events was noted: hatching (H), completion of eye pigmentation (EP), first feeding (FF), yolk exhaustion (E YS), and death of unfed lar- vae (S). Measurements were made of notochord length (NL), dry weight, % ash, % total carbon, % total nitrogen, % total lipid, energy content, oxy- gen consumption, feeding efficiency, and feeding rate. Egg Sources Bay anchovy eggs usually (40 collections) were ob- tained 3-5 hours after spawning (before morula stage) by stationary plankton tows in Pivers Island Channel, near Beaufort, NC. For one series of oxy- gen uptake measurements, eggs and milt were ob- tained by stripping ripe adult anchovies. Black sea bass eggs were stripped from six females (313-672 g), in which ovulation had been induced by injection of human chorionic gonadotropin, and they were fer- tilized artificially (Tucker 1984). Culture Conditions Physical conditions for rearing experiments ap- proximated those in natural habitats in North Caro- hna waters during peak spawning. Temperatures were slightly lower than optimal for growth. Bay anchovies were maintained at 24°C and 32°lm with a 14L:10D photoperiod. Black sea bass were main- tained at 20°C and 34%o with a 12L:12D photo- period. Fluorescent lighting provided 1400 lux at the water surface. Incubation and rearing took place in one to eight (usually six) 10 L black cylindrical fiber- glass tanks of filtered seawater. Initial stocking den- sity was 30 or fewer eggs per liter. First-feeding larval density was reduced to fewer than 15/L. Roti- fers, Brachionus plicatilis, of the same strain inves- tigated by Theilacker and McMaster (1971) were added when larval eye pigmentation was complete. and densities were maintained at about 20/mL. Phy- toplankton, Chlorella sp. or Nannochloris sp., was also added as food for the rotifers. Nutritional qual- ity of starving rotifers diminishes rapidly. Unless well-fed rotifers are added frequently and all of them are eaten quickly, algae must be present in the rear- ing tanks to maintain their quality. Good rotifer nutrition also ensures that they continue to repro- duce, thus maintaining the full size range. Without algae, rotifers not eaten within several hours be- come empty shells. Without reproduction, a rotifer population tends to consist entirely of large adults. Measurements and Calculations The times of eye pigmentation and yolk exhaus- tion were determined by microscopic examination. Starvation mortality was determined in the 10 L rearing tanks (5 times for each species, 1-3 tanks). In addition, three starvation mortality determina- tions were made in 2 L dishes using bay anchovy eggs collected on different nights. For each deter- mination, 25 normally developing eggs were placed in each of eight 2 L black glass dishes; dead eggs and larvae were counted and removed periodically, with 100% recovery. Egg and larval dry weights were determined directly. Daily samples (usually 30 individuals) were taken randomly from the rearing tanks for deter- mination of dry weight. Each group of specimens was rinsed in distilled water and freeze-dried before weighing. Best-fit regression equations were used to predict dry weight at different ages during devel- opment. Notochord length of 10-52 specimens was measured at key times during development. Instan- taneous, or specific, growth rate was calculated as g = (In W„ - In Wo)IT, in which W„ is the final weight on day n, Wq is the initial weight on day 0, and T is the interval in days. I determined energy content of eggs and larvae directly by calorimetry and indirectly by proximate analysis. Eggs and larvae were sampled periodical- ly for ashing, elemental analysis, total lipid assays (black sea bass only), and bomb calorimetry. Ash weights were determined by combustion for 12 hours at 500 °C (0.6-2.0 mg subsamples— anchovies: 9 samples, 1 or 2 replicates, 12 determinations; black sea bass: 9 samples, 1 or 2 replicates, 17 determina- tions). Total carbon and nitrogen contents were determined with a Carlo-Erba^ model 1106 elemen- tal analyzer (0.5-1.1 mg, usually triplicate, subsam- ^Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. 280 TUCKER: ANCHOVY AND SEA BASS ENERGETICS pies). Total lipid content of five sea bass samples (triplicate subsamples) was estimated by the sulpho- phosphovanillin technique (Barnes and Blackstock 1973) using a cholesterol standard. Caloric content of 7-10 h old anchovy eggs and 1-4 h old sea bass eggs (5-12 mg subsamples of five samples for each species— anchovies: 1 or 2 replicates, 8 determina- tions; sea bass: 3 replicates, 15 determinations) was determined by combustion in an oxygen microbomb calorimeter calibrated with benzoic acid. Caloric content of larvae was estimated using in- formation on the proportions of protein, lipid, carbo- hydrate, and ash in larvae. Percent protein was calculated as the product of percent nitrogen and 6.025 (Brett and Groves 1979). Because of sample shortages, the following assumptions were made: 1) Black sea bass carbohydrate content was esti- mated by subtracting % protein, % total lipid, and % ash from 100%. 2) Bay anchovy carbohydrate content was assumed to be the same as that esti- mated for black sea bass 7 h eggs, 28 h eggs, 150 h unfed larvae, and 249 h fed larvae. 3) Starving anchovy % ash was assumed to be the same as for fed anchovy larvae. 4) Total lipid content for an- chovies at four stages was estimated by subtract- ing % protein, % carbohydrate, and % ash from 100%. The effect of these assumptions on estimated caloric values is minimal because of the small per- centages involved; protein is the predominant con- stituent. The average energy equivalents for heat of combustion were used for conversion of weight to energy: 5,650 cal/g protein, 8,660 cal/g lipid, and 4,100 cal/g carbohydrate (Brett and Groves 1979). Oxygen uptake was measured with glass capillary differential microrespirometers (Microchemical Spe- cialties Company), calibrated by the potassium ferri- cyanide-hydrazine sulfate method (Umbreit et al. 1972). The experimental technique was similar to that described by Grunbaum et al. (1955). The ex- perimental and reference flasks each held 0.65 mL of air, a potassium hydroxide saturated filter paper strip for absorption of carbon dioxide, and 0.35 mL of 0.2 txm filtered seawater. Salinity was 32.0"/ofi for anchovies and 34.3"/o(i for sea bass. Tempera- ture was maintained at 24.0 + 0.05°C for anchovies and 20.0 + 0.05°C for sea bass in a water bath. Fluorescent lighting provided 300 lux. Slight agita- tion was provided by the flow of water in the bath. One to six eggs or larvae (number decreasing with age) were placed in each experimental flask. The fish were allowed to adjust for 10-60 minutes, depend- ing on age; the index droplet was stable after 10 minutes, but time was increased to allow for initial- ly greater activity of older larvae to subside after confinement. Measurements were made at all times of the day. To ensure that digestion was essentially complete, measurements with fed larvae began more than 2 hours after feeding had ceased. (Diges- tion time for larger bay anchovy larvae was 1.5 hours; digestion in sea bream up to 100 pig was almost finished by 2.5 hours; Houde and Schekter 1983.) Oxygen consumption was recorded hourly for periods of 3-9 hours (usually 6 hours). The longest that larvae normally would have to go without food is the length of the dark period (10 hours for an- chovies and 12 hours for sea bass). Also, when there is light, larvae expect to eat. Therefore, the mea- surement period for fed larvae was limited to 7 hours. Regression equations relating oxygen uptake to age were fitted. Metabolic energy (energy budget term M) was estimated from oxygen uptake with oxycalorific equivalents 0.00425 cal/pL oxygen for anchovies (24°C) and 0.00431 cal/j^L oxygen for sea bass (20 °C). Because movement of larvae in the flasks was restricted and feeding larvae normally were much more active (chasing rotifers) than nonfeeding larvae, the resulting total metabolism values were multiplied by the factor two for lighted periods for fed larvae (14 h/d for anchovies and 12 h/d for sea bass). This is the same procedure followed by Houde and Schekter (1983). Feeding observations were made in the 10 L rear- ing tanks without handling or otherwise disturbing the larvae. Numbers observed were 160 anchovies, 20 for each day of feeding; 128 sea bass, 5 the day before first feeding, and 10-20 for each day of feed- ing. Individual larvae were observed for 10 minutes. The number of prestrike flexes, strikes, and suc- cessful strikes were recorded and the following ratios were calculated: 1) successful strikes/flexes, 2) successful strikes/total strikes, and 3) strikes/ flexes. Successful strikes/total strikes is referred to as capture success. Feeding incidence is the per- centage of larvae that captured prey within 10 min- utes. Number of rotifers eaten per day was calcu- lated from the mean of observed 10 min feeding rates for each feeding day. Daily ingestion values (energy budget term I) were calculated with the fac- tor 0.000787 cal/rotifer (Theilacker and McMaster 1971). Weight-specific daily ration was calculated using 0.16 )jg/rotifer (best available estimate, from Theilacker and McMaster 1971) and predicted lar- val weights. Energy-specific daily ration was cal- culated from ingestion estimates and estimates of body energy in calorie per individual. Weight of wild zooplankton provided to first feeding bay anchovies by Houde and Schekter (1983) averaged 0.15 ugl 281 FISHERY BULLETIN: VOL. 87, NO. 2, 1989 individual. Detwyler and Houde (1970) found that during the first four feeding days, bay anchovies ate copepod nauplii, copepodites, and adults with daily average widths in the range 50-112 ^m. Because they are much easier to catch than copepods, larger rotifers will be eaten. In this study, growth might have been better with a diverse diet, but rotifers were used because their size range is limited and their nutritional quality is relatively well defined. Energy utilization on a caloric basis was assess- ed separately for endogenous (eggs, prefeeding and starving larvae) and exogenous (feeding larvae) nutrition. Energy budgets were constructed, based on the equation I M + F&U in which I is ingestion, G is growth, M is metabolic needs, F is egestion, and U is excretion. For eggs and unfed larvae, both I and F are near zero. Be- cause energy is needed for embryonic growth and metabolism, and some is excreted, the growth term will be negative. The form G = I - M - F&U may be more appropriate to consider in the con- text of growth and survival. G in calories was estimated from dry weight and proximate analysis data. I was estimated from feeding rate data. Oxygen uptake data provided M. Egested and excreted calories (F&U) were estimated by differ- ence. With endogenous nutrition, G = -M - U, and U = -G - M. With exogenous nutrition, F&U = I - G - M. Because F and U were not estimated separately, assimilation (A = I - F) is not considered. Bay Anchovy I LU 5 > a. a a = fed larvae 1 28 A'a 0.13897 -0.0003895(h ours) R'= 0.914 26 - 24 - A / 22 - /A / ^ 20 - eggs ^g -15,84-0.05469(hours) / 18 - R'= 0.371 Ay/ 16 o "ex ^ A X AA 1^ -^ A / 14 - o\ • /^ 12 - 10 - >^><; A A A A • = unfed larvae 8 _ • ^g =15 82-0.06598(tlou 6 - H FP EYS • s R'= 0.825 4 ' 1 i 1 1 1 1 20 40 60 80 100 120 140 160 180 200 220 240 HOURS AFTER FFRTILIZATION Figure 1.— Dry weight of bay anchovy eggs, unfed larvae, and fed larvae. 282 TUCKER: ANCHOVY' AND SEA BASS ENERGETICS RESULTS Developmental Events Developmental phases were longer for black sea bass at 20°C than for bay anchovies at 24°C (Table 1). Anchovy feeding behavior began within a few hours after EP, and successful feeding was first observed 72 hours after fertilization. Sea bass feed- ing behavior began several hours after EP, and suc- cessful feeding was first observed at 133 hours. An- chovy yolk lasted about 8 hours and sea bass yolk about 47 hours after first feeding. Unfed anchovies in 2 L dishes and 10 L tanks died 5.1 days after hatching (6.2 days after fertilization) and unfed sea bass in 10 L tanks died 8.2 days after hatching (10.2 days after fertilization). Table 1 .—Timing of bay anchovy and black sea bass developmen- tal events (hours after fertilization). Starvation = 100% mortality. Bay anchovy 24-0 Hatching (H) Eye pigmentation (EP) First-feeding success (FF) End of yolk sac (EYS) Starvation (S) 28 60 72 80 150 Black sea bass 20°C 48 110 133 180 245 Length and Weight Throughout development, black sea bass were heavier than bay anchovies of the same age and length, but length at age, and trends in length and weight, were similar. Sea bass egg weight was about twice that of anchovies (32 ^xg vs. 15 jig; Figs. 1, 2); at hatching, sea bass were heavier (22 ^g vs. 14 pig), yet both species were the same length, 2.0-2.1 mm I- o 111 60- 56- 52- 48- 44 40 3 36 32 28 24 20 16 12 8 Black Sea Bass o^eggs ^ig= 32.66-0.09646(hours) R'=0.275 »= unfed larvae ijg = 28 02 e"° °°'*'^5<''°"''*' R'=0 850 & = fed larvae ^g = 5.53 6 R'= 0.956 ,0 007264(hours) H EP EYS _l_ _l_ -L. _1_ _1 1 1 L_ ~5o *5 60 So Tm 120 Tto ToS TaS 200 220 240 2eo 280 300 320 340 HOURS AFTER FERTILIZATION Figure 2.— Dry weight of black sea bass eggs, unfed larvae, and fed larvae. 283 FISHERY BULLETIN: VOL. 87, NO. 2, 1989 NL. Six days after EYS, fed larvae of both species were 4.1-4.2 mm NL (sea bass 58 ^ig, anchovies 19 Hg). After 168 hours past first feeding, sea bass weighed 49 ng and anchovies 22 ^ig. Length of unfed larvae reached a maximum between EP and EYS (anchovies, 3.4 mm; sea bass, 3.2 mm) and a mini- mum between EYS and S (anchovies, 3.0 mm; sea bass, 2.9 mm). Body Composition Similar trends in total carbon and total nitrogen content occurred for unfed and fed larvae of both species (Table 2). Percent nitrogen was relatively constant. Percent carbon and C/N decreased be- tween hatching and yolk exhaustion, and then were relatively constant. Although data are limited, there was an apparent decrease in total lipid content of black sea bass throughout development. Values for the five samples were 1 h eggs, 14.5%; 4 h eggs, 15.6%; 217 h unfed larvae, 12.9%; 186 h fed lar- vae, 12.1%; 298 h fed larvae, 10.4%. Sea bass eggs and larvae contained about 50% more ash than anchovies (Table 3). Bomb calorimetry energy values were similar. Mean values in calories per gram for eggs were anchovies, 5,477 (SD 103, n = 5), ash-free 5,833; sea bass, 5,315 (SD 220, n = 5), ash-free 5,841. Representative calculations of Table 2.— Carbon and nitrogen content of bay anchovy and black sea bass eggs and larvae during growth and starvation. Feed- % Carbon % Nitrogen Age ing (h) day n X SD X SD C/N Bay anchovy Eggs 8 5 48.9 (0.2) 11.7 (0.1) 4.18 24 2 49.6 (0.1) 12.0 (0.2) 4.12 Unfed larvae 34 2 49.1 (0.3) 112 (0) 4.39 98 2 2 43.4 (0.8) 119 (0.1) 3.64 145 4 3 43,1 (0.4) 12.0 (0.2) 3.59 Fed larvae 92 1 2 43.4 (0.1) 116 (0.1) 372 146 4 1 44.5 12.2 3.63 202 6 1 42.4 11.9 3.56 248 8 3 43.8 (0.6) 12.0 (0.3) 3.65 Black sea bass Eggs 2 5 46.2 (1.1) 10.7 (0.2) 4.32 45 3 47.7 (0.7) 11.7 (0.2) 4.09 Unfed larvae 101 1 45.7 11.0 4.14 177 2 2 44.8 (1.3) 12.0 (0.3) 3.74 217 4 1 43.9 11.7 3.75 Fed larvae 183 3 2 44.6 (0.5) 11.7 (0) 3.82 234 5 2 429 (1.6) 11.4 (0.4) 378 298 7 1 43.0 11.7 3.68 caloric content for energy budgets are shown in Table 4. Oxygen Consumption The relationship between age and oxygen con- sumption depended on species, developmental phase, and nutritional status (Figs. 3, 4). In bay an- chovies, oxygen uptake increased continuously. In black sea bass, uptake rose until hatching, dropped, then rose again. Uptake decreased for unfed larvae. Feeding Behavior Black sea bass capture success was consistently higher than that of bay anchovies (Fig. 5). Anchovy capture success increased from 54% on feeding day 1 to 77% on feeding day 8. Sea bass capture success was 70% on feeding day 1, and during feed- ing days 2-8 remained relatively constant at 86-94%. Sea bass feeding incidence was higher than that of anchovies during the first four feeding days (Fig. 5). Anchovy feeding incidence gradually increased from 40% on feeding day 1 to 100% on feeding day 8. Sea bass feeding incidence varied at 85-97% dur- ing the first five feeding days and then remained at 100% for feeding days 6-8. Sea bass larvae had a higher flexing rate but lower strike per flex rate than anchovies. In anchovies, mean number of flexes per hour was 9 at first feed- Table 3.— Ash content of various stages of bay anchovy and black sea bass. Age (h) n Ash (%) SD (0/0) Bay anchovy Eggs '9 5 70 0.7 12 4 6.1 0.6 Fed larvae 85 1 8.7 146 1 9.0 248 3 10.1 2.0 Black sea bass Eggs '2 4 9.5 0.7 13 4 9.0 1.9 Unfed larvae 101 1 11.6 172 1 13.9 217 1 14.6 Fed larvae 186 1 13.2 298 1 17.7 'Values from calorimetry tor comparison only, not used in energy budget calculations 284 TUCKER: ANCHO\T AND SEA BASS ENERGETICS ing, 29 on feeding day 2, and 44 on feeding day 8 (feeding day 2-8 mean = 32). In sea bass, mean number of flexes per hour was 48 at first feeding, 74 on feeding day 2, and 59 on feeding day 8 (feed- ing day 2-8 mean = 63). Anchovy strikes/flexes was 79% at first feeding, 40% on feeding day 2, and 62% on feeding day 8 (feeding day 2-8 mean = 52%). Sea bass strikes/flexes was 38% at first feeding, Table 4.— Calculation of energy content of bay anchovy and black sea bass eggs and larvae. See Table 1 regarding acronyms. Bay ancfiovy Black sea bass Protein Lipid' Carboh,^ Asfi Energy^ Protein Lipid Carboh.' Asfl Energy^ (0/0) (%) (%) (%) cal/g (%) (%) (%) (%) cal/g Eggs Early 70.5 12.5 10,9 6.1 5.512 64,5 15.0 11.5 9.0 5,415 Late 73.5 12.3 8,1 6.1 5.550 70,5 15.0 5.5 9.0 5.508 Unfed larvae Halchling 675 12,3 13,3 6.9 5,424 66,3 15.0 8.4 10.3 5,389 EYS 72.0 99 95 8.6 5.315 71,7 13.2 1,1 14,0 5.239 Starvation 72.0 15,2 3,8 '9,0 5,540 71.7 12.9 0.4 15,0 5,184 Fed larvae First feeding 71 8 9.1 10,5 8,6 5.275 67.4 13.3 7.1 12.2 5,251 7 d after FF 71.8 14,6 3.5 10.1 5,465 69.6 10.4 2.1 17.9 4,919 'Estimated as the difference between 100% and the other components. ^Assumed to be the same as for sea bass at the same age ^Including ash ^Assumed to be the same as for fed larvae at the same age- Bay Ancfiovy < 3 Q > O z z g (- rj to z o o z 111 0.240 0200 160 0.120 0080 040- c= fed larvae ^1 oxygen = 06917 + 0.0004744 (hours) r'=0540 .•V « )jl oxygen^ 007279 + 0. 002750 (flours) R = 906 H EP EYS J i L • = unfed larvae ■< 70 hours: pi oxygen= 0.06494 + 0.0005220 (flours) r'=0 419 > 70 hours: pi oxygen = 1681-0.0008936 (flours) r'=0795 0,0012 0.0010 0.0008 < Z) 9 0006 > D z 0,0004 00002 _1_ J_ _L. J_ _L -L. _1_ 20 40 60 80 100 120 140 160 180 200 220 240 HOURS AFTER FERTILIZATION Figure 3.— Hourly oxygen consumption by bay ancfiovy eggs, unfed lar-vae, and fed larvae. 285 26% on feeding day 2, and 67% on feeding day 8 (feeding day 2-8 mean = 39%). During the first week of feeding, sea bass inspected more rotifers per unit time than anchovies did, but struck at a lower proportion of them. By the end of the week, these differences had diminished. Although obser- vations were made at all times of the day, no trend with time of day was detected. FISHERY BULLETIN; VOL. 87, NO. 2, 1989 Feeding Rate and Daily Ration Black sea bass feeding rates were considerably higher than those of bay anchovies (Fig. 6). Daily consumption of rotifers by anchovies increased from 4/h during the first feeding day to 13/h on feeding day 8. Sea bass rotifer consumption rose from 11/h on feeding day 1 to 17/h on feeding day 2, dropped Black Sea Bass 0.800 0.720 - 0.640 < 0.560 9 > 5 z 3- 0.480 - z g 1- a. 2 m z o u Ui a > X o 0.400 0.320 0.240 0.160 0.080 &=fed larvae ul oxygen =0.03383 e' R'= 0.835 0.009425(hours) •=unfed larvae <1 15 hours: ^1 oxygen=0 3621-0. 005913(hours) + 0.0000327(hours)' i R'= 0.522 >1 IShours; ^1 oxygen=0.21 16-0.0007323(hours) R' = 0.778 o^eggs ^1 oxygen^ 009934+0 002072(hours) R'= 0.884 0.0036 -0.0032 0.0028 -0.0024 0.0020 s U 0.0016 0.0012 0.0008 - 0.0004 20 40 60 80 Too 120 140 160 180 200 220 240 560 580 300 320 HOURS AFTER FERTILIZATION Figure 4.— Hourly oxygen consumption by black sea bass eggs, unfed larvae, and fed larvae. 286 TUCKER: ANCHOVY AND SEA BASS ENERGETICS 100- « UJ O 80- O 03 UJ 60- C 3 < o 40 20 ' 4 .1111 A i * r A A \ 1 1 1 1 1 1 1 CD z 100 80 ^ i 4 .. o UJ UJ u. UJ •< 60 - 1 ^ > A= Black Sea Bass < 40 - # 20 i 1 1 1 • = Bay Anchovy 1 1 1 1 1 2 3 4 5 6 7 8 DAY OF FEEDING Figure 5.— Feeding behavior of bay anchovy and black sea bass larvae (mean ± SE when n > 1). < 9 > o z o cc a o o u. 35 30 25 20 y 15 a. D 01 z o O 10 5- A= Black Sea Bass • = Bay Anchovy 1 I ) I _!_ _1_ _| L 1 J L. 0.028 0.024 0.020 0.016 5 Q > 5 0.012 Z — a O 0.008 0.004 8 2 3 4 5 6 DAY OF FEEDING Figure 6.— Feeding rates of bay anchovy and black sea bass larvae (mean ± SE when n > 1). 287 FISHERY BULLETIN: VOL. 87. NO. 2. 1989 to 12/h on feeding day 4, and then rose to 35/h on feeding day 8. Average daily ration was slightly higher for anchovies than for sea bass. For the 168 hours after first feeding, anchovy daily ration averaged 138% by weight and 126% by calories, compared to 126% by weight and 122% by calories for sea bass. Energy Budgets Total energy ingested during the 168 h period was 0.665 cal by bay anchovies and 1.191 cal by black sea bass (Tables 5, 6). Gross growth efficiency (Kj, G/I) and metaboHc component (M/I) changed with age of larvae. Anchovy G/I increased from - 18% to 15% (overall 9%), while sea bass G/I rose from 9% to 19% then dropped to 12% (overall 14%). Per- cent of ingested energy used for metabolism by anchovies (M/I) decreased from 44% to 21% (over- all 24%), while sea bass M/I increased from 16% to 37% then dropped to 30% (overall 28%). Over- all F&U/I was 67% for anchovies and 58% for sea bass. Four-Day Energy Budgets A striking difference in early growth capability is revealed by restricting the energy budget to the first 96 hours after first feeding. Black sea bass in- gested 1.8 times as much energy as bay anchovies, 0.543 vs. 0.299 cal. Egested and excreted compo- nents (F&U/I = 63% and 68%) were similar. Sea bass metabolic component was slightly lower (M/I = 24% vs. 27%) and gross growth efficiency was higher than that of anchovies (G/I = 13% vs. 5%). DISCUSSION Length of the interval between first feeding and yolk exhaustion is an important factor affecting sur- vivability of fish larvae because it is the period of transition from endogenous to exogenous feeding. Bay anchovies first feed only 8 hours before their yolk is exhausted, and they do not have positive growth until after EYS. Like Pacific sardines, Sar- dinoips caerulea (Lasker 1962), bay anchovies may be particularly vulnerable to food shortages at first feeding. In contrast, black sea bass have 2 days of Table 5. — Energy budget for bay anchovy eggs and larvae during growth and starvation. In the first column, developmental events are indicated in parentheses. G, M, and F&U as percentages of I are given in parentheses. G 1 - M - F&U Egested W/eight and Age Weight change Body Growth Food Metabolic excreted (h) O^g) O^g) calories calories calories calories calories Eggs 7 15.4 0.085 28(H) 14.3 -1.1 0.079 -0.006 0.003 0.003 Unfed larvae 33 13.6 0.074 0,002 60(EP) 11.9 -1.7 0.064 -0.010 0.010 72 11.1 -0.8 0.059 -0.005 0.005 80(EYS) 96 9.5 -1.6 0.051 -0.008 0.009 -0.001 120 7.9 -1.6 0.043 -0.008 0.007 0.001 144 6.3 -1.6 0.035 -0.008 0.005 0.003 150(S) 5.9 -0.4 0.033 - 0.002 0-001 0,001 Fed larvae 72(FF) 11.1 0.059 80(EYS) 96 9.8 -1.3 0.052 -0.007(-18%) 0.039 0.017(44%) 0.029(74%) 120 10.8 1.0 0.058 0.006 (8%) 0.071 0.020(28%) 0.045(64%) 144 12.1 1.3 0.065 0.007 (8%) 0.085 0.021(25%) 0.057(67%) 168 13.6 1.5 0.073 0.008 (8%) 0.104 0.023(22%) 0,073(70%) 192 15.6 2.0 0.084 0.011 (10%) 0.111 0.025(22%) 0075(68%) 216 18.2 2.6 0.099 0.015 (13%) 0.119 0.027(22%) 0.077(65%) 240 22.0 3.8 0.120 0.021 (15%) 0.136 0.028(21%) 0.087(64%) Age: (rom fertilization Weight: measured Weight change: (rom column 2. Body calories = (weight) (estimated caloric content). G = Growth calories = change in body calories. I = Food calories = (average feeding rate) (feeding time) (0 000787 cal/rotifer). M = Metabolic calories (measured ^L 02/h) (0 00425 cal/^L O2) (38 hours). F&U = Egested and excreted calories = t - M - G. 288 TUCKER: ANCHOVY AND SEA BASS ENERGETICS feeding and positive growth before their yolk is ex- hausted. (Neither species had an advantage in sur- vival time after yolk exhaustion; unfed larvae of both species died within 3 days after EYS.) Starving sea bass weighed 50% more than starving anchovies. Time from hatching to starvation for unfed bay an- chovies in both 2 L dishes and 10 L tanks was 122 hours (Table 1). Houde (1974) reported this period to be 126 hours in 35 L aquaria. Growth of larvae might have been slightly reduced by container size; however, anchovies and sea bass were reared in the same tanks and the comparison should not be affected by container size. Although growth in length was similar, sea bass gained weight faster than anchovies (Table 7). At hatching, sea bass weighed 1.6 times as much as an- chovies. After 7 days of feeding, sea bass weighed 2.2 times as much as anchovies (Figs. 1, 2). At a Table 6— Energy budget for black sea bass eggs and lar\,-ae during growth and starvation. In the first column, developmental events are indicated in parentheses. G, M, and F&U as percentages of I are given in parentheses. G 1 M F&U Egested Weight and Age Weight change Body Growth Food Metabolic excreted (h) O-g) (Fg) calories calories calories calories calories Eggs 2 32.5 0.176 48(H) 28.0 -4,5 0,154 -0,022 0,013 0.009 Unfed larvae 62 20.8 0.112 0,008 88 18.6 -2,2 0,099 -0,013 0.010 0.003 IIO(EP) 16.6 -2.0 0.088 -0.011 0,010 0.001 133 14.9 -1.7 0.078 -0,010 0,012 -0.002 157 13.3 -1,6 0,070 -0,008 0,011 -0.003 180(EYS) 181 11.8 -1.5 0.062 -0.008 0.009 -0.001 205 10.5 -1.3 0.055 -0.007 0.007 229 9.4 -1.1 0.049 -0.006 0.006 245(S) 8.7 -0,7 0.045 -0,004 0,003 0001 Fed larvae 133(FF) 14.9 0.078 157 17,3 2,4 0,090 0,012 (9%) 0,138 0022(16%) 0.104(75%) 180(EYS) 181 20.6 3,3 0.107 0-017(11%) 0,155 0.028(18%) 0-110(71%) 205 24.5 3,9 0.126 0,019(14%) 0,132 0,035(27%) 078(59%) 229 29.2 4,7 0.148 0,022(19%) 0.118 0-043(36%) 0,053(45%) 253 34,7 5,5 0.175 0,027(18%) 0.152 0,056(37%) 0-070(45%) 277 41,4 6,7 0.206 0,031(15%) 0,206 0,069(34%) 0.106(51%) 301 49.2 7.8 0.242 0.036(12%) 0.290 0.086(30%) 0.168(58%) Age: from fertilization. Weight: measured Weight change from column 2 Body calories = (weight) (estimated caloric content)- G = Growth calories = change in body calories. I = Food calories = {average feeding rate) (feeding time) (0.000787 cal/rotifer). M = Metabolic calories (measured mL Oz/h) (0 00431 cal/>iL O2) (36 hours). F&U = Egested and excreted calories = I - M - G Table 7.— Percent change in weight and energy content of bay anchovy and black sea bass during developmental phases. Instantaneous growth rates are given. FF = First feeding. Hatching to First feeding Hatching to First feeding to starvation to starvation 168 h after FF 168 h after FF Time Total Inst. Time Total Inst. Time Total Inst. Time Total Inst. (d) (%) (%) (d) (%) (%) (d) (%) (%) (d) (%) (%) Bay anchovy Weight change 5.1 -58 -17 3.2 -47 -19 8-8 58 5 7 98 10 Energy change -57 -16 -44 -18 58 5 103 10 Black sea bass Weight change 8.2 -61 -12 4.7 -42 -11 10.5 120 7 7.0 230 17 Energy change -63 -12 -42 -12 100 7 210 16 289 FISHERY BULLETIN: VOL. 87, NO. 2, 1989 lower temperature (15.5°C), northern anchovies lost 10% of their weight per day during the first 3 days of starvation (Theilacker 1987), versus 17% per day for bay anchovies (Table 5) and 11% per day for black sea bass (Table 6). The greater ash content of sea bass (Table 3) is probably related to their greater size and consequent need for more struc- tural material. Egg and larval caloric values calculated from prox- imate analysis data (Table 4) are similar to bomb calorimetry values for anchovy and sea bass eggs and to published values for other species. Calculated values for eggs were 5,512 cal/g for anchovies and 5,415 cal/g for sea bass (less than a 2% difference from measured values, 5,477 and 5,315 cal/g). Energy content of northern anchovy, Engraulis mordax, eggs was 5,450 cal/g (Hunter and Leong 1981). Calculated values for larvae fed for 7 days were bay anchovies, 5,465 cal/g, 6,079 cal/g ash- free; black sea bass, 4,919 cal/g, 5,991 cal/g ash-free. These numbers are within the ranges given by Thayer et al. (1973) for postlarvae of four marine fish species: 4,904-6,001 cal/g, 5,694-6,418 cal/g ash-free. Ranges of calculated ash-free values were 5,771-6,088 cal/g for bay anchovies and 5,950-6,099 cal/g for black sea bass. The possible effect of vary- ing lipid and carbohydrate content is small. Houde and Schekter (1983) used a constant value of 5,000 cal/g, and Theilacker (1987) used 5,400 cal/g in con- structing energy budgets for larval fish. Patterns of oxygen consumption were generally similar for the two species (Figs. 3, 4). The decrease for black sea bass during the 0.5 day after hatching probably resulted from reduced activity prior to the development of vision. The interval between hatch- ing and EP was shorter for bay anchovies (1.3 days vs. 2.5 days). Lasker and Theilacker (1962) found that oxygen uptake in Pacific sardines in- creased just after hatching, but was variable, depending on activity. On the eighth day of feeding, sea bass consumed oxygen at three times the rate of anchovies. At that stage, sea bass had two and a half times as much respiring tissue, and were more active (Qo of sea bass was 12 fxL 02/mg/h vs. 9 for anchovies). However, at 20 ^g, bay anchovy and black sea bass Qo„ was the same and was inter- mediate among those of other species (Table 8). Early bay anchovy oxygen uptake was similar to that found by Houde and Schekter (1983), who reported mean uptakes of 0.030 nL/hlegg and Table 8. — Comparison of growth characteristics of well-fed larvae of five species. regarding acronyms. See Tables 1 and 5 Northern Bay Bay Black Sea Lined anchovy' anchovy^ anchovy^ sea bass' bream^ sole^ le'C 26°C 24''C 20°C 26°C 28°C Age at FF 3.0 ~1.5 1.8 3.5 ~1.5 '^'2.0 (d after H) Capture success "11 49 54 70 53 69 at FF (%) Capture success "39 60 74 87 61 81 20 >/g (%) Daily ration (weight) 81 281 138 166 198 252 17-22 Mg(%) 1 component 0.06 0,332 0.136 0.155 0.234 0.297 1 7-22 (jg (cal/d) Oxygen uptake 0,134 0.144 178 0.179 0.218 0.240 20 mQ (fi- 02/ind/h) Qo, 6.7 7.2 8.9 89 10.9 12.0 20 Hg O-L Oj/mg/h) M component 0,010 0.025 0028 0.028 0.037 0.041 17-22 Mg (cal/d) M/l 17 8 21 18 16 14 17-22 Mg(%) Inst. Grth. (wt or cal) 19 34 19 17 40 32 1 7-22 ^g (%) G component 0020 0.041 0.021 0.017 0.050 0.038 1 7-22 hQ (cal/d) G/l = K, 33 12 15 11 21 13 17-22 fig (%) (G + M)/l = CU 50 20 36 29 37 27 17-22 Hg(%) 'Theilacker 1987 (except capture success). 'Houde 1974, Houde and Schekter 1980: Houde and Schekter 1983 ^Present study. ■"Hunter 1972, 290 TUCKER: ANCHO\T AND SEA BASS ENERGETICS 0.066 fiL/h/yolk-sac larva, at 26°C, two degrees higher. Sea bass were more active, were more efficient feeders, and spent more time feeding than anchovies (Fig. 5). Sea bass also were more capable predators from first feeding through the eighth feeding day. At first feeding, sea bass were 2.5 days older and were better developed than anchovies. Bay anchov- ies in this study had about the same capture success (Table 8) as bay anchovies and sea bream studied by Houde and Schekter (1980). Black sea bass cap- ture success was similar to that of lined sole. North- ern anchovy larvae feeding on 10-60 BrachionusI mL at 17°-18°C (Hunter 1972) were less successful than bay anchovies in the present study, but they struck more often and therefore consumed more rotifers per hour. Northern anchovy capture success was relatively low, ranging from 11% at first feed- ing to 60% on feeding day 8. For 20 ng larvae, a rate of about 50 strikes/h (Hunter and Thomas 1974) multiplied by 39% capture success gives a feeding rate of 20 rotifers/h vs. 13/h for bay anchovies and 16/h for black sea bass. Daily rations for bay anchovies and black sea bass were intermediate among published estimates from rearing studies using high larval and food densities (Table 8). Theilacker and Dorsey (1980), in a review article, reported weight-specific daily rations of 70- 300% for larvae fed one or more prey mer mL. Houde and Schekter (1983) reported high weight or calorie-specific daily rations of 202-379% for 10-100 ^ig bay anchovies fed 1 copepod nauplius/mL at 26°C. During the first 3 days of starvation, weight and calorie loss were similar for both species; however, sea bass conserved weight and calories better dur- ing the late stages of starvation (Tables 5, 6, 7). Sea bass also gained weight and calories faster when fed (Table 7). Conservation probably resulted partly from a rearing temperature four degrees lower and partly from physiological differences. Better growth probably resulted from a combination of more effi- cient feeding, higher ingestion rate, lower temper- ature, and different physiology. During the first 24 hours after EP, fed anchovies lost more weight and calories than unfed anchovies (-0.015 vs. -0.008 cal). During the first 24 hours after EP, fed sea bass lost about the same weight and calories as unfed sea bass (-0.011 vs. - 0.010 cal). This impHes that an- chovies at first lost more energy to feeding activity than they gained from their food, while sea bass broke even. Overall gross growth efficiencies (G/I) of 9% for anchovies and 14% for sea bass were at the lower end of the known range for early larvae. Published G/I values for larvae fed one or more prey per mL are 11-46% (Theilacker and Dorsey 1980; Houde and Schekter 1983; Theilacker 1987). The decrease in sea bass gross growth efficiency after 4 days of feeding (229 hours, Table 6) may be related to de- creasing suitability of rotifers as food for sea bass (Tucker 1984). After the first few days of feeding, larval growth of both species probably would have been enhanced by the addition of larger prey (Hunter 1980). The effect of small prey on growth may have been greater for sea bass, which have larger mouths and probably can handle larger prey. As a larva grows, the benefit:cost ratio for feeding on constant energy food particles tends to decrease (Theilacker and Dorsey 1980). This principle appears to apply to sea bass, as suggested by reduced feed- ing after the first two days and decreasing growth efficiency after the fourth day. If, in nature, bene- fit:cost (food energy:expended energy) drops close to one, the rule of fast early growth is violated and the larva is vulnerable to a given type of predator for a longer time. Overall M/I values of 24% for anchovies and 28% for sea bass were lower than Brett and Groves' (1979) average of 44% for typical, young, well-fed, fast-growing carnivorous fish; however, M/I is likely to be lower in larvae. One explanation for Houde and Schekter's (1983) lower M/I for anchovies (Table 8) is the high ingestion rate. Hunter and Kimbrell (1980) estimated that 3-5 d old Pacific mackerel, Scomber japonicus, use about 18% of ingested calories for metabolism at 19°C. Overall coefficient of utilization (CU), which is metabolizable energy expressed as a fraction of in- gested energy, (G -i- M)/I, was slightly lower in an- chovies (33%) than in sea bass (42%). The coefficient of utilization for young fish has been estimated at 73% (G = 29%, M = 44%) by Brett and Groves (1979) and 65-75% by Ware (1975). Ingested energy unaccounted for by growth and metabolism, 67% for anchovies and 58% for sea bass, was assumed to have been egested or excreted, F&U/I. These values are higher than Brett and Groves' (1979) mean of 27% for young fish, but similar to values for other larvae. Larvae are not as efficient at using their food energy as larger fish but do not need as much of it for activity and maintenance. The energetics approach can be used to compare adaptations to feeding environments. Although roti- fers are not normally eaten in large quantities by anchovies or sea bass in nature, the results of this study are probably indicative of normal feeding ecol- ogy, especially if larvae encounter patches of food 291 FISHERY BULLETIN: VOL. 87. NO. 2. 1989 organisms of similar nutritional value. The bay an- chovy larva has low feeding and growth efficiencies, but its food (in estuaries and coastal waters) is rela- tively abundant. To compensate for low efficiency, it is obligated to feed in high densities of prey. Fluc- tuations in density of zooplankton prey in estuaries might strongly influence survival and recruitment to anchovy populations. The black sea bass larva feeds and grows more efficiently. It has to because its food (offshore) is not very abundant. The bay an- chovy larva seems to be adapted to the high prey densities, and the black sea bass larva to the low prey densities, that characterize their respective habitats (Theilacker and Dorsey 1980). The results of this study parallel those of Houde and Schekter (1983). ACKNOWLEDGMENTS I thank John Merriner, Dave Peters, Allyn Powell, Bill Hettler, Alex Chester, Dave Colby, Bud Cross, John DeVane, Don Hoss, Jud Kenworthy, Al Kuo, Mike LaCroix. Curtis Lewis, Jack Musick, Pete Parker, Brenda Sanders, Ken Webb, Dick Wetzel, Doug Willis, the U.S. Coast Guard, and anonymous referees for technical and editorial assistance. The study was supported by and conducted at the Beau- fort Laboratory of the Southeast Fisheries Center, National Marine Fisheries Service, during 1978-82. It is part of a dissertation submitted to the College of William and Mary in partial fulfillment of the re- quirements for the Ph.D. degree. LITERATURE CITED Barnes. H.. and J. Blackstock. 1973. Estimation of lipids in marine animals and tissues: detailed investigation of the sulphophosphovanillin method for total lipids. J. exp. mar. Biol. Ecol. 12:103-118. Brett, J. R., and T. D. D. Groves. 1979. Physiological energetics. In W. S. Hoar, D. J. Ran- dall, and J. R. Brett (editors), Fish physiology. Vol. VIII, p. 279-352. Academic Press, N.Y. Detwyler. R., and E. D. Houde. 1970. Food selection by laboratory-reared larvae of the scaled sardine Harengula pensacolae (Pisces. Clupeidae) and the bay anchovy Anchoa mitchilli (Pisces, Engraulidae). Mar. Biol. (N.Y.) 7:214-222. Dovel, W. L. 1971. Fish eggs and larvae of the upper Chesapeake Bay. Univ. Md. Nat. Resour. Inst., Spec. Rep. 4. 71 p. GRUNBAUM, B. W., B. V. SlEGEL. A. R. SCHULZ. AND P. L. KiRK. 1955. Determination of oxygen uptake by tissue grown in an all glass differential microrespirometer. Mikrochim. Acta 1955. 1069-1075. Harpster. B. v., D. E. Roberts, Jr.. and G. E. Bruger. 1977. Growth and food conversion in juvenile southern sea bass, Centropristis melana (Ginsburg), fed commercial and seminatural diets. Proc. World Maricult. Soc. 8:795-809. HiLDEBRAND, S. F. 1963. Family Engraulidae. /n Fishes of the western North Atlantic, Part 3, p. 152-249. Mem. Sears Found. Mar. Res. 1. HiLDEBRAND, S. F., AND L. E. CABLE. 1930. Development and life history of fourteen teleostean fishes at Beaufort, N.C. U.S. Bur. Fish.. Bull. 46:383-488. HOFF. F. H., Jr. 1970. Artificial spawning of black sea bass, Centropristes striatum jnelanus Ginsburg, aided by chorionic gonadotropic hormones. Fla. Dep. Nat. Resour. Mar. Res. Lab., Spec. Sci. Rep. 25, 17 p. Houde. E. D. 1974. Effects of temperature and delayed feeding on growth and survival of larvae of three species of subtropical marine fishes. Mar. Biol. 26:271-285. Houde, E. D., and R. C. Schekter. 1980. Feeding by marine fish larvae: developmental and func- tional responses. Environ. Biol. Fish 5:315-334. 1981. Growth rates, rations and cohort consumption of marine fish larvae in relation to prey concentrations. Rapp. P.-v. R^un. Cons. int. E.xplor. Mer 178:441-453. 1983. Oxygen uptake and comparative energetics among eggs and larvae of three subtropical marine fishes. Mar. Biol. 72:283-293. Hunter, J. R. 1972. Swimming and feeding behavior of larval anchovy, Engraulis mordax. Fish. Bull., U.S. 70:821-838. 1980. The feeding behavior and ecology of marine fish lar- vae. In J. E. Bardach, J. J. Magnuson, R. C. May, and J. M. Reinhart (editors), Fish behavior and its use in the cap- ture and culture of fishes, p. 287-330. ICLARM Conf . Proc. 5. International Center for Living Aquatic Resources Man- agement, Manila, Philippines. Hunter, J. R.. and C. M. Kimbrell. 1980. Early life history of Pacific mackerel. Scomber japo- nicus. Fish Bull.. U.S. 78:89-101. Hunter, J. R.. and R. Leong. 1981 . The spawning energetics of female northern anchovy, Engrauhs mordax. Fish. Bull., U.S. 79:215-230. Hunter, J. R., and G. L. Thomas. 1974. Effect of prey distribution and density on the search- ing and feeding behavior of larval anchovy Engraulis mor- dax Girard. In J. H. S. Blaxter (editor). The early life history of fish, p. 559-574. Springer-Verlag, Berlin. Kendall, A. W., Jr. 1972. Description of black sea bass, Centropristis striata (Linnaeus), larvae and their occurrences north of Cape Look- out, North Carolina, in 1966. Fish. Bull.. U.S. 70:1243- 1260. Kuntz, A. 1914. The embryology and larval development of Bairdieila chrysura and Anchovia mitchilli. U.S. Bur. Fish.. Bull. (1913) 33:1-19. 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. Lasker, R., and G. H. Theilacker. 1962. Oxygen consumption and osmoregulation by single Pacific sardine eggs and larvae {Sardimyps caerulea Girard). J. Cons. int. Explor. Mer 27:25-33. Miller. R. J. 1959. A review of the seabasses of the genus Centropristes (Serranidae). Tulane Stud. Zool. 7:35-68. 292 TUCKER: ANCHO\T AND SEA BASS ENERGETICS 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. Roberts, D. E., Jr.. B. V. Harpster, W. K. Havens, and K. R. Halscott. 1976. Facilities and methodology for the culture of the south- em sea bass, Centropristis melana. Proc. Seventh Ann. Workshop, Worid Maricult. Soc, p. 163-177. Thayer, G. W., W. E. Schaaf, J. W. Angelovic, and M. W. LaCroix. 1973. Caloric measurements of some estuarine organisms. Fish. Bull., U.S. 71:289-296. Theilacker, G. H. 1987. Feeding ecologj- and growth energetics of larval north- ern anchovy, EngrauHs mordnji. Fish. Bull.. U.S. 85:213- 228. Theilacker, G. H., and K. Dorsey. 1980. Larval fish diversity, a summary of laboratory and field research. In G. Sharp (rapporteur). Workshop on the ef- fects of environmental variation on the survival of larval pelagic fishes, p. 106-142. UNESCO, Intergov. Oceanogr. Comm. Workshop Rep. 28. Theilacker, G. H., and M. F. McMaster. 1971. Mass culture of the rotifer Brachionus plicatilis and its evaluation as a food for larval anchovies. Mar. Biol. 10:183-188. Tucker, J. W., Jr. 1984. Hormone-induced ovulation of black sea bass and rear- ing of larvae. Prog. Fish-Cult. 46:201-203. Umbreit. W. W., R. H. Burris, and J. F. Stauffer. 1972. Manometric and biochemical techniques. Burgess Publ. Co.. Minneapolis, 387 p. Ware, D. M. 1975. Growth, metabolism, and optimal swimming speed of a pelagic fish. J. Fish. Res. Board Can. 32:33-41. Winberg, G. G. 1956. Rate of metabolism and food requirements of fishes. Nauchn. Tr. Bashk. SSR. Cos. Univ. V. I. Lenina, Minsk. [Transl. From Russian by Fish. Res. Board Can., Transl. Ser. No. 194 (1960). 253 p.] 293 FACTORS INFLUENCING RECAPTURE PATTERNS OF TAGGED PENAEID SHRIMP IN THE WESTERN GULF OF MEXICO Peter F. Sheridan/ Refugio G. Castro M.,^ Frank J. Patella, Jr.,' AND Gilbert Zamora, Jr.' ABSTRACT Movements of brown shrimp, Penams aztecus, and pink shrimp, P. duorarum, off the adjacent states of Texas (USA) and Tamaulipas (Mexico) in the western Gulf of Mexico were examined by releasing tagged shrimp within 150 l29'W 895 895 0.00 50 5.6 6/5 25°06N, 97°24W 1,592 1,590 0.13 132 8.3 6/6 25°08'N, 97°24'W 1,594 1,581 0.82 139 8.8 6/7 25°26N, 97°14W 1,583 1,579 0.25 218 13.8 6/8 25°45N, 97°05W 1,190 1,188 0.17 163 13.7 5/30-6/8 — — 14,055 13,987 0.48 1,089 7.8 Texas 6/21 26''55N, 97<'17W 2,305 1,931 16.26 163 8.4 6/22 26''15N, 97=03'W 1,995 1,898 436 434 22.9 6/23 26°37N, 97=11 W 3,188 3,024 5.14 583 19.3 6/24 26°45N, 97°15W 991 939 5.25 140 14.9 6/25 26°05N, 97°07W 3,288 3,157 398 873 27.7 6/26 26<'25'N, 97°07W 3,690 3,661 0,79 327 89 6/27 26°37N, 97=11 W 1,784 1,729 3.08 307 17.8 7/7 26°25'N, 97°06'W 2,009 1,993 0.80 242 12.1 7/8 26''47'N, 97°15W 2,083 2,056 0.34 276 13.4 7/9 26''05N, 97°05W 3,817 3,600 0.47 458 12.7 7/10 26°15N, 97°04W 2,447 2,399 1.96 444 18.5 7/11 26°55N, 97°17'W 1,864 1,849 0.80 303 16.4 6/21-7/11 — — 29,241 28,236 3.44 4,550 16.1 Total 43,296 42,223 2.48 5,639 13.4 Table 2.— Comparison of shrimp species compositions for release, for all recaptures regardless of quality of return information, and for recaptures with complete and accurate return information. The latter group formed the data base for analyses of directional movement. N = number examined, ' = significant difference (Mest, a = 0.05) between proportions of brown shrimp released and recaptured (all recaptures excluding unknowns). Release date Release All recaptures Best recaptures N % brown % pink N % brown % pink % unknown N % brown % pink 5/30 297 30.6 69.4 107 15.9* 80.4 3.7 97 13.4 86.6 5/31 294 96.3 3.7 116 70.7* 24.1 5.2 102 74.5 25.5 6/1 289 92.1 7.9 55 92.7 5.5 1.8 49 93.9 6.1 6/2 299 87.0 13,0 48 85.4 10.4 4.2 21 95.2 4.8 6/3 297 37.0 63,0 61 70.5* 24.6 4.9 56 73.2 26.8 6/4 298 93.6 6,4 50 82.0 12.0 6.0 43 88.4 116 6/5 295 586 41,4 132 51.5 45,5 3.0 120 53.3 46.7 6/6 298 85.9 14,1 139 69,1* 23,0 7.9 106 75.5 24,5 6/7 296 31.1 689 218 29,4 62,4 8.3 119 35.3 64,7 6/8 299 59.2 40.8 163 6,1* 87,1 6.7 54 7,4 926 5/30-6/8 2,962 66.9 33.1 1,089 47.1- 47.1 5.8 767 55,3 44.7 6/21 279 21.1 78.9 163 6.7* 79.8 13.5 114 4,4 95.6 6/22 298 429 57,1 434 8.3* 726 19.1 288 9,4 90.6 6/23 298 8.4 91,6 583 2.6* 82.7 14,8 345 2,9 97.1 6/24 299 57.2 42,8 140 12.9* 77.9 9,3 81 19,8 802 6/25 292 229 77.1 873 9.5* 80.9 9.6 482 ,12,2 878 6/26 300 91.7 8,3 327 65.7 8.6 25.7 150 860 14.0 6/27 300 4.0 96.0 307 3.6 88.9 7.5 194 3,1 969 7/7 298 84.9 15.1 242 39.3 11.6 49.2 42 738 26.2 7/8 300 70,7 293 276 51,8* 38.0 10.1 43 58,1 41.9 7/9 299 85.6 14 4 458 57,9* 31,2 10.9 209 598 40.2 7/10 294 44,6 55,4 444 40,8 43,2 16.0 148 45,9 54.1 7/11 296 96.6 3,4 303 79,2 4,6 16.2 169 94.7 5.3 6/21-7/11 3,553 52.7 47,3 4,550 28,9* 55.5 15.6 2,265 29.2 70.8 Total 6,515 59.2 40.8 5,639 32.4* 53.9 13.7 3,032 35.8 64.2 298 SHERIDAN ET AL.: RECAPTURE PATTERNS OF TAGGED PENAEID SHRIMP tude, hereafter called "grids", along the Texas coast. Port agents in Tamaulipas interviewed the cap- tains of all vessels returning to the primary port of Tampico. Unknown, but assumed relatively small, amounts of catch and effort were potentially re- ported in more southerly ports. Interviewers col- lected catch and effort data by depth range and 10 minute lines of latitude between 22°N and 26°N. These data were then recordable either within 9 m depth zones or within grids as was done off Texas. Interviews recorded effort by specific 9 m depth zones (Texas) or by actual depth ranges (Tamauli- pas) per trip. Tamaulipas effort was assumed to fall equally into adjacent 9 m depth zones if more than one zone was covered by the stated depth range. The average fishing depth per trip was then calculated by weighting the hours expended in each 9 m depth zone by the middepth of that zone (e.g., the 10-18 m zone had a middepth of 14 m), summing over all depth zones, then dividing by the total effort ex- pended on that trip. Average fishing depth for each fleet was then compared by a t-test corrected for unequal variances (Sokal and Rohlf 1969) using the average depth for each of 2,008 Texas trips and 505 Tamaulipas trips as observations. Data Analysis Three-factor, model I analysis of variance ( ANOVA) with unequal cell sizes was employed to test hypoth- eses concerning the equivalence of treatment means for several types of observations on recaptured shrimp. The treatment factors were species (brown or pink), sex, and release state (Texas or Tamau- lipas). State was chosen as a treatment because the level of fishing effort off Texas is much greater than that off Tamaulipas (approximately 200 vessels use the port of Tampico, whereas there are nearly 2,000 vessels registered in Texas alone). Four attributes of shrimp movement were examined by ANOVA: 1) distance travelled before recapture, assumed to be a straight line, 2) days at large, 3) apparent speed of movement, and 4) recapture depth. All four variables exhibited nonnormal (skewed) error distri- butions, as indicated by the Shapiro-Wilk test statistic (Shapiro and Wilk 1965), but the effects of nonnormality are thought to be minimal with large sample sizes (Underwood 1981). Variances of all variables were found to be heterogeneous (F-max test for unequal cell sizes; Sokal and Rohlf 1969). Data were log(x -i- l)-transformed prior to ANOVA (Underwood 1981), and F-max tests on transformed data indicated homogeneity of variances. Multiple comparison of treatment means of transformed data employed Fisher's LSD (least significant difference) because of unequal cell sizes (Milliken and Johnson 1984). Circular scale data such as compass directions are a special tj'pe of interval scale data (Zar 1984) that cannot be examined by ANOVA because there is no physical reason for any zero point and high or low values are arbitrary (e.g., 45° is not a "larger" direc- tion than 30°. and the mean of the 45° and 315° is not 180° but 0°). Examination of the raw data indicated that the assumption of unimodal distribu- tions of recapture directions needed for hypothesis testing with the recommended parametric test (Watson- Williams statistic) would be violated. We conducted multisample testing of grouped direc- tional data using contingency tables (Zar 1984). Before analysis, compass direction from release site to recapture site was adjusted downward by 20° off Texas and upward by 20° off Tamaulipas because northerly movement parallel to shore (hereafter termed "north") is 20° west of magnetic north (340°) off southern Texas and 20° east of magnetic north (020°) off northern Tamaulipas (Fig. 1). We grouped the adjusted directional data into eight ar- bitrary 45° divisions (0-44°, 45-89°, etc.) that ful- filled the requirement of having no expected cell fre- quency less than 4 (Zar 1984), with one exception. Only one brown shrimp and one pink shrimp re- leased off Tamaulipas were recaptured between 270° and 359°; thus the contingency tables compar- ing these two data sets employed six 60° divisions (15-74°, 74-134°, etc.) to avoid low cell frequencies. Differences in shrimp movement away from re- lease sites were also t